Skip to main content

Full text of "Transactions of the American Society of Mechanical Engineers"

See other formats


Digitized by tine Internet Arciiive 

in 2010 with funding from 

University of Toronto 



http://www.arcliive.org/details/transactionsof31amer 




rHE AMEBlCAN SOCIETY OF MECHANICAL ENGINEERS 



THE AMERICAN SOCIETY OF 



ffi 



AlKCfi WTOAT. FNOINFFI 



\\\ \\S\rTI()\S 



T^ATT7 M 



UFFTf\'<, 



OFIfi 



T*. 



^ 





<f.^f /^^. t 



THE^ AMERICAN SOCIETY OF 
ME( HANK AL ENGINEERS 



TRANSACTIONS 



VOLUME M 



WASHINGTON MEETING 
NEW YORK MEETING 

1909 





NEW YORK 

PUBLISHED BY THE SOCIETY 

29 West 39th Street 

1910 



I 



Copyright 1910 by 
THE AMERICAN SOCIETY OF MECHANICAL ENGINEERS 



OFFICERS 

THE AMERICAN SOCIETY OF MECHANICAL 
ENGINEERS 

1909 
FORMING THE STATUTORY COUNCIL 

PRESIDENT 
Jesse M. Smith New York. 

VICE-PRESIDENTS 

L. P. BuECKENRiDGE Urbana, 111. 

Fred J. Miller Center Bridge, Pa. 

Arthur West E. Pittsburg, Pa. 

Terms expire at Annual Meeting of 1909 

Geo. M. Bond Hartford, Conn. 

R. C. Carpenter Ithaca, N. Y. 

F. M. Whyte New York 

Terms expire at Annual Meeting of 1910 

PAST PRESIDENTS 

Members of the Council for 1909 

Ambrose Swasey Cleveland, O. 

John R. Freeman Providence R.I. 

Frederick W. Taylor Philadelphia, Pa. 

F. R. Hutton New York 

M. L. Holman St. Louis, Mo. 

MANAGERS 

G. M. Basford New York 

A. J. Caldwell (Deceased) Newburg, N. Y. 

A. L. Riker Bridgeport, Conn. 

Terms expire at Annual Meeting of 1909 

Wm. L. Abbott Chicago, 111. 

Alex. C. Humphreys New York 

Henry G. Stott New Rochelle, N. Y. 

Terms expire at Annual Meeting of 1910 

H. L. Gantt New York 

I. E. MouLTROP Boston, Mass. 

W. J. Sando Milwaukee, Wis. 

Terms expire at Annual Meeting of 1911 

TREASURER 
WiLLiA-M H. Wiley New York 

CHAIRMAN OF THE FINANCE COMMITTEE 
Arthur M. Waitt New York 

HONORARY SECRETARY 
F. R. HuTTON New York 

SECRETARY 
Calvin W. Rice 29 West 39th Street, New York 

V 



PAST PRESIDENTS 



Thurston, R. H 1880-1882 Died Oct. 25, 1903 

Leavitt. E. D 1883 ^ Cambridge, Ma,ss. 

Sweet, John E 1884 Syracuse, N. Y. 

HoLLOWAY, T. F 188.5 Died Sept. 1, 18% 

Sellers, Coleman 1886 Died Dec. 28. 1907 

Babcock, George H 1887 Died Dec. 16, 1893 

See, Horace 1888 Died Dec. 14. 1909 

TowNE, Henry R 1889 New York. 

Smith, Oberlin 1890 Bridgeton, N. J. 

Hunt, Robert W 1891 Chicago, III. 

LoRiNG, Charles H 1892 Died Feb. 5, 1907 

CoxE, EcKLEY B 1892-1894 • Died May 13, 1895 



Davis, E. F. C 

Billings, Charles E 

Fritz, John 

Warner, Worcester R 

Hunt, Charles Wallace . 
Melville, George W . . . . 

Morgan, Charles H 

Wellman, S. T 



.... 1894 Died Aug. 6. 1895 

. . . .1895 Hartford, Conn. 

1896 r BcThlehem. Pa. 

1897 Cleveland, O. 

.... 1898 New York. 

.... 1899 Philadelphia, Pa. 

.... 1900 Worcester, :\Ia.ss. 

.... 1901 Cleveland, O. 

Reynolds. Edwin 1902 Died Feb. 19, 1909 

Dodge, James M 1903 Philadelphia, Pa. 



FAST-FRESIDEXrS AND HONORARY COUNCILORS 

1909 

SwASEY, Ambrose 1904 Cleveland, O 

Freeman, John R 1905 Providence, R. I. 

Taylor, Fred. W 1906 Philadelphia. Pa. 

Hutton, F. R 1907 New York. 

HoLArAN, M. L 1908 St . Louis, Mo. 



According to the Constitution, Article C 27, the five Past-Presidents who 
last held the office shall be members of the Council, with all the rights, privi- 
leges and duties of the otlier members of the Council. 



EXECITIVE ( OMMITTKE OF IIIE ( OUN( IL 



Jesse M. Smith, Cluiimian 
Alex. C. Humphreys 



¥. R. HuTTOiN 

Fred J. Miller 



F. M. Whyte 



STANDING COMMITTEES 
1909 

FINANCE 



ArthikM. Waitt (1), Chairmdii 
Edward F. iSchxuck (2) 

Waldo H. Marshall (5) 



Geo. J. Roberts (3) 
Robert M. Dixon (4) 



HOUSE 
Hknry S. Loud (1), Chairman Bernard V. Swenson (3) 

William Cartkk Dickerman (2) Francls Blossom (4) 

Edward Van Winkle (o) 



LIBRARY 

J(JHN W. LiEB, Jr. (4), Chairmnn 
H. H. SUPLEE (1) 

Chas. L. Clarke (5) 

MEET IN (11^ 
Willis Fl. Hall (Ij, Chairiium 
Wm. H. Bryan (2) 

H. DE B. Parsons (5) 



Ambrose Swasey (2) 
Leonard W^aldo (3j 



L. R. Pomeroy (3) 
Charles K. Lucke (4) 



MEMBERSHIP 

Henry D. Hibbard (1), Chairman 
Charles R. Richards (2) 

Hosea Webster (5) 



Francis H. Stillman (3) 
Geor(;e J. FoRAN (4) 



PUBLIC A TION 

Arthur L. Williston (1), Chairman 
D. S. Jacobus (2) 

Geo. L Rockwood (5) 



H. F. J. Porter (3) 
H. W. Spangler (4) 



RESEARCH 



W. F. M. Goss (5), Chairman 
Jas. Christie (1) 



R. C. Carpenter (2) 
R. H. Rice (3) 



Chas. B. Dudley (4) 

Note. — Numbers in parenthe.se.s indicate length of term in years that the member has yet to serve. 

vii 



SPECIAL COMMITTEES 
1909 

On a Standard Tonnage Basis for Refrigeration 



D. S. Jacobus 
A. P. Trautwein 



John E. Sweet 



E. F. Miller 

On Society History 

Chas. Wallace Hunt 

On Constitution and By-Laws 



Cnas. Wallace Hunt, Chairman 
G. M. Basford 



Jesse M. Smith 



On Conservation of Natural Resources 



Geo. F. Swain, Chairman 
Charles Whiting Baker 



g. t. voorhees 
Philip DeC. Ball 



H. H. Suplee 



F. R. HUTTON 

D. S. Jacobus 



L. D. Burlingame 

M. L. HOLMAN 



Calvin W. Rice 
On International Standard for Pipe Threads 



E. M. Herr, Chairman 
William J. BaLDWiN 



On Thurston Memorial 



Alex. C. Humphreys, Chairman 
R. C. Carpenter 

Fred J. Miller 



Geo. M. Bond 
Stanley G. Flagg, Jr. 

Chas. Wallace Hunt 
J. W. LiEB, Jr. 



On Standards for Involute Gears 



Wilfred Lewis, Chairman 
Hugo Bilgram 



D. S. Jacobus, Chairman 
Edward T. Adams 
George H. Barrus 



Gaetano Lanza 

On Power Tests 

L. P. Breckenridge 
William Kent 
Charles E. Lucke 



On Land and Building Fund 
Fred J. Miller, Chairman 

R. C. McKlNNEY 



E. R. Fellows 
C. R. Gabriel 



Edward F. Miller 

Arthur West 

Albert C. Wood 



James M. Dodge 



On Student Branches 

F. R. HuTTON, Honorary Secretary 

viii 



OFFICERS OF THE (iAS POWER SECTION 

1909 

CHAIRMAN 
F. R. Low 

SECRETARY 
Geo. a. Orrok 

GAS POWER EXECUTIVE COMMITTEE 
F. H. Stillman, Chairman G. I. RocKWOod 

F. R. HUTTON H. H. SUPLEE 

R. H. Fernald 

GAS POWER MEMBERSHIP COMMITTEE 
Robert T. Lozier, Chairman D. B. Rushmore 

Albert A. Gary A. F. Stillman 

H. V. O. Goes G. M. S. Tait 

A. E. Johnson George W. Whyte 

F. S. King S. S. Wyer 

GAS POWER MEETINGS COMMITTEE 
Gecil p. Poole, Chairman E. S. McClelland 

R. T. Kent C. T. Wilkinson G. W. Obert 

GAS POWER LITERATURE COMMITTEE 

C. H. Benjamin, Chairman L. S. Marks 

H. R. Gobleigh T. M. Phetteplace 

G. D. Gonlee G. J. Rathbun 
R. S. DE Mitkiewicz W. Rautenstrauch 
L. V. Goebbels S. a. Reeve 
L. V. LuDY A. J. Wood A. L. Rice 

GAS POWER INSTALLATIONS COMMITTEE 
J. R. Bibbins, Chairynan A. Bement 

L. B. Lent 

GAS POWER PLANT OPERATIONS COMMITTEE 

I. E. Mour.TROP, Chairman H. J. K. Freyn G. H. Parker 

J. D. Andrew N. T. Harrington J. P. Sparrow 

W. H. Blauvelt J. B. Klumpp A. B. Steen 

V. Z. Garacristi G. L. Knight F. W. Walker 

E. P. Goleman J. L. Lyon G. W. Whiting 

G. J. Davidson D. T. MacLeod Paul Winsor 

W. T. Donnelly V. E. McMullen T. H. Yawger 

GAS POWER STANDARDIZATION COMMITTEE 
G. E. Lucre, Chairman E. T. Adams 

Arthur West James D. Andrew 

J. R. Bibbins H. F. Smith 

Louis G. Doelling 



OFFICERS OF STUDENT BRANCHES 



STUDENT BRANCH 



Stevens Inst, of Tech.. 

Hoboken, N. J. 
Cornell University. 

Ithaca. N. Y. 

Armour Inst, of Tech., 

Chicago, 111. 
Iceland Stanford, Jr. 

Universiry, Palo Alto. 

Cal. 
Polytechnic Institute, 

Brooklyn, N. Y. 
State Agri. College of 

Oregon, Corvallis, 

Ore. 
Purdue University, 

Lafayette, Ind. 
Univ. of Kansas, 

Lawrence, Kan. 
New York Univ., 

New York 
Univ. of Illinois, 

L'rbana, 111. 
Penna. State College, 

State College, Pa. 
Columbia University, 

New York. 
Mass. Inst, of Tech., 

Boston, Ma.ss. 
LTniv. of Cincinnati, 

Cincinnati, O. 
Univ. of Wisconsin. 

Madison, Wis. 



.\UrHORIZED 


HONORARY CH.\IR- 


I 

PRESIDENT 


SECRETARY 


BY CO UNCI I. 


M\N 






190S 








Decern bei 4 


Alex. C. Humphreys 


H. II. Haynes 


R. H. Upson 


December 4 


R. C. Carpenter 




C. F. Hirshfeld 


1909 








March 9 


C. F. Gebhardt 


X. .1. Houghton 


.M. C. Shedd 


March 9 


W. F. Duran.l 


P. H. \'an Ktten 


H. L. He.ss 



March 9 W. D. Ennis J. S. Kerins 

March 9 Thos. >L Cardner C.L.Knopf 



March 9 L. ^". Ludy 

March 9 P. F. Walker 

November 9 C. E. Hougliton 
November 9 W. F. M. (loss 
November 9 
Novemlier 9 
November 9 
Noveml)er 
Novend)er 9 



E. A. Kirk 
H. S. Coleman 
Harry Anderson 
W. F. Colman 



Fredk. A. Dewey 



Percy Gianella 
S. H. Graf 

.1. R. Jackson 
.John Ciarver 
Andrew Hamilton 
S. C. Wood 



SUMMARY OF MEMBERSHIP 

Dpcemhor31. 1909 
United States 



Alahama 19 

Alaska 1 

Arizona 5 

Arkansas 2 

California 74 

Colorado 30 

Connecticut 144 

Delaware 18 

District of Columbia 32 

Georgia 19 

Hawaii 3 

Idaho 2 

Illinois 241 

Indiana o9 

Iowa 9 

Kansas 11 

Kentucky 6 

Louisiana 30 

Maine 15 

Maryland 33 

-Massachusetts 339 

Michigan 110 

Minnesota 22 

Mississippi 1 

Missouri 64 

^lontana 10 

Address unknown 



Nebraska 3 

Nevada 5 

New Hampshire lo 

New Jersey 202 

New Mexico 2 

New York 1062 

North Carolina 13 

North Dakota 1 

Ohio 277 

Oklahoma 1 

Oregon 11 

Pennsylvania 4.59 

1 

3 

69 

3 

13 

15 

9 

11 

28 



Philippine Islands 

Porto Rico 

Rhode Island 

South Carolina 

Tennessee 

Texas 

Utah 

Vermont 

Virginia 

Washington 15 

West Virginia 8 

Wisconsin 93 

Wyoming 1 

Total 3619 

6 



Foreign Countries 



Africa 14 

Australia S 

Belgium 5 

Canada 45 

Central America 

China 3 

Cuba 4 

England 45 

Finland 1 

France 10 

Germany 8 

Holland 1 

Hungan.- 2 



India 

Italy 

Japan 

Mexico 

New Zealand . . 

Norway 

Russia 

.Scotland 

South America. 

Sweden 

Switzerland. . . . 



4 
1 
9 

14 
1 
1 
3 
3 

11 



Total. 



207 



SUMMARY OF MEMBERSHIP 

Bt Residence 

December 31, 1909 

Membership in United States 3619 

Foreign membership ... 207 

Address unknown 6 

Total 3832 

By Grade 

December 31, 1909 

Honorary members 15 

Members 2565 

Associates 398 

Juniors 854 

Total membership (including life members) 3832 



MEMBERSHIP OF GAS POWER SECTION 



Alabama. 



United States 
.... 4 Missouri . 



California 7 

Connecticut 5 

District of Columbia 5 

Delaware 1 

Georgia 1 

Illinois 22 

Indiana 9 

Kansas 2 

Maryland 1 

Massachusetts 25 

Michigan 13 

Minnesota 5 



Nebraska 2 

New Jersey 14 

New York 141 

Ohio 29 

Pennsylvania 25 



Rhode Island. 

Vermont 

Virginia 

Washington. . 
Wisconsin 



7 
1 
1 
2 
17 



Total 347 



Foreign Countries 



Belgium. . 
Canada. . 
Germany 



Mexico 

Switzerland. 



Total. 



GAS POWER SECTION 

By Residence 

Membership in United States 347 

Membership in foreign countries 6 

Total membership 353 

By Grades 

Members of the Society 222 

Affiliates 131 

Total 353 

STUDENT BRANCHES 

Armour Institute of Technology 16 

Brooklyn Polytechnic Institute 22 

Columbia University 4 

Cornell Universitj^ 123 

Leland Stanford, Jr., University 13 

Massachusetts Institute of Technologj^ 1 

Pennsylvania State College 35 

Purdue University 3 

State Agricultural College of Oregon 10 

Stevens Institute of Technology 55 

University of Cincinnati 23 

University of Illinois 44 

University of Kansas 7 

University of Wisconsin 26 

Total 382 



ATTENDANCE AT MEETINGS, 1909 

The following figures show the attendance at the several meet- 
ings of the Society daring 1909: 

Jjiuuary 12 New York 168 

February 23 New York 133 

March 9 New York 251 

March 24 New York 625 

April 13 New York 307 

May 4-7 Washixgtox, D. C. Sprinc; Meeting .... Members. . .276 

Guests 333 609 

October 12 New York 192 

November 9 New York 161 

December 7-10 New York. Axxual Meeting Members. . .628 

Guests 435 1063 



CONTKNTS OF VOLUME 31 

Washington, New York and Monthly Meetings, 1909 

Page 

Biography of Jesse M. Sinitli 3 

No. 1229 Monthly Meetings, January to June; Washington Meeting. 5 

Xo. 1230 Carl G. Barth The Transmission of Power by 

Leather Belting 29 

Xo. 1231 F. M. Whyte Safety Valves for Locomotives 105 

Xo. 1232 P. G. Darling Safety Valve Capacity 109 

Xo. 1233 Safety Valve Discussion 129 

Xo. 1234 Ellis C. Soper A Unique Belt Conveyor 151 

Xo. 1235 C. Kemble Baldwin Automatic Feeders for Handling 

Material in Bulk 161 

Xo. 1236 W. H. Kenerson A New Transmission Dynamometer 171 

N^o. 1237 A. Kingsbury Polishing Metals for Examination 

with the Microscope 181 

Xo. 1238 C. L. Straub Marine Producer Gas Power 185 

Xo. 1239 C. W. Obert Operation of a Small Producer- 
Gas Power Plant 209 

Xo. 1240 T. ^I. Phetteplace Offsetting Cylinders in Single- 

Acting Engines 223 

Xo. 1241 Presentation of Portrait of George W. Melville 253 

Xo. 1242 G. A. Orrok Small Steam Turbines 263 

Xo. 1243 E. M. IvENs Compressed Air Pumping Systems of 

Oil Wells 311 

Xo. 1244 C. H. Peahody The Specific Volume of Saturated 

Steam 333 

Xo. 1245 R. C. H. Heck Some properties of Steam 345 

No. 1246 H. V. Wille A New Departure in Flexible Stay- 
bolts 359 

Xo. 1247 Hudson-Fulton Celebration 373 

X'o. 1248 Meetings, October to December; Annual Meeting 381 

X"o. 1249 Annual Reports of Council and Committees 409 

X'o. 1250 Jes.se M. Smith The Profession of Engineering 429 

Xo. 1251 R. C. Carpenter High-Pressure Fire-Service Pumps 

of Manhattan Borough 437 

X'o. 1252 Gaetano Lanza 1 Stres.ses in Reinforced Concrete 

L. S. Smith / Beams 511 

X'o. 1253 Walter Rauten- Design of Curved Machine Mem- 

STRAUCH ber.s under Eccentric Lord 559 

Xo. 1254 C. M. Allen Tests on a Venturi Meter for Boiler 

Feed 589 

No. 1255 G. F. Gkhhardt The Bitot Tubd as a Steam Meter. . 601 



CONTENTS 

Page 

No. 1256 F. H Sibley | Efficiency Tests of Steam Nozzles. . 617 

T. S. Kemble J ^ 

No. 1257 C. C. Thomas An Electric Gas Meter 655 

No. 1258 D. M. Myers Tan Bark as a Boiler Fuel 685 

No. 1259 J. R. BiBBiNS Cooling Towers for Steam and Gas 

Power Plants 725 

JSo. 1260 W. P. Caine Governing Rolling Mill Engines. ... 783 

No. 1261 F. W. Dean An Experience with Leaky Vertical 

Fire Tube Boileis 799 

No. 1262 F. W. Dean The Best Form of Longitudinal Joint 

for Boilers 823 

No. 1263 C.M.Garland \ Testing Suction Gas Producers 

A. P. Kratz / with a Koerting Ejector 831 

No. 1264 J. R. BiBBiNS Bituminous Gas Producers 877 

No. 1265 Walter Ferris The Bucyrus Locomotive Pile Driver 905 

No. 1256 Henry Hess .Line-Shaft Efficiency, Mechanical 

and Economic 923 

No. 1267 A. F. Nagle Pump Valves and Valve Areas 953 

No. 1268 A. F. Nagle A Report on Cast-Iron Test Bars. . 977 

No. 1269 I. N. HOLLIS 1 ^^ . r- . r -r.-^^- 

_, .^ .,. 1 Symposium on Cast-iron Jittmgs 

E. F. Miller r ^ a i 4- .i o* nQo 

. „ ^ tor Superheated Steam 989 

No. 1270 Necrology 1039 

No. 1271 Index 1059 



TRANSACTIONS 

OF 
TPIE AMERICAN SOCIETY OF 
MECHANICAL ENGINEERS 



VOLUME 31-1909 



T 



HIS volume contains the papers and proceedings of The Ameri- 
can Society of Mechanical Engineers for the year 1909, 
covering the thirtieth year of the Society's history. 



The newly-elected President, Jesse M. Smith of New York, was 
introduced, as is customary, at the Annual Meeting of the Society 
in December 1908, an account of which appeared in Volume 30 of 
Transactions, covering the proceedings of 1908. 

The annual report of the Council, presented at the annual meet- 
ing of 1909, gives a record of the work of the year, and follows an 
account of that meeting. 



JESSE MEKKICK SMITH 

Jesse M. Smith, President of the Society for 1909, was born 
in Newark, Ohio, in 1848. He moved to Detroit, Michigan, with his 
father's family in 1862. In 1865 he entered Rensselaer Polytechnic 
Institute, Troy, New York, remaining there three years. The follow- 
ing year he spent traveling in Europe, and entered L'Ecole Centrale 
des Artset Manufactures, Paris, France, receiving after three years 
of study the degree of M. E, in 1872. During his vacations, he trav- 
eled among the manufacturing plants of France, Germany and Bel- 
gium, and attended lectures in the Polytechnic Institute in Berlin. 
After graduation in Paris he traveled three months among the iron 
and machine works of England, 

He began the practice of engineering in 1873, designing and super- 
intending the erection of blast furnaces for smelting iron from native 
ores with raw bituminous coal in the Hocking Valley, Ohio. He 
made surveys of coal mines, opened mines and built coal handling 
machinery for them. He surveyed and constructed railroads from 
mines to furnaces. 

Upon the death of his father in 1880, Mr. Smith returned to Detroit 
and opened an oflEice as Consulting Engineer. He designed and con- 
structed a high-speed center-crank steam engine with shaft governor, 
containing the feature of the modern inertia weight governor, and 
put it in operation driving a Brush dynamo producing 40 arc lights 
in 1883, In 1890 he presented a paper before the Society on this 
governor. 

He represented the United States Electric Lighting Company in 
Ohio and Michigan from 1884 to 1886, during which time he erected a 
number of the early incandescent electric light plants, including 
one of 1000 Ughts in the Stillman Hotel, Cleveland, Ohio, which was 
the first hotel lighted exclusively and continuously by electricity 
from its own plant. He returned to the work of consulting engineer 
in 1886 and continued in it until 1898. During this time he designed 
and erected several power plants and several plants for electric Hght- 
ing and electric railways; also apparatus for steam heating with 
exhaust steam in several large manufacturing plants. 

3 



4 JESSE MERRICK SMITH 

He began in 1883 to be called as an expert witness in the U. S. 
courts in patent litigation. This practice gradually grew and dis- 
placed the work of consulting engineer until 1898, when he moved 
to New York to continue his practice as expert in patent cases 
exclusively. 

Among the notable cases in which he has acted as expert are: steam 
injectors under the Hancock Inspirator patents; cylinder lubricators 
for locomotives; roller mills and middlings purifiers for flour manufac- 
ture; cyclone dust collectors; quick action air brakes under Westing- 
house patents; pneumatic tires for automobiles; automobiles under 
the Selden patent; induction electric motors under Tesla patents; 
pressure filters; incandescent electric lamps; steam heating apparatus; 
typewriters; armored concrete construction; the cabulagi-aph, etc. 

He became a member of the Society in 1883 and was a member of 
the Council as Manager, 1891 to 1894, and Vice-President 1894 to 1896 
and 1899 to 1901. 

He is a charter member of the American Institute of Electrical 
Engineers; La Soci6t^ des Ingenieurs Civils de France; 1' Association 
des Anciens fileves de I'Ecole Centrale des Arts et Manufactures; 
The Detroit Engineering Society; the Society for the Advancement 
of Science; the National Geographical Society; the Engineers Club; 
the Machinery Club and the Ohio Society of New York. 



No. 1229 

MEETINGS rJxVNUARV-.JUNE 

NEW YORK MEETING, JANUARY 12 

The first meeting of the Society for the year 1909 was held in the 
Engineering Societies Building, on the evening of January 12, when a 
paper on The Transmission of Power by Leather Belting was given by 
Carl G. Barth. 

This paper, with the discussion, constitutes one of the most compre- 
hensive presentations of the subject of belting that has been given 
before the Society. Mr. Barth has deduced a theory of belting based 
on the well-known experiments of Lewis and Bancroft and other engi- 
neers who have investigated different factors of the belting problem. 
Moreover, in systematizing manufacturing plants and especially 
machine shops, where the scientific operation of machine tools is 
involved, a careful study was made of the whole belting problem, 
resulting in additional material for his paper. 

Following the discussion pertaining strictly to the subject-matter of 
the paper, there was a general discussion upon the transmission of power 
by electricity and by rope, and by the modern types of chains used for 
power transmission. Written discussions were submitted by the fol- 
lowing: A. F. Nagle, Prof. Wm. W. Bu-d, Prof. 0. H. Benjamin, H. K. 
Hathaway, Prof. L. P. Breckenridge, and Prof. W. S. Aldrich. Oral 
discussions were given by Henry R. Towne, Wilfred Lewis, W. D. 
Hamerstadt, Fred. W. Taylor, Charles Robbins, Geo. N. Van Derhoef, 
Walter C. Allen, Dwight V. Merrick, Fred. A. Waldron, S. B. FHnt 
and A. A. Gary. 

NEW YORK MEETING, FEBRUARY 23 

A meeting was held in the Engineering Societies Builduig on Tues- 
day evening, February 23, the subject for discussion being Safety 
Valves. The meeting was opened by Frederic M. Whyte, general me- 
chanical engineer of the New York Central Lines, with a paper upon 
Safety Valves, giving special attention to locomotive practice. He 



6 SOCIETY AFFAIRS 

was followed by Philip G. Darling, mechanical engineer with Man- 
ning, Maxwell & Moore, with a paper on Safety-Valve Capacity. 

The papers and discussion, which covered locomotive, marine and 
stationary practice, as well as the use of safety valves on heating 
boilers, brought together such late data as were available and em- 
phasized the need of further information for the purpose of establish- 
ing a more rational and uniform practice. 

The papers were discussed by: L. D. Lovekin, A. C. Ashton, A. B. 
Carhart, E. A. May, H. O. Pond, F. J. Cole, Dr. Chas. E. Lucke, Jesse 
M. Smith, G. P. Robinson, Wm. H. Boehm, H. C. McCarty, M. W. 
Sewall, Geo. I. Rockwood, A. A. Gary, Dr. A. D. Risteen, F. L. Du- 
Bosque, N. B. Payne, Frank Creelman. The discussion was ad- 
journed to the Spring meeting. 

NEW YORK MEETING, MARCH 9 

A particularly interesting occasion was the lecture on Modern 
Physics, given by Dr. William Hallock, Professor of Physics, Colum- 
bia University, on Tuesday evening, March 9. 

The lecture included a review of discoveries introductory to the 
X-ray, radio-activity and allied phenomena; experimental demon- 
stration of different forms of radiation, including heat; development 
of the essential identity of radiant light, heat and Hertz waves, 
together with the evidence of the electro-magnetic nature of light 
radiations; differentiations between these forms of radiation and 
those of so-called radio-active material, followed by the bearing of 
the facts developed by radio-activity upon the possible genesis of 
the chemical elements; the kinetic theory of gases and its relation to 
the modern theory of solutions; the moving ion as the determining 
factor ijn electrical conduction; the distinction between the chemical 
and the physical ion; the atom and the relation of its structure to the 
phenomena of radiation and absorption; the principle of relativity 
and its relation to the structure of the atom and the electron; the 
universal application of the /orce, mass, time theory to molecular and 
cosmic phenomena. 

JOINT MEETING ON CONSERVATION 

A meeting of the national engineering societies on the conserva- 
tion of our natural resources was held in the Engineering Societies 
Building on the evening of March 24. Onward Bates, President 



SOCIETY AFFAIRS 7 

Am. Soc. C. E., who was expected to preside, was unable to 
attend, and Dr. James Douglas, Past-President Am. Inst. M. E., 
acted as chairman. At the opening of the meeting, the Chairman 
announced a congratulatory telegram from President Taft, which 
was read by John Hays Hammond, President Am. Inst, M, E. 

In his opening remarks Dr. Douglas said that in a great movement 
of this kind there could be no dividing line between engineers in dif- 
ferent branches of the profession. The great inventions like that of 
the Bessemer process had required a combination of the skill of 
engineers who had specialized in different fields. He said that in 
looking back we must be struck with the advance made in the reduc- 
tion of waste in the use of natural supplies, especially in saving coal, 
both in mining it and in using it in metallurgical work. 

The first address was upon The Conservation of Water, by John 
R. Freeman, Past-President Am. Soc. M. E., Mem. Am. Soc. C. 
E., consulting engineer of the Department of Additional Water 
Supply for the City of New York. He spoke of the relation of 
stream flow to lumbering, emphasizing the importance of accurate 
stream measurements in order to obtain precise knowledge of the 
effect of forests and of the value of water powers. Interesting 
figures were given, comparing the efficiency of turbines of the old 
days with those of the present time. Other phases of the conserva- 
tion of water, such as the purity of the water courses, navigation, 
irrigation, etc., were considered. He recommended the collecting of 
facts by the different States, regarding the notable opportunities for 
power development within their borders, and the making of care- 
ful surveys, thus placing reliable information at the disposal of those 
inclined to take advantage of such natural opportunities for power. 

The address of Dr. R. W. Raymond, Secretary Am. Inst. M. E. 
was upon Conservation by Legislation. He defined true conservation 
as the diminution not of use but of waste. The best method for the 
prevention of waste is by the progressive education of the people, 
rather than by legislation. He urged that government information 
pertaining to natural resources and their conservation should be 
collected with care and not hurried, and stated without bias or argu- 
ment in favor of any measure or policy. Hasty and ill-considered 
legislation, especially if advocated by selfish interests, is a peril. He 
dealt with specific examples of such legislation and urged that the 
work of the departments of the Federal government should be care- 
fully planned in advance instead of expanding without a definitely 
arranged plan. 



8 SOCIETY AFFAIRS 

Charles Whiting Baker, Mem. Am. Soc. M. E., Editor of Engi- 
neering News, spoke on The Waste of our Natural Resources by 
Fire. He gave new statistics upon the fire laws in the United States, 
with the striking illustration that we are burning every year in 
this country a street of buildings a thousand miles long that would 
reach from New York to Chicago. That this destruction is not 
necessary is proved by the experience of European countries where 
the per capita fire loss is in most cases only a few cents annually, 
while in this country it is $2.50. Referring to the destruction by 
forest fires, he said that effective laws for the protection of forests 
must be enacted before capital will be invested in the development 
or preservation of timber lands. 

The last address was by Lewis B. Stillwell, Mem, A. I. E. E., con- 
sulting electrical engineer, upon Electricity and the Conservation 
of Energy. He illustrated by interesting figures the function of 
electricity in the conservation of power resources, showing results 
accomplished in three typical cases, namely, the plants of the 
Niagara Falls Power Co., the Northeast Coast Power System at 
Newcastle-upon-Tyne, and the plants of the Interborough Rapid 
Transit Co., New York. The Niagara plant showed the possibility 
in water power development and the Northeast Coast plant the 
economy resulting from the substitution of large steam-driven units 
for small steam plants, widely distributed. In the case of the Inter- 
borough Company, comparisons were made of the cost under the 
present system of electrical distribution and that which would 
have obtained if locomotives had been used instead. 

ST. LOUIS MEETING, APRIL 10 

A meeting of members of the Society residing in St. Louis and vicin- 
ity was called by Wm. H. Bryan, member of the Meetings Committee, 
and held on Saturday evening, April 10. Prof. E. H. Ohle acted as 
secretary and about twenty engineers were present. This was the 
first monthly meeting of the Society to be held outside of New York 
City. All present expressed themselves in favor of local meetings and 
it was voted that a committee of three, composed of the chairman and 
two others appointed by him, should be formed to lay out a plan of 
organization and to report in sixty days. 

The following topics were discussed: a local organization with occa- 
sional professional and social meetings; 6 increase in membership; c 
contributions to The Journal; d making up a party to attend the 



SOCIETY AFFAIRS 9 

Spring Meeting at Washington, May 4-7; c extending an invitation to 
the Society to meet in St. Louis at some future time; /other means of 
promoting the Society's welfare, not only locally, but generally. 

JOHN FRITZ MEDAL AWARD 

The John Fritz Medal, the only medal which the four National 
Engineering Societies confer, was presented to Charles T. Porter, 
Hon. Mem, Am. Soc. M. E., on Tuesday evening, April 13. The pre- 
sentation took place in the auditorium of the Engineering Societies 
Building, before distinguished invited guests and an audience 
representing the entire engineering profession. The medal was 
conferred upon Mr. Porter for his work in advancing the knowledge 
of steam engineering and for improvements in engine construction. 
Addresses were made by Dean W. F. M. Goss of the University of 
Illinois, upon The Debt of Modern Civilization to the Steam Engine 
as a Source of Power; by Prof. F. R. Hutton of Columbia University, 
Honorary Secretary Am. Soc. M, E,, on The Debt of the Modern 
Steam Engine to Charles T. Porter; by Robert W. Hunt of Chicago, on 
The Debt of the Era of Steel to the High Speed Steam Engine; by 
Frank J. Sprague of New York, on The Debt of the Era of Electricity 
to the High-Speed Steam Engine. 

Henry R. Towne, Past-President Am. Soc. M. E., and Chairman 
of the Board of Award of the John Fritz Medal for 1909-1910, presided 
at the meeting, and in his opening remarks spoke briefly of the 
origin and history of the medal, introducing Dean W. F. M. Goss of 
the University of Illinois. 

At the close of Professor Goss's address, Mr. Towne in a short intro- 
ductory speech recalled that Mr. Porter was the third person and the 
first American to whom was accorded the distinction of Honorary 
Membership in The American Society of Mechanical Engineers. On 
account of this relation, Mr. Porter was introduced by Jesse M. 
Smith, President of the Society, who said, by way of introduction: 

The John Fritz Medal, estabUshed in 1902 by the American engineering pro- 
fession as a meed of recognition for 'notable scientific or industrial achievement,' 
was awarded in the year 1908 by a board representing the four National Engineer- 
ing Societies, to a distinguished mechanical engineer for 'his work in advancing the 
knowledge of steam-engineering and for improvements in engine construction.' 
I present to you, and to this company, the engineer to whom this high distinction 
has been granted. 



10 SOCIETY AFFAIRS 

He is honored because he saw the possibiUties of the high-speed steam engine; 
because his mechanical genius in design made those possibilities real; and because 
he recognized the necessity for, and then applied, the very best mechanical con- 
struction to the realization of his ideals. 

He then introduced into the development of the power plant an idea and an 
influence so revolutionary as to make an epoch in the history of the art of engine 
building; and which has been as world-wide in its effects as the use of the recipro- 
cating engine. 

Many of the present generation of engineers have inherited, without effort and 
often without knowledge of their origin, the results which cost him many years of 
painstaking study and experiment to establish. 

That he may receive the John Fritz Medal awarded to him, I now have the 
honor to present Charles Talbot Porter. 

Mr. E. Gybbon Spilsbury, Chairman of the Board of 1908, by 
which the award was made, said in presenting the medal to Mr. 
Porter: 

Under instructions from the Board of Award of the John Fritz Medal, it is my 
privilege and pleasure to inform you that for your work in advancing the knowledge 
of steam engineering and for improvements in engine construction, you have been 
chosen as the worthy recipient of the medal for the year 1908-1909. 

This medal was instituted in 1902 to commemorate the 80th anniversary of the 
successful and honored career of our beloved colleague John Fritz, and its award by 
a committee selected from the membership of the four great engineering societies 
of the United States is the highest honor which the engineering profession can 
V onfer on any of its members. 

Charles Talbot Porter,in the presence of this distinguished company, I now present 
you this medal, together with an engraved certificate of the award, and confer 
upon you all the rights and honors and the distinction which attach to this emblem. 
May you live long and happily to enjoy the appreciation which is your due at the 
hands of those you have so benefited by your work. 

After the presentation, Mr. Towne read the following telegram 
from John Fritz: 

With all my heart regret my inability to be with my dear friends and associ- 
ates this evening. I cannot be with you in person, but I will be with you in 
spirit. Please convey to my dear friend Porter my sincere congratulations and 
best wishes. 

Congratulatory cablegrams were also read from Wm. H. Maw, Editor 
of London Engineering , from the sons and grandsons of Wm. A. Hoyle, 
with whom Mr. Porter was associated in his early work, the Iron and 
Steel Institute of Great Britain, the Institution of Mechanical Engi- 
neers of Great Britain, E. D. Leavitt, Past-President Am.Soc.M.E., 
an early associate of Mr. Porter, and many others. 

The addresses of the evening followed. 



SOCIETY AFFAIRS 11 



ADDRESS OF DEAN W. F. M. QOSS 



Dean W. F. M. Goss spoke of the debt of modern civilization to 
the steam engine. Dreams of the possibilities of steam belong to 
the days of Addison, Steele, Swift and Defoe; days when there were bril- 
liant men of letters, triumphs in architecture, achievements on the 
battlefield, but when there were no means for performing industrial 
work. There were no large factories in England because there was 
no way by which their machinery could be driven. Mines were 
abandoned because they were flooded with water; women and girls 
were toiling in the mines amid suffering and degradation. The move- 
ment of merchandise by land was laborious and traveling by sea slow 
and dangerous. 

Into the midst of such conditions came the steam engine. It freed 
the mines of England from water, revived dormant industries, intro- 
duced new systems of manufacture, supplied power, water and effective 
means of sanitation to cities, and was later supplemented in all these 
respects by electricity for lighting, power and transportation. 
Steam usurped the place of wind in the propulsion of ships, and through 
the agency of the locomotive has carried civilization to the farthest 
ends of the earth. These achievements are direct contributions to 
the upbuilding of civilization, the key-note of which is service. The 
dwellers on the earth are beginning to see that if one nation suffers 
severely, all are likely to suffer in some degree, and they are learning 
sympathy for their fellow-men. 

ADDRESS OF PROF. F. R. BUTTON 

It was assigned to Prof. F. R. Hutton to speak in detail of the debt 
of the reciprocating steam engine to the pioneer work of Mr. Porter. 
It owes to him the first vision of the advantages that come from 
making the crank shaft turn at a high rate of revolution, whereby 
the weight of the motor per horsepower is reduced. From this 
seed-thought has sprung the modern design of the motor for the 
self-propelled vehicle and for the aeroplane. The high speed in- 
volved the solution of difficult problems, owing to the necessity for 
starting and stopping heavy parts of the mechanism in each revo- 
lution. To Mr. Porter we owe the recognition of these problems. 

Perhaps the most important debt of all is the requirement that the 
standard of mechanical construction in the high-speed engine must be 
of the highest type. We owe to Mr. Porter many manufacturing 
details which now are commonplaces of modern practice. 



12 SOCIETY AFFAIRS 

Mr. Porter created a form of steam-engine condenser to be attached 
directly to the engine and operated at a much higher rate of speed 
than that at which the ordinary pump could be used; and finally, 
invented a sensitive steam engine governor in two forms. 

The address closed with a tribute to Prof. Chas. B. Richards, associ- 
ated with Mr. Porter's early work of designing, and John F. Allen, 
who had conceived many details of the first high-speed engine which 
Mr. Porter combined into a harmonious whole. 

ADDRESS OF ROBERT W. HUNT 

Robert W. Hunt said it was scarcely conceivable that one could 
have witnessed in a single lifetime the remarkable development in 
the steel industry which he had observed since the birth of the Besse- 
mer processes. These accomplishments were made practically pos- 
sible by the discovery of a more rapid power. The early processes 
were deliberate because man was habituated to slow movements. 
The first power came from the slow-turning water wheel; later from 
the slow-speed steam engine. Faster movements were obtained 
through gears and belts. Among the first engineers to attach the 
rolling mill engine direct to its train of rolls were John and George 
Fritz, but the speed of the stroke of their engine was limited. 
Charles T.Porter was the first to give to the rolling-mill a controllable 
direct-connected economical high-speed engine. 

Mr. Hunt referred to two engines in a rolling-mill plant in Troy, 
N. Y.,in 1876. One set of rolls was driven by a walking-beam low-pres- 
sure engine, taken from the 'steamboat Swallow, a Hudson River 
boat, and the other set was driven by Porter-Allen engines. The . 
contrast between the steamboat engine with a slow speed of 35 or 
40 r.p.m., and Mr. Porter's little engines, humming away at high 
speed, and accomplishing much greater results, was an instructive 
sight. 

ADDRESS OF FRANK J. SPRAGUE 

Frank J. Sprague recalled that in 1867, at the French exhibition, 
Charles T. Porter installed two Porter- Allen engines, the only high- 
speed engines exhibited, to drive generators for supplying current for 
lighthouse apparatus. While these engines were not directly coupled, 
it is a curious fact that the piston speeds and revolutions were what 
is common today in isolated direct-coupled plants. In the dozen years 
following, Mr. Porter built many engines with certain common char- 
acteristics, high piston speed and revolutions, solid engine bed and 



SOCIETY AFFAIRS 13 

babbitted bearings, but there was no direct coupliug to dynamos 
until 18S0, when Mr. Porter installed a high-speed engine for Mr. 
Edison in his laboratory at Menlo Park. Shortly after this Mr. Porter 
was invited to construct for the Edison Station at Pearl Street, New 
York, the first of a series of engines for so-called steam dynamos, each 
independently driven by a direct -coupled engine. 

Mr. Sprague likened the relations of electricity and the high-speed 
engine, not to debtor and creditor, but rather to a close partnership, 
an industrial marriage, one of the most important in the engineering 
world, that of the prime mover and the electric generator. Here were 
two machines, destined to be joined together, economizing space, 
increasing economy, augmenting capacity, reducing investments, 
increasing dividends. Primarily and largely due to Mr. Porter, the 
high speed possibilities of the engine were commercially demonstrated. 

BOSTON MEETING, APRIL 16 

A meeting was held in Boston in the auditorium of the Edison 
Building, on April 16, 1909, to discuss the advisability of holding meet- 
ings of the Society in that city. Irving E. Moultrop, Manager of the 
Society, was elected temporary chairman and Ralph E. Curtis tem- 
porary secretary. About 160 members and guests were present, 
including the President and the Secretary of the Society. There was 
a general discussion in which the following participated: Irving E. 
Moultrop, Henry Bartlett, Henry F. Bryant, Vice-President of the 
Boston Society of Engineers, James D. Andrew, Fred R. Low, Paul 
Winsor, Prof. W. W. Bird, Prof. Geo. F. Swain, Prof. L. S. Marks, 
Prof. D. C. Jackson, Prof. Gardner C. Anthony, Francis W. Dean, 
E. G. Bailey, Prof. C. G. Lanza. 

During the discussion there were brief addresses by the President 
and the Secretary. President Smith said that it is the desire of the 
officers of the Society that these meetings be as free and open as is 
consistent with the traditions and the high professional standards 
which the Society has maintained during its thirty years of experi- 
ence. He emphasized the fact that such meetings are meetings of the 
Society as a body rather than of local sections or branches. Papers 
presented would be published in The Journal when accepted, making 
it possible to discuss them in all the cities where meetings are held. 
He emphasized the Society's friendly spirit of cooperation with other 
engineering societies and the particular esteem in which engineers and 
members held the Boston Society of Civil Engineers. 



14 SOCIETY AFFAIRS 

Secretary Rice said it had become evident that two conventions a 
year are not sufficient for a national society and that the holding of 
meetings more frequently in one place does not create a national 
spirit. He said the question before the meeting was, how the 
engineering profession of Boston and vicinity can best get together 
for the common and individual good; and stated that The American 
Society of Mechanical Engineers desires to do what will best serve tlie 
profession. He expressed the hope that that would be accomplished 
by bringing together the various organizations in a common head- 
quarters rather than by the formation of a new organization. Co- 
operation and coordination, he declared, should be the motto of the 
profession. 

ST. LOUIS MEETING, MAY 15 

A meeting was held at the Missouri Athletic Club, St. Louis, on 
May 15, to discuss further the question of holding meetings of the 
Society in St. Louis. William H. Bryan, member of the Meetings 
Committee, presided, and Prof. E. L. Ohle acted as Secretary. 

The report of the committee on organization, recommending that 
the Society cooperate with the Engineers Club of St. Louis in the 
matter of meetings and publication, was presented and discussed by 
the following: M. L. Holman, Past-President Am.Soc. M.E., R. H. Tait, 
Wm. H, Bixby, Thomas Appleton, Professor Westcott, F. L. Jefferies, 
W. M. Armstrong, Prof. H. Wade Hibbard, J. A. Laird, Victor Hugo, 
E. A. Fessenden. In opening the discussion Mr. Holman said that 
the question of enlarging the sphere of usefulness of national engineer- 
ing societies without interfering with local organizations is one which 
has been given much thought by the engineers of the country. It 
has taken years to bring the St. Louis Engineers Club, an earlier 
organization than The American Society of Mechanical Engineers, up 
to its present standing and it could not afford to take steps which 
would interfere with its usefulness or impede its growth and impor- 
tance. This movement, however, was not intended to antagonize 
local clubs but was for the purpose of bringing more engineers into 
the societies, both local and national, thus benefiting both organiza- 
tions. 

Following the discussion, the report was unanimously adopted and 
the meeting concluded with an interesting running account, given by 
Professor F. H. Vose, of the papers and discussion presented at the 
Washington meeting. 



SOCIETY AFFAIRS 15 

BOSTOiN MEETING, JUNE 11 

In accordance with the plans of the prelimhiary meeting of May 
15, a professional meeting was held on June 11, and the paper on 
Small Steam Turbines given by George A. Orrok at the Washington 
meeting was presented for further discussion. 

Prof. Ira N. HoUis, who presided, first outlined the work proposed for 
the meetings of the Society in Boston, saying that the committee was 
planning a number of meetings to be held during the fall and winter. 
The subjects of several unusually timely papers promised for these 
meetings were announced. As indicating the large number of engi- 
neers in the vicinity who might attend, it was stated that notices had 
been sent to 950, of which number 340 were members of the Society. 

The meeting was addressed briefly by Secretary Rice, after which 
Mr. Orrok's paper was read by Prof. E. F. Miller and the following 
joined in the discussion: Dr. L. C. Loewenstein, J. A. London, Chas. 
B. Rearick, F. B. Dowst, Chas. B. Edwards, V. F. Holmes, J. S. Schu- 
maker, Prof. C. A. Read, Prof. I. N. Hollis, Prof. E. F. Miller, John T. 
Hawkins, R. H. Rice, Chas. H. Manning, C. P. Crissey, Chas. B. 
Burleigh. 



IQ SOCIETY AFFAIRS 



SPRING MEETING, WASHINGTON, D. C. 

The 58th meeting of the Society was held in Washington, D. C, at 
the New Willard Hotel, May 4-7. The total registration was 609, of 
which 276 were members. Fewer professional sessions than usual were 
arranged by the Meetings Committee in order that visiting members 
and their guests might avail themselves of opportunities to see places 
of interest at the national capital. 



PROGRAM 

OPENING SESSION 
Tuesday Evening, May 4, «' ^-^^ o'clock 

Reception of the members by the Washington Society of Engi- 
neers and the local members at the New Willard Hotel. Music by 
the Marine Band. 

Address of welcome by Hon. Henry B. F. Macfarland, President 
of the Board of District Commissioners. 

Response by Jesse M. Smith, President of the Society. 

SECOND SESSION 

Wednesday Morning, May 5 

Business Meeting: Reports of committees, tellers of election; new 
business. 

A Unique Belt Conveyor, Ellis C. Soper. 

Discussed by T. A. Bennett, Harrington Emerson, Fred J. 
Miller. 
Automatic Feeders for Handling Material in Bulk, C. Kemble 
Baldwin. 

Discussed by T. A. Bennett. 
A New Transmission Dynamometer, Prof. Wm. H. Kenerson. 

Discussed by A. F. Masury. 
Polishing Metals for Examination with the Microscope, 
Albert Kingsbury. 



SOCIETY AFFAIRS 17 

Wednesday Afternoon 
Special exhibition drill by troops at Fort Myer. 

Wednesday Evening 

Illustrated lecture by Arthur P. Davis, Chief Engineer of theU. S. 
Reclamation Service, on Home-Making in the Arid Regions. 

THIRD SESSION 
Thursday Morning, May G 

GAS power session 

Report of the Standardization Committee. 
Marine Producer Gas Power, C. L. Straub. 

Discussed by Geo. Dinkel, Henry Penton, I. E. Moultrop, H. 
M. Wilson, E. T. Adams. 
I Operation of a Small Producer Gas Power Plant, C.W. Obert 
Discussed by J. A. Holmes, J. H. Norris, W. A. Bole. 
A Method of Improving the Efficiency of Gas Engines, Thos. 
E. Butterfield. 

Discussed by A. M. Greene, Jr., W. 0. Barnes. 
Offsetting Cylinders in Single-Acting Engines, Prof. T. M. 
Phetteplace. 

Discussed by W. H. Herschel, J. H. Norris. 

Thursday Afternoon 

Reception of members by William H. Taft, President of the United 
States. 

Thursday Evening 

Address, The Engineer in the Navy, by Rear-Admiral George W. 
Melville, Ret., Past-President Am. Soc. M. E. 

Address, Rear-Admiral Melville's Service to the Engineer- 
ing Profession and to the Nation, and presentation to the National 
Museum of a portrait of Rear-Admiral Melville; by Walter M. McFar- 
land of Pittsburg, Pa. Acceptance of the portrait by Dr. C. D. Walcott, 
Secretary of the Smithsonian Institution, representing the Nation. 



18 SOCIETY AFFAIRS 

FOURTH SESSION 
Friday Morning, May 7 

PROFESSIONAL SESSION 

Small Steam Turbines, George A. Orrok. 

Discussed by W. D. Forbes, R. H. Rice, Prof. R. C. Carpenter, 
H. Y. Haden, F. D. Herbert, W. E. Snyder, W. T. Don- 
nelly, F. H. Ball, C. A. Howard. The discussion was 
continued at the Boston meeting, June 11. 
Oil Well Tests, Edmund M. Ivens. 

Discussed by F. A. Halsey, S. A. Moss, J. E. Callan. 
Safety-Valve discussion, continued from the ebruary meeting in 
New York: F. L. Pryor, E. F. Miller, G. H. Musgrave, A. B. Car- 
hart, S. B. Paine, M. W. Sewall, A. C. Ashton, A. F. Nagle, J. J. Aull, 
A. J. Hewlings. 

Specific Volume of Saturated Steam, Prof. C. H. Peabody. 

Discussed by Prof. W. D. Ennis. 
Some Properties of Steam, Prof. R. C. H. Heck. 

Discussed by S. A. Moss, G. A. Goodenough. 
A New Departure in Flexible Staybolts, H. V. Wille. 

Discussed by Wm. Elmer, W. E. Hall, Alfred Lovell, F. J. Cole. 

Friday Afternoon 
Trip by boat to Mt. Vernon. 

LOCAL COMMIITEE 

Walter A. McFakland, Chairman 

Gtjstav Ayrbs Hervey S. Knioht 

Albert H. Buckler Walter R. Metz 

Chari^s Eli Burgoon George L. Morton 

Howard A. Coombs Harold P. Norton 

James B. Dillard Willard L. Pollard 

William A. E. Doying John E. Powell 

Charles E. Foster Alfred H. Raynal 

H. A. GiLLis William B. Ridgely 

James Hamilton W. E. Schoenborn 

Frederick E.Healy George R. Simpson 

Herman Hollerith Charles F. Sponsler 

J. A. Holmes Lucien N. Sullivan 

Arthxir E. Johnson Wiluam B. Upton 

Frank B. King Charles V. C. Wheeler 

Earl Wheeler 



SOCIETY AFFAIRS 19 

Committee of the Washington Society of Engineers 

W. A. McFarland, Mem. Am. Soc. M. E., Chairman 

A. E. Johnson, Mem. Am. Soc. M. E. 

A. H. Raynal, Mem. Am. Soc. M. E. 

W. E. Schoenborn, Mem. Am. Soc. M. E. 

W. B. Upton, Mem. Am. Soc. M. E. 

H. W. Fuller, Mem. Am. Inst. E. E. 

John C. Hoyt, Mem. Am. Soc. C. E., Secretary Washington Soc. Engrs. 

D, S. Carll, Mem. Am. Soc. C. E., President Washington Soc. Engrs. 

Chairman of the Ladies' Committee, Mrs. James Loring Lusk 

ACCOUNT OF THE MEETING 

The Convention opened on Tuesday evening with a reception in 
the large assembly hall of the New Willard, followed by dancing, 
with music by the Marine Band. The reception was largely attended 
and the occasion was a brilliant one. As the guests arrived they were 
received by the President and Mrs. Smith, Mrs. W. L. Marshall, Mrs. 
Charles D. Walcott, and Mrs. F. H. Newell. 

D. S. Carll, President of the Washington Society of Engineers, 
called the assembly to order at 9 o'clock, and extended a hearty 
welcome to the Society on behalf of its local members and of the 
Washington Society of Engineers. He then introduced Hon. Henry 
B. F. Macfarland, President of the Board of Commissioners of the 
District of Columbia. 

Speaking on behalf of these same bodies and of the District of 
Columbia, Mr. Macfarland referred especially to the work of engineers 
in the city of Washington, and said in part: There is a particularly 
warm welcome for the Society in the national capital, since engineers 
more than the men of any other profession have made it what it is. 
George Washington, in the year of the birth of the Constitution, 
conceived the idea of a magnificent capital, then ridiculously out of 
proportion to the youth, weakness and poverty of the new nation. 
L' Enfant and Ellicott in the beginning, and a long line of able and 
brilliant engineers since then, chiefly of the United States Army, have 
rendered important service in carrying out his plans. The past nine 
years, the great municipal building period of the city, have been 
occupied with such engineering feats as the installation of the filtra- 
tion plant, the sewage disposal system, the new pumping system, 
the District government railway terminal work, the District govern- 
ment l)uilding on Pennsylvania Avenue and its approaches, the 
Connecticut Avenue bridge, and others of a similar character. Wash- 
ington appreciates engineers. 



20 SOCIETY AFFAIRS 

President Smith in responding for the Society extended the thanks 
of the members for this cordial welcome and their appreciation of 
the interesting program prepared for their pleasure and entertain- 
ment by the committees of the Washington Society of Engineers 
and of the local membeiis. 

Business Meeting Wednesday Morning, May 5 

The report of the tellers of election was received and there being no 
objection the President declared the names presented duly elected 
to membership in the Society. The Ust follows this report. 

Mr. Smith in behalf of the Membership Committee presented the 
following proposed amendments: 

C 10 An Associate shall be 30 years of age or over. He must have been so 
connected with some branch of engineering, or science, or the arts, or indus- 
tries, that the Council will consider him qualified to cooperate with engineers 
in the advancement of professional knowledge. He need not be an engineer. 

The committee recommends the following to be added at the end 
of C 11 of the constitution. 

A person who is over 30 years of age cannot enter the Society as a Junior. 

The report of the Membersliip Committee pubUshed in Transactions, 
Vol. 30, p. 550, gave in full the reasons for desiring the change. 
The proposed amendments were discussed, and in accordance with 
the rules governing the amendments to the constitution, were re- 
ferred to the annual meeting for final action. 

Prof. Ira H. Woolson, who was a member of the Membership Com- 
mittee for five years, heartily commended the proposed change and 
hoped it would become a part of the constitution. 

Prof. F. R. Hutton proposed an amendment to C 45, adding 
"Public Relations Committee" after "House Committee." 

In view of the fact that it has been brought to the notice of the 
Society that a movement is under consideration to increase and 
improve the facilities for the work of the United States Patent Office, 
Prof. F. R. Hutton introduced the following resolution: 

Resolved, That this Society in convention assembled requests the 
Council of the Society to consider the desirabihty of taking some 



SOCIETY AFFAIRS 21 

action in furtherance of the movement to increase the Patent Office 
facilities, and, if deemed advisable, that they request the individual 
members to take steps to urge their influence to this end upon their 
Senators and Representatives. 

The resolution was voted by the meeting. 

Professional Session, Wednesday Morning 

Four papers were presented at tliis morning session, two of which 
related to the conveying of materials. The first was upon A Unique 
Belt Conveyor, by Ellis C. Soper, of Detroit, Mich., and described an 
installation consistiitg of a conveyor one-quarter mile long, so 
located on an incline tliat less power is required to operate it empt}- 
than when loaded. Datajupon performance were given. The 
second was upon Automatic Feeders for Handling Material in Bulk 
by C. Kemble Baldwin, of Chicago, 111. This contained outline 
drawings and descriptive matter upon different designs of feeders, 
to enable the engineer to select the type best suited to his needs. 

The third paper was upon A New Transmission Dynamometer, 
by Prof. Wm. H. Kenerson of Providence, R. I. This is made in the 
form of a shaft coupling. The apparatus contains an oil chamber, 
one side of which is a diaphragm, and is so arranged that pressure 
is brought against this diaphragm directly proportional to the amount 
of power transmitted. A gage or other registering apparatus is con- 
nected with the oil chamber by a small tube which indicates the press- 
ure and the water power transmitted. 

The last paper was upon Polishing Metals for Examination with 
the Microscope by Albert Kingsbury, Pittsburg, Pa., in which he 
described the use of a polishing machine carrying discs faced with 
common paraffin and charged with wet abrasives. This produces 
excellent surfaces on all the harder metals and alloys, but has not 
proved serviceable upon the soft metals, such as lead. 

Wednesday Evening Lecture 

On Wednesday evening Frederick H. Newell, director of the U. S, 
Reclamation Service, was expected to lecture on Home Making in 
the Arid Regions. As he could not be present a lecture on this sub- 
ject was given instead by Arthur P. Davis, Chief Engineer of the 
Reclamation Service. 

The United States Reclamation Service in its seven years of exist- 
ence has undertaken 26 projects situated in 16 different states and 



22 SOCIETY AFFAIRS 

territories of the West. It has invested in construction about S40,- 
000,000. Nineteen projects have been brought to a point where 
some land is now under irrigation. Water is ready for delivery to 
about half a milUon acres. An average of about 10,000 laborers are 
employed on this work, and over 55,000,000 cu. yd. of rock and 
earth have been excavated. Over 2000 miles of canals have been 
built and 56 tunnels have been bored, which have a total length of 
over 13 miles. 

Twelve large earthern dams and one high masonry dam have been 
completed, and two other masonry dam? which will rank among the 
highest dams in the world are in an advanced s^age of construction. 
Many of these projects are in remote localities into which roads had 
to be built, some of which were carved in precipitous rock, or tun- 
neled through mountains. In the aggregate 342 miles of roads 
and 793 bridges have been constructed. 

In some localities, especially on the Pacific slope, the mild climate, 
and the nearly perpetual sunshine, produce remarkable results in 
the growth of fruits, which for color, flavor and physical perfection 
cannot be equaled in a more humid climate. The chemical force 
in sunshine and a perfectly regulated water supply are also evident 
in the yields of vegetables and forage crops. 

The lecture was illustrated by many beautiful slides. 

Gas Power Section 

At this session, F. R. Low, Chairman of the Gas Power Section, 
presided, and Geo. A. Orrok acted as Secretary. Previous to the 
reading of the professional papers reports were received from the 
committees. 

Membership Committee: The report showed a total member- 
ship of 302, of which 177 were members of The American Society 
of Mechanical Engineers and 125 were affiliates. The Membership 
Committee is thoroughly organized with representatives in differ- 
ent cities. 

Literature Committee: Prof. C. H. Benjamin gave a verbal 
report of this committee stating that the committee is organized 
for work and had laid out a tentative program. It was hoped to 
index the books on the subject of gas power and articles in periodi- 
cals dealing with gas power and aUied subjects; also to present reviews 
of new books and abstracts of important articles. There would be 
two fields for work: one, a permanent one, and the other in the line of 



SOCIETY AFFAIRS 23 

current work relating to popular reviews and abstracts for the benefit 
of members. 

Plant Operations Committee: A verbal report offered by living 
E. Moultrop reported progress and stated that standard forms for 
obtaining operating data on gas power plants were in preparation. 
The committee has a large membership and is widely scattered so 
that it had been impossible to arrange a meeting, but the work had 
been advanced as far as possible by correspondence. 

Mr. C. L. Straub presented a report on gas-producer development 
abroad, an abstract of which appears as part of his paper on 
Marine Producer Gas Power, included in this volume. 

Mr. Orrok stated with reference to the work of committees that it 
is conducted with the idea that as the Gas Power Section has been 
formed while the art is young it will be possible to place a record of 
its development on file at the headquarters of the Society. Such 
data in connection with the large library will place at the disposal 
of anyone interested in the industry the available information upon 
the subject of gas power. 

Following the presentation of the reports came the professional 
papers, the first of which was on Marine Producer Gas Power, by 
C. L. Straub of New York. This paper explained the conditions 
opposing the earlier adoption of producer gas power for marine service 
and gave a summary of marine gas power plants in operation at 
present. It compared the updraft and downdraft of producer gas 
apparatus and contained comparative drawings of the steam equip- 
ment and two types of producer gas equipment for a 306-ft. boat. 

The next paper was upon The Operation of a Small Producer Gas 
Power Plant, by C. W. Obert of New York. It presented a general 
description of a producer gas power plant in the Westchester market 
building of Swift & Company, Bronx Borough, New York. The 
author outlined the operating and maintenance system developed 
for keeping producers and engines in proper condition for continuous 
operation. 

A paper was presented upon A Method of Improving the Efficiency 
of Gas Engines, by Thomas E. Butterfield of Philadelphia, Pa. It 
related to the securing of higher efficiency by reducing the clearance 
and increasing the compression and referred especially to a method 
of diluting with an inert gas the charge drawn in during the suction 
stroke of an Otto cycle engine. By this means premature ignition 
and other troubles incident to high compression are avoided. 

The last paper of the session was upon Offsetting Cyhnders in 
Single-Acting Engines by Prof. T. M. Phetteplace of Providence, 



24 SOCIETY AFFAIRS 

R. I. It gave the results of an investigation of this subject in which 
the various factors entering into tlie problem were taken into account. 

Presentation of Portrait of Rear-Admiral Melville 

On Thursday evening, a portrait of Rear-Admiral Geo. W. Melville 
was presented to the National Gallery at a ceremony held in the audi- 
torium of the New Willard. President Smith presided over the large 
audience which assembled and an address was made by Rear-Admiral 
Melville on The Engineer in the Navy. Mr. Walter M. McFarland 
of Pittsburg, Pa., gave an appreciation of Melville as a man and of his 
work for the nation and profession. The portrait was accepted for the 
Nation by Dr. Chas. D. Walcott, Secretary of the Smithsonian Insti- 
tution. 

At the conclusion of the ceremony, President Smith asked that 
Mr. Sigismond de Ivanowski, the Russian artist who had produced 
so admirable a likeness of Melville, be escorted to the platform. 
Mr. de Ivanowski told briefly and simply of his attempt to portray 
the strong characteristics of his subject and displayed evident 
pleasure that his efforts were so warmly appreciated. 

Abstracts of the addresses are published with the professional 
papers in this volume. 

Professional Session, Friday Morning 

Five papers, and a continuation of the Safety Valve discussion 
given at the February meeting in New York, were scheduled for this 
session. The first paper was upon Small Steam Turbines by Geo. 
A. Orrok of New York. The various types of turbines now on the 
market were illustrated and described and a number of steam con- 
sumption curves were given to demonstrate the economy that might 
be expected from machines of this type. 

A paper on Compressed Air Pumping Systems of Oil Wells by 
Edmund M. Ivens of New Orleans, La., was read, in which a descrip- 
tion was given of compressed air plants at Evangeline, La., oil fields, 
and the results of tests upon these plants with different types of 
apparatus. 

This was followed by the Safety Valve discussion, continued from 
the February meeting in New York. 

Two papers were next presented upon the properties of steam: one 
by Prof. C. H. Peabody, of Boston, Mass., on Specific Volume of 



SOCIETY AFFAIRS 25 

Saturated Steam, and the other upon Some Properties of Steam by 
Prof. R. C. H. Heck of New Brunswick, N. J. The former reviewed 
the results of experiments which might form the basis of a computa- 
tion of specific volumes at various temperatures and compared the 
computed results with experimental determinations of the same quanti- 
ties. The latter paper summarized the important work of Holborn 
and Henning and compared the results of other investigators. These 
two papers constituted another step ahead in the work that is now 
being accomplished toward securing accurate information upon the 
properties of both saturated and superheated steam. 

The last paper was by H. V. Wille, Philadelphia, Pa., on A New 
Departure in Flexible Staybolts. This paper proposed the employ- 
ment of tempered spring steel in the manufacture of the stems of 
staybolts, the ends being of soft steel so as to permit riveting over in 
the boiler. 

This session closed with a unanimous resolution extending the thanks 
of the Society to those who had afforded such abundant entertain- 
ment to their visitors. 

Entertainment 

During the convention an information bureau was conducted 
at the Society headquarters by Chairman Walter A. McFarland of 
the Local Committee, where the various excursions were organized. 
These included trips not only to the public buildings, but to govern- 
ment institutions and other points of technical interest, among which 
were the Bureau of Standards, the station of the Potomac Electric 
Power Co., the Union Railway Terminal, the Naval Gun Factory, 
the District pumping stations, etc. 

At the ladies' headquarters in the New Willard, tea was served each 
day from four to six o'clock and the visiting ladies as well as many 
members of the Society, accepted the hospitality extended by the 
ladies at this time. Sight-seeing automobile trips for the ladies were 
also arranged on each day, which were largely patronized and greatly 
enjoyed. 

On Wednesday afternoon alarge number took the trip to Fort Myer 
to see the exhibition drill, and the evolutions performed by the 
several troops, the unusual skill of both riders and drivers and the 
thoroughly trained horses, called forth round after round of applause. 
Two battalions of artillery with guns and caissons went through 
evolutions of great complexity, and two troops of cavalry through 



26 SOCIETY AFFAIRS 

various formations, apparently equally difficult, while a troop of bare- 
back riders captivated the audience by their horsemanship. 

On Thursday afternoon the reception of members and guests by 
President Taft in the East room of the White House was very 
generally attended. 

On Fridaj'- afternoon following the professional session, many went 
by boat to Mt. Vernon to visit the estate and home of Washington, 
and a wreath from the Society was placed on his grave. 

A trip to Fort Myer was also made and the dirigible balloon of 
the Signal Corps of the United States Army inspected. 

ELECTIONS TO MEMBERSHIP 

The following were declared elected to membership in the Society 
upon the ballot of May 5, 1909, and their election reported at the 
Washington Meeting: 



Ahlquist, H., Syracuse, N. Y. Mackenzie, Edmund, Brooklyn, N. Y. 

Babbitt, Edward F., Columbus, O. Mayall, E. L., Racine, Wis. 
Behrend, Bernard A., Milwaukee, Wis. Morat, J., Yonkers, N. Y. 

Billings, A. W. K., Havana, Cuba. Peck, Eugene Colfax, Cleveland, O. 

Blaisdell, Benjamin H., Manila, P. I. Plunkett, Charles T., Adams, Mass. 

Bloemeke, R. B., New York. Puchta, Edward, Chicago, 111. 

Borde, G. U., New Orleans, La. Richardson, L. S., Rosebank, S. I., N. Y. 

Bruckner, R. E., New York. Righter, Addison A., Chicago, 111. 

Burt, Clayton R., Rockford, 111. Riley, Joseph C, Boston, Mass. 

Bushnell, Douglas Stewart, New York. Roberts, Alvin L., Milwaukee, Wis. 

Crockard, Frank H., Birmingham, Ala. See, Alonzo B., New York. 

Davis, Alfred C, E. Liverpool, O. Shaw, James C, Kobe, Japan. 

Duncan , Albert Greene, Boston, Mass. Sheridan, Richard B., Cleveland, O. 

Ennis, J. B., Paterson, N. J. Shouvlin, Patrick J., Springfield, O. 

Fletcher, E. LesUe, Bridgeport, Conn. Smith, Wm. W., Mexico, D. F., Mexico. 

Funk, Nelson E., New York. Stacks, H. Roy, Philadelpliia, Pa. 

Garvin, George K., New York. Still, F. R., Detroit, Mich. 

Hazelton, W. S., Detroit, Mich. Svensson, J. Alfred, Brooklyn, N. Y. 

Hem, H. O., Kansas City, Mo. Thomas, Horace T., E. Lansing, Mich. 

Hogue, Oliver DriscoU, Boston, Mass. Thompson, Sanford Eleazer, Newton 
Hunter, John A., Pittsburgh, Pa. Highlands, Mass. 

Jewett, A. C, Orono, Me. Tiplady, John T., Cleveland, O. 
Johnstone, F. W., Mexico City, Mexico.Tobin, R. P., Boston, Mass. 

Jones, Walter J., New York. Trefts, John C, Buffalo, N. Y. 

Kellogg, Harry F., Chicago, 111. Vail, Jesse A., Beloit, Wis. 

Knight, Alfred H., Ann Arbor, Mich. Waters, W. L., Pittsburgh, Pa. 

Kranz, William George, Sharon, Pa. Wells, Robert G., Kalimati, India. 

McGuire, Charles H., Denver, Colo. Wills, C. Harold, Detroit, Mich. 



SOCIETY AFFAIRS 



27 



PROMOTION TO MEMBER 



Bibbins, James R., E. Pittsburgh, Pa. 
Brown, J. Rowland, Mansfield, O. 
Castanedo, Walter, New Orleans, La. 
Chatard, William M., Baltimore, Md. 
Cooke, Morris L., Germantown, Pa. 
Dale, Orton G., New York. 
Grover, Marcus A., Birmingham, Ala. 
Hawks, Arthurs., Bethlehem, Pa. 
Heisler, F. W., St. Marys, O. 
Hunter, James F., New York. 



Keith, Thomas Marshall, New York. 
Kilgour, Dwight Foster, Boston, Mass. 
Lea, Henry I., Chicago, 111. 
Pomeroy, L. R., New York. 
Robinson, G. P., Albany, N. Y. 
Roe, Mark W., Akron, O. 
Shiebler, M., New York. 
Swan, John J., Cynwyd, Pa. 
Whitted, Thomas B., Charlotte, N. C. 



ASSOCIATES 



Blanchard, A. S., E. Orange, N. J. 
Bryce J. Wares, New York. 
Carpenter, A. O., Franklin, Pa. 
Castle, S. N., London, England. 
Clancy, George W. A., Readville, Mass. 
Clark, Frank S., Cincinnati, O. 
Fuller, Ray W., Scranton, Pa. 
Hart, Robert W., Winchester, Mass. 



Howell, Frank Scott, New York. 
Koenig, Adolph G., New York. 
Pellissier, G. E., New York. 
Sanguinetti, Philip C, New York. 
Shields, George Rex, New York. 
Vincent, Arthur S., Brooklyn, N. Y. 
Willson, Ernest M., Charles City, la. 



PROMOTION TO ASSOCIATE 



Brooks, Paul R., Peabody, Mass. 
Davoud, V. Y., Provo, Utah. 
Dillard, James B., Washington, D. C. 



Idell, Percy C, Hoboken, N. J. 
Marshall, W. C, New Haven Conn. 



Aldrich, Chester S., Boston, Mass. 
Baendcr, F. G., Iowa City, la. 
Bailey, H. Morrell, Johnstown, Pa. 
Bateman, George W., Claremont, N. H 
Bleyer, Charles F., Milwaukee, Wis. 
Bond, Francis M., Washington, D. C. 
Brown, Harry W., Allston, Mass. 
Daugherty, Frank, Philadelphia, Pa. 
Duncombe, Frederic H., New York. 
I-essenden. C. H , Ann Arbor, Mich. 
* afford, B. T., Lebanon, Ind. 
Hamilton, Chester B., Toronto, Can. 
Hildenbrand, Harry, Houston, Texas. 
Home, Harold Field, Mohegan, N. Y. 
Hudson, Robert A., San Francisco, Cal. 
Jenks, Glen Fay, Philadelphia, Pa. 
Kessler, Armin G., Ithaca, N. Y. 
Lawrence, Gerald Peirce, Columbus, O. 
Lawrence, S. E., Galveston, Texas. 
Leahy, Frank E., Clairton, Pa. 



Lee, Ralph A., Brooklyn, N. Y. 
McGlone, R. G., Galveston, Texas. 
Mack, George J., Albany, N. Y. 
■ Meyer, C. Louis, New York. 
Minck, Peter, Town of Union, N. J. 
Newcomb, Robert S., New York. 
Nicholl, John S., Yokohama, Japan. 
Olmsted, George C, Milan, O. 
Otto, Henry S., New York. 
Phelps, Charles C, New York. 
Pinner, Seymour W., Ann Arbor, Mich. 
PuUs, W. Eugene, Saylesville, R. I. 
Rupp, M. E., Culebra, C. Z., C. A. 
Scheel, H. Van Riper, Passaic, N. J. 
Searle, Wilbur C, Worcester, Mass. 
Shenberger, G. H., Lansford, Pa. 
Simpson, William K., New York. 
Singer, Sidney C, Syracuse, N. Y. 
Slee, Norman S., Barberton, O. 
Smith, CD., Pittsburgh, Pa. 



28 



SOClteTY AFFAIRS 



Juniors — Continued 



Smith, Edward S., RoUa, Mo. 
Stanton, Alden D., Brooklyn, N. Y. 
Stanton, F. A. O'C, Hoboken, N. J. 
Taylor, J. W., Massillon, O. 
Thomas, Fred H., Mt. Vernon, O. 
Thompson, Edward C, Boston, Mass. 
Thurston, Edward D., Jr., New York. 



Tuttle, I. E., Brooklyn, N. Y. 

Wegg, David Spencer, Jr., Provo, Utah. 

Whitcomb, L. A., Brooklyn, N. Y. 

Whiting, R. A., New York. 

Wiley, J. M., New York. 

Woolley, Harold O., Dansville, N. Y. 

Wyman, A. H., Milwaukee, Wis. 



By direction of the Council announcement is also made of the elec- 
tion of H. K. Hathaway of Philadelphia, Pa., elected on the ballot of July 
25, 1907, as an Associate Member, but announcement of whose election has 
not previously been made. 



No. 1230 

THE TRANSMISSION OF POWER BY LEATHER 

BELTING 

CONCLUSIONS BASED PRINCIPALLY ON THE EXPERIMENTS OF 
LEWIS AND BANCROFT 

By Caul G. Bauth, Philadelphia, Pa. 
Member of the Society 

In his paper, Slide Rules in the Machine Shop as a Part of the Taylor 
System of Management, read December 1903, the writer referred to an 
improved theory and new formulae developed by him for the pulling 
power of belting, wliich had been applied in connection with the slide 
rules described. He also stated his expectation of presenting his 
theory and conclusions to the Society at some future time. 

2 These conclusions have since been successfully applied in prac- 
tice by the extensive daily use of these slide rules in task-setting for 
machine operations, and the present paper was prepared with the 
general view of submitting this theory to the Society for the criticism 
and consideration of members who are interested in this subject, and 
with the special view of supplementing Mr. Taylor's paper, On the 
Art of Cutting Metals. All the experimental and mathematical data 
for the slide rules were presented in his paper, excepting the data upon 
the pulling power of belting, an important element in these slide rules 
when applied to belt-driven machines. 

3 The theory to be presented is only an additional attempt, and it 
is hoped a fairly successful one, to do something along lines repeatedly 
touched by various other investigators among the members of the 
Society, and the writer is glad to acknowledge his indebtedness to 
nearly all of these, as his work has principally consisted in taking 
advantage either of carefully conducted experiments recorded by 
them, or of suggestions of various kinds that have stimulated his 
interest and set him thinking. 

Prosonted at the New York monthly meeting (January 1909) of The 
t^MEKicAN Society of Mechanical Engineers. 



30 TRANSMISSION OF POWER BY LEATHER BELTING 

4 Most notable is the paper presented by Mr. Wilfred Lewis at the 
Chicago meeting in 1886, in which he recorded a series of experiments 
conducted by himself in the spring of 1885, under the direction of 
Mr. J. Sellers Bancroft and at the works of William Sellers & Com- 
pany, Philadelphia. In his paper was shown for the first time the 
fallacy of the then universal and still common assumption, that the 
sum of the two tensions in a belt is constant for all loads. It was also 
shown that the coefficient of friction between a belt and its pulley is 
considerably higher than was commonly assumed for ordinary work- 
ing conditions, and that this coefficient varies greatly with the velocity 
of sHp, a fact that has also been pointed out by other investigators. 

5 Mr. Lewis did not, however, even attempt to develop either 
empirical or rational mathematical formulae to represent the facts that 
he established, though his experiments, as will subsequently appear, 
contained all that was necessary for a complete mathematical exposi- 
tion of the subject. 

6 An attempt at an empirical mathematical exposition of the 
relations between the two tensions in a belt in accordance with the 
facts established in Mr. Lewis' paper, was later made by Prof. Wm. 
S. Aldrich in a paper read at the New York meeting in 1898. Mr. 
Aldrich made an original layout of a great number of Mr. Lewis' 
experiments in a manner that seemed to the writer to indicate an 
excellent way to investigate the subject, and which resulted in a dis- 
cussion of Mr. Aldrich's paper, the substance of which has formed the 
basis for all of the writer's subsequent work on the subject. ': 

7 But while the writer's complete theory could have been worked 
out without it, its practical application to the running of belt-driven 
machine tools could never have been made in the present satisfactory 
manner without the facts made known by Mr. Taylor in his paper, 
Notes on Belting, read at the New York meeting in 1893. 

8 In this paper Mr. Taylor showed the economy of running belts 
under much lower tensions than those commonly used, and that the 
ultimate strength of a belt, or rather of the joint in a belt, does not 
form a proper basis for the working tension of a belt, since a belt will 
not long retain a tension that is even a small fraction only of its ulti- 
mate strength (see Fig. 4). However, Mr. Taylor's facts and figures 
were derived from comparatively slow-running belts, and he gave 
nothing that could be directly applied to the higher and more eco- 
nomical belt speeds. The modification and extension of Mr. Taylor's 
ideas to include high speed belts have therefore been part of the 
writer's personal work also. 



TRANSMISSION OF POWER B\ LEATHER BELTING 31 

9 A summary of the writer's work on this subject follows; 

a To establish a mathematical formula for the relation be- 
tween the tension in a belt and the stretch due to this 
tension, based on experiments made at different times by 
Mr. Wilfred Lewis, Prof. W. W. Bird and himself. See 
Fig. 1, 2 and 3 and Par. 1-18 of the Appendix. 

b By means of the knowledge of the elastic properties of 
leather belting expressed by this formula to develop a 
formula for the relations between the tensions in the two 
strands of a belt transmitting power, which formula takes 
account of the influence of the sag in a horizontal belt, and 
agrees substantially with the results of the experiments 
made by Mr. Lewis, when plotted in the manner first done 
by Professor Aldrich. See Fig. 6 and Par, 19-38 of the 
Appendix. 

c To establish a formula to express the relation between the 
coefficient of friction between a belt and a cast iron pulley, 
and the velocity with which the belt slips or slides over the 
pulley, as revealed by plotting the results likewise obtained 
by Mr. Lewis. See Fig. 7 and Par. 48 of the Appendix. 

d The construction of a diagram embodying the formula 
expressing the relation between the two tensions in a 
belt, the well known formula for the loss in effective ten- 
sion due to centrifugal force and the likewise well known 
formula for the ratio between the effective parts of the 
two tensions, as determined by the coefficient of friction 
and the arc of contact of the belt on its pulleys. These 
formulae are so correlated on the diagram that problems 
dealing with the contained variables may be solved 
graphically, while a direct algebraic solution is possible 
only for a vertical belt, or what is the same thing, by 
neglecting the effect of the sag of a horizontal belt. 
See Plate 1 and Par. 11-24. 

e Also, by means of the better knowledge gained of the elastic 
properties of leather belting, to develop a formula for the 
creep of a belt on its pulleys due to the difference in the 
tensions in the two strands, along the lines outlined by 
Professor Bird in his paper on Belt Creep, read at the 
Scranton meeting in 1 905 . See Par. 41-44 of the Appendix. 



32 TRANSMISSION OF POWER BY LEATHER BELTING 

/ The construction of diagrams showing the pulling-power 
and other relations of the two tensions of a belt of 1 sq. 
in, cross section and 180 deg. arc of contact at different 
speeds, under certain conditions and assumptions recom- 
mended by the writer. See Fig. 1, 2 and 3, and Par. 38-52. 
Also a modification of these diagrams for extended prac- 
tical use, on which may be read off: (1) The pulling 
power of a belt of any width and thickness and any arc 
of contact between 140 and 180 deg.; (2) The initial 
tensions below which the belt must not be allowed to fall 
in order to confine the slip and the consequent loss of 
efficiency of transmission within certain limits; (3) The 
initial tension to which it is recommended that the belt be 
re-tightened after falling to this minimum limit. See 
Plate 2 and Par. 53-66. 

g Finally, the construction of a slide rule serving the same 
purpose as the diagram just mentioned, but which is 
much handier than the diagram. See Fig. 5. 

10 With these statements the explanation of the diagram Plate 
1 will now be taken up. 

DESCRIPTION AND USE OF THE DIAGRAM PLATE 1 

1 1 Taking the extreme left-hand bottom corner point as the origin 
distances along the bottom line represent the variable tension in the 
tight strand or side of a belt in te'rms of the initial tension, while 
vertical distances measured to any of the bottom group of curves in 
the middle field of the diagram represent the corresponding ten- 
sion in the slack side of the belt, also in terms of the initial tension. 

12 The particular curve against which to read off a certain ten- 
sion depends on the center distance of the pulleys of the belt in con- 
nection with its initial tension per square inch, and is found by con- 
sulting the small diagram directly to the right of this group of curves, 
in the following manner: 

13 Read off the center distance c along the extreme right side of 
this diagram, then follow along the diagonal to the left from this 
reading of c until it intersects the vertical line that extends up from 
the reading of the initial tension t^ on the base of the diagram. 

14 From this point of intersection of c and t^ go horizontally 
to the left to the reading of the corresponding value of the ratio 



TRANSMISSION OF POWER BY LEATHER BELTING 33 

—275 , which leads directly to the proper curve in the bottom group 

*• 

of curves in the middle section of the diagram. 

15 Against this curve there can now be read off any simultaneous 
tensions in the two strands of the belt corresponding to these par- 
ticular values of c and ^o of the belt under consideration. 

16 Having noted this curve, and turning to the extreme right hand 
section of the diagram, the ratio of the effective tensions in the two 
sides of the belt corresponding to the particular coefficient of friction 
4>, and the particular arc of contact a which we wish to count on, 
may be determined. 

17 To this end we read off the arc of contact in degrees on the 
extreme right-hand side of this section of the diagram, follow this 
reading horizontally to the left until it intersects that Une radiating 
from the bottom left corner of this section which is marked with the 
value assumed for 9S at its termination in the extreme top line of the 
section, and then from this point of intersection go vertically up or 
down as the case may be, until we meet the single curve drawn in this 
section of the diagram. From this point in the curve we now go 
horizontally to the extreme left side of this section of the diagram 
and there read off the value of the ratio of the effective tensions, 
which is 

t, 1 

for a belt running so slowly that the centrifugal force has no percept- 
ible influence, and equal to 

L -L 1 



when the centrifugal force reduces the total tensions to the effective 
tensions fj ~ ^c ^^^ h ~ h respectively. 

1 

18 From the point representing we now draw a line to the 

extreme left bottom corner point of the whole diagram. 

19 Any two simultaneous coordinates to this slant line counted 

from its extremity in the base Une of the dl&gri^in, will then also be 

1 
in the ratio 



34 TRANSMISSION OF POWER BY LEATHER BELTING 

20 Passing now to the extreme slanting left side of the diagram, 
we read off the velocity V of the belt, follow this reading horizontally 
to the right until we intersect that radiating line from the extreme left 
bottom corner point of the diagram which is marked with the initial 
tension t^ of the belt per square inch, where it terminates, either in 
the extreme top line of this section of the diagram, or against its 
extreme right side. 

21 From this point of intersection we now go down vertically until 
we reach the long 45-deg. diagonal of the diagram, on which we then 

read off the ratio — , or the loss in effective tension in terms of the 

initial tension, due to the centrifugal force in the belt. 

22 From this point on the long 45-deg. diagonal of the diagram 
we now finally draw a line parallel to the line previously drawn to 

1 
represent the ratio - ^ ^ and extend it to intersect the curve first of all 

e 
determined to represent all possible relations between the two ten- 
sions in the belt. 

23 The coordinates of this point then give the two tensions in the 
belt in terms of its initial tension. By extending the ordinate up to 
intersect that curve in the middle group of curves which is marked 

with the same value of 25 ^^ ^ts terminal in the right side of this 

middle section of the diagram as the curve dealt with in the bottom 
group of curves, we read off the difference of the two tensions; that is, 
the effective pull of the belt, in terms of the initial tension. 

24 By likewise extending the ordinate all the way up to meet that 
curve in the top group of curves which is marked the same as the other 
two curves, we may also read off the sum of the two tensions, in terms 
of the initial tension. 

25 Example. A belt of the pulley center-distance c = 200 in. 
and of 2^ sq. in. cross section, has an initial tension T^, = 175 lb. 
and runs at a velocity V = 2000 ft. per minute. The arc of contact 
of the belt on each pulley may be taken as 180 deg., and the coefficient 
of friction <f) as 0.5. What will be the centrifugal tension T^ = 2.5 
t^ in the belt, what the tension T^ = 2.5 t^ in the tight side, and what 
the tension T^ = 2.5 t^ in the slack side? Also, what will be the effec- 
tive pull P = T^ - T2 and what the sum T^ + T^ of the two tensions? 

26 Solution. The steps have been indicated on the diagram 
by little circles around the points on the several scales of variables 



r— 



FOLDEK No. 1 



TKANSACTIONS THE AMERICAN SOCIETY OF MECHANICAL ENGINEERS VOLUME HI 



TRANSMISSION OK POWER BY LEATHER BELTIXG 




0^<, .1 .2 J .4 05 £ .7 .8 3 1.0 .1 .2 .3 A L5 .6 .7 .8 .9 2.0 .1 .2 .3 A 25 £ .7 .8 .9 30 " ,^^J°J'° c , + 

( TENSION IN TIGHT SIDE or BELT ,n TERMS or INITIAL TENSI0N-8,-^"l INITIAL TENSION PtRSo.l.-t. 

^-LOSS IN TENSION due TO CENTRIFUGAL FORCE ^ ^ 

IN TERMS OF INITIAL TENSION-^ 

PJ'Atk 1 Di.iORAM FOR THu GRAPHICAL Solution of Formhi.ak for Horizontal Belts 



EXPONENT yot 



TRANSMISSION OF POWER BY LEATHER BELTING 35 

that correspond to the values of these variables in the example. 
Thus, in the small diagram at the bottom of the chart, between the 
other two diagrams, circles have been drawn around the points indi- 
cating the unit initial tension 

t, - '-'' = 70 
2.5 

and the center distance 200, giving the resulting value 

c2 200^ , . , , 

1, approximately. 



l2-5 702-5 



"0 

which determines the particular curve in each of the three groups ol 
curves in the middle section of the diagram which apply to the belt 
under consideration. 

27 Also, in the right-hand section of the diagram circles have been 
drawn indicating the coefficient of friction ^ = 0.5, and the arc of con- 
tact 180 deg., giving the resulting value of the ratio between the effec- 

1 
tive tensions -^-^ = 0.208. 

1 

28 From the point ^^ = 0.208 has also been drawn a line to the 

extreme left bottom corner of the whole diagram, the ratio of any 
two coordinates to this line thus being 0.208 also. 

29 Again, in the left-hand section of the diagram circles are drawn 
about points indicating a velocity V = 2000 ft. and the initial ten- 
sion t^ = 70, leading to a resultant value of 

t T 

'« = ^« = 0.202 

to T, 

This means that 0.202 X 175 = 35.35 of the total 175 lb. of initial 
tension in the belt, is made ineffective by the centrifugal force due to 
V = 2000 ft. 

30 From the point — = 0.202 has also been drawn a line parallel 
to the line expressing the ratio 

\ = 0.208 



36 TRANSMISSION OF POWER BY LEATHER BELTING 

SO that the inclination of this line expresses the same ratio between 
the effective belt tensions; or 

r, - r. ^'-^_ 0.208 



31 The point of intersection of this line with the curve previously 
found to express all possible relations between the working tensions 
in the two strands, has also been encircled, and a vertical line has 
been drawn through this point upwards until it intersects that curve 

in the top group of curves which is marked 2.5 = 1> ^'^'^ which 

accordingly expresses all possible values of the sum of the two ten- 
sions in the belt under consideration. 

32 The intersection of this vertical line with that curve in the 

middle group which is likewise marked —275 = 1, and which accord- 

ingly expresses all possible values of the difference of the two tensions, 
has also been encircled. 

33 Taking the readings of the point encircled in the bottom curve, 
we find 

d^ = ^A = 1.81, and d^ = ^-^ = 0.535 



K T, 



We therefore get 



T, = 1.81 X 175 = 316.75 lb. and 
T^ = 0.535 X 175 = 93.63 lb. 

34 From this we again get 

P = T,-T^ = 316.75 - 93.63 = 223.12 lb. 

as the effective pull of the belt. Also, T^ + T^ = 316.75 + 93.63 
■= 410.38 lb. as the sum of the tensions, as against 175 X 2 = 350 
lb., the initial sum. 

35 But usually we would not be interested in the separate values 
of the tensions, and then we would read off directly by the encircled 
point in the middle group of curves, 

P^L = ^l^I^ = 1.276 
f T T 

'ft in i n 



TRANSMISSION OF POWER BY LEATHER BELTING 37 

which gives P = 1.275 X 175 = 223.12 lb., the same answer as 

above. 

:^S6 If also interested in the sum of the tensions, we would read this 

off directly by the point encircled in the top group of curves, 

^.V+ ^2 = ^i-L^2 =,2.345 

which gives Ti + T, = 2.345 X 175= 410. 38 lb., the same answer 
as above. 

37 The solution of problems involving long horizontal belts is thus 
readily enough effected by means of this diagram, but a little considera- 
tion will also make it evident that the difference in results obtained by 
taking the length of a belt into account and by neglecting the same 
is but slight, except for greater lengths and lower initial tensions 
than are usually employed in practice. Ordinarily, therefore, we 
would use merely the very bottom curves in the bottom and top 
groups, and the very top curve in the middle group of curves in the 
middle section of the diagram, which curves are all marked 

^^ = 

/ 2.6 

and thus make no difference between horizontal and vertical belts 
except for exceedingly long belts. 

DESCRIPTION OF DIAGRAMS, FIG. 1, 2, and 3 

38 We will now take up the consideration of diagrams Fig. 1, 2, 
and 3, which form] the basis of the large diagram Plate 2. These 
diagrams are worked out theoretically for vertical belts only, but may 
be applied without hesitation to horizontal belts of the lengths usually 
met with in practice. 

39 In his paper Notes on Belting, Mr. Taylor showed, as already 
mentioned, the economy of running belts under much lower tensions 
than those commonly figured on in proportioning a belt to do a cer- 
tain amount of work. 

40 He also divided the belts with which he dealt into two classes : 
those whose dimensions he could not well increase over what he found 
in use, such, for instance, as the cone belts on lathes and other machines 
provided with a cone pulley in a limited space; and those he could 



38 



TRANSMISSION OP TOWER BY LEATHER BELTING 



20 



300 
90 
80 
70 
60 

250 
40 
30 
20 
10 

200 



QQ 
u- 90 



60 



to 

I 

CO 

to 150 
o 

o 30 



20 



100 



-2L 90 

CD „ 

Q- 80 
70 
60 
50 
40 
30 
20 

























y 


^ 






















^ 


y^ 




















^ 


^ 








k 










[^ 


\^"^ 


>vi . 
\^^ 


"" 










V 






s\J^^JL 


^ 




































































TENSION 


1 1 
IN TIGHT SIDE + 


III. 
iTENSION IN SLACK SIDE=1 


„*tt 


=240 




























— ^ 




























■^ 






_j-f 


















-^ 


^' 




'zpiOfi/ ... 


r 


"^ 


\., 


/ 




f^ 


f:s 


'H 






^ ' ' ' 


-' 




f 






^' 








s^ 










^^-^ 








JV- - 


.-' 






■V 


k 












t-m 






\, 










\ 


V 








^ 






\ 












N 


^ 


/ 


^' 


i^''. 


0,, 




\ 








— - 


.^ 


/ 


/ 








%, 


fe 


\ 


\ 




-^ 




// 


f 


i 








^*U 


^^ 


K 






/ 


y 




h- 


INI 


HALT 


tNS^ 


^ ^ 


\ 


\ 






// 


/ 




/ 


^^-^ 




/ ,' 






\ 


\ 


\ 


/ 


/ 












44 


>^ 






\ 


y, 




/ 














X 


X 




/ 


K, 


Y 








, 












N< 


/ 


\ 


A 








V 








'>^ 


.^^ 


y 


\ 


/ 


V 


\ 






\ 




US^ 


^^% 


r^ 


^ 




/ 


\ 


\ 


\ 






X 


^ "ft 








» 


/ 




\, 


V\ 












ktc 




A 


y 






\ 


s\ 


\ 












,v^,'^ 




^ 








\ 


\\ 






--- 




-<^^>" 


r 






^^ 






\5^ 






:i: 


c<^T 


y^ 






/ 


y 












^ 


-^ 


\ 


\ 




_^ 


^^ 


















^^ 


\ 



1000 



2000 



.„„„ 3000 4000 5000 6000 

VELOCITY OF BELT iN FEET PER M1NUTE=V 

Fig. 1 Diagram showing the Relations of Pulung Power to Tensions 
AT All Speeds, for a Machine Belt of 1 sq. in. Section. (See opposite page 
AND Par. 9/). 



TRANSMISSION OF POWER BY LEATHER BELTING 



39 



rl60 




1000 2000 3000 4000 5000 5557 

VELOCITY OF BELT in FEET PER MINUTE=V 
Fig. 2 Diagram Showing the Relations of Pulung Power to Tensions 
AT All Speeds for a Countershaft Belt of 1 sq. in. Section 

140 
Fig. 1 is plotted for: Arc of contact a=180 deg., coefficient of friction (J) = 0.54— . - , - 

(See Par. 51); and the sum of the tension in the tight side of the belt and one-half the tension 
in the slack side, A = 240 lb. for all velocities. (See Par. 44.) The dotted curve marked tm 
gives the initial tension which for p, the same as that figured for A = 240 lb., corresponds to 
A = 320 lb. This tm, the average value of which is about 185 lb. between practical limits of 
velocities, is the maximum initial tension, to which a machine belt is retightened whenever the 
tension falls to the minimum initial tension <„. 

Fio. 2 is plotted for the same data as Fig. 1, except that here A = 160 lb. for the full-diawn 
curves. (See Par. 44.) The dotted curve marked tm gives the initial tension which, fo'- p the 
same as that figured for A = 160 lb., corresponds to A = 240 lb. This tm, the average value of 
which ia about 142 lb. between practical limits of velocities, is the maximum initial tension to 
which a countershaft belt Ls retightened whenever the tension falls to the minimum initial ten- 
sion <„. 



40 TRANSMISSION OF POWER BY LEATHER BELTING 

readily increase by providing larger pulleys, such as the belts leading 
from line-shafts to the counter-shafts of machines. 

41 For belts in the first class he adopted higher tensions than for 
those in the second class. He also devised a set of belt-clamps pro- 
vided with spring balances, by means of which he could make sure 
that a belt was put up under a specified initial tension, and could also 
ascertain its fall in tension at any time desired. 

42 All this seems to be the first effort made by an engineer to pay 
any systematic attention to the belting in a shop, which even today 
is usually left entirely to the rule-of-thumb method of the machinist or 
millwright. 

43 The reason why Mr. Taylor had adopted, and accordingly 
recommended, lower belt-tensions than usually counted on in propor- 
tioning a belt to do a certain amount of work, was that a belt quickly 
loses its tension if it exceeds a certain amount, and thus in order to 
maintain such a tension, approximately, requires frequent retighten- 
ing, which is a source of too much expense and leads to a rapid destruc- 
tion of the belt. See Fig. 4 and Par. 8. 

44 Taking Mr. Taylor's data as a starting point, the writer has 
finally adopted the rule, as a basis for his use of belts on belt-driven 
machines, that for the driving belt of a machine the minimum ini- 
tial tension must be such that when the belt is doing the maximum 
amount of work intended, the sum of the tension on the tight side of the 
belt and one-half the tension on the slack side will equal 240 lb. per 
square inch of cross-section for all belt speeds; and that for a belt driving 
a countershaft, or any other belt inconvenient to get at for retighten- 
ing or more readily made of liberal dimensions, this sum will equal 
160 lb. 

45 Further, the maximum initial tension, that is, the initial ten- 
sion under which a belt is to be put up in the first place, and to which 
it is to be retightened as often as it drops to the minimum, must be 
such that the sum defined above is 320 lb. for a machine belt, and S40 
lb. for a counter-shaft belt or a belt similarly circumstanced. 

46 The reason for adopting a uniform sum of the tension in 
the tight side and one-half the tension in the slackside, as mentioned 
above, instead of either a uniform initial tension, or a uniform maxi- 
mum tension alone, is that the aim has been to get as uniform periods 
as possible for the retightening of belts at all speeds. 

47 But evidently, while the maximum tension in a belt must be 
the greatest factor in determining the rapidity with which the belt 
will lose its tension as a whole, the accompanying tension in the slack 



TRANSMISSION OF POWER BY LEATHER BELTING 41 

strand or side must also have some influence, though not proportion- 
ally to the same extent; and hence, the idea occurred to the writer of 
taking it into account in the manner and to the extent stated. 

48 On the diagram Fig. 1, various formulae have been plotted 
for 240 lb. as the constant sum at all speeds of the tight tension and 
one-half the slack tension per square inch cross section of belt; for a 
coefficient of friction that varies with the velocity according to For- 
mula 13 in the Appendix; and an arc of contact of 180 deg. The rela- 
tions of the various tensions in the belt for all speeds may there be 
studied to great advantage. It will thus be seen that the centrifugal 
tension completely balances the initial tension at a belt speed of 
6805 ft. per minute. 

49 On the diagram in Fig. 2 some of these formulae have likewise 
been plotted, with the lesser value of 160 lb. as the constant sum of 
the tight tension and one-half the slack tension, but for the same 
values of the coefficient of friction and the arc of contact. Here the 
centrifugal tension balances the initial tension at the speed of 5557 
ft. per minute. 

50 The diagram, Fig. 1, represents the writer's practice in connec- 
tion with machine belts; that in Fig. 2 his practice in connection 
with counter-shaft belts (see Par. 44). Both diagrams were used 
as the basis for the construction of Plate 2, and for the slide rule illus- 
trated in Fig. 5. 

51 In the diagram Fig. 3 are given the horse power outputs per 
square inch of section of belts running under the conditions imposed 
in the diagrams Fig. 1 and 2. 

52 It will be seen that the maximum output is 13.8 h.p. per square 
inch of a belt under the conditions imposed in Fig. 1, and that this 
is for a speed of about 4000 ft. (more exactly 4032 ft.); and that for 
the conditions imposed in Fig. 2, the maximum horse power is 7.46 per 
square inch, and that this is for a speed of about 3250 ft. (more exactly 
3247 ft.). 

DESCRIPTION OF DIAGRAM PLATE 2 

53 In the diagram Plate 2, the data given on the diagrams 
Fig. 1, 2, and 3, for a belt of one square inch of section, and an arc 
of contact of 180 deg., have been so modified that almost any problem 
relating to belting of any size and any arc of contact can be solved. 

54 This will best be illustrated by the following two examples: 

55 Example 1. The maximum cone step on the counter-shaft of 



42 



TRANSMISSION OF POWER BY LEATHER BELTING 



a lathe is 22 in. in diameter and wide enough to carry a 3 in. double 
belt. The speed of the shaft is to be 300 r.p.m. Assuming the thick- 
ness of a 3 in. double belt to be fV in., and the arc of contact of the 
belt to be 170 deg.: (a) What pull can the belt be] counted on to 
exert, and what horse power will it transmit with ^this pull? (6) 
Under what initial tension will the belt first be put up, and retightened 



MACHINE 

BELT 
t+it|=240 




COUNTER 
SHAFT BELT 
t+5tfl60 



1000 2000 3000 4000 

VELOCITY OF BELT in FEET PER MINUTE=V 

Fig. 3 Horsepower Output Corresponding to Belt Pulls in Fig. 1 and 2 



from time to time? (c) And what minimum initial tension must it 
not be allowed to fall below to insure the above-determined pull 
without undue slip? 

56 Solution. To get the answer to question (a), we first turn to 
the small bottom portion of the diagram Plate 2, and on its right hand 
side note the point reading 300 r.p.m. From this we pass horizon- 
tally to the left until we intersect the vertical line from the point 



TRANSMISSION OP POWER BY LEATHER BELTING 



43 



reading 22 in. on the scale of pulley diameters at the bottom line of 
the diagram. From the point of intersection we follow the diagonal 



KO 

ISO 

140 
130 


\ 


i 




120 


\\ 


■•Vs 






110 


1 '\ 


\ 






1 




>> 




1 




\ 


■^ 


in 


100 






^^ \ 














<-*J 




A!'. 




'';— . 






\ 




'■ 


^ 90 
























■ ^ — 




^ 


. 1 \ 
































T 




rm 


-1-II 




















5 80 

Q. 






















































































g 70 






















































































o 
z 






















































































2 50 
to 






















































































LU 

•-40 






















































































30 






















































































20 






















































































10 



































































































































Z 9 16 23 30 

NOVEMBER 

1900 



7 14 21 
DECEMBER 



II 18 
JANUARY 



8 15 22 
FEBRUARY 



I 8 15 32 

MARCH 



1901 



Fig. 4 Experiments Made on the Fall in Tension in Two Belts 5f in. Wide 

BY ^i in.Thick Driving a Large Rotary Planer at the 

Works of the Bethlehem Steel Company 

(See Par, 8 and Par. 4.3) 

The peculiarly high tensions measured on four days, during the latter part of 
February 1901, were probably due to something sticking in the belt scales used. 



line upwards to the bottom line of the main portion of the diagram - 
and there read the velocity of the belt to be about 1700 ft. per min. 



44 



TRANSMISSION OF POWER BY LEATHER BELTING 




H 
W 
H 

a 
o 

o 

02 



O pM 

B 

o 



> 

o 
a 

a 
« 
o 

w 



o 

02 



Wur.i:i. .1 \.i\i: v; ■ .^ >"/• 



FOLDER No. 2 



.Si 



TRANSACTIONS THE AMERICAN SOCIETY OK MECHANICAL ENGINEERS VOLUME ;il 

TRANSMISSION 01' I'OVVKR fiY LEATHER BELTING 



VELOCITY foR INITIAL 
TENSION IN BELT 

MAXIMUM. MINIMUM. 




DIAMETER OF PULLEV 
Plate 2 Geneual Belting Diagram iNconponATiNQ the Author's Pkactice 



TRANSMISSION OF POWER BY LEATHER BELTING 45 

57 We now note the point that corresponds to this belt speed of 
1700 ft. per min. in that scale on this same bottom line of the main 
diagram which is marked " Velocity for Pull of Machine Belt" and 
interpolate a vertical line extending upwards from this point. Then 
leaving this for the time being, we turn to the extreme left hand 
portion of the diagram and there note the point corresponding 
to the belt thickness ^ in. on the vertical scale to the extreme left, 
and also the point marked 3 in, on the scale of belt widths at the 
bottom of the diagram. Following these two points, respectively, 
horizontally to the right and vertically upwards, until they intersect, 
we now follow the diagonal from this point of intersection until it 
terminates against the vertical line marked 170 deg. at the top of the 
diagram, in the field marked "Arc of Contact," and then continue 
horizontally until we intersect the interpolated vertical line for the 
belt speed 1700 ft. already noted. 

58 From the point of intersection we follow the diagonal until we 
meet the vertical scale of pounds, on which we now read the belt pull 
to be 140 lb.; and continuing horizontally until we meet the vertical 
line extending upwards from the point corresponding to the belt speed 
originally found on the scale of belt speeds in this section of the dia- 
gram, and from this line diagonally to the vertical scale of horse power, 
we read ofT the horse power transmitted to be 7.2. All of these move- 
ments are indicated on the diagram by little circles around the vari- 
ous points of intersection. 

59 To get the answer to question (6), we proceed exactly as before, 
with the width and thickness of the belt, except that we follow the 
diagonal across the portion of the diagram headed " Arc of Contact" 
until we meet the border line for 180 deg. From here on we proceed 
horizontally until we meet the vertical line that corresponds to the 
belt speed 1700 ft. in the field marked "Velocity for Maximum 
Initial Tension for Machine Belt." From the point of intersection on 
this vertical line we then pass diagonally to the scale of pounds, and 
there read the maximum initial tension to be 168 lb. Those move- 
ments for this solution on the diagram that differ from those for the 
answer to question (a), are indicated by little filled-in circles around 
the various points of intersection noted. 

60 For the answer to question (c), we proceed in every respect as 
we did for question (b), except that we of course start from the 
point corresponding to the belt speed 1700 ft. in that field of the scale 
on the top line of the diagram which is marked " Velocity for Mini- 
mum Initial Tension for Machine Belt." The answer read off on 



46 TRANSMISSION OF POWER BY LEATHER BELTING 

the vertical scale of pounds is 113 lb. The movements for this 
solution on the diagram that differ from those for the answer to ques- 
tion (6), are indicated by little dotted circles around the points of 
intersection. 

61 Example 2. The counter-shaft in Example 1 is to be driven 
by a belt to run at a speed of 3000 ft. per min. (a) What diameter 
of pulley is required to give this belt speed? (6) What pull must the 
belt transmit? (c) What width of double belt must be used? {d) 
And what will be the initial tension under which the belt must be put 
up, and to which it must be again retightened after falling to the 
minimum? (e) What will be its minimum tension? 

62 Solution, (a) As the counter-shaft is to make 300 r.p.m. and 
the belt is to run at 3000 ft. per min., we turn to the small diagram 
at the right hand bottom corner of the main diagram, proceed hori- 
zontally to the left from the point marked 300 on the scale of revolu- 
tions on the right, until we meet the diagonal line from the point 
marked 3000 on the horizontal scale of velocities. From the point 
of intersection we then go vertically down to the scale of pulley 
diameters, and there read off 38 in. as the nearest even diameter. 

63 (6) To get the pull of the belt we remember that the cone 
belt was found in Example 1 to transmit 7.2 h.p. We therefore note 
that point on the vertical scale of horse powers at the extreme right 
of the main diagram which corresponds to 7.2, and then follow the diag- 
onal from this point towards the left, until we meet the vertical 
line extending up from the point marked 3000 on the scale of velocities 
on the bottom line of this portion of the diagram. From this point 
of intersection we continue horizontally to the left to the vertical 
scale of pounds, on which we then read off the pull 80 lb. 

64 (c) From the point corresponding to this 80 lb. we now 
continue diagonally to the left until we meet the vertical line extend- 
ing up from the point corresponding to the belt speed 3000 on the 
scale marked '' Velocity for Pull of Counter-Shaft Belt" at the bottom 
of this central portion of the main diagram. From this point we 
continue horizontally to the vertical line corresponding to the arc of 
contact, which, not being given, we will assume as 160 deg., and then 
again diagonally in the extreme left hand section of the diagram. 
Any simultaneous readings of width and thickness from points in the 
diagonal along which we are now moving, will then give a proper belt, 
and assuming as in Example 1 a thickness of ^ in., we find the width 
to be 3^ in. 

65 {d) To find maximum initial tension for this belt, we proceed 



TRANSMISSION OF POWER BY LEATHER BELTING 



47 




Ph S 



48 TRANSMISSION OF POWER BY LEATHER BELTING 

exactly as in Example 1, except that we use the scale marked " Veloc- 
ity for Maximum Initial Tension for Counter-Shaft Belt" at the top 
of the middle section of the diagram, and then read this off on the 
scale of pounds as 157 lb. 

66 (e) Similarly, we find the minimum initial tension to be 97.5 lb. 

MEANS OF SECURING AND MAINTAINING DEFINITE TENSIONS IN BELTS 

67 In his paper. Notes on Belting, Mr. Taylor referred to belt- 
clamps provided with spring balances for weighing the tension in 
a belt. In the case of endless belts these scales are put directly on a 
belt in its final position over its pulleys, while in the case of a belt 
with wire lacing, this is cut to length under the required tension on 
the specially designed belt bench illustrated in Fig. 6. As will be seen, 
this bench is provided with a pair of pulleys which can be so adjusted 
that a tape-line will measure the same around these pulleys as over 
the pulleys on which the belt is to run. A belt cut and laced to give 
a certain tension when the bench pulleys have been properly adjusted, 
will then be of a length to assume the same tension over its own 
pulleys. 

68 This indirect way of securing a desired tension in a belt was 
first suggested by our fellow member, Mr. Gullow Gulowsen, who also 
made the drawings from which the first bench and the first improved 
belt scale were made by the Bethlehem Steel Company in the year 
1900. 



APPENDIX 

ELASTIC PROPERTIES OF BELTING 

The only experiment recorded by Mr. Lewis to establish the elastic properties 
of leath. r belting is the following: 

2 " A piece of leather belting 1 sq. in. in section and 92 in. long, was found by 
experiment to elongate \ in. when the load was increased from 100 to 150 lb., and 
only ^ in. when the load was increased from 450 to 500 lb. The total elongation 
from 50 to 500 lb. was 1^ in., but this would vary with the time of suspension, 
and the measurements here given were taken as soon as possible after applying 
the loads. " 



i 






















•^ 






















H^^ — 


















_„,*--' 


" 




94" 
















' — ' 


















... — 




















^..--■^ 




















_^^^ 












i 








-^ 


^-'^^ 












MS 








y 






















^^ 




iti9?») 
















y 


'-+♦ 










93° 




y 




















y 




















1 


^ 




















/ 






















i 


/ 






















/ 
























/ 














































no- 

























100 



200 300 400 

TENSION IN POUNDS per SQ.lN.= t 



500 



Fig. 1 Plot of Experiments on a Piece of Bei/ting 1 sq. in. in Section 
AND 92 in. long. Test Made by Wilfred Lewis at the Works of 
Wm. Sellers & Co., Philadelphia, in 1885 

3 Theae data have been plotted in Fig. 1, in which they are remarkably well 
represented by the formula 

Lt = 92 ( 1 + ^^ 



830 

in which L^ is the elongated length of the belt, and t the load or tension per square 
inch. 

The full development of the mathematical formulae of this paper, with some other related 
mattpr, is given in an unpublished supplement to the Appendix, which is on file in tke 
Library of the Society for the use of any who wish to verify the mathematical work. 



50 



TRANSMISSION OF POWER BY LEATHER BELTING 



4 On the strength of this formula the writer originally established the theorem 
that the sum of the square roots of the tensions in a belt is constant for all loads, when 
no attention is paid to the weight of the belt. 

5 However, he soon realized that it would not be safe to build a theory on a 
single experiment of this nature; and hence, in July 1901, while in the employ of 
the Bethlehem Steel Company, he undertook a series of similar experiments, and 
obtained permission of William Sellers & Company to use their emery testing 
machine for that purpose with the assistance of their shop engineer, Mr. Leonard 
Backstrom. 

G Nine pieces of belting were tested in all. The results upon one of those 
pieces are shown in Fig. 2, which is tj'pical of all of them. Similar diagrams 
representing the other tests are filed with the unpublished supplement to the 
Appendix. In all cases the tests were made as rapidly as the loads could be 
adjusted and the extensometer readings taken.l 




Fig. 2 



100 200 

TENSION IN POUNDS per SQ.lN.=t 

Plot of Experiments on a Piece of Slightly used Double Belting 
3rG IN. Wide by | in. Thick 



Test made by the writer at the works of Wm. Sellers & Co., Philadelphia, in 1901. 
The Supplement contains eight similar plots. 



7 Each piece was several times subjected to a complete cycle of loads between 
two extremes. During the first few cycles the belts invariably showed different 
results, but always gave, eventually, practically the same readings for a number of 
cycles in succession, and these are the readings plotted in the figures. 

8 The small filled-in circles represent the readings for an increasing load and 
the small open circles those obtained for a decreasing load. It is rather astonish- 
ing how much lag is shown by every belt. Unquestionably this has some influ- 
ence on the law of change of tension in a belt, from its minimum to its maximum, 
along its contact with a pulley. This matter has been given some consideration 
in the Suppl ;ment. 

9 On account of this lag, apparently it would have been desirable to subject 
some of these belts to a series of smaller cycles, each between adjacent limits of 
the load. The best the writer could do with the results obtained was to average 



TRANSMISSION OF POWER BY LEATHER BELTING 



51 



the loop of each cycle by means of a parabolic curve, and thus obtain a value for 
the constant E for each belt on the supposition that the formula 



Lt = L 1 + 



E 



[11 



ia approximat<>ly correct. In the various formulae given, however, L is not the 
original 15 in. of length to which the extensometer was originally adjusted for each 
belt, but an ideal length only, for the estimation of the relations between the ten- 
sion and the stretch for values never approaching close to zero. 

10 But the best experiments for ascertaining the relations between the tension 
and the stretch in belts are unquestionably those by Prof. W. W. Bird, published 
in his paper on Belt Creep, read at the Scranton Meeting in 1905. 











1 ! y- 


-jjfj- 


A^ 










1 , ■ J ' j 


\r~ — h — 


1 Pi ' 


r^ i I ' 








■fill 


, /'' ■i-^^\ 




\ i ' 


Lt=29A.3? 


1^^) 




1 


1 \^f' 






fs.n .■ ' f*r/: 


^w 


1 I t 








w/ 1 






/ 




/ 














\J\ ! 










, 


» 1 ' > 








Lt=^9j:52(K^) 


r- . ■ , 






1 ! ! 




Ml! 








y 1 1 








r ' ' 


/ 











295' 



100 200 300 

TENSION IN POUNDS per SQ.lN.=t 

Fig. 3 Plot of Experiments on a Single 4-in., Endless, Running Belt 

Test made by Prof. Wm. W. Bird at the Worcester Polytechnic Institute. 
The Supplement contains a similar plot on a single 6 in. laced belt. See Professor 
Bird's paper on Belt Creep in Volume 26 of the Transactions. 

11 These have been replotted by the writer after making some slight correc- 
tions in the lengths given by Professor Bird, allowing for the influence of the 
sag in the belts and in the tensions given, by the addition of the estimated cen- 
trifugal tension, which was not measured by Professor Bird. The centrifugal 
tension was estimated after obtaining from Professor Bird the information that 
the belts were run at a speed of about 1000 ft. per minute. One of the diagrams 
is shown in Fig. 3 and another has been filed with the Supplement. 

12 The plots made by the writer differ further from those in the original paper 
in that he laid off the tension in pounds per square inch of section of the belts. 

13 It will be noted that the results have been approximated both by a dotted 
line representing a special form of the broadly general formula 



Lt = L 14- 



[2] 



52 TRANSMISSION OF POWER BY LEATHER BELTING 

and by a full line representing this same formula with the special value } for n; 
or in other words, Formula 1. For regularity of results, these experiments are 
remarkable. 

14 While the dotted curves with their more complicated formulae represent the 
experiments more closely, the full curves with their simpler formulae also cover 
the results so well that they may be looked upon as an excellent justification for 
the assumption previously made by the writer on the strength of the experiments 
made by Lewis and himself, namely, that within the limits of ordinary working 
tensions of a belt, the difference between the lengths of a belt at different tensions is 
proportional to the difference between the square roots of those tensions. 

15 This proportion is implied in the general Formula 1, when by L we imply, not 
necessarily the slack length of a belt, but an ideal slack length on the basis of 
which the formula gives reliable results between ordinary working limits of t. 

16 Taking the average of the values of E in all twelve sets of experiments we 
get 895. Leaving out two experiments, one with a value of E exceeding 1000 and 
another for which E was less than 800, and taking the ten remaining experiments 
with values of E between 800 and 1000, we get 890; while if we take the average 
of only the two experiments by Professor Bird we get only 861. 

17 As will be seen hereafter, the writer has adopted 864 as an average working 
value, because this figure, combined with certain other constants, results in the 
simple final constant coefficient 0.04'^in the right member of Equation 5. For 
an average practical working formula on which to build an improved theory for 
the transmission of power by leather belting, we thus have 

in which Lt equals the length of a belt under the unit tension t when its slack 
length is L. 

18 However, it will appear later on, that the value 864 adopted for E has 
significance only in the formulae developed for long horizontal belts, as £/ disappears 
in these formulae when the weight of the belt is neglected. 



LAW OF VARIATION IN THE TWO TENSIONS OF A LONG HORIZONTAL 

BELT 

19 In developing an expression to represent the law of variation in the two 
tensions of a long horizontal belt, the free strands of the belt only are considered, 
and then, for the sake of argument, are assumed to be attached to the ends of 
two double levers fulcrumed in the middle, as shown in Fig. 4 and 5. 

20 That the parts of the belt in contact with the two pulleys remain at prac- 
tically constant length independent of any variation in the tensions of the two 
strands, and thus have no material influence on this variation, will be shown in the 
Supplement. 

21 In Fig. 4 the levers are parallel to each other, and the two strands of belt- 
ing whose normal slack lengths I are supposed to be equal, must form equal cate- 
naries under equal unit tensions <,,, corresponding to the equal initial tensions in 
the strands of a belt continuous over its two pulleys. 



TRANSMISSION OF POWER BY LEATHER BELTING 



53 



22 In Fig. 5 the levers have been moved through equal angles in opposite 
directions, thus tightening the bottom strand to the unit tension <,, and slackening 
the top strand to the unit tension L, with corresponding changes in the respec- 
tive catenaries. 

23 Under this arrangement the sum of the chords of the catenaries remains 
constant under all variations of the tensions, a condition that corresponds to that 
of an actual working belt over its two pulleys, which remain at a constant distance 
apart under all conditions of tension in the belt 

24 In considering the problem, the customary approximations in dealing with 
catenaries were made, in connection with Formula 1 for the elastic properties of 
leather belting, and then the following general formula developed for the relations 
between the unit working tensions <, and t^ in a belt, as dependent on the initial 
unit tension t^, the weight W per square inch of cross section of each free strand, 
and the elasticity constant E in Formula 1. 



W^E I 1 1 




[4] 



FIG. 5 
Fig. 4 and 5 Representation of Change in Tensions and Sags in Strands 
OF A Horizontal Belt 

25 The average weight of a cubic inch of leather belting being about -^-^ lb. 
I 
we may write W = -, in which I is the center distance in inches between the two 

pulleys of a horizontal drive; and by also assuming 864 as the average value of E, 
as already done in Formula 3, we may write more specifically 



Vf. + %/<,-= 2 V<„ + 0.04? ( i + ^^^ - ^^^ 



[5] 



54 TRANSMISSION OF POWER BY LEATHER BELTING 

26 P'or very short belts the last terms in Equations 4 and 5, which are only 
tentatively solvable equations, become very small as compared with the term 
^\/tg, and by neglecting it entirely we write for vertical belts as well as for short 
horizontal belts, or even approximately for all belts: 

V/; + Vi, = 2 VFo [61 

This formula may be enunciated as a new theorem of the relations of the ten- 
sion in a belt, thus: Under any variation of the effective full of a belt, the sum 
of the square roots of the tensions in the two strands remain constant, as against 
the old fallacious supposition that the sum of these tensions remains constant. 

27 However, without entirely neglecting the last term in Formulae 4 and 5 
above, these may be made solvable with respect to <,, by first approximating the 
value of this term by the substitution in it alone of an average relation of the 
tensions t, and ij- As will appear from a study of the diagram Plate 1 in the 
body of the paper, such an average relation of the tensions is 

t^ 

t = - 

h 

28 Substituting this in the last term of Equation 5 and then solving this with 
respect to f, we get 



I, = 



2v/?,-V<; + 0.04P(;^ + i-| 



m 



which gives practically identical results with the original Equation 5 for such val- 
ues of <o and I as fall within ordinary practice. 

29 If we express the tensions ti and t^ in terms of the initial tension t„ by writing 

^ , '2 

5, =• — and 5, = — , Formula 7 reduces to 

to ' t, 

2 - V^, + 0.04 ^ 2- / 5^2 ^ ^^ _ 2 j • [8] 

in which form it is under certain circumstances more readily applied. 

TESTING FORMULA 8 BY THE RESULTS OF LEWIS' EXPERIMENTS 
ON HORIZONTAL BELTS 

30 In Fig. 6 the two tensions simultaneously obtained by Mr. Lewis in one 
series of his experiments have been plotted in terms of the initial tensions of the 
belt, the tensions in the tight side of the belt being laid off horizontally and the 
tensions in the slack side vertically, in the same manner as is done on the diagram 
Plate 1, in the body of the paper. Similar diagrams, representing five additional 
series of experiments made by Lewis, are filed with the Supplement to the Appen- 
dix. 

31 However, as the apparatus used by Mr. Lewis did not measure the cen- 
trifugal tension in his belts, and as he had no occasion to calculate values of thig 



TRANSMISSION OF I'OWER BY LEATHER BELTING 



55 



quantity, these have been calculated for the present purpose, and added to the 
effective tensions tabulated by Mr. Lewis. 

32 Each experiment is represented by one of the small filled-in circles, and 
is numbered the same as in the tables from which the experiments were taken from 
Mr. Lewis' paper. 

33 Unfortunately, but very naturally, the initial tension did not remain con- 
stant throughout a set of experiments, and in plotting the tensions it was there- 
fore necessary to estimate for each individual experiment where the initial tension 
was between the values measured at the beginning and at the end of each set of 
experiments. This was done by assuming that the initial tension measured at the 
beginning of a set of experiments held good for the first experiment, and that the 
initial tension measured at the end of a set of experiments held good for the last 
experiment, and that there was an equal drop for each experiment. 

34 That the initial tension was not constant during each set of experiments 
is the reason why the actual tensions obtained were not plotted, but instead their 
ratios 5, and d^ to their respective initial tensions. 

35 In each figure. Equation 8 is also given with the center distance of the 
pulleys for each particular belt introduced as an approximate value of /. 




3 A H S .1 .S 3 20 i .2 :i A 2i £ J J3 S 3jO J .2 J /» 35 4 

tension in tight side of belt m terms or initial tension=§=|^=|^ 

Fig. 6 Plot op Experiments by Wilfred Lewis 
Horizontal Double Belt 2i in. wide, -j^ in. thick and 32 ft. long. 20 in. Pulleys. 
Average Value of Initial Tension t„ = 70 lb. per sq. in. (The Sup- 
plement contains five similar plots.) 



36 By the introduction of the value obtained for ^g in each experiment plotted, 
a corresponding value was calculated for 5, by the formula mentioned, and this 
value also plotted, and then a curve was drawn to cover the points thus calcu- 
lated. The points themselves are indicated by the little circles drawn around 
them. 

37 The close coincidence between the curves representing Formula 8 and the 
experimental results, is certainly all that can be desired in the way of an experi- 
mental verification of the soundness of Formulae 7 and 8. 

38 The other, lower curve drawn on each diagram represents the relation be- 
tween the tensions in a belt when the influence of its weight is neglected, as given 
by Equation 6, which is also given on each diagram, in the form 

§0.5 4. 5^06 =, 2 [9] 



56 TRANSMISSION OF POWER BY LEATHER BELTING 

BELT CREEP AND ITS INFLUENCE ON THE COEFFICIENT OF FRICTION 
BETWEEN A BELT AND ITS PULLEY 

39 In the paper on his experiments, Mr. Lewis drew the conclusion that the 
friction between a belt and its pulley varies greatly with its velocity of slip, so 
that the greater the slip the greater the friction. But as he did not make a sub- 
stantial study of the elastic properties of leather, upon which the phenomenon 
of belt creep depends, he had no means of distinguishing in his experiments be- 
tween the necessary slip due to the creep of the belt, and the amount that was 
slip pure and simple of the belt as a whole. 

40 In most of his experiments the belt did an amount of work that called for 
much greater friction between the belt and its pulley than that corresponding 
to the creep of the belt alone, and this resulted in additional or true slip that 
produced the friction needed to make the belt exert the pull called for. 

41 By means of Formula 3 it was possible also to derive a formula that gives 
a good idea of the actual creep of a belt in terms of the tensions in its two strands, 
which was the object of Professor Bird's paper on Belt Creep, from which the 
experiments plotted in Fig. 3 of his paper were taken. 

42 This formula, the mathematical development of which is given in the Sup- 
plement, is 

2 864 + \/t 

in which 

V = actual average velocity of the creep of the belt on each of its two 
pulleys. 

F, = velocity of the tight strand of the belt, which is the same as the cir- 
cumferential velocity of the driving pulley, 

y, = velocity of the slack strand of the belt, which is the same as the cir- 
cumferential velocity of the driven pulley. 

43 The total creep of the belt on both pulleys together expressed in per cent 
of F, is then 

^64 + Vt^ 

44 It must be borne in mind, however, that Formulae 10 and 11 take account 
of creep only, and have nothing to do with any additional slip due to a sliding of 
the belt as a whole over its pulleys, though the expression 

t, = ^ (F, - F,) 

taken by itself always represents the total sum of the average creep of the belt 
and the additional sliding of the belt as a whole, over each of its pulleys, when 
such additional sliding does take place. 

45 Considering the matter in this light Mr. Lewis calculated and tabulated 
the velocity of shding from the observed loss in speed between the pulleys in his 
various experiments. 

46 In Tables 1 and 2 appear some of the data thus tabulated by Mr. Lewis. 
However, instead of tabulating merely the effective tensions measured by him, 
the centrifugal tensions have here been figured and allowed for, and then the total 



TRANSMISSION OF POWER BY LEATHER BELTING 57 

tensions thereby obtained subsequently converted into tensions per square inch 
of cross-section. A column has also been added giving the percentage of aver- 
age slip due to belt creep alone, as figured by Formula 11. 

47 It will be seen that in most of the experiments the velocity of sliding greatly 
exceeded the average due to the elastic creep alone, and that thus the belt as a 
whole slid over the pulleys in addition to the elastic creeping, thus showing that 
the friction corresponding to this creep alone was not enough to produce the pull 
the belt was called on to perform. 

4S The relation between the total average velocity of sliding of the belt on 
each pulley, and the corresponding coefficient of friction calculated by Mr. Lewis 
by the formula 

Ratio of Effective Tensions = c^*^ 
and also copied in Tables 1 and 2 of this paper, is plotted in the diagram Fig. 7 
and in a similar diagram of the Supplement. On these diagrams is also shown 
a curve representing the equation 

[12] 

in which 4> is the coefficient of friction and v the total average sliding velocity of 
the belt in feet per minute. These results were obtained from belts that had been 
in active service, and tested without the application of any belt dressing. 

49 As a somewhat conservative average the curve is seen to cover the results 
obtained with the belts in a normal condition in a highly satisfactory manner. 

50 The question now arises, What coefficient of friction ought to be assumed 
in calculating the pulUng power of a belt at any given speed? In view of the fore- 
going it does not seem right to assume an average coefficient for all belt speeds 
Nor would it be right to base it on an average total sliding velocity of a belt cor- 
responding to a fixed percentage of the belt speed, for even a very low percentage 
would mean a very high sliding velocity in the case of a high-speed belt, while 
a high percentage would mean only a moderate sliding velocity in the case of a 
slow-speed belt, and it would seem that the speed with which a belt slides over 
its pulley would principally determine the life of a belt that meets with no accident. 

51 After considerable study over the subject, the writer has assumed a vari- 
able coefficient of friction expressed by the empirical formula 

140 [13] 

^ = 0.54 - 

^ 500 -H F 

in which V is the velocity of the belt in feet per minute 

52 Equating Formulae 12 and 13 we get 

_ 160-1- 0.88 7 
*~ 85-1- 0.03 7 

as the velocity of sliding on each pulley in terms of the velocity of the belt itself 

53 As the percentage of slip between the circumferential speeds of the two 
pulleys of a belt is twice the percentage of the average total velocity of sliding t; 
of the belt over each pulley, we may now write 

200 V 200 160-!- 0.88 7 

X = = . [14] 

7 7 85 + 0.03 7 



58 



TRANSMISSION OF POWER BY LEATHER BELTING 




a, = NOIlDlbJ JOiN3l3UJ303 



TRANSMISSION OF POWER BY LEATHER BELTING 59 

as an expression for the percentage of slip corresponding to the coefficient of 
friction expressed by Formula 1.']. 

54 In Table 3 are listed simultaneous values of (/> and x as expressed by For- 
mulao 13 and 14 and also, for the sake of comparison, for ^ figured by the formula 

400 
^ 800 + F 

V 
which is the value of 6 by Formula 12, for r = — , that is. for a uniform slip of 
V 0' '200 

one per cent at all belt speeds. 

55 A study of Fig. 7 and the similar figures in the Supplement will show 
that the variation in the coefficient of friction with the initial tension of the belt 
is so conflicting as to make it best to leave this out of consideration entirely, and 
so adhere to the customary assumption that the coefficient of friction is inde- 
pendent of the intensity of the pressure. Therefore, as our whole theory of the 
variation of the coefficient with the belt speed rests entirely on the formula 

Ratio of Effective Tensions = «"^" 

t his will be used unhesitatingly, in spite of the fact that its absolute validity will 
be disputed in the Supplement for a number of reasons. 

EFFECT OF CENTRIFUGAL FORCE IN A BELT 

56 While the effect of the centrifugal force in a fast-running belt seems to have 
been fully understood by all who have previously treated the subject before this 
Society, it is still but imperfectly understood by many engineers and mechanics; 
and has even been treated in a wrong way by certain text-book writers who have 
followed the work on belting by the late Professor Ruleaux. It has therefore 
seemed desirable, in the Supplement to the Appendix, to go into details on this 
part of the subject. 

57 Subsequently, the following general formula was developed for the loss in 
effective tension in a belt, due to its centrifugal force: 

w 

ic = V^ 

300 g 

in which 

Ic = loss in effective tension per square inch of cross-section of belt. 

V = velocity of belt in feet per minute. 

w = weight of one cubic inch of belting in pounds. 

g = acceleration of gravity in feet per second. 
Substituting w = -Jjj, as in Par. 25, and g = 32-J^, we have more specifically 

^c = 0.000003454 V^ [15] 

which is substantially the same formula as that given in Mr. Nagle's paper, For- 
mula for the Horse-Power of Leather Belts, read at the Hartford meeting in 1881 . 



60 



TRANSMISSION OF POWER BY LEATHER BELTING 



FORMULA FOR PULLING POWER OF A HORIZONTAL BELT IN TERMS 
OF ITS INITIAL TENSION 



58 For a horizontal belt on pulleys of the center distance c and one square 
inch of cross-section we now have, by substituting c for I in Formula 7, 



t. = 



2V,.-V^ + 0.04cM;i + i-i 



i, - tc 
Ratio of effective tensions = = e'P'^ 

U - tc 



to = 0.000003454 V^ 



= 0.54 - 



140 
500 -F 



[16] 

[17] 

[15] 
[1.3] 



per square inch 
of cross-sec- 
tion of belt. 



in which formulae 

c = center-distance of pulleys, in inches. 
/, = initial tension. 
ti = tension in tight strand or side. 
<2 = tension in slack strand or side. 
tc= centrifugal tension; or, more correctly, loss in 

effective tension due to centrifugal force, 
p — effective pull. 
« = basis of Naperian system of logarithms, 2.71828. 
<i> = coefficient of friction between belt and pulley. 

a — the lesser arc of contact of belt on pulleya, in radians = — X arc in 

180 

degrees. 

V = velocity of belt in feet per minute. 

59 However, an attempt to combine these five equations algebraically to 
obtain an expression of p in terms of t^, V, and a, leads to an equation solvable by 
trial only, and for this reason the diagram Plate 1, the use of which was explained 
in Par. 11 to 24 in the body of the paper, was constructed to effect the solution 
graphically. 

60 For the construction of this diagram Equation 5 was used after substitut- 
ing c for I as above, 



^1 



<. 



, and d. 



as in deriving Formula 8 from 7. It thus became 

c'' / 1 1 
V5, + \/8,-2 = 0.04-- ( -„ + _ - 2 



[18] 



TRANSMISSION OF POWER BY LEATHER BELTING 61 

and was in that form solved tentatively to obtain a aeries of points in a series of 

C" 

curves, each representing a certain value of the factor ^ ^. Onthe diagram these 

y"' 
curves form the bottom field of curves in the middle section. 

61 The equation was in each case first solved to obtain an approximate value 
only of 0,, m terms of an assumed value of ^2, by resorting to the approximation 

J,- for in the right-hand member of the equation, as shown by Formula 8. 

"'■' . 1 

62 Then by substituting this approximate value of S^ m the term ^ ^ in the 

right-hand member of Equation 18 this was again solved for a still closer value 
of 5j. 

63 For the lesser values of , this closer value of d^ differed mappreciably 

from the first approximation, while for the greatest values plotted on the diagram, 
the equation was solved twice to get the values actually plotted. 

64 However, these greatest values of never occur in the practical use of 
belting, and hence the very construction of the diagrams under consideration 
proved the validity of the approximations d^^ for — , which is equivalent to the 
approximation-^ for— resorted to in Par. 27 in modifying Equation 5 to Equii- 
tion 7. 



FORMULAE FOR PULLING POWER OF VERTICAL BELTS IN TERMS OF 

INITIAL TENSION 

65 For a vertical belt the relation between the tensions of a belt is expressed 
by the simple Equation 6, and this can readily be combined with the rest of the 
equations listed in Par. 58 (in the manner done in the Supplement to the Appen- 
dix), which leads to the following formula: 

,/e^«+l I 4e^^« 0. 000003454 V \ , 
"-■'(eT^rm-V „,«_„.+ 1 )'' ['»! 

66 This formula is simple enough, though a great improvement over the 
one derived on the erroneous supposition that the sum of the tensions is constant 
for all loads. 

67 By means of this formula the pulling-power of a belt can easily be deter- 
mined in terms of its initial tension. However, for a uniform unit initial tension 
for all speeds, the unit tension in the tight side would vary so much that belts 
ruiming at different speeds but tightened to a uniform maximum unit initial ten- 
sion, and allowed to run until this had dropped to a uniform minimum tension, 
would require re-tightening at greatly different periods. 

68 As already pointed out the writer has arrived at the conclusion that the 
periods at which belts ruiming at different speeds will have to be re-tightened, 
will b'i nearly constant if they are all made to do their work at such initial tensions 
as under full load will result in the same sum of the tension in the tight side and 



62 TRANSMISSION OF POWER BY LEATHER BELTING 

one-half the tension in the slack side of the belt, at the two extremes of the initial 
tensions, just before and after retightening. 

PULLING POWER OF BELTS IN TERMS OF A CONSTANT SUM OF THE 
TIGHT TENSION AND ONE-HALF THE SLACK TENSION, AT ALL SPEEDS 

Vertical Belts 
69 This condition is expressed by the equation 

ii + i to = A = a constant 
Combining this with Equation 17 (in a manner shown in the Supplement) we get 

(e^a _i) (2A - 0.00001036 V^) 
p=. ^ [20] 

2e"^« + 1 

<. = ^^ r22i 

and 



_ 4 A -p + V(4A-py-9p^ 
"12 

These formulae are the ones plotted in the diagrams Fig. 1 and 2 in the body 
of the paper, for A = 240 and 160 lb., respectively. 

70 By a similar treatment (as shown in the Supplement) we are also able to 
get an expression for the initial tension in a horizontal belt, which gives results of a 
high degree of accuracy. This expression is 






[24] 



which is evaluated by first determining p by Formula 20, and subsequently f, , 
and <2 by P'ormulae 21 and 22, as for a vertical belt, paragraph 09. 

71 However, while this formula is of great theoretical interest, it is hardly of 
much practical vakie; as the initial tension determined by it will differ but little 
from that determined for a vertical belt by Formula 23, except for belts of extra- 
ordinary lengths. 

72 One very interesting general conclusion may now be drawn from Formula 
24; namely, that while actually doing work two horizontal belts of unequal lengths 
may be under precisely the same tensions, but this being the case, when idle the 
longer belt will be under a slightly lower initial tension. 

73 It appears, however, that the popular notion that horizontal belts drive a 
great deal more than vertical belts, is not well founded 



TABLE 1 

EXPEKIMES'TS BY WlLFRED LeWIS, AT THE WoRKS OF Wm. SeLLEEIS & Co., PhILADELPBIA, 

1S85, ON' Single Belt 5i in'. Wide by /t in. Thick and in Ordinary Working Con- 
dition Without Belt Dressing. Belt Speed = 800 ft. per Minote. These Experi- 
ments ARE Plotted i.n Fig. 7 of the Supplement, which see. See also his Paper, 
No. 198, Vol. 2 op Transactions, Table 1 



.So .So 



>>■£ » 



& H 



Q 0) 

o -a 

N - I' 



2 "o 

I w ~ 
,^ II 

•lira 



o. 



o o »- 

■S '^ II 

Cm 

^■^ I I I ^ ^ ^ 

0~-~ Ohi3.2o 
C art! 



•s- £• II :« 

u a> a 

i 5 c2. V ■" 



o 



60 
61 
62 
63 
65 



81.6 lb. 
persquareinch 



125.33 

131.42 
142.00 
152.41 
179.92 



58.67 


0.5 


2.0 


0.251 


0.41 


46:58 


0.9 


3.6 


0.336 


0.53 


42.00 


1.7 


6.8 


0.407 


0.62 


35 . 75 


3.0 


12.0 


0.490 


0.73 


29.92 


12.0 


48.0 


0.610 


0.91 



66 I ' 


177.42 1 


77.42 t 


1 
0.5 1 


2.0 


0.270 


0.52 


68 


198.25 


64.92 


0.8 


3.2 


0.365 


0.69 


69 127.5 1b. 


208.77 


58.67 


1.0 


4.0 


0.418 


0.77 


70 persquareinch 


219.08 


50.75 


1.7 


6.8 


0.472 


0.87 


71 


229.50 


46.17 


2.6 


10.4 


0.545 


0.95 


72 


244.08 


44.08 


3.8 


15.2 


0.569 


1.02 


73 


256 . 58 


39.92 


3.5 


22.0 


0.623 


1.10 


74 


252.42 


35.75 


8.6 


34.4 


0.677 


1.13 


"1 J 


283.66 


33.67 

1 


15.2 ] 


60.8 


0.719 


1.25 



TABLE 2 

Experiments bt Wilfred Lewis, at the Works of Wm. Sellers & Co., Philadelphia, 
1885, ON Double Belt 2} in. Wide bt jj in. Thick and in Okdinart Working Con- 
dition Without Belt Dressing. Belt Speed = 800 ft. pee Minute. These Experi- 
ments ARE Plotted in Fig. 19 of the Supplement, which see. See also his Paper, 
No. 198, Vol. 2 of Transactions, Table 2 



a o 



-a ^ 
m 
II 



<o II 

"I M m 

t3 






_ a 



c 5 



fc. g 



§ 2 ^° ^« 



p. o, 

Ir: 4> ■ 

•30 

= o ■ 



o 00 



105 




104.9 


47.5 


0.3 ! 


1.2 . 


0.263 


0.38 


106 


73.5 lb. 


123.4 


37.5 


0.8 


3.2 


0.395 


0.57 


107 , 


persquareinch 


146.0 


32.6 


1.7 


6.8 


0.511 


0.73 


108 ! 


1 
1 


171.5 


29.5 


4.3 


17.2 1 


0.600 


0.87 



121 
124 
125 
126 
127 



128 
131 
133 
134 
135 



283.0 lb. 
per square inch 



343.5 lb. 
per square inch 



403.0 
450.0 
465.0 
482.2 
497.5 

511.3 
557.0 
589.5 
603.0 
618.0 



175.9 
137.5 

124.8 

113.5 

99.2 

227.0 
187.2 
162.4 
148.2 
134.0 



0.7 
1.5 
2.3 
3.7 
10.1 

0.5 

1.1 
1.8 
2.7 
5.1 



2.8 

6.0 

9.2 

14.8 

40.4 



2.0 

4.4 



10.8 
20.4 



0.267 
0.387 
0.424 
0.469 
0.523 

0.261 
. 3,50 
0.414 
0.450 
0.490 



0.77 
1.07 
1.17 
1.28 
1.39 



0.85 
0.99 
1.30 
1.39 
1.49 



64 



TRANSMISSION OF POWER BY LEATHER BELTING 



TABLE 3 
Relations Between Coefficient of Fbiction, Velocity of Sliding and Belt Speed 

See Par. 54 



V = Velocity in 
feet per minute 


c 


00 M 

X 00 o 

o , o 6 
§ ^+ + 

II O 00 


Velocity of Slid- 
ing V = 
160 + 0.88F 


o 
d 

+ 

00 


n It. 
§ § •« 

o Sh us 
o d 


Velocity of Slid- 
ing; at 1 per cent 
slip 

" =200 


II ^ 
-a- »g + 

"o "o (N -f-'* O 

a c 'I' » 
■%-c 1 ' 
5H ■- to "=- 
gfc, d ° 
o B 







8 


1.8S 


0.260 


0.00 


0.100 


50 




9.432 


2.36 


0.285 


0.25 


0.129 


100 




5.636 


2.82 


0.307 


0.50 


0.156 


200 




3.690 


3.69 


0.340 


1.00 


0.200 


300 




3.010 


4.51 


0.365 


1.50 


0.236 


400 




2.640 


5.28 


0.3S4 


2.00 


0.267 


500 




2.400 


6.00 


0.400 


2.50 


0.292 


600 




2.227 


6.68 


0.413 


3.00 


0.314 


700 




2.090 


7.32 


0.423 


-3.50 


0.333 


800 




1.983 


7.93 


0.432 


4.00 


0.350 


900 




1.889 


8.50 


0.440 


4.50 


0.365 


1000 




1.808 


9.04 


0.446 


5.00 


0.378 


1200 




1.675 


10.05 


0.458 


6.00 


0.400 


1400 




1.566 


10.96 


0.466 


7.00 


0.418 


1600 




1.474 


11.79 


0.473 


8.00 


0.433 


1800 




1.394 


12.55 


0.479 


9.00 


0.446 


2000 




1.325 


13.25 


0.484 


10.00 


0.457 


2500 




1.180 


14.75 


0.493 


12.50 


0.479 


3000 




1.067 


16.00 


0.500 


15.00 


0.495 


3500 




0.974 


17.05 


0.505 


17.50 


0.507 


4000 




0.898 


17.95 


0.509 


20.00 


0.517 


4500 




0.832 


18.72 


0.512 


22.50 


0.525 


5000 




0.768 


19.40 


0.514 


25.00 


0.531 


5500 




0.727 


20.00 


0.517 


27.50 


0.536 


6000 




0.684 


20.53 


0.519 


30.00 


0.541 


6500 




0.646 


21.00 


0.520 


32.50 


0.545 



DISCUSSION 



Henry R. Towne. The earliest investigation of this subject 
was by General Morin, of the Conservatoire des Arts et Metiers, 
who gave, in a volume published, I think, about 1850, the results of 
his experiments to determine the coefficient of friction of belts on 
pulleys, and algebraic formulae to express the power transmitted under 
varying conditions. For many years these formulae were accepted 



TRANSMISSION OF POWER BY LEATHER BELTING 65 

universally. General Morin's experiments were made under labora- 
tory conditions. 

2 In 1867 I made a series of experiments to determine, under 
conditions approximating those of actual use, the coefficient of 
friction and also the tensional strength of commercial belting. These 
experiments, and a discussion by the late Robert Briggs on the 
mathematical conditions involved in the problem, were pubhshed in 
the Journal of the Franklin Institute in 1868. Under the title of 
the Briggs and Towne Experiments, the conclusions thus reached 
were quoted and accepted for many years, by Professor Ranldne, 
Professor Reuleaux, Professor Unwin, and many other technical 
writers. A. F. Nagle, in a valuable paper contributed to the Trans- 
actions of the Society in 1881 (Vol. 2, p. 91), accepted the results 
of the Towne experiments as the basis for his discussion of the 
mathematical problems involved. 

3 The Transactions for 1886 (Vol. 7) contained two important 
contributions to the literature on this subject. One of these is a 
paper by Professor Lanza (p. 347), which first prominently calls 
attention to the importance of syeed of slip as a factor in the trans- 
mission of power by belting. The other is a paper by Wilfred 
Lewis (p. 549) giving the results of a long and elaborate series of 
experiments in the shops of WiUiam Sellers & Co., and demonstrat- 
ing, among other things, that the proposition first enunciated by 
General Morin, and accepted unquestioningly by all subsequent 
authorities, namely, that the sum of the tensions is constant (T^+T^), 
does not hold true in all cases, and is therefore erroneous. 

4 The Transactions for 1894 contains another most valuable paper, 
by Fred. W. Taylor (p. 204), giving the results of his large expe- 
rience covering many years in the use and observation of belting 
under the conditions of actual practice. Many new and important 
deductions based on the investigations of Mr. Taylor are availed 
of by Mr. Barth in the conclusions and recommendations contained 
in his paper. One of the most important facts demonstrated by 
Mr. Taylor relates to the value of increased thickness of belts, and 
the resulting expediency of a larger and more general use of double 
belts. He was also the first to demonstrate and set forth clearly 
the economic gain to be derived from the scientific care of belting. 

5 Finally, Mr. Barth, avaiHng himself, as he has stated, of the 
work of his predecessors, especially that of Mr. Lewis and Mr. 
Taylor, has completed, for the present at least, the study of this 
problem, which has thus extended over some sixty yearSj giving us an 



66 TRANSMISSION OF POWER BY LEATHER BELTING 

elaborate and apparently a conclusive demonstration of the sound- 
ness of the mathematical conclusions finally reached, furnishing work- 
ing formulae for practical use, and presenting a most ingenious appli- 
cation of the slide rule to the problems involved in the practical use 
of leather belting. 

6 The Society is to be congratulated on including in its roster of 
membership the names of all those since General Morin who have 
taken the lead in ascertaining the facts and in determining therefrom 
the rules which govern the application of leather belting to industrial 
uses. 

7 Mr. Earth's system has now been in use for about two years in 
the works of the Yale & Towne Mfg. Co., Stamford, Conn., where it has 
accomplished a substantial increase in economy and efficiency. 

Wilfred Lewis. I am clearly of the opinion that Mr. Barth has 
discovered and formulated principles of the greatest practical value 
in the solution of the problems of the transmission of power by 
leather belting. 

2 It is difficult in a paper of this kind to separate the practical 
from the theoretical without discarding the most valuable part of 
the undertaking. The laborious work done by the author in order to 
reach his conclusions, and recorded in the appendix to this paper, 
is really the basis of the superstructure reared by him and gives 
the reader some idea of the immense amount of patient research and 
good sound reasoning employed in building up a complete analysis 
of the subject. 

3 Mr. Barth is the first, I believe, to analyze the peculiar elastic 
properties of leather, and to demonstrate in a convincing way the 
effects of these properties in the use of belting under varied con- 
ditions. His analysis of the combined effects of elasticity and sag 
is very original and ingenious, and even aside from the results obtained 
his methods cannot fail to interest investigators in other fields of 
research. Difficult and complex problems have been solved by 
making certain assumptions and approximations that are quite allow- 
able as the means to an end, and it is in these short cuts from the 
intricate and unwieldly to the simple and practical that he has dis- 
played such remarkable ingenuity. At the same time, for those not 
enough interested in every step to care to follow a mass of mathe- 
matical formulae, Mr. Barth has presented his conclusions in a form 
available for immediate use. 



TRANSMISSION OF POWER BY LEATHER BELTING 67 

4 Popular impressions, even though well founded, are often 
exaggerated beyond reasonable bounds, and while it is true that 
horizontal belts of considerable length are preferable in the trans- 
mission of power to vertical or shorter ones, it will be a surprise, I 
believe, to engineers, that there really is so little advantage in a 
long horizontal belt over any length of belt in any position. All 
this results from the exposure of the fallacy that the sum of the 
tensions is constant, a belief exploded 23 years ago, although the 
far-reaching effect of the exposure on the transmission of power by 
belting has never before been so clearly expounded. 

5 The author's treatment, also in the unpublished supplement 
to the appendix, of the effect of variations in pulley diameter upon 
the transmission of power, I believe to be absolutely original, and 
his conclusion that a belt will slip on a driven pulley before it will 
slip on a driver of the same diameter indicates a subtlety of analysis 
rarely displayed in our proceedings, and is a fair index of the pains- 
taking care with which the whole paper has been written. Although 
not perhaps of very great practical importance, as a new discovery, 
the analysis might well be included in the appendix to the paper, 
rather than in the unpublished supplement to the appendix. 

W. D. Hamerstadt. The writer has been somewhat closely 
associated -uith work on pulley and belt drives, and recently has had 
occasion to compare the results of some carefully conducted experi- 
ments with the results which might be expected from the use of 
formulae as proposed in Mr. Earth's paper. Considering the many 
variable factors, these comparisons are remarkably favorable, and 
for average conditions of operation, the relationships which have been 
established would appear to hold quite true. 

2 One almost vital point of consideration in the actual design 
of belt drives seems to have been touched upon but lightly, however, 
and then in a manner which, as the author himself has stated, leaves 
some room for discussion — namely, values of the coefficient of fric- 
tion to be used in the formulae given, under varying conditions of 
service. While the value of the coefficient of friction will not affect 
the theory of belt transmission as given, it will seriously affect the 
size of drive required to do a given work, and having now a good 
theoretical basis for work, and being assisted by the observations of 
others, additional experimental work might well be done for the deter- 
mination of such values, using as nearly as possible good average 
leather belting and operating under actual conditions of service. 



68 



TRANSMISSION OF POWER BY LEATHER BELTING 



3 Based partly on the conclusions of Professor Lanza or of Mr. 
Wilfred Lewis, as given in early papers before the Society, and partly 
on the very mechanical reasonableness of the thing as he puts it, 
Mr. Barth assumes that, given a belt and pulley, the value of the 
coefficient of friction to be used in any case will be determined to a 




2 4 6 8 10 12 14 16 IS 20 24 28 3Z 36 40 

Fig. 1 Relation Between Coefficient of Fkiction and Velocity of Slip 






.0 
.9 




























































































— EFf 


ECTl 


\Z RAl 


tGE — 


















s^ 








=a 


ss 


.U 
.7 

6 
.5 
.4 
.3 

2 




















^ 


r^ 


S^ 


■■=■ 


=^ 


^ 


"S^ 




fc= 




















-^ 


^ 


^ 




^ 


>^ 


n 










^ 


















y^ 


^^ 






::::; 


F-« 






== 


= 






















^ 


^ 


^ 


































A 


^ 


^ 
















' 
























y 



























































































































Fig 



1^ 2 

2 Relation Between Coefficient 



3 4 5 

OF Friction and Percentage of Slip 



great extent by the velocity with which the belt sUdes on its pulley. 
Taking then a curve representing average relations between these 
two factors for any convenient speed of belt, values are at once avail- 
able for the coefficient of friction for any speed of belt and any condi- 
tion of sHp desired. 
4 That there exists some ground for such reasoning cannot be 



TRANSMISSION OF POWER BY LEATHER BELTING 



69 



1 




"' c 


V 








c 


c 
















o 


> 

01 








el 


4) 

•a 










^ 






=5 a, 


"S 








c 

a 


« 
S 










E 

'S 


j 




0) Q^ 


a 








o 
O 


2 










1 (0 

s: 


^ 


2 

> 








0) 

e 




(0 

"a 








e 

o 


t 


T3 4) 










3 


c 










o 

42 


H 


c •« 


C3 








< 


c 



a 


r3 








c9 


i tf 


:i 


.£• '• 








6^ 










1 


1 




^ m 


§ 1" 










v 


3 








2 


1 






_ 3 








+ 


O C3 








2 






n 










Ex 












H 




a 1 


o 




O 




o 








•* 






S 


>< 


.2 o 






N 




C4 








« 






C4 


U 


Q ^ 














































































1 C} 

1 a, 




1-J 








'^ 








ri 






-*' 


1 


3 


d 




o 




o 


O 






d 






d 










•o 




























c 












c 






c 




c 


T3 




e! 




•a 


v 






d 






oi 




c 
o 


C 

ei 







C 


1 


§ 1 






01 

-a 
3 








; 


O 


- <1> 




T 


c 


'O o -g 






"b 






>• 


5 




2; 




c 




C 


a 






Q 






Q 




-ts 




H 


; 


-C 






-15 






"C 


P9 


^1 


K 








>< 








X 






X 






■* 




ir 




cs 








lO 






M 




"O 
















•2 


<D 




«> 






c 




_a 




3 






V 

1 


■a 

d 




"3) 




A a 


Ji 3 




"t 


1 


3 






^ 


1 




? 




h] 


Is o 




c 




o 






<d 




o 






O Q 




a 




Q 






o 


c 




» 






















o 






o 


« 


















o 






o 


H H . 












c 






•^ 






t^ 




J t- Z 


o 




c 




o c 






(N 






c^ 












■^ If 

00 ^ 






i 






o 


B. 


Ii4 


















o 






1" 


CO 




















c-^ 






<N 






& 






~f' 




& 


m 










o 

1 








m 






H 




H 














■* 










(N 


: H ^ 






0. 












o 






_2 




01 


: t-' J 










d 






"3 






3 




3 












6 


o 






03 






03 




,* 


• -t; c 


. 




i^ 




o 


»-< 




S 


H 






H 




H 








•g 




c 


1 


i 


-1 








ii 

1 

1^ 




0) 


: H^ 






t3 




o 








-2^ 




1" 


> 


T3 «: 








"3. 

a 




> 

8 

a 


O 






g^ 




IS 


> 

3 










u 

o 
o 




m 




H 




























g 


< 




« 




o 


Q 






w 




b 






D 


























c 


J 





























a - 



0> 




-o 


^ 




«.-i 







o 


i!i 


^ 


s 



o .S 



70 TRANSMISSION OF POWER BY LEATHER BELTING 

denied, but a brief comparison of results actually obtained from 
tests performed from time to time, on belts operating at widely 
different speeds of service, leads one seriously to question its applica- 
tion to practice. Such comparisons rather lead one to expect more 
nearly correct results when drives are designed on the basis of relative 
slip between belt and pulley. 

5 Fig. 1 and Fig. 2 show a series of curves representing, for a 
number of different tests, the relation existing, first between values of 
the coefficient of friction and velocity of slip, and second, between 
values of the coefficient of friction and percentage of slip. Informa- 
tion regarding the data from which these curves were plotted is given 
in the table. Letters designate corresponding tests in either set 
of curves. 

6 Referring to Fig. 1, it will be noted that for each different speed 
of belt there appears to exist a clear and well defined relation between 
values of the coefficient of friction and velocity of slip; at lower 
velocities of slip more especially, the value of the coefficient appearing 
to be higher for slow-speed belts and lower for high-speed belts. 
Obviously then, any curve representing a relation between values of 
the coefficient of friction and velocity of slip holds true only for that 
speed of belt for which it is plotted and cannot be used indiscrim- 
inately for all speeds of belts. The effective range of velocities of 
slip which would be used in the design of belt drives would probably 
be from to 25 ft. per minute, as indicated in Fig. 1, and the error 
which might be incurred then in using either of the extreme outside 
curves shown (even though they do not represent maximum possible 
range of speeds of belts) would vary from about 70 per cent to values 
almost infinitely large. The curves B^ and C^ indicate the relation 
that would presumably have held true between values of the coeffi- 
cient of friction and velocity of slip for belts B and C had those belts 
been such as to have shown a maximum value of the coefficient equal 
to that of belts A, E or F. 

7 From Fig. 2 it will be noted that for any given speed of belt 
the same general relation between values of the coefficient of friction 
and per cent of slip appears to hold true, and belt drives designed on 
such a basis might then reasonably be expected to give anticipated 
results at all speeds of operation. When curves B^ and C^ are plotted 
to represent higher values of the coefficient of friction for belts B 
and C, as in Fig. 1, the similarity in form of these curves is remarkable 
— the more so as they represent tests performed in some cases over 
twenty years apart. 



TRANSMISSION OF POWEK BY LEATHER BELTING 71 

8 As an example of the results to be expected when drives are 
designed on the basis of velocity of slip of the belt, as proposed 
in the paper, let two extreme conditions of service be taken: one a 
drive operating at a speed of 400 ft. per minute and at a slip of 
about 2^ per cent, the other a belt operating at a speed of 5000 ft. 
per minute and at a slip of about 1 per cent. The slow-speed belt 
would then have a velocity of slip of 5 ft. per minute on each pulley, 
the high-speed belt of 25 ft. per minute. Referring to Fig. 7 in Mr. 
Earth's Appendix, it will be found that for such velocities of slip 
the values of the coefficient of friction to be used should be respect- 
ively 0.38 and 0.53; but the maximum value of the coefficient of fric- 
tion at even the highest velocities of slip of 60 ft. per minute, as shown, 
is only about 0.57 and it appears then that the overload capacities 
of all drives is Umited to about that value. This amounts, in the 
case of the slow and high-speed belts given, to about 50 and 7^ per 
cent respectively. 

9 It is safe to say, however, that fully 80 to 90 per cent of all 
high-speed drives are used in connection with electrical machinery, 
and for such work drives must have an overload capacity of at least 50 
per cent of their rated capacity. It would be necessary then, for 
such practice, that belts be originally designed for correspondingly 
lower velocity of slip, amounting in this case to a velocity of about 5 
ft. per minute. The high-speed belt, noted above, at ordinary con- 
ditions of service would then operate at a relative total slip between 
pulley and belt of only one-fifth of one per cent. 

10 While the author has had occasion to observe a large number 
of successful high-speed drives on electrical machines, just the con- 
dition of slip here indicated has never been noted, but in almost every 
case the overload capacity of the drives has been noted as a function 
of the relative slip between belt and pulley. 

11 When it is further considered that, generally speaking, the 
point at which the belt will leave its pulley is a function of the per 
cent of slip and is taken independent of the speed of belt, it certainly 
appears that the relative slip cannot but play an important factor 
in determining values of the coefficient of friction to be used for such 
drives. 

Fred. W. Taylor.* The belt is one of the oldest and most 
commonplace of the elements used in shop practice, so that engineers 
designing new establishments or remodeling old ones, who wish to 
be up-to-date, naturally incline toward the use of the electric drive 

' Discussion abstracted. 



72 TRA.NSMISSION OF POWER BY LEATHER BELTING 

rather than the belt. There is no doubt, however, that this has led 
to the use of the electric drive in many instances where the belt would 
be far more economical and satisfactory in almost every way. 

2 In the average machine shop, for instance, the writer is pre- 
pared to say that for more than half of the machines the belt drive 
can still be used with greater economy and with more satisfactory 
results than the electric drive; only on the assumption, however, that 
the belting is systematically cared for. The most serious objection 
to the belt drive as generally used is the loss of time due to interrup- 
tion to manufacture when retightening and repairing, and to the 
loss of driving power and consequent falling off in output, when the 
belt is allowed to run too slack. Belts can be tightened and repaired 
at regular intervals after working hours, however, with the use of 
spring-balance belt-clamps to get the right tension, causing thus 
practically no interruption to manufacture. 

3 As will be explained later, it has been shown by an accurate 
record kept through a long term of years, that in the average machine 
shop the average cost per belt per year is $2.25. This includes the 
original cost of the belt, plus all labor and materials used in main- 
taining, repairing and cleaning it throughout its life. No similar 
statistics for the maintenance and renewal of the motor drive seem 
to be available, but I think no one will contend that the latter can in 
any way approach this economy. 

4 In a great number of cases the electric drive should be used in 
the machine shop, but in the writer's judgment the burden of proof 
still rests on the motor drive to show in each case that the economy 
in delivery and removal of work more than makes up for the extra 
cost of installation and maintenance, and for the delays incident to 
repairs, blowing out fuses, etc. In large machines economy lies on 
the side of the motor drive in many instances, but with almost all 
small machines the belt drive should still be used. In view of these 
facts, the belt drive is hardly a back number.'T In fact, the manager of 
one pulley manufactory told me recently that even during the dull 
times his company had been selling from twelve to fifteen thousand 
pulleys per month. 

5 Under the rules still in common use, a large proportion of belt 
drives are badly designed, and belts are used under heavier tensions 
than they^should be for all-round economy.P^All who have experi- 
mented with belting or who have been interested in the mathe- 
matics of belting, will be filled with admiration at the remarkable 
analysis which Mr. Barth has made of this diflacult problem. Even 



TRANSMISSION OF POWER BY LEATHER BELTING 73 

Mr. Lewis, whose experiments and scientific conclusions have properly- 
been s;iven first place among writings on this subject, tells us in his 
paper that life is too short to attempt a complete mathematical solu- 
tion of the problems involved. Yet this is precisely the task which 
has been accomplished by Mr. Barth. 

6 The experiments of Messrs. Briggs and Towne and those of 
Messrs. Bancroft and Lewis will remain for many years as classic 
monuments in the development of our scientific knowledge of belt- 
ing laws; but Mr. Barth's remarkable analysis of the work of former 
experimenters, supplemented by his accurate though less voluminous 
experiments on the elastic properties of belting and on the rate and 
extent of the fall in tension of belts, has rendered his conclusions as 
to economical speeds and the proper sizes of belts more reliable than 
those of any previous writer. His final recommendations should be 
accepted, therefore, rather than those in the papers of Messrs. Towne, 
Lewis, or the writer. ^ 

7 It may be of interest to know how the figure of $2.25, quoted 
earlier in the paper as the cost per belt per year, was found. 

8 In the new machine shop of the Midvale Steel Company, 
beginning in the year 1884, the writer experimented^ with all of the 
belts in the shop, in practical use; and upon the comparative values 
of the four leading types of leather belting then in common use. 
This experiment lasted nine years with belting running night and day 
(equivalent to eighteen years running ten hours per day). Exact 
records were kept of all items affecting the life and economical use 
of belting, and at the end of the experiment, among other items, it 
was found that the average belt cost (under the ordinary belt rules 
then in use, as, for example, those used on the cone pulleys of the 
various machines; and on the ten hour basis) $3.34 per belt per year 
for the first cost plus all labor and materials used in maintenance 
and repairs. These are double belts, averaging 29 ft. long by 3.8 
in. wide. 

9 These belts were run under too high ftension lor economy, 
however. They lasted on an average 14 years (ten hours per day). 
The remaining belts in the shop, which proved more economical, 
lasted on an average not far from 28 years (ten hours per day), and 
cost per year per belt less than $2.50 for first cost and maintenance, 

* These experiments are described in a paper entitled Notes on Belting, pre- 
sented before the Society December 1893, and forming part of Volume 15 of 
the Transactions. 



74 TPIANSMISSION OF POWER BY LEATHER BELTING 

etc. And this although they were materially larger than the cone 
belts, averaging 50 ft. long by 4.84 in. wide. The machines in this 
shop averaged much larger than in the average shop, and an investi- 
gation has led me to the conclusion that in the average shop the aver- 
age belt would be about equal to a 3-in. double belt, 20 ft. long. The 
first cost plus the maintenance of this belt would not be greater than 
$2.25 per belt per year. 

10 The care of belting should be entirely taken out of the hands 
of the men who are running the various belt-driven machines, and 
belts should be systematically retightened at regular intervals, with 
belt-clamps fitted with spring-balances, each belt having the tight- 
ening strain carefully figured in advance. Belting should also be 
cleaned at regular intervals, and should be softened with the small 
amount of belt-dressing which is needed to keep it in perfect con- 
dition. A laborer can be quickly trained to tighten and care for all 
the belts in the shop during the noon hours and on Saturday after- 
noons and at other times when the shop is not running. 

11 Two elements of great importance in Mr. Earth's paper are 
The Influence of Pulley Diameters on the Sum of the Tensions of the 
Belt and a condensation of the discussion of the formula. Ratio of 
Effective Tensions e 4>°^. Not only has this discussion a great the- 
oretical interest, but the conclusions have a distinct practical 
value. 



Charles Robbins.^ In applying motors in textile mills where the 
belt has been in use for years and the proposition is essentially that 
of constant and uniform speed we discovered that the capacity of the 
spinning frames was largely increased. This led us to make tests upon 
the loss of speeds, or slip of belts and their lack of uniform operation. 
The net result in using the induction motor instead of the belt is an 
increased production of at least seven per cent, and in some instances 
even ten per cent. Probably some of this increase is due largely to 
the fact that the belting systems tested were not designed in accord 
with Mr. Barth's system; but I believe that a system of belts will never 
approach the uniform and constant speed of an induction motor. 

2 The question of efficiency may be classified as (1) the primary 
efficiency from the engine shaft to the shaft of the driven machine; 



'Charles Robbins, Westinghouse Electric and Manufacturius Company, East 
Pittsbiire. Pa. Discussion abstracted. 



TRANSMISSION OF POW^R BY LEATHER BELTING 75 

(2) the economies which result from the use of tlie electric motor 
drive. These secondary economies, which are undoubtedly the most 
important, will vary with the class of industry to which the electric 
motor is applied. It is greatest in those industries where the load- 
time factor of the installation is lowest and where the inherent charac- 
teristics of the electric motor are of greatest value. These charac- 
teristics are as follows: 

a Ability to adjust the speed according to the demands of the 
work. 

b Absolute certainty of a uniform and constant speed. 

3 While these two characteristics may seem to be opposed, they 
are important factors in the increase of production of different types 
of macliines. As widely separate examples: for a machine tool the 
readiness with which the speed of a motor may be varied to the right 
quantity for the work required contributes to its increase of production; 
on the other hand textile mill service requires an absolutely constant 
and uniform speed, which is obtained from the induction motor. 

4 In determining the value of an electric motor drive the essential 
point is always the secondary, or accruing economies from its use, 
rather than the primary economy, although when the primary is added 
to the secondary the net result will be extremely satisfactory. 

Geo. N. Van Derhoef.^ The author's plan of proportioning 
belts so that the slack will be taken up at approximately regular 
intervals of time, regardless of speed or power transmitted, is excel- 
lent from a theoretical point of view. He is obliged, however, to 
divide belts into two classes — machine belts and countershaft belts, 
under different initial tensions, and therefore with different periods 
between adjustments. I think it will be absolutely necessary to 
provide more classes. In some cases first cost is of greater impor- 
tance, and in other cases the expense or inconvenience of taking up 
belts is the main consideration. With a large belt, running night 
and day, the stopping of the drive to take up the belt is a serious 
matter. In the case of many drives, however, this is a matter of 
small moment. 

2 I have had considerable experience with large quarter-twist 
belts, running from 12 to 20 in, in width, for connecting horizontal 
and vertical shafts, and have seen results that appear incredible in 
view of much of the theoretical data published on belting. These 
belts were under high unit-tension, and always subjected to reverse 

' Discussion abstracted. 



76 TRANSMISSION OP POWER BY LEATHER BELTING 

bending over deflecting idlers. Probably one reason for the success 
of belts of this kind is the automatic regulation, within limits, of the 
slack-side tension, due to the belt worldng up and down across the 
face of the pulley on the vertical shaft. As far as I have observed, 
belt drives of this kind, when properly designed and erected, have been 
as satisfactory as horizontal belts with about the same distance 
between centers. 

3 Possibly the larger unit-stresses frequently used necessitate a 
slight actual slipping of the belt on the pulleys, with some correspond- 
ing increase in the coefficient of friction. This should not necessarily 
be 'regarded as poor practice, but simply as a factor to be weighed 
against savings in first cost, friction losses, etc. There seems to be 
no fundamentaJ objection to slipping within certain limits, pro- 
vided such slip is a constant quantity. All belts are continually 
sliding, to some extent, on the surface of the pulleys, due to the 
theoretical creep caused by the elasticity of the belt. A little more 
would not necessarily be serious. The surface of a well finished 
leather belt is such that sliding on a polished iron pulley will not cause 
much harm provided the heat generated by the slip is dissipated with 
sufficient rapidity to prevent the temperature of the belt surface from 
rising too high. This, of course, involves a loss of energy, as do very 
large belts under low tensions, and the crowning of pulleys. The 
writer desires to emphasize that due consideration should be given 
to all the factors involved. 

4 Spring belt-clamps should be used wherever practicable, and 
ought not to be very expensive if manufactured in reasonable quanti- 
ties. In the majority of cases, however, we shall have to be satisfied 
with figuring belts properly, and leave the actual initial tension to fate. 

5 The idea that the maximum working-stress of a belt should 
not be determined by its ultimate strength is, I believe, correct. 
This becomes more apparent in studying transmission ropes. It is 
a well-known fact that the maximum unit-stress for a manila trans- 
mission rope should be of such amount that the side-pressure between 
the lubricated fibers of the rope will not cause abrasion when the 
ropes bend over the sheaves, and the fibers slide on one another. 
Probably some such internal action takes place in the case of leather 
belts. In transmission ropes the ultimate strength bears a greater 
ratio to the proper maximum working stress than is the case with 
leather belts. Manila rope is therefore a very safe transmitting band. 

6 The constant lengthening of belts in service has its counterpart 
in ropes. Where a rope is simply carried around two sheaves, as 



TRANSMISSION OF POWER BY LEATHER BELTING 77 

in the separate rope system, the general equation of the rope is 
without question similar to that which the author has shown to be 
true of leather belts. 

7 The continuous system of rope transmission, with its automatic 
tension carriage, has the slack-side tension maintained at a minimum. 
This is one of the fundamental reasons why the continuous system 
can transmit the same amount of power at the same rope speed and 
with the same rope life, with less rope than is possible with the sepa- 
rate wrap system. A few years ago the continuous system was looked 
upon by most engineers with considerable scepticism; its enormous 
development in the last quarter of a century is due simply to its 
basis on absolutely sound mathematical principles. 

Walter C. Allen. My contribution to the discussion will 
relate to the practical results obtained from the installation of an 
improved method of caring for belting, rather than to the technical 
phases of the question. In this connection a brief description of 
the working out of the improved system in the works of the Yale & 
Towne Mfg. Co. may prove interesting. 

2 The problem of transmitting large amounts of power by means 
of belting is not a serious one with us, as our power is for the most 
part transmitted electrically; each room is provided with one or 
more motors, and the power is distributed from them through line 
and countershafts to the machines. The great majority of our belts 
are small, and many of them run at high speeds. Altogether we have 
about 4800 belts, so that their proper maintenance is an important 
and somewhat difficult problem. 

3 Early in 1905, at Mr. Barth's suggestion we undertook an inves- 
igation of our belting and the methods employed in its upkeep, 

as a result of which we decided to adopt a system of caring for belting 
recommended by Messrs. Taylor and Barth. For the sake of brevity 
I have divided my notes into comparative statements, of the con- 
ditions before and after the adoption of the new method as'affecting 
each element of this important subject. It may seem that the con- 
ditions existing before the installation of the new plan were dis- 
tinctly bad, but I venture to say that they were as good as those in 
many manufacturing establishments at the present time, if not 
better. The improved conditions, however, are so infinitely superior 
to the old that by comparison the latter appear extremely anti- 
quated and crude. 

4 Tensions. Under the old plan we had no means of knowing 



78 TRANSMISSION OF POWER BY LEATHER BELTING 

with any accuracy the tension of a belt. It was left to the individual 
judgment and experience of those doing the repairing, so that inevit- 
ably the tensions of the belts varied in proportion to the variation 
of judgment of the repair men. 

5 The first step in the reorganization was the building of a belt 
bench and the provision of tension scales such as are shown in Fig. 6. 
These are used now altogether for the determination of tensions. 

6 Records. Under the old regime we had no records whatever 
of our belts. 

7 Under the new plan we have a record of each belt showing its 
location; its type, i. e., whether open or crossed, countershaft or 
machine belt; kind of leather; thickness, width and length. These 
records also show for each belt the dates of inspection. 

8 Organization. Under the old plan our millwrights cared for 
the heavy belts, but the repairing was done only when the belt gave 
way, or stretched so that it failed to transmit the necessary power. 
The small machine belts were cared for by the individual macliine 
operators, rhany of whom knew absolutely nothing about belting, 
and in some cases our investigations showed that ignorant operators 
had attempted to tighten a belt by cutting out a piece, and, finding 
that they had cut out so much that the belt would not go over the 
pulleys, were then compelled to cut out still more and set in a piece 
in order to make the belt long enough to do the work. In these 
cases also the belts were not repaired until they actually gave out 
through breakage or failed to give the necessary pull. 

9 Under the new plan a gang of four men do absolutely nothing 
else but inspect belting and attend to the repairs and retightening. 
A belt room has been provided in which is an annunciator, and a 
series of push buttons are arranged at the telephone central, so that 
in case of an accident to a belt the foreman or gang boss can call the 
belt man easily. In a plant as large as ours the annunciator results 
in a great saving of time. 

10 A tickler system was installed by means of which <;he belt 
gang are notified regarding the belts to be inspected each day. 
After the inspections are made these tickler cards are returned to 
the office where the proper records are made and the ticklers put 
back for the next inspection. 

11 These belt men take their lunch hour from 11 to 12 o'clock, 
working during the noon hour, and are thereby enabled to repair 
many belts which could not be repaired when the works are running, 
without loss of time to other employees. 



TRANSMISSION OF POWER BY LEATHER BELTING 79 

12 Fastening. Under the old plan there was no fixed rule regard- 
ing the fastening, rawhide lacing and belt hooks being used indis- 
criminately. Under the present plan Jackson wire lacing, put into 
the belts by means of a machine, is universally used. For continuous 
belts, under the old plan we used a kind of glue which took from three 
to ten hours to set satisfactorily. Under the present plan we are 
using a special glue which will set hard in thirty minutes. This also 
results in a saving of time in the case of an accident to continuous 
belts. 

13 Belt Dressing. Under the old plan comparatively little belt- 
dressing was used, but in many cases rosin was used through ignorance 
of the fact that it causes the belting to deteriorate rapidly. We 
now use entirely Plomo belt-dressing, which is extremely useful and 
tends to prolong rather than to shorten the life of the belt. 

14 Reclamation of Belts. Under the old plan no reclama- 
tion was attempted, but at the present time we reclaim a consider- 
able amount of belting each year. Belting damaged on the edges 
is cut down and used for narrower belts, short pieces are scarfed 
and glued together and the oil is taken out of oily belting and the 
belts used over again. 

15 Kind of Belting. Several kinds of belting were used under 
the old plan, but we have gradually standardized our belting until 
at the present time practically nothing but a high grade of oak- 
tanned belting is used. 

16 Cost of Up-Keep. Of course there was no method of determin- 
ing the cost of maintenance under the old plan. Our records show 
that during the year 1906 the labor-cost of maintaining our belting 
system was 96 cents per belt. During 1907 it was 73 cents and dur- 
ing 1908, 45 cents. This decrease has of course been due to the 
increased efficiency of the men doing the work and to the fact that 
experience has indicated where inspection periods could be lengthened 
out, and also to the fact that the belting is now in such condition that 
expensive breakdowns seldom occur. 

17 The foregoing statements describe briefly the various features 
of the old and the new plans; a summary of the advantages of the 
new plan follows: 

a Decreased cost of belting. The cost for the year 1907 was 
only about 60 per cent of that for 1906, despite the fact 
that we installed more new machinery in 1907 than in 
1906. 



80 TRANSMISSION OF POWER BY LEATHER BELTING 

6 Increased efficiency of machines, due to the fact that the 
tensions are maintained much more uniformly than for- 
merly. 

c Continuous production by both men and machines, due to 
decreased interference due to belt-breakdowns. 

d Uniform type of belt lacing, decreasing danger to employees. 

e Decreased cost of maintenance. 

f Under the present plan the cost of maintenance appears as 
a separate item where it can be watched and compared 
with that of previous periods to determine the relative 
economies, while under the old plan the figures were 
combined with a mass of others so as to make it impos- 
sible to determine how much it had cost. 



CONCLUSION 

18 When we first commenced to install the new system we had 
all sorts of trouble as is generally the case with any new thing. 
The plan was opposed by foremen, gang bosses and workmen, each 
of whom had an idea that the new tensions were entirely wrong, and 
that the machines would never do the work properly, unless they 
could adjust the belting according to their individual ideas. One 
of the best evidences of the value of the present plan is that this 
antagonism has entirely disappeared, and what was at first con- 
sidered by many an interference and a hindrance is now accepted as 
a help and is believed to be entirely satisfactory by those competent 
to hold an opinion. 

Mr. Taylor. The original experiments at the Midvale Steel 
Works were started in 1884; 17 years later, when all^the machinery 
in that shop was taken out, one of the belts, which was of the type 
of those run under proper rules, that is, approximately the low tension 
suggested by Mr. Barth, had run all that time night and day under 
heavy tension. During this time it had required tightening only nine 
times, and at the end of the equivalent of 34 years of ten-hour service 
that belt came off its pulleys and was immediately put to work on 
another machine, in good condition. This instance of the life of a 
belt properly taken care of and properly tightened will be a sur- 
prise to the man accustomed to see a belt go out of use in from two 
to five years. This statement has just been determined. 



TRANSMISSION OF POWER BY LEATHER BELTING 81 

D WIGHT V. Merrick.^ As I am interested in chain drives, I 
will draw attention to some experiments made by Hans Renold, 
Ltd., of Manchester, England, and embodied in a pamphlet issued 
May 1908, comparing the relative efficiency of chain and belt drives 
on automatic machines. Mr. Renold claims that with the chain 
drive the output was increased 20 per cent, fewer drills and parting 
tools were used, and a better finish was obtained on the work. He 
says: " The tool did its work unflinchingly at every part of the revolu- 
tion of the spindle — no more and no less." He further states that the 
wear and tear on the spindle and countershaft bearing was consider- 
ably reduced. These statements were so striking that the Link- 
Belt Company, with which I am associated, decided to make further 
tests. In one of these which I was detailed to make I maintained a 
constant feed and speed and used the same tools with each drive, 
and in all cases the tool was used until it became necessary to re-grind, 
the object being to cut off as many pieces or drill as many holes as 
possible before this condition was reached. The tool when chain- 
driven did considerably more work than when belt-driven. I quote 
from my report as follows: 

2 These tests were made on a 3 in. by 36 in. Jones and Lamson 
turret lathe, with "blue chip steel" cutting-off tool {% in. wide, cutting 
off cold-rolled shafting 2^ in. diameter, feed 0.012 in. per revolution. 

3 Care was taken in forging, treating and grinding the several 
tools used, to insure uniformity in their cutting qualities; but to 
obviate the possibility of the results being affected by the cutting 
qualities of the different tools, each tool was used with both drives. 

4 One of the tools cut off 16 pieces when chain-driven before it 
became necessary to re-grind, and only 9 pieces when belt-driven. 
The cutting speed in both cases was 94 ft. per minute, feed 0.012 in. 
per revolution, and another tool cut off 8 pieces when chain-driven 
against 5 when belt-driven. In this latter case the cutting speed was 
130 ft. per minute, feed 0.012 in. for chain and belt. 

5 As the cutting periods in the above test were so short, two more 
series of tests were made with longer continuous periods. These tests 
were made on a drill press \vith new | in. carbon steel drills in a soft 
cast-iron block, 3 in, thick. The same drill was used on both drives, 
and was carefully and uniformly ground for each test. 

6 One of the drills when belt-driven drilled 31 holes before it 
became necessary to re-grind, but when chain-driven the same drill 
drilled 57 holes, the cutting speed in both cases was 62 ft. per minute, 

1 Dwight V. Merrick, Link-Belt Mfg. Co., Nicetown, Philadelphia, Pa. 



82 



TRANSMISSION OF POWER BY LEATHER BELTING 



TABLE 1 

Results of Experiments om a 3-in. by 36-in. Jones & Lamson Turret Lathe 0.012 in 

Feed per Revolution. 



^ 






i 


1 








M 

a o 


Kind of 


No. OF 


Metal Cut 


Cutting Speed , 


Condition 


Time 


R.p.M. of 


o 


Drive 


Pieces 


BY Tool 


Ft. per Minute 


op Tool 


Minutes 


Spindle 






Inches 












Beit 


6i 


8.125 


94 


Rmned 


5.91 


143 


2 


Ch^n 


16 


20. 


94 


Ruined 


14.70 


143 




Belt 


9 


11.25 


94 


Ruined 


7.15 


143 




Chain 


7i 


9.843 


128 


Ruined 


4.81 


196 


5 


Belt 


4i 


6.093 


134 


Ruined 


2.72 


203 




Belt 


1 


1.25 


134 


Good 


0.51 


203 


4 


Belt 


i 


0.312 


151 


Ruined 


0.12 


231 


Chain 


1 


1.25 


126 


Good 


0.54 


193 




Chain 


i 


0.937 


151 


Ruined 


0.39 


231 




Belt 


1 


1.25 


129 


Good 


0.53 


197 


1 


Belt 


i 


0.625 


146 


Ruined 


0.15 


223 




Chain 


1 


1.25 


129 


Good 


0.53 


197 


3 


Chain 


1 


1.25 


149 


Fair 


0.44 


228 




Chain 


f 


0.468 


195 


Ruined 


0.15 


299 



Note: A higher cutting speed was obtained by the chain drive. 



TABLE 2 

REfsuLTS of Experiments on a Drill Press with New J-in. Diameter Carbon Steel 

Drills, 0.018 in. Feed per Revolution, in a Soft Cast- Iron Block, 3-in. Thick 



>^ 










Condition 






o 


Kind of 
Drive 


Holes 
Drilled 
Number 


Metal Cut 

BT Drill 

Inches 


Cutting Speed 

R.P.M. 


OF Drill 

AFTER 

Drilling 
Holes 


Time 

Minutes 


R.p.M. or 
Spindlb 


1 


Belt 


31 


93 


62.2 


Started to 


18.91 


273 


1 


Chain 


57 


171 


60.5 


nun 

Comer 

rounded, 

needed 

grinding, 

starting to 

ruin 


35.91 


264 


2 


Chain 


37 


111 


62.2 


Starting to 


22.57 


273 


2 


Belt 


20 


60 


62.2 


ruin 
Starting to 
ruin 


12.20 


273 



Note: A great many more holes were drilled by the chain drive. 



TRANSMISSION OF POWER BY LEATHER BELTING 



83 



TABLE 3 
Results of Experiments on the same Drill Press as in Table 2 with New 1;]-in. Diam_ 
ETER Carbon Steel Drills, 0.018 in . Feed per Revolption, in a vest hard Cast. 
Iron Block, 3 in. Thick 



1^ 




Kind of 
Drive 


Holes 

Drilled 
Ndmbicr 


Metal Cot 

BY Drill 

Inches 


Cdttinq Speed 

R.P.M. 


Condition 
OF Drill 

AFTER 

Drilling 


Time 
Minutes 


R.P.M. OF 

Spindle 










Holes 






3 


Chain 


17 


51 


28.0 


Starting to 


19.84 


148.8 


3 


Belt 


14 


42 


28.4 


ruin 


15.40 


151.4 


4 


Chain 


17 


49i 


28.0 


Started to 

run on 17 th 

hole 1^ in. 

deep 


18.48 


148.8 


4 


Belt 


13 


39 


28.2 


Started to 
niin 


14.44 


150.0 



Note: More holes were drilled by the chain drive, but the percentage of gain was not 
anywhere near as great as in Table 2. 

feed 0.018 in., and another drill at the same speed and feed drilled 
37 holes when chain-driven, against 20 when belt-driven. 

7 The other series of drill tests was made on the same drill 
press, with ff in. carbon steel drills on a very hard cast-iron block, 
3 in. thick. One of the drills when chain-driven drilled 17 holes 
before it became necessary to re-grind, against 14 holes when belt- 
driven; the cutting speed in both cases was 28 ft. per minute, feed 
0.018 in. per revolution, and another drill did 17 holes, chain-driven, 
against 13 belt-driven, same feed and speed as above. 

8 The results were so gratifying that further tests are being made 
on four similar automatic machines, at our plant in Indianapolis, 
two fitted with belt drives and two with chain drives. The same feeds 
and speeds will be maintained with each drive throughout the series 
of tests, but a variety of tests will be made to establish the maximum 
efficiency of both belt and chain drives, to the best of our ability. 
The results will all be tabulated and published in a pamphlet in the 
near future. 

9 The accompanying tables contain the tabulated results of my 
experiments. 



F. A. Waldron, After listening to this paper, one naturally 
asks the question, What is its commercial value? Mr. Allen has 
answered this very well, but I will give a little of my own experience 
with the system. 



84 TRANSMISSION OP POWER BY LEATHER BELTING 

2 At the plant of the Yale & Towne Company, most of the 
responsibility for the condition of belts, prior to the author's going 
there, was placed with me and I am willing to take any criticisms. 
I became an prdent advocate of Mr. Earth's work on belts, however, 
particularly because of the practical results obtained. 

3 After leaving the Yale & Towne Company, I had occasion to 
purchase a Barth bench and spring-balance and apply the elements 
of the system without spending a large amount in replacing counter- 
shafts. I established the system of varying tensions on different 
machines. A light countershaft would not stand as heavy tension 
on the belt as the author originally prescribed. Tensions on belting, 
lengths, taking up, etc., were recorded. A record of complaints 
received in the millwright department for a specified number of coun- 
tershafts and machine belts had been kept, and for ten days or two 
weeks before installation something like 150 complaints came in. 
After complete installation of the Barth bench and scales and the 
Barth system, the complaints dropped to 80 for two weeks, and six 
weeks later to 35, showing the commercial results of systematic care 
of belts. 

4 Belts as low as 1^ in, wide, and some heavy double belts three 
to four inches wide, were the limits on size. 

5 This system was installed almost at the cost of my reputation, 
and on leaving that concern I supposed that the belt bench and 
bench-scales would be relegated to the scrap heap. Having an 
opportunity to put in a belt-bench and scales elsewhere, however, 
I wrote the firm asking if they did not want to sell the bench and 
scale and they said "no." 

A. A. Gary. I was much interested in Mr. Allen's remarks 
concerning the employment of Mr. Barth's system and formulae for 
the selection and proper application of belts to drive the numerous 
machines at the Yale & Towne plant, but explanation of one essential 
point is needed to show how this can be practically accomplished. 

2 As I understand, one important factor required in the formula 
used to determine the proper initial tension to which each belt must 
be subjected when put in place, is the horse-power to be transmitted 
by that belt. It has been stated that 4000 belts are used in this 
plant, operating perhaps one-half that number of machines. I would 
like to know the method employed to determine the power require- 
ments of each of these macliines so as to obtain the required value of 
this factor when the formulae are used. 



Q TRANSMISSION OF POWER BY LEATHER BELTING 85 

3 If we merely guess at the power required, we depart from the 
exact scientific method of determining information in our belt prob- 
lems and recede toward the "rule of thumb" method, as a formula 
is no more exact than is the value of the most uncertain quantity 
employed in its solution. If Mr. Barth can give us any "short cut" 
method for determining the power required by machines to be driven 
by belts, he will funish information that will give his formulae a very 
practical value. 

A. F. Nagle. This paper does not pretend to present new 
facts, but sets forth, in mathematical formulae and diagrams, data 
obtained by, .Messrs. Lewis, Taylor and others. It also diagrams 
some simple arithmetical computations. As a work of mathematical 
study and diagrammatic representation, the paper is admirable, but as 
a practical aid to a busy engineer, it seems to me too compUcated. 
The only part which holds my attention is the diagram in Fig. 3.. 
giving the horse power of belts at different velocities, and of two types 
spoken of here as countershaft and as main di'ive belts, but more 
commonl}^ designated as "single and double thickness." The reason 
for this distinction is that while the stress in the net solid body of the 
leather is taken to be the same in each case, in "single" belts the 
joint is a butt joint and is laced. This cuts away more of the belt 
than where the belt is of double thickness, lapped and cemented, or 
riveted; the difference being in the character of the joint rather than 
in the strength of the belt. 

2 For comparison then, we can take Mr. Earth's estimate of the 
relative strength of these belts, as 160 to 240 or 1.0 to 1.50. Mr. 
Towne found these to be as 1.0 to 1.82, and in my early studies, I was 
inclined to adopt this ratio; later, however, I have used the ratio of 
275 to 400 or 1.0 to 1.45. 

3 The belt problem is very far from being one of pure mathematics. 
As in most engineering problems, there is about 5 per cent of scientific 
knowledge involved, and fully 95 per cent of good judgment based 
upon experience. We rarely know the exact power to be transmitted 
except in the case of prime movers. The arc of contact, the velocity, 
and the stress we are willing to put upon the leather, are all easily 
determined, but we cannot decide upon the coefficient of friction by 
formula. A new leather belt upon an iron pulley may not have a 
coefficient of friction of 25 per cent, while the same belt, well worn and 
well groomed, will give 65 per cent in a clean, dry room ; put the same 
belt in a wet place, hke a tannery, or a dusty place, Uke a stone- 
crushing plant, and we have an entirely different coeflacient. 



86 TEANSMISSION OF POWER BY LEATHER BELIING 

4 It seems to me that the designing engineer, even though he 
understands the mathematics of the belt problem, if ignorant or 
unappreciative of the practical conditions under which the belt works, 
will be liable to make a mistake. On the other hand the engineer 
famiUar with the conditions, but ignorant of the mathematics involved, 
is also liable to error in his conclusions. A cautious man will endea- 
vor to err on the safe side, feeHng no doubt as did our venerable ex- 
President John Fritz, who when remonstrated with for making 
some machinery needlessly strong, replied, " If I do, nobody will ever 
find it out." 

5 On general principles, it is of course desirable to work belts, 
like other members of a machine, with large coefficients of safety, but 
engineering in its last analysis is a question of finance and we must 
"hew as close to the line" as possible. Mr. Towne found the ulti- 
mate strength of laced belts to be 200 lb. per inch width {-h in.) thick, 
and used ^ of this, or 66§ lb. as a safe working stress. Mr. Towne also 
found a coefficient of friction of 42 per cent to be safe. The general 
practice of the day has been quite close to these factors, but if I under- 
stand his diagrams correctly, Mr. Barth has departed far from them. 

6 In 1881 I read before the Society a paper giving for the first 
time, I believe, a belt formula which took cognizance of the effect of 
centrifugal force. The data used therein were based principally 
upon Mr. Towne's experiments. The results obtained were well 
within the safe limits of previous practice for low speeds, but at high 
speeds my formula showed the deviation. Common formulae gave 
results (see Kent, Mech. Eng. Pocket Book, p. 879) as follows: 

For single belt 1 in. wide, 
Velocity (1) (2) (3) (4) Nagle 

600 ft. per min. 1.09 h.p. 0.55 h.p. 0.60 h.p. 0.82 h.p. 0.73 h.p. 
Barth gives only 0.40 h.p. 

For double belt, 

Common Formula Nagle Barth 

1.17 h p. . 1.24 h.p. 0.68 h.p. 

7 For the purpose of giving a clear conception of Mr. Barth's 
deviation from the others, I repeat my formula here: 

C.V.tw{S - 0.012 V^) 



h.p. = 



550 



C is a constant expressing the adhesion of the belt upon the pulley 
under a unit of stress of belt. Its value is expressed by the equation 



TRANSMISSION OF POWER BY LEATHER BELTING 



87 



C = 1 — 10 000''58/a ^vi^gre a is the arc of contact and/ the coefficient 
of friction. The other quantities are as follows: 

V = velocity in feet per second. 

5 = stress upon leather per square inch, which I have taken 

at 275 lb. for laced and 400 lb. for riveted belts. 
t and w are the thickness and width respectively in inches. 
550 ft. lb. = horse power per sec. 



14 
13 


























































































































12 
11 
10 

9 

1 

4 








































^ 


































,»' 






































— 






























/^ 


^ ^ 


*sl^ 


— 




























^ 


;^ 


^ 












I' 






















^ 


r 


















5 


















^ 


^ 














00*2 


oo 


'^ 














^ 


^ 


y 
















- — ' 1 


Nagle 




4 


__-5 


s^"^ 






















/ 


t^ 








:^ 


-"^ — -^Bar 


tb-(D 


>able 










3 


^ 


- 








^ 


^ 


^^ 


^^ 


>^ 






















Si 






A 


s^ ^^ 




























1 


>*^li5*^^^^ 


__J_ — ^ ■ 














J 


^fn^ 


1 





























10 


20 


30 


40 50 60 


70 


80 


90 


100 


600 


1200 


1800 


Feet per Second 

2100 3000 3600 

Feet per Minute 


4200 


4800 


5400 


GOOO 



Fig. 1 Comparison op Different Belt Formula, Based Upon Belts 1 

In. Wide and ^ In. Thfck for Single and \ In. Thick 

FOR Double Belts 

8 To illustrate the solving of this equation, let a = 180 deg. an 
/ = 0.40, then 

180 X 0.40 X 0.00758 = 0.54576 
10-0.54576 _ iQg iQ y^ 0.54576 = 1 X 0.54576 

0.54576 is a logarithm of which 3.513 is the number. This being a 
minus — coefficient, we must take its reciprocal or 0.284; subtracting 
this from 1, we get 0.716. The result could have been obtained by 
subtracting the log 0.54576 from 1, giving 1.45424, and this gives 
0.2846 as its number direct. 



88 TRANSMISSION OF POWER BY LEATHER BELTING 

9 In Kent's Mechanical Engineering Pocket Book, p. 878, tables 
are given based on this formula, which facilitate its use. I wish to 
call attention to the wide divergence of Mr. Earth's conclusions from 
those commonly used as well as from my own, as plotted in Fig. 1 
herewith. I have reduced his figures to the same thickness as mine, 
namely -h in. for single and \ for double. 

10 This work has been done somewhat hastily and I hope the 
author will check it at least so far as relates so the interpretation of 
his diagram. If my work is correct, I am puzzled to understand why 
his tables of belt horse power differ so much from mine. 

Prof. Wm. W. Bird. I feel very much pleased and highly compli- 
mented to see the results of Mr. Barth's mathematical analysis of 
ray experiments on belt creep. On a few points, however, I am still 
in doubt in regard to his general conclusion. I believe: 

a That the elasticity of a belt varies with the velocity and 
that at very slow speeds the sum of the tensions would 
remain constant, while at high speeds, if it were not for the 
centrifugal force, the sum would increase] practically 
with the load. If this is true, a belt will improve with 
higher speeds and will not reach a maximum at 4000 ft. 
per minute as shown in Fig. 3. 

h That in determining the size of a belt for a given load, the 
diameter of the smaller pulley should be considered. A 
belt will do relatively less on a small pulley than on a large 
one, other conditions being the same. 

c That the crowning of the pulleys should be considered as it 
affects the life of the belt. 

d That the carrying capacity of a belt should not be given in 
terms of square inches of cross section, as a double belt 
with one square inch of cross section will not transmit as 
much as a single belt of the same cross section. 

2 I have recently made some tests on compound or rider belts 
and have been somewhat surprised at the relative movements of the 
main and rider belts with various pulley ratios, from which I have 
concluded that if these belts were glued together, as they would be in 
a double belt, numerous internal stresses and strains must be set up 
when the belt passes over a pulley, especially over a small one. 

3 The crown is a very serious matter on a small pulley, as the 
following figures will show. Take a 4 in. pulley with a crown of 0.2 
in. in the diameter, if the belt wraps 180 deg., the length in the 



TRANSMISSION OF POWER BY LEATHER BELTING 89 

middle of the belt will be 0.31 in. greater than on the sides; this means 
a stretch of 0.05 in. per inch or 1000 lb. stress on middle fiber, taking 
modulus of elasticity as 20,000 lb. The belt must slip or be ruined, 
for this stress does not include load or initial tension and is in itself 
enough to stretch the belt beyond the elastic limit. The slipping 
necessary to adjust this stress must influence the friction and slipping 
of the belt as a whole. 

4 I would like to have Mr. Barth answer a question which I have 
been asked a great many times, — why does the sum of the tensions in a 
belt increase with the load? I would also like to have him calculate 
with his slide-rule the size of 'a belt for [the following conditions: 
20-h.p. motor, 6-ih. pulley, 1200 r.p.m., to drive a shaft 12 ft. away 
at 200. Would the same belt last as long if the drive were reversed, 
that is, a shaft running at 200 r.p.m. driving a generator with 6 in. 
pulley at 1200? I would like also to ask Mr. Barth or any engineer 
present whether he would use the same size belt with the 6 in. pulley 
as a driver as with it as a driven, and with the same size of belt; and 
whether in this case it would last longer, other conditions being equal. 

5 Anyone who has undertaken an investigation of the belt prob- 
lem knows that it is almost impossible to keep conditions constant — 
humidity, oil in the belt, surface of pulley, etc., seem to change with- 
out notice and complicate the work. 

6 I wish to congratulate Mr. Barth on his efforts to advance the 
theory of the transmission of power by leather belting, and to agree 
with Mr. Lewis in the conclusion of his paper presented in 1886, 
"That there is still need of more light on the subject." 

Prof. C. H. Benjamin. I have been asked to contribute to the 
discussion of Mr. Barth's paper; technically, I am afraid I can not 
criticise it or add to it, for it leaves but little more to be said. Sen- 
timentally, I can not but regret the gradual disappearance of our 
terra incognita, both geographical and mechanical. Time was when 
large areas on the map bore the encouraging legend "Unexplored 
Wilderness" or "Great American Desert" and left room for the free 
play of the imagination. Today you miss those fascinating areas 
and are tied down to realities. 

2 Not many years ago, the grinding of a lathe tool was an interest- 
ing experiment, attended with much uncertainty, and the matter 
of feeds and speeds offered an alluring field for investigation. Mr. 
Taylor has spoiled all that for us and now our imagination is worked 
by slide-rule. 



90 TRANSMISSION OF POWER BY LEATHER BELTING 

3 Time was when the possibilities of belting were vague in outline 
and when coefficient of friction, slack tension and belt creep were 
rather shadowy phantoms. It was pleasant then to speculate on 
what the belt would do and how long it would do it and the man with 
the longest memory had the advantage. But now, Mr. Lewis, Mr. 
Bird and Mr. Barth have taken all the romance out of it and another 
illusion succumbs to the deadly aim of the slide-rule. 

4 Perhaps I take a malicious pleasure in noting that one or more 
factors of the problem are still out of harness and a trifle intangible. 
Our old friend, the coefficient of friction, is in hiding under the belt 
sporting with those other elusive fairies, modulus of elasticity and 
belt-creep. After all, what does it matter? Aside from the interesting 
theoretical questions involved, what we need to know is, first, how 
wide a belt to use at a certain speed to transmit a certain power, — 
Mr. Taylor has answered this question. Second, how tight to lace 
or cement that belt that it may do the work for a reasonable time 
without relacing, — Mr. Barth has told us that. 

5 I began experimenting on belts 25 years ago and have been at it 
more or less since. With a fixed pulley and a slipping belt, I found no 
difficulty in proving <f> = 0.42 after Rankine, but when I built a 
belt machine and tested belts under running conditions, 4> lost all 
its constancy and might as well have been called x. Working back- 
wards from the measured tension and using the old formula, I found 
^ to vary with the load, the speed, the kind of pulley, the age of the 
belt, the weather and the dominant political party — in fine, to be so 
mysterious and intangible a quantity as to be useless for practical 
purposes. 

6 The sum of the tensions also varied in a manner that did not 
admit of rational explanation. And right here let me say that the 
reasons for Mr. Barth's assumption of constancy for (t^ + ^ Q are 
hardly clear to me. Why call that constant which is not constant? 
Why call anything constant except as it is shown to be so by measure- 
ment? This is not said in criticism but for the sake of information. 

7 There is one aspect of the paper that deserves special attention 
and that is the recognition of the fact that a belt is an elastic connector 
with a variable length and variable tensions. Most writers on the 
subject have treated belting as if it were a non-extensible element 
which could be exactly represented on paper and whose behavior 
was capable of exact mathematical analysis. A belt in action is 
almost like a thing alive, squirming, lengthening, shortening, its 
tension changing back and forth with a variable modulus of elasticity 



TRANSMISSION OF* POWER BY LEATHER BELTING 91 

and a lag in its changes due to its contact with the pulley and the 
short time intei-vals. A belt must be tested to be appreciated and 
theory must wait upon experiment. 

8 I fully appreciate the value of Mr. Earth's analysis and can see 
that his methods will result in marked economies in establishments 
where many large belts are used and where conditions are pre-deter- 
mined. In the smaller shop, where conditions vary, and in isolated 
cases with differing sorts of pulleys, differing kinds of belt, new and 
old, I feel that each case will have to be settled on its own merits. 
Until more experiments are recorded, the average machine-designer 
or millwright will have to be guided largely by his own judgment and 
experience in determining the width and tension of each belt. Let 
us have more experiments. 

H. K. Hathaway.^ To the scientist and machine-tool designer 
the value of Mr. Earth's paper will unquestionably be immediately 
apparent, but the writer feels that the paper does not represent with 
sufficient clearness features of the problem that are of inestimable 
value to the engineer concerned with running a shop. Assuming 
that the designer takes care of the sizes of belts required, and the 
speed at which they should be run, in accordance with the conclusions 
of such eminent authorities as Mr. Taylor, Mr. Earth, and Mr. Lewis, 
a great deal is lost unless the shop-man properly cares for the mainte- 
nance of such belts. As Mr. Earth has pointed out, the care and 
maintenance of belting in the great majority of shops is done by rule 
of thumb, and left entirely to the judgment of the shop millwright or 
the workman operating the machine. 

2 The efficiency of a belt-driven machine largely depends upon the 
tension of the belts being properly maintained at a point above the 
minimum initial tension at which they will transmit the power 
required. This point Mr. Earth has only slightly touched on, whereas 
the writer feels that this subject should have occupied a section fully 
as large as the body of the paper presented. 

3 If a machine stands idle during working hours while the belt 
is being repaired or tightened it produces nothing during that time, 
and there is a distinct loss to the manufacturer. If a machine stands 
idle for one-half hour out of ten hours working time there is a loss 
of 5 per cent in the output of that machine, and if in a shop having 
100 machines, 10 machines out of the 100 lose one-half hour each day 

I H. K. Hathaway, Tabor Mfg. Co.. Philadelphia, Pa. 



92 TRANSMISSION OF POWER BY LEATHER BELTING 

on account of repairs to belts it amounts to a loss of 0.5 per cent 
on the total output of the shop. This feature, however, is probably 
not so bad as the loss in output due to the machine belts being run so 
loose that they cannot begin to take the feeds, speeds, and depths of 
cut for which the machines are designed and that the tools will stand. 

4 The writer has had considerable experience with the system of 
maintenance of belting mentioned in Par. 67 of Mr. Earth's paper, and 
will describe it briefly. 

5 Almost every foreman or superintendent, in attempting to 
bring up the speeds of his machines to something like what he knows 
to be possible, has found that such attempts usually result in the 
belt's slipping or breaking, or the lacing giving out, and knows that 
where the care of belts is left to ^the man on the machine, only in a 
very few cases can the belts be depended upon to do the maximum 
amount of work. If, therefore, the maximum feed, speed and depth 
of cut are to be prescribed and used, as is done by the aid of Mr. 
Earth's shde-rules under the Taylor system, it is essential that belts 
of the best quality and of the proper proportions be used, and that 
they be kept in first-class condition and at the proper tension, so that 
they can be reUed upon to give the pull required. It is also necessary 
that all repairing, tightening, and inspection of belts be done outside 
of working hours that there may be no loss of output from interrup- 
tion to manufacture. In order to accomplish these objects the follow- 
ing system has been evolved. 

6 A record is kept for each belt in the shop on the form shown 
as Fig. 1, on which are given all standard data for each belt in ques- 
tion. 

7 When a new belt is to be put on, or an old belt to be inspected 
or tightened, the special belt fixer's bench developed by Mr. Gulowsen 
is used, together with the belt-tension scales referred to by Mr. 
Earth. With this apparatus it is possible for one man to remove, 
tighten and replace almost any belt in from six to eighteen minutes. 
In putting on a new belt, or tightening an old one, the drums or 
pulleys on the belt bench are set by means of a steel tape to corre- 
spond with the distance over the actual pulleys, as previously deter- 
mined, and shown on the belting record as "Length over Pulleys." 
A roll of belting, of the proper width and thickness, is next placed in 
the open drum, and passed through one pair of clamps of the belt 
scales around the drums or pulleys and through the other pair of 
clamps of the belt scales. The clamps are then tightened on the 
belt and the belt drawn up by means of the screws until the spring 



TRANSMISSION OF POWER BY LEATHER BELTING 



93 









a 





a 
o 






: . S 




o 






< 
a 

z 

o 


z z 






: 










• I <r 




a 
















* H § M 




H 
















" z s ;:; M 




O 


1 














■< S P 




a 

n 
















SH go^ 














: § 


rt 




















a 


1 .<o 






> 


i > 




^ 




■ 1 


z 
•< a 

a s 

« •< 






1 c 
I 


c 

c 




1 

c 

s 


1 


■3 ^ 
S c 

^ 

E- 


ii ^ 




MARKS 

Done 

ONDI- 
r, ETC. 




















a w :a 




















P^ S - " 




h 
















i^«z 

« § g 




Z 


















g 


















Pi 

B. 
















3 s 
























Date 
Dressing 

PUT ON 

During 

Service 






J) 








' 




g 

c 






Date 

TO BE 

Cleaned 

AND 

Greased 
Again 




n 


"i 

PM 
> 




9 


c 


1 

(V 

C 

Q 


C 
g 


5 


• 








5 

1 

c 
o 




2 

c 




1 

& 




a 

i 

"5 


C 

*« 


Date 

leaned 

AND 

REASED 






h^l 


H 


c 


a 


K 


« 







sis g S 

g S g a S < 




















;; z 2 « 5 




















Q«« H< 




" g '^ . 




















D a H 




'! S H 


to 

a 




















l| Z U Q 


!S "3 




















5 ,-5 « 


5 S 




















i «° W z 


is -a 

< M 




















|i "^ ^ 3 




















^ « 5 


E-i 




















!l 2 oPq 


























a 


M 




















' §^i 


£•1 




















a '^ * 


s 












£ 


c 



1 


■3 


en w B, 






;: i 






1 


§ 




•3 




3 

Pi 


"5 
H 
g; 










o 


a 




c3 


■3 


* 








J 




s 


s 


Q 


Q 


3 


a 





94 TRANSMISSION OF POWER BY LEATHER BELTING 

balances between the two pairs of clamps record the tension required, 
after which the belt is cut off so that the two ends will come together, 
and the belt is laced on a belt-lacing machine and put on its pulleys. 

8 A memorandum, which also serves as the belt fixer's order and 
time card, giving him all necessary instruction, is then placed in what 
is called the "tickler," a portfolio having a compartment for each 
day of the year, under the date on which the belt will probably 
require re-tightening, and on that day it will be removed from the 
tickler, together with the memoranda for any other belts requiring 
attention, and sent to the belt fixer for attention during the noon 
hour and after quitting-time. 

9 These belts are then removed from their pulleys, taken to the 
belt bench and tested to ascertain whether they require tightening; 
if the tension is found to have fallen to approximately the minimum, 
they are drawn up to the maximum tension as previously described, 
a piece is cut from one end, the belt is re-laced]^and put back in 
place and these facts are noted on the belt fixer's memorandum, 
which is then returned to the planning department, and entered on the 
belt record; and a new memorandum placed in the tickler under the 
date on which the belt will again require attention. Notices for 
scraping, cleaning and greasing the belts at proper intervals are also 
placed in the tickler. 

10 The length of time a belt will run before the tension will 
fall to the minimum at which it will pull all that is required, has been 
determined from experiments, and a belt seldom requires attention 
before the time set for re-tightening; when this does occur, however, 
a belt-dressing which does not injure the belt, but which will enable 
it to pull properly until noon or the end of the day, is applied, and the 
memorandum is removed from the tickler and another placed under 
its next date for re-tightening. 

11 The system described accomplishes four things of vital impor- 
tance to economical production: 

a Freedom from interruption to production from having 
to repair belts during working hours, by having all belts 
systematically inspected and all breakdown and shppage 
anticipated and prevented before they occur. 

h Possibility of using the maximum feeds, speeds and depths 
of cuts at all times. 

c Increase in life of the belt owing to all belts being of the 
proper dimensions and properly laced and spliced and 
run at the proper tension. 

d Reduction of cost of maintenance to a minimum. 



TRANSMISSION OP POWER BY LEATHER BELTING 



95 



12 Mr. Earth's belting slide-rule is used in determining the 
dimensions of the belts, the maximum and minimum tensions. The 
writer can speak from experience of the great value of the belting 
slide-rule in solving the belting problems that confront the shop 
engineer, and while the mathematical features of Mr. Earth's paper 



Oui 
In 




Order Number 
D L 


Departi 
Day Ra 


nent 

te 






Man's Time 








Max. Tension 


Min. Tension 


Belt Symbol 




Cleaned and Grea 


sed 








Dressed While in 


Use 

It 


Dressing Used 

Length Put In 




Amount Taken 0\ 




Length of Splice. 




Cemftnt TTspd 


Tension in Lbs. 


Indicated by 


' Before 
After. . 


Tightening 


Each Spring Balance^ > 




Workman's Name 


















Entered in 




Pay 

Sheet 


Cost 
Sheet 


Belt 
Record 




T-. TT,^ time 














UAT WORK 

note 



Fig. 2 Belt Fixer's Order and Time Card 



are unquestionably interesting to many, the writer feels that, like 
himself, many will be glad to accept Mr. Earth's figures without 
question provided they can have the slide-rule. 

13 It is a fact that in the average shop very few belts become unfit 
for use through legitimate wear, but rather through accidents or 
improper care. Where the care of the belts is left to the workman, 
the belts are usually far too loose, and when a belt slips it is less 



96 TRANSMISSION OF POWER BY LEATHER BELTING 

trouble for the workman to reduce his speed, feed, or depth of cut, 
or as a last resort to use rosin to make the belt pull. This use of rosin 
will ruin any belt in a very short time. 

14 Very few machinists or even foremen know how to tighten or 
lace a belt properly, the amount to be taken out being usually 
guessed at, and a great deal of time is lost through the machine's 
standing idle while the cutting and trying is going on. The writer 
has seen a good machinist run a cone belt, which he had made too 
tight, on "cross cones," i.e., on steps not in line with each other, 
with the result that it twisted itself up like a corkscrew and was 
practically ruined. 

15 Another cause of premature ruin of belts is improper lacing, the 
ends not being cut square and the lacing on one side stretching more 
than the other, causing the belt to run crooked. 

16 Cemented splices, when properly made, give the best results. 
Machine lacing, using a spiral wire lacing, while not so good as a 
cemented splice, is very satisfactory, however, and more convenient, 
and takes less time for putting on and taldng off belts for the purpose 
of testing and tightening on the belt bench. A belt joined by a 
cemented splice must be tested and spliced in position, which is not 
so convenient as on the belt bench, especially in the case of over- 
head belts. Even where cemented splices are used the belt bench is 
convenient for cutting new belts, or re-tightening to a length giving 
the proper tension, and for repairs. Only one wire joint is used 
in any belt, splices being made if a section becomes damaged so that 
a new piece must be set in. The average belt if cared for under this 
system will last from six to eight years. 

17 The tension on a new belt falls very rapidly, and our present 
practice is to tighten it after 24 hours, then in 48 hours, then in one 
week, then in two weeks, and so on doubling the length of intervals 
until it gets to three months; from this point we must ascertain by 
trial for each belt how much greater the intervals may be. This of 
course depends upon the severity of service the belt is called upon 
to perform as well as the quality of the belt. 

Prof. Wm. S. Aldrich. In the first place, the academic discus- 
sion of the constancy of the sum of the belt tensions under all loads 
is finally set at rest. Now that we really know what is what, by the 
invaluable series of experiments referred to, the wonder is that this 
fallacy of the constancy of the sum of the belt tensions is so per- 
sistent. 



TRANSMISSION OF POWER BY LEATHER HELTING 97 

2 It was doubt of this position that led the writer to analyze 
for himself the experiments on belting then available, those of 
Wilfred Lewis and J. S. Bancroft, undertaken for Wm. Sellei-s & Co., 
and of Professor Lanza, of the Massachusetts Institute of Technology. 
Both of these were recorded in papers read before the Society, and 
published in Vol. 7 of the Transactions. It is remarkable that these 
classic experiments have been before the world thus long, and vet so 
little studied and respected, and, as far as the writer is awar(>, have 
not been superseded by experiments in their special field with more 
modern apparatus. Until they are superseded, Mr. Barth's con- 
clusions must stand, a remarkable instance of the deductive reasoning 
by which it would seem that engineering progress must be made. 

3 On the other hand, Mr. Barth has built up, in characteristic 
fashion, from theoretical considerations more or less influenced by a 
knowledge of the phenomena of belt-transmission, combined with the 
physical properties of belting, certain new and helpful relations that 
must govern in the future. Such is his " new theorem of the relations 
of the tension in a belt," that " under any variation of the effective 
pull of a belt, the sum of the square roots of the tensions in the two 
strands remains constant, as against the old fallacious supposition 
that the sum of these tensions remains constant." (Appendix, Par. 
26). Therefore, 

V7\+ VT^=2VT^ (1) 

4 Now, if we can obtain a similar relation for the difference of the 
square roots of the tensions; then we shall have at once, by the usual 
formula for the product of the sum and difference of two quantities, 
the difference of their squares; that is, in this case, the difference of 
the squares of the square roots of the tensions, which is the difference 
of the tensions, or the pulling power sought. 

5 This much needed "difference of square roots of tensions" has 
been indicated by Mr. Barth (Appendix Par. 14), "on the strength 
of the experiments made by Mr. Lewis and himself, namely, that 
within the limits of ordinary working tensions of a belt, the differ- 
ence between the lengths of a belt at different tensions is proportional 
to the difference between the square roots of those tensions." We 
thus have, 

L,-L,= K {VT, - V¥,) (2) 

in which K is a constant, dependent upon the material of the belt, 
and determined by experiment on the belt. 



98 DISCUSSION 

6 Combining with Equation 1, we have, as abeady indicated 

T,-T, = 2 Vt, (L, -L,) ^ (3) 

It seems to the writer that this might possibly be a helpful deduc- 
tion, though it may be without much practical appUcation; so that 
knowing the initial unit tension T^ and the lengths of the belt under 
the tensions T^ and Tj* together with the constant K, its pulling 
power {Ti — T^) is known. It seems, therefore, necessary to know 
the difference in the lengths of the belt, due to differences in the 
belt tensions, that is, to the different driving powers under which it 
is expected to operate the belt, or in other words, to calibrate the 
belt-performance for this use. 

7 It may be remarked, in passing, that the constant K is to be 
found from the experiments of Mr. Lewis, as analyzed by Mr. Barth 
(Appendix, Equation 3), 



^. = ^(1 + 864' W 

in which L^ equals the length of belt under the unit tension t when its 
slack length is L. From this, by analogy with the above Equation 
2, we have, 

«-8^ (^> 

8 It will no doubt appear that the writer is still inclined to let 
the arc of contact and the coefficient of friction of belts take care of 
themselves, notwithstanding the keen discussion that has centered 
about the fourth conclusion in his paper, referred to by Mr. Barth; 
namely, 

" (4) The ratio of the tensions of a belt-transmitting power 
cannot be calculated with any degree of accuracy by 
means of the well-known belt formula: 

T <f>a 

7f=e (6) 

involving the arc of contact a and the coefficient of fric 
tion 4>:' 

9 This relation is no doubt a guide and a help, indicating the 
way the ratios of belt tensions are most likely to be involved. But 



TRANSMISSION OF POWER BY LEATHER BELTING 99 

it certainly requires a radical modification to adapt it to any reliable 
use in predetermining the ratio of tensions for lacing up belts for 
given pulling power. Mr. Barth has wrought out these modifica- 
tions with excellent results, judged by the adaptabiUty of his slide 
rules, and the closeness of approximation to actual conditions (within 
the limits assigned) of the assumptions upon which they are based; 
namely (Par. 44), "that for the driving belt of a machine the mini- 
mum initial tension must be such that when the belt is doing the maxi- 
mum amount of work intended, the sum of the tension on the tight 
side of the belt and one-half the tension on the slack side will equal 240 lb. 
-per square inch of cross section for all belt speeds; and that for a belt 
driving a countershaft, or any other belt inconvenient to get at 
for re-tightening or more readily made of liberal dimensions, this 
sum will equal 160 lb." 

10 Here, then, is a definite acceptance of things as they are, and 
a straightforward assumption involving additive relations of belt 
tensions of leather belting, as it is made and used, and conformable to 
experience, rather than their ratios agreeable to theoretical formula, 
involving coefficient of friction and arc of contact. This latter rela- 
tion (6) is as elusive as the traction-coefficient in railroad work; and 
engineers probably will have their own opinions about each until 
some genius can predetermine what coefficients of friction are to be 
expected in every instance, and so properly introduce the friction 
for dynamic conditions into a formula based entirely upon a con- 
sideration of statical relations. 

The Author. In reading the unexpectedly numerous discussions 
of this paper, the author is pleased to note the general appreciation 
of it as a contribution to the literature of its kind, but regrets the 
assumption by two or three of the discussors that he considers the 
paper final in its application of the theories developed. All that is 
claimed is that he has taken practical advantage of the experimental 
data at his disposal, and has taken the pains to do mathematical jus- 
tice to them, deriving therefrom excellent results in the scientific 
running of machine tools whose belts have been tightened and 
worked according to the rules thus established. 

2 While the author feels guilty, therefore, of narrowing to a con- 
siderable extent the scope within which Professor Benjamin's imagi- 
nation may still run rampant, so far as the behavior of a leather belt 
goes, he fully agrees that further experiments are needed in order to 
determine the coefficient of friction under all the variable conditions 



100 DISCUSSION 

under which belts are called on to drive; and yet more, in order to 
settle conclusively whether the coefficient of friction is a function of 
the velocity of slip, as he has assumed, or of the percentage of slip, 
as indicated by Mr. Hamerstadt's study of numerous experiments at 
different belt speeds, though the latter seems contrary to the mechan- 
ical principles involved in the phenomenon of slip. 

3 For the special benefit of Professor Bird, the writer will even 
say, that while all he knows about belting could probably be reduced 
to a pamphlet three times the size of his paper, a good-sized volume 
would probably be required to hold all he does not know but would 
like to know about belting, and a small library would be required 
to record all he does not care a straw to know about the subject. 

4 But while the writer agrees with Mr. Hamerstadt as to the 
desirability of further experiments and will look forward to these 
with the keenest interest, he does not see the force of his argument 
about the necessary overload capacity of a high-speed belt, on a 
motor with an overload capacity; surely we need only make the belt 
b'g enough to take care of the overload as a normal load, and be 
satisfied to have it unnecessarily large for the rated capacity of the 
motor; just as a bridge intended for a light normal load must still be 
made strong enough for any anticipated occasional extra load. 
Trouble arises only when we do not know how to design a bridge 
properly, or v.hen we get an occasional extra load which we have had 
no reason to anticipate. 

5 Though the writer had not expected to be forced to express 
himself on the question of belt-drives versus electric motor drives, he 
will say, in view of Mr. Robbins' remarks, that he believes that during 
the past decade hundreds of thousands of dollars, if not millions, have 
been more than wasted by the substitution of motor-drives for belt 
drives. Such a change has often been advantageous, of course, and 
is occasionally recommended to his clients by the writer: the trouble 
has been that the enormous investments of electrical manufacturing 
establishments have forced the electrical salesman more than any 
other to create a demand for his product, so that not only has he 
allowed his enthusiasm to run away with him, but he frequently has 
recommended his product against his own biased judgment; persuad- 
ing the incompetent shop manager or superintendent to accept his 
product as a remedy for a small output that is in reality due to a 
complication of causes that could be cured only by the application of 
a number of remedial measures. 

6 The writer believes, however, that a reaction against this indis- 



TRANSMISSION OF POWER BY LEATHER BELTING 101 

criminate electrification of machine shops has already set in. aside from 
the influence of the industrial depression, and that the electric drive 
will be installed, in the near future, only when conditions make it 
unquestionably more advantageous than the belt-drive. 

7 As touched upon by Mr. Van Derhoef, the elastic properties of 
transmission-rope are probably similar to those of leather belts, and 
it seems to be in order for someone to ascertain them by the neces- 
sary experiments, and subsequently to apply this knowledge by use 
of the writer's methods. 

8 The writer values Mr. Allen's statements of the advantages 
derived from the adoption of the Taylor system, as introduced by the 
author in the Yale & Towne Mfg. Co. 's plant, where 4800 belts are thus 
taken care of. Mr. Allen, in conjunction with Messrs. Tajdor, Hath- 
away and Waldron, has thus supplemented the scant attention paid 
in the paper to the aspect of the subject most practically important. 
The reason for this omission is that in his work with belting the 
writer has derived by far the greatest personal satisfaction from the 
solution of the mathematical problems involved, and he has been 
unable to eliminate entirely the personal interest. 

9 It is not possible to answer here Mr. Gary's question as to how 
to estimate the horse-power required to drive each machine in a 
large plant, but the writer will be pleased to give him, and anybody 
else interested enough to pay a visit to Philadelphia, an idea of how it 
is done, by means of slide-rules especially constructed for the purpose. 

10 As a further answer to Professor Bird's various statements 
and questions, the writer will only say that on a more careful reading 
of the paper, as well as the Appendix and Supplement, he will find 
most of them answered. For instance, the most valuable mathemat- 
ical developments in the Appendix and Supplement answer the 
question why the sum of the tensions of a belt increases with the 
load; and study of this will help him to formulate for himself an 
answer to his non-mathematical questioners. As to the effect of 
crownings a mall pulley, the writer heartily agrees with Professor 
Bird in a general way, though surprised to note with what confi- 
dence the latter estimates the difference between the tensions in the 
middle and edge-fibers of a belt running over such a pulley. 

11. The author is sorry that the considerable trouble to which 
Mr. Nagle has gone to make comparisons with his own earlier formulae 
for the horse-power transmitted by leather belting, is based on a mis- 
understanding of the fundamental basis of the author's work. 

12 As stated in the paper, the author bases his figures on a certain 



102 DISCUSSION 

tension per square inch of belt, independent both of the strength of 
the belt itself and its thickness, and of the strength of the lacing, 
except that the latter must be in excess of the maximum tension 
brought to bear on a belt while delivering power. The author, there- 
fore, makes no distinction between a single and a double belt, but 
merely considers the tension per square inch of section, as it has not 
been definitely proven that the coefficient of friction depends materi- 
ally on the area of the surface presented by the belt against the pulley. 

13 As Mr. Nagle somehow has assumed that the two horse-power 
curves in Fig. 3 are meant respectively for a single and a double belt, 
whereas they stand for something totally different, it unfortunately 
follows that the comparisons made by him of his own and the author's 
ideas as to what power a belt should be counted on to transmit, have 
completely miscarried. 

14 Mr. Nagle says that we cannot decide upon the coefficient of 
friction by formula. This is unquestionably so, but it is also true 
that, having roughly decided, by one means or another, what we wish 
to count on as the coefficient of friction at any one velocity of a belt, 
we may to great advantage make an empirical formula to represent a 
perfectly-graded change of this coefficient with the velocity; and only 
by so doing can we effect a mathematical solution of the belt problem 
that is an improvement on the unquestionably wrong assumption of a 
coefficient independent of the velocity of the belt, such as 0.42, origin- 
ally recommended by Mr. Towne, or 0.28, recommended by the late 
Professor Ruleaux. 

15 The author fully agrees with Mr. Nagle that " a new belt on an 
iron pulley may not have a coefficient of friction of as much as 0.25, 
while the same belt, well worn and well groomed, will give 0.65 in a 
clean, dry room;" and, more than that, knows that this elusive quan- 
tity will vary all the way from almost to 1.50. However, just 
because the author is a practical and practicing engineer, though very 
fond of a little pure mathematics in the handling of practical engineer- 
ing problems, he has adopted something as a standard, this something 
being a variable lying happily between the great extremes, instead of 
being merely a single average between the extreme values. 

16 The author is not at all disappointed because a perfectly new 
belt will not give the output required, at its minimum tension, without 
the resort to a temporary application of some good adhesion-produc- 
ing belt-dressing; nor on the other hand, when a "well worn and 
groomed" belt at times is capable of giving the output required, at a 
little less initial tension than the one he aims at maintaining by the 



TRANSMISSION OF POWER BY LEATHER BELTING 103 

means more fully described in the discussions submitted by Mr. Allen 
and Mr. Hathaway. 

17 Mr. Nagle also remarks that we rarely know the exact power to 
be transmitted except in the case of prime movers, which no doubt 
is true, so far as the work of most engineers is concerned. However, 
in the author's practice at least, the maximum output of every belt 
put up on any machine is known; simply because he personally sets 
the limit, and has means of seeing that the same is never exceeded. 

18 Mr. Nagle refers to his paper read in 1881 as the first one to 
recognize the effect of centrifugal force in a belt. A correct formula, 
however, for the loss of effective tension in a belt, due to its centrifugal 
force, was given by Weisbach, at a much earlier date. This fact does 
not detract, of course, from the value of Mr. Nagle's paper, in which, 
probably for the first time, this matter was presented in a manner that 
made it readily available to the busy, practical engineer. 

19 As regards Mr. Nagle's suggestion that the data have not been 
presented in a sufficiently handy form for the busy engineer, the 
author believes Mr. Nagle has failed to appreciate the slide-rule illus- 
trated in Fig. 5, which contains these data in a form which for handi- 
ness leaves tables and diagrams far behind, while he at the same 
time is not ready to admit that there is anything the matter with the 
various diagrams that give the same data. 



SAFETY VALVES 

No. 1231 

SAFETY VALVES FOR LOCOMOTIVES 

By Frederic M. Whyi , New York 
Member of the Society 

It is the purpose of this paper to present some ideas about safety 
valves for steam boilers and particularly for locomotive boilers. 

2 Just how the capacity of the first valve used on a steam boiler 
was determined, or what relation this capacity had to the generat- 
ing capacity of the boiler, may be recorded somewhere in history, but 
it is doubtful if either fact was recorded or even determined. So far 
as locomotive work is concerned, the same ignorance prevails todav; 
but there is good promise that tliis ignorance will soon be dispelled. 
In marine work certain formulae have been devised for calculating 
the sizes of safety valves, and these formulae have been accepted, 
more or less blindly, it is thought. 

3 The general practice in locomotive work has been to determine 
in an " offhand" way the size and number of safety valves to be used, 
and former practice has guided the determination entirely. If a 
larger boiler is to be used the valve capacity may not be increased, 
depending upon the judgment of the person whose duty it is to deter- 
mine the capacity. Again, the capacity has been indicated in an 
indifferent manner, being expressed as a "size," meaning the diam- 
eter of something more or less uncertain; while the other dimension, 
the lift, which is necessary to give an indication of the capacity, is 
entirely ignored. 

4 But to know the exact capacity of the available valves is not 
sufficient; it is quite as important to know how much steam is to be 
released and in what length of time it should be released. It will 
be comparatively easy to determine the capacity of safety valves, if 
indeed the elaborate tests which have already been made — data 

Presented at the New York monthly meeting (February 1909) of The 
American Society of Mechanical Engineers. 



106 SAFETY VALVES FOR LOCOMOTIVES 

from which it is hoped may be presented in the discussion^ — have 
not already solved part of the problem; more difficult will be that part 
of it which is concerned with the quantity of steam to be released and 
the rate of the release. The subject is of mutual interest to the valve 
manufacturer and the user, — the design of the valve for capacity and 
wear to be worked out by the manufacturer; and the capacity which 
is to be used, both in volume and in number of valves, and the rate of 
release, to be determined by the user with the assistance of the manu- 
facturer. 

ESSENTIALS OF A SAFETY VALVE ON A LOCOMOTIVE 

5 The design of the valve will include the diameter of the con- 
trolling opening and the passages leading to it from the steam volume, 
as well as those leading from it to the atmosphere, the shape and 
material of the seat, the amount of lift of the valve, and the shape 
and material of the valve face, the spring and its protection, the 
adjustment, the muffler, if one is to be used, and the action of the 
valve in lifting and in seating. 

6 It will not be necessary to discuss the diameter of the control- 
ling opening, and of the passages to and from it, in view of the sugges- 
tion here made that instead of indicating the capacity of a valve in 
a very rough way by the diameter of some opening, the capacity be 
expressed in pounds of steam at certain pressures. The shape of the 
seat and of the valve face may or may not be of importance; but this 
will be referred to again. The material in the seat and face will 
naturally be that which will best withstand the effects of the flow of 
steam over them, and the possible pound of the valve when seating. 

7 The reliability of the spring and the effect of heat upon it are 
very important points. Adjustments should be readily made, but 
on the other hand to get out of adjustment should be practically 
impossible. The capacity of the muffler need not be questioned, 
except in extreme designs, but the indicated capacity should be that 
of the valve complete, with or without muffler, according to the in- 
tended use of the valve; then it is important only that it deaden suffl- 
ciently the noise of the escaping steam. 

8 The action of the valve in lifting and in seating, the desirability 
of a forewarning that the maximum pressure is about reached, 
and the operating conditions which bear upon this question of fore- 

' These data are given in the paper "Safety Valve Capacity" which follows. 
They were presented as a discussion'and afterwards published in The Journal as 
a paper under the above title. — Editor. 



SAFETY VALVES FOR LOCOMOTIVES 107 

warning, are correlative. With any Idnd of steam-generating plant 
it ought to be quite sufficient if those immediately responsible for the 
quantity produced, and for its use, know what is available; in station- 
ary and in marine work this is generally true, and steam gages can be 
placed within view of those who should know what the pressure is at 
any time. Unfortunately in locomotive work, however, it has become 
perhaps desirable that others than those within view of the gage in 
the cab know something about the steam pressure, and inasmuch as 
the fireman is wilUng, and sometimes anxious, that they should know, 
he takes the only means at hand to inform them when he thinks that 
the results of his labors are good, and "fires against the pop" so that 
everybody within hearing or sight of the valve knows by the escaping 
steam that the fireman is doing his duty. If when a train is ascend- 
ing a grade the conductor at the rear sees steam escaping from the 
valve he knows the train will get up the grade; on hard grades he 
will watch for the only indication which can be given him, and the 
fireman tries to present this indicator, the escape from the valve, 
the "white feather." 

9 Numerous similar examples might be mentioned, but assum- 
ing that such an indicator of steaming conditions has grown to be a 
necessity, undesirable as it may be, how can it be produced at the 
least expense? Surely not with a valve from 2^ to 4 in. in diameter 
and open to its full capacity. Two devices, at least, are available 
to give the indication at a lower cost than the full open valve: the 
"simmering" valve, which will open slightly for two or three pounds 
about the normal maximum, then open full, and just reversing this 
order in seating; the other, the small pilot valve, which will open at two 
or three pounds pressure below the working valve. The first method 
will have some bearing on the kind of metal to be used in the valve 
seat and valve face and possibly upon the shape of the exterior edge 
of the valve and the opposing surface of the seat. The second 
method means the addition of the small valve, an additional cost for 
which there will be no need if the first method can be developed 
successfully. 

RELATION OP VALVE CAPACITY TO STEAM-GENERATING CAPACITY 

10 There remains for consideration the relation of valve capacity 
to steam-generating capacity, and the unit capacity of the valves 
which will make up the total valve capacity. The fact that in loco- 
motive work the total valve capacity has not been as great as the 
maximum steam-generating capacity should be ample proof that 



108 SAFETY VALVES FOR LOCOMOTIVES 

such valve capacity is not necessary. The reason for this is, 
of course, that on account of using the exhaust steam for producing 
the forced draft, when the demand for steam from the boiler is reduced 
or entirely cut off, the forced draft is automatically reduced or cut off, 
and the generating capacity is reduced so that it is not necessary that 
the safety valves release the full generating capacity. Probably it 
will be largely a question of opinion what per cent of the total generat- 
ing capacity the valve ought to have, although it is possible that as 
attention is centered upon this question some more or less positive 
solution of it may result. 

11 Having fixed upon the per cent of the generating^capacity to 
be provided for in the valves it will be necessary to determine the 
desirable unit capacity of the valves. Some States require that each 
locomotive boiler have at least two valves. Starting with this condi- 
tion, consideration of the maintenance of the valves indicates that 
they should be duplicates and therefore that each have a capacity 
equal to one-half the required generating capacity. If a number of 
boilers of different capacities are to be considered, then the smaller 
ones would probably be provided with the same valves as the larger 
ones for the purpose of duplication. There are some large boilers 
for which three valves might be necessary, because the necessary 
capacity in two units might make the valves abnormally large for 
construction purposes. Also it is worth while to consider whether 
undesirable results would come about from opening almost instan- 
taneously an escape of steam from the boiler to the atmosphere. No 
suggestions are offered on this, but it is hoped that something bear- 
ing on the question may be developed in the discussion. 

12 It is suggested that instead of setting the several valves on 
a boiler at different pressures, all the valves be set at the same 
pressure, with the idea that each of them will operate frequently 
enough to keep all in working condition, rather than run the risk 
of one valve being found out of condition when it is required for 
action. 

13 It is probable that some time it will be found desirable to 
consider the minimum distance above the steam releasing surface of 
the water at which the safety valve seat may be placed. 



No. 1232 

SAFETY VALVE CAPACITY 

By Philip G. Dakung, New York 
Associate Member of the Society 

The function of a safety valve is to prevent the pressure in the 
boiler to which it is applied from rising above a definite point, to do 
this automatically and under the most severe conditions which can 
arise in service. For this, the valve or valves must have a reliev- 
ing capacity at least equal to the boiler evaporation under these 
conditions. If it has not this capacity, the boiler pressure will continue 
to rise, although the valve is blowing, with a strain to the boiler 
and danger of explosion consequent to over-pressure. Thus, with 
the exception of the requisite mechanical reliability, the factor in a 
safety valve bearing the most vital relation to its real safety is its 
capacity. 

2 It is the purpose of this paper to show an apparatus and method 
employed to determine safety valve lifts, giving the results of tests 
made with this apparatus upon different valves; to analyze a few of 
the existing rules or statutes governing valve size; and to propose a 
rule, giving the results of a series of direct capacity tests upon which 
it is based; to indicate its application to special requirements; and 
finally its general bearing upon valve specifications. 

3 Two factors in a safety valve geometrically determine the area 
of discharge and hence the relieving capacity: — the diameter of the 
inlet opening at the seat and the valve lift. The former is the 
nominal valve size, the latter is the amount the valve disc lifts verti- 
cally from the seat when in action. In calculating the sizes of valves 
to be placed on boilers, rules which do not include a term for this valve 
lift, or an equivalent, such as a term for the effective area of discharge, 
assurne in their derivation a lift for each size of valve. Nearly all 
existing rules and formulae are of this kind, which rate all valves of 
a given nominal size as of the same capacity. 

4 To find what lifts standard make valves actually have in prac- 

Presented at the New York monthly meeting (February 1909) of The 
American Society of Mechanical Engineers. 



110 



SAFETY VALVE CAPACITY 



Fixed Center Shaft 
Driven by a siiuiU Motor 




Comicction Tapped iiito 
dilt'erent Places in Valve 
Case Exhaust Pipe etc. 
to determine Back Pressure 







Connection to Boiler 



ConiMctetl to Boiler 
as in.SerN-ice 



Fig. 1 Safety Valve Lift-Recording Apparatus 



SAFETY VALVE CAPACITY 111 

tice, and thus test the truth or error of this assumption that they are 
approximately the "same for the 'same size of valve, an apparatus has 
been devised and tests conducted upon different makes of valves. 
With this apparatus not only can the valve Hft be read at any moment 
to thousandths of an inch, but an exact permanent record of the lift 
during the blowing of the valve is obtained, somewhat similar to a 
steam engine indicator card in appearance and of a quite similar use 
and value in analyzing^the action of the valve. 

5 As appears in Fig. 1 the valve under test is mounted upon the 
boiler in the regular manner, and a small rod is tapped into the top 
end of its spindle, which rod connects the lifting parts of the valve 
directly with a circular micrometer gage, the reading hand of which 
indicates the lift upon a large circular scale or dial. The rod through 
this gage case is solid, maintaining a direct connection to the pencil 
movement of the recording gage above. This gage is a modified 
Edson recording gage with a multiplication in the pencil movement 
of about 8 to 1, and with the chart drum driven by an electric motor 
of different speeds, giving a horizontal time element to the record. 
The steam pressures are noted and read from a large test gage gradu- 
ated in pounds per square inch, and an electric spark device makes 
it possible to spot the chart at any moment, which is done as the 
different pound pressures during the blowing of the valve are reached. 
The actual lift equivalents of the pencil heights upon the chart are 
carefully calibrated so that the record may be accurately measured 
to thousandths of an inch. 

6 In testing the motor driving, the paper drum is started and 
the pressure in the boiler raised. The valve, being mounted directly 
upon the boiler, then pops, blows down and closes under the exact 
conditions of service, the pencil recording on the chart the history 
of its action. 

7 With this apparatus, investigations and tests were started upon 
seven different makes of 4-in. stationary safety valves, followed by 
similar tests upon nine makes of muffler locomotive valves, six of 
which were 3^ in., all of the valves being designed for and tested at 
200 lb. The stationary valve tests were made upon a 94-h.p. water- 
tube boiler made by the Babcock & Wilcox Co. (See Fig. 2.) The 
locomotive valve tests were made upon locomotive No. 900 of the 
Illinois Central R. R., the valve being mounted directly upon the 
top of the main steam dome. (See Figs. 4 and 5.) This locomo- 
tive is a consoUdation type, having 50 sq. ft. of grate area and 
2953 sq. ft. of heating surface. Although a great amount of addi- 



112 



SAFETY VALVE CAPACITY 




Fig. 2 Valve-Lift Apparatus as Used with^the Stationary Tdst Boiler 
AT Bridgeport, Conn. 



SAFETY VALVE CAPACITY 



113 



tional experimenting has been done, only the results of the above tests 
will be quoted here. These lift records show (with the exception of 
a small preliminary simmer, which some of the valves have) an abrupt 
opening to full lift and an almost equally abrupt closing when a certain 
lower lift is reached. Both the opening and closing lifts are signifi- 
cant of the action of the valves. (See Fig. 3.) 

8 The results of the 4-in. iron body stationary valve tests sum- 
marized are as follows: of the seven valves the average lift at open- 
ing was 0.079 in. and at closing 0.044 in., or excluding the valve with 
the highest lifts, the averages were 0.07 in. at opening and 0.037 in. 
at closing. The valve with the lowest Ufts had 0.031 in. at opening 
and 0.017 in. at closing, while that with the highest had 0.137 in. 
and 0.088 in. Expressing the effective steam-discharge areas of the 



r:: 




Fig. 3 Typical Valve-Lift Diagrams 



valves taken at their opening'Jlifts as percentages of the largest 
obtained, the smallest had 31.4 per cent, the next larger 40.8 per cent, 
and the next 46.6 per cent. Of the six 3^-1^. muffler locomotive 
valves the summarized hfts are as follows: average of the six valves, 
0.074 in. at opening and 0.043 in. at closing. Average, excluding 
the highest, 0.061 in. at opening and 0.031 in. at closing. The lowest 
Uft valve had 0.04 in. opening and 0.023 in. closing; the highest, 
0.140 in. opening and 0.102 in. closing. As percentages of the largest 
effective steam-discharge area, the smallest was 36.4 per cent, the 
next larger 39.8 per cent, and the next 46.4 per cent. In both the 



114 



SAFETY VALVE CAPACITY 



stationary and locomotive tests, the lowest lift valve was flat-seated, 
which is allowed for in the above discharge area percentages. 

9 The great variation — 300 per cent — in the lifts of these stand- 
ard valves of the same size is startling and its real significance is 
apparent when it is realized that under existing official safety valve 
rules these valves, some of them with less than one-third the lift and 
capacity of others, receive the same rating and are listed as of equal 
relieving value. Three of these existing rules are given as an illus- 
tration of their nature: the United States Supervising Inspectors 




Fig. 4 Valve Lift Apparatus as Erected for Locomotive Testing at 

BuRNSiDE, III, 

Rule, that of the Board of Boiler Rules of Massachusetts, and the 
Boiler Inspection Rule of Philadelphia. 

BULE OF UNITED STATES BOARD OP SUPERVISING INSPECTORS 

W 
A - 0.2074 X^=r 



A "» area of safety valve in square inches per square foot of grate surface, 
ly = pounds of water evaporation per square foot of grate per hour. 
P = boiler pressure (absolute). 



SAFETY VALVE CAPACITY 



115 



10 In 1875 a special committee was appointed by the United 
States Board of Supervising Inspectors to conduct experiments upon 
safety valves at the Washington Navy Yard. Although the pres- 




FiG. 3 Detail of Lift Apparatus at Burnside, III., Showing Locomotive 

Valve 



sures used in these experiments (30 and 70 lb. per square inch) were 
too low to make the results of much value today, one of the conclu- 
sions reported is significant: 



116 SAFETY VALVE CAPACITY 

a That the diameter of a safety valve is not an infallible test 

of its efficiency. 
b That the lift which can be obtained in a safety valve, 

other conditions being equal, is a test of its efficiency. 

1 1 The present rule of the board, as given above, formulated by 
L. D. Lovekin, Chief Engineer of the New York Shipbuilding 
Co., was adopted in 1904. Its derivation assumes practically a 45- 
deg. seat and a valve lift of 3^ of the nominal valve diameter. The 
discharge area in this rule is obtained by multiplying the valve lift 

D 

— by the valve circumference (tt X D) and taking but 75 per cent of 

the result to allow for the added restriction of a 45-deg. over a flat 
seat. The 75 per cent equals approximately the sine of 45 deg. or 



0.707. This value for the discharge area, i.e.,! 0.75 X tt X ^ ), issubsti- 

P 

tuted directly into Napier's formula for the flow of steam, w= a ^-. 

Thus in the valves to which this rule is applied the following lifts 
are assumed to.exist: 1-in. valve, 0.03 in.; 2-in. valve, 0.06 in.; 3-in. 
valve, 0.09 in.; 4-in. valve, 0.13 in.; 5-in. valve, 0.16 in.; 6-in. valve, 
0.19 in. 

\^2 Referring back to the valve lifts as given in Par. 8, it is seen 
that the highest lift agrees very closely with the lift assumed for 4-in. 
valves in the rule of the Board of Supervising Inspectors. If the 
lifts of valves of different design were more uniformly of this value, 
or if the rule expressly stipulated either that the Uft of 3V of the valve 
diameter actually be obtained in valves qualifying under it; or, if not, 
that an equivalent discharge area be obtained by the use of larger 
valves, the rule would apply satisfactorily However, the lowest 
lift valve actually has but 25% of the lift assumed for the 4-in. valves 
in the rule; the next larger less than 50%; while the average lift of 
these valves, excluding only the highest, is but 50% of the assumed 
lift in the rule for 4-in. valves. 

MASSACHUSETTS RULE 

, IOX70 ,, 
A ^Xll 

A = area of safety valve in square inches per sq. ft. of grate surface. 
w = pounds of water evaporation per square foot of grate surface per 

second. 
P = boiler pressure (absolute) *t which valve is set to Wow. 



SAFETY VALVE CAPACITY 117 

13 One of the most recently issued rules is that contained in the 

pamphlet of the new Massachusetts Board of Boiler Rules, dated 

March 24, 1908. This rule is merely the United States rule given 

above with a 3.2 per cent larger constant and hence requiring a valve 

larger by that amount. The evaporation term is expressed in pounds 

per second instead of per hour and two constants are given instead of 

one, but when reduced to the form of the United States rule it gives 

W 
A = 0.214 X p. Figuring this back as was done above with the 

United States rule, and taking the 75 per cent of the fiat seat area 
as there done, this rule assumes a valve lift of jU of the valve diameter 
instead of the 3^ of the United States rule. This changing of the 
assumed lift from 3^ to ^ of the valve diameter being the only dif- 
ference between the two rules, the inadequacy of the United States 
rule just referred to applies to this more recent rule of the Massa- 
chusetts Board. 

PHILADELPHIA RULE 

22.5 G 
a = 



p + 8.62 



a = total area of safety valve or valves in square inches. 
G "= grate area in square feet. 
p = boiler pressure (gage) . 

14 The Philadelphia rule now in use came from France in 1868, 
where it was the official rule at that time, having been adopted and 
recommended to the city of Philadelphia by a specially appointed 
committee of the Franklin Institute, although this committee frankly 
acknowledged in its report that it "had not found the reasoning upon 
wliich the rule had been based," The area a of this rule is the 
effective valve opening, or as stated in the Philadelphia ordinance 
of July 13, 1868, " the least sectional area for the discharge of steam. " 
Hence if this rule were to be appUed as its derivation from the French 
requires, the lift of the valve must be known and considered when- 
ever it is used. However, the example of its application given in 
the ordinance as well as that given in the original report of the 
Franklin institute committee, which recommended it, shows the area 
a applied to the nominal valve opening. In the light of its derivation, 
this method of using it takes as the effective discharge area the 
valve opening itself, the error of which is very great. Such use, as 
specifically stated in the report of the committee above referred to, 



118 SAFETY VALVE CAPACITY 

assumes a valve lift at least i of the valve diameter, i.e., the practi- 
cally impossible lift of 1 in. in a 4-in. valve. Nevertheless, this is 
exactly the method of use indicated in the text of the ordinance. 

15 The principal defect of these rules in the light of the preceding 
tests is that they assume that valves of the same nominal size have 
the same capacity and they rate them the same without distinction, 
in spite of the fact that in actual practice some have but J of the capac- 
ity of others. There are other defects, as have been shown, such as 
varying the assumed lift as the valve diameter, while in reality with 
a given design the lifts are more nearly the same in the different sizes, 
not varying nearly as rapidly as the diameters. And further than 
this, the lifts assumed for the larger valves are nearly double the 
actual average obtained in practice. 

16 The direct conclusion is this, that existing rules and statutes 
are not safe to follow. Some of these rules in use were formulated 
before, and have not been modified since, spring safety valves were 
invented, and at a time when 120 lb. was considered high pressure. 
None of these rules takes account of the different lifts which exist in 
the different makes of valves of the same nominal size, and they 
thus rate exactly aUke valves which actually vary in lift and reUev- 
ing capacity over 300 per cent. It would therefore seem the duty 
of all who are responsible for steam installation and operation to 
leave the determination of safety valve size and selection no longer 
to such statutes as may happen to exist in their territory, but to 
investigate for themselves. 

17 The elements of a better rule for determining safety valve size 
exist in Napier's formula for the flow of steam, combined with the 
actual discharge area of the valve as determined by its lift. In 
Steam Boilers, by Peabody & Miller, this method of determining 
the discharge of a safety valve is used. The uncertainty of the 
coefficient of flow, that is, of the constant to be used in Napier's 
formula when applied to the irregular steam discharge passages of 
safety valves, has probably been largely responsible for the fact that 
this method of obtaining valve capacities has not been more 
generally used. To determine what this constant or coefficient of 
flow is, and how it is affected by variations in valve design and 
adjustment, an extended series of tests has recently been conducted 
by the writer at the Stirling Department of the Babcock & Wilcox Co., 
at Barberton, Ohio. 

18 A 373-h.p. class K, No. 20 Stirling boiler, fired with a Stirhng 
chain grate, with a total grate area of 101 sq. ft., was used. This 



SAFETY VALVE CAPACITY 



119 



boiler contained a U-type superheater designed for a superheat of 
50 deg. fahr. The water feed to this boiler was measured in caHbrated 
tanks and pumped (steam for the pump being furnished from another 
boiler) through a pipe line which had been blanked wherever it was 
impossible, with stop valves and intermediate open drips, to insure 
against^ any leakage. The entire steam discharge from the boiler 
was through the valve being tested, all other steam connections from 
the boiler being either blanked or closed with stop valves beyond which 
were placed open drip connections to indicate any leakage. A constant 



H^. 




Fig. 6 Akrangement of Valve with Micrometer Spindle Used in the 
Direct-Capacity Testing at Barberton, Ohio 

watch was kept throughout the testing upon all points of the feed 
and steam lines, to insure that all water measured in the calibrated 
tanks was passing through the tested valves without intermediate 
loss. 

19 The valves tested consisted of 3-in., 3^-in. and 4-in. iron 
stationary valves, and l^in., 3-in. and 3^-in. locomotive valves, 
the latter with and without mufflers. These six valves were all 
previously tested and adjusted on steam. Without changing the 
position of the valve disc and ring, the springs of these valves were 



120 SAFETY VALVE CAPACITY 

then removed and solid spindles, threaded (with a 10-pitch thread) 
through the valve casing above, inserted. Upon the top end of these 
spindles, wheels graduated with 100 divisions were placed. Fig. 6 
shows the arrangement used with the locomotive valves, the spindle 
and graduated wheel being similar to that used with the stationary 
valves. By this means the valve lift to thousandths of an inch was 
definitely set for each test and the necessity for constant valve lift 
readings, with that source of error, eliminated. 

20 In conducting the tests three hours' duration was selected as 
the minimum time for satisfactory results. Pressure and tempera- 
ture readings were taken every three minutes, water readings every 
half hour. A man stationed at the water glass regulated the feed 
to the boiler to maintain the same level in the boiler during the test; 
other men were stationed, one at the water tanks, one firing and one 
taldng the pressure and temperature readings. Pressure readings 
were taken from two test gages connected about 4 in. below the 
valve inlet, the gages being calibrated both before and after the series 
of tests was run and corrections applied. In all 29 tests were run, 
of which 15 were 3 hours long, 4 were 2^ hours, 3 were 2 hours, and 
7 of less duration. 

21 Tests numbered 1 to 5 were preliminary runs of but one hour 
or less duration^ apiece, and records of them are thus omitted in 
Table 1, which gives the lifts, discharge areas, average pressure and 
superheat, and the steam discharge in pounds per hour of each 
of the other tests. The discharge areas in the valves with 45-deg. 
seats are given by the following formula which is easily derived 
geometrically : 

a= 2.22 X D X I + 1.11 X J? 
where 

a = effective area in square inches 
D= valve diameter in inches 
^= valve Hft in inches 

In tests 8 and 23, where the width of valve seat was 0.225 in. and 
0.185 in. respectively, and the valve was thus slightly above the 
depth of the valve seat, the area was figured for this condition. 

22 As previously stated, the appHcation of these results is in 
fixing a constant for the flow of Napier's formula as applied to 

p 
sarety valves. The formula is w =a - in which w equals pounds 

discharged per second, P equals the absolute steam pressure behind 



SAFETY VALVE CAPACITY 121 

the orifice or under the valve and a equals the effective discharge 
opening in square inches. This may be stated as E = C X a X P', 
in which E equals the pounds steam discharge per hour and C equals 
a constant. The values of E, a and P being given for the above 
tests, C is directly obtainable. 

23 Figuring and plotting the values of this constant indicate the 
following conclusions: 

a Increasing or altering the steam pressure from approxi- 
mately 50 to 150 lb. per square inch (tests 14 and 10) 
does not affect the constant, this merely checking the 
applicability of Napier's formula in that respect. 

b Radically changing the shape of the valve disc outside 
of the seat at the huddling or throttling chamber, so-called, 
does not affect the constant or discharge. In test 15 
the valve had a downward projecting lip (as in Fig. 1), 
deflecting the steam flow through nearly 90 deg., yet the 
discharge was practically the same as in tests 10 and 14, 
where the lip was cut entirely away (as in Fig. 6), giving 
a comparatively unobstructed flow to the discharging 
steam. 

c Moving the valve adjusting ring through much 'more than 
its complete adjustment range does not affect the con- 
stant or discharge. (Tests 16 and 17.) 

d The addition of the muflBer to a locomotive valve does 
not materially alter the constant or discharge. There is 
but 2 per cent difference between tests 10 and 13. 

e Disregarding the rather unsatisfactory IJ-in. and 3-in. 
locomotive valve tests, the different sizes of valves 
tested show a variation in the constant of about 4 per 
cent when plotted to given lifts. 

/There is a slight uniform decrease of the constant when 
increasing the valve lifts. 

24 The variations indicated in the last two conditions are not large 
enough, however, to impair materially the value of a single constant 
obtained by averaging the constants of all the 24 tests given. The 
selection of such a constant is obviously in accord with the other four 
conditions mentioned. This average constant is 47.5, giving as the 
formula E = 47.5 X a X P. Its theoretical value for the standard 
orifice of Napier's formula is 51.4, of which the above is 92^ per cent. 

25 To make this formula more generally serviceable, it should 
be expressed in terms of the valve diameter and lift, and can be still 



122 SAFETY VALVE CAPACITY 

further simplified in its application by expressing the term E (steam 
discharged or boiler evaporation per hour) in terms of the boiler heat- 
ing surface or grate area. For the almost universal 45-deg. seat the 
effective discharge area is, . with a slight approximation, L X sine 
45 X TT X I), in which L equals the valve lift vertically in inches and 
D the valve diameter in inches. Substituting this in the above 
formula gives E = 47.5 X Lx sine 45 X tt X Z) X P, or £/ = 105.5 
X L X D X P. 

26 The slight mathematical approximation referred to consists 
in multiplying the {L X sine 45) by {n X D) instead of by the exact 
value {k X D -\- ^L) . To find directly the effect of this approxima- 
tion upon the above constant, the values for E, L, D and P from the 
tests have been substituted into the above formula and the average 
constant re-determined, which is 108.1. The average lift of all the 
tests is 0.111 in. Plotting the constants obtained from the above 
formula in each test, as ordinates, to valve Hfts, as abscissae, obtain- 
ing thus the slight inclination referred to in Par. 23 /, and plotting 
a fine with this inclination through the above obtained average con- 
stant 108.1, taken at the 0.111-in. average lift, gives a line which at 
a maximum lift of say 0.14 in. gives a constant of just 105. At 
lower lifts this is slightly larger. Hence 105 would seem to be the 
conservative figure to adopt, as a constant in this formula for general 
use, giving 

E= 105 XLXDXP 

This transposed for D gives: 

pi 
D = 0.0095 X i^^^p 

Note that the nominal valve area does not enter into the use of this 
formula and that if a value of 12, for instance, is obtained for D it 
will call for two 6-in. or three 4-in. valves. For flat seats these con- 
stants become 149 and 0.0067 respectively. 

27 The fact that these tests were run with some superheat (an 
average of 37.2 deg. fahr.) while the majority of valves are used 
with saturated steam, would, if any material difference exists, place 
the above constants on the safe side. The capacities of the stationary 
and locomotive valves, the lift test results of which are summarized 
in Par. 8, have been figured from this formula, taking the valve lifts 
at opening, and in pounds of steam per hour are as follows: 



SAFETY VALVE CAPACITY 123 

Of the seven 4-in. iron body stationary valves, the average 
capacity at 200 lb. pressure is 7370 lb. per hour, the 
smallest capacity valve (figured for a flat seat) has a 
capacity of 3960 lb., the largest 12,400 lb., and of the 
six 3^-in. muffler locomotive valves at 200 lb. pressure, 
the average capacity is 6060 lb. per hour, the smallest 
4020 lb., the largest 11,050 lb. 
28 To make the use of the rule more direct, where the evaporation 
of the boiler is only indirectly known, it may be expressed in terms 
of the boiler heating surface or grate area. This modification con- 
sists merely in substituting for the term E (pounds of total evapora- 
tion) a term H (square feet of total heating surface) multiphed by 
the pounds of water per square foot of heating surface which the 
boiler will evaporate. Evidently the value of these modified forms 
of the formula depends upon the proper selection of average boiler 
evaporation figures for different types of boilers and also upon the 
possibility of so grouping these boiler types that average figures 
can be thus selected. This modified form of the formula is 

D=CX ^ 



L X P 



in which H equals the total boiler heating surface in square feet and C 
equals a constant. 

29 Values of the constant for different types of boilers and of 
service have been selected. These constants are susceptible of 
course to endless discussion among manufacturers, and it is undoubt- 
edly more satisfactory, where any question arises, to use the form 
containing term E itself. Nevertheless the form containing the 
term H is more direct in its application and it is beheved that the 
values given below for the constant will prove serviceable. In apply- 
ing the formula in this form rather than the original one, containing 
the evaporation term E, it should be remembered that these con- 
stants are based upon average proportions and hence should not be 
used for boilers in which any abnormal proportions or relations 
between grate area, heating surface, etc., exist. 

30 For cylindrical multi-tubular, vertical and water-tube station- 
ary boilers a constant of 0.068 is suggested. This is based upon an 
average evaporation of 3^ lb. of water per square foot of heating 
surface per hour, with an overload capacity of 100 per cent, giving 
7 lb. per square foot of heating surface, the figure used in obtaining 
the above constant. 



124 



SAFETY VALVE CAPACITY 



02 c^ 



a 


m 


H 


O 


>< 


. 


H 




< 


w 


(U 


n 


< 


a 


n 


< 


H 


« 


> 


• 


< 


d^ 


I> 


X 






'^ 


J 


g 


^ 


<5 


4 


r/) 






M 




u 


f-i 





W 


n 


hJ 


■< 


m 


« 


<j 


n 


H 















- 














•o 


(U 


A 


a> 






t~ 


•-; "p 














£ 














B 
0] 

o 
> 

■3 

> 


_> 


<B 


> 








"O *hn 




< 

a 

H 






i 


IN 


CO 

£ 
p 


3 


£ 

£ 

p 




> 
o 
S 
o 

o 
o 

H-1 








e 




M 

'^ 


a 

si 




"3 

> 

"o 
<o 
0. 
>> 

H 




3 


m 

£^ 
"o 

c 
.2 


CO c 

£ -B 

p •'~ 

m 

£ Oh 










O ci d 


ci 


u 
c9 




S 

a 
O 


"5 




3 




(D « 

lis 





<u 

a. 

lU 


eu to 










iz; « P5 




_i9 


• — 


"— 


• — 


— - 


s_ 




H 







^.^ 




^0 




^1 




a 


50 


"5 


■* 




o 






CO 


o 


















c!> 




N 


lO 


o 




o 


o 




O) 


00 


CO 




o> 


a> 




la 




•0 




CO 


<N 


t^ 




■* 


o 




■* 


C4 















CO 




CO 


Q 




d- 


d 








o 






«o 


00 












N 




w 






M 










d 


""■ 




d 


d 


l-H 




"^ 


""■ 




"^ 




■^ 


« 2 


»• « 


"o 


g 


o 


^ 


o 




o 


o 




o 


o 























a « 


S p 






(N 




(M 




o: 






t^ 


00 







CM 







t^ 







;; •< 


H n 










O 




<N 






CO 


t^ 


M< 




CO 


CO 




t^ 




a> 


Q gd^i^ 


3 


ra 


3 


00 






t^ 


00 




•* 


CO 


00 




CO 


00 




00 




00 


a, 


■< 




fcl 


(O 


^ 


(N 




O 


o 




e 


(N 


^ 




CO 







CO 




„ 






Ci 


lO 


OJ 




r4 


d 




00 


^ 


06 




r^ 


d 




<N 




d 


P 


n 




■* 


■* 


•* 




■* 


CO 




CO 


■* 


CO 




10 


CO 




11 




iO 


02 






P 




































P 


hi 


a' 


t>. 


■*. 


t^ 




■^ 


t^ 




lo 


CO 


CO 




<N 


■* 




»o 







oo 






t—t 


U5 


lO 




d 


CO 




■N 


d 


CO 




CJ 


CO 




00 




(N 


S 


5 


o* 


U5 


rt< 


CO 




T»t 


■* 




■o 


iO 


T)< 




10 


TT 




CO 




'!< 


<£ 


^ 


02 


■^ 


'^ 






'^ 


"^ 








"^ 






'^ 




"^ 




"^ 




a 




m 


>o 








U3 










10 


















^5 




v 


05 


OS 


o 




"i" 


o 




o 


lO 



























j3 


«o 


CO 


00 




o 


•* 






o 


CO 




■<J< 


Tl< 




•* 




■* 








u 


o 














d 






















^ h3 




a 






































> 






d 


d 


d 




d 


d 




d 


d 


d 




d 


d 




d 




d 










i«^ 










1 










— 










^- 


.s" 










13 










s 






















h^ 










a 










1^ 


























H 
















9 
o 
















1 


a 
5 








a 

g 

p 








, 


'""i 




3 


i 
^ 




, 


, 


CI 








I 


J3 









5 






<< 










■< 
















- 




M « 




<! 






bi 










u 








p 










"r? 




c rt 










C3 










(d 










. 


3 




•< 




^1 
1- 










1 










a 








^ 










Ul 












^ 










4J 








i 
















H 






£? 
























•a 


















g 










tf 








02 










b 








•< 














lU 








<u 




^ 


a 




d 








> 






B 










s 








s 




" 


3 




B 









^ 














H 








H 






^ 




■43 








O 






M 










u 














■^ 




s 








H 






a 










.> 














■5 




02 














o 




























S 








^ 






-^ 


= 


= 




3 


o 

S 




= 


' 


5 




m 


0. 









= 




Q 






(i.' 










o 
u 
o 

Hi 














c^ 


u 


fe 








^ 






(4 






















W 


a) 

J3 


tf 














4 










i 














a S 


.S 
4> 








GQ 






^ 










CO 














02 








•i 


o t^ 




2 




































a 






3 












Hn 














-♦> 










D 




o 


CO 


M 


CO 




CO 


Ol 




CO CO 


CO 




N 


M 




CO 




CO 


Q 




a 






































« 










































H H 










































9 n 
fi a 






«o 


t^ 


00 




a> 


o 




■-^ 


c^ 


CO 




■«)" 


10 




CO 




t^ 












































H p 










































!z: 









































SAFETY VALVE CAPACITY 



125 



.. 


^ 














5 

















8 


_ 














fc" 


■5 

B 


si 






a> 






o 
1 


§ 








0- 








> 












S 


■3 


" 













09 

o 












t> 











^ 






' 




■^2; ^ 












M 00 





CO 





CO 




CO 


CO rl< 


t^ CO 


tf> 


« OS CD 


00 


Tjl 




t~ >-l 


'^ e<3 


<N O 


"5 


lO r»< CO 


c^ 


00 




t^ t^ 


00 


Tl< <N 


(N 


(N >0 --i 


00 


00 




CO ■^ 


00 CO 


d d 


d 


d d rt 


d 


d 




d d 


d d 


U3 O 


IC 


10 










"5 





rH lO 


(N 


t^ (N CO 





t- 




CO 


CO 10 


lO >o 





CT> CO CO 


CO 


CO 




-H 1^ 


03 


C^ •-I 


(N 


-1 Tl« 00 


CO 


CO 




10 CO 


(^ ■<<< 


— Q> 


V 


CO CO 


t^ 


■* 




CO us 


1^ o> 




a 


c • • 















e<i CO 


00 


00 




ui d 


06 co' 


jz; 


^^^ 


Ti< 


•* 




•* IN 


•«< ■* 


« N 


M 


t^ 00 


CO 


^ 




00 •* 


•* CO 


d -^ 


d 


t^ CO 03 


^ 


d 




<N IN 


00' d 


T»< lO 


f-i 


■* ■* CO 


T»l 






•* T(l 


CO CO 


J^ o 


10 


10 


>o 













O CO 


r^ 


r^ 1^ ■* 





CO 




r^ 


CO 01 


rt o 





-H 








1 


"". R 


d d 


d 


ado 


d 


d 




d d 


d d 


bi 




•0 










u 


0) 




V 










V 


S 




a 










m 


a 




S 










s 


S 












s 


sx 




H 










ja 
















'%■ 




« 










'$ 


■3 ' 
< 




5^5 




' 






'4' 

< 


U 




eS 










■l 


cS 














a 






3 












1 




g" 










a 


o 

tf 




tf 
















>. 












CO 




b 




>. 








a> 




oj 




03 






03 






.2 




.2 








D 




CQ 




ca 






H 


> 
O , 

a" 

o 
o 

o 









13 






> 


5 


; t. , 


= 


2 




5 3 


'i^ 




fa 




fe 









Hi 




tf 




« 









a 




■i 




_a 






a 


1-1 




« 




CO 






CO 




•«• 


-» HM 












N -4 


w 


N -H CO 


CO 


CO 




CO Cvl 


CO CO 

1 


00 a> 





^ IN CO 


■* 


>o 




CO r- 


00 a> 




Oi 


Ol (N (N 


(N 


(N 




(N <N 


N (N 

1 



a -- -3 





n 








Q 








X 


^ 


X 


^ 


w 


1^ 




0) 




-5 


"3 


+ 


+ 


> 


"W 


X 


J3 


X 








Q 




—1 




XI 


^^ 


X 

IN 

C^_ 


0. 




+ 


II 


^ 


X 


e 


a 

03 


Q 




j3 


X 



2: a 



O >" rt 



.2 03 

II 



126 SAFETY VALVE CAPACITY 

31 For water tube marine and Scotch marine boilers, the sug- 
gested constant is 0.095. This is based upon an overload or maxi- 
mum evaporation of 10 lb, of water per square foot of heating surface 
per hour. 

32 For locomotives the constant 0.055 was determined experi- 
mentally as explained below. Special conditions to be considered 
in locomotive practice separate it from regular stationary and marine 
work. In the first place the maximum evaporation of a locomotive 
is possible only with the maximum draft obtained when the cylinders 
are exhausting up the stack, at which time the throttle is necessarily 
open. The throttle, being open, is drawing some of the steam and 
therefore the safety valves on a locomotive can never receive the 
full maximum evaporation of the boiler. Just what per cent of this 
maximum evaporation the valve must be able to relieve under the 
most severe conditions can only be determined experimentally. 
Evidently the most severe conditions obtain when an engineman 
after a long, hard, up-hill haul with a full glass of water and full 
pressure, reaching the top of the hill, suddenly shuts off his throttle 
and injectors. The work on the hill has brought the engine steam- 
ing to its maximum and the sudden closing of throttle and injectors 
forces all the steam through the safetj^ valves. Of course the 
minute the throttle is closed the steaming quicldy falls off and it is at 
just that moment that the most severe test upon the valves com.es. 

33 A large number of service tests have been conducted to deter- 
mine this constant. The size of valves u^Don a locomotive has been 
increased or decreased until one valve would just handle the maxi- 
mum steam generation, and the locomotive heating surface being 
known the formula was figured back to obtain the constant. Other 
special conditions were considered, such as the liability in locomotive 
practice to a not infrequent occurrence of the most severe conditions; 
the exceptionally severe service which locomotive safety valves 
receive; and the consequent advisability of jDroviding a substantial 
excess valve capacity on locomotives. 

34 As to the method of applying the proposed safety valve capac- 
ity rule in practice, manufacturers could be asked to specify the 
capacities of their valves, stamping it upon them as the opening and 
closing pressures are now done. This would necessitate no extra 
work further than the time required in the stamping, because for 
valves of the same size and design, giving practically the same lift, 
this would have to be determined but once, which of itself is but a 
moment's work with a small portable lift gage which is now manu- 



SAFETY VALVE CAPACITY ]27 

factured. The specifying of safety valves by a designing engineer 
could then be as definite a problem as is that of other pieces of 
apparatus. Whatever views are held as to the advantages of high 
or low Ufts, there can be no question, it would seem, as to the advan- 
tage of knowing what this lift actually is, as would be shown in this 
specifying by manufacturers of the capacity of their valves. Further, 
as to the feasibility of adopting such a rule (which incorporates the 
valve lift) in statutes governing valve sizes — this would involve the 
granting and obtaining by manufacturers of a legal rating for their 
valve designs based upon their demonstrated lifts. 

35 This paper has dealt with but one phase of the subject of safety 
valves in order that its conclusions might be drawn more clearly. 
The apparatus and tests shown indicate that the lifts and capacities 
of different make valves of the same size and for the same conditions 
vary as much as 300 per cent, and that there is therefore the ha- 
bility of large error in specifying valves in accordance with existing 
rules and statutes, because these rules as shown rate all valves of 
one size as of the same capacit}', irrespective of this variation. 
A simple rule is given, based upon an extended series of direct capacity 
tests, which avoids this error by incorporating a term for the valve 
lift. Finally, the method and advantage of applying this rule in 
practice have been briefly indicated. 



No. 1233 

DISCUSSION UPON [SAFETY VALVES 

The subject of safety valves was extensively discussed at the New York 
monthly meeting, February 1909, when the preceding papers by Frederic M. 
VVhyte and Philip G. Darling were presented. The discussion was continued 
at the Spring Meeting at Washington, May 1909. Full reports of both dis- 
cussions were published in The Journal for April and June 1909. In what 
follows, synopses only are given of the most important engineering data 
presented upon the proportions of safety valves. 

Luther D. Lovekin Miscussed the rules for marine work adopted 
by the Board of Supervising Inspectors. In 1903 the regulations con- 
cerning safety valves prescribed by the United States Board of Steam- 
boat Inspectors were investigated by Mr. Lovekin, for the purpose 
of formulating new rules. The rule in use was based on grate sur- 
face without regard to the amount of coal burned in a given time. 
The rules finally prepared by Mr. Lovekin and adopted by the Board 
of Supervising Inspectors are based on evaporation and on Napier's 
formula for the flow of steam. The formula for the required area of 
discharge for a valve is derived as follows: 
Let 

P = absolute pressure 

W = weight discharged per hour in pounds 

A = area valve opening in square inches 

d = diameter of valve in inches 

a = area of valve of diameter d 
From Napier's rule 

360 A P 

w = -^~ 

For safety valve practice allow 75 per cent of this and restrict the 
lift of valve to ^V diameter. Then 

270 P ndr 
W = — y— X ^ = 4.821 Pa 

W 
a = 0.2074 -^ 

^Chief Engineer, New York Shipbuilding Co., Camden, N. J. 



130 DISCUSSION 

If W represents the weight of water in pounds evaporated per square 
foot of grate surface per hour the above formula will give the area 
of valve required per square foot of grate surface. 

2 A table of safety valve sizes was prepared by the aid of this 
formula, but the board failed to state in their rules that the sizes 
were based on a lift of 1/32 of the diameter. In commenting upon the 
above formula, Mr. Lovekin brought out the following points : 

3 The clear area between the valve and its seat (due to having a 
lift equal to 1/32 of its diameter) is only about 1/11 of the area cor- 
responding to the nominal diameter found by the formula. There- 
fore it would seein that the inlet from the boiler to the safety valve 
need be equal in area only to the free area between the safety valve and 
its seat. This would reduce the opening in the boiler to about 1/11 of 
the area used at the present time. Experiments have shown, how- 
ever, that a free entrance from the boiler to the safety valve is abso- 
lutely necessary to prevent chattering. Just what this relation is has 
not been determined. It would depend entirely on the length of the 
nozzle or pipe connecting the safety valve to the boiler. In most 
cases, safety valves are bolted either directly to the boiler or to a 
casting \shich is bolted directly to the boiler, so there would be very 
little gain in reducing the inlet nozzle to a safety valve. If safety 
valves are connected to the outlets of dry pipes to boilers, it is advis- 
able to have at least 25 per cent excess area through the slots in order 
to prevent excessive pressures in the boilers. 

4 Some rules insist on an outlet area equivalent to the full bore 
of the safety valve, which seems inconsistent if we have only 1/11 of 
the area for the steam to pass through at the valve seat. An outlet 
from a safety valve equal to 1/2 the nominal area of the valve would 
no doubt suffice in all cases. Most of the United States battleships 
are equipped in this manner. While the United States cruiser Tennes- 
see was on trial the main engines were stopped suddenly. All the 
the safety valves responded instantly and though the steam pressure 
went up to 10 lb. above popping point, no trouble was experienced, 
proving that a combined area of outlet pipes equal to 1/2 the area of 
the safety valves was sufficient. 

Albert C. Ashton said that while the tests show what pop safety 
valves would accomplish under certain favorable conditions, they 
have not clearly demonstrated that high-lift valves so made are a 
success on all applications. They certainly have shown many failures 
in locomotive service during the past year and must still be classed 
as an experiment. 



SAFETY VALVES 131 

2 Safety valves should never give such a large and sudden relief 
as to affect the water level in a boiler, neither should they close so 
suddenly as to cause a shock to the boiler by the quick stoppage of the 
flow of steam. High-lift valves which do this are not as practical 
as lower-lift valves which give a somewhat slower and easier 
relief. 

3 The tests which Mr. Darling has explained show an average 
lift of 1/8 in. for the high-lift valves which is about double the lift of 
standard valves. Such high lift seems to be excessive, although there 
may be some virtue in making valves with a little higher lift than the 
common standard of 1/16 in. 

A. B. Carhart, speaking on the proper rating of safety valves and 
their relation to boiler capacity, said that the limit of diameter of 
valves for stationarj' boilers should be 5 in., and for locomotives 3i 
in.; common practice is in accord with this. Valves of 1 sq.in. dis- 
charge area are the largest advisable for locomotives. A total dis- 
charge area of 2 sq. in. for locomotives having 35 sq. ft. grate area, 
and of 3 sq. in. for the largest ones having 50 sq. ft. grate area, has 
been demonstrated to be amply sufficient. The capacity might be 
divided as follows: (a) muffled valve with close adjustment; (6) 
reserve valve regulated for reasonably greater discharge; and (c) an 
emergency valve as the ultimate protection against explosion, the 
other two simply to limit the working pressure under ordinary con- 
ditions. Valves generally discharge more steam than engineers are 
availing to permit. The strain on the boiler is dangerous when the 
opening is too large and sudden; and if water is lifted, it chokes the 
relief through the safety valve and endangers the cylinders. 

2 A smaller valve with high lift is not the equivalent of a valve 
of larger seat diameter and less lift wdth the same discharge area; 
the smaller valve gives a smaller percentage of steam discharge, there 
is greater danger of sticking of the guide wings in opening and more 
trouble from pounding of the seat and leaking, and the outlet area 
becomes too large in proportion to the inlet, causing chattering and 
ineffective relief to the boiler, besides requiring an undesirable exces- 
sive spring compression. The lift should not exceed 0.08 in. for loco- 
motive valves, and 0.10 in. for stationary valves used at lower 
pressures; prudence and economy would reduce rather than increase 
this hmit. Many valves of high lift have been produced in past 
years, but all have been withdrawn because of rational objections 
developed in their use. 



132 DISCUSSION 

3 Every valve has a wide range of lift, which can often be varied 
from 0.04 in. to 0.10 in. by simple adjustment, and to still greater 
limits by a change of springs. In valves as commonly made, limited 
lift is a matter of preference, not of necessity; and such valves are 
giving entire satisfaction in service, with demonstrated safety under 
all conditions, under the present rules and ratings. 

4 All internal work that must be extracted from the escaping 
steam, to accomplish high lift of the disc, reduces the velocity and 
efficiency of the relief and requires an undue throttling of the outlet, 
strangling the discharge instead of relieving the boiler. This con- 
dition is described in the early patent to Richardson, of January 19, 
1869, showing the over-lapping regulating ring. 

5 Napier's formula was used as the basis for calculating safety- 
valve areas as long ago as the tests made by the United States Board 
of Supervising Inspectors of Steam Vessels in 1875; and reports 
of tests of safety valves made at the Massachusetts Institute of Tech- 
nology can be found in Peabody and Miller's text book on Steam 
Boilers printed more than a dozen years ago, showing lifts of 0.07 in. 
and 0.08 in., with an efficiency of 95 per cent of the calculation by 
Napier's formula as there applied. 

E. A. Ma\* spoke on the proper method of rating safety valves for 
low-pressure boilers. A safety valve on a low-pressure boiler is rarely 
called upon to exhaust all the steam-generating capacity, due to several 
conditions : 

a In the majority of heating plants, the full amount of radia- 
tion is almost always in service, caring for a large percent- 
age of the steam generated, and even when the radiation is 
nearly all cut out there is still circulation through the 
piping. 

b Practically every steam boiler used in low-pressure work, 
which rarely calls for gage pressure in excess of 2 lb., has 
its damper regulator which, when properly set, checks com- 
bustion when 2-lb. pressure is reached. 

c Chimney conditions in the majority of heating plants make it 
almost impossible to drive the boiler to its maximum 
steam-generating capacity, i.e., the maximum capacity 
obtainable with every condition exactly right. 

^Mechanical Engineer, American Radiator Company, Chicago. 



SAFETY VALVES 133 

2 In practically all house installations at least two of these condi- 
tions exist, and in a majority all three, so that we would have to select 
a valve out of all proportion to actual requirements in order to 
exhaust all the steam which might be generated by the boiler under 
its full steam generating capacity under ideal conditions. 

3 This brings us to a consideration of maximum capacity and 
how it is established : whether (a) by the heating surface of the boiler 
alone; (b) by the grate surface; (c) by the fuel-carrying capacity; 
(d) by the rate of combustion; or (e) by all combined. Scarcely any 
two manufacturers of low-pressure house-heating boilers agree in 
this particular. One may rate solely on the area of heating surface, 
another on the grate surface, and still another on the amount of fuel 
the grate will carry, but the writer believes that none of those factors 
should be considered alone. 

4 In view of the wide variation in methods employed by manu- 
facturers in ratiiigs of boilers, as well as in the rules employed by users 
of safety valves, it would be difficult to select a proper size valve based 
on grate dimensions only. If valve manufacturers would indicate, in 
addition to the size of the valve, its capacity at different adjustments 
for exhausting steam, it would help materially. Valves could in fact 
be designed and sold on their exhaust capacity without regard to 
specific size, i. e.; owing to variation in design, one valve might have 
a larger diameter with a lesser lift than another, while their capacity 
for exhaust might be identical. 

5 The simplicity of this method will be appreciated by anyone 
considering the rules and formulae in effect at present. If the law 
specified, however, that for a certain evaporative power or rating of 
boiler a certain exhaust capacity should be maintained in the valve 
each manufacturer could determine for himself the proper valve to 
use. 

F. J. Cole quoted from a letter of an Enghsh locomotive builder 
stating that the "Ramsbottom" duplex safety valve is almost uni- 
versally adopted there. It was introduced in 1858 and made 3 in. 
in diameter. Notwithstanding that boilers have since nearly 
doubled in capacity and pressures have been increased 50 per cent, 
this size is still used, which shows the disregard of proportion of 
safety valve to any other part of a locomotive boiler. 

2 It is probable that general foreign practice for locomotive 
safety valves is systematized no more than in England or America. 
On account of the peculiar conditions governing the draft of loco- 



134 DISCUSSION 

motives the same necessity does not exist for safety valve regula- 
tion as in the case of marine or stationaiy boilers, the action of 
the exhaust automatically talcing care, in large measure, of the 
generation of steam. 

3 Mr. Cole stated that he is in favor of a thorough investigation 
looking towards the formulating of definite and authoritative rules 
for the application of safety valves to locomotives, and invited atten- 
tion to the following suggestions for their preparation : 

a The diameter, number and kind of safety valves to be based 
on their capacity for discharging pounds of steam per 
second at different pressures. 

b The maximum amount of steam which the safety valves may 
be required to discharge when the throttle is suddenly 
closed after the fires have been urged to their maximum 
rate, to be based on the square feet of equated heating 
surface, so that the relative values for evaporation for 
various kinds of heating surface, whether of firebox, water 
tubes for supporting arch brick, long and short boiler 
tubes between the limits of 10 and 21 ft. in length, and 
values for different spacing of boiler tubes, will be taken 
into consideration. Or, what would be simpler, some ap- 
proximation of average value of heating surface, equated 
to account for difference in length and spacing of tubes; 
the firebox heating surface in this case to be considered as 
a certain percentage of the whole for all sizes of locomotives. 

4 A great diversity of practice exists in the spacing of flues in 
locomotive boilers. The variation in length ranges in common prac- 
tice from 10 ft. to 21 ft. These two conditions make the use of 
unequated heating surface somewhat unreliable as an absolute guide 
for the amount of water evaporated. It is evident in two boilers hav- 
ing the same diameter and the same length between flue sheets that 
one will contain a much larger amount of heating surface if the flues 
are spaced 11/16 in. apart than the other if they were spaced 1 in., 
and both these figures are within the limits of what is accepted as 
good practice. Furthermore the heating surface of flues of the same 
diameter and, saj', 11 ft. long, will be much more effective per square 
foot than if the flues were 21 ft. long. Firebox heating surface is, of 
course, very much more efficient than tube heating surface, and the 
water-tube heating surface for supporting firebricks is more efficient 
than the ordinary boiler tubes. 



SAFETY VALVES 



135 



5 Tests show that the evaporation of boilers is somewhat inde- 
pendent of the tube spacing, and probably is more in direct relation 
to the cubical contents, as it is a matter of common knowledge that 
the steaming capacity of boilers does not vary in direct proportion to 
the amount of heating surface if a great variation exists in the spacing 
of the tubes. 

6 The evaporation per square foot of heating surface in locomo- 
tives is a variable quantity, ranging from 6 lb. or even less to 15 or 16 
lb. per square foot of heating surface per hour. On the authority of 
Professor Goss, from Purdue University tests, it may be stated that 
the evaporation in a very general way, and the draft produced by the 
blower and exhaust in terms of inches of water, will be approximately 
as follows: 



1-in. draft will evaporate 3.0 lb. per foot of heating surface per hour 



2-in. " 




6.0 " 


3-in. " 




8.2 " 


4-in. " 




10.0 " 


5-in. ' 




11.4 " 


6-iii. " 




12.G " 


7-in. ' 




14.0 " 


8-in. ' 




15.0 " 



Dr. Chas. E. Lucke stated that another element in the safety valve 
question, of minor importance, perhaps, is the time element. He 
had experimented for many years with rapidly rising and rapidly 
falling pressm-es, and believed that increase in pressure in a chamber 
may go momentarily far beyond what a safety valve is set for. Be- 
cause this excess is only momentary and measured in fractions of 
seconds, it should not be considered of no consequence ; it is indeed 
of far more consequence, as a suddenly applied load cannot be resisted 
by the metal under stress as well as a steady load. 

2 If then by any romote series of circumstances the pressure in the 
boiler suddenly rises, as it may, the time element will enter in, the 
pressure will go higher than the safety valve is set for, before the valve 
opens, and will suddenly stress the entire structure. This subject 
should be stu died eperimentallj^, with the others involving the steady 
rate of steam discharge, to discover if it is of any consequence in practi- 
cal safety valve work. Although Dr. Lucke had never seen the pres- 
sure rise in a steam boiler in this way, he believed it could so rise, pro- 
ducing the effects described. 



136 DISCUSSION 

Jesse M. Smith thought Dr. Lucke had touched on a point which 
needed investigation. Another point along the same line is the danger 
of having a safety valve too large, particularly if it be of the "pop" 
kind. If a boiler be stored with water at a temperature corresponding 
to 150-lb. pressure, and that pressure be suddenly reduced, a portion 
of the water will instantly flash into steam and the boiler may be 
greatly strained and may explode. There is danger from having 
a safety valve too large as well as from having it too small. 

2 Those who have had to do with the investigation of boiler 
explosions, and particularly those being questioned with regard to 
them in the courts, will realize the necessity for rules based upon 
scientific investigation and reason, instead of rules having no special 
reason for their existence except that they have existed for hundreds of 
years or more. 

Garland P. Robinson stated that the commission with which he 
is connected had collected reliable data on about 7500 locomotive 
boilers, and he had recently calculated the valve capacity of 1000 of 
these boilers for the purpose of finding the average practice for safety- 
valve equipment. The greatest variations were noted. For in- 
stance, boilers using 180-lb. pressure with valves of re-in. lift have 
two 3-in. valves to take care of the evaporation from 1750 to 3350 
sq. ft. heating surface. It was also found that two 2^-in. valves were 
used to take care of 900 to 1900 sq. ft. of heating surface. These 
cases represent whole classes and not individual boilers. Therefore 
it would appear that no rule has been followed to determine the size 
of valve required. 

2 The heating surface, all things considered, is the best unit of 
measurement for determining the size of safety valves for locomotive 
boilers. In his opinion a formula based on the heating surface and 
providing for 50 per cent of the maximum evaporation of the boiler, 
will give satisfactory results for locomotives. A formula for size 
of safety valves for locomotive boilers can be derived in the manner 
shown in Mr. Darling's paper on Safety Valve Capacity. 

3 For locomotive valves with 45-deg. valve-seats, Mr. Robinson 
used the formula 

^ heating surface 

and for locomotive valves with flat valve-seats, the formula 

^ heating surface 

^ = "°^^ LXP 



SAFETY VALVES 137 

4 He had checked 1000 boilers and found the average constant to 
be 0.0441 for present practice. Included in the 1000 boilers, however, 
are a number which evidently have valves of insufficient size, as the 
constant in their case is only 0.024. Eliminating this class of boilers, 
the constant for average practice is about 0.05, as given in the formula, 

H. C. McCakty^ said that his experience had clearly proved that 
safety valves with U7iusual discharge, resulting from increased lift of 
valve, cause a violent disturbance in the water level, especially on the 
large modern locomotive boilers, and in proportion to this disturbance 
is the volume of water passing the throttle valve, and hence to the 
steam chest and cylinders, increased. Railroads will be relieved of 
many expensive repairs by reversing these conditions, and thus pro- 
duce the driest steam possible for the engine. To this end, the 
throttle valve, as is well known, is located at the highest possible 
point in the boiler. Further to secure greater locomotive efficiency 
in this direction, the safety valve should be at as high a point on the 
boiler as clearance limits permit, and with an independent short con- 
nection of ample dimensions to the boiler. 

2 Mr. McCarty said his further observations have been that the 
location of the valve on a boiler has much to do with the normal crest 
of the water. Air-brake shocks in train and similar effects, in conjunc- 
tion with high-lift valves, have been a frequent cause of locomotive 
failures through the combination of undesirable conditions, all of 
which cause a greater volume of water to pass through the throttle 
valve and safety valve. 

3 In the speaker's experience in locomotive service he had never 
had even a suggestion of the necessity or the advisability of increas- 
ing the lift of the valve; on the contrary, the reverse condition, from 
a service standpoint, presents itself. The possible limited economy in 
first cost of a slightly smaller valve having increased lift, to accom- 
phsh increased discharge, compared with the next larger size valve 
with normal lift, is deceptive, as the short life and expensive mainte- 
nance of the high-lift valve make it not only an expensive burden to 
the railroads, but an unreliable device. 

M. W. Sewall said that if the evaporative capacity of the boiler and 
the delivery capacity of the safety valve were adapted to each other, 
no difficulty need arise from the use of high-lift valves. The areas 

'President Coale Muifler and Safety Valve Co., Baltimore, Md. 



138 DISCUSSION 

of approach to the valve and discharge from it should then be such 
as practice has already shown to be essential. The usual diameters 
have been mentioned in the discussion as if the}- could not be changed, 
and the high and low lifts have been spoken of as related to those 
diameters. As the high-lift valve has a greatly increased discharge 
capacity, however, it should be reduced in diameter and an entirely 
new adaptation of diameters to " pounds of steam discharged " 
should be made. A manufacturer could then adopt any desired com- 
bination of diameter and lift and the valves would be rated on the 
pounds of steam delivered per second. 

A. A. Caki:, in discussing safetj'-valve springs, said that the ratio 
between the pitch diameter of the spring and the diameter of the wire 
composing it should not be less than 5 to 1, but 7 to 1 is a better mini- 
mum proportion. A pop- valve spring should not be wound to a 
smaller proportion than 7 to 1, and with such a spring coiled to a 
smaller ratio he had found a considerable breakage resulting. 

2 One matter deserving careful attention in the design of pop- 
valve springs is the shape of the section of wire used. Unquestion- 
ably, the best and safest wire for springs is that of round section. The 
principal stress occurring in the wire of a helical spring is that of 
torsion, and in a wire of square section the greatest fiber stress occurs 
at the corners of the square, which are the most distantly removed 
from the center of the section. 

3 The only advantage gained by the use of square-wire springs is 
a slight reduction of the space required for a spring having the same 
resistance to compression. 

4 The most durable of all is the helical spring designed to resist 
extension, known as an extension spring. When this spring is prop- 
erly applied, the load is carried directly along the line of the spring's 
axis, thus doing away with the "buckling" which so frequently im- 
poses harmful bending strains (in addition to the torsional strain) 
in the wire composing compression springs. The use of compression 
springs for pop valves has become almost universal, but there is 
no reason why extension springs of good design cannot be used for 
this purpose. 

F. L. DuBosQUE thought that the formula in the U. S. marine 
laws has the serious defect that some of its factors are left to the 
opinion of any one of a great number of persons concerned in its use. 
The factor W is made up of two quantities, the calorific value of the 



SAFETY VALVES 139 

fuel and the amount burned per square foot of grate surface, and the 
\ahio of these factors can with reasonable judgment be varied so as 
to vary the size of the safety valve at least 50 j)er cent. It is now im- 
possible for a tlesigner to specify the size of a safety valve on a marine 
boiler without first obtaining from the United States inspector his 
opinion on the value of these two factors, notwithstanding the fact 
that the inspector who is compelled to decide this question cannot 
i:)ossibly have as much information to assist him as the designer. 

2 This new formula, therefore, has not in any way improved the 
Rules of the Steamboat Inspection Service and, as pointed out above, 
has added only a complication. As to results produced by it, it is 
easy to see that by selecting proper proportions for the two factors 
that make up W, — and these factors both may be within reasonable 
limits, — the same result will be obtained as by the old formula. The 
old formula at least gave the designer a certain basis to work on, and 
if he was designing his work with the proper regard for safety he had 
the privilege of deviating from the formula if he felt it did not pro- 
vide a valve of large enough size. 

3 This new formula is also similar for cylindrical and water-tube 
boilers. Practical operation shows that a safety valve on water-tube 
boilers should be much smaller than on cylindrical boilers of equal 
evaporative power. A sudden release of steam pressure in a water- 
tube boiler with its limited water-line area causes more damage by 
lifting the w^ater within the l:)oilef than can be caused by a moderate 
increase in steam pressure. 

L. D. LovEKiN, in replying to the remarks by Mr. DuBosque, said 
that he was not aware of the trouble he had caused marine engineers, 
and still further, he saw no reason for such trouble. He had dis- 
cussed the matter fully wnth a number of engineers and showed them 
the new formula which he proposed submitting to the Board, and all 
agreed that his formula was based upon common sense. 

2 Any safety valve based on one square inch of opening for three 
square feet of grate area for a Scotch boiler, and one square inch of 
opening for six square feet of grate area for a water-tube boiler, is 
absurd, and yet this was the formula used by the United States 
Inspectors for many years. 

3 Xelson Foley, of England, states that safety valves may be 
made capable of hfting, say | of their diameter; that a high lift is 
useless and may be an evil if anything gives way; that the waste steam 
pipe, when not under the Board of Trade rules, may be equal in area 



140 DISCUSSION 

to the opening with the lift just mentioned, i. e., the area of the waste 
steam pipe would be one-half the gross cross-sectional area of the 
valve. 

4 Our United States Navy Steam Engineering Department, with 
all their experience in connection with boilers, have agreed with sev- 
eral prominent authorities abroad on a lift of | the diameter of the 
valve. It does not follow, however, because a valve has provision 
for a lift equal to | of its diameter, that it ever lifts this amount. It 
is simply a provision in case the valve is required to be lifted by the 
safety-valve hand-operating gear usually provided on all ships. 

5 The area of waste steam pipe on all our recent naval vessels is 
made | the gross cross-sectional area of the valve, which accords 
with the statements of Mr. Foley. 

6 It is a coincidence that while the present rule might give an ex- 
cessive lift on sizes above 4|-in. diameter, it averages up closely to the 
sizes recommended by many manufacturers for valves below 4|-in. 
diameter. The rate of evaporation of 180 lb. in the present rule 
almost coincides throughout with the Board of Trade formula for 
safety valves under natural draft. 

7 It appears therefore that the Board of Trade thought it wise to 
keep all boilers under natural draft at the same rate of evaporation, 
i. e., all boilers worked under natural draft are assumed to be capable 
of evaporating 180 lb. of water per square foot of grate surface, which 
seems to be a safe maximum rate for any marine boiler under natural 
draft. 

8 When forced draft is used, under the Board of Trade regulations, 
the area of the safety valve must not be less than that found by the fol- 
lowing formula. 

(estimated consumption of coal\ 
per square foot of grate, in I = area of valves required, 
pounds per hour -=-20 / 

A equals the nominal area of the valve, based on its diameter, as found 
from the table of safety valve areas under the Board regulations. 

9 The results of experiments on safety-valve lips illustrated in 
the figures may be of interest to the members of the Society. These 
experiments were made by Nelson Foley to determine the effect of 
adjusting the lip on safety valves. 

Nathan B. Payne believed the most important point brought out 
by the discussion was that there is no proper standard of measurement 



SAFETY VALVES 



141 



rJK-2 Fig.3 




Fig. 1 Results of Experiments on Safety-Valve Lips 



VALVES ROSE AT 81 LB. AND LIFTED ABOUT ^ IN. RATIO OF VALVE AREA TO 
GRATE AREA § SQ. IN. TO 1 SQ. FT. FIG. 1 VALVE CLO.SED IN J MIN. AT 79 LB., 
AND VIBRATED CONSIDERABLY. FIG. 2 BLEW STEADILY, WITHOUT CLOSING. 
FIG. 3 CLOSED IN 1 MIN. AT 80 LB. PRESSURE DROPPED STEADILY. FIG. 4 
SAME AS IN FIG. 3. CLOSED IN f MIN. AT 79^ LB. LESS VIBRATION THAN 
IN FIG. 1. 



142 DISCUSSION 

for the safety valve's capacity. Whether a high-lift or a low-lift 
valve is selected, what we must have is some way of measuring the 
relieving capacity. When we buy a 4-in. valve, for instance, we want 
to know whether that valve has relieving capacity for a 100-h.p. or 
a 200-h.p. boiler, or what size it is suited for. 

2 We have been thinking with regard to the relieving capacity 
of safety valves that we need consider only one dimension, but it is 
absolutely impossible to determine the amount of relieving capacity 
in a given time without knowing the lift of the valve off the seat, 
so as to get the effective area of opening. The question for the 
user to decide is how much relief he can get from a given make and 
size of valve. If one maker offers a safety valve having :^-in. lift, 
and another offers a i-in. lift, each should state how many pounds 
of steam per hour his valve will relieve. 

H. 0. Pond said the question of high-lift and low-lift valves seemed 
to be one simply of capacity. If the low-lift valve will deliver a cer- 
tain number of pounds of steam at a given pressure and temperature, 
and its capacity under these conditions is known, this is the principal 
thing required. The same test applies in the case of the high-lift 
valve, the essential point, however, being to know how much steam 
the valve will discharge. Undoubtedly a high-lift valve will give 
more capacity than a low-lift valve having the same diameter of 
opening. This being so, we could use a smaller valve of the high- 
lift type, which would be an advantage in many ways. 

F. L. Pryor summarized the results obtained from tests made to 
obtain the blowing-off pressure of safety valves when tested with 
water and with steam. 

2 A standard 4-in. pop safety valve, set for 125 lb., was mounted 
on a 4-in. pipe and so connected that either steam pressure or water 
pressure could be admitted to the valve. One set of tests was made 
over a period of 15 days, the test of one day being with steam and the 
following day with water. In a second series of tests, the valve was 
tested at three different settings on the same day, viz. 104, 131 and 
159 lb. The third series of tests was made with the valve at a num- 
ber of different settings, from 105 to 165 lb., one measurement being 
made directly after the other and no precaution taken to insure that 
the valve had returned to its normal temperature after the preceding 
test, except that before operating with water pressure a considerable 
amount of water was flushed through the valve. 



SAFETY VALUES 143 

3 The results obtained in all the tests were in practical agreement, 
antl indicated that the l)lo wing-off pressure with steam and with water 
did not differ to any great extent, although the pressure to blow off 
with water was higher than with steam. 

4 In the case when the valve was allowed to cool 24 hours, the 
water pressure required to open it was about 3| lb. higher than the 
steam pressure. In the tests where the valve was cooled thoroughly 
with water, the pressure with water was about 3 lb. higher than with 
steam. In the rapid change test the water pres'sure amounted to 
about 2.6 lb. more than the steam pressure. 

5 In all tests the steam and water pressure record was that at 
which the valve was in full operation. In the case of the steam pres- 
sure test there were two distinct points below full open pressure which 
could also have been noted: when the valve began to leak, which 
occurred about 2 lb. below the final blowing-off pressure, and when 
the rate of flow suddenly increased, which was about 1 lb. below 
maximum. 

A. B. Cakhart, speaking on the proportions of safety-valve parts, 
said that the specifications which require valve seats to be made of 
non-corrosive metal, and the rules which compel every valve to be 
tried and lifted by the lever every day, aim to overcome the ever- 
present danger that the valve may stick upon its seat and fail to open 
at the critical moment. But the greatest cause of the sticking of the 
valve, when it does occur, is not corrosion of the seat face, but the 
binding friction of the disc-guides against the side of the well or 
throat of the valve. This cocking or binding effect can be decreased 
by any modification of design which will reduce the diameter of the 
cylindrical guide, or which will bring the guiding surface close to the 
plane of the seat, both of which will reduce the moment of the friction 
or cocking stress. 

2 Any device which reduces the lift of the disc and the spring 
movement to the least possible amount will also reduce the eccentric 
spring action and its effect, and any valve design which contemplates 
an unnecessarily large lift or compression disadvantageously magnifies 
this effect. 

3 An early and still common form of safety valve has the seat 
opening beveled at an angle of 45 deg. The effective steam passage 
is therefore measured by the sine of 45 deg., which is approximately 
only 0.7 of the actual compression of the spring when the valve 
opens, so that the spring must necessarily compress about Ih times 



144 DISCUSSION 

the effective lift. Even this does not always afford a free passage for 
the steam tojthe air where there is vertical overlap of the regulating 
ring against the lip of the disc in order to increase the lift against the 
greater pressure of the shortening spring. 

Edw. F. Miller presented the following method of obtaining the 
valve discharge area based on the rate of fuel consumption. The 
weight of steam flowing through an orifice with a slightly rounded 
entrance may be figured quite accurately by Napier's formula (some- 
times called Rankine's formula). Its accuracy for commercially 
dry steam has been shown by tests made under pressures varying 
from 30 to 150 lb. There are a number of papers on this subject in 
the earlier volumes of the Transactions. According to the formula 
the weight of steam discharged per second through an orifice with 

FP 

slightly rounded entrance is — where F is the area of the orifice in 

square inches and P is the pressure in pounds absolute on one square 
inch. 

2 The discharge per second through an orifice with a sharp edge 
at the entrance, as would be the case in a safety valve, has been found 
from actual tests on valves to be 0.95, the amount figured from this 
formula. The opening needed in a safety valve may be figured as 
follows: 

G = grate area. 

U = rate of coal consumption per square foot of grate per 
hour. 

9 = probable evaporation per pound of coal under actual 
conditions. 

G X iil X 9 

= weight of steam made per second. 



3600 
Equate this to the preceding expression and solve for F: 

GXRX9 ^ FXP 

3600 ' 70 

p^GxBX9X70 
3600 X P X 0.95 

3 The area of the opening through a safety valve is equal to 
the inner circumference of the seat times the effective lift. For a valve 
with the seat at an angle the effective lift is equal to the lift multi- 
plied by the cosine of the angle the seat makes with a horizontal. 



SAFETY VALVES 145 

4 For a 45-deg. angle the effective lift is 0.707 X lift. Calling 
D the inner diameter of the valve, the opening is 

~ X D X lift X 0.707 

Substituting this for F, 

.i)X lift X 0.707= ^X^><^X^O 
3600 X P X 0.95 

If the lift of the valve is -j^,, in., 

^^ GX/2X9X70 _ GR 

3600 X P X 0.95 X ^ X 0.707 X 0.1 ~P X 1.206 

5 If the lift is 0.05 instead of 0.10, then the valve diameter D 
is doubled. Doubhng the pressure will make the same valve with 
the same lift take care of double the weight of steam. For illustration : 

Grate area = 2'". 

Coal consumption = 18 lb. per square foot per hour. 

Pressure = 120 lb. absolute. 

^^ ^X18_3^ 
120 X 1.206 

Pressure, 150 lb. absolute. 

Grate area, 50 sq. ft. 

Coal consumption, 25 lb. per square foot per hour. 

50X25 



150 X 1.206 



= Z)=6.9in. 



A valve as large as this would be replaced by two of equivalent capacity. 
The circumference = 3.14 X 6.9 

6.9 
Two smaller valves of diameter '- = 3.45 will give the same cir- 

cumference and the same discharge with the same lift. 

George H. Musgrave^ The function of the safety valve is two- 
fold: (a) it gives notice of the highest pressure permissible; (6) it 
gives the alarm that more water or less fuel is needed. He had been 
told by engineers in the marine service, that through the use of safety 
valves with excessive lift and quick discharge, their engines had been 
plugged by taking over water. He had known of numerous occasions 
in locomotive service where there have been very disastrous results. 

'General Sales Agent, Star Brass Mfg. Co., Boston, Mass. 



146 DISCUSSION 

2 If the same principle is to be introduced in high-lift locomotive 
safety valves that is now used in injectors to raise water, what is to 
prevent the syphoning of the water to the throttle valve, and its 
flowing through the dry pipe and into the cylinders? From his long 
experience, originally in locomotive service, afterwards in marine and 
stationary service and at the present time, on safety valves for all 
uses, he would suggest that the medium-discharge type is the safest 
and most satisfactory valve to use. Any valve that will materially 
disturb the water level and have a tendency to raise it is dangerous. 

M. W. Sew ALL, speaking on exact regulations for valve propor- 
tions, said that the public authorities and insurance companies 
should establish means of regulation in regard to the following : 

a Flange diam.eters for various rates of discharge. 

b Requirements as to minimum discharge of pounds of steam 
per second within given ranges of pressure. 

c Requirements as to non-corrosive seats or other operating 
parts, strength of parts, means of operation by hand, and 
security against being put out of adjustment by ill-dis- 
posed persons. 

Albert C. Ashton was opposed to high-lift valves, since they open 
and close so suddenly as to injure the boiler and its connected fittings, 
as well as the valve itself. 

2 He thought that any revised rule for the size of pop safety valves 
should not prescribe a capacity of relief that could be obtained only 
with a high-lift valve, as suggested by Mr. Darling and Mr. Lovekin, 
who approve of a lift equal to J5 of the diameter of the valve. 

A. F. Nagle presented the following table on the size of safety 
valves for boilers of a given power, based upon the following data: 

a A boiler horsepower is the term used to express the evapora- 
tion of 34.50 lb. of water per hr. from and at 212 deg. Fahr. 

b A spring safety valve can and should be depended upon to 
lift M of its diameter. 

c The flow of steam follows closely Napier's formula, reduced 
to 92| per cent by Mr. Darling's experiments (Par. 24). 



SAFETY VALVES 



147 



r/ The fornmla usod in the computation is h.p. = 0.0951 D'-P, 
transformed from Mr. Darling's formula, where 

h.p. = boiler horsepower. 

D = diameter of valve in inches. 

P = absolute steam pressure. 

2 In using this table, allowance must be made for what is likely 
to be the maximum horsepower of the boiler and not its normal rating. 
Fifty per cent overload is not unusual, and double the rating, while 
not impossible, is not liable to pass through the safety valve. 

HORSEPOWER OF BOILERS AND SIZE OF SAFETY VALVES 



Steam 






Safety Valves 






Pressure 
Pounds 












2 In. 


2Jln. 


Sin. 


3J in. 


4 in. 


4iln. 


100 


44 


68 


98 


134 


175 


221 


125 


53 


83 


120 163 


213 


269 


150 


63 


98 


141 I 192 


251 


318 


175 


72 


113 


162 221 


289 


366 


200 


82 


128 


184 


250 


327 


414 


225 


91 


142 


205 


280 


365 


462 


250 


100 


167 


227 


309 


403 


510 



Note. — Roughly every 4 lb. of coal burned per hour represents one boiler h.p. 

Jerome J. Aull thought that the proposed rule should include a 
term for a fixed lift rather than a variable one, for the reason that with 
the latter would result a hopeless confusion of safety-valve openings 
in boilers of the same size. Thus under Mr. Darling's rule a boiler 
of a certain size might be provided with a safety-valve connection 
varying from 2| in. to 4 in. in diameter, depending upon the make of 
valve specified. It would be far more convenient and satisfactory 
to standardize safety-valve connections so that any valve having the 
required capacity could be used. To do this it would be necessary 
that the valves themselves be standardized within certain set limits 
and this could be done only by a body of disinterested engineers, 
properly authorized to investigate the subject. 

2 Mr. Aull condemned high lift as it made the seats and spring- 
bearings subject to a severe pounding action; there is more danger 
of chattering; close adjustment is not possible; there is danger of 
lifting water; and th e boiler seams are sometimes strained to the open- 
ing point. 



148 DISCUSSION 

Philip G. Darling, in closing, dealt with the many different values 
at present being advocated for safety-valve lifts. Recent articles 
place this maximum limit variously at 0.05 in., 0.06 in., 0.08 in., 0.09 
in., and 0.14 in., for the same size valves. 

2 It is well known by those in t!ouch with foreign manufacturers 
that valve lifts, spring compressions and other valve elements which 
are radically different from what has been the general practice in 
this country, are being used successfully, and in some places univer- 
sally. 

3 Two cases will illustrate this. The springs on 3|-in. triplex 
valves of the Thornycroft design, used widely in English marine 
practice, are not only of the exposed type, but have, when set for a 
designed pressure of 250 lb., a compression of 4 in. These are regular 
safety valves of the same principles as our own duplex valves. To 
those who would condemn a compression of § in. to | in. as radically 
high and unsafe this instance should be suggestive and help to broaden 
their conceptions of the possibilities of safety-valve design. Again, 
in London Engineering, February 26, 1909, reprinted in Power, March 
30, 1909, J. H. Gibson tells of exceptionally good results obtained in 
a valve having 0.21-in. lift. He says: "We think we are justified 
in the assumption .... that anything tending to reduce the 
size of these important fittings (safety valves), which have been 
growing to abnormal proportions of late, is a step in the right direc- 
tion. " 

4 High lift is not synonj^mous with excess safety-valve capacity. 
A boiler's evaporation absolutely determines the necessary safety- 
valve capacity. In a given boiler the pounds of steam per hour which 
the valve should be able to relieve can be definitely figured and all 
that is further needed, in making the correct valve specification, is the 
capacity of the safety valves. 

5 It is not a question of lift for itself, but of requisite relieving 
capacity, and if this is obtained with a 3-in. instead of a 4-in. or 4|-in. 
valve there is a positive, real advantage, not only in original cost but 
in the maintenance and better action of comparatively small rather 
than large valves. 

6 It is thus not a uniformity in the lifts of different valves which 
the engineering public should demand, but rather the practice of 
stating relieving capacities, based on the actual lifts existing in the 
valves themselves. If the capacities were stamped upon the valves, 
as already done by one maker, it would give a rational basis for use 
in the application of safety valves to boilers. 



SAFETY VALVES 



149 



7 It has been objected that capacities thus published could not 
be verified without actual capacity runs, such as the Barberton tests 
recorded in the paper, on the ground that in some valves the effective 
area of discharge at the seat, upon which the formula is based, is not 
the smallest discharge area; or even if it is, that there is a material 
throttling or holding back of the steam flow. Valves containing 
the original Richardson adjusting ring have been cited as designs in 
which this choking occurs. 

8 In order to secure information upon this matter prior to con- 
ducting the direct capacity tests at Barberton referred to in the paper, 
the effective discharge areas at the seat and at the most contracted 
passage between the lip and adjusting ring were figured and plotted 
for the different valves tested at different lifts. Further, a 3^-in. 
valve was constructed having this Richardson ring and projecting 
disc lip design, and for the same valve another disc and ring in which 
the projecting Hp was cut entirely away. In the former the discharg- 
ing steam was deflected through practically 90 deg., and in the latter 
the steam had a free straightaway passage. These two designs were 
radically different and fairly represented the extremes of what on the 
one hand seemed to be a choked or impeded steam discharge passage 
and on the other a free open one. 

9 The most effective discharge areas of the two taken at the seat 
and at the most contracted part of the passage between the lip 
and rings are given in the table in square inches for different lifts. 

TABLE 1 EFFECTIVE DISCHARGE AREAS 



Valve with Projecting Lip and Richardson Ring 



Valve without the Lip 







Most Contracted 




Most Contracted 


Lift 


At Seat 


Point Beyond 
Seat 


At Seat 


Point Beyond 
Seat 


0.02 


0.16 


1.20 


0.16 


2.01 


0.06 


0.47 


1.40 


0.47 


2,14 


0.10 


0.79 


1.76 


0,79 


2,29 


O.U 


1.11 


2.27 


1.11 


' 2.43 



These areas, taken with Napier's formula, give a method of figuring 
the theoretical pressure existing in the "throttling chamber " under the 
disc lip; that pressure being to the boiler pressure as the effective dis- 
charge area at the seat is to the most contracted area between the lip 
and ring beyond. The highest pressure thus indicated in the throt- 
tling chamber is less than 50 per cent of the corresponding boiler pres- 



150 DISCUSSION 

sure. This pressure in the "throttHng chamber" being the discharge 
pressure of steam passing over the valve seat, and the full flow of 
Napier's formula being practically unaffected by any discharge pres- 
sure less than 60 per cent of the original or boiler pressure, the theoreti- 
cal conclusion is that the discharge from neither of these valves would 
be affected by the disc design or discharge areas outside of the valve 
seat. 

10 In replying to the references made to disastrous results to boilers 
such as the opening up of seams and fittings due to the sudden release 
or cutting off of steam by the safety valve, Mr. Darling discussed 
the sudden change of pressure due to opening and closing of throttle 
valves and blow-off valves, concluding that the shock to the boiler 
from this source would far exceed that due to the closing of the safety 
valve. 

11 The larger the safety valve compared with the boiler the 
greater the shock to the boiler due to its action, if such shock exists. 
A 5-in. valve mounted directly upon a94-h.p. test boiler would increase 
or accentuate this tendency to strain over say a 3i-in. valve on an 
800-b.h.p. locomotive surely 12 or 13 times. Yet with a most sensi- 
tive boiler pressure test gage graduated to pounds and mounted upon 
this 94-h.p. test boiler, absolutely no recoil of the gage hand upward 
either at the opening or closing of a 5-in. valve is perceptible. It 
would seem that some increase of pressure such as would be indicated 
upon the gage would be positively necessary to transmit a strain to 
the boiler. 

12 Two cases had recently come to his notice in which loco- 
motive safety valves had loosened from their spud connections 
and had blown off while the boiler was under its full steam pressure. 
One was a 3i-in and the other a 4-in. valve, which therefore opened 
areas of 9.6 and 12.6 sq. in., respectively, while the maximum cor- 
responding safety-valve discharge area could be but a little over 
one square inch. Yet no damage to the boilers was experienced. 
The blowing off of 2-in. locomotive whistle connections had been 
cited as a not infrequent occurrence. The steam-relief in such acci- 
dents is of course more sudden than with a safety valve, and the 
fact that this opening of ten to twelve times the maximum discharge 
area of the corresponding safety valves results in no further incon- 
venience than the replacing of the fittings raises some question as 
to the actual disaster impending in the use of valves having a dis- 
charge area of but 1 sq. in. 



No. 1234 

A uniqup: belt conveyor 

I'v E. (.'. SopER, Detroit, Mich. 
Member of the Society 

It is quite possible that a description of a belt conveyor a quarter 
of a mile long, and requiring more power to operate empty than 
loaded, will be interesting to some of the members and since its 
installation and operation are at variance from the prescribed rules 
of conveyor design, we beg to submit the following: 

2 The belt conveyor was built during the summer of 1908 in one 
of the large portland cement plants of the South. It consists of a 
24-in. 8-ply canvas belt in two sections, one section about 1000 ft. 
between centers, and the other with 1100 ft. between centers, its 
function being to convey the shale used in the manufacture of the 
cement, from the shale quarry to the plant. The shale deposit is 
located on a mountain about 247 ft. above the shale storage tanks, 
as shown in profile, Fig. 1. The two sections intersect at an angle 
of 140 deg. 40 min., so that the blasting from the limestone quarry 
does not interfere with the operation of the belt. The belt conveys 
the shale around the limestone quarry, as shown in plan. Fig. 1. 

3 The belt is flat and carried by rollers, the top row having 4 ft. 
between centers and the return idlers 12 ft. between centers. Guide 
rollers are placed with about 40 ft. between centers along both upper 
and lower belts. (See Fig. 2.) The majority of manufacturers of 
belt conveyors recommend the maximum length between centers of 
a single belt to be about 700 ft. to 800 ft. 

4 Referring to Fig. 1, the belt conveys the material down-hill, 
and to this fact is due the apparently parodoxical results in power 
required to operate, shown in Tables 1 and 2. 

5 Because of the extreme length of the belt, and the fact that 
there is no roof or other covering, it was necessary to install some 
system for taking care of the expansion and contraction, in addition 

Presented at the Spring Meeting, Washington, May 1909, of The American 
Society of Mechanical Engineers. 



152 



A UNIQUE BELT CONVEYOR 



to the ordinary stretch of the belt, which is taken up in the majority 
of installations by 24-in., 36-in."^or 48-in. takeups, according to 
length of belt. A set of 36-in. takeups, (Fig. 3) was installed at 
the upper end of each of these belts to maintain alignment and equal 























Shale Quarry 


400 




















^'^1.446.65 
























300 
















^ 


















El.268^38^^^ 










200 








El.227.73 

1 


"ElTlolT 











10 12 14 

ELEVATION 




PLAN 
Fig. 1 Profile Showing Elevation and Plan of Conveyors 

tension on each edge of the belt. The system installed acts as a 
tension carriage and makes it less often necessary to cut out the 



-4'o— 



Tojp or Carrying Belt \ 



Lower Belt- 




FORWARD IDLERS 



-36'^ 



RETURN IDLERS 



Fig. 2 Details op Forward and Return Idlers 



slack in the belt , and in cool and wet weather the belt adjusts itself, the 
increased tension due to contraction raising the weight in the tower. 
A 10-h.p. motor drives each section. The lower section has a 6-ft. 
drop and requires approximately 5. Ih.p. to operate empty and 5. Ih.p. 



A UNIQUE BELT CONVEYOR 



153 




« 
o 

s 

O 



O 






154 



A UNIQUE BELT CONVEYOR 




Fig. 4 View of Discharge from Fig. 5 Looking Down onFirst 



Upper to Lower Section 



or Lower Section 




Fig. 6 "^Side View of Upper Section 



A UNIQUE BELT CONVEYOR 



155 



to carry a load of 1200 lb., as shoveled by ten men. (See tests which 
follow.) The discharge from the upper to the loAver sections through 
a chute is shown in Fig. 4. There is no spilling of material at any 
point of the travel, and pieces of shale a cubic foot in size are carried. 
The upper section is driven, contrary to practice, at the upper end, 
the pull being on the lower or slack side of the belt, but in this case, 
due to the pull of gravity on the top side, the belt was found to work 
better with the pull on the lower side. 




Fig. 7 General View Showing Both Belts 



6 The several halftones give views of the belt taken from different 
points. In clearing a way through the woods, the poles obtained were 
utilized for trestling and the planking was obtained from the scrap 
pile of concrete-form lumber. 

7 Fig. 6 is a side view of the lower end of the upper section, show- 
ing the two depressions in the belt, and though these depressions do 
not conform closely to the prescribed radius of 300 ft., there is no 
lifting of the belt from the carrying idlers. 

8 Power tests were made on the two sections after the belt had 
been operating a few days, with the following results; the speed of 



156 A UNIQUE BELT CONVEYOR 

the belt of the lower section, which has a grade of 2.4 per cent for 665 
ft., or 0.024 X 665 = 16 ft., was 146 ft. per min.; the belt was driven 
by a 10-h.p. direct-current Westinghouse motor, and was loaded 2.2 
lb. per ft. for a distance of 550 ft., or 1210 lb.; this load fell 16 ft. in 
5 min. Then 

= 3520 ft.-lb. of work exerted bv load 

5 

or, 

= — h.p. (approx.) helping to pull the belt. 

33,000 11 ^ ^^ ^ ^ ^ 

When the belt was loaded as above, a test of the motor showed that 
16 amperes, 239 volts, or 5.1 h.p., were required. There was no 
appreciable difference in the ammeter and voltmeter readings, when 
belt was empty or loaded, as in test. 

9 When the belts were installed, after trying them out and ascer- 
taining how easily they could be operated, a sprocket was placed on 
the tail-shaft of the lower section and also one on the head-shaft of 
the upper section, and the two sprockets were connected by a vertical 
quarter-twist chain. The idea was to drive both belts by a 10-h.p. 
motor at the head of the lower belt section, after all shafts had 
become well seated in the bearings and the stiffness had disappeared 
from the belt and it was in good operating condition. This was also 
necessary in order to take up the slack in the upper section when 
starting, and the speeds were such that the top side of the belt ran 3 ft. 
per min. faster than the lower side. The results of a series of tests are 
given in Tables 1 and 2. 

TABLE 1 POWER TESTS OF BELTS UNDER CONDITIONS NOTED IN TEXT 



Time 


Volts 


Amperes 


Watts 


H.p. 


Notes 


(A.M.) 












9:50 
10:08 


207 
210 


12 
12 


2484 
2520 


3.3 
3.3 


/ Belts chained together 
\ Eight men loading 


10:09 


208 


14 


2912 


3.9 


/Connecting chain off, 10-h.p. 


10:11 


210 


14 


2940 


3.9 


1 motor only 


10:15 
10:20 
10:35 


200 
200 
200 


14 
15 
16 


2800 
3000 
3200 


3.7] 
4.0[ 
4.2J 


Gradual increase in electrical load 
due to decrease in shale load 



Note: Low voltage due to very small mains and long distance (2500 ft.). 



A UNIQUE BELT CONYEVOR 



167 



TABLE 2 SECOND SERIES OF TESTS 



Time 

(P.M.) 


Volts Amperes 


Watts h.h. 

i 


Notes 


2:00 

2:15 
2:25 
2:35 
3:45 

3:50 


194 

180 
182 
186 
195 

185 


16 

16 
18 
18 
14 

19 


3104 

2886 
3275 
3348 
2730 

3515 


4.1 

3.8] 
4.4 [ 
4.4J 
3.6 

4.7 


Empty. Connected to lower belt by 
chain 

All readings at motor and not in- 
cluding line loss 

Loaded by seven men 

Loaded as before, but with connect- 
ing chain off. 10-h.p. motor only 



Note : Readings taken on motor at upper end of upper belt-section. 

Initial and Operating Costs 

10 Tables are given herewith upon the first cost of the equipment 
(Table 3) and the cost of operation and maintenance (Table 4). 
Table 4 is based upon a capacity of 200 tons conveyed in ten hours. 
Inasmuch as the capacity is directly proportional to the speed, if it 
was desired to increase the capacity of the conveyor, it would only 
be necessary to increase the travel of the belt per minute, and from 
experience, it is quite possible that by doubling the load the power 
required to operate would be reduced 50 per cent. 

11 The operation costs given in Table 4 are taken from actual 
practice. Doubling the capacity per day and assuming above costs 
to be approximately the same -reduces the actual cost of conveying 
to S0.0038 per ton. Interest and depreciation, $0.0063, or a total 
of $0.0101. 



TABLE 3 COST PER FOOT OF COMPLETED BELTS INCLUDING ELECTRICAL 
MOTORS. TRESTLING. ETC. 



Uatebials 


Total Cost 


Cost pbb Ft. 


Lumber 


S 496.34 
5361.52 

1435.77 
637.11 
193.20 
962.20 


$0,238 


Belt 


2.58 


Castings 


0.69 


Electrical equipment, including two 10-h.p. motors 

Miscellaneous: nails, bolts, screws, iron, etc 

Labor 


0.316 
0.093 
0.46 








S9106.16 


$4.37 



Note: Length of first section, center to center, 998 ft.; second section, 1082 ft.; total, 
2080 ft.; takeup, 15 ft. 

Cost of castings includes machine work, etc. 



158 A UNIQUE BELT CONVEYOR 

12 Regarding the operation of the belt: after the stiffness had 
disappeared there was very httle slipping at the head or drive pulleys, 
and there was sufficient lubrication in the shale itself to form a water- 
proof covering about J-in. thick on the belt, thereby protecting it not 
only from wear but from the action of the elements, and proving a 
very good dressing to keep the belt pliable. Because of the slow 
speed, etc., there are very few repairs necessary to the belt, and in 
this instance, being coated as described above, the belt should last 
several years. 

TABLE 4 COST TO OPERATE AND MAINTAIN BELT CONVEYOR 

PER 10 HR. pgjj ^Q^ 
DAY 

Power 

lOh.p. at$0.004perh.p.-hr $0.40 $0,002 

Labor 

Boy oiling, etc $0 . 75 

Taking up slack once in 7 days, 2 men, 3 hr. at 

$0.20perhr 0.171 0.92 0.0046 



Supplies 

Belt Lacing . 10 

Waste, Resin, etc 0.10 0.20 0.001 



Total $1.52 $0.0076 

Oil (no charge, using waste oil from large crushers) . 

Interest, etc. 
Interest, Depreciation, Renewals, 10 per cent on 

investment of $9200 2.52 0.0126 



Grand Total $4.04 $0.0202 



DISCUSSION 



T. A. Bennett. Mr. Soper's paper, while giving a practical descrip- 
tion of a certain installation, hardly seems to warrant the word 
" unique." There are many conveyors in regular practice just as 
long — conveyors which run downhill — in fact, conveyors that need 
a brake. As for size, there is a 36-in. movable belt conveyor in New 
York over a thousand feet long, used for filling in the refuse from the 



A UNIQUE BELT CONVEYOR 159 

city on Hiker's Island. The driving arrangement mentioned, from 
the receiving end of the conveyor, is also common practice. 

2 Regarding the maximum length of conveyors, with a flat belt, 
as in tliis instance, the limit would be merely the cost of installation as 
practically any tensile strength desired can be obtained by increasing 
the number of plies of the belt. With a troughed belt the hmit of 
length would be the tensile strength of the thickest belt that would 
conform to the trough of the idlers. This limit approaches somewhat 
the Umits the author mentions, although such belts have been put in, 
as above, for lengths of a thousand feet or more. 

3 The take-up has been in use for over five years in belt-conveyor 
practice. There is one installation at Bilbao, Spain, handling iron 
ore, which runs down an incline of 13 deg. and needs a brake, and has 
a counterweighted take-up working in a vertical plane. The take-up 
and drive are located on the return belt near the foot of the incHne. 

4 The tonnage of the conveyor is so small that the cost of mainte- 
nance per ton is also misleading. The wear of a belt is occasioned by 
the material coming in contact with it when dehvered to it. A 
narrow stream of material permits each particle to come in contact 
with a small proportion of the total width, whereas a wide stream 
utilizes the full width of the belt and furthermore carries a large part 
of the material on top of the belt without ever touching it. I beUeve 
the capacity of this conveyor is something like 20 tons per hour, 
whereas such a conveyor should easily handle 200 tons per hour. 
Therefore the average cost for maintenance of the belt per ton carried 
is high. 

Harrington Emerson. The last words in this paper are "the 
belt should last several years." In the last Hne of Table 4, it is 
stated that interest, depreciation and renewals amount to 10 per cent 
on an investment of $9200. Now, if the belt is to last only a few 
years, 10 per cent is not sufficient to cover interest, depreciation and 
renewal. Assuming that the belt lasts four years, the depreciation 
account alone would be S2300. That would increase the cost per 
ton from $0.02 to about $0,035. 

Fred J. Miller. It might be well to consider that there are other 
things that constitute part of this plant as well as the belt. As I 
understand, it is stated that the belt may last only a few years, but 
the rest of the plant may last enough longer than ten years to make 
10 per cent a fair total charge for depreciation. 



160 DISCUSSION 

The Author.* The belt conveyor in question has now been in 
operation about eighteen months, in which time less than $3 in 
repairs has been expended. The belt itself shows little wear and 
should last ten years. Of course the driving and carrying mechan- 
isms will last indefinitely under ordinary conditions, as there is little 
wearing of the working parts, due to the slow speed of the belt. 

2 The installation, as stated previously, though ample for a 
capacity of 200 tons per hour, is required to carry not over 20 tons 
per hour, and certainly the cost per ton for maintenance and other 
charges is out of proportion to what it would be were the belt carry- 
ing an5rthing like full load. 

3 The motor at the receiving end of the upper belt has been taken 
out and the belt has been driven for several months by the "lower" 
10-h.p. motor, the "upper" belt being driven by the "quarter-twist" 
chain mentioned in the paper. At the time of writing the paper, 
this was to our knowledge, the longest single-driven belt conveyor 
(about 2150 ft.) in operation. 

4 The writer has since learned of a slightly longer belt carry- 
ing grain. Its speed is 1800 ft. per min., and hence a much greater 
power is necessary. As to the power required to operate it, it 
is reasonable to assume that if it takes about 6 h.p. to operate 
the belt when empty, and 3 h.p. when loaded by 20 men, the 
belt will practically run itself when loaded by 40 men, and will 
require a band-brake when loading is increased above this number. 

^ Abstracted. 



No. 1235 

AUTOMATIC FEEDERS FOR HANDLING 
MATERIAL IN BULK 

By C. Kbmble Baldwin, Chicago, III. 
Member of the Society 

In the writer's paper on the Belt Conveyor, read before the Society 
in June 1908, mention was made of the advisability of using some 
type of automatic feeder when feeding a conveyor from bulk, for 
example, from a storage bin. This brief mention of the automatic 
feeder brought so many inquiries for information ^regarding feeders 
for various materials that this paper has been prepared in order to 
present a brief description of the various types now in use. The 
cuts are not intended to show the construction, but to illustrate the 
principle involved, so that they may be compared. 

2 Careful study of this subject reveals the fact that a particu- 
lar type of feeder has been developed in a certain industry, or local- 
ity, and is little used elsewhere. The types illustrated and described 
below are only those which have come under the writer's personal 
observation in many processes and locations within the past fifteen 
years. There may, therefore, be many other types. 

3 When handling dry, free-flowing material such as grain, from 
a storage bin to a conveyor, a feeder is not required, as a simple gate 
may be set to give the desired opening, thus allowing the proper 
quantity to flow from the bin. Should the material be of varying 
size, such as mine-run coal, a simple gate is not satisfactory unless 
constantly attended; even then it is impossible to get the same con- 
stant, regular feed that a properly designed feeder gives. If the gate 
is raised high enough to allow a large lump to pass, there usually 
results a rush of fine material, which floods the conveyor before the 
gate can be closed. The automatic feeder, therefore, not only saves 
the expense of an attendant, but insures a constant and regular feed, 
irrespective of the size of the material. 

Presented at the Spring Meeting, Washington, May 1909, of The American 
SociETT OF Mechanical Engineers. 



162 AUTOMATIC FEEDERS FOR HANDLING MATERIAL 




*-UrtDERCUT G>\T£ 



Fig. 1 Undercut-Gate Feeder 




Fig. 2 Lifting-Gate Feeder 




SHftFT FOR GtftR OR SPRQCKE 



Fig. 3 Screw-Conveyor Feeder 



AUTOMATIC FEEDERS FOR HANDLING MATERIAL 163 

4 Fig. 1 shows the undercut-gate feeder, with a body either of 
cast-iron or steel plate. Pivoted near the top is the undercut gate — 
which is swung back and forth by a connecting rod from crank or 
eccentric. This type of feeder is best adapted to fine-sized, free-flow- 
ing material. Material containing lumps is likely to bridge. As the 
feed is intermittent, the feeder is generally used in connection with 
chain or bucket conveyors, the strokes being timed to feed material 
between the flights, or into the buckets. The capacity may be 
changed only by changing the length or the number of strokes. As 
the length of stroke is more easily changed, it is preferable to use a 
crank with an adjustable throw rather than an eccentric. P^lliptic 
ger.rs are sometimes used to give a quick return, but in practice 
this quick return has not been found of sufficient value to justify 
the t-pi'cial and more expensive gears. 

5 The lifting-gate feeder, shown in Fig. 2, also gives an inter- 
mittent feed and is therefore used only with a chain or bucket con- 
veyor. The chute is hinged, so that when down, the material will flow 
out of the hopper, but when raised above the angle of flow of the 
material, the discharge is stopped. The moving of the chute may be 
accomplished by a connecting rod receiving motion from either crank 
or eccentric. This feeder will handle material regardless of size, but 
it must be free-flowing material, so that it will move by gravity when 
the chute is lowered to the angle of flow. The capacity may be 
adjusted by varying the number of strokes, also, in a measure, by 
increasing the length of the stroke, thus increasingthe maximum angle 
of the chute and causing the material to flow more quickly. 

6 The screw-conveyor feeder, illustrated in Fig. 3, will deliver 
a constant stream of material, but in this case also it must be of such 
a nature that it will flow by gravity to the screw. The capacity can 
be changed only by altering the speed of the screw shaft. This type 
of feeder has a large field in the handling of pulverized material, such 
as coal, cement, etc. 

7 The roll feeder, shown in Fig. 4, is extensively used in the mineral 
industries for handling both large and small materials. The roll is 
so located under the hopper that the material will not flow when the 
roll is stationary, but when rotated it will carry the material forward. 
The capacity is determined by the speed and width of the roll, and 
the thickness of the stream, as fixed by the adjustable gate. 

8 The roll feeder has been successfully used in handling iron-ore, 
coke and stone from the bins to the weigh cars for furnace changing. 
Edison used this type for feeding ore and stone from bins to crush- 
ing-rolls. The disadvantage is the head-room required, owing to 



164 



AUTOMATIC FEEDERS FOR HANDLING MATERIAL 








/ 




♦ 


' 








ROLL-> = 




E 







rsc-i 




j^--**--jr 


= 








.,--^^— -^-^-v-^ 





Fig. 4 'Roll Feeder 

the large roll necessary to satisfactory operation. For handling mine- 
run material, the ^oll should be 6 ft. to 8 ft. in diameter and in many 
cases it is not possible to obtain this space. 

9 The rotary-paddle feeder, Fig. 5, acts not only as a feeder, but 
as a measuring device. It is used for fine material which flows 
readily from the blades. The capacity is fixed by the speed of the 
paddle shaft. 

10 The revolving-plate feeder, shown in Fig. 6, is used mostly 
for feeding stamp-mills. The inclined plate driven by gears, placed 
either above (as shown) or below, moves the material out of the hopper 




Fio. 5 Rotatino-Paddle Feeder 



AUTOMATIC FEEDERS FOR HANDLING MATERIAL 



165 




RtVOLVING Pl-BTt 



Fig. 6 Revolving-Plate Feeder 

where it is scraped off by the skirt-board. When the skirt-board is 
made adjustable, sticky material may be handled by this feeder 
because the curved plate will scrape the material off the revolving 
disc and into the chute. The capacity is fixed by the speed of the 
plate and the location of the adjustable gate. 

11 Fig. 7 illustrates the apron-conveyor feeder used for handling 
material of all sizes. The conveyor may be of any of the various 
types of apron flights, depending on the nature of the material handled. 
The chain should be provided with rollers or wheels traveling on 




CUWVCO ftPBOM FLICHT 



Fig. 7 Apron-Conveyor Feeder 



166 AUTOMATIC FEEDERS FOR HANDLING MATERIAL 

track to prevent the apron from sagging. The capacity is fixed by 
the speed of the apron and the position of the adjustable gate. 

12 The disadvantage of this type is the inherent disadvantage 
of the apron conveyor. If the flights become bent or buckled, 
the material leaks through or catches between them. It has an 
advantage over other feeders in that it may be used to carry the mate- 
rial a greater distance. 

13 A rubber or canvas belt may be used in place of the apron, in 
which case the belt is supported by idlers placed close together. 

14 The swinging-plate feeder, shown in Fig. 8, is used for handling 
coal and such material of all sizes. It consists of two castings pivoted 
at their tops and swung alternately so as to move the material forward 
on the bottom plate. The plates are moved by connecting-rods 





\ COHMtCTIM 



Fig. 8 Swinging-Plate Feeder 

from a crank or eccentric through a rocker shaft. The capacity is 
fixed by the length and the number of strokes, but as it is limited to 
the amount of material displaced by the plates, a wide range is not 
possible. 

15 The disadvantages are the lack of adjustability and the 
tendency of the material to pack. It will also be noted that the 
feeder is not self-cleaning, so that the bottom plate always contains 
material which is very liable to freeze in winter. 

16 The plunger feeder, illustrated in Fig. 9, is similar in operation 
to the swinging-plate feeder in pushing the material along the bottom 
plate. The plunger may be built either in one or two parts, moving 
ahead alternately and driven through a rocker shaft, as in the case 
of the one previously described. The capacity is fixed by the number 
and length of the strokes and the location of the adjusting gate. 



AUTOMATIC FEEDERS FOR HANDLING MATERIAL 



167 



This type has the same disadvantages as the swinging plate feeder, 
the most serious being that it is not self-cleaning. 

17 Fig. 10 shows the reciprocating-plate feeder, consisting of a 
plate mounted on four wheels forming the bottom of the hopper. 
When the plate is moved forward, it carries the material with it, and 
when it is moved back the plate is withdraAvn from under the material, 
allowing it to fall into the chute. The plate is moved by a connect- 
ing rod from crank or eccenti'ic. The capacity is determined by the 
length and number of strokes and the location of the gate. The 
disadvantages are the lack of adjustment and the inability to clear 
the feeder of material. 

18 The shaking feeder. Fig. 1 1, consists of the shaker-pan located 
under the opening in the bottom of the hopper at such an angle that 
the material will not flow when the pan is stationary. When given 



HOPPER 


./ 


nOJUSTflBLE GATE 



DISC CHW1K I RBIL3 J ^ 



SKIRT B0ffRD5 



CONWECTinG ROD 



Fig. 9 Plunger Fekder 



a reciprocating motion by the crank and connecting-rod, the material 
is moved forward on the pan. The front end of the pan is carried by 
a pair of flanged wheels; the back end is suspended by two hanger- 
rods, each being provided with a turn-buckle so that the angle of the 
pan may be varied. The crank having an adjustable length of stroke, 
there are three variables, viz: number of strokes, length of stroke; 
and inclination of the pan. As the number of strokes is difficult to 
change, and the others easily changed, the feeders are usually de- 
signed for about 75 strokes per niin., a number determined by 
experiment. The angle of the pan is fixed by the capacity desired 
and the nature of the material handled. For coal, stone, ore, etc., 
8 deg. to 10 deg. is sufficient, while clay and other sticlcy substances 
require from 15 deg. to 20 deg. The length of stroke varies from 4 in. 
to 12 in., so that a large range is possible. 

19 A feeder designed to handle 400 tons per hr. of mine-run coal 



168 



AUTOMATIC FEEDERS FOR HANDLING MATERIAL 




Fig. 10 Reciprocating-Plate Feeder 

was changed in five minutes to deliver 30 tons per hr., by shortening 
the length of stroke and lowering the pan until nearly horizontal. 

20 Not only has this feeder the widest possible range in capacity, 
but it is self-cleaning, a very important feature. From the cut it 
will be noted that the pan is placed under the opening and the 
material rests directly on the pan, so that when the pan is moved the 
material in the hopper is moved, which prevents the material from 
bridging. 

21 The shaking feeder has none of the disadvantages of the 
other types for general use, and possesses many advantages which the 
others lack. Owing to its great flexibility it is more easily standardized 




Fig. 11 Shaking Febdeb 



AUTOMATIC FEEDERS FOR HANDLING MATERIAL 169 

and will successfully handle practically any material, regardless of 
size or condition. If desired the bottom plate may be perforated to 
screen out the fine material, thus acting as both screen and feeder. 
This is not possible with any of the other types. 

22 The power required by all of the types is so small that it is 
not an important consideration. The shaking feeder mentioned 
above, which handled 400 tons of coal per hr., required but 3.5 h.p. 

23 The preceding cuts and descriptions will give a general idea 
of the different types and their possible uses, so that an engineer may 
readily choose the best type for the work to be done. The point that 
should be kept in mind is, that it is always advisable to gear the 
feeder to the conveyor, crusher, or other machine which it feeds so 
that they will both start and stop simultaneously. 

DISCUSSION 

T. A. Bennett. An automatic feeder is absolutely necessary in 
some installations; in others it is demanded for economic reasons. 
For example, run-of-mine coal, on account of constriction in the chute, 
requires properly a 36-in. belt, although it can be handled on a 30- 
in. belt. With an automatic feeder it is possible to use a 24-in. con- 
veyor provided the capacity will permit. In handling damp sand, 
a large chute opening is necessary and this usually requires at least 
a 16-in. conveyor belt. With a feeder, however, this width can easily 
be reduced to 12 in. A 12-in, conveyor has capacity to take care of 
nearly every problem in handling damp sand. 

2 The feeder is also economical in filling in the blank spaces on 
the belt. The loading is usually intermittent, the belt being either 
over-loaded or under-loaded intermittently; the feeder can be regu- 
lated to give a uniform maximum loading, greatly increasing the 
capacity. Intermittent loading also increases the wear on the belt. 
As the only wear worth considering is that done by the material 
coming in contact with the surface of the belt in the delivery of ma- 
terial to it, the larger the load the less is the proportionate wear on 
the belt. 

3 A type of feeder has been developed similar to that in Fig. 8, 
but doubled; that is, a hopper and either one or two swinging plates at 
each end and pushing to an outlet at the center part of the skirt 
boards. This type is now working very satisfactorily in two large 
plants, the Hudson Company's power house at Jersey City, N. J., 
and the Illinois Steel Company at Joliet, III. The chief advantage 
over any other is the saving of headroom. Where there is sufficient 
headroom, the shaker feeder is correct practice and is in general use. 



170 DISCUSSION 

The Author. Mr. Bennett's discussion emphasizes three impor- 
tant points with reference to the use of automatic feeders in connection 
with belt conveyors: 

a The installation of the feeder frequently permits the use of 

narrower belts. 
b Delivery of an even and continuous stream of material en- 
ables the conveyor to operate at its maximum capacity. 
c By loading the conveyor to its full capacity, a smaller pro- 
portion of the load comes in contact with the belt, there- 
by reducing the wear per ton carried on the belt at the 
loading point. 

2 The type of feeder mentioned by Mr. Bennett as similar to that 
illustrated in Fig. 8 is a variation of that type, used where material 
is taken from two hoppers; instead of tAvo swinging plates placed side 
by side, there is a single plate under each hopper. These two plates 
are connected by rods, so that when one plate is in the forward stroke 
the other will be in the back stroke. It is therefore nothing more 
than two single-plate feeders so connected that they operate together. 

3 Mr. Bennett is mistaken regarding this type saving headroom 
over any other type. It requires about the same headroom as the 
plunger feeder (Fig. 9) and the reciprocating plate feeder (Fig. 10). 
The great drawback of the swinging-plate feeder is that it is not self- 
cleaning, so that if exposed in winter the material will freeze to the 
bottom plate. 

4 This type of feeder was originated by Mr. Lincoln Moss of New 
York, and the two installations mentioned by Mr. Bennett were 
designed by Mr. Moss under the writer's direction. 



No. 1230 

A NEW TRANSMISSION DYNAMOMETER 

By Pkof. Wm. H. Kenerson, Providence, R. I. 
Member of the Society 

The author has received from time to time many requests for a 
simple transmission dynamometer, and has himself often felt the 
need of one which would be more generally applicable than those now 
in use. These continued requests, together with the requirements 
of a definite problem whose solution demanded a rigid transmission 
dynamometer in the form of a coupling, led to the design and con- 
struction of the instrument described below. The accompanying 
illustrations show the construction of the dynamometer and its 
method of application and use. In Fig. 2 and Fig. 4 the correspond- 
ing parts of the dynamometer are given the same letters and are 
referred to in the text. 

2 The couplings A and B, each keyed to its respective shaft, are 
held together loosely by the stud bolts C. The holes in the flange A 
are larger than the studs C, so that these studs have no part in trans- 
mitting power from one shaft to the other. The power is trans- 
mitted from A to B through the agency of the latches L, four of 
wliich are arranged around the circumference of the flange B. These 
latches are mounted and are free to turn on the studs E. The two 
fingers of the latches engage the studs F on the flange A. On the 
ends of each latch are knife-edges parallel to the stud about which 
the latch turns. For either direction of rotation of the flange A 
the latches L, which are in effect double bell-crank levers, will exert 
a pressure on the disc G, tending to force it axially along the hub of 
the coupling B, and this pressure, it will be seen, is proportional to 
the torque. 

3 Between the end-thrust ball, or roller, bearings M M, is held 
the stationary ring S, which is the weighing member. is a thrust- 
collar screwed on the hub of B, and P is its check nut, which is ordi- 

Presented at the Spring Meeting, Washington, May 1909, of The American 
Society of Mechanical Engineers. 



172 



A NEW TRANSMISSION DYNAMOMETER 



narily pinned to the hub when in position. The stationary member 
S, in the form of a ring surrounding the shaft, is prevented from 
rotating by fastening to some fixed object the attached arm shown 
in the view (Fig. 1) of the assembled instrument. In this ring is an 
annular cavity covered by a thin, flexible copper diaphragm D, 
against which the ball-race of one of the thrust bearings presses. 
The edge of this ball-race is slightly chamfered to allow some motion 




Fig. 1 Dynamometer for 2-in. Shaft, Weight 60 lb. 



to the diaphragm. The cavity is filled with a fluid, such as oil, and 
connected by means of a tube to a gage. The oil pressure measured 
by the gage is proportional to the pressure between the thrust- bear- 
ings, which in turn is proportional to the torque. 

4 The instrument may be calibrated in the torsion-testing machine 
or by means of a sensitive friction brake. Fig. 6 is an actual cali- 
bration curve for a small instrument, obtained by hanging standard 



A NEW TRANSMISSION DYNAMOMETER 



173 



weights at proper distances from the shaft on a horizontal lever 
attached to the shaft, and reading the pressures indicated by the 
gage for the various torques shown in the diagram. For ordinary 
purposes, however, it is not necessary to calibrate the instrument by 
actual trial, since computations of the oil pressures for the various 
torques from the lengths of the lever-arms and diaphragm area 
check very closely those thus obtained. 

5 It will be seen that the weighing means is similar to that 
employed in the Emery testing macliine, which is recognized as being 
extremely accurate. It will be possible to employ the Emery flexible 




Fig. 2 Dynamometer Shown in Section 



steel knife-edges on the levers, if desired, but this has been found in 
practice an unnecessary refinement. 

6 The construction makes the coupling as nearly rigid as materials 
will permit, the movement of the diaphragm being extremely small. 
The only flow of oil through the copper connecting pipe is that suffi- 
cient to alter the shape of the Bourdon tube, if that be the form of 
gage employed. As soon as the normal position of the gage is reached 
this flow ceases, hence there can be no fluid friction. It is possible 
therefore, to use as long and as small a tube as desired, without intro- 
ducing error. Where the gage is placed at a distance above or below 
the coupling, correction should of course be made for the static head. 



174 



A NEW TRANSMISSION DYNAMOMETER 




E^ 




A NEW TRANSMISSION DYNAMOMETER 



17.5 



7 Other means than the gage shown may be employed to measure 
the fluid pressure. Where extreme accuracy is desired it will be well 
to employ the weighing device used with the Emery testing machine. 
The manograph has been used in this connection to measure varia- 
tions in torque too rapid for indication by the ordinary gage. For 
example, the variations in torque in a single revolution of the shaft 
of a 3-cylinder gasolene engine have been recorded with its aid. 

8 Where the rate of rotation of the shaft is variable and it is 




Fig. 5 Dynamometer Placed between Flanges in Machine-Shop Drive 

3-lNCH SHAFT. SPIRAL RUNNING TO THE WALL IS OIL PIPE TO GAGE 



desired to indicate the horsepower direct, the combination of gage 
and tachometer shown in Fig. 7 is employed. The hydraulic gage 
is connected to the coupling described, its pointer therefore indicat- 
ing torque. The pointer of the tachometer shows the number of 
revolutions per minute. Being a function of the revolutions per 
minute and the torque, the horsepower will be indicated by the inter- 
section of the two pointers and suitable curves on the dial as shown. 
Arrangements for recording or integrating the work done may also be 
attached to the coupling. 



176 



A NEW TRANSMISSION DYNAMOMETER 



50 

■g ^ 

a 

"i *^ 

u 
K 35 

'^-•30 

f« 

Sf 20 

1 




















1 




















/ 


















[/ 


















/ 
















( 


/ 
















( 


y 


















/ 


















/ 














S 15 

p. 






/ 
















o 10 




1/ 


















5 


/ 





















250 500 750 1000 1250 1500 1750 2000 2250 2500 
Torque, iiic'h-poumls 

FiQ. 6 Calibration Curve for Transmission Dynamometer, 




Fig. 7 Combination Pressure Gage and Tachometer Indicating Torque 
Revolutions per Minute and Horse Power 



A NEW TRANSMISSION DYNAMOMETER 177 

9 A summary of some of the more important characteristics of 
the instrument follows: 

o The instrument is compact. The example shown in Fig. 
3 and Fig. 4, which is designed to transmit 30 h.p, at 500 
r.p.m., is about 5f in. in diameter and weighs about 25 
lb. That shown in Fig. 5 driving a 3-in. shaft is about 13 
in. in diameter and weighs about 160 lb. 

b It is as rigid as an ordinary flange coupling. 

c It may be made in the form of a coupling, and will then 
occupy about the same space as the usual flange coupling, 
or it may be made in the form of a quill on which a pulley 
is mounted. This form may be made in halves for appli- 
cation to a continuous shaft. 

d It will indicate for either direction of rotation of the shaft. 

e The torque may be read and recorded or the work inte- 
grated at a considerable distance from the coupling. 

/ The readings do not require correction for different speeds 
of rotation. All parts containing oil are stationary, hence 
are unaffected by variation in speed. Other parts are 
likewise unaffected by centrifugal action. 

g It may be made very sensitive and accurate. The construc- 
tion lends itself very easily to variation of range of appli- 
cation and to varying degrees of sensitiveness, since the 
oil pressure, and hence the sensitiveness of the instrument, 
depend upon the area of the diaphragm, the relative 
lengths of the arms of the latches L, and the diameter of 
flanges. Its accuracy is dependent mainly on the degree 
of accuracy of the means employed to measure the fluid 
pressure, of which a number of forms, other than the 
usual pressure gage, are available. 

h The only power absorbed is the small amount due to the 
friction of the ball, or roller, bearings, and this can be 
determined from the pull of the retaining arm. It is 
unnecessary to make correction for this, however, since 
the amount is so small as to be negligible. 

i Since the only wearing parts are the ball, or roller, bearings, 
which may be lightly loaded, the instrument should not 
be deranged easily. Because of the very small volume of 
oil contained in the weighing chamber, ordinary tempera- 
ture changes do not affect the calibration. All parts con- 
taining oil are stationary, hence all joints may be soldered 
and leakage entirely prevented. 



178 DISCUSSION 

j With suitable material and ordinary workmanship, it is 
believed that there is little likelihood of failure of any 
part of the instrument. It is conceivable, however, that 
the balls or rollers, although lightly loaded, might -crush; 
the diaphragm might shear; or the stationary member, 
although bearing only its own weight and lubricated, 
might seize to the hub. Remote as are any of these 
possibihties, should any or all of them occur, the worst 
that coul'l happen would be the tearing-off of the oil pipe 
and retaining arm, when the whole would revolve as a 
solid coupling. In no case can the coupling fail to drive 
the shaft because of its variation from the standard form, 
since, in addition to the driving latches employed to 
carry the load normally, the same number of connecting 
bolts may be employed as in the ordinary coupling, which 
will still hold the coupling together should the latches 
fail. Since, however, these latches are farther from the 
shaft, they should, if properly constructed, be less likely 
to fail than the connecting bolts usually emploj^ed. 
10 It is believed that uses for the instrument here described will 
suggest themselves, and it is with the hope that the device will prove 
of some interest to those who deal with the use and transmission of 
power that the matter is presented to the Society. 

DISCUSSION 

A. F. Masury. I want to say a few words as to how the Kenerson 
dynamometer may be applied to the betterment of design in motor 
vehicles. In the first place, we must have exact data regarding the 
effect of road irregularities, wind pressures, and the resistances set 
up by grades and speed. These figures are absolutely necessary in 
order to determine the best torque to apply on motor gearing and 
equipment, such as tires, etc., to overcome the existing conditions in 
each particular car. 

2 At present we have two recourses: first, the figures procured 
by Mr. S. F. Edge at the Brooklands track in England with his Napier 
car. He first calibrated his motor and then made the test on the 
track. These figures must necessarily include many errors. Second, 
the dynamometer at the Automobile Club of America in New York. 
Here again satisfaction is not entirely procurable as all the road con- 
ditions are obtained artificially by attachments on the machine. 



A NEW TRANSMISSION DYNAMOMETER 179 

3 With the Kenerson dynamometer we can certainly get exact 
readings while the machine is working on the road. These can even 
be made graphic if desired. Manufacturers wll thus have available 
means of getting information which should result in more perfect 
design. 

4 There is one thing more, the dynamometer of the Automobile 
Club of America cost, I believe, in the vicinity of S15,000 by the time 
it was completely installed, while the price of the Kenerson machine, 
around $500, will make it possible for even a small manufacturer to get 
his own reading. 

The autlior desired to present no closure. — EoiToit. 



No. 1237 

POLISHING METALS FOR EXAMINATION WITH 
THE MICROSCOPE 

Bt Albert Kingsbuby, Pittsbubq, Pa. 
Member of the Society 

In 1902 the writer made experiments to find the most suitable 
method of polishing samples of metals for microscopic examination 
The polisliing of the surface is one of the most important as well as 
most troublesome details of metallography, particularly when high 
magnification is required. 

2 At the outset, trials were made of all the methods of which 
descriptions have been published. Some of those methods have been 
successfully employed by various metallographists, as shown by 
numerous reproductions of excellent micro-photographs in different 
publications. Nevertheless the writer did not find any of these free 
from objectionable features. The ideal method should produce a 
fairly flat surface, free from excessive relief of the harder constituents, 
rounded edges at flaws, or scratches and smearing of the metal. The 
method should be simple, the materials employed readily available, 
and the process as rapid as consistent with the first-named requisites. 
None of the published methods embodied all these requisites, nor is 
a perfect method likely to be found. However, the method finally 
developed by the writer appears to him superior. 

3 The preliminary trials were made with rotating discs covered 
with various materials, including canvas, felt, silk, leather, chamois, 
parchment, paper, wood, pitch, asphalt, resin, shellac, beeswax, etc. 
The polishing powders included commercial abrasives, such as emery, 
carborundum, tripoli, crocus and jewelers' rouge; also precipitates, 
such as carbonates and sulphates of the alkaline earths. Attempts 
were made to obtain fine finishing powders by the levigation process 
from commercial abrasives. These abrasives were tried both wet and 
dry and with various speeds of the discs. Hand polishing was also tried. 

Presented at the Spring Meeting, Washington, May 1909, of The American 
Society of Mechanical Engineers. 



182 POLISHING METALS FOR EXAMINATION 

It is needless to detail the objectionable features encountered, which 
are probably familiar to all metallographists. 

4 The method finally adopted was the result of two distinct dis- 
coveries: {a), that ordinary paraffin Avax makes a good polishing bed; 
(6), that excellent polishing powders of certain grades are commer- 
cially available. 

5 The paraffin is used as a facing for rotating discs of metal, 
preferably brass, about 8 in in. diameter. The discs are grooved on 
the flat face for anchoring the wax. To prepare the discs, they are 
warmed to about 100 deg. cent., and laid flat, and the melted paraffin 
is poured on them to a depth of about ^ in., a removable ring or band 
retaining the melted wax. The whole is tli^n covered to exclude 
dust and allowed to cool. After the wax has solidified it may be 
dipped in water to hasten the hardening. Since the wax has very 
little viscosity when melted, all hard foreign particles, which might 
) )roduce scratches in the samples, settle out before the wax hardens, 
the elimination being practically complete. No advantage in this 
respect was gained by keeping the wax in a fluid condition on the disc 
for several hours in an oven. After the hardening of the wax the 
discs are placed on the spindle of the polishing machine and the face 
of the wax is turned true and flat by a hand-tool. 

6 In the writer's machine the spindle was horizontal and four 
discs were used for abrasives of progressive fineness, two discs being 
placed back to back at each end of the spindle. The disc used for the 
final polishing should not be perforated and the wax should be con- 
tinuous to the center of the disc, as that part is best for the finishing 
touches to the sample. This latter disc should be at the right-hand 
end of the spindle. The speed of rotation should be about 200 r.p.m.; 
a higher speed throws off the polishing powder with the water used, 
and a lower speed makes the work too slow. A stationary sheet- 
metal strip about 3 in. wide bent over the discs serves as a screen. 

7 The polishing powders, in the order used, were as follows: (a) 
commercial flour of emery; (6) washed Naxos emery, 3/0 grade; (c) 
washed Naxos emery, 7/0 grade; (d) soft optical rouge, light grade. 
These were obtained from the George Zucker Co., New York, except 
the first, which is available everywhere. 

8 The emery powders were mixed to a paste with water in tall 
glass jars provided with covers; the paste was apphed to the rotating 
discs with small brushes as required, the brushes being kept in the 
jars when not in use. The rouge was in cake form, best applied by 
holding a small piece in the hand, wetting both the rouge and the wax, 
and pressing the rouge lightly against the rotating surface. 



POLISHING METALS FOR EXAMINATION 183 

9 A small quantity of water is required throughout the polishing 
process, but water cannot be used very freely without wasting the 
powders. The water is best applied as required, from an ordinary 
chemist's wash-bottle, held in the left hand while the right hand 
manipulates the sample. No water pipes or drains are required for 
the polishing machine. Distilled water may be used if available. 
If tap water is used, it should be drawn into large jars provided with 
covers and siphons, and allowed to stand a day or more before use, 
in order that all gritty particles in suspension may be deposited. 
The inner ends of the siphon tubes should be at least 3 in. above the 
bottom of the jars. 

10 The treatment of the samples is as follows: the samples are 
first di-essed to shape and size' by any convenient method, the surface 
to be polished made flat by an emery wheel or file, and the sharp 
edges rounded to prevent cutting into the wax. The dimensions of 
the samples should depend to some extent upon the coarseness of 
structure. For normal iron and steels, and for much other work, 
a f-in. cube is a convenient sample. Massive castings sometimes have 
grains an inch or more in diameter, and correspondingly large samples 
are rec^uired. The samples are held flat against the waxed discs, 
which are kept well covered by the polishing paste, using successively 
the flour of emery, the 3/0 emer}^, the 7/0 emery, and the rouge, on the 
several discs. At each grinding with emery the sample should be 
held without rotation and with a slow transverse motion across the 
face of the disc until the grinding marks show over the entire surface. 
The sample may then be given a quarter turn, so that the new marks 
cross the old ones, and so on. The discs must be kept wet continu- 
ally while grinding. With each grade of powder the grinding should 
continue for some time after the marks of the last previous grade 
have disappeared, especially with soft metals, since the scratches 
cause a flow or disturbance of the metal to a minute depth below the 
surface, and if this disturbed metal is not ground off, the deep effect 
of the scratches becomes apparent on etching. In the final polishing 
on the rouge disc, the sample should be continuously rotated; this is 
most readily done by moving the sample nearly in a circle about the 
center of the disc in an opposite direction from the rotation of the disc. 
This keeps the direction of the grinding marks constantly changing, 
and avoids grooving. The finishing should be done near the center 
of the disc, the slower motion being most effective for very fine polish- 
ing. After grinding with one grade of powder and before proceeding 
to the next, the samples and the operator's hands should be thor- 



184 POLISHING METALS FOR EXAMINATION 

oughly washed; and the hands and the apparatus should be kept free 
of dust or dirt, to secure a polish free from scratches. 

11 The most important item to be noted by the beginner is the 
liability of the paraffin to adhere to the samples when the grinding 
is begun, particularly in the case of the rouge disc. When the sample 
is first brought into contact with the disc, especially if the latter has 
been freshly prepared, the paraffin nearly always smears over the 
surface of the sample in a second or two, and if the sample is not 
removed and cleaned at once the result is a roughened disc, requir- 
ing re-turning with the hand-tool and re-application of the paste. 
Therefore the sample should at first be touched very lightly to the 
disc, and at once removed and wiped with the finger, or with a cloth. 
If this is repeated several times, the surface of the sample will no 
longer become coated with paraffin but can be ground continuously, 
except when a fresh coating of paste is required by the disc. One 
great advantage of the paraffin disc over discs covered with cloth or 
felt, is that if the disc becomes roughened or cut, it can readily be 
turned smooth and true again. 

12 For cleaning the samples after polishing, the best material is 
a stock of old linen or cotton cloth well-laundered and cut to 3-in. 
squares. These small pieces are preferable to larger ones, since they 
can be discarded for fresh ones after once using. The old cotton or 
linen is also the best material for cleaning the lenses and mirrors of 
the optical apparatus, being superior to chamois for this purpose. 

13 The time required for polishing a sample varies somewhat 
with the hardness. A single sample of normal steel, cast iron, or 
wrought iron, may be finished in fifteen minutes; a set of five or six 
such samples may be finished in an hour. Hardened steels require 
a slightly longer time. The method has not thus far proved service- 
able for very soft metals and alloys, particularly lead, owing to the 
persistent adhesion of the paraffin to the surface of the sample. 
The harder alloys polish well by this process. The finished surface 
presents a minute relief of the harder constituents, but much less 
than is produced by the use of felt or other very soft materials. 

14 The paraffin beds are more durable than might be supposed; 
on long standing at summer temperatures the surfaces become dis- 
torted by the flow of the wax, but they can always readily be made 
true by the turning tool. The harder paraffin (ceresin) offers no 
advantages over ordinary paraffin, except that it flows less at sum- 
mer temperatures. It is serviceable for use with the emery powders 
but too hard for best results with the rouge. 



No. 1238 

MAKINE PRODUCER GAS POWER 

A COMPARISON OF PRODUCER-GAS AND STEAM EQUIPMENTS 

By C. L. Steaub,* New York 

Non-Member 

So much interest is exhibited both by the engineering profes- 
sion and the general public in the application of producer gas power 
to marine, commercial and naval service, that a brief summary of 
recent progress in this field appears timely. 

2 Any innovation which makes for improvement in present 
practices, surely, though sometimes slowly, achieves its end. Pro- 
ducer gas power, on impartial analysis, offers so many benefits to 
marine service that it appears strange indeed that more rapid prog- 
ress has not been made in its adoption. The delay appears to be 
due to several causes. 

3 The marine public, which since the days of the Clermont has 
exclusively associated the term "motive power" with steam, has 
every reason for demanding exact and conclusive evidence of the 
superiority of gas power or any other power, before adopting it 
in lieu of present methods. This evidence is only now slowly 
coming forth. Many who have been credited with authority by the 
engineering profession and others, either through ignorance or through 
misinformation, have beset the way of marine gas power with 
numberless imaginary obstacles, ridiculous in proportion to the real 
difficulties, but sufficient nevertheless to instill some doubt of the 
possibilities of the system into the minds of the waiting public. 

4 Only recently has such progress been made in the development 
of gas power for marine work, as to warrant its early adoption in 
commercial service. Two years ago, less than 300 h.p. in the aggre- 
gate was being developed by marine producer'gas power installations; 
these were experimental in nature and were of the German Capitaine 

' With the Loomis-Pettibone Co., New York. 

Presented at the Spring Meeting, Washington, May 1909, of The American 
Society op Mechanical Engineers. 



186 MARINE PRODUCER GAS POWER 

type. There are now installed and accepted 23 Capitaine marine 
plants, aggregating 2035 h.p., a partial list of which follows: 

a Emil Capitaine: Launch, 60 b.h.p.; 4-cylinder single-acting, 4-cycle 

engine; boat 60 ft. long, 10 ft. beam, 4ft. draft; ran an average speed 

of 10 miles for 10 hr. on 412 lb. of anthracite coal. 
h Rex: Sea-going Swedish boat; 102 ft. long; 22 ft. beam, carries 350 tons 

on 9-ft. draft; fitted with a 3-cylinder single-acting, 45-h.p. engine at 

.300 r.p.m. 
c Capitaine: Tow boat at. Genoa; length 47 ft., beam 12 ft., draft 7 ft.; 

fitted with a 3-cylinder, single-acting, 4-cycle engine, 105 b.h.p. at 

240 r.p.m. 
d Duchess: Canal barge; length 71 ft., beam 7 ft. 1 in.; carries 20 tons 

cargo on 42-in. draft; fitted with double-cylinder, single-acting, 4-cycle 

engine of 25 b.h.p. 
e Dusseldorf: Tug at Hamburg; fitted with a 4-cylinder, single-acting, 

4-cycle engine, 60 b.h.p. at 240 r.p.m. 
/ Isee: Tug, fitted with a 3-cylinder, single-acting, 4-cycle engine, 45 

b.h.p., 300 r.p.m. 
g Wilhclm: Combination freight and passenger Rhine boat, fitted with a 

5-cylinder, single-acting engine, 175 b.h.p. at 240 r.p.m. 
h Badenia: Rhine freight boat, fitted with a 2-cylinder, single-acting, 

4-cycle engine of 30 b.h.p. 
i Katrina: Canal freight boat, fitted with a 3-cylinder, single-acting, 

4-cycle engine, 45 b.h.p. 
i Marie: Canal freight boat; fitted with a 3-cylinder, single-acting, 

4-cycle engine, 45 b.h.p. 
k Hoffnung: Combination freight and passenger Rhine boat, fitted with 

a 5-cylinder, single-acting, 4-cycle engine of 210 b.h.p. 

I Amersie: Volga freight boat, fitted with a 4-cylinder, single-acting, 

4-cycle engine of 60 b.h.p. 
m No. 58: Canal freight boat, fitted with a 4-cylinder, single-acting, 
4-cycle engine of 60 b.h.p. 

5 In addition to the above there were a number of freight boats, 
the dimensions and names of which we were unable to obtain, but 
whose power plants varied in capacity from 30 to 175 h.p. each. 

II H. M. S. Rattler: An old gun boat, 165 ft. long, 29 ft. beam, originally 

fitted with a triple expansion engine. The gas engine is 5-cylinder, 
single-acting, 4-cycle. Cylinders 20 in. diameter by 24 in. stroke, 
developing 500 b.h.p. at 120 r.p.m. This engine is started by means 
of a mixture of gas and air which is pumped into the cylinders at a 
pressure of about 95 lb. per sq. in. This complete plant was designed 
entirely in the Capitaine Works at Diisseldorf. The total weight 
of the entire plant, including the donkey boiler for working the pumps 
and auxiliaries, is 94 tons, as compared with 150 tons in the case of 
the displaced steam engine. A consumption of 1525 lb. of coal was 
made for a measured distance of 45 knots on an average speed of 



MARINE PRODUCER GAS POWER 187 

10^ knots per hr. The cost per mile for fuel with coal at 15s. Qd. per 
ton is $0,064 U. S. currency. This boat made a maximum speed of 
11.3 knots per hr. against a 1^ knot current at 110 r.p.m. of the 
engine shaft. 

6 All of the above plants by their design and construction are 
restricted to operation on anthracite coal, coke or hard-burned char- 
coal, and any plant so restricted by its design to one class of fuel is 
seriousl}' limited in its scope of application. The development of a 
simple marine gas-producer for use with any class of solid fuel is a 
necessity, if the system is to be considered seriously by the marine 
profession. 

7 The writer is fortunate in having been associated with some 
recent American developments both in stationary and marine gas- 
power plants, a brief survey of a portion of which will enable us to 
draw more clearly the comparison between a typical steam and a 
possible gas installation. 

8 There are in commercial operation in this country today, two 
distinct types of stationary power gas-producers which are suited by 
their design for operation on almost any class of solid fuel. They may, 
by their systems of operation, be qualified as up-draft and down- 
draft producers. 

9 In the up-draft producer, the fuel is charged into the generator 
through an air-tight mechanism at the top, while air and steam, or 
air and products of combustion are admitted at the bottom of the 
fuel bed, and passing upward, leave the generator at the top in contact 
with the fresh fuel. Almost all of the hydrocarbons are unfixed 
when leaving the generator with the hot gas, and are condensed later 
in the gas coolers or scrubbers and gas mains, forming large amounts of 
tar, which, if not removed to a minute degree, will positively prevent the 
operation of the engine. The removal of tliis tar is troublesome and is 
accomplished at a loss of power and efficiency. The fuel in the upper 
zone of the bed in the up-draft producers cokes and cakes so seriously 
as to require continuous poking of the fuel bed, either mechanically or 
by hand. These features and others in this type of apparatus contrib- 
ute to limit the rates of combustion per sq. ft. of grate to a relatively 
low quantity. All things considered, therefore, this type of appara- 
tus has not lent itself agreeably to modification for marine service. 

10 In the down-draft type of apparatus, the fuel is charged by 
hand through a large door at the top of the producer, which is nor- 
mally in an open position, allowingthe operator unrestricted inspection 
of the whole upper zone of the fuel bed. The hydrocarbons con- 



188 MARINE PRODUCER GAS POWER 

tained in the fuel are driven off in the upper zone, mixed with air and 
almost completely burned, and the burnt products, passing downward 
through the relatively deep bed of fuel, are decomposed and regener- 
ated into carbon monoxid and hydrogen gases. All of the tar and the 
lighter hydrocarbons are completely fixed in this process, and no tar 
is found in condensation in any portion of the plant after cooling. 
Coking or caking of the fuel bed is not detrimental, but on the other 
hand assists in keeping the fire in the open porous condition, which is 
desirable and necessary where high rates of combustion obtain. This 
feature eliminates the poldng necessary in the up-draft apparatus. 
The gas leaves the bottom of the producer through brick-lined connec- 
tions, and a portion of the sensible heat is extracted in passing through 
an economizer. The gas is then cooled and washed and passed 
through an exhausting mechanism, whence it is delivered under 
pressure to the engine. 

1 1 This type of apparatus lends itself admirably to the high rate 
of fuel combustion, which for the sake of economy in space and weight 
is desirable in marine service. There are in actual commercial opera- 
tion today, a number of plants of this type having an average fuel 
consumption of over 40 lb. of good bituminous coal per sq. ft. of grate 
per hr. These producers are sold on a rating of from 18 lb. to 20 lb. of 
fuel per sq. ft. of grate per hr., which is almost 100 per cent greater 
than the average rating of the up-draft type of producers. 

12 Undoubtedly a better method of measuring the ability or 
success of these two systems, is to make note of the number and capa- 
city of plants of each type in actual operation on engine service. A 
report of the committee on gas engines of the National Electric Light 
Association, spring of 1908, showed that in gas-engine power plants, 
of capacities of over 300 h.p. each, there were in operation 32 plants of 
both types having a total capacity of 57,225 h.p. Of these, 4 plants 
were of the up-draft type, having an aggregate capacity of 4050 h.p., 
and 28 plants were of the down-draft type, with an aggregate capacity 
of 53,175 h.p. The latter contain the Loomis-Pettibone gas-generat- 
ing apparatus, some of which has been in operation on engine service 
for 13 years. 

13 Three years have been devoted to the modification of these 
stationary plants for marine service. The work involved a reduction 
in the size and weight of the generators; complete revision of the 
scrubbing, gas cleansing and exhausting mechanism; elimination of all 
gas holders, storage receptacles, mixing chambers, etc. 

14 The plant as modified to date has a light compact producer, 



MARINE PRODUCER GAS POWER 189 

which while retaining the same rate of combustion as the stationary 
apparatus, has materially reduced dimensions and weight of the shells, 
brick lining, fittings, etc. The economizer boilers which were used on 
stationary work have been abandoned, and replaced with light air- 
heating economizers. The gas coolers no longer contain any coke or 
brokenmaterial, or wooden trays, and are built of very hght, non-corro- 
sive sheet metal, and arranged for eitlier vertical or horizontal posi- 
tions, the latter arrangement being convenient for space which would 
be otherwise wasted in the vessel. The cooled and partially cleansed 
gas is drawn through the above portion of the plant by a centrifugal 
gas-cleaning exhauster, driven by direct-connected motor. The gas 
passes directly from the exhauster under pressure, through an auto- 
matic pressure-regulating valve, to the engine manifold. 

15 That the plant is adaptable for marine service with regard to 
space occupied and weight, may be seen from the following conserva- 
tive estimate: 

Plants of from 100 to 500 h.p. each occupy from 0.4 to 0.5 sq. ft. 
per h.p., and weigh from 70 lb. to 90 lb. per h.p., including all 
auxiliaries, piping, etc. ; plants of from 500 h.p. to 1000 h.p. 
occupy from O.SOsq. ft. to 0.45sq. ft. per h.p., and weigh from 
401b. to 70 lb. per h.p., including all auxiharies, piping, etc. 

16 Undoubtedly the rational opportunity at the present time for 
marine gas power lies in commercial service, in which regard the most 
rapid advancement in America has been made in the freight, ore and 
fuel carriers of the Great Lakes. 

17 We have therefore taken for our example a ship built from 
the designs of Messrs. Babcock & Penton within the last year. 
For the sake of clearness, the views show only the machinery space; 
all of the ladders, stairways and grates have been omitted from the 
plans, and the piping is shown only on the gas installation. The 
machinery installation proper is all there, however, and while the 
parts eliminated are merely accessory, the contrast between the two 
plants would be all the more striking were they included. 

18 The boat is a modern lake freighter and represents the best 
standard practice in this service. She is 306 ft. long over all, 45 ft. 
beam and 24 ft. deep. Her present power equipment consists of a 
single-screw, triple-expansion, three-crank condensing engine, 18-30- 
50 by 36-in. stroke, which indicates 1050 h.p. at 90 to 95 r.p.m. 
The vessel is fitted with a jet condenser and has independent 



190 MARINE PRODUCER GAS POWER 

steam-dri'ven reciprocating, bilge, vsanitary and feed pumps. The 
complete engine room weight, including piping and alJ auxiharies, is, 
in round figures, 182,000 lb. 

19 The boiler room equipment consists of two single-ended Scotch 
boilers 11 ft. 10 in. mean diameter, 11 ft. length over heads, operat- 
ing on a working pressure of 180 lb. per sq. in. Each boiler is fitted 
with two 42-in. corrugated furnaces and has tM'O hundred and forty- 
four 2f-in. tubes. The grate surface is 36 1 sq. ft. and the heating 
surface 1642 sq. ft. in each boiler. 

20 The boilers are fitted with forced draft from a 66-in. steam- 
driven fan. The air for the draft is taken from the stoke hole and the 
fan is located in the engine room. The fan discharge passes through 
air heaters in the up-take and thence through ducts to the under side 
of the grates. The complete boiler-room weight, including water in 
the boilers, but not fuel, is 170,000 lbs. These weights are actual 
rather than mere estimates. 

21 The coal bunker extends from the main deck to the tank top 
and is arranged atliwartship. It has a capacity of 170 tons. The 
bunker doors face the stokers on the stoke hole floor. The bunker is 
6 ft. fore and aft at the stoke hole. The distance from the forward to 
the after bulkhead in the boiler room is 24 ft. in. The distance from 
the forward to the after bulkhead in the engine room is 22 ft. in., 
maldng a total over-all length for the plant, including bunkers, of 
52 ft. in. 

22 The coal consumption on this vessel is from 1.80 lb. to 2 lb. 
per i.h.p. lir. This coal is of approximately 13,500 B.t.u. per lb. 

23 The problem of substitution of gas for steam, aside from the 
design -of the construction of the gas producers or cylinders of the gas 
engines, has been thoroughly worked out by Messrs. Babcock & Pen- 
ton, of Cleveland. The illustrations show two different arrangements 
of gas producers with the same engine. The proposed gas engine is a 
four-cylinder, double-acting, reversing type, having cylinders 24 
in. bore by 36 in. stroke, delivering 1000 b.li.p. at 100 r.p.m. The 
reversing is accomplished by means of compressed air, which is used 
to shift the cams from the head to the stern position. Compressed 
air is admitted to the cylinders by timed cams in proper cycle. The 
crank shaft of the engine is rigidly coupled to the tail shaft of the screw. 

24 The illustrations show a column-framed engine. Since making 
this layout, the design of the engine has been modified to meet all of 
the present marine conditions now found in marine engine design on 
the Lakes. In fact, with the exception of the condenser shown on the 



MARINE PRODUCER GAS POWER 191 

steam drawings, the gas-engine frame will be very similar to that of 
the steam engine. 

25 For the generation of current to drive the auxiliaries, there 
will be installed a double-cylinder, double-acting gas engine, direct- 
connected to a 50-lv^v. direct-current generator. All of the pumps 
and auxiliaries will be motor-driven. A smaller direct-connected 
unit operating on oil will be used for pumping, blowing fires, or 
other service, when the gas plant is down. Allowing a distance of 
4 ft. 3 in. between the forward bulkhead in the engine room and the 
forward side of the flywheel, which distance is one foot greater than 
that in the steam installation, we have an over-all distance between 
forward and after bulkheads in the engine room of 19 ft. 6 in. 

26 As previously stated, two arrangements of the producer room are 
shown. The first, the four-generator plant, consists of four 6 ft. by 
9 ft. generators, each fitted with independent economizers. The for- 
ward pair and the aft pair are connected independently to two 
horizontal gas scrubbers which are shown slung under the main deck 
beams. The gas passes from these scrubbers to independent motor- 
driven centrifugal gas-cleaning fans, whence it is delivered, either 
through common connection to a purge or blow-off pipe which also 
acts as a by-pass, or through two gas pressure regulator valves to the 
air and gas mixing valve at the engine manifold. The 6 ft. generators 
require only one cleaning door each. As a I'esult a single cleaning 
space suffices for the four macliines, allowing them to be grouped with 
reference to athwartship space, so as to give ample room on each side 
of the vessel for coal bunkers. The total space occupied by the pro- 
ducer plant is 21 ft. 10 in. athwartship, and 15 ft. between forward and 
after bulkheads. The producer room weight, including generators, 
economizers, piping, and scrubbers, complete, of the four-generator 
set, is 110,000 lb. This weight is estimated, but has been carefully 
checked and completely covers all the mechanism. In addition to the 
above mechanism, there will be a heating boiler which is shown on the 
main deck. This boiler will serve to furnish low-pressure steam for 
heating the vessel and supplying hot water for washing down decks, 
etc. This boiler, with water, will weigh about 8000 lb. 

27 The two-generator producer plant, which will undoubtedly be 
the one installed, will consist of two 8 ft. diameter by 9 ft. 6 in. generat- 
ors, connected to indepenrlent air economizers and each fitted with 
an independent horizontal scrubber, located athwartship under the 
main deck beams. The gas outlet at the scrubbers will be connected 
with a cross-over, so that either exhauster may operate either or both 



192 MARINE PRODUCER GAS POWER 

producer plants. The exhausters are installed in duplicate and are 
connected with common purge or blow-off and common gas outlets 
leading either through one pressure-regulator valve, or through a by- 
pass direct to the air and gas mixing valves at the engine manifold. 

28 On account of the fact that the 8-ft. generators require two 
cleaning doors set at 120 deg. the double generator unit plant will 
require the full athwartship space in the producer room. The approxi- 
mate floor space occupied, therefore, will be ;30 Jft. athwartship and 
15 ft. between forward, and^ aft ^bulkheads. The producer-room 
weight, including generators, economizers, piping and scrubbers 
complete for the two-generator set, is 82,000 lb. jThis weight is 
estimated, but has been carefully checked and completely covers all 
of the mechanism. As in the case of the four-generator plant, a 
low-pressure boiler for heating service will be installed. In the two- 
generator plant, however, this boiler will be located on the producer- 
operating floor, so that one set of firemen may suffice for both. 

29 The only guide we have for estimating the probable fuel con- 
sumption for this service is found in the large number of stationary 
producer gas power plants now in operation. Fortunately, in marine 
service, the load factor will be uniformly much higher than that found 
in any stationary service to which gas power is applied at the present 
time. The builders of this apparatus are prepared to guarantee one 
brake horse power per hr. on one lb. of good bituminous coal, averag- 
ing 13,500 B.t.u. per lb. 

30 Messrs. Babcock & Penton, the engineers who designed and 
built the steam plant, and who have spent years on the problem of the 
substitution of gas for steam, have suggested that the coal bunker, 
which will be placed above the charging deck of the producer, should 
have a capacity of about 80 tons of coal. These bunkers will run from 
the charging deck to the deck-house and will have doors opening 
closely adjacent to the charging doors of the generators, so that little 
or no coal passing on the operating deck will be required. 

31 In making the comparison shown in the table, it is unnecessary 
to go into the cost of fuel, labor, hours of service, etc., as these ele- 
ments vary with every class of service. In this particular proposition, 
it will suffice to state that the engineers who have been working on 
this substitution problem have conservatively figured that with the 
saving in fuel and the increased cargo carried, the cost of the com- 
plete plant will be saved in two years of operation. 

32 While the gas plant here described has neither been constructed 
nor ordered at this writing, its forthcoming will not be long delayed, 



MARINE PRODUCER GAS POWER 



193 



TABLE I COMPARISON OF POWER PLANTS FOR GREAT LAKES 
FREIGHT-CARRIER 

Length over all 306 ft. in. Displacement. . 6000 tons gross, 6600 

Beam 45 ft. in. tons net. 

^«P*^ 24ft. 0. in. (.^^g^ 4200 tonsnet, 18 ft. draft 

Speed, 12 statute miles per hr. on 900 
i.h.p. 
Steam Gas 

engine room engine room 

3-cylinder triple-expansion, condens- 4-cylinder, 4-cycle, double-acting, gas 

ing, 18-30-50 by 36 in., 1050 i.h.p. at engine, 24 in. diam., by 36 in. stroke 

90 to 95 r.p.m. 1000 b.h.p. at 100 r.p.m. 

AuxiUaries steam-driven Auxiliaries motor-driven 

Length between bulkheads, 22 ft. in. Length between bulkheads, 19 ft. 6 in. 

Engine room weights, including auxili- Engine room weights, 105,000 lb. 
aries and piping, 182,000 lb. 



BOILER ROOM 

2 single-ended Scotch boilers fitted 
with economizers, forced draught. 
Length each boiler, overheads 11 ft. 
Oin. 

Mean diameter, each, 11 ft. 10 in. 

Two 42-in. furnaces each 

244 2|-in. tubes, each 

Grate surface, each, 36.75 sq. ft. 
Heating surface, each, 1642 sq. ft. 
Boiler room weight, water in boilers, 

no fuel, 170,000 lb. 
Length boiler room 24 ft. in. 
Length boiler room, includes bunkers, 

30 ft. in. 



Square feet boiler room, including 

bunkers, 900 
Square feet per h.p., 0.9 

Bunker capacity, 340,000 lb. 

Total weight, machinery and fuel, 692,- 
000 lb. 

Total length of machinery space includ- 
ing bunkers, 52 ft. in. 



PRODUCER ROOM 



Two down-draft gas producers and 
auxiliaries 



Diameter shell, each generator, 8 ft.O in. 
Inside diameter, lining generator, 6 ft. 

Sin. 
Height shell, each generator, 9 ft. 6 in. 
Grate surface, each generator, 30.67 

sq. ft. 
Producer room weights, no water, no 

fuel, 82,000 lb. 

Length producer room, includes bunk- 
ers, 15 ft. in. 

Square feet producer room, 450 

Square feet per h.p., 0.45 

Square feet producer room with four 
smaller generators, 330 

Square feet per h.p., four generators, 
0.33 

Bunker capacity, 160,000 lb. 

Total weight, machinery and fuel, 
347,000 lb. 

Total length of machinery space, 34 ft. 
6 in. 

Saving in weight, 355,000 lb. 

Saving in fore-and-aft length, 17 ft. 6 in. 

Saving in cubic space 17 ft. 6 in. by 32 
ft. beam by 20 ft. high, 11,200 cu.ft. 



194 



MARINE PRODUCER GAS POWER 




MARINE PRODUCER GAS POWER 



195 






^ai}=^. 




'fj CO 



S O 

X o 






^5 2 

CO 



- -A 
O ^ 






t^ 



196 



MARINE PRODUCER GAS POWER 




MARINE PRODUCER GAS POWER 



197 




198 



MARINE PRODUCER GAS POWER 




MARINE PRODUCER GAS POWER 



199 




200 DISCUSSION 

and this comparison, while somewhat premature, is made to present 
the possibilities of marine producer gas power to those interested in 
its future. 

33 A marine bituminous gas plant, similar in construction and 
operation to the one described, but of 300-h.p. capacity, has been in 
commercial operation driving a six-cylinder, single-acting, reversing 
marine gas engine for over a year. The results obtained give ample 
security for the statements made in this paper, and point to the early 
adoption of this type of prime mover for our marine commercial ser- 
vice. 

DISCUSSION 

C. L, Straub. We have received reports from abroad of progress 
in the marine gas-producer field, a summary of which I hope will 
prove of interest. 

2 In Holland teaming is practically unknown and local freight 
trains are never run, the canals being used for moving freight from 
one city to another or between different parts of the same city. 
The canal barges range from 40 ft. in length, with an 8-ft. beam, 
drawing 3 ft. of water, to 150 ft. in length, with a 20-ft. beam, drawing 
6 ft. of water. The majority of the barges are hand-propelled, about 
9 per cent have steam equipment, while about 6 per cent are pro- 
pelled by gas engines. Of the last-named, a few — about | of 1 per 
cent — use gas producers, the others using liquid fuel. 

3 The gas-engine barges range in size from 40 ft. to 130 ft., the 
engines from 10 h.p to 300 h.p., and the engine speed from 250 r.p.m. 
for the larger to 400 r.p.m. for the smaller engines. The reason for 
the small number of power boats is the great scarcity of fuel. Holland 
is without coal mines or any natural growth of timber. Hence coal 
is expensive and difficult to obtain, though wood is more plentiful as 
large quantities of lumber are shipped in from Germany. Peat is 
used a great deal, while compressed peat and some domestic and 
imported briquettes are burned to some extent. 

4 The gas-engine boats using liquid fuel are more handicapped by 
fuel conditions than the steamboats are. The Standard Oil Co. 
supplies fuel in a few large cities such as Amsterdam and boats of 
large power can work around these fuel depots only. This limit to the 

Note.— =-At the time of the presentation of his paper Mr. Straub also pre- 
sented a report on gas producer development abroad, with special reference 
to marine work. An abstract of this report is here given as a part of the 
discussion on his paper. — Editor. 



MARINE PRODUCER GAS POWER 201 

radius of action prevents the extensive equipment of boats with gas 
engines operating on liquid fuel. 

5 The producers installed on the comparatively small number of 
boats include principally modifications of standard German stationary 
types. They are not successful and apparently cannot be made to 
give continued good results with the fuels available. Producers that 
will operate economically on peat, wood, briquettes or poor coal, and 
require little attention, are in great demand. The engines are giving 
satisfaction. Boat owners want gas-engine equipments but not for 
use with liquid fuel. There is a great demand for gas-producer equip- 
ment of capacities up to 300 h.p. for inland waterway traffic, and for 
capacities up to 600 h.p. for general inland and foreign traffic. 
Holland thus offers an attractive field, because success with these 
small capacities will open the way for larger equipments. 

6 A test run of two tug-boats was made from Hamburg to Kiel 
and return, one boat having a steam equipment, and the other a gas- 
producer equipment. The results are shown in Table 1. The 
weather was rough and the speed maintained 8^ knots an hour. 
The coal consumption for a period of 10 hours was as follows: Gastug, 
530 lb. anthracite coal; Elfrieda, 1820 lb. bituminous coal; an econ- 
omy in coal consumption of nearly 3.5 to 1 in favor of the gas-pro- 
ducer equipment. 

TABLE 1 COMPARISON OF STEAM AND GAS-PRODUCER EQUIPMENTS 





Gas 


Steam 


Name 

Length 

Beam 

Horsepower 

Towing meter pull 


Gastug 
44 ft. 3 in. 
10 ft. 6 in. 
4 cyl., 70 b.h.p. 
2140 lb. 


Elfrieda 

47 ft. 

12 ft. 

Triple-Exp., 75 b.h.p. 

2020 lb. 



7 Herr Korting has established a marine department at Kiel, and 
has practically a monopoly of the government submarine work. He 
is experimenting and preparing to take up larger work. The Ger- 
man government has decided to adopt gas-engine boats for all war- 
ship jaunches, pinnaces, videttes, portable torpedo boats, and the like. 

8 The Niirnberg Company (Vereinigte Maschinenfabrik Augs- 
burg und Maschinenburg Gesellschaft Niirnberg A. G.) have installed a 
number of small Diesel oil engines for marine use. Both suction and 
pressure producers are built by them. As to fuel, the company states 
that anthracite, coke, charcoal and lignite briquettes are the most 



202 DISCUSSION 

suitable for producers. Efforts to gasify raw pit coal have not been 
entirely successful, at least in medium and small-sized plants. 

9 In Great Britain, Vicker Sons & Maxim, Ltd., have built 40,000 
h.p. of marine high-speed gas engines. Plans have been drawn for an 
oil-engine torpedo-boat destroyer of about 30 knots speed. With the 
same dimensions and speed, the oil-engine destroyer saves enough 
weight and space so that the armament may be increased from the 
one 12-lb. and five 6-lb. rapid-fire guns of the steam destroyer to four 
33-lb. and two 6-lb. rapid-fire guns. Moreover the quantity of 
ammunition (number of rounds per gun) is the same for l30th boats, 
although the guns and ammunition per round of the oil-engine 
destroyer are much heavier than those of the steam-engine destroyer. 
Sufficient fuel capacity is provided to allow a speed of 30 knots an 
hour with a radius of action 6^ times greater than that of the steam- 
driven destroyer. Also more space below decks allows superior 
accommodations for the crew. 

10 A marine gas-producer equipment built by the Beardmore 
Company under the Capitaine patents was tried out with satisfactory 
results. The capacity of the plant was 70 h.p., the engine having 
four cylinders 8 J in. in diameter and of 11-in. stroke. The weight 
of the whole was shghtly over 13,440 lb. The equipment was installed 
in a launch 60 ft. long and of 10-ft. beam. A 10-mile run was made 
in one hour without recharging the producer. 

George Dinkel. I would like to ask the author if there are any 
gas-producers working with the small grades of the steam sizes of 
anthracite coal, such as No. 3, 2 and 1 buckwheat, especially as he 
states in Par. 8 that there are in commercial operation in this country 
today two distinct types of stationary power gas-producers suited by 
their design for operation on almost any class of solid fuel. Where 
are those producers being used on the same grade of steam sizes, and 
what has been the result? 

Henry Penton.^ As the members probably know, freight is 
carried on the Great Lakes at a lower cost than anywhere else in the 
world, and over 75 per cent of the merchant steam tonnage of the 
United States is built on the Great Lakes. We are, of course, con- 
stantly seeking methods of reducing carrying costs, and so far as the 
ship is concerned the most important item of expense is that of power. 
We have for some time been firm believers in gas power: power for 

1 Henry Penton, Babcock & Penton, Cleveland, O. 



MARINE PRODUCER GAS POWER 203 

propulsion, however, is only one of the problems to be met; the prob- 
lem of the application of gas power to the auxiliary service has given 
us more concern. At present our ships are handled entirely by steam, 
both in port and out, and we must have power available at all times. 

2 We believed we could depend upon the combustion engineer to 
perfect the producer- gas engine when the opportunity arrived; but 
it has arrived and while we believe the producer to be satisfactory 
we are not yet satisfied with the engine. 

3 We first considered the installation of gas power in one of our 
modern ships where the average horsepower (indicated) is in the 
neighborhood of 2000, but subsequently decided to select a smaller 
type, believing that we should practice creeping before walking. 
While the ship to which the author has alluded represents the best 
standard practice in design, she is not a representative lake steamer 
in that she is relatively small. Her engines develop approximately 
1000 i.h.p. Her carrying capacity is about 4080 tons on 18 ft., 
while our modern ships run to 12,000 and 13,000 tons on the same 
draft. The fuel is necessarily bituminous coal, fairly uniform in heat 
value and averaging about 13,500 B.t.u. Anthracite and coke are out 
of the question both as to delivery and cost- 

4 It should be noted that the fuel consumption given by the 
author includes fuel used for all purposes aboard ship charged against 
the actual indicated horsepower. This is the customary method of 
stating the consumption and is used merely for purposes of compari- 
son. The propelling engines actually do their work on an average of 
1.5 lb. to 1.65 lb. per i.h.p. per hr., including the auxiliaries. 

5 In Par, 25 mention is made of the installation of a 50-kw. direct- 
current generator. I think, however, it will be necessary to use at 
least two of this size, depending somewhat on the method adopted 
for handling the windlass, which calls for the largest individual motor 
of any of the apparatus. There must be no such thing as a generator 
shutdown. As just stated, every function of the ship, including 
propulsion, is now performed by steam, and power must be available 
every instant from the time the ship goes into commission in the spring 
until she is laid up the following winter. If we take out steam we 
must provide something just as available in its place. While a great 
part of the time the output will be small, we must be able to take care 
of the maximum requirements. 

6 It may be wondered why, as stated in Par. 26, the auxiliary 
boiler is required for heat and for washing down decks. These ships 
run until well into the winter and when the weather becomes severe 



204 DISCUSSION 

they ice up badly, and it is not uncommon to make port with 200 or 
300 tons of ice aboard. The quickest method of clearing away is 
with the hot-water hose. I hope we shall be able to use the exhaust 
gases for generating steam at sea and thus operate the boiler without 
the use of coal; but I do not know whether this is yet feasible. 

7 With reference to figuring the elimination of cost in two years, 
the operation is brought about in this way: some of our ships get in 
more, some not so many, but the average is not far from 20 round 
trips per year; and taking into consideration the reduction in weights, 
which means additional revenue-producing cargo capacity; reduced 
space, which in some trades is also additional cargo capacity; and 
reduction in fuel, which is both reduced expense and additional cargo, 
I have succeeded in convincing myself that we can get even in about 
1^ seasons; but two seasons is perfectly satisfactory, and you can 
readily see that the addition of one or more trips in the year wUl go a 
long way toward the extinction of that cost. 

Irving E. Moultrop. Examination of Table 1 gives some very 
interesting information. It is rather surprising that the total weight 
of the complete gas-power plant is so much less than that of the steam 
plant. Of course, the steam plant has a number of auxiliaries which 
the gas plant does not require, and these auxiliaries are quite heavy, 
yet up to the present time, in stationary practice, the total weight of 
a gas engine has been verj'- much in excess of that of a steam engine for 
the same power. The two prime movers discussed in this paper oper- 
ate at the same speed; one would naturally assume, therefore, 
that the extra weight in the gas engine would go far toward making 
up for the saving in weight due to the omitting of a number of steam 
engine auxiliaries. As the total weight of the gas engine room machin- 
ery is only about 60 per cent of that of the steam engine room machin- 
ery, one naturally wonders if the factor of safety in the gas engine 
design has not been reduced to save weight, or if this is not the case, if 
some special weight-saving features have not been introduced in the 
gas engine design, which might have been used to equal advantage in 
the steam engine. 

2 Comparing the producer room with the boiler room it is inter- 
esting to note that the total grate area of the producers is only about 
five-sixths of that of the steam boilers. Stationary practice has 
shown that the best producer results are obtained at a very much 
lower rate of combustion per square foot of grate than in good steam 
boiler practice. It would be interesting to know how the engineer of 



MARINE PRODUCER GAS POWER 206 

the gas-power plant expects to obtain capacity out of its producers 
when the full consumption per square foot of grate area in the pro- 
ducer will exceed what is considered good practice on the grate of 
steam boilers. 

3 In comparing the total weight, machinery and fuel, in the gas- 
power plant with the steam plant, and also the total space occupied, 
it should be noted that in the gas plant the bunker capacity is less 
than half that in the steam plant. 

Herbert M. Wilson.^ Perhaps it would not be a breach of con- 
fidence for Mr. Straub to tell us something concerning the gas producer 
for the new non-magnetic vessel of the Carnegie Institute. This 
vessel is being constructed with as little iron as possible, for use in 
magnetic surveys; and I understand the gas producer and gas engine 
were selected for auxiliary power because of the small weight of metal 
required and the possibility of substituting bronze and other non-mag- 
netic metal for iron and steel. Their vessel is about to be launched, 
and perhaps Mr. Straub could tell something of the gas producer engine 
plant which is under construction for actual operation. 

E. T. Adams. Great changes in the weight required have come 
about in the past few years. The first designs in any line of manufac- 
ture uniformly carry unnecessary weight in the various parts and it is 
sale to say that the gas engine of today, built with the same factor of 
safety, would be 25 per cent lighter than any engine of the same horse- 
power built three years ago. 

2 This applies to engines in use for ordinary commercial purposes, 
as electric lighting or power. In view of this fact the statements 
of the author on this point are not at all surprising. The increasing 
use of steel and the modification of design based on experience have 
led to great reduction in weight of engines for special purposes, such 
as are here specified. 

The Author. With regard to the question of the speed of vessels 
depending on the character of power equipment, taking a given hull 
and a given power equipment, the craft will go at a certain speed, the 
character of the equipment notwithstanding. The comparative 
space required by steam and by gas equipments can be best shown 
by an example given by Capt. A. B. Willits, U. S. N., in an article. 
Gas vs. Steam for Marine Motive Power, printed in the United 
States Naval Institute Proceedings, December 1908. Mr. Willits 

* With U. S. Geological Survey, Washington, D. C. 



206 DISCUSSION 

states that the floor space occupied by the boilers in the New Hamp- 
shire equals 0.33 sq. ft. per h.p., and that the weight per b.h.p. is 
110 lb. The power of these boilers is rated at their forced capa- 
city. If we install a producer plant and rate it at its forced capa- 
city at 40 lb. of fuel per square foot of grate, it will occupy 1/10 
square foot per b.h.p., for such a plant as the New Hampshire would 
require, and weigh 30 lb. per b.h.p. 

2 From these figures it is apparent that a marine producer-gas 
plant can be installed in at least the same space and of certainly 
not greater weight than a modern marine steam boiler plant. 
These figures refute Capt. Willits' figures of the Westinghouse gas 
plant, which occupied 1 sq. ft. per b.h.p., and weighed 28.5 lb. per h.p., 
and the R. D. Wood producer which occupied 1.84 sq. ft. per b.h p. 
and weighed 194 lb. 

3 Mr. Dinkel asked regarding the small anthracite coals in gas 
producers. I can refer him to the generator of the R. D. Wood plant 
at Jersey City, which has been operating at the plant of the Erie 
Railroad for five or six years on a mixture of No. 1 and No. 2 buck- 
wheat coal. The Lehigh Coal and Navigation Companv has in- 
stalled a gas producer which has operated on rice coal and has 
been running for almost two years. We have two plants in opera- 
tion, one at Hartford, Conn., and one near Philadelphia, running on 
a fine grade of anthracite coal, mixtures of Nos. 1, 2 and 3 buckwheat. 

4 Regarding the factor of safety in gas-engine parts, which Mr. 
Moultrop brought up, the six-cylinder single-acting engine referred 
to in Par. 33 weighs less than 30 lb. per b.h.p. Of course, that was 
a comparatively high-speed engine and delivered 300 b.h.p. at 320 
r.p.m. The 1000-h.p. engine will be fitted with cast-steel parts, 
in almost every instance where cast iron was used on the steam 
plant, and this makes for a big reduction in weight at a very slightly 
increased cost. 

5 The producer and equipment for the Carnegie, about which Mr. 
Wilson asked, is almost finished and will be in the boat when she is 
launched on Maj^ 10. The producer shell is about 6 ft. in diameter 
and is of copper. The pipe and scrubber are of composition metal, 
containing no iron or steel. The only steel or iron parts on the pro- 
ducer are manganese steel grates, doors and door frames, near the 
hot portion of the fire in the producer equipment, and this manganese 
steel is less than one per cent magnetic, when compared to mild steel, 
so that of it we have been allowed the use of 1500 lb. The engine 
will have bronze cylinders and will not be lined with cast iron, as the 



MARINE PRODUCER GAS POWER 207 

published reports indicate. We are going to run cast-iron pistons in 
the bronze cylinders. The cylinders are comparatively so thin and 
so close to the water jackets, that we anticipate no trouble from dete- 
rioration. The only steel or iron parts about the engine will be the 
cams, rollers and valves. The valves will be of cast iron. The steel 
cams and rollers will be hardened and ground. On a commercial 
basis, using terms equivalent to mild steel, as far as magnetic force 
is concerned, we will have less than 200 lb. total of iron or steel in 
that vessel. The published reports make further detail unnecessary. 
The boat will be ready to sail July 1. 



No. 1239 

OPERATION OF A SMALL PRODUCER GAS- 
POWER PLANT 

By C. W. Obert, New York 
Associate Member of the Society 

It has been the practice of the packing house of Swift & Company 
of Chicago, in the distribution of meats and provisions to retailers, to 
establish in different cities distributing depots with the necessary power 
equipment for the handling and refrigeration of the products. Some 
of these branches in the larger cities are establishments of consider- 
able size, and with the extensive cold storage facilities required for 
the large stocks carried, require comparatively large power installa- 
tions. The new Westchester market, which the company has recently 
built in New York at 152d Street and Brook Avenue in the Bronx, is 
a notable installation of this kind, involving a 400-h.p. producer gas- 
power plant for the operation of both refrigerating and electric 
generating machinery, which supplies similar service to a number of 
adjoining depots of other houses. 

2 The refrigerating duty at present required embraces the opera- 
tion of a total cooling system containing over 46,000 ft. of 2-in. pipe, 
which reaches a maximum of over 100 tons of refrigeration per 24 
hours under the most severe summer weather conditions. Two 65-ton 
refrigerating machines were installed for this service, with equip- 
ments in duplicate, owing to the great importance of continuity of 
refrigeration, particularly in hot weather. A maximum of nearly 90 
h.p. is required for compression machines of this size and engines of 
100 h.p. were selected for driving them, to provide sufficient capacity 
for unfavorable or overload conditions. 

3 The electrical load, which includes the operation of several 
electric elevators, fluctuates ordinarily between 30 kw. and 50 kw. 
but occasionally reaches a maximum of over 60 kw. For this service, 
duplicate 75-kw. generators were installed, with driving engines of 100 

Presented at the Spring Meeting, Washington, May 1909, of The Amer- 
ican Society of Mechanical Enqineeys. 



210 OPERATION OF PRODUCER GAS-POWER PLANT 

h.p. This was done to secure uniformity of size and detail in all four 
of the driving-engines. 

4 For gas making, two producer equipments were installed, also in 
duplicate. One of these is a 200-h.p. producer, intended for the 
supply of one refrigerating machine and one generator engine when 
operating at maximum capacity. The other is of 150-h.p. capacity 
to permit of closer adjustment of the producer capacity to the load at 
other times. 

5 The plant arrangement consists of an engine room in the easterly 
end of the sub-basement of the market building, and a producer room 
adjoining, the entire power equipment occupying a total space, includ- 
ing fuel storage, of 48 ft. by about 55 ft. Headroom for the machin- 
nery and piping is afforded by the depression of the sub-basement 
floor to a level 16 ft. below the street, and the omission of the base- 
ment floor in this section, giving thus a clear headroom of 18 ft. The 
machinery space was originally laid out as a single room, but as a 
result of the requirements of the underwriters, the producer space has 
been separated from the rest by a 6-in. hollow- tile fire wall, forming a 
producer room 20^ ft. by 24 ft. maximum dimensions. Under the 
152d Street sidewalk, there is an 11 ft. by 29 ft. room containing 
pumps and auxiliaries for the power equipment and the building 
heater; and adjoining this, an 11 ft. by 30 ft. space for fuel storage. 
The latter has capacity for over 150 tons of coal, which is dumped into 
it through sidewalk coalholes from wagons in the street. 

6 The engines are Rathbun vertical, three-cylinder units, of 100 
h.p., rated at 280 r.p.m., built by the Rathbun-Jones Engineering 
Company, Toledo, Ohio. The two for the electrical service are direct- 
connected to 75-kw. generators and the other two through silent chain 
drives to the ammonia compressors of the refrigerating equipment. 
They are all of the four-stroke cycle, single-acting, enclosed type, 
and have 12f in. by 13-in. cylinders, designed for the above rating 
when operating on producer gas of not less than 125 B.t.u. per cu. ft. 
These engines are throttle-governed, a special form of centrifugal 
flyball governor being used, and have each a one-ton fly-wheel at 
both ends of the crankshaft. 

7 The gas is generated for the engines in a duplicate equipment of 
Smith suction producers built by the Smith Gas Power Company, 
Lexington, Ohio. Each equipment consists of a simple shell pro- 
ducer, a wet scrubber and a dry purifier. While the producers differ 
in rated capacity to permit of more accurate adjustment of their 
capacity to the power requirements at different seasons of the year, 



OPERATION OF PRODUCER GAS-POWER PLANT 



211 




m 
z 

o 
o 
a 






o 
o 

t-H 

o 

a 

03 



212 



OPERATION OF PRODUCER GAS-POWER PLANT 



the scrubbers and purifiers have a maximum capacity of 200 h.p., 
which permits the smaller producer to operate up to the maximum 
plant capacity of 200 h.p., if required to do so temporarily. The small 
and large producers have 6-ft. and 7-ft. shells respectively, both 12 ft. 
in height, their internal diameters being 4^ ft. and 5J ft. respectively, 
and they are fitted with shaking grates on the up-draft principle for 
operation with anthracite coal. They are not fitted with attached 
vaporizers or air pre-heaters, but have an automatic control attach- 
ment for regulation of the amount of water vapor to conform to the 
power requirement and consequent rate of gasification. The scrub- 
bers for cleansing the gas are vertical cylindrical tanks, each 4 ft. in 
diameter by 15 ft. high, and the dry purifiers have 4-ft. shells 6 ft. in 
height. 



irat Floor | [ I 




?i^'^f:&^k^.-r^l-ii.^:^' "'■■ '"--" :\. h-^^^^^" 



23 



Fig. 2 Elevation of Machinery Room in Cross-Section 



8 The piping of the plant was somewhat involved by the arrange- 
ment of the engines relative to the producers, and, in the Smith 
producer system, by automatic vaporizers in the exhaust connections 
to utilize the waste heat of the engines for the vaporization of the 
water. The vaporizers are located close to the engines and attached 
to each vaporizer is an automatic device, through which air is admitted 
and superheated for the producer. The air is conducted to the pro- 
ducers from these devices by a 10-in. pipe main, extending through the 
engine room, and heavily covered with magnesia insulation. 

9 The gas is delivered from the producers by 8-in. pipes connecting 
from the top of the producer to the bottom of the scrubber shell and 
each scrubber has a triplicate connection to its corresponding purifier, 
which is a three-part filter. From these the gas is conducted to the 
engines through a 5-in. line, with a S^-in, branch to each. The exhaust 
connections from the engines to the vaporizers are 5-in. lines and from 
the latter, individual discharge pipes are carried up for each engine 
through a pipe shaft in the corner of the building to a roof outlet. 



OPERATION OF PRODUCER GAS-POWER PLANT 213 

It is to be noted that this arrangement of exhaust connections is 
effective in so muffling the noise of the escaping gases that they can- 
not be heard from the adjoining street and are only barely noticeable 
when on the roof close to the outlets. 

10 The electrical generators are 75-kw. General Electric direct- 
current machines, each rigidly coupled to the driving engine. They 
are wound to deliver current at 220 volts, the distribution for both 
lighting and power being on the two-wire system. The electrical 
circuits are controlled on a three-panel switchboard which contains 
the usual equipment of indicating and recording instruments, field- 
rheostat switches and generator and feeder switches. The building 
is wired separately for lighting and power circuits, and recording watt 
meters are connected into the feeder circuits for measurement of the 
power delivered. It is to be noted that separate bus bars are provided 
for both power and lighting feeders, as well as a switching arrangement 
by which the lighting service may be supplied from a generator other 
than that carrying the power load, incase the fluctuations of the latter 
should interfere with the voltage regulation. This provision has been 
found unnecessary, however, as the speed regulation of the engines and 
generators is satisfactory under all fluctuations of loading due to ele- 
vator operation. 

11 The refrigerating equipment was installed on the direct 
ammonia expansion system, a feature of which is the connection of all 
coils in the coolers in series with those in the freezers, whereby all 
ammonia not thoroughly evaporated in the freezer coils will be in the 
cooler coils (temperature, 36 deg. fahr.), which permits carrying 
the freezer temperature at from deg. to -H 5 deg. without frosting 
the compressor. The compressors were built by the Hutteman & 
Cramer Company, Detroit, Mich., and are horizontal single-cylinder 
double-acting machines, with 14-in. by 30-in. cylinders, each driven at 
a speed of 60 r.p.m. by a Renold silent-chain connection from its 
driving engine, with a speed reduction of about five to one. 

12 The ammonia condenser is located on the roof of the building 
and provided with the usual water-cooling sprays. The water supply 
for it is obtained from a well extending into water-bearing soil under 
the basement floor, and the drainage from the sprays is subsequently 
utilized in the scrubbers and in the engine cylinder jackets. One of 
the compressor units normally handles the load alone, which leaves 
one equipment always in reserve, to provide against the serious emer- 
gency of a complete stoppage of the refrigerating service during hot 
weather. 



214 OPERATION OF PRODUCER GAS-POWER PLANT 

13 In operation this plant has proved particularly economical, 
largely due to the continuous character of the service resulting from 
the operation of the refrigeration plant 24 hours a day, seven da3^s a 
week, thereby eliminating standby losses. The average load range 
of the plant is ordinarily from 50 per cent (100 h.p.) to full rated load 
(200 h.p.), the high and low load factors occurring during the summer 
and winter months respectively, when the refrigeration requirements 
are maximum and minimum. With the heavier load factor during 
the summer months, the fuel consumption has ranged between 3400 
and 4800 lb. per 24 hours, the larger figure having been exceeded on 
only two days in 11 months, and the consumption per horsepower- 
hour as calculated from station fuel records and observed loads, 
ranged from 1.4 to 2.0 lb. of coal. The fuel rate has dropped during 
periods of continuous high loads, to about 1 lb. per horsepower-hour, 
as based on observed loadings, but the daily average under conditions 
of ordinary commercial operation is usually greater. 

14 The operating conditions during the heavy-load season are indi- 
cated in the table at the end of the paper, in which the relation of 
fuel consumption to load carried is shown for two weeks of similar 
duty. The variations in the amount of fuel charged from day to 
day are due chiefly to the differing conditions of the fuel bed in the 
producer, the removal of a particularly large amount of ashes on any 
day necessitating a heavy fuel charge. No account is taken of cost of 
water used in the scrubbers and cooling jackets, as the supply is 
obtained from a well on the premises without cost other than that of 
pumping. 

15 The fuel used is No. 1 buckwheat anthracite that has been 
passed over a f-in. mesh and through a yVi^^- mesh screen, with 5 per 
cent fineness, and costs $3.50 pergross ton delivered in cargo lots. It is 
charged only at the regular cleaning periods, at each of which from 
400 to 900 lb. of coal are fed, after the fire has been cleaned down and 
the ashes removed from the grate. The fire is cleaned periodically 
twice every shift,or four times per 24 hr., and requires about an hour 
per cleaning on the average. 

16 In this connection it is interesting to note the comparatively 
short time required to start a producer into service from the cold, 
which has been done repeatedly on short notice in about five hours; 
on December 12 when the 150-h.p. producer was placed in operation 
to relieve the larger unit, the kindling wood was lighted at 10 a.m. and 
the gas supply turned onto the engine at 2 p.m., with only about 12-in. 
of fire zone in the fuel bed. The reliability of a suction producer 



OPERATION OF PRODUCER GAS-POWER PLANT 215 

operating under a continuous and exacting service of this character 
is well shown by the duty of the 200-h.p. producer during the summer 
season of 1908, which when taken out of service on December 12, had 
been continuously in service 24 hours per day and seven days per week 
since April 22, a continuous run of 235 days. During that time, it had 
received no more attention than the four cleanings and chargings per 
24 hours. 

17 The operating force for the power plant consists of an engineer 
and an assistant engineer and two producer tenders, who work in two 
shifts. This force is able to maintain the plant equipment in satis- 
factory operating condition, as well as the refrigerating and electri- 
cal equipment of the depot, and it is worthy of note that the plant has 
not been shut down for any reason since it was started on February 1, 
1908, a period of 15 months. Experience during this period indicates 
that, contrary to the general opinion, no more attention is required 
than for a first-class steam plant, the necessary attendance comparing 
very favorably with that of a high-grade steam plant of the same 
capacity. CleanHness of all parts of both producer and engine equip- 
ments, and careful adjustments, especially of the latter, are imperative 
and are the keynotes of successful operation. In order to maintain the 
equipment in such condition, a thorough and comprehensive operating 
system has been developed which may be of interest. 

18 The operating system involves a detailed and thorough 
inspection routine that keeps the force well informed as to the condi- 
tion of the entire equipment, and a division of duties tending to favor 
the maintenance work. To the day operating force is assigned the 
inspection and adjustments of the engines and repairs to igniters, 
batteries, etc., while the night force has the work of cleaning all 
machinery. 

19 The regular routine of the day force is in detail as follows: 
First upon coming on duty at 7 a.m., an examination is made of all 
moving parts of the two engines in operation, and also of oil levels in 
lubricators and conditions of water jackets and ignition systems. 
There are always two engines in operation, one being a generator 
engine and the other a refrigerating engine, which in the periods of 
heavier loadings in summer time have a combined load of about 140 
h.p. of which fully 75 h.p. is taken by the refrigerating system. Next 
the water regulation for the steam supply is noted and then the con- 
dition of the suction draft on the producer and also on the scrubber 
and purifier, there being three U-shaped draft gages provided for this 
purpose, one connected to the gas suction Hne to the engines, the 



216 OPERATION OF PRODUCER GAS-POWER PLANT 

second to the gas connection from the scrubber to the purifier and the 
third to the connection between the producer and scrubber. A 
uniformity of suction of from 2 in. to 3 in. of water in these three 
gages indicates a proper condition of the three units, while any unusual 
suction in any of the connections would indicate an obstruction 
needing immediate attention. The latter is always clearly indicated, 
as an obstructed condition in the producer, for instance, will raise 
the suction to as high as 9 in. or 10 in. of water. 

20 Next an inspection is made of the producer, the temperatures 
of different portions of the fire being determined to ascertain the con- 
dition of the fuel bed, the existence of cracks or fissures or pockets of 
unburned coal. To do this, a j\ -in. iron rod is pushed into the fire 
through the side peep holes in the producer shell, held there exactly 
one minute and then withdrawn, the temperature within being noted 
from the color of the rod. If the latter is at a uniform cherry red 
temperature throughout its length, this is taken as an indication of an 
even fire; but if at a brighter heat or dull in some portions of the rod, 
there is evidence of unnecessarily high local temperatures due to rapid 
combustion in fissures in the fuel bed, or of a stagnant condition in 
dirty or unburnt portions of the fire. The rod is first inserted in the 
lowest hole and then successively into the upper holes, in order to 
explore the fire in zones. On withdrawing the rod the operator 
notes graphically the condition of the fire by marking a line with 
chalk on the shell of the producer even with the hole, a straight line 
indicating an even temperature, and a broken line sho\ving the dirty 
condition, etc. This operation is continued for the four holes and a 
fuel curve drawn from it which gives a practical idea of how the dirt 
lies in the producer and shows what quality of gas can be expected. 
Finding the producer in good order, the scrubber, purifier and connec- 
tions are examined for unusual temperatures, condition of water flow, 
etc. 

21 In the maintenance work, each engine is shut down after every 
seven days work of 160 hours for general inspection and cleaning, 
and thus on Monday mornings it is necessary to start up the two 
reserve units and transfer the respective loads to them. Before 
starting up either reserve engine, its igniters are cleaned, which takes 
about one hour. With the igniters clear and everything in good 
order, the attendant looks at the draft gage, which is equal in impor- 
tance to the gage of a steam boiler, to see what gas the engines in 
operation are drawing and whether the start can be made without 
interfering with their suction. If there are any doubts the gas is 



OPERATION OF PRODUCER GAS-POWER PLANT 217 

enriched temporarily by putting about four pails of water in the ash 
pit of the producer and slicing the fire to work down some hot coals, 
which, by turning the water into vapor, increase the hydrogen content 
of the gas and enable the third engine to be started without inter- 
fering with the others. After getting the engine warmed up, the load 
is thrown on and the other engine is shut down. The extra pull on the 
producer, due to overload from running the three engines and the 
hydrogen added, has usually so enriched the gas that on cutting out a 
unit the quality of gas is too rich for the two units operating alone. 
To counteract this, it is necessary to give additional air to each of the 
units that remain and then, as in starting, there will be no varia- 
tion in speed of operation. 

22 After the engines have been shut down their inspection is begun 
by the removal of the back crank case covers and examination of the 
bearings, crank pins, wrist pins, etc., for necessary adjustments. 
Besides this the exhaust valves are cleaned and the ignition system 
checked. This requires about two days, as but one thing is done at a 
time and then only at times when the load on the plant is not heavy. 
While the engineer is performing this work, the producer tender pre- 
pares to clean and coal the producer, as follows: 

23 The method of cleaning is to rake off the ash from the grate 
table and then poke down around the shell from the top poke holes. 
Having before him the fuel chart which was noted graphically on the 
producer shell on coming on watch, the attendant knows what part 
of the bed requires most poking. Before opening the ash pit doors 
about the shell, water is placed in the ash pit as before and the hot 
ashes, dropping down, form sufficient steam to mix with the air 
coming through the ash pit door and offset any bad effect therefrom. 
This enables the cleaning to be done without affecting the engines. 
Having cleaned and poked the fire thoroughly and worked down all 
the ash so as to leave it as clean as possible, the coaling is then begun, 
count being taken of each hopper of coal charged. The coal is cleaned 
by screening if very fine or dirty. Having coaled, the operator slices 
across the grate so as to relieve the center of the fire and again puts 
water in the ash pit, this time to cool off the grate after cleaning and to 
offset the effect of any air that may have gotten in during the opera- 
tion. The cleaning usually occupies one hour, the amount of coal 
put in ranging up to 900 lb. After giving the producer time to settle 
down the ashes are withdrawn from the ash pit, an average of 1^ ash 
cans (about 3 bushels) being removed after each cleaning. During the 
cleaning operation the operator is always on the lookout for any 



218 



OPERATION OF PRODUCER GAS-POWER PLANT 



change in the engine speed due to weak gas on account of opening the 
ash doors. Should this occur he immediately cuts the air supply to 
the engine, resulting in a combustible mixture without noticeably 
reducing the speed. The producer is now good for 6 hours' operation, 
after which the cleaning is repeated. 

24 The refrigerating engines are operated for periods of 84 hours 
and then gone over. One exhaust valve is taken out of an engine 
each week, thoroughly cleaned, and regTound if necessary, thus 
insuring attention to each valve once in every three months. Igniters 
are cleaned weekly and the batteries and ignition system checked. 
The temperature of the fuel bed of the producer is taken twice a day 
and a gas analysis is made once a week or oftener if necessary. The 
average calorific value per cubic foot of gas is 134 B.t.u., based on 
analysis: COj, 8.6 per cent; O, 0.6 per cent; CO, 20.2 percent; H, 
18.5 per cent and N, 52.1 per cent. 



TABLE 1 RECORD OF LOAD AND FUEL FOR TWO HEAVY WEEKS 



ELECTRICAL LOAD jRcfrigera- 
'ting Load^ 



Kw. 

hoursi 



B.h.p. 

hours^ 



Total 
Load 

B.h.p. B.h.p. 

hours hours 



Coal 

Charged 
Pounds 



Sunday, July 25, 1908. . . 

Monday, July 26 

Tuesday, July 27 
Wednesday, July 28 

Thursday, July 29 

Friday, July 30 

Saturday, July 31 

Sunday, August 22, 1P08 

Monday, August 23 

Tuesday, August 24 

Wednesday, August 25 

Thursday, August 26 

Friday, August 27 

Saturday, August 28 

Totals 



332 

400 
404 
390 
410 
415 
403 



3.^8 
386 
393 
392 
397 
391 
393 



5434 



556 
600 
606 
585 
615 
622 
605 



2010 
2030 
2020 
2020 
2030 
2040 
2020 



549 
679 
589 
588 
596 
586 
590 



2030 
2020 
2020 
2010 
2010 
2020 
2020 



2566 
2630 
2626 
2605 
2645 
2662 
2625 



3600 
3900 
3540 
3180 
4020 
4320 
4080 



2579 
2599 
2609 
2598 
2606 
2606 
2610 



3660 
3420 
3600 
3540 
3840 
3540 
3720 



8266, 



28300 



36566 



51960 



1 Recorded by watt-hour meters. 

2 Deduced from kilowatt-hours by assuming 80 per cent efficiency for the generator during 
light-load periods and 90 per cent for the remaining time. 



OPERATION OF PKODUCER GAS-POWER PLANT 219 

DISCUSSION 

J. A. Holmes. The success of the small producer plant using 
anthracite coal is practically assured. Not long since (1905), in 
visiting a number of these plants in Cologne, Germany, I found a 
newspaper press that had been operated entirely for more than a 
year by a small gas-producer plant burning small-sized anthracite 
coal; one of the larger hotels there had been using such a plant for a 
longer period with entire satisfaction to supply all its electric light 
and power; in a large commercial house, electric lamps, elevators 
and all other machinery connected with the establishment were 
operated by one of these plants. In each of these cases the producer, 
engine-driven generators and other equipment in the power room, 
were all operated by one man, and the plant was regarded as a success 
in efficiency and economy of labor and fuel. In the United States, 
also, many producer plants have been using anthracite coal for some 
years. In our own investigations at the Government testing station, 
anthracite coal has been regarded as a fuel so simple and so easily 
regulated that we have done little work on it, turning our attention 
mainly to the bituminous coal producer problems. 

2 In regard to producer work with bituminous coal, we have in- 
vestigated fuels rather than different types of producers. Using every 
imaginable grade of bituminous coal and lignite in making short-time 
tests, we have encountered many difficulties due to a lack of famil- 
iarity with the special manipulations required by certain fuels. These 
difficulties, one being to secure a uniform quality of producer gas, 
would not be met in using the same fuel year after year. In early 
work, with the Taylor producer, we could get gas of absolutely uni- 
form character not more than an hour at a time, and the variation in 
24 hours was at times from 125 B.t.u. to more than 200 B.t.u. per 
cu. ft. of gas, these variations being largely due, no doubt, to inex- 
perience in the handling of any special fuel. During the past three 
years, however, with more experience, the regularity and efficiency 
of this gas have been greatly increased. 

3 Another difficulty, and one not entirely separable as yet, is the 
slagging or clinkering of the ash in the producer. The ash in certain 
coals slags more readily than in others; and different ashes slag more 
readily at different temperatures. One of the greatest needs in pro- 
ducer development at the present time is that of a regular mechanical 
feed of coal and removal of the ashes which now accumulate in some 
producers, to be cleaned out after the producer has cooled down. We 



220 DISCUSSION 

have sometimes found the slag from certain coals, burned at high 
temperatures, accumulating irregularly on the brick walls lining the 
producer, at the rate of 6 in. to 10 in. during a week's run. If me- 
chanical arrangements can be devised, by which the ash may be re- 
moved from the base of the producer as regularly as from the base of 
a boiler, then the use of a double producer will be largely avoided. 
Decided progress is being made in overcoming this difficulty. 

4 Still another line of progress is in the reduction of weight and 
bulk of the producer making its use possible instead of that of steam 
boilers for propelling ships. Mr. Straub's paper indicates what is 
being accomplished along this line. Already the anthracite producer 
and gas engine have been reduced in size and weight to less than 
those of the steam boiler and reciprocating engine; and the outlook 
is hopeful for the producer burning bituminous coal. 

John H. Norris. I have been connected with the manufacture 
of gas engines for a number of years, and the principal trouble we 
have had in the operation of gas engines of any size is to overcome the 
notion that a gas engine needs no care. At the present time, however, 
gas engines are running successfully because in most installations they 
receive proper attention. I am glad to see put on record the state- 
ment that a gas engine installation needs as close attention as a steam 
engine installation. 

William A. Bole. The Westinghouse Machine Company has 
been working on the gas-producer problem as well as on the gas- 
engine problem for some time, and now believes itself ready to offer 
gas producers that will be as practical and as easily manipulated 
and capable of as long-continued runs as any boiler plant. A pro- 
ducer plant of 175-b.h.p. capacity has been in operation at our works 
for practically a year, without pulling down the fires. During that 
period all sorts of runs have been made, continuous runs at full capac- 
ity for ten days or two weeks, and the more ordinary runs in which 
producer and engine are shut down at night; and the producers have 
burned not only the comparatively good coals of the Pittsburg dis- 
trict and the better coals of the Pocahontas region, but several of 
the Western and Southwestern lignites and even peats from New 
England. The latter have not been so successfully burned, but the 
success in burning Colorado lignites has been very decided. This 
producer was shut down and cleaned out, simply by shoveling the 
ashes out of the water seal, and observations of the condition of the 



OPERATION OF PRODUCER-GAS POWER PLANT 221 

interior walls showed that it might just as well have been operated 
continuously for five years instead of one, or as many years as the 
firebrick lining would last. The requirements for continuous per- 
formance seem to have been admirably met in this design. 

2 This producer is designed for the burning of bituminous coal 
alone, and resembles a small producer inverted and placed on top of a 
large one, making a double-zone producer especially adapted for the 
gasification of bituminous coal without passing tar of any descrip- 
tion out of the producer shaft. Apparently the only solid material 
emitted from the gas is a small amount of lamp-black which is success- 
fully removed by the use of a static or stationary scrubber, and the 
cleanliness of the gas is proved by the fact that practically all the 
gas was converted into brake horsepower by being employed in the 
actual operation of a gas engine, without troublesome deposit of any 
kind upon the ports or other parts of the gas engine. 

3 Whether such a producer would be available for marine pur- 
poses I do not know ; the only question seems to be whether the motion 
of the ship would interfere seriously with the descent of the fuel from 
top to bottom. The producer has been subjected to practically every 
test, and we believe we are about ready to offer it for both large and 
small plants. 

The author desired to present no closure. — Editor. 



No. 1240 

OFFSETTING CYLINDERS IN SINGLE-ACTING 

ENGINES 

By Prof. Thurston M. Phetteplacb, Providence, R. I. 
Member of the Society 

A great deal has been said recently about the offsetting of cylinders 
in single-acting engines and many claims of superiority are made by 
those who employ this form of construction. 

2 About twenty-five manufacturing establishments in the United 
States are building engines in which the cylinders are offset, chiefly 
those of the automobile type, and one company is formed for the pur- 
pose of making engines in which the offset is equal to the crank radius 
and the connecting rod length is about 3f times the crank radius. 

3 Among the claims made by manufacturers for offset engines 
are: greater power, less side-pressure of the piston on the walls of 
the cylinder, better turning effort, less vibration, smoother running 
qualities, and when one cam shaft is used, a more convenient mechan- 
ical arrangement. 

4 On account of the importance of this subject and the lack of 
information concerning it, a complete discussion is desirable and is 
here presented. 

5 The cylinder of an engine is said to be offset when its center- 
Une is not in a plane through the center of the crank shaft. The 
practice is not new and is applied to both steam and gas engines hav- 
ing one or any number of cylinders. 

6 In the diagram, Fig. 1, AB represents the stroke, OE the crank 
radius, DE the connecting rod, 6 the crank angle, and OC the offset. 
It should be noticed that is the angle the crank makes with a line 
through the center of the crank shaft parallel to the center-line of the 

The full development of the mathematical formulae of this paper, with some 
other related matter, is given in an unpublished Appendix, which is on file in 
the Library of the Society. 

Presented at the Spring Meeting, Washington, May 1909, of The American 
Society of Mechanical Engineers. 



224 



OFFSETTING CYLINDERS IN SINGLE-ACTING ENGINES 



cylinder, and not the actual angle passed over from the inner dead 
point. The length of the stroke is 



AB = R (1/ (a + 1 y- ¥ - v. (a - 1)^ - P) 

which is greater than 2R. 

R = crank radius. 

a = L/R. 

L = connecting rod length. 

k = offset divided by R. 




Fig. 1 Diagram of Crank and Connecting-Rod Train 

Thus for 3-in. crank radius the strokes would be as shown in Table 1 . 
The distance to the end of the stroke farther awaj^ from the center 
of the crank shaft is shortened, thus OM is less than DE + EO or 
L -\- R which affects the height of the engine. 



TABLE l^^LENGTHS OF STROKES FOR DIFFERENT OFFSETS, 3-IN. CRANK 

ENGINE 



Ratio L/R 



Offset 
R 



Stroke 



Increase 
Per cent 



Any 


eero 


6.00000 


0.00 


3 


0.10 


6.00375 


0.06 


3 


1.00 


6.42279 


7.04 


4 


1.00 


6.21180 


3.50 


5 


1.00 


6.12930 


2.15 


6 


1.00 


6.08730 


1.01 



7 The dead points are not opposite each other, -so that the crank 
angle swept over while the piston makes the out-stroke is less than 
that for the in-stroke, causing a quick return motion and an average 
velocity for the in-stroke or compression and exhaust strokes greater 
than for the out-stroke, or explosion and suction strokes. 



OFFSETTING CYLINDERS IN SINGLE-ACTING ENGINES 



225 



8 An expression for the piston position in terms of the crank angle 
6 is developed in the usual way and is 



X/R = V {a-\-iy -k^ - coBd - V a" - {k-sin Oy 

in which X = the piston displacement from the end of the stroke 
farther from the crank shaft. 

9 The force of inertia due to the reciprocating parts is equal to 
the weight multiplied by the acceleration divided by 32.2. The value 
for the acceleration is found by differentiating the expression for the 
piston displacement twice with respect to the time. This is done by 
expanding the radical V a^ — (k — sin 6y by the binomial theorem, 
into a convergent series and then dropping all terms containing a 
with a negative exponent of 3 or larger in order to get an expression 
that can be easily differentiated. This gives 



[a^ - {k - sin)2 ] 



i = n - 



i a-* k^ + a-» k sin 6 - i a"* sin^ 



This approximate expression for the radical differs from the radical 
for different values of k, a and 6, as shown in Table 2. 



TABLE 2 DIFFERENCE BETWEEN EXACT AND APPROXIMATE EXPRESSIONS 



l/o=!- (fc - 



6)2 



— ia"i sin' 6 



Difference 



1 


6 


90 


6.000000 


6.000000 


zero 


1 


6 





5.916079 


5.916666 


+ .000587 


1 


6 


45 


5.992762 


5.992867 


+ .000105 


.5 


6 


90 


5.979130 


5.979166 


+ .000036 


.5 


6 





5.979130 


5.979166 


+ .000036 


.5 


6 


45 


5.996428 


5.996429 


+ .000001 


.5 


3 


90 


2.958039 


2.958333 


+ .000294 


.5 


3 





2.958039 


2.958333 


+ .000294 


.5 


3 


45 


2.992849 


2.992859 


+ .000010 


.6 


4i 


90 


4.472136 


4.472222 


+ .000086 


.6 


4i 





4.472136 


4.472222 


+ .000086 


.5 


44 


45 


4.495236 


4.495239 


+ .000003 



10 The greatest difference has no significant figure until the fourth 

decimal place is reached and this is when A; = 1, which is an unusual 
value. Hence it is readily seen that the error introduced by this 
approximate form is slight. 

11 Substituting this value of the radical in the expression for the 
piston displacement and differentiating twice with respect to the time 
gives 

F/A = 0.00034 W/A N^R (cT^k sin d -i- cos 6 + oT^ cos 2d) 



226 OFFSETTING CYLINDERS IN SINGLE-ACTING ENGINES 

which is the expression for the inertia force per square inch of piston 
head area when there is an offset. 

A = area of piston head. 

W = weight of the reciprocating parts. 

iV = revokitions per minute. 

R = crank radius in feet. 

This differs from the similar expression when there is no offset by 
the addition of the term a"% sin 6, so that tables for inertia factors 
for no offset may be used by adding the value of this term. 

12 The expression for the tangential pressure or the turning force 
for any offset is 

T = Pl sin^+ cos^ sin^-A: 



1 



i a-^ ¥ + a"' kQmd-\ oT^ sin^ 

i n which P is the pressure on the piston pin in the direction of the cen- 
ter of the cylinder. This is a long expression to solve and a graphical 
solution may be followed if preferred. The work of solving the ex- 
pressions for inertia force and tangential pressure may be somewhat 
lessened by tabulating the quantity a~' k sin 6 which appears in these 
expressions. 

13 The derivation of the preceding formulae and tables is shown 
in the appendix. 

SIDE PRESSURE OF PISTON ON CYLINDER WALLS 

14 A reduction of the side pressure of the piston on the cylinder 
walls is one of the advantages claimed for offsetting. 

15 There are two ways in which the side pressure may affect the 
single-acting engine : (a) The maximum value of the side pressure de- 
termines the length of piston to keep the maximum pressure per square 
inch of projected area below a value which is assumed as not too great 
to destroy the oil film between the rubbing surfaces; (b) The average 
value of the side pressure produces the friction between the sliding 
surfaces causing a mechanical loss and some wear of the parts. The 
loss in mechanical efficiency is more important than the wear, especi- 
ally in the small high-speed automobile engines. 

16 The average side pressures may be found by adding all of the 
areas between the axis and the curve of side pressures and dividing 
by the total length, or the areas themselves may be taken for com- 
parison, as they represent the work done on the side of the cylinder 



OFFSETTING CYLINDERS IN SINGLE-ACTING ENGINES 



227 



by the piston, which is lost work and should be kept as low as possible. 
Curves of side pressures of the piston on the cylinder walls were con- 
structed, it being necessary (a) to assume a gas card, (6) to assume 
engine dimensions, (c) to calculate inertia forces and plot curves, 




13 3 4 5 6. 

Fig. 2 Offsetting Ctunders in Singlb-Actinq Engines 
Gasolene Card Compression, 70 lb.; Maximum Pressure, 259 lb.; Pressure 
ratio, 3.77 

(d) to combine inertia forces with gas pressures, obtaining the force at 
the piston pin, and then (e) to determine the side pressure component 
perpendicular to the center-line of the cylinder for the changing angu- 
larity of the connecting rod. The gasolene card chosen is shown in 
Fig. 2. As it seemed desirable to investigate two similar cases, one 

TABLE 3 CASES INVESTIGATED 



High 



Specifications Slow 

R.p.m 450 1500 

W/A lib. 0.70 1b. 

R 6 in. = 0.5 ft. 2i in. = 0.208 ft. 

O.OOOSiW/A mR 34.4 111.38 



228 



OFFSETTING CYLINDERS IN SINGLE-ACTING ENGINES 



for high speed and the other for slow speed, the dimensions given in 
Table 3 were chosen. 



TABLE 4 PISTON POSITION FACTORS 
Calculated by the Formula 



X/R = V(a + 1)2 - A;2 - cos 6* - a + i a "1 A;2 - a -ifc sin0 + i a "isin^© 

From Beginning of Stroke Towards the Crank Shaft. Multiply by Crank Radius 
TO Find Piston Position. (Note: Crank Radius is Not One-Half of the Stroke) 



Crank Angle 


L-^R = Z 


L -s-fi = 4i 












Offset = 0.30 iJ 


Offset = 0.50 5 


Offset = 0.30 2? 


Offset = 0.50 B 


4°18' 











7°11' 











307' 











5°13' 











15 


0.023 


0.012 


0.026 


0.018 


30 


0.129 


0.103 


0.130 


0.111 


45 


0.309 


0.269 


0.303 


0.275 


60 


0.525 


0.491 


0.527 


0.492 


75 


0.804 


0.746 


0.782 


0.742 


go 


1.070 


1.010 


1.046 


1.005 


105 


1.321 


1.263 


1.299 


1.26 


120 


1.525 


1.491 


1.527 


1.49 


135 


1.723 


1.683 


1.717 


1.69 


150 


1.861 


1.835 


1.862 


1.84 


165 


1.955 


1.944 


1.958 


1.95 


180 


2.0037 


2.0102 


2.0018 


2.0049 


188°38' 


2.0114 








194°29' 




2.0321 






184°55' 






2.0047 




188''13' 








2.0131 


195 


2.0065 


2 .0303 


1.992 


2.007 


210 


1.961 


2.001 


1.929 


1.954 


225 


1.865 


1.918 


1.810 


1.846 


240 


1.699 


1.779 


1.643 


1.684 


255 


1.515 


1.585 


1.429 


1.475 


270 


1.270 


1.343 


1.179 


1.227 


285 


0.997 


1.068 


0.911 


0.957 


300 


0.699 


0.779 


0.643 


0.684 


315 


0.471 


0.504 


0.407 


0.432 


330 


0.229 


0.269 


0.197 


0.222 


345 


0.075 


0.098 


0.061 


0.075 


360 


0.0037 


0.0102 


0.0018 


0.0049 



17 The piston position factors and inertia factors are given in 
Tables 4 and 5, and Fig. 3 to Fig. 8 give the curves of inertia forces. 



OFFSETTING CYLINDERS IN SINGLE-ACTING ENGINES 



229 



The full lines represent the slow-speed and the mixed lines the high- 
speed cases. Inertia curves and side-pressure curves were plotted 
for ratios of l^jR of A.\ and 3, and for offsets of zero, 0.30 i2, and 0.50 i2 
for both high and slow speeds, making twelve cases in all. When 
there is an offset the inertia curve must be plotted for 360 deg. instead 
of 180, since for the return stroke it is not the reverse of that for the 
forward stroke, as is the case when there is no offset. 

TABLE 5 INERTIA FACTORS 
Calculated by Formula (o"*&siii0 + cos^ + o ■* cos 20) 



Angle 


L/R 


= 3 


L/R = 3i 


L/«=4 




L/R = 4i 


K = 0.30 


K = 0.50 


K = 0.20 


if = 0.30 


K = 0.30 


K = 0.40 


K = 0.50 


15 


1.280 


1.297 


1.229 


1.200 


1.175 


1.181 


1.187 


30 


1.083 


1.116 


1.037 


1.028 


1.010 


1.021 


1.033 


45 


.778 


.825 


.747 


.760 


.754 


.769 


.785 


60 


.419 


.477 


.406 


.440 


.447 


.465 


.485 


75 


.067 


.131 


.066 


.104 


.131 


.152 


.174 


90 


-.233 


-.166 


-.229 


-.175 


-.156 


-.134 -.111 


105 


-.450 


-.386 


-.451 


-.404 


-.385 


-.364 -.344 


120 


-.580 


-.523 


-.594 


-.560 


-.553 


-.535 -.515 


135 


-.636 


-.589 


-.666 


-.654 


-.660 


-.645 ! -.629 


150 


-.650 


-.617 


-.695 


-.704 


- .722 


-.711 i -.699 


165 


-.653 


-.635 


-.703 


-.731 


-.757 


-.751 -.745 


180 


-.667 


-.667 


-.714 


-.750 


-.778 


-.778 -.777 


195 


-.703 


-.721 


-.733 


-.769 


-.791 


-.797 -.803 


210 


-.750 


-.783 


-.752 


-.778 


-.788 


-.799 1 -.810 


225 


-.778 


-.825 


-.747 


-.760 


-.754 


-.769 1 -.785 


240 


-.753 


-.811 


-.692 


-.690 


-.669 


-.687 


-.707 


255 


-.644 


-.708 


-.561 


-.548 


-.513 


-.534 


-.556 


270 


-.433 


-.500 


-.343 


-.325 


-.288 


-.310 


-.333 


285 


-.127 


-.191 


-.044 


-.040 


.003 


.018 


-.040 


300 


.246 


.189 


.308 


.310 


.331 


.313 


.293 


315 


.636 


.589 


.667 


.654 


.660 


.545 .629 


330 


.983 


.950 


.981 


.954 


.944 


.933 .917 


345 


1.228 


1.211 


1.199 


1.164 


1.141 


1.135 1.129 


360 


1.333 


1.333 


1.286 


1.250 


1.222 


1.222 1.222 



18 Comparing Fig. 3 with Fig, 6 a slight hump is noticed at the 
right-hand side in the former but not in the latter. This is probably- 
due to error in the formula, for the small value of L/R since the force 
could not be higher near the end of the stroke than at the end. 

19 The general effect of offsetting on the inertia curve is shown 
in Fig. 9, where the curves for L/R = 3, offsets = zero and 0.50 R, 
are compared, the curve for no-offset being in full lines. 



230 



OFFSETTING CYLINDERS IN SINGLE-ACTING ENGINES 




SLOW SPEED 



/ 



/ 

/ 



/ 



/ 



/ 



HIGH SPEED 



/ 




Fig. 3 Offset = Zero- 



away FROM 
CRANK SHAFT 



Fig. 4 Offset = 0.30 R. 



Curves of Inertia Forces on Piston Position Base 

W 
Slow Speed: r.p.m. = 450; R = 6; ^ = 1 lb.; L -^ R = 3. 

W 
High Speed: r.p.m. = 1500; R = 2^ in.; ~ = 0.7 lb.; L ^ R = 3. 

Full Lines, Slow Speed; Mixed Lines, High Speed. 




Fig. 5 L -^ R = 3. Offset = 0.50 R. Fig. 6 L ^ R = 4i. Offset = zero. 
Cdkvbjs of Inertia Forcbs on Piston Position Bask 



OFFSETTING CYLINDERS IN SINGLE-ACTING ENGINES 231 



/ 



/: 



^.^^ / TOWARDS 


^ / 

/ 
/ 


CRANK SHAFT / 

/ 


V 


/ . 




l^^^ 


___--' 


7 


/ 


AWAY FROM 


CRANK SHAFT 


_,./ 







7 


TOWARDS 


" / 




:RANK SHAFT / 


/ 




/ 

/ 


/ 




/ 




/ 


/ 

AWAY FROM 




CRANK SHAFT 


\y 







Fig. 7 L ^ R= 4^. Offset= 0.30 R. Fig. 8 L ^ R = 4^. Offset = 0.50 R 
Curves of Inertia Forces on Piston Position Base 




Fig. 9 Inertia Force Curves Showing Effect of Offsetting 
L-5-R = 3, High Speed. Full Lines, Offset = Zero; Mixed Lines, Offset - 0.60 R 



232 



OFFSETTING CYLINDERS IN SINGLE-ACTING ENGINES 



20 The curve for an offset is a little flatter near the end of the out 
stroke and the hump is increased near the beginning of the return 
stroke, which is probably due to inaccuracy in the formula. 

21 The curves of side-pressures are shown in Fig. 10 to 15. The 
maximum side-pressure, its cause (whether combined gas and inertia 
pressure or inertia force alone) and its location are given in Table 6. 



TABLE 6 MAXIMUM SIDE PRESSURES, CAUSE AND LOCATION 



MAX. SIDE-PREBSUBE 



Ratio L/R Offset 



Slow 



High 



Slow 



High 



Slow 



High 



4i 





25 


26 


Gas 


Gas 




1 


4i 


0.30 iJ 


17 


24 


Gas 


Inertia 




2 


4J 


0.50 B 


12 


28 


Gas 


Inertia 




2 


3 





35i 


45 


Gas 


Gas 




1 


3 


0.30 B 


23 


39 


Gas 


Inertia 




2 


3 


0.50 « 


20 


51 


Inertia 


Inertia 


2 


2 



22 From this for L/R = 4J, slow-speed, maximum pressure is 
lowest with 0.50 R offset, and if the offset were further increased the 
maximum side-pressure would probably not be reduced as the values 
at the beginning of the second and fourth strokes would increase, and 
now they are already 11 so that any increase would soon cause an 
increase in the maximum value instead of a decrease. In the case 
of L/R = 3 the lowest maximum value occurs when the offset is 
0.50 R, but in this case it is possible that the offset is already a trifle 
large, as the maximum value occurs at the beginning of the second 
stroke, although it is not much greater than that in the first stroke, 
being 20 in the former case and 18 in the latter. Hence for the slow 
speed the best offset would seem to be about 0.50 R as far as the max- 
imum value of side-pressure is concerned. 

23 In the ca.se of L/R = 4^, high speed, the maximum side-pres- 
sure due to inertia force at the beginning of the second stroke seems 
to increase with the amount of offset, while the maximum value due 
to the gas pressure in the first stroke seems to decrease with the in- 
crease in offset. These values are shown in Table 7. L/R = 4^. 

TABLE 7 COMPARISON OF SIDE PRESSURES FOR L/R = 4i 



Offset 


Zero 


0.30 fi 


0.50B 


Side pressures due to Gas pressure, 1st stroke. . . 

Inertia, 2d stroke 


26 17 13 
15 j 24 ! 28 











OFFSETTING CYUNDERS IN SINGLE-ACTING ENGINES 233 




^ 



/ 



Fig. 10 Curve of Side Pressukes on Piston Position Base 
C -T- R = 3, No Offset. Full Lines, Slow Speed; Mixed Lines, High Speed 





Fig. 11 Curve of Side Pressures on Piston Position Base 
L - R = 3, Offset = 0.30 R. FuU Lines, Slow Speed; Mixed Lines, High Speed 





Fig. 12 Curve of Side Pressures on Piston Position Base 
L -f- R = 3, Offset = 0.50 R. Full Lines, Slow Speed; Mixed Lines, High Speed 



234 



OFFSETTING CYLINDERS IN SINGLE-ACTING ENGINES 




Fig. 13 Curve of Side Pressures on Piston Position Base 
L ^ R = 4J, Offset = Zero. Full Lines, Slow Speed; Mixed Lines, High Speed 




Fig. 14 Curve of Side Pressures on Piston Position Base 
L ^R = 4|, Offset =- 0.30 R. Full Lines, Slow Speed; Mixed Lines, High Speed 




7f^ 



/^ 



Fig. 15 Curve of Side Pressures on Piston Position'/Base 
L -^ R = 4i, Offset = 0.50 R. Full Lines, Slow Speed; Mixed Lines, High Speed 



OFFSETTING CYLINDERS IN SINGLE-ACTING ENGINES 



235 



24 Plotting curves of these values, the most favorable offset as 
far as maximum side-pressure is concerned is 0.16 R when L/R = A\, 
This curve is shown in Fig. 16. 

25 See Table 8 for values L/R = 3. This would place the best 
offset for L/R = 3, as far as maximum side-pressure is concerned, 



TABLE 8 


COMPARISON OF SIDE PRESSURES FOR 


l/r = z 






Offset 


i 
Zero 


0.30K 1 


0.50 K 


Side pressures due to 


Gas pressure, 1st stroke . . . 


45 
25 


30 
39 

1 


20 
51 









as 0.20 R, which would seem to indicate that a greater offset would 
be desirable as the ratio L/R decreased. It remains to determine if 
possible the best offset as far as the work done in side-pressm-e is con- 
cerned. 

26 The work done is proportional to the areas included between 
the axis and the curve of side-pressures. It seems to make no dif- 
ference whether a larger amount of work is done on one side than 
on the other, or in other words there seems to be no advantage in 
having the work done on each side the same, unless at some time it 
might be desired to rebore the cylinder, in which case wear occurring 
all on one side might have left the walls too thin or might necessi- 
tate the removal of much more metal. 





Fig. 16 Curve Showing Variation of Side Pressure with Offset 

27 Table 9 shows the results of measuring the areas, namely the 
ratio of work done in one case to that in each of the other cases, and 
also the actual average side-pressures. For the slow-speed case there 
seems to be a decrease of work done on one side, and an increase of 
work done on the other side, resulting in a decrease in the total work 
done with the offset, which would indicate that the greater the off- 
set, the less the loss in work. The average side-pressure also decreases 
with the offset, although it is less with no-offset when L/R = 4^ 
than with 0.50 R offset when L/R = 3. 



236 



OFFSETTING CYLINDERS IN SINGLE-ACTING ENGINES 



28 In the case of the high speed the areas on one side decrease with 
the offset while those on the other side increase, but the totals for 
L/R = 4^ decrease and then increase, while for L/R = 3 they con- 



TABLE 9 



Ratio 
L/R 



Offset 



BLOW BPEBD 



+ Area 



— Area 



Total 
Area 



Average 
I Side 
Pressure 



HIGH SPEED 



H-Area 



-Area 



Total 
Area 



I Average 

Side 
Pressure 



4i 

4i 

44 

3 

3 

3 





0.30 fi 
0.50 iJ 


0.30R 
0.50R 



1.61 
1.09 
0.74 
2.59 
1.55 
1.11 



2.14 
1.99 
1.67 
3.51 
2.70 
2.50 



6.6 


2.33 


6.2 


1.78 


5.2 


1.62 


11. 


4.11 


8.4 


3.01 


7.8 


2.47 



-1.11 

-1.58 
-1.90 
-1.45 
-2.76 
-3.40 



3.44 
3.36 
3.52 
5.56 
5.77 
5.87 



10.7 

10.5 

11. 

17.3 

18.0 

18.3 



tinue to increase, and of course the same is true for the mean side- 
pressure. 

29 This would seem to indicate that there is little if anything to 
be gained by an offset in regard to work done by the piston on the 
walls of the cyhnder when the inertia force is very high, since what 
is gained on one side is more than made up in loss on the other side. 

30 If it is of sufficient importance to have the work done on each 
side of the cylinder the same, we may plot curves of the work done 
on each side and note where they intersect, as in Fig. 17 . In the case 
of the slow speed we would have the work done on each side equal 
when the offset was about 0.40 R and in the high speed this point 
would be about 0.36 R. 

31 Thermal Cycle. Offsetting increases the length of stroke, 
which gives increased expansion to the gas, and increases the piston 
velocity on the in-stroke, giving greater inertia to the gas on the 
exhaust and reducing the amount of leakage by the piston on the 
compression stroke. This refers to the 4-cycle gas engine. 

32 Lubrication. The curves of side-pressure show the manner in 
which the side-pressure changes sides, which is a good thing for 
lubrication. This changing sides would be about the same for off- 
set or no-offset except in the case when the offset is equal to the crank 
rachus. Here the pressure is almost continually on one side of the 
cylinder so that oil would with difficulty be introduced between the 
surfaces. Other tilings being even, except for this extreme case, 
the reduction in amount of side-pressure should make lubrication 
more satisfactory. 



OFFSETTING CYLINDERS IN SINGLE-ACTING ENGINES 



237 



33 Vibration and Balance. Revolving masses and reciprocating 
masses may cause vibration in gas engines. Offsetting the cylinders 
would not affect the revolving masses at all but does change the curves 
of inertia forces, as already shown in Fig. 3 to 8. These inertia- 
force diagrams are now combined in different ways according to dif- 
ferent arrangements of cylinders, and are compared with similar 
curves when there is no offset. 



SLOW SPEED 




HIGH SPEED 




Fig. 17 



.:30 .38 

Curves Showing Offset when Work is same on Each Side 
OF Cylinder 



34 The following discussion applies only to the 4-cycIe type of 
gas engine, whose arrangements are: 

a Single cylinder. 

b Two-cylinder vertical. 

c Two-cylinder opposed. 

d Three-cylinder vertical. 

e Four-cylinder vertical. 

/ Four-cylinder double-opposod. 

g Six-cylinder vertical. 



238 



OFFSETTING CYLINDERS IN SINGLE-ACTING ENGINES 



35 For this comparison the high speed case, when L/R = 4-^, was 
chosen, the offset being equal to zero and one-half the crank radius, 

36 Fig. ]8 shows the inertia curves for a single-cylinder engine. 
These curves must be shown for 360 deg. of crank angle, for they dif- 
fer on the return and forward strokes. The curves for no-offset are 
shown in full lines and for 0.50 R offset in dotted lines. The difference 
between the two curves is apparent. 




Fig. 18 Curves of Inertia Forces 
ON Piston Position Base. Single- 
Cylinder Engine 

L -- R = 4J. Full Line, Offset = 
Zero; Dotted Line, Offset = 0.50 R. 
High Speed Case 



Fig. 19 Curves of Free Unbal- 
anced Inertia Forces. Two-Cyl- 
inder Vertical Engine 

L -f- R = 4^; Full Line, Offset = Zero; 
Dotted Line, Offset =0.50 R. High 
Speed Case 



37 Fig. 19 shows the curve of free inertia forces for a two-cylinder 
vertical arrangement. These curves are not so very different; the 
one for an offset being nearly the same as the other but moved along 
a little instead of being symmetrical with a center line perpendicular 
to the axis. The maximum values of the forces are about the same. 
The vibrations when there is an offset would have unequal periods 
but about the same amplitudes. For the four-cylinder vertical case 
the ordinates of these curves could be doubled and the same general 
difference would exist. 

38 In the case of the two-cylinder opposed motor with cranks 
at 180 deg., the inertia forces^wouk^be balanced whether the cylin- 
ders were offset or not, but in the case of an offset a new couple in 



OFFSETTING CYLINDERS IN SINGLE-ACTING ENGINES 239 

a plane perpendicular to the axis would be introduced due wholly to 
the offsetting, which cannot be balanced. The couple in an axial 
plane due to the cylinders being not in line would be the same, offset 
or not, but with an offset there would be added another couple in this 
plane due to the offset, which would not be balanced. 

39 In the case of a four-cylinder double-opposed motor the forces 
would be balanced and also the coujDles in an axial plane, but the 
couples in the plane perpendicular to the axis would be doubled while 
those in the axial plane due to the offset would be balanced. 

40 The case of a three-cylinder vertical arrangement can be dis- 
cussed by considering the formula for the inertia forces, 

F/A = 0.00034 W/A N^'R {a-'k sin 6 -\- cosd + a-' cos 2 6) 

Let the cranks be at 120 deg.; then the crank angles will he 6,6 + 120, 
and 6 + 240. Substituting these values in the formula, the part in 
brackets reduces to zero, showing that the inertia-forces are balanced. 
However, the moments resulting from these forces are not balanced. 
By placing two three-cylinder vertical engines together so that the 
two middle cranks are in the same plane the six-cyhnder engine is 
obtained, in which the inertia forces and couples are both balanced. 

41 From this discussion it follows that offsetting the cylinders 
has no effect on the vibration due to the reciprocating parts, except 
in the case of the 2-cylinder opposed and 4-cylinder double-opposed 
arrangements of cyhnders. In these cases the offsetting increases 
the unbalanced inertia-force couples by adding new ones. 

42 Vibration may be felt from the irregularity of the turning- 
effort curves, which for four different cases are shown in Fig. 22. 
There is such a slight difference here that it can be neglected, espe- 
cially since the turning-effort curve depends so directly on the shape 
of the gas card, which may vary considerably. The conclusion in 
regard to vibration would be that offsetting does not affect the vibra- 
tion appreciably except in the case of a two-cylinder opposed or a 
four-cylinder double-opposed motor. 

GENERAL CONCLUSIONS 

43 The following are perfect^ general conclusions, to be followed 
by a more definite comparison of actual engines: 

a The length of stroke for a given crank radius increases as 
the offset increases. 



240 OFFSETTING CYLINDERS IN SINGLE-ACTING ENGINES 

h The length of stroke for a given crank radius for any off- 
set decreases as the ratio of L/R increases. 

c The increase in length of stroke causes an increase in aver- 
age piston speed. 

d Offsetting the cyUnders makes the crank and connecting 
rod train a quick return mechanism. 

e When the cylinders are offset the crank passes over an 
angle greater than 180 deg. during the out-stroke of the 
piston, and less than 180 deg. during the in-stroke. 

/ The average velocity of the piston is greater on the exhaust 
and compression than on the explosion and suction strokes. 

g Offsetting the cylinders reduces the angularity of the con- 
necting rod on the out-stroke and increases it on the in- 
stroke. 

h When there is an offset, the side-pressure of the piston on 
the cylinder walls does not change sides at the end of the 
stroke, but just after the beginning and just before the 
end of the out-stroke. 
{ The place where this change of side-pressure occurs ap- 
proaches the middle of the stroke as the amount of offset 
approaches the crank radius. 

/ With no offset, liigh inertia forces do not greatly increase 
the maximum side-pressure during the explosion stroke, 
but do increase it considerably during all of the other 
strokes, and this effect is slightly greater as the ratio of 
L/R decreases. 

k With no offset the work done increases with the inertia- 
force and as the ratio of L/R decreases. 

I For low inertia forces, as far as the maximum value of 
side-pressure is concerned the best offset is one-half the 
crank radius. 

m Considering the maximum value of the side-pressure only, 
the most favorable value for the offset decreases as the 
inertia-forces increase, for any ratio of L/R, but does not 
decrease as rapidly, for smaller values of the ratio L/R. 

n For low inertia forces, the work done by the piston on the 
cylinder walls decreases as the offset increases, but of 
course is greater for smaller values of L/R. 

For very high inertia forces, the work done decreases 
slightly with the offset up to 0.40 of the crank radius for 
value of L/R = 4^. 



OFFSETTING CYLINDERS IN SINGLE-ACTING ENGINES 



241 



p For very high inertia forces, and small values of L/R, 
there is no advantage in an offset, as far as the work done 
by the piston on the cylinder walls is concerned. 

q The thermal cycle is slightly benefited by offsetting and 
the benefit increases with the amount of offset. 

r Lubrication should be sUghtly improved by offsetting the 
cylinders. 

s Vibration due to the free inertia forces is no different ex- 
cept in the case of a two-cylinder opposed or four-cylinder 
double-opposed motor. 

44 Table 10 gives data of gas engines that have been constructed 
and put in operation. The average crank radius is about 2| in., the 

TABLE 10 DATA OF GAS ENGINES HAVING CYLINDERS OFFSET 







Ii Length 




\ 


1 






Ratio 


No. 


R Crank 


of Con- 


Ratio 


Offset 


Offset 


Length of 


Diam. 


Piston 


Radius 


necting 


L/R 


Amount 


Per cent 


Pbton 


of Bore 


Length to 






Rod 












Diameter 


1 


7 


231 


3.37 


7 


100 


14 


9.47 


1.47 


2 


2A 


8it 


3.48 


1 


24. 


6 


5i 


1.09 


3 


2i 


8i 


3.77 


i 


16.6 


6 


4 


1.5 


4 


2i 


94 


3.8 


1 


15 


51 


5 


1.125 


5 


■ 2i 


10 


4.0 


i 


15 


6i 


4i 


1.37 


6 


3 


121 


4.08 


A 


18.76 


5J 






7 


2f 


9f 


4.1 


i 


21 


5J 


5 


1.075 


8 


2i 


lOi 


4.2 


i 


35 


5A 


41 


1.098 


9 


2i 


12 


4.36 


i* 


34 


6i 


5i 


1.28 


10 


2i 


lOi 


4.66 


i 


38 


5i 


4i 


1.29 


11 


2i 


12 


4.8 




40 


6 


4J 


1.26 


12 


2i 


12 


4.8 




40 


6i 


5 


1.25 


13 


2i 


12 


4.8 




40 


6 


4J 


1.26 


14 


2i 






li 


50 








15 


21 


i 




U 


40 




j .... 


.... 


16 


3 






1 


33 ^ 




.... 




17 


2* 


i^J'l„ 


'.'.'.'. 


f 


30 




.... 


.... 



Westinghoxise standard engine has an offset of 50 per cent of crank radius. 

ratio of L/R varies from 3,37 to 4.8 and the percentage of offset varies 
from 15 to 50. The average diameter of cyhnder-bore is 4.81 in. and 
the average ratio of length of piston to diameter is ] ,24. 



242 OFFSETTING CYLINDERS IN SINGLE-ACTING ENGINES 

45 For comparison of engines the following dimensions were 
taken : 

Crank radius = 2^ in. 
Diameter of bore = 4| in. 
R.p.m. = 1000 

Weight of reciprocating parts per square inch of piston head area 
= 0.6 lb. 

Ratios of L/R = 2>\, 4, and 4^ and an offset, for each of the values 
of L/R, the largest amount practicable. These offsets are: 

L/R = 4i Offset = zero 

« = 4^ " =0.40 i2 

" = 4 " =0.30 i? 

" = 3J " =0.20 R 

46 Tables were prepared for each of the above cases, and values 
calculated for crank angles varying^by increments of 15 deg. each. 
Each of these tables contained values for the crank angle, piston posi- 
tion factor, the actual piston position, the gas pressure, inertia factor, 
inertia force, piston pin pressure, tangential factor, and the turning 
force from which the inertia curves and turning effort curves were 
plotted. 



Fig. 20 Curve of Side Pres.sures on Piston Position Base 

W 
R.p.m. = 1000; j- = 0.60 lb.; R = 2\ in. Full Line, L -r- R = 4^; Offset = Zero 

Broken Line, L -^ R = 4J; Offset = 0.40 R. Mixed Line, L -^ R = 4; 
Offset = 0.30 R. Dotted Line, L -5- R = 3^; Offset = 0.20 R. 

47 Careful comparison of the curves in Fig. 20 will show a slight 
difference between them, but not enough to warrant the trouble of 
plotting them separately for use in connection with the gas pressures 
to find the piston-pin forces from which the side-pressures are deter- 
mined. 



OFFSETTING CYLINDERS IN SINGLE-ACTING ENGINES 



243 



48 The inertia forces shown in Fig. 21 were combined with the gas 
pressures and the curves of side-pressures plotted as before, with 
the results shown in Fig. 20. 

49 The maximum values for the side- pressure were determined 
and the areas representing the work done by the piston on the cylinde- 




Fig. 21 Curves of Inertia 
Forces ox Piston Posi- 
tion Base 
R.p.m. = 1000; 
R=2i m;? =0.60 lb. 

1 L -i- R = 4i Offset = Zero 

2 L ^ R = 4^ Offset = 0.40 R 

3 L -T- R = 4 Offset = 0.30 R 

4 L -=- R = 3^ Offset = 0.20 R 



Fig 22 Turning Effort Curves on 
Crank Angle Base 

Full Line, L -- R = 4^ Offset = Zero 
Mixed Line, L -^ R = 3^ Offset = 0.20 R 
Dotted Line, L-T-R=4i Offset = 0.40R 
Broken Line, L -f- R = 4 Offset = 0.30 R 



walls were carefully measured and recorded (see Table 11). As far as 
these quantities are concerned, the best value is L/R = 4|, offset = 
0.40 R. The turning-effort curves (shown in Fig. 22) are so nearly 
alike that the difference is hardly worth mentioning. 



TABLE 11 



SIDE PRESSURES AND WORK DONE ON CYLINDER BY PISTON OF 
TYPICAL ENGINE 







1 

MAX. SIDE 


PRESBDBB 


WORK 


DONE IN AREA 


UNITS 


L/R 


Offset 












One Side 


Other Side 


One Side 


Other Side 


Total 


4J 


zero 


24 


7 


1.55 


0.60 


2.15 


4i 


0.40 


12 


12 


0.93 


0.87 


1.80 


4 


0.30 


17 


12 


1.17 


0.82 


1.99 


3* 


0.20 


i " 


1 12 

1 


1.53 


0.81 


2.54 



50 Table 12 gives a comparison of the four cases chosen. This 
table explains itself, but a short discussion may bring out the impor- 
tant points more clearly. 



244 OFFSETTING CYLINDERS IN SINGLE-ACTING ENGINES 

51 There is a slight increase in the length of stroke, but less than 
one-half of one per cent, so that it amounts to very little. The angle 
passed over by the crank during the out-stroke is slightly greater than 
180 deg. and the greatest gain is 1.7 per cent, which is small. The 
first great difference occurs in the length of connecting rod. No. 4 
effecting a saving of 2.50 in, or 22.2 per cent. 

52 Referring to the next line, the distance from the center of the 
crank-shaft to the position of the center of the piston pin at the end 
of the stroke, is a measure of the height of the engine and shows a 
decrease corresponding to the value of L/R. 

53 The maximum side-pressure decreases with the offset and 
increases with the decrease in value of the ratio L/R, so the best case 
would be No. 2, where L/R is largest and the offset is also largest. 
Here a reduction of 50 per cent is gained, which reduced the necessary 
length of the piston 44 per cent. No. 4 is the worst case, L/R very 
small and the offset also small and then the side-pressure is a trifle 
less than it is with no offset. The maximum value of the side-pres- 
sure affects the length of the piston and consequently the height of the 
engine, and the length of the cylinder, and so the weight of the cyl- 
inder and engine, and the weight of the piston and the correspond- 
ing weight of the reciprocating parts, and so the inertia force. The 
length of the piston is reduced 43.7 per cent in No. 2, 24.4 per cent 
in No. 3, and 3.6 per cent in No. 4. The ratio of length of piston to 
diameter is rather small in No. 2 but is not unusual in the other cases. 

54 If it is not desired to take advantage of the maximum value of 
the side-pressure by reducing the length of the piston, it can be made 
1.20 times the diameter, a usual value as is seen in Table 10, which 
would reduce the pressure per square inch of projected area and so 
increase the chances of satisfactory lubrication. The reduction of 
this pressure per square inch of projected area is shown in the next 
row. 

55 In order to find the exact resulting height of the cylinder up 
to the top of the piston at the end of the in-stroke it is necessary to 
calculate the position of the piston pin in the piston. This is done 
in the next row, by making the sums of the products of the areas with 
the distance from the piston pin center to their centers balance on 
each side of the piston pin. In case No. 2, with L/R = A.\, a 50 per 
cent offset might have been used without interference and this would 
give better results than 0.40 R offset, but in case No. 4, 0.20 R is 
undoubtedly about as much as could be used although it would be 
desirable to use more if a very low engine were wanted. 



OFFSETTING CYLINDERS IN SINGLE-ACTING ENGINES 



245 



^ 



- - kO ^ .3 

O ■-" o^ 0> vp 

US lO M »-i U5 o^ 

• • • • (N CO 

O (N OO ■* • • 

.-H CO 



65 



■<1< fe^ vO o> • 

CD >-c •* (S~^ CO 00 

• -MO • .H 

N >0 ■ • I I 



ft; OS 



a; CO « 

OS 00 CO 



^ 



^ 



— 


r,^!? 


CO 


CO Tt< 


CM IN (N 6^ 


r-l ■* 


• • ■ OS 


e^i 


-H ^ 05 <N 





■^ ^ CO 

-^ <=^. '^ 65 

CV| (N iH •"ll 



OS >o 



CM c t^ 

o — ■ r^ CO 

• O ■^ ,-1 

CO • • 

r- o c^ t^ CM lO 



Oi CO 
CM CO 

c<i la 



O 11 
< I 



^ £S 






«65 



^ 



CO 
CD 


CO 




^ 


CO 


5S 

CO 


^ 


05 


CM 


CM 




CO 





^s^ 



CM CO 

i-l CO 



CO rt 

CO 00 

CM CM CO 



o C^ t> 
O • ■ 

CO .-H CO 



00 .-I 



tf s 



^•2 



(D 

13 en 2 
in u a 

o o ■". 



Id g 



S tf 



^2 s i ►^ 

S g .2 I .2 



iJ w 



S -o 



j3 p "O 



Q Q 



1-3 



^ ^ 



246 OFFSETTING CYLINDERS IN SINGLE-ACTING ENGINES 

56 The next two rows show the distance from the center of the 
crank-shaft to the end of the piston at the end of the in-stroke, which 
is a measure of the height and the length of the cyhnder and also of 
the w'eight of the cylinder. As regards the height a gain of 15 per 
cent may be had in No. 4. No. 2 gives the shortest cylinder, 25 per 
cent shorter than No. 1, wMle No. 4 gives one only 2 per cent shorter. 
It must be borne in mind that these values are for 1000 r.p.m. and 
that the value of the maximum side-pressure will increase with the 
speed. However, the low value of the pressure per square inch of 
projected area, 15 lb., allows a considerable increase before a dan- 
gerous value is reached. 

57 The total amount oi lost work is shown in the next line. 
No. 2 gives the best value, a saving of 16 per cent, while No. 4 gives 
a loss of 18 per cent. 

58 In worldng out a satisfactory solution it would seem that one 
of two predominating ideas should be followed. Either a very low 
engine should be aimed at in which everything is sacrificed to height, 
or else the important object is to reduce to a minimum the side-pres- 
sure and the work lost due to friction resulting from side-pressure. 

59 In the first case, let L/R = 3^, offset as much as possible with- 
out interference, and a reduction in height of 13 to 15 per cent may 
be had. This means a reduction and a saving in weight of the con- 
necting rod, cylinder, valve stems, exhaust pipes, inlet pipes, and 
piston. The actual saving in length in the case above is 2f in. There 
will be some increase in the work lost in friction due to the increased 
average pressure of the piston on the cylinder walls. 

60 If a reduction in height is not of primary importance, then a 
ratio of L/R = 4^ and an offset of 0.40 R to 0.50 R would seem to 
give the best results. This gives a reduction in total height of 8 or 
9 per cent, a reduction in piston length of 44 to 45 per cent, a reduction 
in cylinder length of about 20 per cent, and a saving in lost work of 
about 16 per cent. These reductions would cause a further reduction 
in weight of piston, weight of cylinder, weight of valve stems, 
weight of exhaust and inlet manifolds, and a reduction of inertia 
effects as well as an increased life to the piston, piston rings and cyl- 
inder. In this case it might not be desirable to take full advantage 
of the reduction in length of piston, maldng it less than the stroke 
because the oil hole in the side of the cylinder, if one were used, 
would be uncovered at one end of the stroke or the other. 

61 In concluding this comparison, the most desirable offset seems 
to be as much as can be practically obtained with ratios of L/R = 4 



OFFSETTING CYLINDERS IN SINGLE-ACTING ENGINES 247 

and greater, with a decided gain over an engine with no offset for 
speeds less than 1400 to 1500 r.p.m. The subject may be summed 
up as follows: 

62 Improvements due to offsetting, (1) in the thermal cycle, (2) 
in the mechanical arrangement, (3) in the turning effort curve, and 
(4) in lubrication, are very shght and may be neglected. The real 
advantages are: 

a A reduction of the frictionai losses due to the pressure of 
the piston on the walls of the cyhnder, resulting in a 
slight increase in mechanical efficiency and less wear of 
the piston, piston rings, and cylinder, and consequently 
longer life. 
h A reduction of the maximum value of the side-pressure 
of the piston on the walls of the cylinder, allowing the 
use of shorter connecting rods, shorter pistons, and 
shorter cylinders, resulting in a shorter and lighter engine 
and in lower inertia-forces due to the reciprocating parts. 

The most important of these advantages would be a considerable 

saving in weight. 

63 The disadvantage of offsetting lies in the fact that the reduc- 
tions in average side-pressure and maximum side-pressure grow less 
as the speed and inertia-force increase, so that for a speed of 1400 to 
1500 r.p.m. there is either no reduction at all or an increase. 

Principal Conclusions 

64 Offsetting increases slightly the length of stroke and the crank 
angle passed over during the stroke toward the crank shaft. 

65 The maximum value for the side-pressure of the piston on the 
cyhnder walls decreases as the offset increases up to a value of one- 
half the crank radius for any ratio of L/R. 

66 The work lost in friction due to the side-pressure of the piston 
on the cylinder walls decreases as the offset increases up to a value 
of 50 per cent of the crank radius. 

67 Both the maximum value of the side-pressure and the work 
lost in friction increase as the value of the ratio L/R decreases. 

68 Offsetting decreases the height and weight of the engine. 

69 Offsetting increases the life of the cylinder and piston. 

70 Offsetting improves the thermal cycle. 



248 DISCUSSION 

71 The turning-effort curves when the cylinders are offset differ 
but slightly from those for no-offset. 

72 The advantages of offsetting as regards the maximum side- 
pressure and work lost may be zero or negative for high inertia-forces 
resulting from speeds of 1500 r.p.m. or more. 

DISCUSSION 

WiNSLOw H. Herschel. In December 1907, 1 had occasion to inves- 
tigate the question of offsetting cyHnders for large-sized gas engines, 
and as the conditions are somewhat different from those of the auto- 
mobile engines considered by Professor Phetteplace, the results may 
be of interest. I shall consider only the variations in maximum and 
average pressure on the cyhnder walls, since, as indicated in Par. 62 
of the paper, these are the main questions at issue. 

2 For the sake of simplicity I used the graphical method men- 
tioned in Tolle's Die Regelung der Kraftmaschinen, page 32. The 
computations were based on an actual card from a four-stroke cycle 
producer-gas engine, and upon the following data: 

R.p.m. = iV = 225 

W/A =4.18 

22 = 12 in. =1 ft. 

L/R = 5 

0.00034 W/AN^R = 71.8 

This last value, 71.8, happens to be very nearly the average of the 
two corresponding values, 34.4 and 111.38, given by the author in 
Table 3. Computations were also made using speeds of 450 and 1000, 
giving inertia constants of 289 and 1430 respectively, but it soon 
became evident that there could be no gain from offsetting under 
these conditions, and the investigation was restricted to the speed of 
225. 

3 As I understand the paper, the author has considered only 
vertical engines, or horizontal engines where the weight of the recipro- 
cating parts is so small that its direct effect in increasing or decreas- 
ing the pressure on the cylinder walls need not be taken into account. 
In the present case, however, a distinction must be made between 
vertical and horizontal engines. For the latter, when the side pres- 
sure acts downward, due to gas pressure or inertia forces, the weight 
of the piston must be added, but when the side pressure acts upward, 



OFFSETTING CYLINDERS IN SINGLE-ACTING ENGINES 



249 



the piston weight is subtracted. It should be noted that for a vertical 
engine, for a given value of 0.00034 W/AN^R, that is, for a given 
inertia constant, as I have called it, it would make no difference 
whether this value were obtained with a large value of W/A and a 
small value of A^, as in my case, or with a small value of W/A and a 
large value of A^, as in the cases used by the author. But on the other 
hand it would make considerable difference for a horizontal engine 
where the value of W enters into the computation apart from the 
inertia constant. 

4 By using the same indicator card as for the four-stroke cycle 
computations, and disregarding the second and third strokes, I 




■iO^i 40^ 50^c 60^ 70f 
Offset in Per Cent of Crank 



90/ 



Fig. 1 DiAGKAM Showing Variation in Maximum Pressure on Cylinder 
Walls, Due to Offsetting Cylinder 



obtained, somewhat approximately, the side pressures for a two- 
stroke cycle engine. 

5 Fig. 1 shows the variation in maximum pressure on cylinder 
walls, or side pressure, due to variations in offset. The ordinates 
above the base line ST are proportional to the side pressures. The 
line AB shows the maximum side pressure for a vertical engine at 
about the middle of the first stroke. As the offset increases the 
angle of the connecting rod for this middle position decreases, while 
the angle at the end of the stroke increases, so that for large offsets 
the maximum side pressure is found at the end of the stroke, with 



250 



DISCUSSION 



values as shown by line BC. Similarly for a horizontal engine we get 
the lines D£J and EF. For the maximum side pressure due to inertia 
forces we have hne GH for a vertical, and Une JK for a horizontal 
engine. 

6 It will be noticed that line JK is not parallel to line GH. The 
reason for this is that for the hne JK we must use the inertia force 
near the end of the second stroke, which gives a downward pressure 
on the cyhnder walls to be added to the weight of the piston, and this 
downward pressure is not as large as the upward pressure near the 
beginning of the second stroke, which was used for the Une GH. 

7 In the case of a two-stroke cycle engine, where we must use 
the fourth instead of the second stroke, our maximum inertia force 
will be near the beginning of the fourth stroke acting upward, so that 
the weight of the piston must be subtracted, giving hne QR. 

8 If we imagine a somewhat earlier ignition than that shown in 



TABLE 1 OFFSET AND PER CENT REDUCTION IN SIDE PRESSURE 



Curves 


Dominating Factors 


Cass 


% 
Offset 


Gain 


AB and GH 


Gas middle first, inertia beginning 2nd 


Vertical late ignition, 2-4 Cycle 


37.0 


41.5 


ABanALM 


Gas middle first, gas beginning 1st 


Vertical early ignition, 2-4 Cycle 


28.7 


31.6 


DE and JK 


Gas middle first, inertia end 2nd 


Horizontal late ignition, 4 Cycle 


37.7 


34.7 


DE and QR 


Gas middle first, inertia beginning 4th 


Horizontal late ignition, 2 Cycle 


61.3 


54.3 


DE and JK 


Gas middle first, inertia end 2nd 


Horizontal early ignition, 4 Cycle 


37.7 


34 7 


DE and NP 


Gas middle first, gas beginning 1st 


Horizontal early ignition, 2 Cycle 


40.6 


37.2 



Fig. 2 of the paper, the maximum side pressm-e at or near the begin- 
ning of the first stroke will be increased. Whether for this reason or 
not, I found that with large offsets the maximum side pressure of the 
first stroke was at the beginning of the stroke, acting upward, with 
values as shown by line LM for a vertical, and line NP for a horizon- 
tal engine. 

9 Fig. 1 corresponds to Fig. 16 of the paper and may be used in the 
same way to determine the most favorable offset for the various con- 
ditions. Table 1 gives the offset and the per cent reduction in side 
pressure in each case. 

10 The author (Par. 22) finds the most favorable offset to be 
50 per cent of the crank length for slow speed, and in Par. 24 and 
Par. 25 he finds it to be about 20 per cent for high speed. These 



OFFSETTING CYLINDERS IN SINGLE-ACTING ENGINES 



251 



values may be compared with the first fine of Table 1 ; for the case of a 
vertical engine with late ignition, the offset is 37.0 per cent, which is 
nearly the average of 20 per cent and 50 per cent, as might have been 
expected from the inertia constants. 

11 Fig. 2 shows the decrease or increase in average side pressure 
or total loss of work from side pressure, in per cent of lost work with 
zero offset. While the use of a different indicator card with a later 
ignition might have made some difference, it obviously could not 
have changed the results so materially as in the case of maximum side 



Fig. 2 



40S^ 




























o.Cl^ 


^ 












a, 30^ 








\etv 






X 






y y^ 


„ 4-t 


;ycle, Hi)iizonfni 






\ 


IO5; 


/y^^ 


1 " vJrf.r,, 


^^""^ 1 J.'V) 1 


!i2£2 








r"-\^ 


\ 


5 10^ 

a 




— I^o~ 


SS^- 


^"ca; 




1 




^ 




















\ 



10^ 



•iwi 



i% 



30ji 40!4 5U5J 6US^ 1% 

Offset in Per Cent of Crank 

Diagram Showing Change in Average Pressure on Cylinder 
Walls Due to Offsetting Cylinder 



pressure. Thus the curve marked four-cycle vertical may be fairly 
compared with the result in Table 9 of the paper, that the most favor- 
able offset lies between 30 and 50 per cent. 

12 The curves marked 450 and 1000 r.p.m. show the results of 
the few computations concerning these speeds not considered in Fig 1 . 

13 Both Table 1 and Fig. 2 appear to indicate that more could 
be gained from offsetting with a tAVo-cycle than with a four-cycle 
engine. But at present it is difficult to make general statements 
about this type of engine, and whether or not this advantage will be 
attained will depend upon the inertia constant and indicator card 
shown by these engines. 



252 DISCUSSION 

PRINCIPAL CONCLUSIONS 

a An offset cylinder may be employed with least benefit on a 
high-speed four-stroke cycle vertical engine. 

h It may be employed with most benefit on a slow-speed 
two-stroke cycle horizontal engine. 

c The maximum advantageous offset is limited by the side 
pressure near the beginning of the first stroke. 

John H. Norris. I have been designing and building both two- 
stroke and four-stroke cylinder engines with offset cranks for a num- 
ber of years. Our concern was so impressed with the advantage 
that in 1886 they bought the patent right to apply the offset stroke 
to gas engines. We are still building a few small sizes with offset 
crank, but there is no practical gain and as fast as we can re-design 
the engines we find we can get better economy and a more convenient 
engine, by placing the cylinder directly over the crank shaft. If you 
want to reverse the direction of rotation of an engine with an offset 
crank, you are in trouble at once. We have built an offset engine as 
large as four-cylinder, 14 by 22, and scrapped it. I would like 
to say, in connection with the large engine above mentioned, that it 
was built in 1896 and was one of a pair that were to run a suburban 
electric railroad in the West, on a suction gas producer. We buUt at 
that time a suction producer that was reasonably satisfactory. We 
have had a great deal of success, however, with our small single and 
double-cylinder engines with offset cranks, of which we have built a 
large number, though we are just putting on the market engines to 
replace them, in which we have placed the cylinder directly over the 
center. We used an offset of from 20 to 25 per cent of the stroke. 
With this offset the side strain is quite sufficient on the upstroke. 
These were all vertical engines. 

The Author would suggest that the fact that the cylinders were 
offset was not the real cause for scrapping the 4-C5dinder 14 by 22 
engine mentioned by Mr. Norris. Of course if stock engines were built, 
some to rotate in one direction and some in the other, or if it is desir- 
able to build engines that may be reversed, offsetting may not prac- 
tically be taken advantage of, as Mr, Norris points out. Furthermore, 
in spite of Mr. Norris' experience and his desire to eradicate offsetting 
from his product, this practice, in small vertical 4-cycle automobile 
engines, at least, seems to be increasing. 



No, 1241 

PRESENTATION OF PORTRAIT OF GEORGE 
AV. MELVILLE 

At the Spring jNIeeting, \Yashiiigton, INIay 1909, of The American Society 
OF Mechanical Engineers, a portrait of Rear-Admiral George W. Melville, 
U. S. N., Ret., painted by Sigismond de Ivanowski, was presented by friends to 
the National Gallery. Previous to the presentation of the portrait an address 
was made by Admiral Melville on The Engineer in the L'''. S. Navy. This is 
given in abstract, together with the addresses of presentation by Walter M. 
McFarland, Mem.Am.Soc.M.E., and of acceptance of the portrait for the 
Nation by Dr. C D. Walcott, Secretary of the Smithsonian Institution. 

THE ENGINEER IN THE U. S. NA\^ 

By Rear- Admiral Geo. W. Melville, U.S.N., Ret. 
Past-President of the Society 

Ten years ago my presidential address before the Society had 
almost the same theme as my remarks this evening. At that time 
the Personnel Law was passed, which amalgamated the engineer 
corps with the line or executive officers of the navy, with the under- 
standing that thenceforth engineering was to be the function of 
these Une-officers. In his report as Chairman of the Personnel Board, 
ex-President Roosevelt, then Assistant Secretary of the Navy, said, 
"On the modern war vessel every officer has to be an engineer 
whether he wants to or not." It is well that these lines should be 
constantly in mind, for they set forth the only justification for the 
Personnel Law. 

2 Remarks have been made to the effect that a line-officer 
charged with all these duties would be a hybrid or Jack-of-all- 
trades. It is to be noted, however, that our naval officers have to 
perform definite duties. The curriculum of the naval school can 
be planned to give them a thorough and specific training for the 
work they have to do. In this respect these young men have a 
decided advantage over the students of any of our great technical 
schools, who can receive instruction only in general principles because 
they themselves rarely know the particular line of engineering 
work which they will follow. It is, in my judgment, just as ridicu- 
lous to speak of our modern line-officers, specially trained for the work 



254 PRESENTATION OF PORTRAIT OF ADMIRAL MELVILLE 

they have to do, as Jacks-of-all-trades, as it would be to apply 
this designation to a blacksmith, a lawyer or a doctor. 

3 I have not forgotten that I am talking of men who go to sea 
and that the line-officers are responsible for the handling of the ships; 
in other words, that seamanship is an essential element of the 
training. It must be remembered that the modern navy has entirely 
dispensed with sails and that it is a misnomer to call the modern 
man-of-warsman, a sailor. He is not a sailor, because there are 
no sails for him to handle. He is a seaman, because he goes to sea. 
Seamanship is an art, proficiency in which comes almost entirely 
from practice, so that officers who are given the other portions of 
the training in the classroom, laboratory and workshop, get the 
requisite proficiency in seamanship from the practical exercises dur- 
ing their career as midshipmen at the naval school, and in the 
handling of vessels after they graduate. 

4 It is natural to inquire how the amalgamation has worked 
out in practice. On January 21 of this year, the Chairman of the 
House Naval Committee quoted from the remarks of the officer 
who commanded the battleship fleet which cruised around the 
world, to the effect, " When I got to California, without any engi- 
neers, my fleet was in better condition than when it started." It 
would seem, however, to have agreed much better with the avowed 
intention of the Personnel Law if he had said, "Our cruise was 
a great success because every officer was an engineer." 

5 The Chairman of the House Naval Committee further said, 
" It is the opinion of our naval officers, in command of our fleet 
and ships, that this consolidation has been a splendid thing for the 
navy, because it makes the man in command of the ship, the master 
of the ship, a man who understands all the workings of the sliip. 
Before, the command of the sliip was in the hands of the engineer. 
We had to make a change in the curriculum of the Naval Academy 
whereby the officer of midshipmen there must acquire a knowledge 
of engineering, further adding to that the experience which he 
must obtain in the engine room as a watch officer. By reason of 
these facts, the entire ship is toda}^ under the command of an engi- 
neer officer, a man who understands all the duties of engineering 
and who is complete master of the ship." 

6 I have been told by officers who have recently served on 
board ship that one great benefit has resulted from the amalgama- 
tion: namely, tliat the idea just expressed in the above quotation 
is true; that the commanding officer is now the master of the entire 
ship. In my early days, few commanding officers felt any interest 



PRESENTATION OF PORTRAIT OF ADMIRAL MELVILLE 255 

in the machinery beyond their demand that it should always be 
ready for service. If anything went wrong, they washed their 
hands of the responsibility, which was naturally upon the special 
body of engineers. They now feel the same keen interest in the 
machinery that they do in the guns or any other part of the ship, 
and the chief engineer of the ship (still so-called) is generally looked 
upon as the officer next in importance after the captain. 

7 The part of the new regime about which I have felt misgivings 
is that thus far there has been no systematic effort to assure train- 
ing and experience for every line-officer in connection with the 
motive power. Every young officer should be required to serve an 
apprenticeship in the engine and fire-rooms, just as he does on 
deck, but so far as I have been able to learn there has never been 
careful attention given to this point. 

8 Having touched upon the general conditions of the executive 
side of engineering as affecting the operation and integrity of the 
machinery at sea, it is now pertinent to consider the prospects with 
regard to present and future designs. Thus far, this work has 
remained in the hands of officers specially trained. Unfortunately, 
the same condition is found here as mentioned in the previous divi- 
sion of the subject. An effort has been made to arouse the interest 
of some of the younger line-officers by a course of special training 
for this most important work after the present highly competent 
and experienced men have retired. There has, however, been no 
settled policy for this, and the attempt that was started was inter- 
rupted by the cruise of the Atlantic Fleet. 

9 I am very glad indeed to bear testimony to the fact that 
the recent designs of the Bureau of Steam Engineering have been 
highly creditable in every way. In saying this, I feel a touch of 
personal pride for the reason that the men who have been doing 
this work were formerly my assistants and received most of their 
experience during my term of office. I am naturally pleased that 
the record which was made during my own term is being maintained. 

10 When such praise as this can be given in simple truth, what 
can be thought of the official who plans to discredit the men who 
have made such a record, and destroy the autonomy of the Bureau 
by subordinating it to the Bureau charged with the design of hulls? 
I believe you will agree that m}'- service of a lifetime in the Navy 
and my record as the head of a great Bureau in the Department, 
the longest since the Civil War, entitles my opinion to some weight, 
and I want to register my earnest conviction that any such scheme 
of consolidation can only bring inefficiency, retrogression and waste. 



256 PRESENTATION OF PORTIIAIT OF ADMIRAL MELVILLE 

11 There is still another side to engineering in the navy; namely, 
the work of the navy yards, which has been prominently before the 
pubhc during the regime of the last Secretary of the Navy. Changes 
have been made abolishing the separate departments in the navy 
yards and consolidating their administration under one officer, 
whose work, while a vital element in the building of a ship, was 
certainly not the only important part, and moreover was so different 
in its nature from the other departments which were absorbed, 
that it is obvious he could not be an expert on these other lines. 
To me it was so marvelous as to be almost beyond belief that in 
this age of specialization a movement so absolutely counter to the 
spirit of the age should take place in the name of economy and 
reform. If the methods of great shipyards in civil life, or of the 
great manufacturing establishments, or the dockyard administra- 
tion of other countries, had differed from the methods employed 
in our navy yards, a change would at least have been indicated; 
if, in these places a system somewhat like the one which it has 
been attempted to introduce in our navy yards, had been in vogue, 
there could be some understanding of the change; the facts are 
however, that in its essential featm'es our navy yard administration 
was along the very lines which obtain in foreign dockyards, in the 
great shipyards at home and abroad, and in our great manufactur- 
ing estabhshments. 

12 I am led to believe that the present Secretary is giving the 
matter very careful consideration Avith a view to undoing the 
tremendous harm brought about by his predecessor, and I trust 
he will be well-advised and will restore the yards to their former 
efficiency. It ought to be said, however, as a matter of record, that 
these changes were made without any consultation between the 
late Secretary and the officers most competent from long experience 
to know what was best. Indeed, by his own statement, the scheme 
was evolved from his own inner consciousness. 

13 Our modern navy is essentially an engineering affair. The 
vessels themselves are the product of the engineer's brain, and their 
successful maintenance and utilization depend entirely on engineer- 
ing skill. Ten years ago I said that the change which had been 
made, absorbing the engineer corps into the line of the navy and 
making every officer an engineer, was a tremendous step forward, 
provided a sincere and earnest effort was made to carry out the 
scheme which was thus outlined. 

14 From what I have said this evening, it will be clear that 



PRESENTATION OF PORTRAIT OF ADMIRAL MELVILLE 257 

I am not as yet satisfied that this has been brought about. Undoubt- 
edly the responsibility for the machinery of our vessels, guns, motive 
machinery, electrical machinery, torpedoes, etc., is upon the line- 
officer of the navy. He is charged with this duty by law. If the 
older officers of the navy had taken hold of this matter with enthu- 
siasm, I believe that it would now have been settled and there 
would be no question as to the great success of the new officer, the 
line-officer of the twentieth century. I am not willing to believe 
(and indeed hardly willing to consider) the possibility that naval 
officers will neglect any duty with which they are charged, and I 
still hope that the scheme will be worked out to a great success. 

15 Not much is ordinarily said about the machinists who are 
doing good work on board our vessels. They look after the routine 
work of repair and adjustment on board ship, but they are without 
the scientific training which is required for engineers who are 
really qualified for the duties comprehended by that title. If the 
line-officers of the navy do not maintain engineering efficiency, it 
will then, as the organization now stands, fall upon these machinists 
to perform the work of the engineers. In other words, in an organi- 
zation whose efficiency is absolutely dependent upon the skill of 
engineers, the men relied upon for such work would be relegated to 
a position of inferiority so low that they are hardly counted. This 
is utterly un-American and can only be matched by absolute mon- 
archies or countries as unprogressive in the mechanic arts as Spain. 
The speedy destruction of her navy in the war of 1898 was due 
to her utter incompetence even more in engineering than in gunnery. 
It is inconceivable that self-respecting men in a free country like 
ours will attempt to perform work of such vital importance with- 
out adequate recognition in the way of rank and position. 

16 I will not permit myself to believe that we shall have to 
consider this as a practical • question^'because I cannot conceive 
that naval officers would fail in their duty, but I feel that both sides 
of the question, so vital to our naval efficiency, should be presented. 

17 During my entire career in the navy, it was my constant 
endeavor to show by my work the importance of the engineer and 
to encourage that spirit in my subordinates. I trust I will not be 
accused of vanity if I say that I believe my record as engineer-in- 
chief added a little to the reputation of engineering. My active 
work in the navy is done, but so long as I live my interest will never 
slacken and my voice will always be raised to encourage efficiency 
in every branch of the service. 



258 PRESENTATION OF PORTRAIT OP ADMIRAL MELVILLE 

PRESENTATION OF PORTRAIT 

By Walter M. McFarlanb, Mem. Am. Soc. M. E. 

The honor of being invited to pay a tribute to my dear old Chief, 
Admiral Melville, is one which I appreciate highly, as well as the 
allied one of acting as spokesman for the donors of the splendid 
portrait which is to be presented to the National Gallery this evening. 
I admire the Admiral as the fine flower of a splendid type of man- 
hood, and his kindness to me for many years has been so like a 
father's that with a son's affection I rejoice at this splendid testi- 
monial to his personality and his work. 

2 Too often the pathway to greatness and fame is marked by 
the wreckage of competitors, and even friends, who have been ruth- 
lessly thrust aside in the egoism of selfish ambition. Then, there 
may be a grudging admission of ability, but there is no love, no true 
admiration. When, on the other hand, the hero has always been 
the helper and friend of his companions, when he has cheerfully 
acknowledged their aid to his success, we have such greatness as 
we are celebrating tonight. Then, every member of the profession 
feels that the fame of the leader is reflected on the whole body, and 
they love the man while they rejoice in his reputation. 

3 George Wallace Melville is such a man. He has been one o/ 
the famous men of engineering so long that we find it hard to remem- 
ber a time when his name was not synonymous, as it is now, with 
all that represents progress and achievement in our profession. 

4 It is a matter of delight to all of us who love him that the 
artist, in the picture which is to be presented to the National 
Gallery this evening, has faithfully depicted the chief character- 
istics which have made him great. These are, in my judgment, 
indomitable courage and unbending honesty. It is possible for a 
man to have great mental ability and yet fail of true greatness 
if he lack these essentials. 

5 You all know Melville's Arctic record, which first brought 
him an international reputation; there he displayed a heroic 
courage which has never been surpassed, and for which Congress 
advanced him a grade in the Navy. This, however, was only 
one instance of the absolute fearlessness that began with his 
earliest days in the service. When he became Engineer-in-Chief, 
the same courage, but rather on the moral than the physical 
side, was shown. Beginning with his first annual report, he 



PRESENTATION OF PORTRAIT OF ADMIRAL MELVILLE 259 

spoke out fearlessly, setting forth the truth as he saw it and 
striving always for advancement and efficiency. Complaint was 
made to President Cleveland of the plain speech in this first report, 
but that strong man read it himself and said, " We want more 
such men." 

6 His professional courage is also remarkable, and, moreover, 
a faculty, I believe which is characteristic of all great men, 
having once made his decision he does not worry about the result. 
Able men of minor rank are always fearful that something may 
go wrong and their reputation be injured. The really big man 
does not believe himself infallible. He knows that all men who do 
things will make some mistakes, and he is strong enough not to 
dread them. A notable instance of this kind in Melville's career 
was his use of triple screws for the Columbia and Minneapolis. I 
saw letters from some of his friends, for whose professional opinion 
he had the highest regard, urging him not to make the experiment, 
but he had studied the problem carefully, was satisfied with the 
correctness of the solution, and persevered. The result was perhaps 
the greatest triumph of his professional career. 

7 His ability as an executive is of a very high order. The 
feature of deciding a case and then refraining from worry is an 
evidence. He had a rare talent for choosing able assistants, and 
having proved them he left in their hands all the detail work, 
thereby giving himself time for careful study of the 'larger problems. 
The effect of this was very marked in stimulating the entire staff 
to the highest efficiency and zeal. I have known them all personally, 
and every man counted it a pleasure to work, without regard to 
hours, for the credit of the "Chief" and the glory of the Service. 

8 With respect to his professional work, it is notable that his 
career as Engineer-in-Chief of the Navy, from 1887 to 1903, is the 
longest on record. It covers the building of the "new navy," and 
the Spanish war. During this time he was responsible for new designs 
of machinery for about 120 vessels, among which were 24 battleships 
and 41 armored vessels. Best of all, there were no "lame ducks," 
and no failures. 

9 I will mention briefly some details of his more important work. 
He was the first to use water-tube boilers in large war vessels and to 
determine the actual coal consumption by trials. He was also the 
first to use the method of determining trial-speeds, known as the 
"standardized screw," which is the simplest, most accurate and 
inexpensive, and fairest to the contractor as well as to the gov.ern- 
ment. 

^ 10 It is^to him also that we owe our first high-speed battleship. 
When'* in 1898 the proposals for the Maine, Missouri and Ohio 



260 PRESENTATION OF PORTRAIT OF ADMIRAL MELVILLE 

were being prepared, he stood alone in his demand for 18-knot 
ships. If he had not persisted, we should have been three years 
longer behind the other navies of the world in battleship speed. 

11 It is very interesting to note also that only a little after this 
he proposed an "all-gun one-caliber ship," in other words, what is 
now called the "Dreadnought" type. Before I left the Service, I 
had often heard him talk of this big ship with ten twelve-inch or 
twelve ten-inch rifles and nineteen or twenty knots speed; and about 
1899 he submitted a sketch plan of the battery of such a ship to the 
Board on Construction. Possibly the same influence which almost 
prevented the eighteen-knot battleships prevented consideration of 
this more advanced type. At all events Melville was in advance of 
the general naval mind, and our country lost the credit which it might 
have had for the introduction of this revolutionary improvement 
several years before the Dreadnought was produced. 

12 During the war with Spain he brought out the repair sliip 
and the distilling ship. The idea of the former was not new, but 
the Vulcan was by far the most complete vessel of the kind equipped 
up to that time. The latter furnished fresh feed-water to the boilers 
and enabled a vessel with a storage bunker capacity of 3000 tons 
to supply 60,000 tons of water. 

13 A clever piece of work at this time was the fitting of new 
boilers to some of the old Civil War Monitors to enable them to be 
used for harbor defense. For years Melville had advised the Navy 
Department that new boilers must be supplied before these vessels 
could be used. When the destruction of the Maine made the out- 
break of hostilities seem probable, the makers of water-tube boilers 
submitted estimates of time and cost for the work. Boilers were 
promised in 30 days, but it was necessary to use the standard land 
type. As these vessels were not to go to sea, however, this was 
satisfactory. The worn-out boilers were cut up and passed out 
through the smokepipe (because the armored deck could not be 
taken up; the new boilers were passed down the smokepipe in sections 
and erected on board; finally each of the boats was given a steam 
trial, which was entirely successful. 

14 A great deal of experimental work was done under his direc- 
tion, all of which is published in his annual reports. The last of 
such experiments was a series of tests of oil as fuel, probably the 
most comprehensive ever made. 

15 My brief sketch of this famous man would be incomplete if 
I failed to speak of his personality. The lion-like head and the 
frank speech have led some to say that he is one of the old " Vikings," 



PRESENTATION OF PORTRAIT OF ADMIRAL MELVILLE 201 

spared to us a thousand years after the others have gone; but if 
this leads any to think that he is harsh and cold, there could be 
no greater mistake. Like all strong natures, he is pronounced in 
his feelings, but he is a man of warm affection, and when he has 
once taken you into his heart, you are sure of an abiding-place 
there as long as you are worthy. It is often said that no man is 
great to his intimates, but I have been with him, day by day, for 
years; have seen him under all conditions; and my admiration and 
love for him have simply increased as the years go by. I have no 
ambition to be a Boswell and I have not kept notes of his doings; 
but I have seen the daily workings of a great, kind heart, tender 
for the humble yet fearless toward the great; and I can truly say 
that I count it a privilege and an inspiration to have been a trusted 
friend and helper of this noble man, who has exemplified the highest 
type of manhood and added new luster to the profession of engineering. 

ACCEPTANCE OF PORTRAIT 

By De. C. D. Walcott 

It gives me pleasure, speaking for the Smithsonian Institution as 
the custodian of the National Collection of Art, to accept from you 
for the people of this country this fine portrait of Rear-Admiral 
George Wallace Melville, to be exhibited in the gallery of portraits 
of Americans who have achieved eminence in their life work. 

2 Among the men who have rendered distinguished service to 
their country in literature, science, or art, in war or in peace, in 
professional or civil life — few have won such well-merited distinc- 
tion in so many lines of duty as Admiral Melville. He stands high 
in the regard of the Nation as a naval hero, as an engineer of excep- 
tional ability, and as a wise and resourceful administrator and 
advisor. It is only to be regretted that, under the operation of 
law governing retirement, Admiral Melville was obliged to retire 
from active duty in 1903, but it is to be hoped that the country 
which he has so efficiently and actively served may long be per- 
mitted to enjoy the benefits of his counsel. 

3 The portrait of Admiral Melville is a most appropriate addition 
to this National collection and it is peculiarly fitting that his serv- 
ices should be emphasized in this happy manner by a Society which 
embraces so distinguished an array of men in the engineering pro- 
fession, a Society that for nearly thirty years has exercised a power- 
ful influence toward unity of interest and harmony of purpose in 
the broad field of American engineering. 



No. 1242. 

SMALL STEAM TURBINES 

By George A. Orrok, New York 
Member of the Society 

The papers upon steam turbines which have been presented before 
the Society have dealt mainly with the larger types of apparatus and 
have been written to show the reliability, efficiency and general 
desirability of this type of prime mover. 

2 This paper treats of the smaller sizes of steam turbines from 
the standpoint of the designing and operating engineer, describing 
the commercial machines in sufficient detail, with reference to the 
service to which they have been applied, and giving certain facts con- 
cerning their operation which may be of advantage to the engineer- 
ing profession. Curves of steam consumption are given which show 
in a general way what may be expected of these machines under cer- 
tain conditions. 

3 At the present time seven machines are on the market and can 
be obtained in various sizes from 10 h.p. to 300 h.p. with reasonable 
deliveries. These are the De Laval, Terry, Sturtevant, Bliss, Dake, 
Curtis and Kerr turbines. Three other machines are nearly at this 
stage of development and patents have been applied for on several 
others. 

4 Many thousand horsepower of these turbines have been sold 
and are in successful commercial service. The following reports of 
total sales of sizes from 10 h.p. to 300 h.p. have been obtained from 
the manufacturers: 

For further discussion of Steam Turbines, consult Transactions as follows: 
Vol. 10, p. 680, Notes on Steam Turbines, J. B. Webb; vol. 17, p. 81, Steam Tur- 
bmes, W. F. M. Goss; vol. 22, p. 170, Steam Turbines, R. H. Thurston; vol. 24, p. 
999, Steam Turbines from the Operating Standpoint, F. A. Waldron; vol. 25, p. 
1056, The De Laval Steam Turbine, E. S. Lea and E. Meden; vol. 25, p. 1041, 
The Steam Turbine in Modern Engineering, W. L. R. Emmet; vol. 25, p. 782, 
Different Applications of Steam Turbines, A. Rateau; vol. 25., p. 716, Some Theo- 
retical and Practical Considerations in Steam Turbine Work, Francis Hodgkinson. 

Presented at the Spring Meeting, Washington, May 1909, of The American 
Society op Mechanical Engineers. 



264 



SMALL STEAM TURBINES 



De Laval, De Laval Steam Turbine Company 70,000 h.p. 

Curtis, General Electric Company 70,000 h.p. 

Terry, Terry Steam Turbine Company 15,000 h.p. 

Kerr, Kerr Turbine Company 10,000 h.p. 

Sturtevant, B. F. Sturtevant Company 

Bliss, E. W. Bliss Company 

Dake, Dake- American Steam Turbine Co 

5 All of these machines are of the impulse type: that is to say, 
the steam is expanded in a nozzle and the kinetic energy of the jet is 
absorbed by passing one or more times through the buckets of the 




Fig. 1 High and Low-pressure De Laval Turbine 



turbine rotor. In the De Laval turbine only one moving element 
and one steam pass are used, which necessitates a very high bucket 
velocity. In the Terry, Sturtevant, Bliss and Dake turbines a series 
of return passages is provided. The steam returns two or more 
times to the same rotor and the bucket speed is much lower. In the 
Kerr turbine the steam is used in stages with one bucket wheel in a 
stage; while in most of the Curtis machines two or three stages are 
used with two or three rows of moving buckets, separated by station- 
ary guide blades, in each stage. Compound machines of the other 
types have been made but are not as yet produced commercially. 

6 By far the larger number of these machines is used in connection 
with extra high-speed electric generators, the next larger application 



SMALL STEAM TURBINES 



265 



being to centrifugal fans for high pressures. Centrifugal pumps 
adapted to high rotative speeds have been rather generally introduced 
in the last few years and it is becoming usual to connect small turbines 
direct to these machines. The small space required and the simplic- 
ity obtainable in a 100-h.p. turbine at speeds of from 800 to 1200 
r.p.m. have been important factors in their introduction. 

7 The first of the small turbines to be put on the market was the 
De Laval, made by the De Laval Steam Turbine Company of Trenton, 
N. J., and introduced in this country about 1896. This machine is of the 




Fig. 2 Terry Steam Turbine, 36-in. 



pure impulse type, the steam being expanded in the nozzle down to 
the exhaust pressure, and the resultant velocity transferred to the 
wheel in one steam pass. The bucket speed is high, ranging from 
600 to 1300 ft. per sec. Eight sizes of wheels are made, generat- 
ing from 10 h.p. to 500 h.p., with one nozzle in the smallest size and 
eight or more in the 500-h.p. size. 

8 The high bucket speed necessitates the use of gears of special 
construction, which have been very successful. The design, construc- 
tion and economy of this type have been discussed in vol. 25 of 
Transactions, p. 1056. 



266 



SMALL STEAM TURBINES 



9 The Terry turbine, made by the Terry Steam Turbine Company 
of Hartford, Conn., has been manufactured for about ten years, 
although the commercial machine has been on the market only for 
about four years. This turbine is of the impulse type, but the steam 
passes through the buckets a number of times before its energy is 
absorbed. The case of the machine is parted on a horizontal plane 
through the shaft and at right angles to the wheel. The nozzles and 




Fig. 3 Terry Turbine Showing Construction 



return passes are bolted to the inside of both parts of the casing. 
The nozzles are in the plane of the side of the wheel. The return 
passages are of brass and are separated by partitions. The wheel 
itself is built up of two steel discs held together by bolts over a steel 
center. The buckets are built of steel punchings, fitting in grooves 
cut in the discs, as shown by the figures. The sizes of wheels manu- 



SMALL STEAM TURBINES 



267 



factured at the present time are 12, 18, 24, 36 and 48 in., and the 
number of nozzles varies from two on the 12-in. wheel to eight or ten 
on the 48-in. wheel. 

10 The Sturtevant turbine, made by the B. F. Sturtevant Com- 
pany of Hyde Park, Mass., has been in the development stage for 
three or four years and quite a number of machines have been sold. 
The present type of turbine may be called " standard, " however, and 
four sizes of wheel are built, 20, 25, 30 and 36-in., developing from 




Fig. 4 Sectional View of Terry Turbine 



3 h.p. to 300 h.p. The turbine is of the multiple-pass type similar 
to the Riedler-Stumpf. The casing is cast solid with one end. The 
nozzle and return chamber ring are inserted from one side and the 
wheel is milled from the solid. The return passages are from eight 
to twelve in number and are milled on the inside of the return cham- 
ber ring. They are partitioned and are similar in shape to the 
buckets. The nozzle lies in the plane of the side of the wheel. 

11 The Bliss turbine, formerly known as the American, made by 
the E. W. Bliss Company of Brooklyn, N. Y., is of the same type as 



268 



SMALL STEAM TURBINES 




Fig. 5 Wheel and Casing of Stttrtevant Ttjkbine 




Fig. 6 Sturtevant Steam Turbine, 30-in. 



SMALL STEAM TURBINES 



269 



the Terry and Sturtevant and has been on the market only a few 
months. The casing and steam chamber are cast solid with one side 
and the nozzle and return chambers bolted in. The wheel is milled 
from a steel casting, or forging in the smaller sizes, and the partitions 




Fig. 7 Section of Stxjrtevamt Turbine 



separating the buckets are inserted and held in place by three bands 
of steel shrunk on the face of the wheel. The return passages are 
peculiar in having no partitions. Two sizes of wheel have been built, 
the 42-in. and .30-in., but designs have been developed for the 12, 18, 
24, 36, 48 and 60-in., covering powers from 10 h.p. to above 600 h.p^ 



270 



SMALL STEAM TURBINES 




Fig. 8 Bliss Turbine, 30-in. 




Fig, 9 Dake Steam Turbine, 24-in. 



SMALL STEAM TURBINES 



271 



12 The Dake turbine, made by the Dake-American Steam Tur- 
bine Company of Grand Rapids, Mich., is a single-stage impulse tur- 
bine. The wheel is made of two bucket discs, with milled buckets 
and inserted partitions, bolted together over a wheel center. In their 
Headlight turbine the governor is enclosed between the sides of the 
wheel. The nozzles and return-passages are placed between the 
bucket discs. The machine is built in sizes of from 5 h.p. to 100 h.p., 
the diameter of the smallest wheel being 12 in. 




Fig. 10 Parts of the Bliss Turbine 



13 Coincident with the development of the large Curtis turbines> 
the General Electric Company, at their Lynn Works, have developed 
and placed on the market a line of small generating sets ranging from 
5 kw. to 300 kw. This range is covered by eight sizes, the smaller 
machines being single-stage with two or three passes per stage. The 
buckets and nozzles are of the well-known Curtis type. 

14 The Kerr Turbine, made by the Kerr Turbine Company of 
Wellsville, N. Y., is of the compound impulse type. It is generally 
built in from two to eight stages. The buckets are of the double 



272 



SMALL STEAM TURBINES 



Pelton type, inserted like saw teeth in the wheel disc. Five sizes of 
of wheels, 12, 18. 24, 30 and 36-in., are made and cover a range of from 
10 h.p. to 300 h.p. The nozzles are in the plane of revolution of the 




Fig. 11 Sectional View of Buss Turbine 



wheel and are screwed into the stage partitions and held in place by 
a lock nut. 

15 As in large turbines, details of these small turbines, to which 
reference has been made, show the skill and knowledge of the designer, 
and that the same problem may be solved in different ways is well 
illustrated by the sections here reproduced. 



SMALL STEAM TURBINES 



DESCRIPTION OF DETAILS 



273 



16 Nozzles. The diverging nozzle is used by all makers except 
Kerr, whose multi-stage wheel requires a converging nozzle. In the 
De Laval, Sturtevant and Kerr turbines, the nozzles are screwed into 
their seats; that of the Terry is held in place by a bolt. The nozzles 




Fig. 12 Section of Dakb Headlight Tuhbine; Exterior Shown in Fig. 9 

of the Curtis, Dake and Bliss turbines are reamed out of the solid. 
The larger sizes of the De Laval machine which have been put on the 
market lately have a large number of reamed nozzles instead of the 
older construction. 

17 Buckets. The constructions employed in the Curtis and De 
Laval wheels are well known and have been described many times. 



274 



SMALL STEAM TURBINES 




Fig. 13 Curtis Turbine 50 h.p. 




Fig. 14 Curtis Turbine in Process of Assembly 



SMALL STEAM TURBINES 



275 



The Terry, Dake, Bliss and Sturtevant buckets are practically semi- 
circular in form. The Terry bucket is constructed entirely of steel 
punchings assembled between grooves in the two steel discs forming 
the sides of the wheel. The Sturtevant wheel is milled out of a steel 
casting. The Bliss buckets are milled out, but the partitions are 




Fig. 15 Section of Cubtis two-stage, Non-Condensing Turbine, 160 h.p. 



inserted and held in place and steel rings are shrunk on. The Dake 
buckets are turned out of the solid, the recesses for the partitions 
milled out and the partitions inserted; the wheel is then bolted 
together. The Kerr buckets are very similar to the original Pelfon 
buckets and are inserted in the wheel in a manner similar to the 
De Laval buckets. 



276 



SMALL STEAM TURBINES 




Fig. 16 Curtis Turbine, 200 h.p. 




Fig. 17 Revolving Element of Curtis Turbine in Bearings 



SMALL STEAM TURBINES 



277 





278 



SMALL STEAM TURBINES 



1 i 


L 


^^^7> ^^1^1 


K 


J^JSSk - "^^^^1 


^v 


iLdM^BS i^^^^HHI-i '^^^^^^^^^^^1 


^^ 


^■.^ 





Fig. 19 Kekk Turbine, 18-in, 

18 Return Chambers. The Sturtevant returns are milled out of 
the solid ring. Bliss casts them in the nozzle piece and finishes them 
by hand; Terry casts each one separately, finishes by hand and assem- 
bles with bolts; Dake casts the return chambers solid, mills the pas- 
sages and covers them with a shrouding. 




Fig. 20 Complete Rotating Part, 18-in., 7-Stage, Kerr Steam Turbine 



19 Wheel Centers. De Laval, Curtis, Sturtevant and Bliss make 
the wheel centers of steel castings or forgings integral with the wheel. 
Terry uses a steel casting but bolts the wheel disc to it. Kerr uses 
a screwed coupling, the inner part cut in three pieces and keyed to 



SMALL STEAM TURBINES 



279 




280 



SMALL STEAM TURBINES 



the shaft with round keys, clamping the wheel disc. Dake's wheel 
centers are an integral part of the wheel in small sizes, but in the 
larger machines are steel castings, in some cases a part of the shaft. 
20 Governors. De Laval, Terry, Sturtevant, Bliss, Dake and 
Kerr use a fiyball governor on the shaft end, which actuates the 
throttle valve through a system of levers. Curtis uses the fiyball 
governor on the shaft for small sizes and slower-speed spring-con- 
trolled governors of different forms for the larger sizes. The Sturte- 
vant, Bliss and Curtis machines are provided with an emergency stop 
governor as well as the throttling governor. 




Fig. 22 One Stage op Kerr Turbine, Showing Nozzles and Wheel 



21 Glands. For non-condensing machines glands are not trouble- 
some, as the difference of pressure between the casing and atmosphere 
is rarely more than a few pounds. Terry uses a bronze ball-and- 
socket gland with a long loose fit on the shaft. Sturtevant and Dake 
use a set of ring packing, either cast-iron or bronze. Bliss has a laby- 
rinth packing without contact. Kerr has a floating bronze bush with 
soft packing behind it. Curtis uses a metallic packing held in place 
by a. gland ring, and for condensing service a carbon-ring packing, 
steam-sealed. 

22 Clearance. In none of these machines is clearance an impor- 
tant factor. The clearance between buckets and guide passages on 
A 24-in. wheel is usually from ^ in. to ^ in. when hot. Striking or 
rubbing is practically unknown. 



SMALL STEAM TURBINES 



281 



23 Thrust. Theoretically, there should be no thrust in any 
turbine of these types. Practically, there is always a very small 








jy 




DE LAVAL 




BLISS 



Fig. 23 Typical Turbine Buckets 



thrust one way or the other. This thrust is usually taken care of by 
small thrust collars or washers next to the bearings. Thrust from 



282 



SMALL STEAM TURBINES 



the outside is prevented by the use of a flexible coupling between the 
turbine and the machine it drives. 

24 Bearings. The bearings are always ring-oiled with large oil 
reservoirs, sometimes, on the larger sizes, provided with water jackets 




Fig. 24 Section op Wilkinson Steam Turbine, 20-in. 



or water cooling pipes for an emergency cold-water circulation. The 
lubrication of the thrust is obtained at the same time. 

25 Operation. These machines are nearly automatic in their 
operation. When the machine is once properly set, the coupling 
properly adjusted and the bearings supplied with oil. the machine may 



SMALL STEAM TURBINES 



283 



run for years without an overhauling. The bearings must be looked 
after to see that no heating takes place and that the ring is carrying 
the oil to the shaft. The coupling should be examined from time to 
time to make sure that no thrust is communicated through it to the 
turbine. With these precautions a three months' continuous run is 
common and a number of turbines have to my knowledge run more 
than eighteen months without a cent spent on them for maintenance. 
Apparently there is no wear in nozzles, buckets, or return chambers 







\s^ 




y^ ^ 












i - ^^ -y^ 


ri^-^-s \->z,u3 




y\^ ' 


>'ater Rat^ 






.^y. 




'* 








.<^ 


-y^^ 

'^^0^^' 


,5.^ 

5^'^ 










,'-> 

.' 












^ 















1500 I 



30 40 50 

Brake Horse Power 



Fig. 25 Steaai Consumption Curvks, Terry Turbine 

24-IN. WHEEL, 150-LB. PEE88UKE, NO SUPERHEAT, NON-CONDENSING. TESTED BY WSSTINOHOT7BB 
MACHINE CO., PITTSBUKa, PA. 

The only wearing parts are the bearings and these are generously 
proportioned. 

26 These machines may be taken apart and reassembled in half 
a day; some of them in two hours. The over-hung machines may be 
overhauled in an even shorter time. 

27 'New Turbines. The Hachenberg turbine, made by Wm. 
Gardam & Son, New York, is a compound impulse turbine resem- 
bling in construction the Dow turbine so frequently illustrated twenty 
years ago. Some experimental machines have been built, one of 
which was tested at Columbia University, and the commercial ma- 
chine will soon be on the market. 

28 James Wilkinson, of Providence, R. I., has a small steam 
turbine nearly in the commercial stage. A number of these machines 
are running, and witliin the next few months it is expected they will 
be on the market. 



284 



SMALL STEAM TURBINES 




10 15 

Brake Horse Power 



Fig. 26 Steam Consumption Curves, Sturtevant Turbine 

20-IN. WHEEL, SINGLE-STAGE, NON-CONDENSING, 2400 R.P.M.] 

29 The Church turbine, lately completed by the Watson-Still- 
man Company and tested at Stevens Institute, is another promising 
turbine. 



T-5000-, 10000 




lOOOi 2000 



Brake Horse Power 



Fig. 27 Steam Consumption Curves, Bliss Turbine, Non-Condensing 



TESTED BY F. L. PRYOR AT HOBOKEN, N. f. 

O = Two-nozzle, X = Four-nozzle 



STEAM ECONOMY 



30 The curves of steam economy have in most cases been obtained 
from the manufacturers. For the Curtis turbine speed-economy 



SMALL STEAM TURBINES 



285 




10 



20 30 40 50 

Brake Horse Power 



60 



Fig. 28 Steam Consumption Curves, 50 h.p. Curtis Turbine 

ONE-PUESSDRE-STAOE, THREE ROWS OF BUCKETS, 25|-IN. WHEEL, CURVES CORRECTED TO 
150-I.B. BOILER PRESSURE, NO SUPERHEAT, ATMOSPHERIC EXHAUST 

curves are given for the 50 h.p. and 200 h.p. sizes. These curves 
represent the average of a large number of tests and have been cor- 
rected to bring them to standard conditions. The averages were 
consistent, and the variation from the average in any case was not 
large. 




ICO SCO 

Turbine 3rat3 Horse Power 

Fig. 29 Steam Consumption Curves, 200-h.p. Curtis Turbine 

THREE-STAGE, 36-IN. WHEEL, CORRECTED TO 165-LB. ABS. BOILER PRESSURE, NO SUPERHEAT, 
NON-CONDENSING 

31 The curves for the Terry turbine were plotted from fourteen 
tests made at East Pittsburg by the Westinghouse Machine Company. 
The curves for the Bliss turbine were plotted from twenty- four tests 



286 



SMALL STEAM TURBINES 



made at Stevens Institute by Prof. F. L. Pryor. The curves for the 
Kerr turbine were plotted from tests made by the Kerr Turbine Com- 
pany in their testing plant at Wellsville, N. Y. 




100 150 

Brake Horse Power 



Fig. 30 Steam Consumption Curves, 24-in. Kerr Turbine 

SIX-STAGE, CONDENSING, VARYING VACUUM, 70-LB. GAGE PRESSURE 




7000 



20 40 60 80 100 120 liO 160 180 230 

Brake Horse Power 

Fig. 31 Load Curves of Kerr Turbine 

24-IN. WHEEL, 8-STAGE 175-LB. GAGE PRESSURE, NON-CONDEN8IVQ 



32 There seems to be no change in steam economy use. It 
may be too early to make this statement, but machines running 
regularly for three years have shown no increase in steam consump- 
tion. 

33 The field of the small steam turbine is somewhat narrow when 



SMALL STEAM TURBINES 287 

compared with the high-speed steam engine. The small turbine has 
its place, however, and with the development of a more economical 
machine at the lower speed ranges, will have a much wider field. The 
turbine-driven centrifugal fan, for both high and low pressures, will 
have an increasing use, and the turbine-driven centrifugal pumps 
have marked advantages over reciprocating apparatus because of the 
absence of shock on the pipe line and their adaptation to space 
conditions. 

34 The promise of development on these lines has led many manu- 
facturers to enter the small-turbine field and the great expansion of 
the large-turbine business without doubt presages a like future for 
the small steam turbine. 



DISCUSSION 

Charles B. Rearick. Small turbines are being used extensively 
for hot-well service for surface condensers, the turbine driving a 
centrifugal pump which carries the condensed water away from the 
condenser, supplanting in these cases the usual reciprocating pump. 
They are more efficient than the reciprocating pump, they take less 
space, there are fewer parts to maintain, and the service seems to 
be very popular. They have also proved their worth in larger 
installations for driving boiler-feed pumps of the multistage turbine 
type. 

2 Some of the newest work taken up by turbine drive is for circu- 
lating pumps for surface and jet condenser work. We have a represent- 
ative lot of such installations using turbines from 50-h.p. to 250-h.p., 
operating such pumps under low heads and at speeds as low as 600 
r.p.m., direct-coupled to the rotor shaft of the turbines. 

3 The question of economy is not touched upon to any great 
extent in Mr. Orrok's paper, except to give some curves which cover 
only one condition of service in most cases and cannot very fairly 
be compared for the different makes. High economy in small tur- 
bine units is in many instances of minor importance. Reliability of 
service is most important of all. Under the very low speeds for 

The paper on Small Steam Turbines was discussed both at the Washington 
meeting, May 4-7, and at the Boston meeting, June 11, 1909. The discussion 
is here given in abstract only, eliminating the matter presented at the Boston 
meeting which duplicated that given at the Washington meeting. The com- 
plete discussion was published in the September 1909 issue of The Journal. 



288 DISCUSSION 

driving circulating pumps and similar pump work the economy can- 
not be especially good; but in nearly all these large power plants 
the exhaust steam is all utilized in feed-water heaters, and approxi- 
mately 80 per cent of the heat is returned to the boilers. 

4 There is only one class of service in which high economy is 
absolutely necessary, and that is, when the unit becomes the prime 
mover or the main unit for a plant. In that case the turbine, both 
condensing and non-condensing, compares well with the engine; for 
such work is usually driving dynamos and other high-speed apparatus 
and the speed can be chosen for the best economy. 

W. D. Forbes. Mr. Francis B. Stevens of Hoboken, who died a 
year ago in his ninety-fifth year, seeing in my shop some small Pelton 
water wheels, told me that in 1854 he had seen a steam turbine in a 
candy establishment in New York, which ran 1200 revolutions for 
some twelve years, with little or no attention. It drove a small 
fan. Mr. Stevens described the machine as practically the same as a 
Pelton water wheel, except that the bucket was cut in two, each half 
being placed on a separate disc, and the steam was led to these two 
buckets by a " splitter" between them, which of course was stationary. 
Each bucket was made fast to the disc, which was of course keyed to 
the shaft. 

2 What seems strange to me is, if steam turbines are such excellent 
things and have been known so long, that they are not more generally 
used. 

Richard H. Rice. The author describes the construction of 
seven different turbines, which may be divided into four classes ac- 
cording to the method of using steam, as follows: 

a Single-stage, expanding nozzle, one bucket row, one velocity 
extraction: De Laval. 

b Multistage, conveying nozzle, one bucket row per stage, 
one velocity extraction per stage: Kerr. 

c Single-stage, expanding nozzle, one bucket row, multiple- 
velocity extraction: Terry, Sturtevant, Bliss, Dake. 

d Single or multistage (depending on capacity), two to three 
bucket rows, mutiple velocity extraction but only one per 
bucket row: Curtis. 

2 The value of these various methods of using steam is clearly 
set forth in the curves presented by the author (Fig. 25 to Fig. 30), 



SMALL STEAM TURBINES 



289 



giving the water (steam) consumption of the various turbines in 
Classes b, c and d. In the^diagram in Fig. 1 all these curves are drawn 
to the same scale so that they may be readily compared. 

3 It will be seen that "with one exception, a very small machine, 
which suffers somewhat from this fact, the water-rates of all the tur- 
bines in Classes b and c fall rather closely together, while Class d^ 







































Citrvc 
A 
B 
C 

c' 

D 
E 
E' 


Type R.P.M. Rated H.P. 

Sturte^^tult 2,400 20 
Terry 2,350 60 
Bliss 2,000 :00 
2,600 200 
IveiT 2,800 150 
Curtis 3,000 50 
2,000 200 

Steam Press.- 150 Lb. 
Dry Steam 
Atmospheric Exhaust 
























80 
















70 

f-l 

w 


A 


^ 










^^ 




^ 




















Pi 

K 60 

w 














^ 
















B 

C 

— G- 

D 


\ 


^ 






















A 


i 50 

f 


^ 
























"v. 


^- 


^ 


^ 


::::::;:; 
















1 

I40 
1 


E 


"-^ 








^ 


-^ 












B 
C 


























^ 


D, 

c' 


30 


e' 





























E 




























e' 


20 






























i 


i 








i 






4 


i 






^. 


i 



Load 



Fio, 1 Economy Curves op Small Turbines 



even with a 50-h.p. machine, shows considerably better results. This 
is much more jaarked in the case of the 200-h.p. machine. Class b 
is represented by an eight-stage machine'of 150-h.p. and shows slightly 
better results than the machines of Class'c, particularly at light loads. 
This result is insignificant as compared with the complication of the 
large number of wheels, diaphragms, packings, and length of machine, 



290 



DISCUSSION 



and this complication is therefore apparently not justified. The rea- 
son such multistage machines do not give better waterratesisdue to 
a considerable extent to the high frictional losses caused by rotating 
wheels in a dense atmosphere of steam at comparatively high pres- 



sure. 



C2 
60 
58 
56 
54 




























L 


\ 






























^ Keciprocatiug Engine Tests 

A Sturtevant Turbine 

X Teir>- Turbine 

Q Bliss 

X Kerr " 

® Curtis " 


























































Pi 
n 

|48 

§ 46 

o 

1 44 

P 

|48 
140 
38 
36 
34 
32 
30 


























































\/ 
























































L 


] 






















(§)e 






f 


^ 




c 


] 
















(f 


5)3 


4 






m 












































^(^ 


















® 
















d 


lA 
















































i 





! 

1 



20 40 60 80 :00 120 140 160 180 200 230 
Rated Load B.H.P. 

Fig. 2 Comparison op Economy op Reciprocating Engines 
AND Turbines 



4 The turbines of Class c labor under two disadvantages, due to 
using the steam repeatedly in the same Ijuckets; either (as in the case 
of the Terry turbine) the steam has to be turned at high velocity 
through an angle of 180 deg. in the return chambers, or it has to be 



SMALL STEAM TURBINES 291 

used in buckets which have in general the wrong angle of entrance; 
for it is easy to see that if the bucket angle is correct for receiving 
the jet at its highest speed, it cannot be correct when the jet has con- 
siderably slowed down. 

5 The general principle employed in these turbines was used first 
by Professors Riedler and Stunipf in the years 1902-1903, and a 
number of these machines were built and put into service. The com- 
pany which exploited them, however^ abandoned the principle about 
the latter year and since then has built under a Curtis license, this 
step having been taken by reason ofthe superior economies obtainable 
by the Curtis construction, which the paper seems fully to confirm. 

6 At the Detroit meeting, June 1908, Messrs. Dean and Wood 
presented a paper giving results of tests on high-speed engines of sizes 
comparable with the turbine figures given by Mr. Orrok. Fig. 2 
shows the results compared with the full-load water rates given by 
the author. 

7 Messrs. Dean and Wood tested engines which had been in ser- 
vice for some time and many of which had evidently seriously deteri- 
orated in efficiency due to wear and leakage. Mr. Orrok confirms 
the statement made b}^ the writer, that the turbine does not fall off 
in efficiency after similar length of service and in further confirma- 
tion of this is the test on a 75-kw. Curtis turbine made by Professor 
Carpenter. 

8 This fact has also been established by many tests made by the 
writer on turbines which have been in use for considerable lengths of 
time, and is subject only to the qualification that when steel bucket 
constructions are used, as in all the turbines described except the 
Curtis, wear may be expected under certain conditions of wet steam 
and light loads which will increase steam consumption after a very 
moderate length of service. The use of bronze buckets of the proper 
composition to resist this deterioration is therefore essential to secure 
the best results under all conditions. 

Prof. R. C. Carpenter. During the past year I have given con- 
siderable study to the results from the use of small turbines, arriving 
at practically the conclusion of the author (Par. 33), that the field of 
the small turbine is somewhat narrow as compared with the high- 
speed steam engine. This conclusion applies to small turbines run- 
ning non-condensing, however; when large turbines are operated con- 
densing the economy is very high, and I think the results will usually 
be superior to those obtained with piston engines. 



292 



DISCUSSION 



2 On testing one turbine, which I think had been in use for three 
years, I was pleased to find that my results practically agreed with 
those obtained when the turbine was first installed. This seems tf' 
indicate that so far as that turbine is concerned, there was no deteri - 
oration from use. The general results which I obtained in the econ- 
omy tests were substantially those shown on these curves, andindica1(j 
that the econom}' is not good compared with the piston engine; the 
advantages of the small steam turbine must therefore be other than 
simply that of economy. The results of the tests of this machine aie 
shown in the table. 

3 I have recently had an opportunity of getting figures from a 
small turbine operated with a high degree of superheat and running 

TEST OF CURTIS TURBINE, 75-KW. CAPACITY, OPERATED NON-CONDENSlxNG 

October 13, 1908 



Test No 1 

Electric load, kilowatts ' 57 . 7 

Pressureat throttle, pounds gage 121. 

Pressure at nozzle, pounds gage 108.4 

Back pressure, pounds gage 0.24 

Barometer reading, inches 30 .0 

Total water to boiler, pounds 14971 

Wet steam to turbine, pounds 14121 

Quality of steam, per cent 98 . 3 

Dry steam to turbine, pounds 13881 

Dry steam to turbine per hour, pounds 3085 

Dry steam to turbine per kw-hr. , pounds 53 . 5 

Equivalent peri. h.p. (provided 1 kw. = 1.6 i.h.p.) i 33.5 




Note — The pressure at the throttle is practically the same as at the boiler, which stood about 
80 ft. away. 

non-condensing. The results of that test were satisfactory in many 
ways: 350 deg. superheat seemed to have aboutthe same effect as 18 
in. of vacuum, and a machine having a water rate given as approxi- 
mately 50 lb. per b.h.p. went down to about 22 lb. per l).h p. The 
small steam turbine has special advantages for many kinds of work 
where a high rotative speed and small torque are desirable; for those 
kinds of work I believe it will ultimately supersede the small piston 
engine. 



H. Y. Haden. a field for small turbines not touched by this excel- 
lent paper is that of installations where exhaust steam can be utilized 
to advantage. There is an installation in Pittsburg of a 150-h.p. 
turbine connected to centrifugal pumps, which operates under very 



SMALL STEAM TURBINES 



293 



unusual conditions. Primarily, it takes the exhaust of reciprocating 
pumps, without any regenerator, and develops the full power when 
exhausting into a vacuum of 25 in., but it is also capable of automatic- 
ally talcing high-pressure steam at 125-lb. pressure, should the supply 
of exhaust steam fluctuate too much or be entirely cut off. The same 
turbine also operates taking steam at 125-lb. pressure and exhausting 
into the atmosphere, and it had further to be guaranteed by the manu- 
facturers to take high-pressure steam at 65-lb. pressure when exhaust- 
ing freely. I believe the above four conditions could not be met by 
any reciprocating engine. Of course maximum economy will not be 
attained under each, but it is attained under two: that of exhaust- 
pressure condensing and high-pressure condensing. 

2 The turbine is of the De Laval type, which is particularly adapt- 
able for changing steam conditions, and is the only machine in which 
any desired plan of operation can be attained by simply changing nozzle 
ratios, without any change in the angle of the bucket or velocity of 
the pump, and without sacrificing either capacity or efficiency. I 
believe there is a large field for turbines operating under conditions 
such as the foregoing, whether connected to generators or centri- 
fugal pumps — a field where one can depend upon a unit irrespective 
of the supply of exhaust steam and without a regenerator or other 
expensive auxiliary. 

Fred. D. Herbert. Mr. Orrok did not mention tests made on the 
Terry turbine at the New York Edison Company's plant some years 
ago, in which the steam consumption is much below that shown in 
the Terry curve (Fig. 25). The accompanying curve and tables 
show the water rates of 50-h.p., 25-h.p. and 12-h.p. Terry turbines 
respectively. 



TABLE 1 TESTS OF A 25-H.P. TERRY TURBINE 



Test 


Steam 


Back 


Quality or 


Speed 


Load 


Per Cent 


\ 1 
ofTotal WaterWater per 


No. 


Pres. 


Pres. 


Superheat 


R.P.M. 


B.H.P. 


Rating 


PER HR. 


H.P.-HR. 








Degrees 












1 


90 





62.50 


2500 


25.52 


102.0 


1 1068 


41.85 


5 


90 





71.70 


2500 


18.60 


74.5 


851 


45.72 


2 


90 





82.42 


2500 


12.39 


49.5 


680.5 


54.93 


10 


90 





50.00 


2100 


24.29 


97.2 


1068 


44.00 


11 


90 





60.26 


2100 


15.816 


63.4 


808.25 


51.09 


6 


90 





45.60 


1800 


22.90 


92.0 


1077.25 


47.10 


7 


90 





79.86 


1800 


10.21 


40.8 


638.58 


62.54 


9 


90 





44.70 


1800 


15.69 


62.4 


807 


1 51.76 



294 



DISCUSSION 







TABLE 2 TESTS OF A 


12-H.P 


. TERRY TURBINE 




Test 


Steam 


Back Quality or 


Speed 


Load 


Per Cent oi 


Total Water Water per 


No. 


Pres. 


Pres. Superheat 


R.P.M. 


B.H.P. 


Rating 


PER HR. 


H. P.-HR. 






Degrees 












1 


136.5 


! 58 


2501 


13.65 


113.6 


577.3 


42.0 


2 


136.8 


15 


2513 1 


12.08 


100.7 


547.2 


, 45.3 


3 


128.5 


12 


2490 


11.70 


95.6 


545.2 


46.6 


4 


130.9 





2501 


7.23 


60.2 


421.5 


68.3 



X 3600 



SO •= 































































1 
















































































































































































































































































































































































-+-1 




















































































^ 














































































































































































-^ 


















































































^^>^ 


^ 






















































































A 


















































































K0- 


^ 






























































^ 






















Back Pressure .68 ^ 
Superheat 60° 


- 










\ 




































































N 




















J 


































R 


P,M. 2500 


























1 


K>' 




















































































.J 


















































































s 






^1 


















































































^ 


\ 














































































^ 








V 


























































































\ 




























































































\ 
































































































k 




























































































' 
































































































> 


































































































'^0, 


























































































1^ 


J 
























































































1 






^ 




P 


























































































>J 




u 






























































































~-^ 


P^ 


































































































-- 


- 
































































































"■ 


— — 




__ 

















































































































































































































































































Fig. 1 Curves of 90 h. p. Terry Tuhbine 

2 Regarding the statement that turbines are built of the single 
stage only, the Terry Company has in operation and under construc- 
tion several two-stage condensing turbines, the largest of which is 600 
h.p. and the smallest about 125-h.p. Because of the successful re- 
sults obtained, the builders claim that the turbine is superior to the 
reciprocating engine as an operating proposition, and in none of these 
turbines now running has a bucket been replaced. 



W. E. Snyder. Almost all the emphasis has been laid on steam 
economy. Another point which should receive careful consideration 
is the lower cost of maintenance, particularly where turbines are used 
for boiler feed, replacing the direct-acting boiler-feed pumps ordinarily 
used; for pumping circulating water to condensers; or for driving cen- 
trifugal pumps pumping water at comparatively low heads. In all 
this work the turbine replaces direct-acting pumps which are very 



SMALL STEAM TURBINES 295 

expensive to operate, not only from the standpoint of steam-consump- 
tion, but from the cost of supi^hes, such as cyUnder oil, pacldng and 
pump valves. I have in mind one central condensing plant served by 
a direct-acting pump where the costs are from $75 to SlOO a month for 
packing, cylinder oil and valves. A similar condensing plant served by 
a small turbine involved practically no expense for these supplies. 

2 The first question, three or four years ago, before our company 
had installed any of these machines, was not a question of steam 
economy so much as of reliability of operation. If we put in a tur- 
bine to pump water for a central condensing plant, into which exhaust 
a large number of steam engines of various kinds used for varied 
service, would the turbine break down, just when it was most needed? 
For the purpose of investigating this point I went to a plant which 
had a small unit in operation driving a generator and which had been 
in use for about four years. The gears showed no appreciable wear, 
and there had been practically no shutdowns. 

3 The result of that investigation was the adoption of turbines in 
central condensing plants in a number of works with which I have 
to do, and also later for boiler feed and for pumping water under low 
heads; and they have proved generally reUable regardless of make. 
This is also true in regard to small turbine air-compressor units for 
the cupola, etc., in steel works. 

4 I think it is, therefore, in the displacement of the direct-acting 
pump, always expensive to operate apart from steam consumption, 
that the small tiu-bine will find its greatest field of usefulness. In the 
Waterside Station in New York, the turbine-driven boiler feed 
pumps have been continuously operating for months, running almost 
automatically, and requiring practically no attention or supplies. 
Steam economies are of course important, yet in large plants where all 
of the large units are condensing, the steam from the auxiliaries is 
needed to heat the feed-water, and a few per cent more or less in 
steam consumption of turbine auxiliaries does not materially change 
the total economy of the plant. 

5 Other advantages in favor of the small turbine as compared 
to direct-acting pumps, are the small'space required and the fact that 
they can always be kept clean and present a good appearance. 
Direct-acting pumps are usually'very difficult to keep in presentable 
condition on account of water leakage and of the excessive use of lubri- 
cants. The turbine and pump are entirely enclosed, the case can be 
wiped very conveniently, and it presents nothing of the unsightly 
appearance which is so often characteristic of the direct-acting 
pump. 



296 DISCUSSION 

F. H, Ball. The conclusions of the author regarding the future 
of the small steam turbine may fairly be questioned. On the score 
of efficiency the showing made by the several types, even when tested 
by the parties interested in their success, is very poor. From these 
test figures, it appears that the best performance ranges from about 
30 lb. per h.p. per hr. to nearly 70 lb. 

2 It must be noted also that very high steam pressure is generally 
used, and in some cases superheat. Under these conditions any 
good reciprocating engine, even of the single-valve type, run as a non- 
condensing compound, would easily develop power on 20 lb. or less: 
the user of one of the non-condensing turbines described must 
therefore expect to increase his coal bill from 50 to 200 per cent over 
a single-valve non-condensing compound engine having the simplest 
possible form of valve gear. 

3 Moreover, the speed of these turbines, from 2000 to3600r.p.m., 
will generally be considered objectionably liigh. Buyers of electric 
motors generally prefer motors of moderate speed, even at the extra 
cost, and generally speeds above 1000 r.p.m., even for motors as small 
as 10 h.p., are considered objectionable. This same objection must 
inevitably be urged against speeds of 2000 to 3600 for engines of con- 
siderable power. 

CuAS. A. Howard, As far as a comparison of the merits of differ- 
ent turbines goes, it must be remembered that the economy is affected 
by the bucket speed even more than by steam pressure. In all of 
these tests the bucket speeds are different, and any attempt to make a 
comparison of the steam economy, as in the diagram by Mr. Rice, 
would thus lead only to an erroneous view. While his curves 
show in general what may be expected from turbines of this size, 
no correct comparisons can be drawn between individual machines. 

W. J. A. London.* With reference to Richard H. Rice's compari- 
son of Fig. 25 and Fig. 28, showing the steam consumption of the 
Terry turbine and that of the Curtis turbine, if the curve of the 
Curtis turbine be produced and the peripheral speed of the two 
types be made the same, the curve of the Terry turbine will cross 
that of the Curtis type at about 1950 r.p.m. See Fig. 1. There is 
therefore not much room for discussion of the difference of efficiency 
of the two types. Moreover, with a large turbine an increase of a 
pound on the steam consumption would increase the cost bill from 

* Terry Steam Turbine Co., Hartford, Conn. 



SMALL STEAM TURBINES 



297 



$2000 to $4000 a year, but with a small turbine it would mean an 
increase of only from $10 to $25 a year, which would be offset by the 
difference in first cost. 

2 The greatest value of the paper lies in the fact that, better than a 
salesman, it shows to men having reciprocating engines, the great 
simplicity of construction of the turbine.- Men famihar with recipro- 
cating engines know what to do in case of breakdown, but with a 



50 
i8 
IC 
U 
12 
40 
38 
3G 
3i 
32 




\ 






































X 


\ 










































N 












o Indicates Curtis Turbine 
*■ " Terry " 


















^ 


\ 


























\ 


^ 








































N 










































\ 


k 










































\ 












































^ 


k. 










































"-- 


-^ 














































" — 






30 








Carve A 17C 

1 1 


























^ 



17 18 19 2W0 21 22 23 21 



20 27 28 39 3000 31 32 33 31 35 36 



Fig. 1. Comparison op Tests in Fig. 25 and Fig. 28 Reduced to the 
Same Wheel Velocity. 

turbine a failure means a shutdown for several days. In a few years, 
however, every engineer will thoroughly understand the construction 
of a steam turbine and will be able to make his own repairs. Another 
fact which hindered the more frequent use of the steam turbine was 
that generators, pumps, blowers and other machines had to be 
designed specially for operation with turbines. That this is now 
being done is an acknowledgment that the steam turbine is here to 
stay. 



Chas. B. Rearick. The matter of speed for turbines driving 
centrifugal pumps is often a compromise, as the pump speeds 
are not always ideal for the turbine. Especially is this true for circu- 
lating work where the heads are often only 15 or 20 ft. and the deUvery 
20,000 gal. or more per min. It may be necessary to sacrifice some 
efficiency of the pump in order to run at a speed suited to the turbine. 
In such instances all the exhaust steam may be used to advantage 
in heating feed water, while the low cost of^operation and the saving 
in oil and supplies will overcome the cost of increased steam con- 



298 DISCUSSION 

sumption. When efficiency is of importance, as in isolated dynamo 
work, the speeds are usually such as to give results quite as good 
as those of the reciprocating engine, and in many cases better. 
The turbine has demonstrated its ability to give economical results 
in all cases where the speeds ire favorable. 

2 Regarding the rating of turbines, there is only one point of 
maximum efficiency of any design, so far as I loiow, and that is its 
maximum load. It is very similar to a gas engine in that respect. 
If we want maximum efficiency the turbine must carry its maximum 
load. That can be brought about in some cases by the nozzle system 
of design in which better economy is obtained at light loads by shut- 
ting off nozzles. But where t"bis is done by hand regulation there is 
always danger of the load coming on without notice or without the 
engineer having opened any of the hand adjustments. Turbines 
should therefore be designed to eliminate hand regulation, and to 
accomplish this some builders provide for automatic operation of 
these valves. While this is successful within a limited degree, these 
valves may become leaky in service, and when once leaky the control 
of the turbine is beyond the operator, which may result in over-speed- 
ing to a bursting point in case the load is suddenly thrown off. 

3 It follows that the number of controlling valves should be 
reduced to a minimum. If there is only one valve to control in a 
machine there is but one valve to look after and to keep tight. Few 
operators of turbines appreciate how serious the leakage of the con- 
trolling valve is to the proper governing of the turbine, especially on 
very light loads or no load. On the other hand, if turbines are kept 
well loaded these leaky valves are not noticed and as a result steam 
consumption is increased through their use rather than diminished, 
unless all the valves are wide open and the turbine is working up to 
its full capacity. The turbine which eliminates these dangers, it 
would seem, is the better machine, and in small units the difference 
in economy is entirely outweighed by the complications and dangers 
above cited. 

F. B. DowsT. The B. F. Sturtevant Company have for years built 
reciprocating engines — single engines, multiple-single engines, and 
multistage engines. Later we built direct-current and alternating- 
current motors. We first came into touch with the steam turbine 
principle in 1883 when our attention was first called to the Wise steam 
motor. This motor was an impulse wheel, with four jets, I think, 
the steam impinging on buckets, no endeavor being made to expand 



SMALL STEAM TURBINES 299 

the steam in nozzles. One of our engineers left us at that time to 
exploit the Wise steam motor. He returned after a year's sad experi- 
ence in connection with the amount of steam that would flow through 
a small opening. 

2 Our next experience was with the Dow steam motor. Mr. 
Dow came to us early in the nineties, I think it was, backed by Mr. 
Chisholm of the Chisholm Shovel Works of Cleveland, O. This tur- 
bine had previously been developed and used successfully to drive 
the flywheel in the Howell torpedo. Associated with Mr. Dow was 
Mr. Howard, for some time connected with the Fore River Ship and 
Engine Company, as it was then called. The Dow turbine was built 
in the Sturtevant works and was really a meritorious machine. It 
was what might be called an inward-flow reaction turbine. A motor 
of bronze was built and the method which Mr. Dow developed is now 
used in balancing our rotors. 

3 A little later, during the development of the Curtis turbine, 
a representative of Mr. Curtis came to us for journals for high-speed 
work. He had trouble in finding a journal box capable of withstand- 
ing the high speed of his rotor shaft. A box that we used was very 
successful in solving the problem. We knew there was a somewhat 
limited field for high-speed motors for use with our fans, however, 
and were not quite ready to take up the Curtis turbine commercially. 

4 A completed Dow turbine was frequently connected with one of 
our No. 6 blowers. The governing device was not developed, but 
that was not necessary in order to connect the turbine with a fan. 
I believe this turbine was tested in our works, and afterwards tested 
at the Massachusetts Institute of Technology for water consumption, 
which was found to be high. I have always thought that the Dow 
turbine possessed great possibilities and wondered why someone did 
not develop it. 

5 Experiments with a steam turbine of our own for use with our 
fans resulted in the turbine described as the Sturtevant turbine. Mr. 
Orrok's description of the rotor is substantially correct, except that 
he omitted the fact that in the manufacture of this part we use an 
open-hearth steel forging of the best quality. Among the uses of the 
turbine are direct connection with generators, with fans for blowing 
blast furnaces, and with multivane fans for work on shipboard. Four 
fans with geared connection, recently built for a heating system in 
the West, have done good work. 

6 An interesting problem was recently presented when the Navy 
Department insisted on fans for forced draft for the new torpedo boats 



300 DISCUSSION 

where oil fuel is to be used. This of course demanded a slow-speed 
turbine, but we think we have worked out a satisfactory combination 
by effecting a compromise between the fan and the turbine element. 
It is interesting to recall that a few years ago the Navy Department 
did not consider any motive power except a reciprocating engine. 
Later the electric motor came into use and now many of the new ships 
are equipped with forced -draft fans driven by electric motors. 

7 There is not the slightest doubt that the general type of turbine 
discussed in Mr. Orrok's paper is here to stay. Engineers like it and 
engine builders must get ready to furnish turbine engines. 

Chas. B. Edwards.^ We are more particularly interested in the 
development of the large marine turbine, but our attention has been 
recently directed to the smaller turbines owing to the Navy Depart- 
ment specifying them for blowers; there is also a possibility of their 
use for circulating pumps and other auxihary machinery on board 
ship.. The Navy Department has increased the steam allowance for 
turbine installations over what it was a few years ago when it was 
limited to 50 lb. per h.p. hour. 

2 The great problem in marine installations, particularly for 
naval purposes, is that of weight. The navy contracts specify a cer- 
tain weight of machine and if we exceed that weight we must pay for 
it at the rate of about $500 a ton. In considering the turbine propo- 
sition, therefore, we must look at it not only from the mechanical side 
but also from the standpoint of weight. One of the difficulties, of 
course, is that of the exhaust. The weight of piping, fittings, valves, 
etc., runs up rapidly and it is therefore desirable that turbine auxiliaries 
should be placed as close to the condenser as possible; and in order 
to secure economical results it is also desirable to secure a low veloc- 
ity of exhaust in the pipe lines. 

V. F. HoLMES.2 The DeLaval Company has recently brought out a 
combination high-and-low-pressure steam turbine. Many plants 
where condensing water is available have an excess of exhaust steam 
from auxiliaries and the question has arisen whether a machine could 
not be devised for this class of service. That would necessitate stor- 
ing up energy in times of an excess of exhaust steam to carry the 
macliine over the periods of limited exhaust steam, and would involve 
expense and complications. What is desirable is a machine in which 

* Chief Engineer, Fore River Shipbuilding Co., Quincy, Mass. 
'Power Equipment Company, Boston, Mass. 



SMALL STEAM TURBINES 301 

both the exhaust and the Hve steam can be used economically with- 
out regenerators and other heat-storing devices. 

2 The DeLaval combination high-and-low-pressure turbine is 
built with two nozzle compartments, one for high pressure and the 
other for low pressure. Each compartment is furnished with nozzles 
having the proper ratio of expansion for the conditions under which 
they operate. Some of the nozzles are furnished with shut-off 
valves for regulation under variable conditions. 

3 Two steam connections are provided, one for high -pressure 
steam and the other for low-pressure steam, each connection leading 
to its own governor valve, which in turn is operated by a separate 
governor. The operation is entirely automatic, the low-pressure 
governor being set for a speed slightly higher than that of the high- 
pressure governor. On the total or partial failure of the low-pressure 
steam supply the machine will automatically draw from the high- 
pressure steam supply the steam necessary to make up the deficiency. 
Also in case of the complete failure of the low-pressure steam sup- 
ply, the machine will operate on high-pressure steam, and under this 
condition will give practically the same economy as a high-pressuro 
steam turbine. 

4 The combination high-and-low-pressure turbine is built for 
conditions where continuous operation is essential and where the 
supply of low-pressure steam is intermittent or is apt to fail com- 
pletely. The regulation when changing from cue steam pressure to 
the other varies from 2 to 3 per cent, this being on an instantaneous 
change from one condition to the other, such as seldom occurs in 
actual service. A by-pass valve allows the admission of high-pres- 
sure steam into the low-pressure compartment, for operation under 
full-load conditions non- condensing. This by-pass valve is not 
automatic, and is simply to enable the machine to carry full load in 
case of failure of, or during repairs to, the condensing apparatus. 

5 Both the low-pressure turbine and the combination high-and- 
low pressure turbine are built for steam conditions varying from 5 
lb. pressme above atmosphere to 10 in. of vacuum at the steam inlet. 
They are also built for low vacuums for conditions where the temper- 
ature of the circulating water or existing condensers prohibits the 
maintenance of a high vacuum. The steam consumption of the 
machine varies somewhat with the sizes and operating conditions; 
the average machine operating with steam at atmospheric pressure 
and exhausting into a vacuum of 27 in. to 27^^ in. will use from 28 
lb. to 32 lb. of steam per b.h.p-hour. 



302 DISCUSSION 

6 The DeLaval Company is also building a high-speed, low-pres- 
sure turbine particularly adapted for direct connection to centrifugal 
pumps and blowers. This class of machine is built in both the low- 
pressure and combination high-and-low-pressure types, and consists 
of the DeLaval wheel direct-connected to the machinery to be driven. 
On account of the direct connection of the wheel and the elimination 
of the usual DeLaval reduction the machine can be economically 
operated only at high speed, and for this reason is not suited to direct- 
current generator work, but is particularly adapted for high-speed 
pumping and blower work, such as power-plant auxiliaries, boiler- 
feed pumps, elevator pumps, etc. 

J. S. ScHUMAKER. An error has crept into these figures that I am 
sure was not intended. That is, the figures given for the economy 
of the Terry steam turbine were obtained from a turbine with nozzles 
designed for 100 lb. pressure. But the steam pressure used on the 
test was, I believe, 150 lb. One other point that may in fairness be 
brought out is that the Terry turbine tests as offered here were made 
without representatives of the Terry Steam Turbine Company being 
present, while in the majority of the other cases cited the tests are 
shop tests. 

Prof. Carleton A. Read. I am interested from the fuel side of 
the question in the use of non-condensing turbines in small manu- 
facturing plants of from 75-kw. to 300-kw. capacity, where there is 
an excess of exhaust that can be used only for feed-water heating and 
heating the buildings in cold weather. We all agree that it is well 
not to have oil in the exhaust if the condensation is to return to the 
boilers, but man}^ plants have a good and cheap water supply and 
after using as much of their exhaust as possible still have some going 
to waste. Nearly all of the tests quoted are from the manufacturers 
and without doubt are correct for the conditions under which they 
were made, but data as to coal consumption under actual working 
conditions would be of interest to the man buying an equipment for a 
small plant. 

Prof. Ira N. Hollis. One aspect of the subject impresses me as 
important. The curves of efficiency used for comparing different 
turbines relate particularly to the thermodynamic efficiency of the 
machine or the number of pounds of steam per horsepower developed. 
It seems to me that where a steam turbine is connected with a pump, 



SMALL STEAM TURBINES 303 

such as one used for feeding a boiler or for circulating water in a con- 
denser, the machine ought to be treated as a whole. From this point 
of view, the number of gallons of water delivered per pound of steam 
or per pound of coal is an important factor and should be given in 
every case. Naturally the pressure against which the water is pumped 
is another factor. Ordinary reciprocating engines driving feed 
pumps are very uneconomical machines. I have had experience with 
pumps that used 100 lb. of steam per i.h.p. or even more. However, 
the efficiency of the pumps as a whole for delivering water into a 
boiler was never worked out. 

2 It may be that the steam turbine is to replace the steam engine 
for all purposes about a power station, particularly if the high-pres- 
sure centrifugal pump can be developed into a highly efficient machine 
in connection with the turbine. It seems to me, therefore that it 
would be useful in connection with all tests of turbines used to drive 
pumps, to give the combined efficiency of the unit. 

Prof. Edw. F. Miller. In looking through these figures of steam 
economies it will be noticed that the greater the load the smaller the 
amount of steam per horsepower. All the turbines I have had to 
do with would stand considerable overload, in some cases 80 per 
cent. I would like to know what decides the maker in rating his 
turbine. Apparently the economy line runs down as the overload 
goes on. Why not rate the turbine higher and get better economy? 

John T. Hawkins. I was a pretty old engineer when the turbine 
was born and consequently know little about it except what I have 
learned by reading and observation. I am not going to try to impart 
information but I wish to ask a question. It seems to be a well- 
known fact that with the turbine engine, the higher the load the greater 
the efficiency within its limits. To what is the fact due that the 
turbine is more efficient under high load ? 

Richard H. Rice. Just a few words in explanation of why the 
turbine water rates decrease as the load goes up and of the effect 
of the various governing mechanisms on this action. The impulse 
turbine is essentially a turbine of partial admission. In a multi- 
stage condensing turbine of this type the buckets in the last stage 
are usually designed of the right height and the nozzles of the right 
proportion to use the entire circumference of the wheel. In a four- 
stage machine, the next to the last stage will use perhaps one-half 



304 DISCUSSION 

the circumference of the bucket wheel for steam admission. It 
could be designed to use all the circumference, but that would 
involve undue shortening of buckets. In the second stage there is 
a further shortening of the arc of steam flow, and in the first stage, 
a still shorter arc is used, perhaps 90 deg. 

2 In non-condensing turbines, if we were to attempt to use the 
entire circumference of the wheel the buckets would be so small 
that the machine would be impracticable and inefficient. We must 
therefore use a short arc, decreasing the cost of the governing 
mechanism and making a reasonable bucket speed possible. It is 
evident that the bucket speed must depend on the size of the machine 
and that, in connection with the operating speed, it is the prime 
consideration in the cost of the machine. It would be a mistake to 
make a 25-kw. machine with the same bucket speed as a 300-kw. 
machine, because the former would be so large in diameter and so 
expensive as to be impracticable. 

3 It follows that one reason why the turbine increases in economy 
as the load increases is that a larger circumference of bucket wheel 
is used; a smaller percentage of the total power is wasted and there- 
fore efficiency increases. Therefore, if the governing mechanism 
works by throttling we have this condition : the steam pressure and 
area of the nozzle system determine the amount of steam that can 
be used in the turbine. In a machine with nozzles wide open, the 
latter must be so designed that the turbine will carry maximum 
load, as otherwise the turbine would shut down at maximum load. 
It follows that nozzles designed for full pressure at maximum load 
will be greatly throttled when running with light load, and conse- 
quently the efficiency will decrease. Therefore it is advisable to 
govern the nozzle system in such a way that nozzles can be designed 
for full boiler pressure. By using a larger or smaller number of 
nozzles, and hence a larger or smaller arc of wheel, the full economy 
of the nozzles is obtained and only the proper number of nozzles are 
open for a given load. 

Chas. H. Manning. The diagrams confirm the opinion I had 
formed of the small steam turbine to the effect that it is a steam 
thief. But that its virtues will outclass its sins I am thoroughly 
convinced. Recent developments of high-speed centrifugal pumps, 
fans and generators open a field for the turbine in which it is sure to 
succeed. 

2 A small practical point is that almost all of these small high- 



SMALL STEAM TURBINES 305 

speed machines use the ring oiler, which is in general bad practice. It 
has a very small contact on the shaft and any small thing will stop 
its running. Furthermore, the rings frequently break. If for the 
ordinary ring oiler a chain with a lai'ge arc of contact is substituted, 
preferably a window-cord chain, it will never fail and will bring up 
ten times as much oil as a ring oiler. While this is a small point, 
any machine depends more on the perfection of its detail than it 
does on the theory on which it is built. 

C. P. Crissey.^ There is one type of the small turbine to which 
the author has given scant space; that is, the small condensing 
machine. While, perhaps, the field is not so wide for this type as it 
is for small turbines exhausting at or above atmosphere, it cannot 
be ignored. Practically all marine work requires condensing prime 
movers, and many small stations use this type. Only one example 
of a condensing machine is referred to by the author, the results of 
tests being shown in Fig. 30. It would be a mistake to consider this 
curve as representative of small condensing turbines. 

2 Small turbines as well as large derive great benefit in economy 
from high vacuum, and a vacuum of 28 in. is easily obtained on small 
machines of proper design. In a well-designed small turbine the 
vacuum shows no greater tendency to fall with the increase of load 
than in large machines. Why the Kerr turbine show^s a loss in 
vacuum as the load and hence steam flow increase, we are unable to 
tell definitely from this paper. It will be noted that the steam is 
discharged from the buckets on each side of the wheels. It is there- 
fore necessary for one-half of the total flow to pass about the wheels 
in order to reach the succeeding nozzles. Excessive velocities and 
throttling will occur in the low-pressure stages where the volumes 
encountered are great, unless large areas are provided for this steam. 
I understand that in the Kerr turbine this throttling is obviated as 
much as possible by providing holes in the wheels. These holes, how- 
ever, increase the windage loss. 

3 One of the reasons for abandoning the Riedler-Stumpf turbine 
in Germany was the inability of its buckets to handle large volumes 
of steam successfully. The same objection holds against all machines 
of the Riedler-Stumpf type. 

4 The only small turbines having buckets capable of caring 
efficiently for large volumes of low-pressure steam are the De Laval 
and the Curtis types. The DeLaval turbine is seriously handicapped 

' General Electric Co., West Lynn, Mass. 



306 DISCUSSION 

by its high rotative speed, while the Curtis turbino, due to its pressure 
and velocity stages, is capable of moderate speeds. In order to show 
that the results of Fig. 30 are not typical of all small condensing tur- 
bines, I will say that Curtis condensing turbines of from 100 to 200 
h.p. give economies of 18.5 to 15.5 lb. of steam per b.h.p. hour when 
operating with 150 lb. dry steam and 28 in. vacuum. 

5 Regarding the curves of this paper, I believe they should be 
compared at rated speed, because the bucket angles are designed for 
this speed. The rated speed may be taken as the maximum stated 
upon the curves. 

W. J. A. London. I wish to add ^ something to my remarks 
in connection with the curve mentioned by me earlier in the discus- 
sion; namely, a comparison between Fig. 25 and Fig. 28. In making 
this curve, I used only absolute tests according to the figures men- 
tioned and made no deductions whatever except in the question 
of relative peripheral speeds. If the curves plotted in Fig. 28 are 
reproduced for a series of full-load points on a speed basis, a positive 
curve will be formed. On this curve is plotted the two full-load 
tests shown in Fig. 25. Now, then, as there are only two points 
given in the Terry tests, it is impossible from these tests to obtain the 
nature of the curve, but the point I particularly wished to bring 
forward was that these two points practically coincide, — one test, 
as a matter of fact, being better and one worse, — both of them be- 
ing so near the Curtis curve as to make little difference. They are not 
so far away as Mr. Rice would have us believe from his diagram. 

2 The point has also been raised as to whether emergency gover- 
nors were fitted on other makes of turbines besides the Curtis. Par- 
ticularly with the Terry turbine and I beheve with the majority of 
the other makes, an emergency governor is not provided for the 
reason that a positive type of governor is fitted on the main shaft. 
The worst that can happen is the breaking of a spring, which would 
immediately close the valve. With a governor driven by a gear 
shaft an emergency governor is more essential for the reason that 
the gears are likely to break; hence some form of governor is used 
on the main shaft. If the direct-connected governor on the main 
shaft is likely to get out of order, why then is the emergency governor 
not likely to get out of order when placed in the same position? Up 
to the present time the Terry Turbine Co. has not had a machine 
burst, and with the type of governor adopted and the speeds em- 



SMALL STEAM TURBINES 307 

ployed, the designers consider an emergency governor an unneces- 
sary luxury. 

R. H. Rice. Mr. London claims that the steam consumption of 
the Terry turbine is the same as that of the Curtis turbine, when 
operating the Terry turbine at its designed speed and reducing the 
speed of the Curtis turbine to two-thirds of its designed speed of 
3600 r.p.m. The inaccuracy of this comparison can be readily 
understood when it is known that the angles of the buckets in the 
Curtis turbine would be radically changed if designed to run at two- 
thirds of the present rated speed. 

2 In discussing emergency governors, it must be realized that 
we are dealing with much higher speeds than those usual with recip- 
rocating engines. It has been found best in many plants to install 
emergency governors on reciprocating engines. If this is desirable 
on slow-speed apparatus, how much more desirable, and even neces- 
sary, is it on high-speed apparatus like steam turbines. Many other 
accidents besides the breaking of a spring can happen to a positive 
type of governor fitted to the main shaft, and any one of these is suffi- 
cient to cause a dangerous increase in speed of the turbine. An 
emergency governor can be made to possess the utmost certainty 
and reliability of action, since its function is to shut down a machine 
and not to regulate its speed. 

J. H. LiBBEY. The applications of small steam turbines men- 
tioned by the author, except for driving high-pressure fans, refer to 
uses with auxihary apparatus in a central power station. For this 
purpose a small steam turbine must be considered in competition 
with a reciprocating engine, and in general the choice will be decided 
by the following considerations: First cost, attendance required, 
maintenance and repairs, space, economy and influence on design 
of power station. 

2 First Cost. At present, when the service permits operation at 
speeds approximating those for which the turbine was designed, the 
cost of the turbine is somewhat lower than that of a corresponding 
reciprocating engine. 

3 Attendance Required. The attendance required for a small 
steam turbine is less than that required for any other type of steam 
machinery. It approaches very closely that required for an electric 
motor. 

4 Maintenance and Repairs. In Par. 32, the author indicates 



308 DISCUSSION 

that small turbines have been running for only three years. Ob- 
viously in this time no data of great value could be obtained to enable 
a decision to be made in regard to maintenance and repairs. The 
evidence, however, strongly indicates that they will be materially 
less than for a reciprocating engine. 

5 Space. Steam-turbine-driven apparatus is generally charac- 
terized by the small space required. In a great many cases', a tur- 
bine unit can be installed where a reciprocating engine would be 
impossible. 

6 Economy. An inspection of Fig. 28 and Fig. 29 shows that for 
the best conditions a turbine can deliver a horsepower with as little 
steam as, or less steam than, the same size reciprocating engine. In 
installations where the conditions are not favorable, the economy is 
reduced. Unfavorable conditions for a steam turbine are low super- 
heat, low steam pressure, high back pressure or reduced speed of the 
turbine, on account of the characteristics of the driven machine. 
The last condition is the most likely to cause reduction in the economy. 

7 It should be borne in mind that in the ordinary large central 
station where fairly large generating units are installed, the steam con- 
sumption of the auxiliaries does not in general amount to more than 
10 per cent of that of the main generating units. In such cases, the 
auxiliary exhaust will heat the feed water to about 175 or 180 deg. 
fahr. A considerable increase of steam consumption of the auxil- 
iaries can be permitted before there is sufficient exhaust to heat the 
feed water to 212 deg. fahr. In most cases, therefore, the steam 
consumption of these small auxiliaries is a matter of secondary con- 
sideration. 

8 In the various auxiliaries generally used the inherent conditions 
which would affect the steam consumption would be in general as 
follows: 

Exciter, favorable- 
Circulating pump, speed low for best results. 
Hot- well pump, favorable. 
Forced or induced-draft fans, speed low for best results; special 

design of fan required. 
Feed pump, speed low; however, the steam consumption of 

a turbine-driven multistage centrifugal feed pump is much 

lower than that of a reciprocating pump of the same 

capacity. 

9 In this connection Mr. Orrok's statement in Par. 32, that there 



SMALL STEAM TURBINES 309 

seems to be no change in steam use with length of service is of impor- 
tance as it is well known that the steam consumption of engines or 
pumps increases greatly with wear of valves, rings, pistons, cyhnders, 
etc. 

10 Many central stations are toda}^ operating the auxiliaries 
with superheated steam. Very few changes are required in the 
structure of a steam turbine to adapt it to superheated steam by the 
use of which the economy is improved. The reciprocating engine 
gains in economy from superheat, but greater changes in the design 
are required to obtain satisfactoi-y operation, and the expense of the 
engine is therefore increased. 

11 Influence of Design on Power Station. Turbine-driven exciters 
are generally light in weight and compact. They can be set on the 
engine-room floor without a heavy foundation or resulting vibration. 

12 Circulating-pump units are Of simple design. In many cases, 
a combination of auxiliaries may often be effected. There is on the 
market a jet condenser, the centrifugal pump and air pump of which 
are on the same shaft with the turbine. When in surface-condenser 
work the conditions are such that the speed of the circulating pump 
is subject to little variation, the turbine, circulating pump, and hot- 
well pump can be mounted on the same shaft. One manufacturer 
is prepared to add a rotary vacuum pump to these, either direct- 
connected or chain-driven. This arrangement gives practically one 
auxiliary for a surface condenser in place of three. 

13 Future Designs. The small steam turbine has sufficiently justi- 
fied its existence. The future will undoubtedly show types with 
improved economy, especially at reduced speeds, simplicity of design, 
rugged characteristics, ability to operate without attention, interior 
construction that is easily accessible and such that few repairs due 
to wear are required. 

The Author. The author is greatly pleased with the reception 
accorded his paper and the amount of discussion which it brought 
out. He must take exception to a comparison of water rates plotted 
on percentages of load as abscissae, and a new diagram has been pre- 
pared. Fig. 1, showing the results of all the water-rate curves plotted 
with bucket speed or peripheral velocity as abscissae, obviously a 
much better measure of the performance of these machines. The 
author would like to take up the question of improper entrance and 
discharge bucket angles in machines of the Riedler-Stumpf type, 
as well as the fluid friction question, both mentioned in the discussion 



310 



DISCUSSION 



of Mr. Rice ; but these should be the subject of a mathematical paper and 
are not of serious importance in a small turbine. The author feels 
that Mr. Ball has failed to grasp the fact that with small rotating 
masses speeds of from 600 to 3000 r.p.m. are not as objectionable as 
a speed of 150 r.p.m. in a modern four-valve engine, or 100 double 
strokes per minute in a direct-acting pump. 

2 Replying to Prof. Hollis' discussion, the author knows of many 
power plants where entire reliance is placed on turbine-driven multi- 




10,000 15,000 ^,000 

Peripheral Speed —Ft. per iiiiii. 



25,000 



Fig. 1 Steam Consumption of Small Turbines Plotted with Peripheral 

Speeds as Abscissae 



stage centrifugal pumps for feed-water service. He knows of no 
case where an attempt has been made to find the coal consumption 
of the feed pumps directly; in other words, the duty of the pumps. 
It has usually been obtained through the steam consumption with a 
knowledge of the evaporation constant of the plant. The use of 
Venturi meters in the feed lines and in the steam connections to the 
turbine-driven feed pumps would give this duty directly, and a partial 
installation of this nature has been made at the Waterside Station of 
the New York Edison Company. 



No. 1243. 

TESTS UPON COMPRESSED AIR PUMPING 
SYSTEMS OF OIL WELLS 

By Edmund M. Ivbns, New Orleans, La. 
Junior Member of the Society 

When the Louisiana oil fields at Evangeline were in full operation, 
they offered exceptional opportunities for the study of air lifts. 
Nearly every known method of piping the wells was in use. The 
air plants originally installed were the crudest affairs imaginable, 
having been erected in feverish haste during the boom several years 
ago. When the production of the fields began to decrease, and the 
price of oil also declined, it was realized for the first time that the 
operating expenses were abnormal, and that unless greater economy 
were practiced, disastrous results would follow. Few changes were 
made, however, up to eighteen months ago, beyond the purchasing of 
additional equipment. 

2 Each concern has a central station or air plant and all the 
compressors therein are connected to a manifold from which the air 
lines lead to the various wells on the property held by that concern. 
The manifold design is such that by manipulating the valves, any 
machine may be made to operate any of the wells. 

3 Often the air lines reach the wells by a roundabout way, and 
have innumerable short bends, valves, double swings to avoid pipe 
cutting, and plugged tees instead of elbows. All of this tends further 
to decrease the economy of the operation, and taking all things into 
consideration, it is little wonder that the eflniciencies of the plants were 
low. The size pipe used for these air lines is designed neither for the 
amount of air to be transmitted nor for the distance it is to be carried, 
but is with one exception 2 in. in diameter. 

4 The boilers of the air plants are of 40 h.p., of a portable con- 
tracted waist type, and few were covered with asbestos. The boilers 
were so set that one-fifth of their lengths projected into the open, as 

Presented at the Spring Meeting, Washington, May 1909, of The American 
Society of Mechanical Engineers. 



312 



COMPRESSED AIR PUMPING SYSTEMS OF OIL WELLS 



indicated in Fig. 1, in order to avoid the necessity of perforating the 
roof to receive the stacks and to provide cooler boiler-rooms, regard- 
less of the heat wasted. 

5 The redeeming feature in all the plants is the type of compressor 
in general use. These compressors are generally of high grade, and 
display remarkable endurance. It is common for a machine designed 
for 350-lb. pressure to operate under a pressure of 500 lb., and at 
speeds far in excess of those for which it was designed. The most 
popular type of compressor has the duplex steam end and compound 
or two-stage air end. The steam cylinders are fitted with Meyer 
adjustable cut-off valves and the air cylinders in some instances with 
piston and in others with Corliss intake valves and poppet discharge 
valves. Plain speed governors are used and the capacities of the 




Fig. 1 A Typical Air Plant 



compressors range from 100 to 1000 cu. ft. of free air per minute and 
operate at pressures of from 150 to 750 lb. per square inch. The 
machine best adapted to the purpose, however, is the 500-cu. ft., 
500-lb. type. 

TERMS 

6 An explanation of certain terms to be used may not be out of 
place. 

"Submergence in feet" refers to the number of feet below 
the surface of the fluid (after the well has been pumped 
down, and is operating under its normal conditions) 
that the air under pressure is admitted. 

" Per cent of submergence" is the submergence in feet divided 
by the total number of feet of vertical discharge line, 
measured from the point of admission of the air to the 
point of discharge of the fluid. 



COMPRESSED AIR PUMPING SYSTEMS OF OIL WELLS 313 

" Volumetric efficiency" of the compressor is the actual amount 
of free air that is compressed and discharged by the cyl- 
inder, divided by the cubical contents of that cylinder. 

"Free air" is air at standard temperature and pressure. 

" Pumping head" is the vertical distance in feet (after the well 
is pumped down, as before stated) from the fluid level 
in the well to the point of discharge. 

Gal. per minute X pumping head in feet 



"The Constant" 
"The Ratio" = 



Cu. ft. of free air per minute 
Cu. ft. of free air per minute 
Cu. ft. of fluid per minute 



DESCRIPTION OF SYSTEMS 

7 Fig. 2, 3, 4, and 5, illustrate the air lift systems that are and 
hi-ve been in use on the oil fields. 

8 Fig. 2 shows the Straight Air or Sanders system. The well 
top is sealed as shown at A. Compressed air is forced through the 
pipe B into the space between the discharge or eduction pipe C, and 
the well casing D. 

9 When without air pressure, the fluid in the well will stand at 
some point- such as E, the level in the air space and the discharge 
line being identical. When air is forced through B, the level of the 
fluid in the air space is gradually forced down until the end of C is 
uncovered. Instantly some of the air escapes into the discharge 
pipe C, lowering the air pressure in the air space F. This cau^^es 
the fluid to rise in and up the air space and discharge pipe until a 
point is reached where air and water pressure balance. Then, more 
air coming in, the pressure again rises, the fluid level is forced down 
as before, more air escapes into the discharge pipe, and thus the cycle 
is repeated. As may be readily seen, the air that rushes into the 
discharge line carries the "slug" of water that has just previously 
entered. 

10 Fig. 3 shows what is commonly known as the Central Pipe 
system. The discharge line A is placed inside of the well casing as 
before and inside of the discharge is suspended a small air line usually 
H in. in diameter. The end of the 1 i-in. Une is plugged and a number 
of ^-in. holes are drilled inclining upwards in the last joint of pipe. 
Air is forced down through the small air line shown, passes out of 
the ^-in. holes, and mingles with the fluid carrying it out through 
the discharge line A. It is generally supposed that the fluid in this 



314 



COMPRESSED AIR PUMPING SYSTEMS OF OIL WELLS 



case is discharged because of the aeration of the fluid in the discharge 
pipe which in turn is caused by the intimate comminghng of air and 
fluid. The weight of the fluid column inside of the discharge pipe 
is therefore less in pounds per square inch than that without and the 
energy due to this difiference in weight is utilized to lift the fluid and 
overcome the various losses. 




Fig. 2 Fig. 3 . Fig. 4 

Straight Air Lift Central Pipe Return Bend 
System System System 



11 What is commonly known as the Open End system of air 
lift was at one time in quite extensive use on the field. It is 
similar to the system just described except that the small air line is 
open at the lower end, and of course there are no holes drilled in the 
air line. 

12 Fig. 4 illustrates a form of the Return Bend system. It is 
clj^imed by the inventor that: " It consists in improved processes and 



COMPRESSED AIR PUMPING SYSTEMS OF OIL WELLS 



315 



apparatus whereby the compressed air is delivered in bulk into the 
lower end of the water eduction pipe, and the water and air are 
caused to ascend through said pipe in distinct alternate la3'ers of 
definite dimensions." 

13 The use of this system has been discontinued in Evangeline 
because, as the field managers told the writer, it failed to produce as 
large a quantity of fluid as that produced by other systems. 




^—3 



SEiCTION f}-j9 



Fig. 5 System Combining Features of other Systems Described 

14 Fig. 5 shows a patent system which in reality is a combination 
of the several systems already described. The claims of the inventor 
are: less submergence, and hence less air pressure necessary, decreased 
air consumption, or with an equal amount of air, increased fluid yield. 

15 Compressed air is forced through a down into the foot piece, 
which is placed at that point of submergence shown by test to be 
most economical. The well top is sealed and air under pressure 



316 



COMPRESSED AIR PUMPING SYSTEMS OF OIL WELLS 



is also admitted between the casing and discharge pipe on the 
water head by means of the branch shown at h. This forces the 
fluid to a higher level in the discharge pipe and also prevents fluid 
in the air space or chamber from vibrating and foaming. This is 
quite an advantage in oil well pumping as the liability of making 
"riley oil" is thereby greatly lessened. 

16 The footpiece shown in section is made of cast brass and is in 
two parts. The air on reaching the foot piece divides and goes up 
through the hollow prongs / and g and out the nozzle n. The nozzle 
is adjusted to receive the quantity of air to be used by screwing the 
upper part s of the footpiece, in or out as the case may be. To 
increase the velocity of the fluid in the discharge hne, the footpiece 
is restricted and formed into a "venturi" as shown at v. 




Fig. 6 Type of Compressor Used 



Test No. 1 



17 The Crowley Oil and Mineral Company was the first to take 
active steps for the improvement of their plant and pumping equip- 
ment. They decided to install the patent air lift last described 
(Fig. 5). A test of the old system was first made to determine the 
amount of compressed air used and the fluid yield. The new equip- 
ment was next installed and a similar test made of the same duration 
and under the same conditions. The tests and installation were 
conducted on Well No. 32, 1805 ft. deep, and located 542 ft. from the 
compressor operating it. The air to the well was controlled by means 



COMPRESSED AIR PUMPING SYSTEMS OF OIL WELLS 



317 



of a manifold in the plant and was conveyed to the well top through a 
2 in. pipe line which as usual was in poor condition and badly 
designed. 

18 The system of pumping was that illustrated in Fig. 2. The 
well casing was 6 in. in diameter, suspended inside of which was a 
4-in. discharge line. 

19 The compressor was a duplex steam and compound air type 
made by the Ingersoll-Rand Company and designed to compress 
1000 cu. ft. of free air per minute to 350 lb. pressure. The steam 
end was fitted with Meyer adjustable cut-off valves and the air end 
with Corliss intake and poppet discharge valves. The machine is 
shown in Fig. 6. 

20 The discharge pipe from the well top was run up into a steel 
tank of known dimensions and the amount of fluid pumped during the 




Fig. 7 



test ascertained by direct measurement. Air gages, previously tested, 
were placed both at the compressor and at the well top thereby 
making it possible to determine the friction losses in the manifold 
and air line and also the actual pressure at the well top. Simultane- 
ous indicator cards were taken from the steam and air ends of the 
compressor, and from these cards were obtained the volumetric and 
mechanical efficiencies, the steam and air horse powers, and the 
.steam consumption (theoretical) of the machine. 

21 The volumetric efficiency .was assumed to be the ratio of 
the piston travel during admission stroke to the total piston db- 

placement, {— on the indicator card. Fig. 7). Sometimes this 
ce 

method is inaccurate and unsatisfactory (1) because the enter- 
mg air at atmospheric temperature and pressure is heated by 



318 COMPRESSED AIR PUMPING SYSTEMS OF OIL WELLS 

contact with the cylinder walls and piston; and (2) because of leak- 
age of air from the compression side of the moving piston to the suc- 
tion side. Neither the expansion resulting from the first condition 
nor the reduction in volumetric efficiency resulting from the second 
are observable on the indicator card. 

22 The first inaccuracy was partially overcome by placing a 
recently calibrated thermometer as far down in the intake pipe of the 
compressor as possible, noting the temperature and making the 
necessary corrections as will be observed in the log of results. 

23 The method of ascertaining the volumetric efficiency, that the 
writer would have used, but for his inability to obtain the necessary 
apparatus, was in brief as follows: 

Connect the air discharge of the compressor to an enclosed 
tank. From this tank^ connect to a cooler and from 
thence to a second enclosed tank of known dimensions. 

Place a regulating valve between the first tank and the 
cooler, setting the valve to maintain the pressure in the 
first tank at that point at which the efficiency is to be 
determined. 

Attach test gages to both tanks and a reliable thermometer 
to the second tank. 

Start the compressor and note the temperatures of the intake 
air and of the air in the second tank both at the beginning 
and end of the run. Note also the initial and final air 
pressures and the reading of the barometer, and the speed 
in revolutions per minute of the compressor. 

The volume of air compressed is then determined from the 
formula: 

273 + T/29.92 X P, _ 29.92 X P 
~ ^ R V 273 + r, 273 + T, 

where 

V = Volume of air compressed. 
Vi = Cubical contents of the second tank. 
T = Room, or intake air, temperature. 
Ti = Initial temperature of the air in the tank, 
y, = Final temperature of the air in the tank. 
R = Reading of the barometer in inches of mercury. 
P and Pi = Initial and final air pressures in the 
tank. 



COMPRESSED AIR PUMPING SYSTEMS OF OIL WELLS 319 

The volume of air thus obtained divided by the total piston displace- 
ment equals the volumetric efficiency. 

OBSERVATIONS 

24 Every thirty minutes for a period of six hours, readings were 
taken of the boiler pressure gage, the air gage at the compressor 
and at the well top, the r.p.m. of the compressor, the temperature 
of the intake air and of the barometer. A set of indicator cards, and 
also a sample of the fluid pumped from the well, were taken at each 
interval ^^,' |" 

25 The temperature of each sample of fluid was noted; it was then 
placed in a proper receptacle, and at the end of the test, the weight of 
a gallon was ascertained, together with the specific gravity of the oil. 
The amount of fluid pumped was determined, as before stated, by 
direct i^ easurement, due allowance having been made for the samples 
that were withdrawn. 

26 When these tests were run, no attempt was made by the writer 
to re-design the air lines, or to correct in any manner the numerous 
other defects. The old system was tested just as it had been operated, 
and the new system was installed and tested under the same adverse 
conditions. After both systems had been tested, some few of the 
defects were corrected in the manifold and air line design, thereby 
insuring more economical operation in the future. 

TABLE 1 SUMMARY OF RESULTS 

The Crowley Oil and Mineral Company, Evangeline, La. 

Old System New System 

Duration of test, hours 5.5 6.0 

Meani.h.p 122,56 89.19 

Mean water h.p 9.97 10.36 

Meanairh.p 107.38 79.16 

Gallons of fluid per second 0.542 0.608 

" " " " hour 1953.6 2188.2 

Barrels of fluid per hour 46 . 51 52 . 1 

Weight of 1 gal. of fluid 8.7 8.69 

Mean temperature of fluid, deg. fahr 111.5 113.2 

Percentage of salt water in fluid 87 . 3 86 . 7 

sand " 2.2 1.9 

crudeoil " 10.5 11.4 

Barrels of oil per hour 4 .86 5 .94 

Barometer reading, inches of mercury 29 .95 29 .94 

Specific gravity of oil 0.9 0.9 



320 COMPRESSED AIR PUMPING SYSTEMS OF OIL WELLS 

Old System New System 

Constant 63 . 1 101 .8 

Size of air line, inches 1 . 25 

Total depth of well, feet 1805.0 1805.0 

Size of casing, inches 6.0 6.0 

Height above ground to which the fluid was pumped, 

feet 18.5 18.5 

Size of discharge line used, inches 4 4 

Total length of vertical discharge line 1513 . 5 1513 . 5 

Total length of vertical air line in well 1513.5 . 1493.0 

Dimensions of compressors, inches* 10x22x16x20-7^x18x16x20 

Number operated 1 1 

Kind of fuel usedf Crude oil Crude oil 

Gallons of fuel used per hour 48 . 36 35 . 27 

Barrels of fuel used per hour 1.15 0.835 

Price of 1 bbl. of oil at time of test, dollars 0.90 0.90 

Costoffuelforproducinglbbl. of fluid, dollars 0.0222 0.0146 

Cost of fuel for producing 1 bbl. of crude oil, dollars. ... 0.212 0.126 

♦Type of compressor used, Rand Drill Co. Imperial Type X, Steam Cylinders, compound 
air cylinders. 

tType of boiler, oil well supply, portable contracted waste. 

Test No. 2 
wells no. 12, 30, and 32 op the crowley oil and mineral 

COMPANY 

27 The saving in air volume accomplished by the new system 
led those interested to endeavor to operate two wells with one 
machine, something before considered impossible in the field. 

28 Well No. 30 was forthwith tested, though not with sufficient 
accuracy to warrant the publication of the results, and the approxi- 
mate pumping head and submergence established. The new system 
was then installed with the requisite pipe to equalize the submergence 
(hence working pressure) of this well with that of No. 32. How 
successfully the working pressures of the two wells were equalized 
may be seen by reference to Table 2 of the Appendix. 

29 The two wells in question were then connected to one air 
compressor with gratifying results. No trouble was experienced in 
starting, and the machine furnished air in abundance for steady 
operation. 

30 Preparations were being made to run the usual test when the 
compressor operating Well No. 12 "went dead." This last named 
well had been previously tested and equipped with the new system. 
This shutdown, of course, would mean a loss of at least a day's pro- 



COMPRESSED AIR PUMPING Si'STEMS OF OIL WELLS 321 

duction from the well, amounting to quite an item, so the writer 
advised that this well be also connected to the machine already 
operating No. 30 and No. 32. By speeding the machine up a few 
revolutions, the additional load was easily taken care of as may be 
more fully noted by reference to the accompanying log (Table 2). 

TABLE 2 SUMMARY OF RESULTS 

Wells No. 12, 30, 32, Crowley Oil and Mineral Company 

Duration of tests, hours 6.0 

Mean (total) i.h.p 151 . 1 

w.h.p 25.14 • 

a.h.p 129.05 

Total gallons of fluid per hour 6168.0 

« barrels " " " " 146.87 

" " " oil " " 16.17 

Well No. 12 

Weight of 1 gal. of fluid 8.5 

Temperature of fluid 118 . 5 

Per cent of salt water in fluid 87 . 2 

" " " sand " " 1.3 

" " " crudeoU " " 11.5 

Barrels of oil per hour 6 . 44 

Specific gravity of oil . 87 

Total depth of well in feet 1705 . 00 

Size of casing, inches 6 . 00 

" " discharge line, inches 4 . 00 

Well No. 30 

Height above ground to which fluid was pumped, feet 17.5 

Total length of vertical discharge line 1025 . 5 

airline 992.58 

Weight of 1 gal. of fluid 8 .65 

Temperature of fluid 120.2 

Per cent of salt water in fluid 88 . 3 

" " " sand " " 1.5 

" " " crude oil " ".... 10.2 

Barrels of oil per hour 4.83 

Specific gravity 0.9 

Total depth of well in feet 1920 .00 

Si«e of casing, inches 6 .00 

" " discharge line, inches 4.00 



322 COMPRESSED AIR PUMPING SYSTEMS OF OIL WELLS 

Height above ground to which fluid was pumped, feet 18.00 

Total length of vertical discharge line 1516 .3 

Total length of vertical air line in well 1494 .2 

Well No. 32 

Weight of 1 gal. of fluid, pounds 8.7 

Temperature of fluid 114 .5 

Per cent salt water in fluid 86 . 9 

I. " " sand " " 1.8 

' " " crude oil " " 11.3 

Barrels of oil per hour 4 .90 

Specific gravity 0.9 

Total depth of well, feet 1901 .00 

Size of casing, inches 6 .00 

" " discharge " 4 .00 

Height above groimd to which fluid was pumped 18 . 5 

Total length of vertical discharge line 1513 . 

Total length of vertical air line in well 1493 .0 

Size of air lines in well, inches 1 .25 

Barometer reading, inches of mercury 29 .95 

Dimensions of compressor, inches* 10x22x16x20 

Number operated 1 

Kind of fuel usedf Crude oil 

Barrels of fuel used per hour 1 . 45 

Price of 1 bbl. of oil at time of test, dollars . 90 

Cost in fuel of producing 1 bbl. of oil, dollars .074 

♦Type of compressor used, Rand Drill Co. Imperial Type X, duplex steam cylinders, com- 
pound or two stage, air cylinders. 

tType of boilers, oil well supply, portable contracted waste. 

Test No. 3 
well no. 2, mamou power company 

31 This test was run in the same manner as those preceding 
except that the fluid field was ascertained by means of a two-foot 
rectangular weir placed between the earthen fluid and oil pits, the 
salt water bleeds of the former having been closed. The old system 
used was that illustrated in Fig. 3. 

32 The depth of fluid over the crest of the weir was measured by 
means of the ordinary hook gage calibrated to read accurately in 
hundredths of a centimeter. The weir constant was previously 
determined by testing in the usual way, using a sample of the fluid 
as pumped from the well. 



COMPRESSED AIR PUMPING SYSTEMS OF OIL WELLS 323 

TABLE 3 SUMMARY OF RESULTS 
Well No. 2, Mamou Power Company 

Old System New System 

Duration of tests, hours 10.0 10.0 

Mean i.h.p 99.1 62.8 

" w.h.p 9.85 13.36 

" a.h.p 82.5 50.4 

Gallons of fluid per second 0.694 0.849 

" hour 2499 .6 3056 .4 

Barrels of fluid per hour 54 . 75 72 . 77 

Weightofl gal. of fluid, pounds 8.72 8.75 

Mean temperature of fluid, deg. fahr 118.3 117.9 

Percentage of salt water in fluid 87 . 7 86 . 1 

" sand " " 1.2 1.6 

"crude oil in fluid 11.1 12.3 

Barrels of oil per hour 6.08 8.95 

Specific gravity of oil 0.9 0.9 

Barometer reading, inches of mercury 29 .94 29 .93 

Weir constant 24 .39 24. 39 

Pumping constant 97 . 1 202 .9 

Total depth of well in feet 1901 .0 1901 .0 

Size of casing, inches 6.0 6.0 

Height above ground to which fluid was pumped, feet . 3.33 3 . 33 

Size of discharge line used, inches 4.0 4.0 

Sizeof air line in well, inches 1.25 1.25 

Total length of vertical discharge line 1500 . 1500 . 

Total length of air line in well 1489.5 1489.5 

Dimensions of compressor, inches* 7^x18x16x16 — 7^x18x16x16 

Number operated 1 1 

Kind of fuel usedt Crude oil Crude oil 

Gallons of fuel used per hour 44.22 30.53 

Barrelsof fuel " " " 1 .05 0.727 

Price of 1 bbl. of oil at time of test, dollars 0.85 0.85 

Cost in fuel of producing Ibbl. of fluid, dollars 0.0163 0.0085 

Cost in fuel of producing 1 bbl. of crude oil, dollars .... . 128 . 069 

*Type of compressor used, Hall Steam Pump Co .Duplex steam cylinders, compound air 
cylinders. Plain "D" valves on steam end, poppet valves on air end. 

tType of boiler, 72'xl8' horizontal return tubular, manufactured by the Loclcout 
Boiler Co. 

Conclusion 

33 A careful examination of the tests brings out several points 
that may require explanation. 

34 The loss of air pressure by friction in the small l|-in. air 
line in the well, to which the footpiece of the new system was attached. 



324 COMPRESSED AIR PUMPING SYSTEMS OF OIL WELLS 

was approximately determined as follows: Pipe connections were 
made at the well top, so that by the manipulation of various valves, 
the air from the main line could be sent either through the 1^-in. 
air line or into the space between the well casing and the discharge 
line. By noting the pressure gage readings in each instance, the 
friction loss (assuming that there is no loss by friction when air is 
forced between casing and discharge) is represented by the difference 
in the readings. Corrections were made, of course, for that part of 
the discharge line below the footpiece. 

35 It was impossible to obtain the actual friction loss in said 1^- 
in. line by other means more accurate than those employed. While 
some little error may be involved in assuming no friction loss in the 
one instance, a comparison of the loss thus obtained with the theoret- 
ical loss is quite favorable, the former loss being the greater. 

36 Reference to Table 3 will show that the working submergence 
of the new system is less than that of the old, in spite of the fact that 
there is the same amount of pipe in the well in each case. This is 
due to the additional drop in pumping head caused by the increase 
of fluid yield. All calculations of submergence and pumping head 
were made from the observed air pressures after correcting for fric- 
tion losses, etc. The mean of these calculations was verified as far 
as possible by actual measurement. This was done by shutting 
down the compressor after the well had been in steady operation for 
several hours and pulling the discharge line. The point at which the 
fluid stood, while the well was being pumped, was plainly defined on 
the pipe. The time required after shutting down the compressor 
to pull the first "triple" from the well was a fraction less than two 
minutes. Comparison of the actual pumping head and submergence 
thus obtained with those obtained by calculations from the pressure 
gage readings was in each case very close, a difference of 10 ft. 2 in. 
being the maximum. 

37 Acknowledgment of valuable aid during tests is hereby made 
to the following who checked the writer in his various observations: 
on Well No.'32,'to'Mr. B. Brand, of the Crowley Oil and Mineral Co.; 
on Wells Nol 12, 30 and 32, to Mr. Brand, Mr. J. Murphree and Mr.S. 
Bolin, of the Crowley Oil and Mineral Co.; on Well No. 2 of the Mamou 
Power Co., to Mr. J. A.Sonet of that company and his able assistants; 
and especially is the writer grateful to Mr. J. W. Smith for courtesies 
extended during the former's sojourn on the field. 



COMPRESSED AIR PUMPING SYSTEMS OF OIL WELLS 



325 



X 






XMao aaj I «>; w t^ « m 

AONaiOMJ^ ONIdFIQjJ I » C3 05 OJ OS 

i.^jao 
Had aoNaoHaHHQg 

HONaoHawaflg 



«> 00 CO N 00 

C5 00 00 Oi 00 



t- CO 



X W 00 to -H 

£1 CO c^ ci M 

<N CV| <N CS (^5 

•« lo u5 X; 00 

«0 CO CO T)< o 



00 <0 l>. (N. 00 

N CO CO CO co' 

N N <N C5 N 


00 « 

a 


346.6 

358.05 

359.15 

359.66 

361.28 


u) o 
«0 00 

CO CO 




aaanaij I -^ m co ■>!< o 



1%%1% 


CD « 


S 2 !2 2 2 

CO Tl< ■^' ijj lO 




<0 h- 00 0> O 


r^- 1 



326 



COMPRESSED AIR PUMPING SYSTEMS OF OIL WELLS 



i 2 



ij 


fl 


h) 


« 


a 


S 


ts 


Jz; 


§ 


1 


H 






>> 


W 


oa 


H 


^ 




ai 


rH 


^ 


H 




i-J 




n 




<! 




H 







O N 00 00 CO 


CO 00 <N ^ «> 


CO us •* 


CO CO CO CO CO 


CO CI CO CO CO 


CO CO eo j 


XNao 


0> 0> CD «D CO 


CO •* 0> -"It CO 


CO CO CO 


t^ t^ r~ t^ t^ 

IN N (M IN IM 


t^ CO b~ 00 t^ 
<N (N (N <N IN 


t^ t^ t- 

CI IN N 


xaaj 
Ni aoNaaaawang 


05 O) 0) 
to lO i-H ^ ,-1 


05 

rH •^ US ■* 05 


05 05 05 


CO CO ^ -1 -H 


;:^?52g5:l 


555 


•M -d •» 


CO lO ■* ■* lO 

t^ t^ t^ I> l> 


us CD us us -^ 
t~ t^ I> t- t~ 


■* us us 

t~- t^ t^ 


•d 'H 'V 


CO ^ ■* •* 00 

05 00 Ifl lO t» 

t~ r- 1^ t^ t^ 


t' ■* t~ 05 05 

t^ O 00 00 US 
t^ 00 t^ t~ t^ 


us 00 US 

CO CO t>. 

t^ t^ t> 


•d -H -Al. 


CO CO ■* ■* Tl( 


^ CO CO CO ^ 


■*■*•<*( 


o o o o o 


o o o o o 


o o o 


•d -H 1 


to CO t~ t~ m 

T-l 05 CO CD 05 
05 00 00 00 00 


■* us O) t^ CO 
05 (N o d 1^ 

00 05 00 05 00 


87.9 
88.5 
89.1 


avajj ONidWQj 


US us rt ,-( rH 


rt CO us CO .-H 


.-1 iH 1-4 


CO CO ^ rt -H 

t^ t^ 00 00 00 
o o o o o 


^H 03 CO 05 1-t 
00 CD t^ CD 00 

o o o o o 


III 


K 
D 

(D 
(D 

« 

•< 


IPAilV 


<N (N O O O 
O O O O O 
N (N IN (N IN 


O us <M US O 

o o o o o 

<N IN (N IN <N 


IN IN (N 


jajoooja^ui 


t~ t- t^ t^ l^ 

t^ r- r- t> i^ 


t~ t~ t» t^ t» 

r- t^ t^ t^ t- 


r^ t^ t» 
t^ t^ ^• 


aSjBqosiQ 


f- t^ us us 10 

o o o o o 

N N C^ (N <N 


us O t^ O us 

O -^ O rH O 
CI (N (N IN M 


us us us 

o o o 

IN N N 


oiivg 


IN 00 US US OS 

.-I 05 r>^ t^ c» 
00 t> r- t^ t^ 


O) •<* O 05 00 

ess .-1 d 05 t>^ 
t^ 00 00 W b- 


78.6 
78.1 
79.5 


aiQNij*! aaj 


t^ r~ t> t~ t^ 

00 W 00 00 00 


t»- i^ t» r» r- 

00 X 00 00 00 


t^ r- t^ 

00 00 00 

•* •^ •.^ 


axiiNTi\[ 
Had SNO^^vo 


t^ t^ I^ t^ t- 

^ ^ ^ ^ ^ 

CD CO CO CD CD 

CO CO CO CO CO 


i^ t>. r- r^ t^ 

■* n< Ti< •>i< •* 

CO CO CO CO CO 


36.47 
36.47 
36.47 


HNVX 


1ft. 6 in. 




'. i .s 
: : "* 


(avni 
-OV) axaNij^ aaj 
aiy aaa^ .lo 'i^ "QO 


00 •* O O 05 

us CO t^ t-^ 00 
05 00 t^ t^ CO 

CO CO CO CO CO 


00 ,-( rH O) t^ 

CO CO 05 00 05 
00 05 00 00 t^ 
CO CO CO CO CO 


382.7 
383.9 
387.4 


•HHv J saaaoaQ 
aiy JO aaaxvaadwax 


O O O O CO 

IN (N IN IN .H 

00 00 00 00 00 


us CO CQ CO o 

00 00 00 00 00 


81.0 
80.5 
80.5 


iONaioMiia xNao 
aaj oiaxawQiOA 


C3 us O <S us 

ei r^ d d T^ 

05 05 05 05 05 


us C us us us 

.-I (N ^ ^ d 

05 05 05 O 05 


91.2 
90.2 
91.0 


(xNawaovidsiQ) 
axaNij\[ aaj 
HIV aan^ .jo "xj -qq 


■* US CO CO US 
00 (N CD CD C-i 

^ ^ CO CO ^ 


us T«i US us CO 

IN 00 Cq IN CO 

tJ^ ^ ^ ^ CO 

Tt* Tjt ■* ■* •* 


436.6 
442.5 
442.5 


awij, 


o o 

O CO o o o 

-1 -1 'H ^ (N 


o o o o o 

CO O CO O CO 
IN CO CO ■* ■* 


o o o 

O CO o 
us lis as 




HaawQU 


iH (N CO •* us 


CO t^ 00 05 o 


rt (N CO 
>H i-l iH 



COMPRESSED AIR PUMPING SYSTEMS OF OIL WELLS 



327 



iDHSIOIAA^ ONIdRQJ 



a 


£ 


<5§ 


e » 


H S 


0^ M 





OQ 



2S-ON 



08 'ON 



ZI ON 



28 "ON 



OS 'ON 



ZX -ON 



•d -H -M. i«>ox 



Z8"ON 



08 "ON 



0> O CO U5 ■* 

00 0> 0> OS o 

-< — I — c ,-1 (N 

^ lO i-H t-( lO 



t^ 0> t>. CO lO 

O 00 O OS 03 

lO «o rt ^ m 



U3 cc u} 

O 0> 03 



o» 


o 


^ 


^ 


to 


to 


oj 


^ 


^ 


to 


o> 


to 


fft 










1^ 


t^ 




</l 


r/1 


t^ 


to 


t^ 
















o 


o 


o 


o 


o 


o 


o 




























t~ 


00 


00 


t^ 


1^ 


r- 


00 


1^ 


t^ 


t^ 


00 


t^ 


t» 


t^ 


o 


o 


r^ 


r^ 


r- 


o 


r^ 


1^ 


t^ 


o 


I~- 


r>. 




1- 


r- 


t^ 


i^ 


r^ 




r^ 












o 


o 


o 


o 


o 


o 


o 


o 


o 


o 


o 


o 


o 


•-H 


»— ' 


1—* 


r~* 


f—* 


r^ 
















o 


05 


o» 


a> 


Oi 


in 


05 


05 


05 


lO 


05 


o> 


o> 



to o 
G5 a 
to to 



00 t^ 

(N <N 

O lO 

00 00 

05 Oi 



t^ 00 r^ t~. r^ 



00 00 00 00 00 



00 r~ t^ 

(N <N N 

lO o o 



lOioioioic tomiO"5Ti< 



N rt 05 00 (N 

CO CO <N IN N 



NCS— lOCO 05(N05.-<0 



oo 00 00 00 r~ 

LO lO 

il O -H (N rt 



t^ 00 r » CO oo_ 
lO _ »o 

o> o i-H e^ (N 




© to t- r^ to 



Til IQ lO lO lO 
<N CQ M CQ N 



to to t^ r- to 



G> 0> 0> Oi > 



SI 'ON 



aaaj JO -i j -no OOI 
[ aiQNip^ saj SNoaavo 



« M 



Z8 


•ON II^Ai %Y 


in e-j o o (N 

o o o o o 

(N <N C^ (N Cl 


IN m O Q N 

o o o o o 

IN IN (N IN IN 


IN M IN 


08 


■ON IPM »V 


CI lO lO OJ IM 

o o o o o 

(N (N IN (N IN 


(N in IN IN <N 

o o o o o 

IN IN (N (N IN 


(N IN N 


ZI 


•ON IPAi ?V 


OO XI CC CO CO 
N (N <N W !N 


CO 00 00 OO CO 
CO (N IN (N <N 


00 00 00 
N IN CI 


J3IOO0J3}a| 


ic lo in lO lO 
«o to to to to 


in in in in in 

to to CO to to 


in in lo 

to to to 






t^ lO in m in 


in r^ in in in 


in in in 



9Sj'eT{08IQ 



OUVJJ 



aiQNiw HS£ oiaij .JO 'Xjj -no 



axQNij\[ Haj SNoaiTj) iviox 



ZT •ON 



ininci'-i-'i' o)ino5co>H ,-ico>h 



in in lo in ^ 

'n< Tjl to 00 «5 



to to to to t- 



Tii in ■* m in 

O Tjt 05 to us 

t^ to to to to 



in •* in ■* 



in in m •* d 

t^ t- t>. r- t» 

coco'coeoeo coeocococo 



t- i> t- t» t^ 



0000000000 0000000000 



00 00 00 

ci ci N 









Z8 -ON 


o o o o o 

cO CO CO CO CO 


O O O O O 

CO CO CO CO CO 


O O O 
CO CO CO 


OS •ON 


N N N CI N 


CJ CI Cl CI Cl 


C< N N 


CO CO CO CO CO 


CO CO CO CO CO 


CO CO CO 




C» N (N N IN 


N IN CI N IN 


CI « N 



05 05 O 05 O) 



(avQoov) aiiNij^ 
Haj HIV aaaj .io -xj -qo 

•HHyj saaHoaci 'Hiy 
aHVXNi J.O anniVHadwax 

fKvajv) 
iONaiOLJjg oiHxawQqoA 



(XNawaovidsiQ) axQNij^ 
aaj arv aanj jo 'xj 'qq 

awix 
Haana|i~ 




to to to to t- 
t^ r^ r^ t^ t~ 


b* F^ to to to 

r- t^ t^ t~ w 


S S K? 
r^ t>« t* 




I o • • • 

; 03 • _• • 


: : : 


05 05 CI ■* 00 


to <3> to CI •* 


•* CI ■* 



05 05 O CI •* 

rt -H -H O 00 

^- r~ t-- t^ to 



rt rt CI CI fH 

^i-HTeTco •* lo 



o r^ CO t^ t^ 



11 N 
to t^ 



328 



COMPRESSED AIR PUMPING SYSTEMS OF OIL WELLS 



o 


g 


CO 

o 


►J 




(M 


w 


•< 


b 


fa 
n 


> 


03 
2 








M 




n 


fa 


H 


H 


a 






0) 




ft 




CO 





!>, 


» 


PL, 


OJ 


h3 


P 




PQ 


O 


o 


■xl 


S 




H 


< 





iONaiOlJjg ONHWQfJ 


"«< ic r- «c N 


t^ r~ O IN 1-1 CO 1 


CO M O -H O 

-H -H rH -1 rH 


T-H 1-1 CO CO CO d 


XNHO 

aaj 'aoNaoHawaag 


00 00 CO CO CO 

00 00 (n' (N (M 

CO CO •* •* ■* 


CO CO 05 00 00 00 

(N N 00 00 00 00 
■* ■* CO CO CO CO 


laa^ 
aoNaoaawaag 


00 00 CD «D to 
(N (M 00 00 CO 

CO CO CO CO CO 

00 00 CO CO CO 

lO lO CD «0 to 


CO CD CO 00 00 00 
00 00 111 IN C^ IN 

CO CO ■* CO CO CO 

CO CO 00 00 00 00 

CD to 1(5 X5 X5 lO 


1 

•w •<! -a 


00 •* CO O IM 

00 05 O O 1-t 

-1 rt rH 


00 O (N ■* ■* •* 

OS O OS OS OS »H 


•d -H "V 


CO CO •^ "5 in 


CO t^ «5 ■* CO iH 


lO 1(3 03 1-1 ■* 
t^ t> 00 00 O 


(N 1-1 r^ to to i> 
00 CO r- t^ t' OS 


•d -H -M 


1-1 e^ to lo to 

O O 0) O 05 


to CO ^ 1-1 O O 

c» OS d <d d d 


•J -H •! 


OO CO 05 00 t^ 
^ co' 00 00 lO 

05 05 O 05 *-< 

1-1 I-l 


!N lO X5 CO CO 1-1 

1-1 1-1 in CO CO t^ 

O O OS OS OS o> 


avajj ONidivnj 


(N C^ ■* >* •* 

r~ t^ 1-1 1-1 I-l 


Tf rjl t^ N N (N 

rH 1-1 in t- t» t>- 


CO CO CO to CO 
^ —I CO CO CO 
OS 03 00 00 00 


CD CO in to to CD 

CO CD 1-1 ,-1 rH rt 

00 00 OS <33 OS 05 


Air Pressures 


IPAV *V 


iC lO ■*■<*' -"J" 


■* Tji o >n m in 


(N (N t^ f^ t^ 

lO lO 1^ ■* Tl< 
(N IM IN e^ (N 


t^ t- CO C^ C<1 N 

Tj< •* m m in in 

IN IN IN (N (N N 


jajoooja^nj 


o o o o o 

00 00 00 00 00 


O O (N <N (N IN 

00 00 00 00 00 00 


aSjBqosja 


o o »o lo lo 

to CD m lO IC 
IN M IN (N (N 


in in o o o o 

in in CO to to CO 

IN N IN IN (N (N 


oixvy 


lO 1(5 Tjt lO OS 


iH in OS 05 ■* o 


-H Tl< CO 00 (^ 
CO CO t^ CD 1> 


05 00 (N ■* m o 

CD CO CO to to 00 


axamp\r aa<j 
aiaaj jo -xj 'qq 


t^ rt h- ■* 1-1 

U5 CD lO to CD 


t^ t^ t> 1^ •* •* 

in in in in in in 


US U? K) KJ US 


in in in in in in 


' aXQNIJ\[ 

aa<j SK[0^^v][) 

i 


U5 CD lO 

r^ OS t^ ^ 05 


CD CD 

r~ t~ t^ 1^ •<i< ^ 


•WO 'aia^y^ h3ao 
aIa^^^I .■lo iixjhq 


-1 C^4 -H O IN 

t^ t^ ^~ t» t» 


_ rt _ rt o o 
t» t^ t^ t^ t^ t^ 


^^^^^ """""" 


(avaxov) 
axamj^ aaj aiy 
aaaj ao uj -no 


to Tft rt ^ CO 
Tjt O K3 O lO 

M (N c:s CD CD 
■* to O CO CO 
CO CO •* w ■* 


(N in CO •* T)< ^ 
in in 00 o o c^ 

d i-< d IN M* CO 

00 00 in CD CO >* 

CO CO CO CO CO ■^ 


•UHVjj saaaoaQ 
aiy 'SO aaoxvaadnax 


I-l lO 00 <N t^ 
U5 b-.' t^ 00 00 

1^ t^ t~ t^ t^ 


00 CO CO in N o 
00 1-1 iH t^ CO in 

t^ 00 00 t^ t^ W 


AONaiow^a 
oiaxawaao^ 


(N r- O o 1-1 

ui •<ii U5 U3 CD 

00 00 00 00 00 


CO m in t^ t^ 1-1 
in ■^ ■^ -^ -^ ui 

00 00 00 00 00 00 


(XNawaovidsxa) 
axnNii\[ aa<j 
aiy aaaj jo -xj -no 


CO 00 IN O ■* 


C« O ■<1< 00 00 00 


CO 1-1 00 O CO 
i-( ■* cs r~ (N 

■* 1j< Tjt 1^ lO 


01 o c^i -H rt in 
CO r^ CO ■* ■«< CO 


1 

j awix 


t^ <» 6s' o rt 


u^ in in in m m 
^ ^ ^ ^ ^ ^ 

1^ 1^ (N CO ■* in 


! 


aaftwntj 


-H N CO •* «S 


to t^ 00 05 O *H 

1-1 1-1 



COMPRESSED AIR PUMPING SYSTEMS OF OIL WELLS 



329 



00 ^ Tl< N (O 


to o a> <-> lO oi 


t- r~ t^ CO t~ 

<N C<1 (N (M IN 


n in ui t> 00 <o 

M (N M M IN (N 



i-< 00 es) M 64 



CO CO CO M ^5 CO 
CO CO CO CO CO CO 



CO •* tH Tl< -^ 



'J" lO U5 lO ■* ^ 



"S CO o ■* ■* 
CO CO CO CO CO 



■* CO N CO us Tj« 
CO CO CO CO CO CO 



00 00 M CO CO 
OJ 05 CI t^ t^ 



C^ M C» N W 



« N (N N N « 



i^ t^ CO r^ CO 
CO CO ■* CO Tji 

e^ (N (N (N (N 



CO O OS -< T»< 1-1 
M ffl lO ^ CO CO 
(N (N N (N N N 



O O N O N 


<N eo CO <N O (O 
t^ CO CO CO CO CO 


US CI lO CO 00 

00 a> o> r^ 00 
•* ^ 'I' ^ •«ji 


-H « t» CO CO e» 

5 S S 5 5J S 



r» 


in 


CO 


t~ 


^ 


CO 


00 


us 


C4 


00 


t- 


a 


o 


f_H 


o 


^H 


,^ 


in 


in 


,_, 


nn 


lO 








la 




t~ 


CO 


CO 


CO 


us 


(O 


N 


CO 


(O 


<o 


CO 


to 


N. 


r». 


t^ 


(O 


CO 


CO 


w 


'^ 


'-' 


"-I 


rH 


o 


o 


o 




1-1 


CO 


lO 


U5 


>o 


lO 


>n 


r^ 


r- 


r^ 






o 


uo 


» 


Oi 


Oi 


05 


on 


00 


CO 




d> 


a> 


OS 


Oi 


o 


o> 


c» 


a> 


03 


o> 


OJ 


a> 


la 


CO 


o 


o 


o 


o 


lO 


US 


lO 


o 


o 


CO 


00 


M< 


■* 


•<J< 


■* 


r^ 


r^ 


r^ 


•^ 


























IM 


(N 


M 


N 


(N 


(N 


M 


(N 


(N 


(N 


N 


o 


o 


o 


o 


o 


O 


o 


o 


o 


o 


o 










00 




00 


00 


00 


00 


00 . 


o 


CO 


o 


o 


o 


o 


us 


in 


in 


o 












(N 








IN 














N 


w 


(N 


C'J 


<N 


w 


(N 


a> 


o 


CO 


a> 


00 


CO 


w 


C4 


US 


« 


**< 


rf 


■* 


U5 


•* 


U5 


^ 


M 


00 


US 


•* 


00 


CO 


CO 


CO 


CO 


CO 


•* 


CO 


CO 


CO 


CO 


CO 


^ 


^ 


Th 


00 


00 


^ 


^_^ 


f/l 


^ 


1< 


1 










t^ 


00 


00 


r- 


00 


w 


00 


CO 


(O 


CO 


CO 


CO 


o 


CO 


o 


CO 


CO 


CO 



CO 


CO 


US 


00 


C4 


O 


FH CO 


■H CO 


00 




C4 


K 


f2 


■* 
t^ 


IS 


CO 00 


t^ CO 

1^ t^ 


04 


CO 


CO 


t» 


CO 


t^ 


rt 


us us 


CO CO 


lO 


to 

00 


CO 
00 


us 
00 


CO 
00 


s 


CO 

00 


CO CO 
00 00 


us <o 

00 00 


s 


o 


o 


'Jf 


o 


<* 


■«J< 


<N (M 


TJ. O 


(M 


IM 


IM 


o> 


(M 

00 


i 


CO 


O O 

CO CO 


r- IM 
o » 
(M Ci 


O 
CO 


o 

(N 


s 


o 


o 


o 

<M 


o 


N a 


N <N 


s 


t~ 


00 


a> 


o 


i-H 


esi 


•H evj 


CO ■««< 


UJ 



-<e4co^us cot^oooO'H 



330 DISCUSSION 

DISCUSSION 

F. A. Halsey. The air-lift pump is one of those things of which 
our knowledge is almost exclusively experimental. Its analysis pre- 
sents such serious difficulties that rational equations for the perform- 
ance of the apparatus have not been derived by anyone to my 
knowledge. Under these circumstances, we are reduced to experi- 
ment for the determination of the fundamental laws of the perform- 
ance of the apparatus, and these experiments should cover a wide 
range in the conditions which lead to variation in the performance 
of apparatus, namely, depth of submergence and lift. 

2 The most complete experiments now on record were presented in 
a paper before the British Institution of Civil Engineers in 1906,* 
these experiments being of sufficiently wide scope to supply a satis- 
factory guide for the design of these pumps under a considerable 
range of conditions. 

3 One curious feature attends the behavior of the air-lift pump and 
the Taylor hydraulic air compressor, which are essentially the same 
apparatus, reversed in action. In each case we have a pair of vertical 
pipes communicating at the bottom, one filled with water and the 
other with a mixture of air and water — a sort of suds. If the pipes 
are of indefinite length the columns will take levels corresponding with 
their respective gravities, but if the suds pipe is cut off below this 
level it will overflow, the suds will rise, and we have an air-lift pump; 
while on the contrary, if the water pipe is cut off below this level, it 
will overflow, the column of water will rise, and we have a Taylor 
hydraulic air compressor. The point to which I refer is that while the 
first of these constructions has a low, the second has a high efficiency. 
The highest figure given in this paper for the efficiency of the pump is 
about 28 per cent, and while 40 per cent has been reached in exceptional 
cases, the average is probably not more than 20 per cent. On the 
other hand, the efficiency of the hydraulic air compressor is in the 
vicinity of 75 or 80 per cent. 

Dr. Sanford A. Moss. The method used by the author for computa- 
tion of volumetric efficiency is subject to serious inaccuracies, as he 
points out. For the purpose of the tests in question, where relative 
rather than absolute volumetric efficiencies are desired, the method 
is quite proper. The volumetric efficiencies given must be understood 
to be relative and not absolute, however, as the errors may be quite 

♦Abstracted in the American Machinist, August 16, 1906. 



COMPRESSED AIR PUMPING SYSTEMS OF OIL WELLS 331 

serious. I have used the method which the author states he would 
have preferred for ascertaining volumetric efficiency, Par, 23, and 
found it very good. Everything operated as would be expected and a 
reputable value of volumetric efficiency was obtained. In one case, 
the method as given in Par. 23 was checked by use of calibrated ori- 
fices and also by use of a single tank with rise of pressure, the rate 
of increase of pressure when the pressure reached the rated value 
giving the desired volumetric efficiency. All three methods checked 
very closely. The volumetric efficiency in this case was 58 per cent, 
showing the serious losses which can occur due to the inaccuracies 
mentioned. 

J, G. Callan.* Par. 21 to Par. 23,* with diagram (Fig. 7), refer to 
the method of determining the efiiciency of the air compressor, and 
Mr. Ivens states in Par. 23 the method which he would have preferred 
for determining this quantity. 

2 The method of pumping up pressure in a large tank is open only 
to the objection that the large tank is rarely obtainable, particularly 
when the pressures which it is proposed to use are high, and the vol- 
ume large. A more available and almost equally exact method is 
afforded by the use of calibrated orifices of which the discharge coeffi- 
cient has been duly determined. These orifices can be used in con- 
nection with well-lcnown formulae, which involve a knowledge of 
corrected pressure and temperature of air delivered to orifice, and of 
velocity and temperature of emergent jet, all factors which can be 
readily and accurately determined. 

3 The losses of a compressor are so considerable and so complex 
that the assumption of volumetric efficiency from the indicator card 
may be in error by extremely large percentages and is quite as likely 
to be misleading as enlightening, unless results are strictly compara- 
tive rather than absolute. The value of the present tests would appar- 
ently be somewhat affected by the errors in this method of determin- 
ing delivery, since certain elements of compressor loss would assume 
different percentage values with different discharge rates. 

4 This brings up the desirability of some well-understood stand- 
ardized method of testing compressors and other apparatus for defiver- 
ing air. Commercial loss is certain to arise from the considerable 
variations in terminology, and a definition of terms by an authorita- 
tive body such as this Society would be of marked advantage. For 
example, volumetric efficiency undoubtedly should mean the ratio 

' General Electric Co., Schenectady, N. Y. 



332 DISCUSSION 

between the air actually discharged (when restored to standard atmos- 
pheric conditions of temperature and pressure) to the air which would 
have been discharged had the compressor cylinder been completely 
filled on each stroke with air under standard atmospheric conditions, 
and had all the air displaced been completely discharged. The exact 
method of determining the approximation of real to theoretical 
discharge obviously should be only a secondary consideration in such 
a definition, but could properly be taken up as corollary. 

5 The segregation and determination of losses in air compressors 
of ^iifferent types form an interesting and useful line of investigation 
upon which little has been pubhshed. A considerable amount of 
work with which I am familiar indicates that these losses are usually 
materially greater than the customary assumptions, particularly in 
compressors which have been in service for some time. Various 
methods h.'ive been devised for independently estimating loss due to 
heating of intakc; leakage of inlet, discharge valves and piston, clear- 
ance loss, throttling, and losses due to improper setting of mechanical 
valve gears, but usually these determinations will involve more labor 
than a direct measurement of air output by orifice and impact tube. 
It is my belief, however, that the computation which neglects all of 
these losses except the clearance loss, particularly where the pressure 
is high and the compressor is somewhat worn, is almost sure to be 
very gravely in error. 

The Author. The discussions of the paper on oil-well pumping 
presented at the Washington meeting were centered on the method 
of ascertaining volumetric efficiency of air compressors. The object 
of the paper was rather to compare the relative efficiencies of various 
air-lift systems when being operated on extremely high pumping 
heads. However, as all the discussions were agreeable to the writer's 
views and methods, there is no room for argument. 



No. 1244 

THE SPECIFIC VOLUME OF SATURATED 
STEAM 

By Prof. C. H. I'eabody,' Boston, Mass. 
Non-Member 

For many years the specific volume of saturated steam has been 
computed from the thermodynamic equation 

AT dp 
dt 

in which the quantities have the following significance: 

s is the specific volume, for example the volume in cubic 

meters of one kilogram. 
r is the heat of vaporization in calories. 
A is the heat equivalent of a unit of work. 
T is the absolute temperature obtained by adding 273 to the 
temperature by the centigrade thermometer. 

-^ is the differential coefficient of the pressure with regard to 

the temperature, the pressure being in kilograms per 
square meter. 
o is the specific volume of water i,0.01 cubic meters per kg.) 

2 For this paper French units are used because the original data 
are given in them and comparison with experimental values is con- 
venient. 

3 All the quantities entering into this equation are now deter- 
mined with a certainty and precision that must be considered satis- 
factory for engineering purposes and a comparison with experimental 
determinations of the specific volume shows an exceptionally good 
concordance. 

' Professor Naval Architecture and Marine Engineering, Mass. Inst. Tech. 

Presented at the Spring Meeting, Washington, May 1909, of The Amebican 
Society of Mechanical Engineers. 



334 SPECIFIC VOLUME OF SATURATED STEAM 

4 To make the exposition of this statement clear it is necessary 
to review the experimental data and to state the jDrecision that can 
properly be attributed to them. 

5 The mechanical equivalent of heat as determined by Rowland' 
may be taken as 427 meter-kilograms (778 foot pounds) at 15 deg. 
cent., which corresponds nearly with 62 deg. fahr. There have been 
more recent investigations which on the whole confirm this result, 
though there is some indication that it is a trifle small. The uncer- 
tainty may be one in a thousand or one in two thousand. 

6 Callendar^ gives for the absolute temperature of freezing point 
273.1 deg. cent., with a probable error of one in two thousand. 

7 For the range of temperature from 30 deg. to 100 deg. Henning^ 
gives the equation 

r= 94.210 (365 - t) "■^''^'^ 

in calories at 15 deg. cent. In English units the equation may be 
written 

r= 141.124 (689 - t) ^-^^^^ 

Experiments by Dieterici,^ Griffiths^ and A. C. Smith^ confirm his 
results and extend the equation to freezing point. The probable 
error of this equation is one in one thousand. 

8 In his paper, The Total Heat of Saturated Steam, read at the 
Annual Meeting, 1908, Dr. Harvey N. Davis gives for the total heat 
of steam from 212 deg. to 400 deg. fahr. 

H = 7/212 + 0.3745 {t - 212) - 0.000550 (t - 212)- 

Transformed into French units this may be written 

H = 638.9 + 0.3745' (t - 100) - 0.00099 (t - lOOy 

provided that the constant term be taken as the sum of Henning's 
value for r at 100 deg. cent, and the heat of the liquid be taken as 
100.2, according to a consideration to be taken up later in this paper. 
To conform with the conditions already accepted, this equation 
should give the total heat in calories at 15 deg. cent., while Dr. Davis 
used for the calories 1/100 of the heat required to raise one kilogram 

^ Proc. Am. Acad., vol. 15 (n.s. 7), 1879. 

' Phil. Mag., Jan. 1903. 

'Annalen der Physik, vol. 21, p. 849, 1906. 

* Annalen der Physik, vol. 16, p. 912, 19Q5. 

•Phil. Tians., 180, p. 261, 1895. 

' Physical Review,vol, 25, 1907. 



SPECIFIC VOLUME ©F SATURATED STEAM 335 

of water from freezing to boiling point. The difference amounts to 
2/1000, as indicated by the heat of the Hquid just mentioned 
(q = 100.2). Now the total heats at 100 deg. and 200 deg. cent, 
are 638.9 and 666.5, and their difference is 27.6 calories, so that 
the total effect is less than one-tenth of a calorie. 

9 As for the heat of the hquid we have the three following sources 
of information: 

a Barnes" determinations of the specific heat of water from 
deg. to 95 deg. cent. 

b Dieterici's^ determinations of the same property from freez- 
ing point to very high temperatures. 

c Regnault's^ determinations of the heat of the liquid. 
Barnes' experiments were made by an electrical method for which 
great relative precision is claimed, and they showed a good concor d- 
ance with RoAvland's work on the mechanical equivalent, which in 
reality was an investigation also of the specific heat. Dieterici's 
investigation consisted essentially in heating water in a quartz tube, 
which was then transferred to the ice calorimeter. His results appear 
to be systematically larger than Barnes'; at 95 deg. cent., the dis- 
crepancy is y\ of 1 per cent. 

10 In 1907 the author endeavored to join Regnault's values for 
the heat of the liquid to those deduced from Barnes' values of the 
specific heat. Now Regnault's experiments consisted in running hot 
water into a calorimeter partly filled with cold water and noting the 
rise of temperature in the calorimeter. There were 40 tests in all, 

cattered irregularly from about 100 deg. to 190 deg. cent, for the 
temperature of the hot water; there were in a way three groups of 
tests, one near 110 deg., one near 160 deg., and the third near 190 
deg. cent. 

11 The average rise of temperature in the calorimeter for the first 
group was not far from 9 deg. cent., which item appears to account 
for the considerable irregularity of results at that place. The 
experiments with the highest temperatures had nearly twice that 
rise of temperature in the calorimeter and about half the dispersion 
of results. 

12 In order to use Regnault's results his values for the heat of 
the liquid were recomputed, allowing for the true specific heat of the 
water in the calorimeter, and then a diagram was plotted as shown 

» Phys. Review, vol. 15, p. 71, 1902. 

* Annalen der Physik, vol. 16, p. 593, 1905. 

* Memoirs de Tlnstitut de France, vol. 26. 



336 



SPECIFIC VOLUME OF SATURATED STEAM 



by Fig. 1, in which the abscissae are temperatures and the ordinates 
are values oi q — t. 

13 This allows of the use of a large vertical scale which much 
accentuates the apparent scattering of points. A curve was then 
drawn to join a curve from deg. to 100 deg. cent., from Barnes' 
results for the specific heat of water. This curve passes near the 
highest group of points, above the middle group and below the lowest 
group. 

14 It should be said that Barnes' results were fii'st transformed 
to allow for the use of 62 deg. fahr. for the standard temperature, 
instead of 20 deg., which he had taken in his report; also that his 



- 










' / 












v7 


—3.0 










7 


- 








> 


^ 


- 








/ 




—1.0 






/•*. 






- 


t 


. 


•/ *. * 









. 


•* V 








- 


•;' 


y 








— OtO 












- 


100 




150 




200 


- 1 1 


1 1 1 


1 1 


1 1 ' 


1_ 


1 1 1 



Temperature Centigrade 

Fig. 1 Recomputation of Regnault's Experiments on the Heat of the 

Liquid of Water 

values were slightly increased at temperatures approaching 100 deg. 
so as to avoid a break in the curve. The last had the effect of increas- 
ing the heat of the liquid at 100 deg. by one one-thousandth. 

15 Finally a table of specific heats was drawn off for temperatures 
from deg. to 220 deg. cent., which served as the basis of a graphical 
integration for the value of g — ^. Fig. 2 gives the curve represent- 
ing the final value of this quantity and also a curve representing 
values that would be obtained if Dieterici's values for the specific 
heat were accepted. 

16 The author is of the opinion that the full curve in Fig. 2 shows 
verj'- nearly the true value of the property under consideration, and 
he has used it to determine heats of the liquid. 



SPECIFIC VOLUME OF SATURATED STEAM 



337 



17 The maximum deviation of a single point from the curve in 
Fig. 1 is 0.8 of a calorie, which amounts to | of 1 per cent of the 
heat of the liquid at that point. If we could consider that an error 
of 0.02 deg. might be attributed to the temperatures in the calorim- 
eter it would account for one-third of that deviation. But to take 
the most pessimistic view of the situation and charge an error of 0.8 
of a calorie against the method, we may still consider that for tem- 
peratures above boiling-point the heat of the liquid is always asso- 
ciated with the heat of vaporization, and that their sum is more than 



-4.80 



—5.20 



-1.60 




100 



150 



300 



Terupe-rature Centigrade 
Fig. 2 Values op the Quantity (q-t) 

THE FULL CURVE SHOWS THE QUANTITY DEDUCED FROM THE AUTHOR'S COMBINATION OP BARNES* 
EXPERIMENTS ON THE SPECIFIC HEAT OP WATER WITH REGNAULT'S EXPERIMENTS ON THE 
HEAT OP THE LIQUID, WHILE THE DOTTED CURVE SHOWS EBSDLT8 FROM DIETERICI'B BXPERI- 
MENT8 ON THE SPECIFIC HEAT OF WATER. 



630 calories, so that the deviation in this light amounts to J of 1 
per cent. 

18 A more just view is clearly to take the deviation of the worst 
group of points. This occurs at 1 17 deg. and is about 0.3 of a calorie, 
that is, 0.25 per cent of the heat of theUquid. The most favorable 
view is to consider that the upper end of the curve is well fixed by Reg- 
nault's experiments, which were then under the most favorable con- 
ditions, and that the lower end is tied to Barnes' values, which have 
all desired precision. This matter is discussed with some detail be- 



338 SPECIFIC VOLUME OF SATURATED STEAM 

cause the original experimental results needed to be entirely recast 
for the present purpose. 

19 But while important from some aspects, the quantities with 
which we are dealing are not affected by uncertainties that concern our 
main investigation, i.e., the specific volume of saturated steam, for 
the maximum variation between the author's value for the heat of 
the liquid, and a value determined from Dieterici's investigation, 
amounts to 0.8 of a calorie at 200 deg. cent. This is only J of 1 per 
cent of the total heat at that place. However, we need for our specific 
volume the heat of vaporization, and the discrepancy then becomes 
i of 1 per cent. 

20 Recent determinations of the pressure of saturated steam have 
been made by Holborn and Henning,^ with all the resources of modern 
physical methods including the platinum thermometer. They claim 
a precision of 0.01 deg. in the determination of temperature and that 



25 
20 
15 
10 

5 

1100 [110 |120_.--flS0 lUO H50 1160 ilTO IISO 1190 |200 i210 \ 
. 




Fia. 3 Curve to Extrapolate Pressure op Saturated Steam to 220 Dpo 

Cent. 

their results reduced to the thermometric scale have a probable error of 
not more than 0.02 deg. at 200 deg. cent. Their own experiments cover 
the range of temperature from 50 deg. to 200 deg. cent. (122 deg. to 
392 deg. fahr.), and they have extrapolated results to 205 deg. cent. 
Below 30 deg. they have made use of experiments by Thiesen and 
Scheel to extend results to freezing points; these experiments were 
not made with the same degree of precision as those by Holborn and 
Henning. 

21 In order to extend calculations to 220 deg. cent., as has been 
the habit in computing steam tables, the author made use of a dia- 
gram shown by Fig. 3, in which the abscissae are temperatures centi- 

' Annaleu der Physik, vol, 26, p. 383, 1908. 

Note. Since these results may not be easily accessible, it may be of interest 
to say that they have been transferred directly to Table 3, of the author's Steam 
and Entropy Tables, edition of 1909. 



SPECIFIC VOLUME OP SATURATED STEAM 339 

grade and the ordinates are differences between Holborn and Hen- 
ning's value and pressures computed by the following equation: 
log V = 5.457570-0.4120021(9.997411296 - 10)* "^o^ + 

(7.74168 -10) (9.997411296 - 10)* "^oo 
which was chosen as a matter of convenience and because it gave a 
curve which crossed the axis near 220 deg. cent, when produced. It 
is thought that the extrapolated values are not much in error, 
though there is no means of determining this question. Fortunately 
this part of the range of temperature, as well as that below 30 deg. 
cent., is not so important to engineers. 

22 The degree of precision attained by Holborn and Henning in 
the determination of the pressure of saturated steam is far beyond 
any direct technical requirement, since pressures are seldom deter- 
mined closer than one-tenth of a pound; it is, however, requisite, if the 

differential coefficient -f is to be determined with certainty and 
at 

accuracy. 

23 Since their results are presented in a table without attempting 

to represent it by an equation, it becomes necessary to replace by 

dt 

J p . 

-^ which can be most readily obtained as follows: for a given tem- 
J t 

perature, for example 100 deg., we may compute the ratio by taldng 

two adjacent temperatures, such as 98 deg. and 102 deg., finding the 

difference of pressure, which is to be divided by the difference of 

temperature; and the result is to be multiplied by 13.5959, because 

that is the pressure of one millimeter of mercury on one square meter. 

This result is 

^=13.5959 ^'"-^-^°^-^ =369.1 
Jt 4 

24 A number of elements entered into the determination to use 
this method and to take an interval of 4 deg. If the relation of the 
pressure to the temperature could be represented by a second-degree 
curve, that is, if such a curve were a parabola with its axis parallel 

to the axis of pressure, the ratio -f for any interval would be pre- 

Jt 

dv 
cisely equal to ^. A table of values that could be represented by 

such a curve would have constant second differences; by second 
differences are'meant the results obtained by taking (a) the differences 



340 SPECIFIC VOLUME OF SATURATED STEAM 

of successive tabular values, and (6) the differences of these differences. 
An examination of the second differences of Holborn and Henning's 
values showed great regularity between 50 deg. and 100 deg., i.e., for 
their own determinations. The second differences increased slowly; 
for intervals of 4 deg. the increase was imperceptible, for 6-deg. inter- 
vals the increase was barely perceptible, but for 10-deg. intervals 
it was very apparent. 

25 Now the possible precision of reading the height of a column 
of mercury, including allowance for variations of density, is better 
than the determination of temperature; consequently the prob- 
able error to be considered is that attributed to the determina- 
tion of temperature, namely 0.01 deg., consequently the probable 

^ J) 

error of a single determination of the ratio ~-^. To diminish the 

At 

effect of local variations this ratio was computed for each degree of 
temperature and the regularity of the results thus obtained was tested 
by taking first and second differences. Where the second differences 
showed irregularity, the values of the ratio were changed to the 
extent of 1/1000 in order to improve the regularity of the second dif- 
ferences. This process is equivalent to drawing a smooth or fair 
curve to represent physical properties obtained by observation. 

Aj) 

26 Having values of the ratio —^ for each degree of temperature 

At 

the specific volumes were computed by the thermodynamic equation 
in Par. 1. They were in turn tested for regularity by taking first 
and second differences: and again the values were changed when 
necessary to the extent of 1/1000 to improve the regularity of the 
second differences. The combined effect of both fairings is esti- 
mated not to exceed 1/500 in any case and the author believes that the 
probable error of the final determinations of the specific volumes is 
not greater than that amount for the range of 50 deg. to 200 deg. 
cent. 

27 It may further be said that having computed the values of A'pu 
at each fifth degree and plotted the results on a large diagram, no indi- 
vidual values were found to vary from a fair curve more than 1/750. 

28 Fortunately there are extant experiments on the specific 
volume of saturated steam by Knoblauch, Linde and Klebe,* made 
with such a degree of precision as to give a satisfactory check on the 
computations made by the method described. These experiments 

^Mitteilunyen iiber Forschungsarbeiten, vol. 21, S. 33, 1905. 



SPECIFIC VOLUME OP SATURATED STEAM 341 

consisted in measuring the temperature and pressure of superheated 
steam at constant volume, and the results were so treated as to give 
the volume at saturation by a sl.'tight-line extrapolation with great 
certainty. The experimenters give the following equation to repre- 
sent the properties of both superheated and saturated steam; 



p V = BT - p (l + o p) 



T 



B = 47.10; a = 0.000002; C = 0.031; D = 0.0052, 
volumes being in cubic meters per kilogram, pressures in kilograms 
per square meter, and the absolute temperature being on the centi- 
grade scale. 

29 For English units the equation may be written 



p V = 85.85 T - p {1 +0.00000976 p) 



150,300,000 _ ^^333 



the volumes being in cubic feet, the pressures in pounds per square 
foot and the temperatures in degrees fahr. 

30 Knoblauch claims for this equation a mean probable error of 
1/500, though admitting individual discrepancies of twice that amount. 
This equation applied to the computation of specific volumes of satur- 
ated steam shows a good concordance with results, computed by the 
thermodynamic equation, the greatest discrepancy being 1/300 at 165 
deg. cent. (329 deg. fahr.). 

31 Not satisfied with this apparent concordance, which after all 
was with an empirical equation which on examination showed some- 
what larger variation from individual experimental values at satura- 
tion, the author had a diagram drawn of the 32 values of the specific 
volume reported by the experimenters. The diagram was drawn to 
a very large scale, using temperatures for abscissae and logarithms 
of volumes for ordinates, and a fair curve was drawn by aid of a stiff 
spline. From readings on this curve the volumes were determined 
at 5 deg. intervals, and are set down in the accompanying table 
together with values computed by the thermodynamic equation. 

32 The greatest deviation of values in this table is 0.2 per cent, 
which is precisely the probable error assigned by the experimenters 
for their work. It may therefore be concluded that between the 
limits of temperature in this table and probably from 30 deg. to 200 
deg. cent. (86 deg. to 392 deg. fahr.), the probable error of computa- 
tions by aid of the thermodynamic equation is not in excess of 
1/500. 



342 



DISCUSSION 



COMPARISON OF EXPERIMENTAL AND COMPUTED VALUES OF THE SPECIFIC 
VOLUME OF SATURATED STEAM 



0. 


VoLDME, Cubic 


Metebs 


MPEKATURE 


Volume, Cxtbic Meters 


Experi- 




Per cent 


Experi- 




Per cent 


H 


mental 


Computed 


deviation 


^ 


mental 


Computed 


deviation 


100 


1.674 


1.671 


: +0.18 


145 


0.4458 


0.4457 


+0.02 


105 


1.421 


1.419 


+0.14 


150 


0.3927 


0.3921 


+0.15 


110 


1.211 


1.209 


+0.17 


155 


0.3466 


0.3463 


+ 0.09 


115 


1.036 


1.036 


i 0. 


160 


0.3069 


0.3063 


+ 0.20 


120 


0.8894 


0.8910 


-0.18 


165 


0.2724 


0.2729 


+ 0.18 


125 


0.7688 


0.7698 


-0.13 


170 


0.2426 


0.2423 


+0.12 


130 


0.6670 


0.6677 


-0.10 


175 


0.2168 


0.2164 


+ 0.19 


135 


0.5809 


0.5812 


1 -0.05 


180 


0.1940 


0.1941 


-0.05 


140 


0.5080 


0.5081 


-0.02 











33 This conclusion carries with it the attribution of at least the 
same degree of precision to all the properties entering into the ther- 
modynamic equation. A little consideration will show that this con- 
clusion covers all the properties given in steam-tables including the 
entropy. As an apparent exception we have the heat of the liquid 
at high temperatures which may be uncertain to the extent of ^ of 
1 per cent of itself, but as that quantity is then associated with the 
heat of vaporization the influence of such an error will be of no con- 
sequence in computations. 

34 It may therefore be expected that steam tables based on the 
present information will have permanence. 

DISCUSSION 



Prof. William D. Ennis. The reason for the extrapolation of Fig. 
3 is not quite clear to me. The ordinates of this diagram are differ- 
ences between the Holborn and Henning values for the pressure of 
saturated steam, and those given by the Peabody formula, log p =a — 



,t-100 



+ c/3 



t-ioo 



The diagram is extended to include temperatures 



above 205 deg., the Holborn and Henning limit. Why would it not 
be just as satisfactory, if the Holborn and Henning values are satis- 
factory, to extrapolate directly the curve expressing their results? 

2 There seems to be little room for uncertainty in any of the 
properties of saturated steam, excepting, possibly, the heat of the 
liquid. The maximum divergence in values for the former, comparing 
the Dieterici and the modified Regnault values adopted by Professor 



SPECIFIC VOLUME OF SATURATED STEAM 343 

Peabody, occurs at the highest temperatures: at 220 deg. (an extra- 
polated point) it is 1.31 cal. or 0.584 per cent. Now if Dieterici does 
not claim an accuracy exceeding 0.5 per cent at this temperature, and 
since Professor Peabody admits a possible fractional percentage of 
error, the true value may be within the limit of estimated error in 
both computations. The result of taldng Dieterici's values would be 
to make the computed specific volume 0.01 per cent less at 100 deg., 
0.02 per cent less at 140 deg. and 0.045 per cent less at 165 deg., than 
those tabulated by the author. The deviation from the Knoblauch, 
Linde and Klebe results would be thus generally decreased. There is 
still a possibility that the Dieterici values may be more nearly correct. 
If the preponderance of error in values of the other quantities enter- 
ing into the volume formula were in such a direction as to make the 
computed volumes too small, the lower heats of the liquid used by 
Professor Peabody might lead to an apparently better result because 
of a balancing of opposite errors. It does not seem safe to say defi- 
nitely that such may not be the fact. The uncertainty in the heat of 
the liquid at 200 deg. is 0.36 per cent rather than 0.25 per cent. The 
same uncertainty applies to the entropy of the liquid, and a possible 
error of about 0.16 per cent to the entropy of vaporization. The 
entropy of saturation at this temperature may then be wrong to as 
great an extent as 0.23 per cent, or -gr • This would introduce a barely 
noticeable error into computations involving vapor cycles. 

3 It is questionable whether permanence, in a matter of this kind, 
is as desirable as a standard, a flexible standard. Would it not be 
within the scope of the Society to cooperate with national engineering 
organizations abroad in the preparation of an international steam 
table for saturation and superheat, embpdying the generally accepted 
values and subject to modification whenever an undisputed conclu- 
sion is reached on the one or two remaining doubtful quantities? 

The Author. In reply to Professor Ennis I will first explain 
that Holborn and Henning give a table of pressure for each degree 
of temperature, and make no use of an equation except as a means 
of fairing their results, for which purpose they chose Thiesen's 
equation which gives divergent values for higher temperatures. 
Now it happens that the equation which I chose gives values which 
converge toward Holborn and Henning's values so that it is possible 
to draw an extrapolation diagram as shown in my paper. 

2 Secondly, I wish to say that Dr. Henning has kindly sent me 
in advance of publication an abstract of results which he has recently 



344 DISCUSSION 

obtained for the heat of vaporization of water from 100 deg. to 180 
deg. cent. His memoir will soon be available and will show that 
the results which I have deduced from Dr. Davis' values of the total 
heat, show a close concordance with these new experimental values. 
It is not unlikely that we may have a conclusive determination of the 
remaining quantity, heat of the liquid, but as stated in my paper the 
concordance of all quantities involved in the computation of steam 
tables is even now very satisfactory so that there is no reason to 
anticipate any necessity for changing tables for engineers. 



No. 1245 

SOME PROPERTIES OF STEAM 

By Prof. R, C. H. Heck, New Brunswick, N. J. 
Member of the Society 

The purpose of this paper is to present some recent experimental 
results as to two of the fundamental thermodynamic properties of 
water and steam, and to make certain comparisons between these 
determinations and the older values used in our steam tables. The 
two properties considered are, the relation between pressure and 
temperature of saturated steam, and the specific heat of water. 

THE PRESSURE-TEMPERATURE RELATION 

2 This relation is, from the point of view of experimental deter- 
mination, the simplest of the properties of steam, and with accurate 
instruments and adequate skill can be very precisely measured. For 
this reason, the results obtained by various experimenters differ by 
relatively small amounts, and in discussing them we take up a ques- 
tion in the realm of scientific accuracy rather than one concerning 
effectively correct values for ordinary tex3hnical use. For certain 
purposes, however, it is most important that this relation be truly 
and accurately known. 

3 In Annalen der Physik, 1907, vol. 22, p. 609 to 630, is published 
a paper by F. Henning, On the Saturation Pressure of Steam, in 
which are gathered together all the determinations that have been 
made on this relation, from Magnus and Regnault down to that time. 
These are compared by means of curves, which show, to a large scale, 
their departures from an assumed standard of reference. This stand- 
ard is the formula of Thiesen, 

(t + 273) log / = 5.409 (t - 100) - 0.508 X lO"' [(365 - t)* - 265*] 
760 

where t is centigrade temperature and p is pressure in millimeters 
of mercury. From the comparison and discussion the conclusion was 
reached that up to 100 deg. cent, this formula is to be accepted. 

Presented at the Spring Meeting, Washington, May 1909, of The American 
Society op Mechanical Engineers. 



346 SOME PROPERTIES OF STEAM 

while above 100 deg. the determinations of Regnault are best — not as 
set forth by his formula, but as worked over by Henning, from a 
selection of his more reliable observations. 

4 A new and very accurate determination by Holborn and Hen- 
ning, over the range from 50 deg. to 200 deg. cent., is fully described in 
Annalen der Physik, 1908, vol. 26, p. 833 to 883, in a paper, On the 
Platinum Thermometer and the Saturation Pressure of Steam, while 
in Zeitschrift des Vereins deutscher Ingenieure, February 20, 1909, is 
given a brief presentation and comparison of results. Exceedingly 
close agreement is shown between these new observations, the recom- 
puted Regnault values, and the work of Knoblauch, Linde, and Klebe 
— see Table 3 in Zeitschrift article. The final result is a table giving 
p for every degree from deg. to 205 deg. cent., which follows Thiesen's 
formula up to 50 deg., and embodies the authors' work from that point. 

5 This table is here reproduced in Table 1, but with pressure con- 
verted to pounds per square inch and interpolated for every degree 
fahrenheit from 32 deg. to 402 deg., or to just past 250 lb. abs. Later 
the writer hopes to extend this table, carrying forward the line of the 
Holborn-Henning determination in comparison with the observa- 
tions of Regnault and others. This can be done even up to a pressure 
of 1000 lb. with sufficient accuracy for all practical purposes. 

6 In the work of conversion and interpolation, it was necessary to 

carry the numbers to a higher degree of apparent accuracy, or to use 

more significant figures, than any experimental precision would call 

for. Without a mathematical formula, a function of this sort can be 

carried forward only by carefully smoothing out the differences until 

those of the second order follow a continuous rate of change. In this 

operation, the first differences were brought to a sufficient degree of 

smoothness to furnish effectively accurate values of the rate of change 

dx) 
of pwithf; and this differential coefficient, —is also given in Table 1. 

It may be considered absolutely correct (as a derivative) within about 
four or five units in the last place, while as between successive values 
the closeness is much better. This is less precise than might be 
desired, but it is accurate enough for use in calculating specific volume, 
since the thermal data there involved are not of any greater degree of 
reliability. 

7 In Fig. 1 is given a comparison between the pressures in Table 1 
and some hitherto generally used values. The base is temperature 
fahrenheit, the ordinate the difference between the other value of p 
and that in Table 1. Curve 1, for the range up to 225 deg. fahr., is 



SOME PROPERTIES OP STEAM 



347 



3 + 





















































































CL/ 
















•/ 


<-♦ * 














* 


A 










S-, 






»/"" 












cc 


s 


!d 




* 












X 


■^ 


<f 
















'X 


tt 


f , 
















^H 




















S. 


















\ 


^ 


















^ 


















a^ 


^ 


















\ 


















\ 




















V 


















\ 


















\ 


















V 




-^ 














\ 






^ 


















\^ 


















\ 




















• 


































■ 




































^ 


















/ 


















/ 


















"/ 


















r 

































































































































































P z 



+ I 



2 Q. 3 



348 SOME PROPERTIES OF STEAM 

drawn to the large scale at the left, and shows how Regnault's formula 
drops below the new determination. The curves at 2 have the ordi- 
nate scale at the right, only one-tenth as large as that for 1. The 
letter R marks the "standard" Regnault curve, here plotted from 
the table in Roentgen's Thermodynamics, which happened to be the 
most convenient in its manner of expression: note the abrupt change 
at about 380 deg. fahr. Curve P shows Peabody's values, which are 
based on Regnault, but with revised computations, and depart quite 
decidedly from the older table above 325 deg. The scattering of the 
points above that temperature is due to the coarseness of numerical 
expression, Peabody giving but one decimal place for the higher pres- 
sures. The curve is simply sketched through this band of points. 

8 Holborn and Henning do not attempt to devise a formula, but 
base their table on a method of graphical interpolation. It will be 
noted that Curve 1 shows a faint waviness, indicating some departure 
from perfect mathematical smoothness; but the extreme smallness of 
the irregularities is really a proof of the skill with which the original 
interpolation was made. 

THE SPECIFIC HEAT OF WATER 

9 In Fig. 2 are plotted several important curves for the specific 
heat of water — the true or instantaneous, not the mean value. 
Curve R shows Regnault's formula, which in fahrenheit units is, 

c = 1 + 0.0000222 (t - 32) -|- 0.000000278 (t - 32)^ 

This curve differs radically from the newer and true determination of 
the specific heat over the lower part of the range, as shown by the 
other curves. 

10 Curve B represents the experiments of H. T. Barnes and 
associates; these are described briefly in Proceedings Royal Society, 
1900, vol. 67, fully in Phil Trans. Roy. Soc, 1902, vol. A 199; while 
in Physical Review, 1902, vol. 15, there is a description of the 
determination on supercooled water, which was carried to — 5 deg. cent., 
and the tabulated values for the whole range up to 95 deg. cent. The 
body of this work was done by a continuous method, water flowing 
through a small tube and absorbing heat which was electrically sup- 
plied and measured; for the range below freezing, a method of mixing 
was found necessary. 

11 Curve P, which begins at 140 deg. fahr., shows the values used 
by Peabody above this temperature; below it he accepts the work of 
Barnes. Peabody's line — it is almost straight — is based on Reg- 



SOME PROPERTIES OF STEAM 



349 




350 SOME PROPERTIES OF STEAM 

nault's experiments: but it hardly seems reasonable to make c thus 
an almost straight-Une function of t. 

12 Curve D shows the very important experiments of Dieterici, 
described in Annalen der Physik, 1905, vol. 16. In these a small 
body of water, pure and free from air, was sealed in a tube of quartz. 
This little cartridge was heated to a certain desired temperature, 
then dropped into a Bunsen ice calorimeter, where the heat given off 
in its cooling to deg. cent, is measured. The highest temperature 
reached was about 300 deg. cent. The drawback in tliis method is 
the relatively large heat capacity of the quartz tube, which has to be 
very carefully determined. From 100 deg. fahr. upward, Dieterici 
finds that Ms results conform very well to a parabolic equation like 
that of Regnault, which for fahrenheit units has the constants, 

c = 0.99827 - 0.0000576 (t - 32) + 0.00000064 (t - 32)^ 

Below 100 deg. fahr., tabulation from graphical interpolation is pref- 
erable to expression by formula. A numerical comparison of the 
several curves is given in Table 2. 

DIFFERENT HEAT UNITS 

13 Before discussing these data, something must be said as to the 
unit of heat measurement. Regnault intended to use the heat 
capacity of water at 15 deg. cent, as the heat unit — in other words, the 
15-deg. calorie — but it was not until long after his time that the true 
manner of variation of the specific heat over the lower range of 
ordinary temperatures was either clearly perceived or accurately 
measured. Barnes' values are based on unity at 16 deg. cent., and 
it will be noted that the B curve on Fig. 2 crosses the base-line at 
just about 16 deg. cent, (the two short vertical cross-lines near 60 deg. 
fahr. are at 15 deg. and 16 deg. cent.). The now generally used 
numerical values of the mechanical equivalent of heat, 427 m-kg. or 
778 ft. lb. are based on a heat unit at 15 deg. cent or 59 deg. fahr. 

14 Dieterici's results are expressed in the mean calorie, which is 
one one-hundredth of the heat required to raise 1 kg. of water from 
deg. to 100 deg. cent.; and his specific heat values check up to an 
average of unity over this range. Graphically, on Fig. 2, his curve 
cuts the 15-deg. cent, ordinate at 0.0012 below the unity base-line. 
In a special expei iment, with electrical measui ement analogous to that 
used by Barnes, he made the mechanical equivalent of the mean 
calorie bear to our standard Rowland value for the 15-deg. calorie the 



SOME PROPERTIES OF STEAM 351 

ratio of the numbers 419.25 to 418.8, or 1.0011 to 1.00000. Dis- 
regarding some micertainties which may exist in the minds of physi- 
cists as to the finality of this determination, it seems reasonable, for 
engineering purposes, to use this 0.0011 or 0.11 per cent correction 
in order to change from one system of units to the other. 

15 The amount of attention here paid to this small point is justi- 
fied by the importance given to it through the introduction of the 
mean calorie to the Society in the recent paper on The Total Heat of 
Saturated Steam, by Dr. H. N. Davis. Personally, I think we had 
better transform heat values in this unit by means of the ratio just 
offered, rather than change our mechanical equivalent of heat from 
778 to 778.9. 

16 Now the specific heat is the ratio of a certain absolute quantity 
of heat to an assumed unit quantity. If we use a larger unit, the 
ratio will be smaller, and vice versa. Assuming that the mean calorie 
is 1.0011 of the 15-deg. calorie, we change Dieterici's values to the 
15-deg. unit if we increase them by 0.11 per cent. This would raise 
his curve to the dotted position on Fig. 2, and change his formula to 
c = 0.99938 - 0.00005766 (t - 32) + 0.0000006407 (t - 32\) 

SPECIFIC HEAT OF WATER — CONCLUSION 

17 It is pretty safe to say that the Holborn-Henning results for 
pressure and temperature, set forth in Table 1, are final, and that 
this relation is now known surely and accurately enough for all pur- 
poses of practical science. But in regard to the specific heat of water 
we are yet confronted by one of the annoying uncertainties which 
have so long surrounded many parts of this subject. Dieterici 
claims an experimental accuracy ranging from 0.1 per cent at low 
ranges to 0.5 per cent at high ranges of temperature; but his method 
is open to the objection that two heat-capacities have to be measured 
and their difference used. 

18 In spite of some small doubt as to the accuracy of Dieterici's 
results, and a faint suspicion that his curve may rise too rapidly, 
I am of the opinion that his determination is to be accepted instead of 
Regnault's. Further, the idea of an increasing rate of increase in c, 
as expressed by a second-degree equation, seems to be far more rea- 
sonable than that of a nearly constant rate of increase. 

19 It is hardly probable that the heat capacity of water will ever 
be so accurately determined that the heat for the external work of 
expanding the water will be more than a small fraction of the prob- 
able error in heat measurement. 



352 



SOME PROPERTIES OF STEAM 
TABLE 1 THE PRESSURE-TEMPERATURE RELATION 



t \ p '• dp/dt 


« ' p dp/dt t 


p j dp/dt t 


p dp/dt 




76 0.4433 0.01467 121 

77 0.4582 0.01510 122 


1.7362 0.04815' 166 
1.7849 0.0493 167 


5.459 0.1277 


32 0.08860.003575 


5.588 0.1302 


33 0.09220.00371 


78 0.4735 0.01554 123 


1.8348 0.0505 168 


5.719 0.1327 


34 0.0960( 


). 003845' 


79 1 0.4893 0.01600 124 


1.8859 0.0517 169 


5.853 0.1353 


35 


0.0999( 


). 003985' 


80 J 0.5055 0.01646 125 


1.9382 


0.05295 170 


5.990 0.1380 


36 


0.1039 


J. 00413 


81 0.5222 0.01694, 126 


1.9918 


0.0642 1 171 


6.129 0.1407 


37 


0.1081 


3.00428 


82 0.5394 


0.01742 127 


2.0466 0.05545 172 


6.271 0.1434 


38 0.1125 


3.00443 1 


83 0.5570 


0.01792 128 


2.1027, 0.05675 173 


6.416 0.1462 


39 0.1170 


0. 004585; 


84 0.5752 


0.01844 129 


2.1601 0.0581 174 


6.564 0.1490 


40 0.1217 


0.004745 


85 


0.5939 


0.01898 130 


2.2189 0.05945 175 

! •< 


6.714 0.1519 


1 


0.00491 


86 


0.6132 


0.01952 131 


2.2790 0.0608 176 


6.868 0.1548 


42 i 0.1315 


0.005075 


87 


0.6330 


0.02008 132 


2.3406 0.06216 177 


7.024 0.1677 


43 0.1367 


0.00525 


88' 


0.6533 


0.02065 133 


2.4033 0.06355 178 


7.183 0.1607 


44 0.1420 


0.00543 


89 


0.6743 


0.02123 134 


2.4675 0.06496 179 


7.346 0.1637 


45 [ 0.1475 


0.00561 


90 


0.6958 


0.02182 136 


2.5332 0.06646 180 


7.511 


0.1668 


46 


0.1532 


0.00580 


91 


0.7179 


0.02243 136 


2.6004 0.0680 181 


7.679 


0.1699 


47 


0.1591 


0.00600 


92 


0.7406 


0.02305 137 


2.6692 0.0696 \ 182 


7.850 0.1730 


48 


0.1652 


0.00620 


93 


0.7640 


0.02368 138 


2.7396 0.0712 : 183 


8.025 0.1762 


49 


0.1715 


0.00641 


94 


0.7880 


0.02432 139 


2.8116 


0.0728 184 


8.203 0.1794 


50 


0.1780 


0.00663 


95 


0.8127 


0.02498. 140 


2.8861 


0.0744 185 


8.384 0.1827 


51 


0.1847 


0.00685 


96 


0.8380 


0.02566 


141 


2.9603 


0.0760 186 


8.568' 0.1860 


62 


0.1917 


0.00708 


97 


0.8640 


0.02635 


142 


3.0371 


0.0776 187 


8.756 0.1894 


53 


0.1989 


0.00731 


98 


0.8907 


0.02705 


143 


3.1155 


0.0793 188 


8.947 0.1929 


64 


0.2063 


0.00754 


99 


0.9181 


0.02776 


144 


3.1956 


0.0810 189 


9.142 


0.1964 


55 


0.2104 


0.00778 


100 


0.9462 


0.02849 


145 


3.2775 


0.0828 190 


9.340 


0.1999 


66 


0.2219 


0.00803 


101 


0.9751 


0.02923 


146 


3.3612 


0.0846 


191 


9.542 


0.2036 


57 


0.2301 


0.00829 


102 


1.0047 


0.02999 


147 


3.4467 


0.0864 


192 


9.747 


0.2072 


58 


0.2385 


0.00856 


103 


1.0350 


0.03077 


148 


3.5341 


0.0883 


193 


9.956 


0.2109 


59 


0.2472 


0.00883 


104 


1.0662 


0.03157 


149 


3.6233 


0.0902 


194 


10.169 


0.2147 


60 


0.2661 


0.00911 


105 


1.0982 


0.03240 


150 


3.7141 


0.0921 


195 


10.385 


0.2185 


61 


0.2653 


0.00939 


106 


1.1310 


0.03325 


151 


3.808 


0.0940 


196 


10.606 


0.2224 


62 


0.2749 


0.00968 


107 


1 . 1647 


0.0341 


152 


3.903 


0.0960 


197 


10.830 


0.2263 


63 


0.2847 


0.00998 


108 


1.1992 


0.0350 


153 


4.000 


0.0980 


198 


11.058 


0.2303 


64 


0.2948 


0.01029 


109 


1.2347 


0.0359 


154 


4.099 


0.1001 


199 


11.291 


0.2343 


65 


0.3053 


0.01061 


110 


1.2711 


0.03685 


155 


4.200 


0.1022 


200 


11.527 


0.2384 


66 


0.3161 


0.01094 


111 


1.3084 


0.03775 


166 


4.303 


0.1043 


201 


11.767 


0.2425 


67 


0.3272 


0.01127 


112 


1.3466 


0.0387 


157 


4.408 


0.1064 


202 


12.013 


0.2467 


68 


0.3386 


0.01161 


113 


1.3858 


0.0397 


158 


4.516 


0.1086 


203 


12.261 


0.2509 


69 


0.3504 


0.01196 


114 


1.4260 


0.0407 


159 


4.625 


0.1108 


204 


12.514 


0.2562 


70 


0.3625 


0.01232 


115 


1.4671 


0.0417 


160 


4.737 


0.1131 


205 


12.771 


; 0.2596 


71 


0.3750 


0.01269 


116 


1.5093 


0.0427 


161 


4.852 


0.1154 


206 


13.033 


0.2639 


72 


0.3879 


0.01307 


117 


1.5525 


0.04375 


162 


4.968 


0.1178 


207 


13.29S 


0.2683 


73 


0.4012 


0.01345 


118 


1.5968 


0.0448 


163 


5.087 


0.1202 


208 


13.566 


0.2728 


74 


0.4148 


0.01384 


119 


1.6421 


0.0459 


164 


5.209 


0.1227 


209 


13.84£ 


0.2783 


75 


0.4289 


0.01425 


120 


1.6886 


. 0.0470 


165 


5.332 


0.1262 


210 


14,124 


0.2819 



SOME PROPERTIES OF STEAM 



353 



TABLE 1.— Continued 



t 


" 


dp/dt 


1 ' 


P 


dp/dt 


t p 


dp/dt 


1 t 


P i 


dp/dt 


211 


14.408 0.2866 


256 


33.085 


0.5677 


301 67.99 i 1.016 


346 


127.67 


1.675 


212 


14.697 0.2914 


257 


33.657 


0.5758 302 69.01 1.027 


347 


129.35 


1.693 


213 


14.991 0.2962 


258 


34.236 


0.5840 


303 70.06 


1.0395 


348 


131.05 


1.711 


214 


15.290 0.3011 


259 


34.824 


0.6922 


304 71.09 


1.052 


349 


132.77 


1.729 


215 


15.594 


0.3061 


260 


36.420 


0.6006 


305 [ 72.16 


1.065 


350 

1 


134. 6li 


1.746 


216 


15.902 


0.3111 


261 


36.026 


0.6088 


306 


73.22 


1.0775 


1 

361 


136.26' 


1.764 


217 


16.215 


0.3162 


262 


36.638 


0.6172 


307 


74.31 


1.090 


352 


138.04 


1.782 


218 


16.534 


0.3214 


263 


37.259 


0.6266 


308 


76.40 1.103 


353 


139.83 


1.800 


210 


16.858 


0.3266 


264 


37.888 


0.6341 


309 


76.51 1.116 


354 


141.64 


1.818 


220 


17.187 


0.3319 


265 


38.526 


0.6426 


310 


77.64 1.129 


356 


143.46 


1.836 


221 


17.621 


0.3372 


266 


39.173 


0.6513 


311 


1 
78.77 1.142 


356 


145.31 


1.866 


222 


17.860' 0.3426 


267 


39.828 


0.6600 


312 


70.92 1.155 


357 


147.17 


1.874 


223 


18.205 0.3480 


268 


40.492 


0.6688 


313 


81.08 1.169 


358 


149.06 


1.893 


224 


18.556 0.3535 


269 


41.165 


0.6777 


314 


82.26 1.182 


369 


160.96 


1.912 


225 


18.913 


0.3591 


270 


41.848 0.6868 


316 


83.44 1.195 


360 


162.88 


1.031 


226 


19.275 


0.3648 


271 


42.54 


0.6960 


316 


84.66 1.209 


361 


154.82 


1.961 


227 


19.643 


0.3705 


272 


43.24 


0.7052 


317 


86.86 1.223 


362 


156.78 


1.970 


228 


20.017 


0.3763 


273 


43.95 1 


0.7145 


318 


87.09 1 1.237 


3&3 


158.76 


1.990 


229 


20.396 


0.3821 


274 


44.67 , 


0.7239 


319 


88.34 1.261 


364 


160.76 


2.010 


230 


20.781 


0.3880 


276 


45.40 


0.7334 


320 


89.60 1.266 


365 


162.78 


2.029 


231 


21.172 


0.3940 


276 


46.14 , 


0.7430 


321 


00.87 I 1.280 


366 


164.82 


2.049 


232 


21.568 


0.4000 


277 


46.88 


0.7527 


322 


92.16 1.296 


367 


166.88 


2.069 


233 


21.970 


0.4061 


278 


47.64 


0.7625 


323 


03.46 1.309 


368 


168.96 


2.089 


234 


22.379 


0.4123 


279 


48.41 


0.7725 


324 


94.78 1.324 


369 


171.06 


2.108 


235 


22.794 


0.4185 


280 


49.10 


0.7826 


326 


96.17 1.339 


370 


173.18 


2.128 


236 


23.216 


0.4248 


281 


49.98 


0.7926 


326 


07.46 1.364 


371 


175.31 


2.148 


237 


23.644 


0.4312 


282 


50.77 


0.8028 


327 


08.81 1.369 


372 


177.47 


2.168 


238 


24.079 


0.4377 


283 


51.58 


0.8131 


328 


100.19 ! 1.384 


373 


179.65 


2.189 


239 


24.520 


0.4442 


284 


52.40 


0.8236 


329 


101.68 i 1.400 


374 


181.85 


2.210 


240 


24.967 


0.4508 


286 


53.23 


0.8340 


330 


102.99 i 1.416 


375 


184.07 


2.231 


241 


25.421 


0.4575 


286 


54.07 


0.8446 


331 


104.41 1.430 


376 


1 
186.31 


2.262 


242 


25.882 


0.4643 


287 


54.92 


0.8663 


332 


106.86 1.445 


377 


188.58 


2.274 


243 


26.350 


0.4711 


288 


56.78 


0.8661 


333 


107.30 1.461 


378 


190.86 


2.296 


244 


26.825 


0.4780 


289 


66.65 


0.8770 


334 


108.77 1.477 


379 


193.17 


2.318 


245 


27.307 


0.4850 


290 


57.63 


0.8880 


336 


110.26 1.493 


380 


105.50 


2.341 


246 


27.795 


0.4920 


291 


58.42 


0.8991 


336 


111.76 1 1.509 


381 


197.86 


2.364 


247 


28.290! 0.4991 


292 


59.33 


0.9103 


337 


113.27 1.525 


382 


200.23 


2.387 


248 


28.793 0.6063 


293 


60.25 


0.9216 


338 


114.81 1.542 


383 


202.63 


2.410 


249 


29.303i 0.5136 


294 


61.17 


0.9330 


339 


116.36 ' 1.568 


384 


205.05 


2.433 


250 


29.820 0.5210 


295 


62.11 


0.9445 


340 


117.02 1.674 

I 


386 


207.49 


2.466 


251 


30.345 0.5285 


296 


63.06 


0.9661 


341 


1 

119.50 j 1.601 


386 


209.06 


2.479 


252 


30.877 0.5361 


297 


64.03 


0.9678 


342 


121.10 1.607 


387 


212.46 


2.502 


253 


31.417 0.5438 


298 


65.00 


0.9796 


343 


122.72 1.624 


388 


214.96 


2.525 


254 


31.965 0.5517 


299 


65.98 


0.9916 


344 


124.36 1.641 


389 


217.50 
220.06 


2.548 


2ft6 


32.621 0.6596 


300 


66.98 


1.0036 


345 


126.00 ! 1.658 


390 


2.671 



354 



SOME PROPERTIES OF STEAM 



TABLE 1. — Continued 



t p 1 dp/dt 


t 


P 


dp/dt 


t 


P 


dp/dt 


t 


p j dp/dt 


391 222.64 2.594 

392 225.24 2.617 

393 227.87 2.641 

394 230.52 2.664 


395 
396 
397 


233.20 
235.90 
238.62 


2.68/ 
2.71*1 
2,735 


398 
399 
400 


241.37 
244.14 
246,93 


2.759 
2.783 
2.807 


401 
402 


249.75 
252.60 


2.832 

2.857 



TABLE 2 THE SPECIFIC HEAT OF WATER 



Tempehatukh 


Regnault 


Dieterici 


Barnes 


Peabody 


Cent. 


Fahr. 


-5 


23 
32 
41 
50 
69 

68 
77 
86 
95 
104 

122 
140 
158 
176 
194 
212 

248 
284 
320 
356 
392 

428 
464 
500 
536 
672 






1.0158 

1.0094 

1.00530 

1.00230 

1.00030 

0.99895 
0.99806 
0.99759 
0.99735 
0.99735 

0.99800 
0.99910 
1.00035 
1.00166 
1.00305 
(1 .0044) 







1.00000 


1.0075 
1.0037 
1.0008 
0,9987 

0.9974 
0.9970 
0.9971 
0.9972 
0.9974 

0.9983 
0.9995 
1.0012 
1.0032 
1.0057 
1.0086 

1.0167 
1.0244 
1.0348 
1.0468 
1.0605 

1,0758 
1.0928 
1.1115 
1.1318 
1.1538 




+5 




10 


1.00049 




15 




20 


1.00116 




25 




80 


1.00201 




.'iC 




U) 


1.00304 

1.00425 
1.00564 
1.00721 
1.00896 
1.01089 
1.01300 

1.01776 
1.02324 
1.02944 
1.03636 
1.04400 

1.05236 
(1.06144) 
(1.07124) 
(1.08176) 
(1.09300) 




so 




60 
70 

80 

90 

100 

120 


0.99940 
1.00150 
1.00415 
1.00705 
1.01010 

1.01620 


140 




1.02230 


160 




1.02850 


180 





1.03475 


200 





1 .04100 


220 




1,04760 


240 






260 


1 


280 




300 











Resnault: from formula, par. 9, Above 200 deg, cent, hia formula is an extrapolation, 

Dieterici: from table in original publication, computed by formula from 40 deg, cent, upward. 

Bamea: from Phyeical Review, with last value extrapolated. 

Teabody: from Steam and Entropy Tables, p. 10. 

Dieterici: values in mean calor •^ 'heat units), others in 15 deg. cent, units. 



SOME PROPERTIES OF STEAM 355 

DISCUSSION 

Dr. Sanford A. Moss. In Regnault's original paper on pressure 
and temperature of saturated steam was given an empirical formula, 
first suggested by Roche, but reconstructed by Regnault and called 
"Formula K." This represented Regnault's results more closely 
than any other single formula and has also been shown to represent 
other experimental results, particularly those for very high pressures 
and temperatures. This formula was discussed by Ramsey and Young 
in the Philosophical Transactions, vol. 183, 1892, page 111, and also 
in London Engineering, vol. 83, January 4, 1907. I have given 
some discussion of this matter in the Physical Review, vol. 25, no. 6, 
December 1907. It would be interesting if Professor Heck would 
give a comparison of the values of his table with those computed by 
this formula. If it can be demonstrated, as is the conclusion in the 
papers above mentioned, that this formula represents all of the experi- 
mental results very closely, it is highly desirable that it be used. 
Thermodynamic computations can be carried on very readily, if we 
have a single formula for the entire range. 

Prof. G. A. Goodenough.* Since the experimental results of 
Holborn and Henning have been generally accepted, it seems highly 
desirable to have an analytical relation that will express these 
results with a sufficient degree of accuracy. For this purpose the 
formula of Bertrand, 

T 

logp = k -n log ^^-^ [1] 

seems to be quite suitable. In this formula k, n and a are constants 
and T denotes the absolute temperature. 

2 The values of the constants can be so chosen that the formula 
will give fair results throughout the ordinary range of temperatures 
with one set of constants. A closer agreement, however, is obtained 
by dividing the temperature range into three parts. The values of 
the constants are: 

For t = 32 deg. - 90 deg., k = 6.23167, a = 140.1 n = 50 
For t = 91 deg. - 237 deg., k = 6.30217, a = 141.43 n = 50 
For I = 238 deg. - 400 deg., k = 6.27756, a = 140.8 n = 50 

'Associate Professor Mechanical Engineering, University of Illinois, Urbana, 111. 



356 



DISCUSSION 



3 The derivative — takes the simple form 
dt 



dp 
dt 



from which follows 



dp 
lif 



pna 
T{T -a) 

pna 
T-a 



[2] 



[3] 



This relation is important since the product T ^ appears in the Clapey- 
ron-Clausius formula for steam volume. Table 1 shows a comparison 

of the values of p and calculated from Formulae 1 and 2, respect- 
dt 

ively, with the values obtained by Professor Heck. 

4 It will be seen that the maximum difference between the cor- 
responding values of p is about 0. 1 per cent. The maximum difference 
between the corresponding values of the derivative is greater. 



COMPARISON OF VALUES OF p AND 



dp 





t 


P 
Beck's value 


P 
Bertrand'b 

FORMULA 


1 

dp/dt 

Heck 


dp/dt 

Bertrand's 

FORMULA 




32 


0.0886 


0.0885 


0.003575 


0.003588 




50 


0.1780 


0.1781 


0.00663 


0.006625 




75 


0.4289 


0.4288 


0.01425 


0.014245 




100 


0.9462 


0.9455 


0.02849 


0.028575 




125 


1.9382 


1.9387 


0.05295 


0.052921 




150 


3.7141 


3.7144 


0.0921 


0.092041 




175 


6.714 


6.711 


0.1519 


0.15165 




200 


11.527 


11.523 


0.2384 


0.23843 




225 


18.913 


18.921 


0.3591 


0.35985 




250 


29.820 


29.834 


0.5210 


0.52039 




275 


45.40 


45.382 


0.7334 


0.73248 




300 


66.98 


66.94 


1.0035 


1.0027 




325 


96.11 


96.07 


1.339 


1.3390 




350 


134.51 


134.50 


1.746 


1.7489 




375 


184.07 


184.18 


2.231 


2.2394 


_ 


400 


246.93 


247.20 


2.807 


2.8167 



Evidently this is due to the fact that the values of — were obtained 

dt 

by Professor Heck (and also by Professor Peabody) by a method 
involving some approximation; and it is likely that the values calcu- 
lated from Formula 2 are the more reliable. 



SOME PROPERTIES OF STEAM 



357 



5 A great advantage of Bertrand's formula lies in the ease with 
which it may be used in calculation. This is shown by the following 
scheme, which is copied from calculations recently made for the 
engineering experiment station of the University of Illinois : 

< = 311 312 313 314 

7 = 770.58 771.58 772.58 773.58 

r-a = 629.78 630.78 631.78 632.78 

logT= 2.8868177 2.8873810 2.8879435 2.8885052 
2.7998779 



log T-a 

T 
log 



50 log 



log p = A; - 50 log 



T-a 

T 
T-a~ 
fc = 

T 
T-a^ 



2.7991889 
0.0876288 

4.381440 
6.27756 . 
1.896120 



0.0875031 



4.375155 



2.8005659 
0.0873776 

4.368880 



2.8012527 
0.0872525 

4.362625 



los 



pna 
T-a 



p= 78.726 

Iogna= 3.8475727. 

log pna= 5.7436927 

dp 

logT— = 2.9445038 
^ dt 



1.902405 

79.874 



1.908680 
81.036 



1.914935 
82.212 



5.7499777 
2.9500998 



5.7562527 
2.9556768 



5.7625077 
2.9612550 



6 In conclusion, it may be stated that Bertrand's formula seems 
to have a wide range of applicabiUty, and with proper choice of con- 
stants can probably be used equally well for nearly all saturated vapors. 



The Author. The preparation of this paper was an early step in 
a special study of the properties of steam upon which the writer 
has been engaged for some months; the object of this investigation 
has been to formulate more precisely our knowledge of the subject 
up to the limit of accurate experiment, at about 400 deg. fahr., and 
to extend the relations into the higher ranges, even up to the critical 
temperature. It has been found that Thiesen's formula, by modifi- 
cation of the two coefficients or constants, can be made to fittheHol- 
born-Henning table exactly from 100 deg. to 200 deg. cent., and that 
this modified equation holds good up to about 450 deg. fahr. Above 
that point the curve begins to rise more rapidly, and even crosses the 
original Thiesen curve. Below 212 deg., the relation can be very 
exactly extended by means of a small adjustment from the Thiesen 
values. 

2 The formula presented by Professor Goodenough fits the data 
very well; but for precise calculation it has the disadvantage that at 



358 DISCUSSION 

each point of change in constants there will be a break in the curve 

dp . 

of the derivative --rT, The Thiesen formula has a forbiddingly com- 
plex expression for its derivative, and would be much less convenient 
for regular use, if the calculation of — were an operation to be under- 
taken frequently. 

3 The writer has tried formula K , which is recommended by Dr. 
Moss. In the very high range it runs low in p, dropping beneath the 
modified Thiesen formula, which itself is not high enough for the 
more reliable data. 

dp 

4 The values of — in Table 1, found simply by the method of 

differences, are not precise enough for close work with Clapeyron's 
equation, the irregularities resulting from the use of them running as 
high as nearly one-fifth of one per cent. 



No. 1246 

A NF W DEPARTURE IN FLEXIBLE STAY-BOLTS 

By H. V. WiLiiE, Philadelphia, Pa. 
Member of the Society 

There is practically no literature on the subject of stay-bolts, and 
this is particularly true of flexible stay-bolts. The increasing size 
and pressure of boilers make this subject of vital importance to rail- 
roads and to those responsible for the management of that type of 
boiler in which the firebox is stayed by a large number of bolts. 

2 The boiler of the consolidation locomotive, now the prevailing 
type in freight service, contains about 1000 bolts less than 8 in. long 
and about 300 of greater length. The large types of Mallet compound 
locomotives now meeting with much favor have a much larger num- 
ber, there being 1250 short and 300 long bolts in locomotives recently 
constructed. 

3 In recent years some form of flexible stay-bolt, that is, one 
having a movable joint, has been very extensively used in the break- 
ing zone of locomotive boilers, but their high cost and the difficulty of 
applying them, their rigidity from rust and scale, and the fact that 
their use throws an additional service on the adjacent bolts because 
of lost motion, has militated against their more general use. 

4 It is well known that stay-bolts fail, not because of the ten- 
sional loads upon them, but from flexural stresses induced by the vibra- 
tion resulting from the greater expansion of the firebox sheets than 
of the outside sheets, but notwithstanding the general acceptance of 
this theory, engineers have designed stay-bolts solely with respect to 
the tensional loads. It is quite general practice, it is true, to recess 
the bolts below the base of the thread, and this has effected a 
slight reduction in the fiber stress, but practically no effort has been 
made to design a bolt to meet the flexural stresses or even to calculate 
their magnitude. This is su^p^^"•ing in view of the simplicity of the 
calculations to which the ordinary formulae for flexure apply. 

Presented at theSpring Meeting, Washington, May 1909, of The American 
SociBTT OF Mechanical Engineebs. 



360 DISCUSSION 

5 Let 

F = fiber stress. 

E = modulus of elasticity. 

I = moment of inertia. 

D = diameter. 

A^ = deflection. 

L = length. 

W = load. 

We then have 

2 F I 

W=^-^ (1) 

D L 

N = ~-^ (2) 

Z E I 

Substituting 

2F U 

''-ZED ''' 

F=3^-°-^* (4) 

2 U 

This formula shows that the stress increases in direct proportion to 
the diameter and decreases as the square of the distance between the 
sheets. 

6 The application of the formula to service conditions gives the 
following stresses: 

Conditions : Bolt spacing, 4 in. centers. 

Assumed expansion, 4/100 in. 
Length of bolt, 6 in. 



Type 



Diameter of Bolt 



Flexural Stress 



Iron li in. 

Iron 1 " 

Iron i " 

Spring steel i 1 in. ends /^ in. stem 



51,500 
45,000 
39,400 
19,700 



7 Iron is universally employed in the manufacture of these bolts 
and it is not good practice to exceed a fiber stress of 12,000 lb. per 



* Testing of Stay-bolt Iron. H. V. WiUe, A. S. T. M.. vol. 4, 1904. 



NEW DEPARTURE IN FLEXIBLE STAT-BOLTS 



361 



square inch. It is apparent that stay-bolts in the zone which meets 
the expansion of the sheets are stressed above the elastic limit and 
must necessarily fail from fatigue. Fractures always originate at 
the outside sheet at the point where the bending moment due to the 
movement of the furnace sheets is greatest. 

8 The fractures are in detail, usually starting from the base of a 
thread and gradually extending inward. Manufacturers of stay- 
bolt material have endeavored to minimize failures and to meet the 
unusual conditions of an iron stressed beyond its elastic limit by the 
supply of specially piled iron arranged with a view to breaking up 
the extension of the initial fracture. For this reason iron piled with 




Fig. 1 Section of Firebox 
Showing Stay-bolts 



Fig. 2 Faggott Piling for 
Iron for Stay-bolts 



a central section of small bars and an envelop of flat plates has met 
with much success for this class of service. In a further efifort to 
secure an iron specially adapted to this class of work various forms of 
shock, vibratory and fatigue tests have been imposed. No design 
has yet been produced however which permits the employment 
of material of elastic limit sufficiently high to resist the flexural 
stresses, although a large class of material particularly adapted to the 
purpose is available. 

9 It is obvious that the remedy does not lie in the use of a slow- 
breaking material but in the employment, of material of sufficiently 
high elastic limit to meet the conditions of service. It is also possi- 



362 



DISCUSSION 



ble to reduce the diameter of the bolt greatly by the use of such a 
material, thus proportionately reducing the fiber stress in flexure. 
10 Stay-bolt material however must possess sufficient ductility 
to enable the ends to be readily hammered over to make a steam- 
tight joint and to afford additional security against pulling through 
the sheets. To meet these conditions the bolt illustrated in Fig. 
3 has been designed. The stem is of the same grade of steel as that 
used in the manufacture of springs. It is oil-tempered and will safely 
stand a fiber stress of 100,000 lb. per square inch. Its high elastic 
limit makes it possible to reduce the diameter to f or ^ in. or even 
less. The ends are of soft steel, and it is thus possible to apply and 
head up the bolt in the usual manner 

jll The employment of a stem of the diameter indicated reduces 
the fiber stress in flexure to less than one-half that in the ordinary type 
of bolt and it is of material capable of being stressed to a high degree. 
It has hitherto been impossible to employ in stay-bolts any of the 
steels containing chromium, nicke), vanadium or other metaloids 




Fig. 3 Flexible Spring Steel Stay Bolt 



possessing properties expecially adapted to this class of work, but 
these steels can readily be used in the stem of the bolt described. 
^12 The stem of the bolt can be flexibly secured to the end in one 
)f the customary ways, but the flexibility of the bolt does not depend 
upon a flexible connection. A type of bolt with a relatively inflexi- 
ble connection, usually one in which the stem screwed into the ends 
with a running fit, met with the most favorable consideration. Such 
a bolt is flexible as a spring is flexible, in that it can be deflected to 
meet the requirements of service without exceeding the elastic limit. 
In fact the stem^may be of a number of pieces, either of plates or 
small rods, thus increasing its flexibility. 

il3 The actual breaking strength of |the bolt sizes ordinarily 
employed is shown in the following statement. These bolts were 
recessed to the base of the thread and tested in the same form as that 
in which they are employed in service. For comparison the approxi- 
mate weights of the usual length of bolt are also given. These 



NEW DEPARTURE IN FLEXIBLE STAY-BOLTS 



363 



weights are for bolts over the entire length, including the squared 
ends for screwing the bolts into the sheets. 



ACTUAL BREAKING STRENGTH OF 


STAY-BOLTS 




Type 


Nominal Diameter 


Actual 
Breaking 


Weight 


Vibrations 




. . i 1 in. 


32,500 
24,500 
32,000 


20 'oz. 
15 " 
10/' t 

1 


6,000 




• • 1 I " 


5,200 


Spring steel stem. . . 


. . 1 in. ends ,'„ in. stem 


500,000 



14 The vibrating test was made by clamping one end of theboltin n 
machine and revolving the other end through a radius of ^ in. , the spec- 
men bsing 6 in. long from the end of the right head to the center of 
the rotating head. A tensional load of 4000 lb. was also applied i(» 
the bolts. The best grades of iron bolts break on being subjected to 
from 5000 to 6000 rotations, whereas the spring steel bolts wore 



■ 


lllll 


^ma^m^^^^^mmmmmmm 


1 


mm: 











Fig. 4 Spuing Flexible and Regular Iron Bolts of Sa.me Tensile SxRENOTri 

vibrated 500,000 times without failure, and on some of them tlie 
test was continued without -failure to 1,000,000 vibrations. These 
tests demonstrated that the bolt is not stressed beyond the elastic 
limit under these severe conditions and that the probability of its 
failure in less severe conditions is very remote. 

15 The extent of the expansion which can take place in the fire- 
box of a boiler can readily be calculated. 

Distance between stay-bolts, 4 in. 
Temperature of inside sheet, 400 deg. fahr. 
Temperature of^outside sheet, 100 deg. fahr. 
Coefficient of expansion, 0.0000066. 



364 



NEW DEPARTURE IN FLEXIBI.E STAY-BOLTS 



Then the expansion between two bolts will equal: 0.0000066 X (400 — 
100) X 4 = 0.0079, and each bolt will deflect 0.00395 in. It has 
been shown that this amount of deflection will stress the usual type 
of bolt beyond the elastic limit. In practice however one bolt may 
hold rigidly, throwing the entire deflection on the adjacent bolt, or 
neither bolt may deflect and the sheet will then buckle. Under this 
condition the neutral axis will assume the form ABC and the length 
AB will equal 2.00395 in. and the sheet will buckle to an extent, 
BD = 1/2.00395^ — 2^ = 0.125 in. It is obvious that the repetition of 
a force suflScient to buckle a sheet ^ in. must ultimately lead to a 






Fig. 5 Showing Manner in which Plates Buckle with Eigid Stays 



crack in the furnace sheets. If, however, the bolt deflects, allowing 
the sheet to expand normally, the latter will be relieved of these extra- 
neous loads. 

16 A bolt of sufficient flexibility to deflect under the forces follow- 
ing expansion, and of material which will not be stressed beyond the 
elastic limit in resisting these forces, will greatly assist in reducing 
the cost of boiler maintenance by eliminating broken stay-bolts and 
reducing the stresses in the furnace plates. If in addition the bolt 
has a smaller diameter the life of the furnace plates should be fur- 
ther increased, as such a bolt will interpose less obstruction to the 
circulation of the water in the water legs. 



NEW DEPARTURE IN FLEXIBLE STAYBOLTS 365 

DISCUSSION 

William Elmer. The decrease in the diameter of a staybolt, from 
15/16 in. or 1 in., which I believe is the present practice, to a small 
diameter, as 7/16 in., even if the tensile strength of the material is 
increased, brings to mind at once the possibility of twisting off these 
small bolts in their mechanical application to the boiler. The writer 
hopes Mr. Wille will say something about this, 

W. E. Hall. Any one who has had the care of the locomotive 
type of boiler appreciates that staybolt maintenance is a source of 
intense anxiety, and that the fact that more disastrous results are not 
forthcoming reflects great credit on the vigilance of the motive power 
departments of our railroads. This result, however, is accomplished 
only at high maintenance cost, which, fortunately, has always been of 
secondary consideration. 

2 Flexibility and length, of which Mr. Wille speaks, no doubt have 
considerable influence on the breakage of these bolts. It would be 
interesting to have details showing the construction of the bolts, 
and just how they were held, in his vibratory tests, 

3 It should be noted, however, that his construction calls for a 
studbolt of larger diameter than the ordinary staybolt, and in addi- 
tion this studbolt must project into the leg of the boiler to give suffi- 
cient length of thread-contact of the staybolt proper in the studbolt; 
in other words, he increases the diameter (the studbolt), introduces 
two threaded surfaces to the strain of flexure, and shortens the stay- 
bolt an amount equal to the projection of the studbolt into the leg 
of the fire-box. Is this not contrary to his own deductions? Assum- 
ing that this construction would decrease, but not eliminate, broken 
staybolts, would this construction facilitate their detection, diflScult 
under the best conditions, or make it more difficult? The reduced 
length might be relieved by making the fit of the thread of the bolt 
in the studbolt looser than that of the studbolt in the sheet, thereby 
always throwing the point of flexure upon the studbolt at the side of 
the outer sheet. But this is a risky procedure hardly deserving of 
consideration in boiler practice. 

4 The breakage of staybolts is confined almost exclusively to the 
upper rows of bolts of the legs of the fire-box. The fracture usually 
starts from the top, sometimes from the side, not infrequently around 
the circumference and occasionally from the bottom. The fracture 
is almost always close to the outer sheet, but a break close to the 



366 



DISCUSSION 




Fig. 1 Blank and Completed Ends of Flexible Staybolt 



NEW DEPARTURE IN FLEXIBLE STAYB0LT8 



3 67 



inside sheet is not unknown. The fracture is always in detail, bar- 
ring shamefully defective material or workmanship, or at least up to 
the point where the remaining area in contact is not sufficient to with- 
stand the strain to which that reduced area of the bolt is subjected. 
More or less irregularity of the line of fracture is to be expected. Con- 
ditions are not always the same for every bolt : all bolts do not fit the 
sheets alike, there is more or less variation in the upsetting, and the 
buckling and warping of the inner sheet is not the same for each bolt. 
These, together with other minor conditions, representing reasonable 
refinement in practice, preclude uniformity of the fracture. 

5 But the important points are, that these bolts always break in 
detail and always at the root of the thread. Have we any reason to 
expect that it would be otherwise? For example, if we wish to break 
a piece of metal we first grip it in a vise, notch it close to the jaws of 
the vise and bend it back and forth. We do the same with a stay- 




FiQ. 2 Flexible Statbolt with End shown in Fig. 1 

bolt when we cut the thread with a sharp die (and in this respect the 
U. S. Standard is inferior to the Whitworth), hold it firmly in the 
outer sheet and subject it to the bending due to the expansion, con- 
traction and warping of the inner sheet. Cutting the thread produces 
a constructive defect. This feature is necessary as the bolts are now 
used and its correction is beyond the scope of mathametics. 

6 Fig. 1 and Fig. 2 show designs for bolts made in January 1886. 
In Fig 1 the object was to make the bolt of iron or steel, and by a 
gradual reduction from the ends towards the center to make the body 
of the bolt more flexible; and to make the length of the end such that, 
after making sufficient allowance for upsetting, the thread could m^t 
project beyond the face of the inside of the sheet. The flare on 
the end of the bolt was shown merely as a more ready means of 
upsettmg the 'end, especially when the bolt is of steel. As the 



368 DISCUSSION 

number of bolts breaking at the face of the inside sheet is so 
small as to be negligible, and to accommodate for the difference in 
widths in the leg of the fire box, the bolts could be made of vary- 
ing lengths on a bolt machine and kept in stock. By this method, 
if made by upsetting, it would not be necessary to turn the shank 
of the bolt. In all cases, however, it is preferable that the thread- 
ed end in the heavier sheet should not project into the leg of the 
fire box. This construction permits of better circulation, and of 
a somewhat longer bolt, and better braces the threaded length 
against deflection. 

Alfred Lovell. The durability ot stay bolts in high-pressure 
locomotive boilers is a prominent factor in the cost of locomotive 
maintenance and has an important bearing on the safety of the public. 
Any innovation intended to increase their durability or reliability is 
therefore worthy of the most careful demonstration and service trial. 
The staybolt described by Mr. Wille is of this character, and his paper 
clearly shows the desirability of providing a staybolt of high elastic 
limit and great flexibility. 

2 This is accomplished by the use of metal having the requisite 
qualities in a high degree, and yet maintaining in the direction of the 
bolt's axis, the desirable features of continuity and rigidity, thus 
avoiding the liability of unequal tension of adjacent bolts. This 
is undeniably a departure in the right direction, and one that will 
receive prompt attention in railway mechanical engineering. 

3 One other important feature to be considered, however, in 
adopting a new form or in selecting a new material for staybolts, 
is ability to resist corrosion under the conditions of use, a feature even 
more important with staybolts of the proposed high tensile strength 
and small diameter, than with staybolts of ordinary diameter. For 
example, the paper gives the actual breaking strength of a 1-in. iron 
bolt as 32,500 lb., and of a 3^ -in. spring-steel stembolt as 32,000 lb. 
If it is assumed that each is reduced ^ in. in diameter by corrosion, 
this will reduce the breaking-strength of the 1-in. bolt by 23.56 per 
cent, and of the ^-in. bolt by 48.98 per cent. The breaking-strength 
of the iron bolt is then 24,843 lb. and that of the spring-steel stem- 
bolt is 16,327 lb. Another J-in. reduction in diameter would bring 
the strength of the iron bolt to 18,278 lb., and of the spring-steel stem- 
bolt to 5878 lb. 

4 Formula 4 of Mr. Wille's paper shows that the flexural stress, the 
ordinary cause of failure, decreases directly as ihe diameter, yet it is 



NEW DEPARTURE IN FLEXIBLE STAYBOLTS 369 

evident that if the bolts are affected by corrosion and in equal amount, 
this will eventually reduce the tensile strength of the small bolt to a 
point below the proper factor of safety, while the larger bolt has still 
tensile strength to spare. 

5 Since the new type of staybolt makes it possible to employ 
various steels and alloys which cannot be used in the ordinary stay- 
bolt, it fortunately provides for a wide range of selection, and quite 
probably a steel-alloy or mixture may be provided that will resist 
corrosion equally well or better than the iron ordinarily used. 

6 The impurities in the water used also very greatly affect corro- 
sive action, and it is probable that with the new type of bolt it would 
be advantageous to examine carefully the water of the locality where it 
is to be used and to select a metal with reference to the character of 
the water. Information regarding the resistance of various steel alloys 
to corrosion, in water of ordinary purity, and experiments determining 
the metal that will give the least corrosion in water having various 
impurities, are highly desirable, since the new bolt makes possible the 
use of that metal which combines in the greatest degree flexibility, 
high elastic limit and resistance to corrosion. 

7 A combination of these three qualities, in a staybolt of the type 
described, will be an innovation much to be desired in locomotive 
practice. 

F. J. Cole. In reading this paper the following occurred to me: 

a Have any of these staybolts been applied to locomotive 
boilers? 

b What is their approximate cost? 

c With 1-in. ends is there any danger of overstraining the |-in. 
diam. spring-steel center in screwing the bolt to place? 
It would seem that the torsional stress on the compara- 
tively small central part in screwing the bolts to place 
when they fit tightly in the sheets would be excessive. 

d In upsetting the soft ends, does the |-in. diam. center 
afford sufl&cient resistance to avoid injury to the threads 
from hammering? It is customary in upsetting the ends 
to use a hammer or heavy weight at the other end of the 
bolt to afford the necessary resistance and to prevent 
injury to the threads. 

e In Par. 7 Mr. Wille speaks of 12,000 lb. fiber stress being 
good practice. I presume he has in mind the combined 
stress due to tension and bending. Current locomotiv* 



370 DISCUSSION 

practice is represented fairly well by 5500 to 6000 lb. for 
tension alone. 

The Author. Mr. Hall has fully described the general charac- 
teristics of a staybolt broken in service. The able experiments of 
Boushinger and Wohler have shown that a detail fracture occurs in 
parts which have been stressed by flexure above the elastic limit. If 
the parts be threaded the stress may be localized at the base of the 
thread which will increase the possibility of failure of parts stressed 
at or near the elastic limit, but if the parts be not stressed above the 
"fatigue" elastic limit of the material, it will not fail irrespective of 
whether it be threaded or of plain section. 

2 The bolt described is made of two kinds of material: 

a The stem of high-carbon tempered steel. 
b The two ends of soft steel. 

This arrangement permits the use of a material having such a high 
elastic limit and provides an ample factor of safety and still allows 
the soft ends to be ri vetted over to make a steam-tight joint. The 
bolt referred to by Mr. Hall does not fulfil these conditions, as if it 
were made of steel having a sufficiently high elastic limit it would be 
impossible properly to rivet the ends. 

3 The ends of the composite bolts are not larger than a normal 
bolt, those in service being made 1 in. in diameter when used in new 
boilers and if in. and 1^ in. when used to replace f-in. and 1-in. 
bolts respectively in old boilers. The stems are not made a tight fit 
into the ends but are threaded so that they can be tightly screwed into 
the ends by hand thus affording some flexibility in the joint. It is 
not necessary however to thread the stems as they can be upset and 
the ends closed around them, making a ball joint, thus having all the 
advantages with none of the disadvantages of the usual type of flexi- 
ble bolt. 

4 The bolts have been in service for two years in a district notori- 
ous for the trouble experienced with corrosion and the bolts are still 
intact. Mr. Lovell suggests the possibility of providing special steels 
to resist corrosion. The high-carbon tempered steel being dense and 
hard is an admirable material to resist corrosion; furthermore as the 
stems are oil-tempered they have a coat of enamel which prevented 
the slightest amount of pitting even after two years' service. 

5 There are a large number of these bolts in service in locomotive 



NEW DEPARTURE IN FLEXIBLE STAYBOLTS 371 

boilers and no difficulty has been experienced either from twisting of 
the stem in screwing into the boiler or from injury to threads in rivet- 
ting. 

6 The cost of material entering into the construction of the spring 
flexible bolt is very much less than that of the usual type of staybolt 
since the weight is much less and the materials are the relatively cheap 
commercial grades of steel instead of the high-priced irons usually 
employed in the manufacture of staybolts. The labor cost to man- 
ufacture is however somewhat higher. 

7 There have been about 3000 of these bolts put into service 
and no danger has been experienced of overstrainin the f-inch 
tern in screwing the bolt in the boiler; nor has any difficulty been 
sxperienced because of insufficient resistance to avoid injury to the 
e bread when the bolts are hammered up. 



No. 1247 

THE HUDSON-FULTON CELEBRATION 

In keeping with the celebration of the discovery of the Hudson 
River and the successful application of steam to navigation, the 
House Committee of the Society appointed a subcommittee, Edward 
Xan Winkle, Chairman, to prepare an exhibit of models, drawings, 
letters, books and other items related to early steam navigation, of 
interest to the general public as well as the engineer. 

The exhibit was the only participation of any engineering organi- 
zation, as such, in the celebration, and much credit is due Mr. Van 
Winkle for the time and attention devoted to its preparation. The 
exhibit was held in the rooms of the Society and was open from 9 a.m. 
to 5 p. m. every week day during September and October 1910. A 
list of the exhibits follows : 

Model of Fulton's boat, the Clermont, as she was at the time of her 
maiden trip. After making several trips during the fall of 1807, the Cler- 
mont was docked at Browne's shipyard and fitted up for regular passenger 
traffic. Loaned by the Smithsonian Institution. 

Model of Stevens' boat, the Phoenix, at the time of making her notable 
New York-Philadelphia trip — the first ocean voyage made by a steam vessel. 
Loaned bj' the Smithsonian Institution. 

Model of a steamboat built by John Fitch in 1786, having a peculiar 
arrangement of oars dipping into the v.ater something like a canoe paddle. 
Despite its clumsy appearance the boat made a trip of 20 miles on the Dela- 
ware River. Loaned by the Smithsonian Institution. 

Model (9-ft.) of the Deutschland, showing the remarkable development in 
steam navigation during the last one hundred years. Loaned by the Ham- 
burg-American line. 

Portrait of Fulton painted by himself while a pupil of Benjamin West. 
Presented to the Sociely by Mrs. R. Anna Gary. 

Fulton dining table, of mahogany, 85 ft. long and 5 ft. wide. Presented to 
the Society by Thomas Egleston in 1891, who received it from his brother, 
George Egleston, into who-se hands it had passed from Mrs. Egleston 's House, 
to whom Robert Fulton gave it. 

Two autograph drawings bj' Fulton, one of them a brush dr.awing of a high- 
level canal, presented by Cornelia J. Carll, showing his artistic ability as well 
as his mechanical genius. The other drawing is of the Sound steamer Fulton, 
built in 1813, and was presented by Louisa Lee Schuyler. 

^ Drawing of the Fulton reproduced on a bronze tablet made in connection 
with the Fulton monument erected by the Society over Fulton's grave in Trin- 
ity churchyard. The tablet also gives a short description of the Clermont. 



374 



SOCIETY AFFAIRS 




ROBEKT FtJLTON 

FROM A PAINTING BY HIMSELF, PRESENTED TO THE AMERICAN SOCIETY OP MECHANICAL 
ENGINEERS BY MRS. R. ANNA CART, NOVEMBER 30, 1897 



SOCIETY AFFAIRS 375 

Photograph of a letter by Fulton describing his experimental boat on the 
Seine, with copies of other letters by Fulton and his friends, presented by 
Henry Harrison Suplee, member of the Committee on Society History. 

Hudson River Guide, published in 1850, describing the various points of 
interest on the Hudson and giving the time of sailing of the Hudson River 
boats of that period. Loaned by H. J. Gclien. 

Oil painting of James Watt, a copy of a portrait by deBreda, now in tlie 
possession of John Scott of Hawkhill, Greenock, presented by past members 
of the Council. 

Oil portrait of Capt. John Ericsson, painted by Ballin of Stockholm, show- 
ing the designer of the Monitor at the age of 59. A bust of Ericsson, by Knee- 
land, was presented to the Society by James Mapes Dodge, Past-President, 
Am. Soc. M. E. 

Solid silver model of the Half Moon, loaned by Tiffany and Company. 

Model of the Monitor, presented by Thomas F. Rowland, who built the Mon- 
itor at the Continental Iron Works. 

Models, exhibited by Ericsson at the Centennial Exposition, loaned by the 
United Engineering Society. 

Early books on steam navigation, from the Library of the Society. 

Copy of a letter from Fulton to Boulton & Watt in 1810, ordering an engine 
for another boat. 

Copies of letters by Fulton's workmen. 

Drawings showing the comparative size of the White Star S. S. Olympic, 
Clermont and Half Moon; of New York's early water works, corner of Centre 
and Reade Streets; a water works note issue in 1774: all loaned by Daniel 
Arthur. 

Color print of the Half Moon. 

Fac-simile of Rules and Regulations for passengers in the Clermont. 

Sketch of James Watt discovering the condensation of steam. 

Photographs of a letter written by Fulton to Boulton & Watt, describing the 
Clermont and ordering an engine for another steamboat. Contributed by 
the Smithsonian Institution. 

Print of the original oil painting by W. F. HalsoU of the battle between the 
Monitor and the Merrimac in the collection of the late Thomas Fitch Rowland, 
the builder of the Monitor. 

Bills, receipts, etc., written by John Fitch; stock certificate issued by John 
Rumsy; copy of New York Herald of 1815 containing items of steamboats; 
copy of Washington Gazette of 1821 attacking Fulton: all loaned by Dr. C. S. 
Bullock of Stratford, Conn. 

Silver Hudson-Fulton Medal, loaned by Dr. Geo. F. Kunz. 

Descriptive Guide to the Grounds, Buildings and Collections of the New 
York Botanical Garden. 

List of Prints, Books, Manuscripts, etc., relating to Henry Hudson, the Hud- 
son River, Robert Fulton and Steam Navigation, at the Lenox Branch of the 
New York Public Library. 

The Indians of Manhattan Island and Vicinity, by Alanson Skinner of the 
department of anthropology of the American Museum of Natural Histor}^ 

The Wild Animals of Hudson's Day and the Zoological Park of our Day, by 
W. T. Hornaday, Sc.D., Director of the New York Zoological Park. 

All of these loaned by Geo. F. Kunz. 



376 SOCIETY AFFAIRS 

Considerable interest was manifested in this exhibit and it was well 
attended. The total namber of visitors who registered was 355, but 
400 is a fair estimate of the total. The greatest registration for any 
one day was 52 on Monday, September 27. A wide extent of territory 
was represented, including Connecticut, Delaware, Illinois, Maine, 
Massachusetts, Maryland, Michigan, Ohio, Pennsylvania, Rhode 
Island, Tennessee, Washington, D. C, Wisconsin, Canada, Japan 
and Switzerland. 

On the morning of September 24, in the presence of officers and 
members of the Society and of the Pennsylvania Society, wreaths 
were placed on the Fulton Monument erected by the Society in 1901 
in Trinity Churchyard, New York. The Pennsylvania Society was 
represented by Robert Mazet, Vice-President, and Barr Ferree, Sec- 
retary. J. M. Schroeder, Commissioner from Pennsylvania, was also 
present. 

Members of both societies assembled in the lobby of the church 
and, preceded by Dr. Manning, rector of Trinity Church, marched to 
the Fulton Monument. Dr. Manning offered prayer and wreaths 
were placed on the monument, after which the Lord's Prayer was 
recited, followed by the benediction. "Taps" sounded by a Seventh 
Regiment bugler brought the exercises to a close. 



SOCIETY AFFAIRS 



377 




378 



SOCIETY AFFAIRS 




a a 
ii a 



SOCIETY AFFAIRS 



379 




o 

E-i 



380 



SOCIETY AFFAIRS 




Model of John Fitch's Boat 

BUILT IN PHILADELPHIA AND TRIED ON THE DELAWARE RIVER IN 1786 




The Phoenix, Built by John Stevens. In 1809 She Made an Ocean Trip 
from New York to Philadelphia 

rROU PHOTOGRAPHS MADE IN THE ROOMS OF THE SOCIETY 



No. 1248 

MEETINGS, OCTOBER-DECEMBER 

NEW YORK MEETING, OCTOBER 12 

At the meeting of the Society held October 12 in the Engineering 
Societies Building, Prof. R. C. Carpenter presented his paper on The 
High-Pressure Fire-Service Pumps of Manhattan Borough, City of 
New York. President Jesse M. Smith presided. The attendance was 
192. 

Secretary Calvin W. Rice read an invitation from the Institution 
of Mechanical Engineers to hold a joint session with them from 
July 26 to July 29, 1910. The President then introduced the follow- 
ing Japanese commissioners visiting the United States to study 
various industries: Dr. Ryota Hara, doctor of engineering and chief 
engineer of Yokohama; Rinnosuke Hara, of the Japanese Architec- 
tural Society; Junkichi Tanabe, of Tokyo, of the Institute of Japanese 
Architects; and Narazo Takatsuji, director of a large spinning fac- 
tory. A telegram of regret was received from Kojiro Matsukata^ 
the leading shipbuilder of Japan. 

Those taking part in the discussion of the paper were: Prof. Geo. 
F. Sever, WiUiam M. White, Geo. L. Fowler, John H. Norris, J. R. 
Bibbins, J. J. Brown, Geo. A. Orrok, Frederick Ray, H. Y. Haden, 
Thos. J. Gannon, Henry B. Machen, Richard H. Rice, Chas. A. Hague, 

A. C. Paulsmeier, Prof. W. B. Gregory, Wm. 0. Webber and Chas. 

B. Rearick. 

At the close of the discussion Mr. White showed a number of lantern 
slides giving efficiency curves of various pumps designed by the 
I. P. Morris Co., Philadelphia, Pa. 

Mr. Fowler, described the work of centrifugal pumps in dredging, 
and exhibited the following lantern slides, as evidence of the great 
suction capacity of these pumps : 

A piece of shaft weighing 70 lb. raised and passed by a 15-in. dredging pump; 
improvement of New York Harbor, Steamer Reliance. 

A piece of tree root raised and passed by a 12-in. pump from 14 ft. of water at 
Miami, Fla.; Florida East Coast Railway Company improvements. 



382 SOCIETY AFFAIRS 

A piece of pig iron measuring 11| in. by 4f in. by 3j in. and weighing 35 lb., 
raised and passed by an 8-in. special cataract wrecking pump from 15 ft. of water 
from the wreck of a canal boat sunk at Puas Dock, Yonkers, N. Y., by the Bax- 
ter Wrecking Company, New York. 

ST. LOUIS MEETING, OCTOBER 16 

A meeting of The American Society of Mechanical Engineers and 
the Engineers Club of St. Louis was held at the rooms of the latter 
organization at 8.15, Saturday evening, October 16, under the direc- 
tion of William H. Bryan, Chairman, M. L. Holman and E. L. Ohle, 
Secretary, of the local joint committee. 

A letter from President Jesse M. Smith was presented, indicating 
the sentiment of the Society towards local meetings. This was briefly 
responded to by President E. E. Wall, of the Engineers Club of St. 
Louis, who also emphasized his belief in the advantages of cooperation. 

Prof. R. C. Carpenter of Cornell then presented in abstract his 
paper on The High-Pressure Fire-Service Pumps of Manhattan 
Borough, City of New York, accompanying it by running comments 
and comparisons. 

He was followed by Horace S. Baker, assistant engineer of the City 
of Chicago, who presented with the aid of illustrations the results of 
recent study with a view to adopting high-pressure service. E. E. 
Wall, assistant water commissioner. City of St. Louis, outlined the 
plan proposed for high-pressure fire-service in St. Louis. He was 
followed by H. C. Henley, chief inspector, St. Louis fire prevention 
bureau, and vice-president of the National Fire Protection Associa- 
tion, expressing views of the fire insurance authorities, entirely favor- 
able to the installation of such systems when properly designed and 
operated. Chas. E. Swingley, chief of the St. Louis fire department, 
on invitation made a few brief remarks to the effect that such systems 
were of undoubted advantage in the congested dictricts of large cities, 
and expressed the^hope that something might soon be done along this 
line in St. Louis. There was further brief discussion by^Edw. Flad, 
Prof. H. W. Hibbard, and H. C. Toensfeldt. 

Luncheon was served by the Engineers Club of St. Louis. The 
attendance was 100. 

BOSTON MEETING, OCTOBER 20 

On Wednesday evening, October 20, a joint meeting of the Society 
with the Boston Society of Civil Engineers was held in the latter 
society's rooms, Tremont Temple, Boston, Mass. 



SOCIETY AFFAIRS 383 

Chas. T. Main, vice-president of the Boston Society of Civil Engi- 
neers, presided. Following the routine business of the Society of Civil 
Engineers, Mr. Main read a letter from Jesse M. Smith, President of 
The American Society of Mechanical Engineers, regretting his inabil- 
ity to be present at the meeting, and wishing the Boston members 
success for their coming meetings. 

A paper by Cav. Gaetano Lanza, professor, and Lawrence S. 
Smith, instructor, at the Massachusetts Institute of Technology, on 
Stresses in Reinforced Concrete Beams, was read by the former. 
Following the presentation of the paper, a discussion by J. R. 
Worcester of Boston, Mass., was read. Sanford E. Thompson, 
Fred S. Hinds, Henry F. Bryant and Geo. F. Swain contributed 
oral discussions. 

The total attendance at the meeting was 180, of whom 60 were 
members of the Society of Civil Engineers, 50 members of The Amer- 
ican Society of Mechanical Engineers and 70 guests. 

NEW YORK MEETING, NOVEMBER 9 

At the meeting of^the Society in New York on November 9, Pro- 
fessor Lanza presented his paper on Stresses in Reinforced Concrete 
Beams, and Professor Rautenstrauch his paper on Design of Curved 
Machine Members. The discussion on both papers proved valu- 
able, the lantern slides shown in the discussion of Professor Lanza's 
paper adding much to its interest. Those participating were Sanford 
E. Thompson, E. P. Goodrich, Prof. Walter Rautenstrauch, Prof. W. 
H. Burr, B. H. Davis, of the Lackawanna Railroad, who showed slides 
of a number of concrete arches in railroad work, C. B. Grady of the 
New York Edison Company, who showed slides of beams and floor 
slabs under test, F. B. Gilbreth, who showed slides of the longest 
concrete beam of a rectangular section ever built in a roof, as well as 
other beams which had successfully passed through the fire and earth- 
quake of the San Francisco disaster. Contributed discussions by 
Prof. J. C. Ostrup, E. L. Heidenreich, and.C. E. Houghton were also 
presented. Professor Rautenstrauch's paper was discussed by a num- 
ber of authorities on machine tool design, as follows : Professor Lanza, 
Chas. R. Gabriel, George R. Henderson, Professor Burr and Carl G. 
Barth. Those submitting written chscussions were: C. E. Houghton, 
A. L. Campbeh, H. Gansslen, F. I. Ellis, E. J. Loring and John 
S. Myers. 



384 SOCIETY AFFAIRS 

ST. LOUIS MEETING, NOVEMBER 13 

At the meeting of the Societj'^ at St. Louis, November 13, with the 
Engineers Club of St. Louis, a description of the new plant of the 
Heine Safety Boiler Company of Boston was presented by E. R. Fish, 
Secretary of the Company, under the title, A Modern Boiler Shop. 
There was also further discussion of Professor Carpenter's paper on 
High-Pressure Fire-Service, continued from the October meeting. 

BOSTON MEETING, NOVEMBER 17 

A successful meeting of the Society was held at Boston in the Lowell 
Building, Massachusetts Institute of Technology, Wednesday even- 
ing,' November 17. Two hundred and forty were present at this meet- 
ing and the Low-Pressure Steam Turbine was the topic of discussion. 

Henry G. Stott of the Interborough Rapid Transit Company gave 
an interesting account of the difficulties encountered as well as the 
very fine results obtained from an installation recently made at the 
59th Street Station of his company, New York. W. L. R. Emmet, 
engineer of the lighting department of the General Electric Company, 
described the low-pressure turbine situation from his viewpoint and 
pointed out the advantages of this type of prime mover for many mill 
installations and industrial works in New England. H. E. Longwell, 
consulting engineer of the Westinghouse Machine Company, and 
Edward L. Clark, manager of their Boston office, both spoke on the 
work that company is doing in this field. Max Rotter, turbine engi- 
neer of the Allis-Chalmers Company, pointed out in a humorous way 
a number of situations where the low-pressure turbine was not a desir- 
able proposition. Professor Miller of the Massachusetts Institute of 
Technology also discussed the subject. 

ST. LOUIS MEETING, DECEMBER 11 

A meeting was held Avith the Engineers Club of St. Louis on Sat- 
urday evening, December 11, at the rooms of the latter society. The 
meeting was called to order by William H. Bryan, member of the 
Meetings Committee of the Society and chairman of the joint com- 
mittee of the two societies at St. Louis. Prof. E. L. Ohle acted as sec- 
retary. There were present fifty-five members and guests. 

The paper of the evening was by G. R. Parker of the General Elec- 
tric Company, on The Relation of the Steam Turbine to Modern Cen- 



SOCIKTT AFFAIKS 385 

tral Station Practice, in which the underlying principles of modern 
steam turbines were discussed, together with the design of various 
prominent types on the market, and the developments made in recent 
years in improving capacity and efficiency. Attention was called to 
the large turbine capacity which may now be obtained within limited 
floor space; to the question of low-pressure turbines and their avail- 
ability in supplementing standard reciprocating engines, increasing 
both their capacity and economy; also to the work already done in 
this direction at the plant of the Union Electric Light & Power Com- 
pany in St. Louis, and to prospective work along similar lines in the 
same plant. The address was illustrated by lantern slides. 

Discussion followed by Chairman Bryan, Prof. H. W. Hibbard, 
L. R. Day, E. R. Smith and Prof. E. L. Ohle. 

On Saturday afternoon an excursion was made to the Ashley Street 
plant of the Union Electric Light & Power Company, for the inspec- 
tion of the apparatus and equipment, on the invitation of John 
Hunter, chief engineer. 

BOSTON MEETING, DECEMBER 17 

On Friday evening, December 17, a goodly number of engineers 
of Boston and vicinity gathered on invitation of the local members of 
The American Society of Mechanical Engineers to discuss the Effect 
of Superheated Steam on Cast Iron. The meeting was called to order 
by Prof. Ira N. HolHs. 

The committee which has been in charge of the meetings, consist- 
ing of Messrs. Hollis, Moultrop, Miller, Mann and Libbey, was con- 
tinued. 

The papers on the subject for the evening were then presented by 
their authors. Prof. Edward F. Miller of Boston, Arthur S. Mann of 
Schenectady, and Prof. Ira N. Hollis of Boston, and were discussed 
by B. R. T. Collins, George A. Orrok, Chas. H. Bigelow, W. K. 
Mitchell, John Primrose, L. B. Nutting, Wm. E. Snyder and others. 
The general purport of the discussion was rather reassuring to the 
users of cast-iron pipe and fittings, and to those who are interested 
in the extension of the use of superheated steam, in indicating that 
superheated steam per se has no injurious effect upon cast iron fittings, 
but that if the pipe lines are properly designed for the greater ranges 
of temperature, the fittings made adequate to the pressure and fluctu- 
ations in temperature avoided, the use of superheated steam intro- 
duces no piping difficulties which can not be easily overcome. 



THE ANNUAL MEETING 

The thiitieth annual meeting of The American Society of Mech- 
anical Engineers was held in the Engineering Societies Building 
December 7 to 10, with an attendance of 628 members and 435 
guests. For the first time at such a meeting the arrangements for 
the entertainment features were entirely in the hands of the local 
coEomittee, the members in New York and vicinity acting as hosts, 
this method of handling an important part of the annual meeting 
being fully justified by the results. A feature of the meetmg was 
an afternoon trip on Wednesday, through the new Pennsylvania 
Terminal, which brought out a large body of members and guests. 
The attendance at the reception on Thursday evening, held in the 
ball-room of the Hotel Astor, was nearly 600, 

PROGRAM 

OPENING SESSION 
Tuesday, December 7, 8.30 p.m., Auditorium 

THE president's ADDKESS 

The Profession of Engineering, by Jesse M. Smith 

ELECTION OF OFFICERS 

Report of Tellers of Election of Officers and introduction of the 
President-elect. 

RECEPTION 

The President and President-elect, with their ladies, received the 
members and guests in the rooms of the Society. Music and refresh- 
ments followed the reception. 

BUSINESS MEETING 
Wednesday, December 8, 9.80 a.m., Auditorium 

Annual business meeting. Reports of the Council, Tellers of Elec- 
tion of membership, standing and special committees and Gas Power 
Section. Amendments to the Constitution. New business. 

Luncheon was served to members and guests. 



SOCIETY AFFAIRS 387 

Wednesday afternoon 

Excursion to points of engineering interest. Hosea Webster, 
Chairman Sub-Committee on Excursions. 

LECTURE 

Wednesday, 8.15 p.m., Auditorium 

The Era of Farm Machinery, L. W. Ellis, of the Bureau of Plant 
Industry of the United States Department of Agriculture at Wash- 
ington, D. C. Illustrated by lantern slides. 

PROFESSIONAL SESSIONS 

Thursday, December 9, 9.30 a.m., Auditorium 

measurement of the flow of fluids 

Tests on a Venturi Meter for Boiler Feed, Chas. M. Allen. 
Discussed by F. N. Connet, Clemens Herschel, Dr. Sanford A. 
Moss, Prof. L. S. Marks. 

Efficiency Tests of Steam Nozzles, F. H. Sibley and T. S. Kem- 
ble. 

Discussed by A. F. Nagle, A. R. Dodge, Prof. C. C Thomas, J. 
A. Moyer. 

The Pitot Tube as a Steam Meter, Geo. F. Gebhardt. 
Discussed by Walter Ferris, A. R. Dodge, Prof. W. B. Gregory. 
An Electric Gas Meter, C. C. Thomas. 

Discussed by Prof. W. D. Ennis, E. D. Dreyfus, A. R. Dodge, 
Prof. L.S.Marks. 

Luncheon was served to members and guests at the conclusion of 
the session. 

Thursday, 2 p.m., Auditorium 

STEAM engineering 

Tan Bark as a Boiler Fuel, David M. Myers. 

Discussed by A. A. Cary, Prof. Wm. Kent, Prof. L. P. Brecken- 
ridge. 



388 SOCIETTf AFFAIRS 

Cooling Towers for Steam and Gas Power Plants, J. R. 
Bibbins. 

Discussed by Geo. J. Foran, W. D. Ennis, H. E. Longwell, B. H. 
Coffey, E. D. Dreyfus, F. J. Bryant, Carl G. deLaval. 

Governing Rolling Mill Engines, W. P. Caine. 
Discussed by H. C. Ord, James Tribe. 

An Experience with Leaky Vertical Fire Tube Boilers 
F. W. Dean. 

Discussed by R. P. Bolton, Prof. Wm. Kent, J. C. Parker, O. C 
Woolson, A. A. Gary, Prof. A. M. Greene, Jr., E. D. Meier, D. M. 
Myers. 

The Best Form of Longitudinal Joint for Boilers, F. W. 
Dean. 

Discussed by R. P. Bolton, Carl G. Bartb, E. D. Meier, Prof. A. 
M. Greene, Jr., W. A. Jones, Prof. S. W. Robinson, Geo. I. Rock- 
wood, Sherwood F. Jeter. 

Thursday, 2 p.m., Lecture Hall 

GAS POWER section 

Business meeting and election of officers. 

Testing Suction Gas Producers with a Koerting Ejector 
CM. Garland, A. P. Kratz. 

Discussed by Prof. R. H. Fernald, G. M. S. Tait, H. H. Suplee, L. 
B. Lent, S. C. Smith, W. B. Chapman, Edw. N. Trump. 

Bituminous Gas Producers, J. R. Bibbins. 

Discussed by, G. M. S. Tait, Prof. R. H. Fernald, W. B. Chapman, 
H. M. Latham, H. H. Suplee, Edw. N. Trump, H. B. Langer, S. C. 
Smith, Prof. Walter Rautenstrauch, G. D. Conlee. 

RECEPTION 

Thursday, 9 'p.m., Hotel Astor 

The Members of New York and vicinity received the membership 
of the Society, their ladies and guests, at the Hotel Astor. Dancing 
and refreshments followed the reception. 



SOCIETY AFFAIRS 



389 



PROFESSIONAL^SESSION 

Friday, December 10, 9. SO a.m. 

The Bucyrus Locomotive Pile Driver, Walter Ferris. 

Discussed by O. K. Harlan, A. F. Robinson, L. J. Hotchkiss. 

LiNESHAFT Efficiency, Mechanical and Economic, Henry Hess. 

Discussed by T. F. Salter, Prof. R. C. Carpenter, C. A. Graves, 
O.K. Harlan, C. J. H. Woodbury, Walter Ferris, Fred J. Miller, A. C. 
Jackson, C. D. Parker, Oliver B. Zimmerman, Geo. N. Van Der- 
hoef. 

Pump Valves and Valve Areas, A. F. Nagle. 

Discussed by Prof. Wm. Kent, A. B. Carhart, Prof. R. C. Car- 
penter, E. H. Foster, Chas. A. Hague, I. H. Reynolds, F. W. Sal- 
mon. 

A Report on Cast-Iron Test Bars, A. F. Nagle. 

Discussed by A. A. Cary, T. M. Phetteplace, Prof. W. B. Gregory, 
Geo. M. Peek. 

COMMITTEES OF THE ANNUAL MEETING 

MEETINGS COMMITTEE 

Willis E. Hall, Chairman 



William H. Bryan 
L. R. Pomeroy 



Charles E. Lucke 
H. de B. Parsons 



LOCAL COMMITTEE 



William D. Hoxie, Chairman 



F. A. Scheffler, Secretary 



Wm. L. Abbott 
H. P. Ahrnke 
Louis Alberger 
L P. Alford 
G. H. Barbour 
G. M. Basford 
Edgar H. Berry 
Francis Blossom 
William H. Boehm 
Reginald P. Bolton 
Geo. M. Bond 
G. I. Bouton 



L. P. Breckenridge 
Wm. H. Bryan 
R. C. Carpenter 
H. R. Cobleigh 
F. H. Colvin 
W. C. Dickerman 
Robert M. Dixon 
F. L. DuBosque 
Frank E. Eberhardt 
Harrington Emerson 
A. Falkenau 
W. H. Fletcher 



George J. Foran 

E. H. Foster 
H. A. Foster 
Geo. L. Fowler 
R. E. Fox 

F. L. R. Francisco 
John R. Freeman 
H. L. Gantt 
Fred J. Gubelman 
Willis E. Hall 

F. A. Halsey 
Geo. F. Hardy 



390 



SOCIETY AFFAIRS 



Henry S. Hayward 
G. R. Henderson 
F. V. Henshaw 
M. L. Holman 
W. R. Hulbert 
Alex. C. Humphreys 
William F. Hunt 
F. R. Hutton 
F. E. Idell 
H. S. Isham 

E. B. Katte 
R. S. Kent 
Walter C. Kerr 
Chas. Kirchhoff 
J. W. Lieb, Jr. 
Henry S. Loud 
Fred R. Low 
Chas. E. Lucke 
R. C. McKinney 

F. E. Matthews 
E. D. Meier 
Fred J. Miller 



B. M. Mitchell 
Chas. A. Moore 
I. E. Moultrop 

D. M. Myers 
W. W. Nichols 
J. H. Norris 

H. de B. Parsons 

E. H. Peabody 
L. R. Pomeroy 
H. O. Pond 

H. F. J. Porter 
W. P. Pressinger 
Calvin W. Rice 
A. L. Riker 
Fred. E. Rogers 
H. W. Rowley 
W. J. Sando 
E. F. Schnuck 
Jesse M. Smith 
Leo H. Snyder 
Albert Spies 
E. G. Spilsbury 



J. E. Starr 
Theo. Stebbins 

A. F. Stillman 
F. H. Stillman 
H. G. Stott 
H. H. Suplee 
Ambrose Swasey 

B. V. Swenson 
F. H. Taylor 

F. W. Taylor 
Stevenson Taj'lor 
Edw. Van Winkle 

G. T. Voorhees 
A. M. Waitt 

F. A. Waldron 

C. M. Wales 
Arthur West 
F. M. Whyte 
W. H. Wiley 
A. L. Williston 
Ira H. Woolson 



W. L. Clark 

W. C. Dickerman 



Sub-Committee on Finance 
C. A. Moore, Chairman 



Alex. C. Humphreys 
W. C. Kerr 



EXCURSION COMMITTEE 



George J. Foran 
Percy C. Idell 



Hosea Webster, Chairman 



Alfred F. Masury 
Frederick A. Scheffler 



BUREAU OF INFORMATION 



F. E. Idell, Chairman 



Charles C. Phelps 



James V. V. Colwell 



ENTERTAINMENT COMMITTEE 



Dr. D. S. Jacobus, Chairman 



SOCIETY AFFAIRS 



391 



Sub-Committee on President's Reception 
Tuesday Evening 
Col. E. D. Meier, Chairman 



W. C. Dickerman 
Bernard V. Swenson 
Francis Blossom 
Edward Van Winkle 
Louis Alberger 
G. M. Basford 
Harrington Emerson 
Dudley Farrand 
W. H. Fletcher 
C. H. Foster 



Geo. L. Fowler 
Willis E. Hall 
Geo. F. Hardy 
Alex. C. Humphreys 
F. R. Hutton 
Chas. Kirchhoff 
J. W. Lieb, Jr. 
R. C. McKinney 
Fred J. Miller 
B. M. Mitchell 
Charles A. Moore 



W. W\ Nichols 
H. F. J. Porter 
Albert Spies 

E. G. Spilsbury 
John E. Starr 

F. H. Stillman 
H. H. Suplee 
Stevenson Taylor 
A. M. Waitt 

Ira H. Woolson 



Sub-Committee on Reception and Dance 
Thursday Evening 



Prof. Arthur L. Williston, Chairman 



Edgar H. Berry 
Wm. H. Boehm 
A. P. Boiler, Jr. 
Reginald P. Bolton 
G. I. Bouton 
H. R. Cobleigh 
Frank E. Eberhardt 



Harrington Emerson 

E. H. Foster 
Henry S. Hay ward 

F. V. Henshaw 
F. E. Matthews 
David M. Myers 
J. H. Norris 

E. H. Peabody 



H. O. Pond 
L. H. Snyder 
Theodore Stebbins 
A. F. Stillman 
J. Stewart Thomson 
Edw. Van Winkle 
F. A. Waldron 



Assignments for the Reception of Members 



Tuesday 



Afternoon 

F. A. Halsey, Chairman 
H. P. Ahrnke 
H. A. Foster 
H. S. Isham 



Morning 

Albert Spies, Chairman 
Percy Allan 
C. G. de Laval 
F. H. Taylor 



Evening 

Chas. Whiting Baker, Chairman 
W. P. Pressinger 
Fred. E. Rogers 
H. M. Rowley 



Wednesday 

Afternoon 

G. A. Orrok, Chairman 
Geo. B. Caldwell 
A. Falkenau 
W. R. Hulbert 



Evening 

H. G. Stott, Chairman 
F. H. Colvin 
R. E. Fox, Jr. 
J. P. Sparrow 



S92 



SOCIETY AFFAIRS 



Thursday 



Morning 

F. L. Du Bosque, Chairman 
L. P. Alford 
Geo. H. Barbour 
Anson W. Burchard 



Afternoon 

James T. Whittlesey, Chairman 
John J. Boyd 
N. B. Payne 
R. P. Bolton 



Friday Morning 

Fred R. Low, Chairman 
G. R. Henderson 
E. B. Katte 
Geo. Dinkel 



ladies' reception committee 



Mrs. Herbert Gray Torrej^, Chairman 



Mrs. H. C. Abell 
Mrs. Charles W. Baker 
Mrs. G. H. Barbour 
Mrs. A. R. Baylis 
Mrs. E. H. Berry 
Mrs. Wm. H. Boehm 
Mrs. R. P. Bolton 
Mrs. G. I. Bouton 
Mrs. Edward Ciardi 
Mrs. J. VanV. Colwell 
Mrs. H. Emerson 
Mrs. G. L. Fowler 
Mrs. F. J. Gubelman 
Mrs. F. A. Hall 
Mrs. N. H. Hiller 
Mrs. H. F. Holloway 



Mrs. G. S. Humphrey 
Mrs. A. C. Humphreys 
Mrs. C. W. Hunt 
Mrs. F. R. Hutton 
Mrs. F. E. Idell 
Mrs. P. C. Idell 
Mrs. D. S. Jacobus 
Mrs. J. E. Jones 
Mrs. J. A. Kinkead 
Mrs. G. L. Knight 
Mrs. J. W. Lieb, Jr. 
Mrs. H. S. Loud 
Mrs. Fred R. Low 
The Misses Meier 
Mrs. C. W. Obert 
Mrs. G. A. Orrok 
Miss Eugenie Price 



Mrs. Calvin W. Rice 
Mrs. E. N. Sanderson 
Miss Marion R. Scheffler 
Mrs. Horace See 
Mrs. Jesse M. Smith 
Mrs. J. P. Sneddon 
Miss Jean Sneddon 
Mrs. H. H. Suplee 
Mrs. Stevenson Taylor 
Mrs. Edward Van Winkle 
Mrs. S. E. Whitaker 
Dr. Lucy O. Wight 
Mrs. Wm. H. Wiley 
Mrs. A. L. Williston 
Mrs. Jas. Edw. Wilson 
Mrs. Ira H. Woolson 



Chairmen Committees for the Several Days 

Tuesday Afternoon Mrs. John W. Lieb, Jr. 

Wednesday Morning Mrs. James Edward Wilson 

Wednesday Afternoon Mrs. Ira H. Woolson 

Thursday Morning Mrs. J. P. Sneddon 

Thursday Afternoon Dr. Lucj"^ O. Wight 

„ . , ,, . /Mrs. C. W. Hunt 

Friday Mornmg (Mrs. F. A. Hall 



SOCIETY AFFAIRS 393 

Executive Committee 

Mrs. Herbert Gray Torrey, Chairman 

Mrs. F. A. Hall Mrs. John W. Lieb, Jr. 

Mrs. G. S. Humphrey Mrs. J. P. Sneddon 

Mrs. C. W. Hunt Dr. Lucy O. Wight 

Mrs. J. E. Jones Mrs. James Edward Wilson 

Mrs. Ira H. Woolson 

ladies' guides 

Mrs. G. S. Humphrey \ chairmen 
Mrs. J. E. Jones J 

Mrs. Edward Ciardi Mrs. J. P. Sneddon 

Mrs. John W. Lieb, Jr. Mrs. S. E. Whitaker 

Miss Jean Sneddon Dr. Lucy O. Wight 

ACCOUNT OF THE ANNUAL MEETING 

OPENING SESSION, TUESDAY EVENING 

Vice-President Fred J. MiJler called the session in the auditorium 
to order and presented President Jesse M. Smith, who delivered his 
address on The Profession of Engineering in which he dealt with 
the need of cooperation among engineers, looking toward the mainten- 
ance of high standards in engineering practice. 

Following the address, Theodore Stebbins, chairman of the Tellers 
of Election, presented to the President the report on the election of 
oflficers and the following were thereupon declared elected: For presi- 
dent, George Westinghouse ; for vice-presidents, Charles Whiting 
Baker, W. F. M. Goss, E. D. Meier; for managers, J. Sellers Bancroft, 
James Hartness, H. G. Reist; for treasurer, William H. Wiley. 

President Smith then called on Past-Presidents Worcester R. 
Warner, Geo. W. Melville and Samuel T. Wellman to escort Presi- 
dent-elect George Westinghouse to the platform. 

After his notification of election and introduction to the members, 
the president-elect spoke as follows: 

When Mr. Warner, the Chairman of your Nominating Committee, after first writ- 
ing on the subject, came to Lenox to ask me to accept the nomination for president 
of this great Society, I had already decided that it would be impossible for rae to 
have the privilege of accepting; but after he had explained to me the desires of hir 



394 SOCIETY AFFAIRS 

associates and had represented to me that it was the unanimous wish of all of the 
members of your Nominating Committee to honor me at this particular time, and 
in so doing to express an appreciation of my efforts and accomplishments in the 
engineering field, I with much hesitation consented to accept the nomination and 
promised if elected to do everything in my power. 

Whether two mistakes have been made — one in yielding to the persuasive words 
of Mr. Warner, and the other in my election as your president — the forthcoming 
year will determine. I trust I may be able to fulfil your expectations by adding 
something to the worldwide reputation of The American Society of Mechanical 
Engineers. 

With these remarks, I now accept with feelings of deep gratitude the honor 
which the members of the Society have tonight unanimously conferred upon me. 

There never was a time in the history of the world when honest, wise and con- 
servative action is more strongly demanded of us and of all men than now, if \vc 
have any desire to preserve the right to comfortably carry on our various 
affairs. 

I thank you, and I ask your cooperation in my efforts to perform my duties as 
your president. 

The meeting was then adjourned to the rooms of the Societ}- whein 
the members and guests were introduced by Secretary Calvin AA'. 
Rice, to the President-elect and Mrs. Westinghouse, who were assisted 
in receiving by President Jesse M. Smith and Mrs. Smith, and' Honor- 
ary Secretary F. R. Hutton and Mrs. Hutton. 



WEDNESDAY EVENING LECTURE 

As already stated the lecture on Wednesday evening was on the 
Era of Farm Machinery,by L.W.Ellis, of the Bureau of Plant Industry 
of the United States Department of Agriculture at Washington, D. C. 
The lecture was illustrated by lantern slides. Mr. Ellis first gave an 
idea of agricultural progress, by describing some of the most sti'iking 
mechanical achievements found on Western farms of the present day. 
He first described early farm implements and told briefly of the 
transition from hand to machine methods. In 1 SOO wheat was sown 
broadcast by hand, after the ground had been plowed with a heavy, 
clumsy, wooden plow, requiring as many as eight oxen to pull it. 
Sickles cut the grain, and it was bound by hand. During the suc- 
ceeding winter it was threshed out either by a flail or by driving 
animals over it as it lay in heaps. It was finally winnowed by hand. 

Corn cultivation was by the hoe, or a rude shovel plow. The 
stalks were cut and the ears husked out by hand. Shelling was done 
by scraping the ears against the handle of a frying pan, a bushel in 
one hundred minutes. 



SOCIETY AFFAIRS 395 

Hay was cut with a scythe and was pitched by hand from ground 
to cart, and cart to haymow. BaHng and shipping were practically 
unknown. Hand methods prevailed in the dairy, the stable, the cot- 
ton fields, the potato patch, in fact in every phase of production. 

From 1855 to 1894 the human labor consumed in producing a bushel 
of corn by the best available methods declined from four hours and 
thirty minutes to forty-one minutes, and for shelling it from one hun- 
dred minutes to one minute. In 1830, three hours and three minutes 
of human labor were requii'ed to raise and thresh a bushel of wheat; 
in 1896 ten minutes. Eleven hours were required to cut and cure a 
ton of hay in 1860, and but one hour and thirty-nine minutes in 1894. 

Power corn shelJers now used have a capacity of from one hundred 
to eight hundred ])ushels per day. The cobs are carried to a pile and 
the shelled corn delivered into sacks or wagons. The fuel value of 
the cobs pays the cost of shelling. 

Though hand methods still prevail in some sections, the mower is 
now practically the universal means of cutting the hay crop. This 
is a modification of the early reaping machines with such factors elim- 
inated as are not necessary for cutting the grass. The steel self- 
dump rake, the side-deHvery rake and the hay loader, the stacker, and 
the baling press are other developments for hay harvesting. 

In the extreme West there has been developed the combined har- 
vester which seems to represent the greatest possible saving of human 
labor. This machine, drawn by from twenty to forty horses, under 
control of a single driver, cuts, threshes, recleans, and delivers into 
sacks the grain from forty to fifty acres per day. Two men are re- 
quired for sewing the sacks. The straw, including all weed seeds, is 
distributed over the ground as the team proceeds. On level land the 
horses may be replaced by the steam engine, which furnishes power 
sufficient to cut a swath up to forty feet in width and to cover from 
seventy-five to one hundred and twenty-five acres per day. 

For general farm work the internal-combustion tractor may be 
said to be rapidly supplanting the steam engine, which, however, has 
a great field of usefulness in sections where it is desired to bring large 
areas rapidly under cultivation. In older sections, in order to com- 
pete successfully with the horse, tractors must bring the cost of ope- 
ration close to the cost with horses and at the same time be capable 
of a great variety of work. The internal-combustion tractor meets 
these conditions better than the steam engine, and is being introduced 
at a rate estimated anywhere from two thousand to five thousand 
per year. 



396 SOCIETY AFFAIRS 

The automobile is rapidly finding a place in the business manage- 
ment of the farm. It takes from the heavy draft horse the necessity 
for long, exhausting trips to town on light errands. 

In general, machinery has reduced the cost of producing farm pro- 
ducts. It has improved the quality of products by condensing crop 
operations within the period when the most favorable conditions pre- 
vail. By increasing the acre effectiveness of a man it has reduced 
the labor necessary to produce the nation's food supply, leaving it free 
to assist in development along other lines. At the same time it has 
thrown upon the cities the burden of providing work for an ever- 
increasing army of non-producers. It has increased the investment 
necessary for the proper organization of a farm, this and the price of 
land making it more difficult for a person of small capital to engage in 
farming. 

As a nation we have occupied nearly all of our naturally productive 
area and are confronted with the necessity of providing food for an 
increasing population with a constant acreage. In the past, machin- 
ery has encouraged extensive rather than intensive farming. Hence- 
forth the reverse should be true. If he who makes two blades of 
grass grow where one grew before, is a pubhc benefactor, then none 
the less is he a pubhc servant who puts into the farmer's hands the 
machinery for making such a course attractive. 

BUSINESS MEETING 

The business session on Wednesday mommg was called to order by 
President Jesse M. Smith. Secretary Calvin W. Rice read the annual 
report of the Council. The Secretary then read the report of the 
Tellers of Election of members, including 166 apphcants for mem- 
bership and 21 for advance in grade. 

The next in order was the consideration of the proposed amend- 
ments to the Constitution. The first amendment relates to C 10 
on associate membership, which reads as follows: 

C 10 An Associate shall be 26 years of age or over. He must either have the 
other qualifications of a member or be so connected with engineering as to be com- 
petent to take charge of engineering work, or to cooperate with engineers. 

The proposed amendment reads as follows: 

An associate member shall be thirty years of age or over; he must have been so 
connected with some branch of engineering, or science, or the arts, or industries, that 
the Council will consider him quaUfied to cooperate with engineers in the advance- 
ment of professional knowledge. 



SOCIETY AFFAIRS 397 

Another amendment relates to the clause on Junior Membership 
which now reads as follows: 

C 11 A Junior shall be 21 years of age or over. He must have had such engi- 
neering experience as will enable him to fill a responsible subordinate position in 
engineering work, or he must be a graduate of an engineering school. 

The following addition is proposed by the Committee on Constitu- 
tion and By-Laws: 

A person who is over 30 years of age can not enter the Society as a Junior. 

Both these amendments have been approved by the Committee on 
Membership. It therefore remains for the members to vote on them 
by letter ballot. 

A third proposed amendment to the Constitution relates to the 
formation of an additional standing committee. This was presented 
at the Washington meeting in the form of a resolution, as follows: 

Resolved, That we recommend to the Council the appointment of a Public 
Relations Committee, to investigate, consider and report on the methods whereby 
the Society may more directly cooperate with the public on engineering matters 
and on the general policy which should control such cooperation. 

It was moved and seconded that this also be referred to the mem- 
bers for letter ballot. 

Dr. D. S. Jacobus, Chairman of the Committee on Power Tests, then 
made a verbal report. This committee was appointed to revise all 
the codes relating to power tests, some of which did not agree with 
others, or were not up to date. It had been decided to blend the 
whole into one report rather than present a series of reports, as on 
engine testing, boiler testing, etc. The first part of the report will 
deal with tests in general, caHbration of apparatus, units, etc., while 
the second part will be subdivided for the various classes of machines 
and apparatus. 

Dr. Jacobus also made a verbal report for the Joint Committee on 
a Standard Tonnage Basis for Refrigeration. This committee had 
made a preliminary report in 1904 and suggested certain units for 
measuring the refrigerating capacity of the machinery. They had 
also suggested a standard set of conditions under which a machine 
should be tested to obtain the refrigerating capacity of that machine. 
Later on, the work of the committee was extended, and they were 
asked to recommend a method of testing the machines. A prehm- 
inary report was also prepared on this portion of the work and had 
been before the Society. 



398 SOCIETY AFFAIRS 

Though the committee had received some favorable discussion on 
the report they felt that it was not a complete piece of work, and they 
wished that some one would give the committee additional light on 
how the report could be made. Furthermore, there were many places 
in the report where the committee could not make any definite recom- 
mendations, because they did not have enough data at hand. 

A resume of the work that has been done by the Committee on Re- 
frigeration was prepared and sent to the Congress of Refrigerating 
Industries, held in Paris in the fall of 1908, with the request that it be 
discussed. In making this resume certain questions were asked, on 
which the committee wished to obtain specific information. This 
was done in a semi-official way, and after taking up the matter with 
the Secretary of this Society, Dr. Jacobus, speaking on the behalf 
of the committee, concluded the communication to the International 
Committee as follows: 

The policy of The American Society of Mechanical Engineers has always been 
for the advancement of the arts, and whereas it is only natural that it should take 
pride in participating in advancements, it will never look except with satisfac- 
tion upon activities of other bodies, even in the subjects on which it has worked. 

I feel safe in saying, therefore, that any criticism by the members of this organi- 
zation on the work which has been done in connection with the subject at hand 
will be gladly received. Criticism leads to the establishment of better and more 
up-to-date methods, and what The American Society of Mechanical Engineers is 
after, and what I am sure we are all after, is to work hand in hand for the good 
of the cause. 

I also feel safe in saying that The American Society of Mechanical Engineers 
will cooperate in every way in the endeavor to establish some standard set of 
rules which shall conform with the views of such able experts as are gathered in 
this meeting. It is certainly hoped that the matter presented in this paper 
will receive a thorough discussion, irrespective of whether those who take part 
agree or disagree with the findings of the committee. 

About the same time, a request was made by the committee that 
it should be allowed to cooperate with a committee of the American 
Society of Refrigerating Engineers, so that if this general committee 
recommended certain units, they would really be used by both socie- 
ties. A committee of five was appointed by the American Society 
of Refrigerating Engineers to cooperate with the committee of five 
of The American Society of Mechanical PJngineers. This combined 
committee had already held one meeting and sent out a circular letter 
to a number of refrigerating engineers, reA'iewing the units that had 
been recommended by the Society, and asking for an opinion regard- 
ing these specific units. A great number of replies had been received, 



SOCIETY AFFAIRS 399 

showing how much interest there is in the subject. Most of the re- 
pKes said either that the units were acceptable to those who had read 
the letter, or that they would leave the selection of the units entirely 
in the hands of the committee. The committee therefore has a very 
good working basis, and hopes within a comparatively short time to 
be able to present the results of its work. 

Dr. C. E. Lucke then abstracted the report of the Gas Power Stand- 
ardization Committee, of which he is chairman. The report was dis- 
cussed by Dr. D. S. Jacobus, Prof. R. H. Femald, A. A. Gary, Edwin 
D. Dreyfus and L. B. Lent. 

The report of the Gas Power Plant Operations Committee was pre- 
sented by F. R. Low in the absence of I. E. Moultrop, chairman of the 
committee. The report was discussed by Prof. R. H. Femald, Ed- 
win D. Dreyfus, and Arthur J. Wood. 

THURSDAY MORNING SESSION 

The Thursday morning session was devoted to papers on the meas- 
urement of the flow of fluids. 

The first paper presented was on Tests on a Venturi Meter for Boiler 
Feed, by Prof. C. M. Allen, of Worcester Polytechnic Institute. The 
object of these tests with the venturi meter was to determine how 
well adapted it would be for use in measuring the feed to a boiler, in 
view of the variety of conditions under which it might have to oper- 
ate such as the methods of pumping the wate^ through the meter, the 
different temperatures of the water pumped, various and fluctuating 
pressures and velocities of flow, etc. The results obtained indicate 
that such occurrences have practically no effect on the satisfactory 
perfoiTnance of the meter. Though there are limits to the satis- 
factory operation of a meter, the tests indicate that the venturi 
meter is sufficiently accurate for the majority of commercial or engi- 
neering requirements. 

The paper was discussed by F. N. Connet and Clemens Herschel, 
Dr. Sanford A. Moss and Prof. L. S. Marks submitting written dis- 
cussions. 

The next paper. Efficiency Tests of Steam Nozzles, by Prof. F. H. 
Sibley of the University of Alabama, was read by Prof. C. C. Thomas 
of the University of Wisconsin. The object of the test was to deter- 
mine the efficiency of various shaped nozzles with steam flowing from 
a given initial pressure to a known vacuum; also to determine 
the effect on the efficiency of changing the angle of divergence. 
Two methods were tried out for finding this efficiency: (a) by 



400 80CIETT AFFAIRS 

the pressure in the nozzle by means of a search tube placed axiaUy 
in the nozzle; (6) by the reaction of the nozzle by suspending 
it in an air-tight box at the end of a flexible steel tube. The deflection 
of the tube caused by the reaction of the nozzle was measured by a 
calibrated spring. The results of the tests indicate: (a) that the 
reaction is affected by a difference Jh pressure between the muzzle of 
the nozzle and the medium surrounding the nozzle; (6) that the effi- 
ciencies of the various nozzles were determined within a probable 
error of 2 percent; (c) that the efficiency is affected more by the smooth- 
ness of finish on the inside of the nozzle than by the exact contour of 
the nozzle. 

A. F. Nagle, A. R. Dodge and Professor Thomas discussed the 
paper, J. A, Moyer submitting a written discussion. 

George F. Gebhardt's paper on The Pitot Tube as a Steam Meter 
was read by the Secretary in the author's absence. The application 
of a pitot tube system as described in the paper is an accurate means 
of determining the velocity of steam at any point in a pipe, provided 
the values of the various influencing factors are known; and for straight 
lengths of piping with continuous flow, under these conditions, it is 
an accurate means of determining the weight of steam flowing. Under 
average commercial conditions in which the pressure and quality 
of the steam fluctuate and an average value must be taken for the 
density of the self-adjusting water column, only approximate results 
can be obtained, the extent varying with the degree of fluctuation. 

Walter Ferris and A. R. Dodge discussed the paper, a written dis- 
cussion by Prof. W. B. Gregory being read by the Secretary. 

The paper on An Electric Gas Meter was presented by the author, 
Prof . Carl C. Thomas, of the University o f Wisconsin. The paper 
describes a meter for measuring the rate of flow of gas or air, which can 
be adapted for use as a steam meter or as a steam calorimeter. The 
operation of the gas meter depends upon the principle of adding elec- 
trically a known quantity of heat to the gas and determining the rate 
of flow by the rise in temperature of the gas (about o deg. fahr.) 
between inlet and outlet. The adoption of this principle of operation 
permits the construction of a very accurate and sensitive autographic 
meter of large capacity containing no moving parts in the gas pas- 
sage; independent of fluctuations in pressure and temperature of the 
gas; and capable of measuring gas or air at either high or low pres- 
sures or temperatures. The electrical energy required is about 1 kw. 
per 50,000 cu. ft. hourly capacity, at the pressures ordinarily used in 
gas mains. 



SOCIETY AFFAIIto 401 

Prof. W. D. Ennis, E. D. Dreyfus and A. R. Dodge discussed the 
paper, a wrirten discussion from Prof. L. S. Marks being also read. 

THURSDAY AFTERNOON — STEAM ENGINEERING 

At the Thursday afternoon session Vice-President L. P. Brecken- 
rid c presided. Five papers were presented deahng with different 
pha.ses of steam engineering. The first paper, Tan Bark as a Boiler 
Fuel, by David M. Myers, described the results obtained by burning 
spent hemlock tan bark, the average fuel value of which is about 9500 
B.t.u. per lb. of dry matter, which is about 35 per cent of its total 
moist weight in the fireroom. The available heat value per pound 
as fired is 26G5 B.t.u. One ton of air-dry hemlock bark produces 
boiler fuel equal to 0.42 tons of 13,500 B.t.u. coal. A. A. Gary, 
Prof. Wm. Kent and Prof. L. P. Breckenridge took part in the dis- 
cussion. 

J. K. Bibbins then presented his paper on Cooling Towers for Steam 
and Gas Power Plants, which contained a critical study of different 
types of towers with a description of their distinctive features. The 
paper also describes a simple inexpensive type of tower employing a 
lath-mat cooling surface and offers suggestions for a combination of 
natural-draft and forced-draft types. 

The paper was discussed by Geo. J. Foran, W. D. Ennis, H. E. 
Longwell, B. H. Coffey, E. D. Dreyfus and F. J. Bryant. A written 
discussion by Carl G. de Laval was read by the Secretary. 

W. P. Caine's paper, Governing Rolling Mill Engines, was read by 
Richard H. Rice. The paper describes and gives indicator cards 
and speed curves of a Coriiss engine driving a three-high mill under 
two different conditions of governing, (a) under the widest range of 
adjustment of cut-off, (6) under a limited range, increasing the econ- 
omy and making the engine run much more smoothly and safely. A 
table gives the power required for rolling in the mill and the momen- 
taiy source of energy, whether from the cylinder or flywheel. A 
description is also given of the tachometer used to take the speed 
curves. Written cUscussions by H. C. Ord and James Tribe were 
read by the Secretary. 

The next paper was that by F. W. Dean on An Experience with 
Leaky Vertical Fire-Tube Boilers. The author discussed the diffi- 
cult.es experienced with some large vertical boilers, somewhat over 
10 ft. in diameter, and containing over 6000 sq. ft. of heatmg surface. 
The boilers leaked badly very soon after being started and nothing 



402 SOCIETY AFFAIRS 

that was done improved their condition until the water legs were 
lengthened from 2 ft. to 7 ft. 2| in., the boilers thus being raised 5 ft. 
2f in. Before they were raised the lower ends of the tubes would 
cover with very hard cUnker and become stopped up. This clinker 
could be removed only by cutting it off when the boilers were cold. 
After the boilers were raised, a Ught clinker that could be blown off 
foiTned about the tubes; by removing this by blowing every three or 
four hours the leaks were stopped and they have never returned. 

Those taking part in the discussion were R. P. Bolton, Prof. Wm. 
Kent, J. C. Parker, 0. C. Woolson, A. A. Gary, Prof. A. M. Greene, Jr., 
E. D. Meier and D. M. Myers. A. Bement submitted a written dis- 
cussion. 

Mr. Dean's second paper. The Best Form of ^Longitudinal Joint for 
Boilers, dealt with the defects of the usual form of butt joint used on 
the longitudinal seams of boilers, in which the inside strap is wider 
than the outside strap. It gave some history of the joint and dis- 
cussed some of its defects and suggested a substitution for this form. 

The paper was discussed by R. P. Bolton, Carl G. Barth, E. D. 
Meier, Prof. A. M. Greene, Jr., W. A. Jones, Prof. S. W. Robinson, 
Geo. I. Rockwood. and Sherwood F. Jeter. 

GAS POWER^SECTION 

The session^of the Gas^Power Section was held on Thursday after- 
noon. Chairman F. R. Low presiding. In his address, the Chan-man 
referred briefly to the work of the various committees of the Section 
and stated that during the year the membership had increased from 
247 to 378, a gain of over 50 per cent. Mr. Low also dealt with the 
development in the gas-power field during the year, mentioning some 
experiments with gas turbines. Gas-engine design, the use of by- 
product gases, the development of the bituminous producer, the gas- 
ification of peat, and the gas engine in marine work, were also briefly 
dealt with. 

The report of the Tellers of Election, Edw. Van Winkle, Prof. Walter 
Rautenstrauch and J. V. V. Colwell, was then presented by Prof. 
Rautenstrauch, the results being as follows: for chairman J. R. Bib- 
bins 107; for member of the Executive Committee, F. R. Low 108. 

The report of the Gas Power Plant Operations Committee was then 
presented by James D. Andrew, and discussed by J. C. Parker, J. N. 
Norris and H. H. Suplee. Prof. C. H. Benjamin reported verbally 
for the Literature Committee, outlining the work of the committee in 



SOCIETY AFFAIRS 403 

bringing gas-power literature to the attention of the members. H. 
R. Cobleigh and Professor Rautenstrauch also spoke on the work of 
this committee, the latter suggesting a plan for better organization 
of the committee to deal with literature on the subject. 

L. B. Lent reported for the Gas Power Installations Committee 
that two forms had been prepared and sent to manufacturers, and 
while a good deal of information had been received, not enough was 
on hand for a complete report. The committee hoped to have the 
material in shape at an early date. 

Prof. W. F. M. Goss then presented the paper on Testing Suction 
Gas Producers with a Koerting Ejector, by C. M. Garland and A. P. 
Kratz. The paper describes a method of testing the suction gas pro- 
ducer which is independent of the engine. The engine is blanked off 
from the producer and a Schutte & Koerting steam ejector is inserted, 
which draws the gases from the producer and delivers them to a scrub- 
ber in which the steam used by the ejector is condensed. The gases 
then pass to a meter for measuring their volume. Complete data of 
calculations and results are given in appendices. 

The paper was discussed by Prof. R. H. Femald, G. M. S. Tait, H. H. 
Suplee, L. B. Lent, S. C. Smith, W. B. Chapman and Edw. N. Trump. 

The paper on Bituminous Gas Producers was then presented by the 
author, J. R. Bibbins. The paper describes a double-zone type of 
producer and the results obtained in gasifying bituminous coal. Con- 
tinuous operation was secured with tar-free gas of reasonable heat 
value and producer efficiency and an over-all plant economy of about 
one pound of fair bituminous coal per brake horsepower (proportionate 
economies for poorer grades). The efficiency and general effec- 
tiveness of operation of the producer on low-grade fuel, lignites, 
etc., was practically as high as with the higher grades. The following 
took part in the discussion: G. M. S. Tait, Prof. R. H. Femald, W. B. 
Chapman, H. M. Latham, H. H. Suplee, Edw. N, Trump, H. B. 
Langer, S. C. Smith, Prof. Walter Rautenstrauch, and G. D. Conlee. 

FRIDAY MORNING 

The session on Friday morning opened with the paper by Walter 
Ferris on The Bucyrus Locomotive Pile Driver. This paper describes 
a new railway pile driver, the leading feature of which is a very power- 
ful propelling apparatus and a large boiler, enablmg it to act as a 
locomotive and haul its own train of tool cars, boarding cars, etc., 
over the road. A special turntable, consisting of hydraulic lifting 



404 SOCIETY AFFAIRS 

apparatus and a large ball-bearing, enables the entire pile driver, 
including trucks, to be turned end for end or crosswise of the tracks. 
0. K. Harlan discussed the paper, A. F. Robinson and L. J. Hotch- 
kiss submitting written discussions. 

The paper by Henry Hess on Lineshaft Efficiency, Mechanical and 
Economic, deFcribed the test of the relative efficiency of a lineshaft 
of 2]^ in. diameter, making 214 r.p.m., with bearing load due to 
the weight of the parts plus the tension of the belts subjected to known 
stress by counterweighting, when running in ring-oiling babbitted 
bearings and when mounted in ball bearhigs. The savings in power 
consequent on this change ranged fi'om 14 to 65 per cent, with 36 
and 35 per cent under average conditions of good practice, due to 
belt tensions of 44 lb. and 57 lb. per inch width of single belt respec- 
tively. The paper gives data for determining the power savings that 
may be expected in various plants, by the use of ball bearings. 

Those discussing the paper were T. F. Salter, Prof. R. C. Carpenter, 
C. A. Graves, O. K. Harlan, C. J. H. Woodbury, Walter Ferris, Fred 
J. Miller, A. C. Jackson, C. D. Parker and Oliver B. Zimmerman. 
Geo. N. Van Derhoff submitted a written discussion. 

A. F. Nagie's paper on Pump Valves and Valve Areas, called the 
attention of engineers to the need of reviewing the common notion 
that " valve-seat area " is synonymous with " velocity of flow. " The 
purpose of specifications for pumping engines is to secure a low veloc- 
ity of flow through the valves, thus reducing the head required to 
force water through the pump; but to accomplish this purpose, special 
and intelligent attention should be given to the springs of the valves, 
rather than to valve-seat areas. If that be done, valve-seat areas 
need not be greater than the plunger area' for the vertical triple- 
expansion pumping engines so largely used in city pumps. Prof. 
Wm. Kent, A. B. Carhart, Prof. R. C. Carpenter and E. H. Foster 
discussed the paper. Contributed discussions were by Chas. A. 
Hague, I. H. Reynolds and F. W. Salmon. 

Another paper by Mr. Nagle, A Report on Cast-Iron Test Bars, 
brought out the fact that test pieces, whether cast in separate molds 
or in the same mold as the main casting, are not perfect indications 
of the character of the iron in the main casting. The results obtained 
by the author would indicate a probable variation of 15 per cent 
where uniformity might be expected. A. A. Car}' and T. M. Phctte- 
place discussed the paper, contributed discussion being by Prof W. 
B. Gregory and Geo. M. Peek. 



SOCIETY AFFAIRS 405 

The meeting closed with the following resolutions, offered by Luther 
D. Burlingame: 

Whereas The American Society of Mechanical Engineers at its 
Annual Meeting, December 1909, desires to express its appreciation 
to those who have provided opportunities for entertainment an d on 
behalf of the visiting members and their guests thanks for the cordial 
welcome extended by the local members and their friends of New York 
and vicinity, 

Be it Resolved that the Secretary extend the thanks of the So( lety 
and express the appreciation of its members and guests to the local 
committee for their untiring el'forts, to those who have sent inAita- 
tions to visit technical and engineering works and places of inteiest, 
to Mr. Geo. Gibbs, chief engineer of the Pennsylvania Tunnel and 
Terminal Railroad Co., and to Mr. Walter Kerr, president of the West- 
inghouse. Church, Kerr & Co., and their associates, for the opportu- 
nity to inspect the new Pennsylvania Railroad station; to Dr. B.T. Gal- 
loway, chief of the Bureau of Plant Industry, Department of Agricul- 
ture, for the very instructive and entertaining paper on The Era of 
Agricultural Machinery, and especially to those ladies who have so 
efficiently assisted by extending a generous hospitality to their guests. 

EXCURSIONS 

As usual at conventions of the Society there were numerous ex- 
cursions to points of interest in New York and vicinity, which con- 
stituted an important feature of the program for the entertainment 
of visiting members and guests. Invitations for these excursions 
were generously extended by many firms and individuals, and through 
the efforts of the Excursion Committee, Hosea Webster, Chairman, 
trips to various plants and industries were arranged, to the represen- 
tatives of which the grateful appreciation of the Society has been 
expressed. 

A list of excursions follows: 

Pennsylvania Railroad Terminal and Passenger Station: Invitation by George 
Gibbs, Chief Engineer, Pennsylvania Tunnel Terminal R. R. Co., and member of 
the Society; Henry R. Worthington Hydraulic Works, Harrison, N. J., by William 
Schwanhausser, Chief Consulting Engineer of International Steam Pump Co., 
member of the Society; Ha^-rison Lamp Works of General Electric Co., Harrison, 
N. J., by George H. Morrison, General Manager; Interborough Rapid Transit Co., 
central power station at 59th St., New York, by H. G. Stott, Superintendent of 
Motive Power, Manager of the Society; Edison factories and Edison Laboratory 
at Orange, N. J., by Frank L. Dyer, President of National Phonograph Co., asso- 



406 SOCIETY AFFAIRS 

ciate member of the Society; De La Vergne Machine Co., New York, by Adolf 
Bender, President; New York Telephone Co.; Gramercy and Stuyvesant Central 
Offices, by E. F. Sherwood, Superintendent of Traffic; Crocker-Wheeler Co., 
Ampere, N. J., by S. S. Wheeler, President, member of the Society; Westinghouse 
Lamp Co., Bloomfield, N. J., by Walter Carey, General Manager; New York Edi- 
son Co., Waterside Stations Nos. 1 and 2, by John W. Lieb, Jr., 3d Vice-President, 
member of the Society; Astoria Light, Heat & Power Co., Astoria, N. Y., by Wil- 
liam H. Bradley, Chief Engineer, Consolidated Gas Co., member of the Society; 
BrookljTi Rapid Transit Co., Williamsburg Power Station, by C. E. Roehl, Elec- 
trical Engineer; Rockland Electric Co., Hillburn, N. Y.; Singer Building, New York, 
by Singer Mfg. Co.; Trenton Iron Co., Trenton, N. J.; Watson-Stillman Co., Am- 
pere, N. J.; Metropohtan Life Insurance Building, New York. 

Every possible courtesy was extended to the visiting parties in each 
case and in some instances special transportation facihties were pro- 
vided. At the Edison Laboratory visitors were met by Thomas A. 
Edison, Hon.Mem.Am.Soc.M.E., who personally explained many 
points of interest about the plant. The Information Bureau, located 
in the foyer of the building, under the chairmanship of F. E. Idell, 
was of material aid in this connection with the trips. 

ENTERTAINMENT FEATURES 

The Ladies' Reception Committee, composed of ladies resident in 
and about New York, vmder the chairmanship of Mrs. Herbert Cray 
Torrey, contributed much to the pleasure of members and guests 
(if the Society. Tea was served from four until six o'clock on Tues- 
elay, Wednesday and Thursday afternoons during the convention, 
in the ladies' headquarters, located in the reception rooms of the 
Society on the eleventh floor. Mrs. George H. Westinghouse was 
the guest of the committee on Wednesday afternoon. 

A number of excursions to shops and hotels were arranged and suc- 
cessfully carried out under the guidance of members of the committee 
The kindness of Mr. and Mrs. John W. Lieb, Jr., made possible several 
enjoyable automobile rides through Central Park and Riverside Drive- 



No. 1249 

THE ANNUAL REPORT OF THE COUNCIL AND 
COMMUrTEES 1909 

REPORT OF THE COUNCIL 

The Society entered upon a distinct epoch in its history when the 
Council approved the recommendation of the Meetings Committee 
that meetings of the Society be held periodically in cities other than 
New York, thus satisfying a long-felt desire on the part of the 
membership, as well as of the Council and the Meetings Committee, to 
extend as fully as possible the benefits of membership in the Society. 

As a result meetings have been successfully held in Boston and St. 
Louis. In the former place, the attendance has been even larger in 
some cases than the meetings in New York. The spirit of coopera- 
tion has been developed and although these are meetings of the 
Society, fellowship in the profession has been promoted in each center 
by the participation in the meetings of the membership of local 
engineering societies and engineers generally. 

Inquiries are constantly being received from other centers for in- 
formation respecting the holding of meetings, and every encourage- 
ment is being rendered and assistance pledged by the Society to make 
it possible for groups of the members in any locality to hold meetings. 

Through the policy of conducting these meetings as meetings of 
the Society rather than of sections or branches, the solidarity and nat- 
ional character of the Society is at once developed. All meetings 
are conducted in all places on the same basis with an equally high 
standard and before publication in The Journal all papers and dis- 
cussions thereon must be approved by the same committee, viz., 
the Meetings Committee; and no papers may be read or discussed 
that are not of a uniformly high grade and suitable and worthy 
of publication for the benefit of the entire membership. 

STUDENT. BRANCHES 

The nmnber of student branches affiliated with the Society which 
have been formed in colleges and universities during the past year 



408 SOCIETY AFFAIRS 

show the importance of another movement. Seventeen of these 
branches have been established and the reports of their meetings 
which have appeared at intervals in The Journal indicate a keen 
interest on the part of these organizations and show that here is a 
work that the Society may well foster. The basis of affiliation of 
these student societies with The American Society of Mechanical En- 
gineers is a broad one, and provides for the maintaining of each branch 
under its own by-laws subject only to limitations set by the Council of 
the Society. The Journal is furnished to each member for the nomi- 
nal sum of $2 a year and, in addition, advance copies of papers to be 
presented before the Society are supplied gratis for discussion at meet- 
ings. Papers for local representation may also be printed and sup- 
plied at cost to the affiliated branches. A list of the branches 
includes : Stevens Institute of Technology, Hoboken, N. J. ; Cornell 
University, Ithaca, N. Y.; Armour Institute of Technology, Chicago, 
111.; Leland Stanford Jr. University, Palo Alto, Cal.; Polytechnic In- 
stitute of Brooklyn, Brooklyn, N. Y. ; State Agricultural College, 
Corvallis, Ore.; Purdue University, Lafayette, Ind.; University of 
Kansas, Lawrence, Kan.; New York University, New York; Univer- 
sity of Illinois, Urbana, 111.; Pennsylvania State College, State Col- 
lege, Pa.; Columbia University, New York; Massachusetts Institute 
of Technology, Boston, Mass.; University of Cincinnati, Cincinnati, 
0.; University of Wisconsin, Madison, Wis.; University of Mis- 
souri, Columbia, Mo.; University of Nebraska, Lincoln, Neb. 

HUDSON-FULTON EXHIBIT 

The Society's part in the recent Hudson-Fulton celebration was 
the preparation of an interesting exhibit of steamboat models, draw- 
ings, portraits, books, manuscripts, and other material related to the 
development of steam navigation. In making this exhibit, the So- 
ciety had the hearty cooperation of the Smithsonian Institution and 
of the Hamburg-American line, and was able to place on view 
models of early and modern steamboats, the American Museum of 
Natural History loaning show cases for this purpose, and members 
and friends of the Society also helping to make the exhibit of interest 
by loaning or presenting manuscripts, books and drawings. The 
American Society of Mechanical Engineers was the only engineering 
organization as such taking part in the celebration of engineering 
achievement. 

Representatives of the Society, together with the Pennsylvania 
Society, on September 24th placed a wreath on the Fulton monument 



SOCIETY AFFAIRS 409 

erected by this Society in Trinity churchyard. The Rev. Dr. William 
T. Manning, Rector of Trinity Church, conducted the service. 

A description of the improvements in the decorations and the re- 
arrangement of the rooms of the Society is contained in the Annual 
Report of the House Committee. 

The same report contains also a description of the mahogany desk 
formerly belonging to Edwin Reynolds, Past-President of the Society, 
donated to the Society by Mrs. Reynolds. 

THURSTON MEMORIAL 

As stated in the Transactions of last year, permission was obtained 
from the Alumni Committee of Sibley College, Cornell University, 
to place in the rooms of the Society, a bronze replica of the Thurston 
memorial tablet at Cornell University. Arrangements for its execu- 
tion were made with the sculptor, H. A. MacNeil, a personal friend of 
Dr. Thurston, ^.nd the tablet is now in place in the entrance hall. 
The figure is about three-quarter life size and below it is the inscrip- 
tion. 

1839 ROBERT HENRY THURSTON 1903 

FiusT President 

AMERICAN SOCIETY MECHANICAL ENGINEERS 

The committee having the matter in charge were: Dr. Alex. C. 
Humphreys, Chairmin, Dr. R. C. Carpenter, Charles Wallace Hunt, 
J. W. Lieb, Jr., Fred J. Miller. 

The Society was represented by Honorary Vice-Presidents on the 
following occasions: 

Commencement Exercises of Columbia University, Jesse M. Smith; Inaugu- 
ration of Richard Cockburn MacLaurin as President of Massachusetts Insti- 
tute of Technology, Worcester R. Warner and Calvin W. Rice; National Con- 
servation Congress, Seattle, Wash., R. M. Dyer, M. K. Rogers, W. F. Zimmer- 
mann; American Mining Congress, Goldfield, Nevada, Dr. J. A. Holmes; Inter- 
national Association for Testing Materials, Chas. B. Dudley; funeral of Edwin 
Reynolds, E. T. Adams, F. M. Prescott, E. T. Sederholm, W. J. Sando and 
James Tribe ; funeral of F. H. Boyer, G. H. Barrus, F. W. Dean, Gaetano Lanza, 
G. H. Stoddard, Dr. C. J. H. Woodbury. 

The following resignations were accepted during the fiscal year: 

VV. S. Auchincloss, G. W. Blanchard, C. E. Brown, Chas. J. Carney, R. T. 
Close, Fred Collins, M. T. Conklin, S. G. Colt, B. J. Dashiell, H. H. Dixon, 



410 



SOCIETY AFFAIRS 



W. L. Draper, Saml. W. Dudley, Thomas Farmer, Jr., W. Flint, M. L. Foucard, 
Alex. Gordon, M. M. Green, E. B. Gutherie, O. V. de Gaigne, E. E. Hanna, 
W. L. Hedenberg, Jas. Inglis, T. A. Holies, Edmund Kent, C. W. Kettell, C. C 
King, F. C. Kretschmer, A. G. Linzee, J. W. Loveland, Jas. H. Massie, F. Mack- 
intosh, Alfred Marshall, L. M. Northrup, A. T. Porter, A. S. Pritchard, H. S. 
Richardson, L. C. Schaeffer, E. L. Ross, L. N. Sullivan, Marshall L. Whitney, 
R. H. Whitlock. 

Membership of the following has lapsed during the fiscal year : 

M. L. Abrahams, Chas. B. Bruger, H. M. Coale, H. S. Deck, F. H. Davis,.C. M- 
Einfeldt. Robt. P. Fritch, J. M. Garza Aldape, A. A. Hale, M. J. Hammers, R. 
R. Harkins, L. E. Harper, B. U. Hills, L. A. Holeman, O. H. Klein, D. H. Lo- 
pez, Harry G. Manning, Chas. F. Mantine, E. S. Matthews, W. J. P. Moore, 
Wm. H. Moulton, A.' W. Mellowes, C. W. Marx, F. J. McMahon, E. C. Patter- 
son, F. D. Potter, J. A. Prescott. J. L. Ranch, Fred L. Ray, F. S. Ruttmann, 
G. T. Simpson, H. W. Stacy, R. L. Shipman, O. P. Sells, w'm. E. Toelle, W. O. 
Teague, Geo. B. Wilson, H. W. Woodward, Chas. H. Young. 

The membership has increased during the fiscal year as here indi- 
cated : 



1 






LOSSES 




ADDITIONS 


1 1 




GRADE 


1908 


Transfer 


Resig- 
nation 


Lapsed 


Death 


Trans- 
fer 


Elec- 
tion 


INCREASE 


1909 


Honorary 


15 








1 




1 




15 


Members 


2357 




18 


10 


23 


35 


142 


126 


2483 


Associates 


366 


11 


5 


4 


4 


11 


42 


29 


395 


Juniors 


786 


35 


10 


11 


4 




108 


48 


834 


Total 


3524 


46 


33 


25 


32 


46 


293 


203 


3727 


AflaUates of C 


Jas Power 
)tudent Br 


Section . . 












50 
194 


150 


Affiliates of £ 














194 



















The losses by death reported during the fiscal year number the 
following: 

Honorary Member: Gustav Canet; Members: W. M. Allen, W. H. Bailey, 
F. H. Boyer, A. J. Caldwell, K. Chickering, D. H. Gildersleeve, H. F. Glenn, 
Thomas Gray, .C. L. Hildreth, W. E. Hill, Robert Hoe, W. S. Huyette, E. H. 
Jones, J. Landsing, R. B. Lincoln, Alex. Miller, A. W. K. Pierce, F. A. C. Per- 
rine, W. T. Reed, E. Reynolds, R. H. Soule, Geo. W. West, A. R. Wolff; Asso- 
ciates: Thomas H. Briggs, Geo. W. Corbin, G. Eberhardt, E. L. Jennings; 
Juniors: Albert K. Ashworth, Archibald W. Blair, J. R. Rand, A. E. Wellbaum. 

The membership has doubled in the last 11 years. The number of 
applications favorably reported during the year 1909 was 290 for 
admission. 45 for transfer. 



SOCIETY AFFAIKS 411 

With the number of men eminent in the profession this is a rela- 
tively small increase and on account of the benefits which accrue to 
membership and the importance of extending the Society's influence 
the members might very properly bring to the attention of engineers 
of attainment the desirability of securing membership in the Society. 

An amendment to C 45 of the Constitution, involving the appoint- 
ment of a Standing Committee on Public Relations, to investigate, 
consider and report on methods whereby the Society may more 
directly cooperate with the public on engineering matters, and on the 
general policy which should control such cooperation, was proposed at 
the Spring Meeting and has been approved. 

The Committee on Revision and Extension of the Code for Test- 
ing Gas Power Machinery, Chas. E. Lucke, Chairman, E. T. Adams, 
George H. Barrus^ D. S. Jacobus and Arthur West, was discharged, 
and a Committee on Power Tests was appointed by the President, 
consisting of D. S. Jacobus, Chairman, Edward T. Adams, Geo. H. 
Barrus, L. P. Breckenridge, William Kent, Chas. E. Lucke, Edw. 
F. Miller, Arthur West and Albert C. Wood. The purpose of this 
committee is to revise the present testing codes of the Society relating 
to boilers, pumping engines, locomotives, steam engines in general, 
internal-combustion engineS; and apparatus and fuel therefor; to ex- 
tend these codes so as to apply to such power-generating apparatus, 
as is not at present covered, including water-power apparatus, and to 
bring them into harmony with each other and with the best practice 
of the day. The committee is empowered to confer with other engi- 
neering bodies for the purpose of cooperation. 

The Committee on Boiler Code, consisting of J. W. Lieb, Jr. and 
Fred. W. Taylor, reported a revision of the Standard Code for Boiler 
Tests as desirable in view of the progress made in the art since the 
formulation of the code. 

A Committee on Standards for Involute Gears, consisting of Wil- 
fred Lewis, Chairinan, Hugo Bilgram, E. R. Fellows, Chas. R. Gab- 
riel and Gaetano Lanza, was appointed to formulate standards for 
involute gears and report to the Council. 

The following were appointed members of the Research Committee • 
W. F. M. Goss, Chairman, James Christie, R. C. Carpenter, R. H. 
Rice, Chas. B. Dudley. 

The report of George H. Barrus, P. W. Gates and W. F. M. Goss, 
members of the Government Advisory Board on Fuels and Structural 
Materials, U. S. Geological Survey, was received and placed on file. 

Worcester R. Warner, Chairman, Walter M. McFarland, Morgan 



412 SOCIETY AFFAIRS 

Brooks, David Townsend and Francis W. Dean were appointetl a 
Nominating Committee. 

The request of a number of members of the Society for the organiza- 
tion of a Machine Shop Section was received and referred for action 
to the Meetings Committee, with the suggestion that a sub-committee 
to treat the subject be formed rather than a section of the Society. 

The invitation extended to the Society by the Institution of Mechan- 
ical Engineers of Great Britain, for a joint meeting in England in 1910 
has been accepted and a large number of members have already sig- 
nified their intention of attending. Arrangements will probably hv 
made for the transportation of the party in a single steamer. 

The courtesies of the library and rooms of the Society were ex- 
tended b}^ the President and Secretary to the Japanese Honorary 
Commercial Commission and the professional members attended a 
meeting of the Society. 

At a gathering of representatives of the four national engineering 
societies on April 13, the John Fritz Medal was awarded to Charles 
T. Porter, Honorary Member of the Society, for his development of 
the high-speed steam engine. 

The Society also took a prominent part in the bringing together 
in a joint meeting of the four national engineering societies for the 
discussion of our national resources. This was the first meeting of 
its kind and it is to be hoped that many other occasions will be offered 
for cooperation. 

FINANCES 

The Finance Committee has carefully guided the affairs of the So- 
ciety so that notwithstanding increased activities the excess of income 
over expense is $4232.79. Of this amount $3010.77 represents 10 per 
cent of the reserve fund which for some considerable time in accord- 
ance with a resolution of the Council has been transferred annually 
from the reserve to the income account. 

It is a source of satisfaction to report that the Society is now so 
strong that this transfer will be discontinued. 



SOCIETY AFFAIRS 41;^ 



REPORTS OP' STANDING COMMITTEES 

Report of the Finance Committee 

The Committee submits the statements of the financial condition 
of the Society, together with the report of Peirce, Struss & Co., of 
New York, certified public accountants, who have audited the books 
and accounts. 

Peirce, Struss & Co. 

Certified Public Accountants 

37 Wall Street, New York 

November 8, 1909 
Mr. Arthur M. Waitt, 

Chairman Finance Committee 
Dear Sir: 

In accordance with your instructions, we have audited the books and accounts 
of The American Society of Mechanical Engineers for the year ended September 
30, 1909. 

The results of this examination are presented in three exhibits, attached hereto, 
as follows: 
Exhibit A Balance Sheet, September 30, 1909. 

Exhibit B Income and Expense?; based on Cash receipts for year ended 
September 30, 1909. 

Exhibit C Receipts and Disbursements for year ended September 30, 1909. 
We beg to present, attached hereto, our certificate to the aforesaid exhibits. 

Respectfully submitted, 
Peirce, Struss & Co. 

Certified Public Accountants 

Peirce, Struss & Co. 

Certified Public Accountants 

37 Wall Street, New York 

November 8, 1909 
Mr. Arthur M. Waitt, 

Chairman Finance Committee 
Dear Sir: 

Having audited the books and accounts of The American Society of Mechanical 
Engineers for the year ended September 30, 1909, we hereby certify that the 
accompanying Balance Sheet is a true exhibit of its financial condition as of 
September 30, 1909, and that the attached statements of Income and Expenses, 
and Cash Receipts and Disbursements, are correct. 

Peirce, Struss & Co. 

Certified Public Accountants 



414 SOCIETY AFFAIRS 

EXHIBIT A 
Balance Sheet, September 30, 1909 

ASSETS 

Equity in Societies Building (25 to 33 West 39th 

Street) $353 346.62 

Equity, one-third cost of land (25 to 33 West 39th 

Street) 180 000.00 

$533 346.62 

Library Books $13 700.60 

Furniture and Fixtures 2 966 . 96 

16 667 56 
New York City 3J % Bonds 1954, Par, $35,000 .... $30 925 . 00 
Cash in Bank representing Trust Funds 12 918 . 39 

43 843 39 

Stores including plates and finished publications 11 600 . 00 

Cash in Bank for general purposes $7 444 . 83 

Petty Cash on hand 250 . 00 

7 694.83 

Accounts Receivable 

Membership dues $4 924 . 50 

Initiation fees 285 . 00 

Sale of publications, advertising, etc 4 334 . 55 

9 544.05 

Advances account of land subscription fund 7 960 . 94 

Advanced payments 2 214 . 15 

Total assets $632 871 . 54 

LIABILITIES 

United Engineering Society (for cost of land) $81 000 . 00 

Funds 

Life membership Fund $35 151 . 07 

Library Development Fund 4 902 . 71 

Weeks Legacy Fund 1 957 .00 . 

Land Fund Subscriptions 1 227 . 88 

Robert H, Thurston Memorial Fund 399 . 13 

Subscriptions to Annual Meeting 205 . 60 

43 843.39 

Current Accounts Payable 11 163 . 00 

Membership dues paid in advance $494 . 50 

Initiation fees paid in advance 50 . 00 

544.50 



SOCIETY AFFAIRS 415 

Initiation fees uncollected $285 . 00 

Reserve (Initiation fees) 24 596 . 97 

Surplus in property and accounts receivable 471 438 . 68 



Total Liabilities $632 871 . 54 

EXHIBIT B 

Income and Expenses based on Cash Receipts for Year Ended Sep- 
tember 30, 1909 

INCOME 

Membership dues, current $50 273 . 79 

Membership dues, arrears 2 355 . 00 

Sales gross receipts 8 847 . 39 

Advertising receipts 11 997 . 50 

Interest and Discount 1 234 . 68 

ReserveFund, 10% 3 010.77 

$77 719.13 

expenses 

Finance Committee Office Administration including 

Salaries $19 971.91 

Finance, United Engineering Society As- 
sessments 6 000 . 00 

Finance, miscellaneous 983 . 56 

^ $26 955.47 

Membership Committee 2 392.36 

Increase of Membership Committee 147 . 94 

House Committee* 1 192.43 

Library Committee 2 699. 17 

Meetings Committee 

Annual Meeting $2 074 . 24 

Spring Meeting 1 410 . 52 

Monthly Meetings 2 278.19 5 762.95 

Publication Committee 

Advertising Section The Journal .. $7 026.06 

The Journal, except Advertising. ... 13 134 . 80 

Pocket List 1 599.59 

Revises 523.93 

Transactions, Vol. 30. 6 533 . 87 

YearBook 1401.30 

History 43.65 

30 263.20 



' From Current Income $1192.43 

Reserve Fimd 2500.00 



Total Expenses 3892.43 



416 SOCIETY AFFAIRS 

Research Committee $0 . 58 

Committee on Power Test . . 11.25 

Sales Expenditures 4 060.99 

$73 486.34 

Excess of Income over Expenses 4 232 . 79 



$77 719.13 
EXHIBIT C 

Receipts and Disbursements foh Year Ended September 30, 1909 

receipts 

Membership dues $50 832 . 70 

Initiation fees 6 460 . 00 

Membership dues and initiation fees paid in advance.. 551 . 00 

Sales of publications, badges, advertising, etc 20 833.25 

Subscriptions to Land Fund 3 251 .00 

Subscriptions to Expense of Annual Meeting 2 188 . 00 

Interest 2 072 . 24 

John Fritz Medal 123 . 74 

Cash Exchanges per contra 575 . 92 



. $86 887.85 
Cash in Banks and on hand, September 30, 1908 13 708 . 98 



$100 596.83 



DISBURSEMENTS 

Disbursements for general purposes $76 167 . 69 

Interest on Mortgage on land 3 240 . 00 

Cash Exchanges per Contra 575 . 92 

$79 983.61 
Cash in Banks and on hand, September 30, 1909 20 613 . 22 



$100 596.83 

The Committee also submits as called for by the By-Laws a detailed 
estimate of the probable income and expenditure of the Society for 
the Fiscal year ensuing. This estimate has been submitted to the 
careful consideration of each committee concerned and the Finance 
Committee has been assured in each instance that the appropria- 
tions asked for in the estimate include all needed expenditures to carry 
out the work of the different committees as now planned and author- 
ized. 

It will be noted that the Society is not being operated for profit, 
but that practically all of the money received is appropriated for the 
development of the Society's various interests, and to enable giving to 
eaf^h member a constantly increasing return for his membership dues. 



SOCIETY AKFAIK.S 417 

ESTIMATE, 1909-1910 
Current Income Current Expenses 

Dues, Current $52000 Finance Committee $26000 

Dues, Arrears 2000 Membership Committee 2400 

Reserve Fund, 10 % 2800 Increase Memb. Committee . . 500 

Sales gross receipts 5000 House Committee' 1150 

Interest 800 Library Committee 2880 

Advertising 21000 Meetings Committee 8050 

Publication Committee 34900 

$83600 Research Committee 500 

Executive Committee'"' 600 

Power Tests Committee 500 

Sales Expenditures 3000 



Excess of income over expense 3120 



$83600 

1 In addition J3000, to be appropriated from the Reserve Fund for the House Committee 
for betterments for 1909-1910. 

^The appropriation for the Executive Committee for the foreign meeting to be $3000, to be 
divided from Current Income at not less than $600 per year for a term of years until can- 
celled. 

Especial attention of the Council is called to the fact that in con- 
nection with entering upon our occupancy of the present refined and 
dignified headquarters, a large sum was advanced from the Society's 
working capital, known as the Reserve to the Land and Build- 
ing Fund from which fund by vote of the Council the interest on the 
mortgage for the land is paid. Admittedly the Society cannot afford 
to pay for the present headquarters out of its current income unless 
the Society is freed from debt ; and it was with the understanding that 
sooner or later this debt would be raised, that the Society was justi- 
fied and enabled to accept the gift from Mr. Carnegie. During the 
past year the total receipts to the Land and Building Fund have 
been practically used up for paying the interest on the mortgage, with- 
out decreasing the total of the mortgage to the extent of one dollar. 

The Finance Committee observes that it has been the custom, by 
ruling of the Council, to take 10 per cent of the Reserve Fund 
each year to be applied to the payment of current expenses; and^they 
recommend to the Council that this custom be discontinued, and -that 
the total payments into the Society of initiation fees, which go to 
make up the Reserve Fund, shall remain in the Reserve, and that only 
by special vote of the Council shall money be expended from this 
Reserve. 



418 SOCIETY AFFAIRS 

The Finance Committee trusts that the time is opportune for the 
Land and Building Fund Committee to take steps during the coming 
year to raise a portion if not all of the indebtedness amounting to 
about S90,000. 

It is highly desirable in view of plans for broadening the work 
of the Society that our income available for such extension of work be 
increased. The organization of our Society is such that the Finance 
Committee is charged solely with the responsibility of administering 
the Financial affairs of the Society as they find them and not to pro- 
duce revenue. All the remaining activities of the Society are for 
the expenditure of revenue. The Finance Committee suggests there- 
fore that it would be in keeping with good management if a special 
committee was appointed to consider the essential feature of the 
Society's broader life, viz: the income side, and to see that it is in- 
creased to provide for the reduction caused by the discontinuance of 
taking 10 per cent annually from the Reserve for operating expenses 
and to provide for a broader work in the future, 

Respectfully submitted 

Arthur M. Waitt, Chairman 

Edward F, Schnuck 

George J. Roberts \- Finance 

Robert M. Dixon [ Committee 

Waldo H. Marshall J 

Report of the House Committee 

The House Committee reports that it has endeavored to make 
the headquarters of the Society more attractive, by a rearrangement 
of the rooms and by additions to the furnishings. 

When the Society entered its new headquarters nearly three years 
ago, provisional furnishings were purchased sufficient to carry on the 
business of the Society but with no attempt at decorative features. 

The original plans of the rooms provided for a large reception hall 
which visitors enter from the elevators. In common with the other 
floors of the building this hall was open to the main stairway. 

A partition cutting off this stairway and another partition separat- 
ing the offices has converted this hall into an excellent reception 
room. 

Sliding doors have been arranged so that the Council Room, the 
Library and the Secretary's office give the effect of one large and spa- 
cious room. 



SOCIETT AFFAIRS ' 419 

The walls have been retinted, and new rugs cover the floors. Com- 
fortable furniture has been placed in the reception room. There will 
be portieres between the rooms, draperies at the windows, and more 
comfortable chairs and divans added to the library and Council cham- 
ber. 

The Committee has aimed to make the rooms homelike and com- 
fortable, to make a place which the members will use freely for their 
own convenience and for meeting other members or friends for social 
or business engagements. 

In addition to the large rooms referred to, a small room is especially 
reserved where members may attend to their correspondence or hold 
private conferences. 

Photographs of the Past-presidents have been placed on the walls 
of the Library and by order of the Council a similar portrait of each 
succeeding President will be added as he retires from office. Name- 
plates have been placed on the portraits, paintings and historical 
objects, and a very complete catalogue of all these objects of historical 
interest has been prepared. This catalogue represents the result of 
long and painstaking research on the part of Mr. Edward Van Winkle 
of our Committee. 

Respectfully submitted, 

Henry S. Loud, Chairman ' 

W. C. DiCKERMAN 

B. V. SwENSON )■ House Committee 

Francis Blossom 
Edward Van Winkle 

Report of the Library Committee 

During the past year further steps have been taken in the evolu- 
tionary process of administering the libraries of the American Insti- 
tute of Mining Engineers, the American Institute of Electrical Engi- 
neers and that of our own Society, as far as practicable, as a unit. 

This process has involved the development of a comprehensive plan 
whereby the libraiy of each society maintains only books on sub- 
jects in which its membership is particularly interested, treating 
all other publications in its library as duplicates. To the American 
Institute of Mining Engineers' have been assigned the subjects of 
mining engineering, geology, mineralogy, chemistry, metallurgy and a 
part of chemical technology. To'the American Institute of Electrical 



420 SOCIETY AFFAIRS 

Engineers the subjects of electrical engineering, electricity, physics, 
mathematics and pure science; and to this Society the' subjects of 
general engineering, railroad engineering, mechanical engineering, 
civil engineering and a part of chemical technology. This plan has 
given satisfaction as a temporary working basis enabling each organi- 
zation to complete or supplement imperfect sets from the collections 
of the others. 

During the year a union card catalogue has been estabhshed, cover- 
ing the libraries of the three Founder Societies, which enables readers 
to find at a glance all the literature on any subject that may be con- 
tained in any of the libraries. 

A Library Conference Committee, consisting of the Chairmen of 
the Library Committees of the three societies, has under considera- 
tion a further important step toward the unification of the three 
libraries, involving the organization of the library of the United Engi- 
neering Society, to which the three societies shall bear the same rela- 
tion as do the Founder Societies in the holding of the United Engi- 
neering Societies building and property. Such a plan will enable 
gifts of books or periodicals not specifically designated for one society 
to be received and taken care of and it may eventually result in the 
purchase of books jointly in which the three Societies would have a 
common ownership. This plan avoids purchases in tripUcate or 
duplicate and concentrates the purchasing power and extension of the 
library in a way that will be of undoubted advantage to all who may 
have occasion to consult a comprehensive library of engineering liter- 
ature, covering all branches of the profession and having available 
promptly after publication all the important books. 

It is probable that these improvements will necessitate the carry- 
ing out of the original building plans for the library, providing addi- 
tional shelving in the library room proper, so that all of the volumes 
may be readily accessible. 

The present status of the Library of The American Society of Mech- 
anical Engineers is as follows: 

The following titles have been catalogued to date: 

Durfee library 570 vol. 

A. S. M. E. library 7237 " 

Withdrawal of duplicates (not accessioned) 800 " 

Pamphlets 1339 " 

Total 9946 " 



SOCIETY AFFAIRS 421 

The additions provided for 1908-1909 and included in tlie above 
are as follows: 

By gift 168 vol . 

By purchase 95 " 

By binding of exchanges 197 '' 



Total accessions 460 

Respectfully submitted, 



J. W. LiEB, Jr., Chairman 
C. L. Clarke 
h. h. suplee 
Ambrose Swasey 
Leonard Waldo 



Library 
Committee 



Report of the Meetings Committee 

To facilitate the work of the present Committee, and it is hoped, of 
succeeding committees, a record has been made of its policies and 
decisions, some of the more important of which are given below: 

The policy of the Committee shall be: 

1 Further condensation of papers by the elimination of all superfluous and 
irrelevant matter, or matter'previously printed, and of such statements of fact as 
are of common knowledge in the profession. 

2 The solicitation and selection of such papers, together with the plan of 
their presentation at meetings, as may make the Transactions a historical and 
up-to-date record of the progress of all branches of mechanical engineering. 

3 The presentation of a subject, whenever possible, in such a way as best to 
permit of a general and thorough discussion ; and to this end to extend invitations 
to those, whether members or otherwise, whose experience has been such as to 
bring out the most valuable discussion of the subject. 

4 At the Annual and Semi-Annual Meetings, a reduction, when possible, of 
the number of professional sessions, and of the number of papers assigned thereto 
in order that more opportunity may be given for satisfactory discussion and for 
social intercourse between the members. It is the opinion of the Committee that 
the professional sessions heretofore have been too crowded. 

5 For the sake of uniformity, the adoption of a few rules for the guidance of 
authors, these to be based on the experience of the Committee and of the edi- 
torial department of the Society, and to offer a review of the rules governing 
similar organizations. 

6 The adoption of rules tending towards greater uniformity in the actions of 
the Commitee; these rules to be such only as concern actions within the juris- 
diction of the Committee and subject to such exceptions as in the opinion of the 
Committee may seem desirable. 



422 SOCIETY AFFAIRS 

During the past year, the Committee has submitted to the Council 
a number of suggestions relative to changes in some of the methods of 
conducting such affairs of the Society as relate to the Meetings Com- 
mittee. All of these, with slight modifications, have been accepted 
and endorsed by the Council and so far as possible placed in operation. 

The selection of a local committee to take charge of all entertain- 
ment, apart from the professional sessions, was tried at the last Annual 
Meeting with satisfactory results, which we believe long-established 
practice will make even better. This is creating greater interest 
among the local members, and a feeling of some responsibility for the 
entertainment of the visiting members, and places the Annual Meet- 
ing upon the same basis as the Spring Meeting, thereby eliminating 
what has been heretofore a somewhat inconsistent situation. The 
Social and Entertainment Committee will for the first time this year 
collect and disburse the fund for this purpose, which will be kept sep- 
arate and apart from the^ funds of the Society. This phase of the 
arrangement cannot be otherwise than satisfactory. 

The resolution of the Committee submitted to the Council, rela- 
tive to meetings in mid-season in cities other than New York, was 
put into operation immediately upon approval by the Council. In 
the opinion of the Committee, this movement is progressing very 
satisfactorily and seems to be assuming a natural, healthy growth. 
Successful meetings were held at Boston, April 16, June 11, October. 
20, and November 17; and at St. Louis, April 10, May 15, October 
16, and November 13. This movement, as was desired and antici- 
pated, is bringing before the Society much valuable material in the 
form of papers and especially of discussion that would otherwise be 
inaccessible to tht members. It has resulted in an exchange of 
papers, which promises to become more extensive in the future. 

The Council's amendment to the Committee's resolution, "subject 
to the approval of the Council, " we find from experience to be cum- 
bersome. To facilitate these meetings, the Committee must act 
promptly upon request from members residing in places other than 
New York. With the appropriations for these meetings decided upon 
the Committee urges that the Council modify its instructions to the 
effect that the Committee may have full authority in compliance with 
the original resolution submitted by the Committee to the Council. 

The Committee's interpretation that B 21 did not include th§ 
vouchering of bills covering the expenditures of the appropriations 
for its work, has been confirmed by the minutes of the Council of a 
few years ago, when the details of such expenditures were placed in 



SOCIETY AFFAIRS 423 

the Secretary's hands as business manager. The rules governing 
office procedure have, however, been changed to define more clearly 
this interpretation, resulting in some simplification of the work of 
the accounting department. 

Last spring a number of members of the Society requested a meet- 
ing or conference on the subject of Smoke Abatement. This peti- 
tion and the action of the Committee were referred to the Council on 
May 28, 1909. This request was for a National Conference with the 
elimination of the engineering features as far as possible. After 
due consideration the Committee declined to take favorable action. 

Subsequent to the above, the Committee received a second peti- 
tion asking for a National Conference, but along strictly engineering 
lines. In the absence of precedent relative to such a Conference, the 
Committee referred the question to the Council. The Committee 
has not received, but would gladly receive and carefully consider, a 
paper on the subject of Smoke Abatement, if presented along strictly 
engineering lines. 

We believe the best interests of the Society make necessary a close 
working arrangement between the Research and Meetings Committees. 

A plan was inaugurated early in the year which it is thought will 
bring before the Society more new material than has been heretofore 
available. This is accomplished by correspondence with those inter- 
ested in original research. 

The usual number of meetings were held by the Society during the 
past year, all of which are now on record. The Committee begs to 
express its appreciation for the assistance and cooperation during the 
year of the officers and the several departments of the Society. 

Willis E. Hall, Chairman 

William H. Bryan 

L. R. PoMEROY \ Meetings 

Charles E. Lucke j Committee 

H. deB. Parsons J 



Report of the Membership Committee 

During the current year the Membership Committee has held seven 
meetings, at which a total of 361 applications for membership have 
been considered with the following results: 



424 SOCIETY AFFAIRS 

Applications void and withdrawn 11 

Applications deferred 11 

Recommended for membership 339 

There were two ballots during the year on which the applicants 
recommended by the Committee were voted for. These were at the 

Washington meeting 148 

New York meeting 187 

Total 335 

In addition to the most careful consideration which the Secretary 
and the Membership Committee can give to the applications for mem- 
bership, the cooperation of the whole voting membership is needed 
in order to maintain the high standing of the Society. In several 
instances during the year action by certain members in giving infor- 
mation to the Committee has caused reconsideration of apphcations, 
with the result that they have been indefinitely deferred. 

A member should not agree to act as proposer or seconder for an 
applicant unless he actually knows from his own personal observa- 
tion enough of the latter and his work to be able to answer favorably 
all the questions on the reference blank regarding him. 

The Committee has endeavored to maintain under the By-Laws 
the standard of qualifications of applicants for whom they have rec- 
ommended to be voted. 

The work of the Committee has been greatly facilitated and expe- 
dited by the complete and admirable way in which the cases have 
been arranged by the Secretary and his staff for presentation to them. 

Respectfully submitted, 

Henry D. Hibbard, Chairman 

Charles R. Richards 

Francis H. Stillman >■ Membership 

George T. Foran Committee 

HosEA Webster 

Report of the Publication Committee 

The Publication Committee submits herewith the annual report of 
its work and of the activities under its control for the past year. 

The Committee has held frequent meetings and has earnestly en- 
deavored not only to maintain the high standard for the publications 
of the Society which has previously been set, but also wherever pos- 



SOCIETY AFFAIRS 425 

sible to raise the standard to a new level. In its work upon Volume 
30 of the Transactions which contains the record of the spring 
and winter meetings of 1908, the Committee has given careful study 
to the available papers with a view of selecting for that volume only 
those of greatest value for permanent record. After due consider- 
ation several papers have been omitted and others have been edited 
or revised with the approval of the authors. Discussions also have 
been edited and in some cases considerably condensed in order to 
separate material of permanent value from that which had but a tem- 
porary or passing interest. 

In compliance with the Resolutions passed by the Council in April 
1909, the Publication Committee has undertaken the general super- 
vision of The Journal in addition to its other duties, and has adopted 
the following general plan for the conduct of this work: 

As a general policy, The Journal should be regarded as the news- 
paper of the Society and reports of committees, reports of meetings, 
professional papers of the Society as a whole or of sections, book 
reviews. Society items, etc., should be published as requested by com- 
mittees in their official capacity when approved by this Committee, 
without charging to the committees or activities concerned any 
expense for publication. The Journal has its own expense account 
and the appropriation for The Journal should be sufficient to cover 
editing and publication of this material. 

•No papers, whether for the meetings of the Society as a whole, or 
for sections, technical, student or geographical, are to be published 
except as formally authorized b}'' the Meetings Committee. 

Material from standing committees offered officially will, in general, 
be published in the form which these committees desire. 

Reports of meetings of the Society and of sections, except when con- 
taining strictly professional papers and discussions will, in general, 
be published in condensed form. 

All matter presented at meetings other than the professional papers 
provided by the Meetings Committee, including all discussions, will 
be edited under the direction of the Publication Committee. As a 
general policy, discussion will be condensed, commercial matter 
removed, with a view to presenting only engineering data, opinions 
based on experience, historical notes and similar material of value for 
permanent record in Transactions. 

The advertising section of The Journal which began with the 
number of^September 1908, has proven successful. The income from 
this source has increased steadily until at the present time there is a 



426 . SOCIETY AFFAIRS 

gross annual income from it of $21,000; and through the action of the 
Council this increased income may be applied to the improving of 
the quality of, and to the development of The Journal. Plans for 
such development are under consideration, and it is the purpose of 
the Committee to make improvements as rapidly as conditions may 
warrant. 

But the most effective work upon The Journal and that which will 
be of greatest benefit to our membership at large is the careful pre- 
paration for publication of the professional material presented at the 
regular meetings of the Society, and at the meetings of the different 
sections. In this great fund of material there is always some that is 
unimportant and irrelevant, and much more that could be made of 
greater value by skilful editing or by condensation. During the 
past year the Committee has done much in this direction that has 
resulted in the improved quality of our paper, and also in a consider- 
able economy of money, and the papers now appearing in The Journal 
are suitable, with little or no alteration, for publication in the Trans- 
actions. 

In addition to the volume of the Transactions and The Journal the 
Committee has issued the annual Year Book of the Society and the 
Pocket List of Members. 

Respectfully submitted, 
A. L. WiLLisTON, Chairman 
D. S. Jacobus 

H. F. J. Porter \- Publication 

H. W. Spangler Committee 

g. i. rockwood 

Report of the Research Committee 

The Research Committee was formally notified of their appoint- 
ment under date of April 7, 1909, and at the suggestion of the Presi- 
dent, the members were requested to meet during the Spring meeting 
of the Society at Washington. Notice was given a short time in 
advance of the meeting, and only Prof. R. C. Carpenter and R. H. 
Rice were present. These members, however, together with the 
President of the Society, Jesse M. Smith, and Charles W, Hunt, 
Past-President, and originator of the suggestion that a Research 
Committee be appointed, engaged in an informal conference. 
S'^A second meeting was called for Wednesday, June 23, 1909, to be 
held in New York. There were in attendance the President, Jesse 



SOCIETY AFFAIRS 427 

M. Smith, R. H. Rice, James Christie, W. F. M. Goss, and the Secre- 
tary, Calvin W. Rice. Dr. Coss was chosen Chairman. The Secre- 
tary of the Society was recognized as the secretary of the Committee. 
The minutes of the informal meeting held in May were read for the 
information of the members. After a considerable discussion as to the 
scope of the work of the Committee, it was agreed that the Committee 
should have information concerning the laboratories of the various 
colleges, and other public institutions in America, in which work of 
engineering research is proceeding, and to this end tli3 Secretary 
was directed to develop a process which would result h: t ! le establish- 
ment of such a record in the office of the Society. 

It was agreed that the Committee should consider the question of 
safety valve efficiency. Arrangements were made for gathering in 
existing information upon this general subject, and steps were taken 
which will, it is believed, result in a satisfactory outline from which 
actual work may proceed. Several other subjects for research, re- 
ferred to the Committee by the Council, were laid on the table for 
future consideration. 

Respectfully submitted 

W. F. M. Goss, Chairman 

James Christie 

R. C. Carpenter \- Research 

Richard H. Rice j Committee 

Charles B. Dudley J 



No. 1250 

THE PROFESSION OF ENGINEERING 

PRESIDENTIAL ADDRESS 1909 

By Jesse M. Smith, New York 

President of the Society 

Great engineering works existed in raanj^ parts of the world long 
before Columbus discovered America. We have but to consider the 
ruins left by the Incas in South America and the Aztecs in Mexico to 
realize the great work done on this continent in engineering. In 
Asia the great wall of China, the temples of Japan, China, Babylonia 
and Assyria bear record of the presence of the engineer. 

2 In Africa, the vast pyramids of Egypt and the temples on the 
Nile are evidences that great engineers existed long before the Chris- 
tian era. We marvel still when contemplating the pile of immense 
blocks of stone forming the pyramids and try to imagine what form 
of apparatus could have been used in placing those great stones one 
upon the other. 

3 In Europe the Greeks and Romans did marvelous work in roads, 
bridges, aqueducts, and various mechanical structures which the 
modern engineer may well ponder upon and admire. While we read 
much in history of the emperors and kings who reigned when these 
great engineering works were produced, we learn little of the men 
who produced them, men whom we now call engineers. 

4 While engineers have existed for thousands of years it is only 
within a comparatively recent time that they have begun to form 
themselves into societies for their mutual education and the advance- 
ment of the profession of engineering. 

5 In England, as early as 1771, Smeaton and his contemporaries 
came together to form the Smeatonian Society of Engineers, which, 
therefore, according to the calculations of a noted English engineer, 
is five years older than the United States. The Institution of Civil 
Engineers of Great Britain came into existence in 1818, and was 



An address delivered at the Annual Meeting, New York, (December 1909) 
of The American Societt op Mechanical Engineers.. . 



430 THE PROFESSION OF ENGINEERING 

followed by its sister society, the Institution of Mechanical Engineers, 
in 1847. La Soci^te des Ingdnieurs Civils de France was founded in 
1§48. Der Verein Deutscher Ingenieure was organized in 1856. 

6 In this country the Boston Society of Civil Engineers began 
its work in 1848. Our elder sister among national societies, the Ameri- 
can Society of Civil Engineers, was organized in 1852. The next 
member of the family, the American Institute of Mining Engineers, 
was born in 1871. Our own Society came into existence in 18S0, 
and our younger and very vigorous sister, the American Institute 
of Electrical Engineers, came along in 1884. 

7 Each of these four national societies, the American Society of 
Civil Engineers, the American Institute of Mining Engineers, The 
American Society of Mechanical Engineers and the American 
Institute of Electrical Engineers, has grown greatly since its organiza- 
tion, and each continues to thrive. During the process of upbuilding 
of these four great national societies, several other national societies 
of specialists in engineering and many local societies of engineers 
have been formed, and all of these are also active and thriving. 

8 The four greater national societies have an aggregate member- 
ship at this time of over 19,000 members. Twelve national socie- 
ties of engineering specialists contain more than 13,000 members. 
Twenty-three local engineering societies in different cities of the 
United States count over 8,600 in their membership. 

9 What does this great army of over 40,000 engineers, organized 
into many different societies, all for purely professional purposes, 
mean? It means that the engineering profession is making Itself 
felt in this country of ours, that it proposes to take a prominent place 
in the great activities by which the country is being developed, that 
it will take its place in public affairs, that it is coing into its own. 

10 The national societies are not antagonistic to each other ;^ on 
the contrary, they support and give confidence to each other. The 
national societies' of specialists are not at war with the other 
national societies; they supplement them. 

11 The local societies are not in opposition to the national socie- 
ties; they extend their influence; they are the outposts of the great 
Mrmy. The specialists do not interfere with each other. We are all 
specialists to a greater or less extent; but we are all engineers. 

12 In the legal profession, some men practice in the criminal 
courts; others devote themselves to titles in real estate; others are in 
corporation law; others hi patent causes; they all sciuabble with each 
other in their practice; but when they meet in their bar association 
they arc aU lawyers; they stand by each other and their profession; 
they are a power in the world. 



THE PROFESSION OF ENGINEERING 431 

13 The medical profession is made up of surgeons, oculists, 
aurists, general practitioners, specialists of the skin, the heart, 
the lungs and every other part of the human anatomy; but when they 
come together in their general medical associations they are all 
doctors; they also stand by each other and their profession; they 
also are a power in the world. 

14 In the engineering profession why may not the men who 
practice in steam engineering, in machine construction, in hydrauhcs, 
in railroad, bridge, mining, electrical and chemical engineering, in 
metallurgy, refrigeration, heating! and every other specialty in engi- 
neering, come together, stand by each other and their profession, 
become known as engineers and be a power in the world? 

15 When, in 1SS9, the Institution of Civil Engineers of Great 
Britain invited the four national American societies of civil, mining 
mechanical and electrical engineers to visit it in London, there was 
inaugurated a spirit of friendship and cooperation in the engineering 
profession which has grown stronger and stronger as the years have 
passed. Following the visit in London, La Soci6t6 des Ing^nieurs 
Civils de France, in the same year, invited the American societies to 
Paris. 

16 Those who were fortunate enough to participate in those 
memorable demonstrations of hospitality cannot fail to realize how 
greatly the seed of cooperation sown in that year has fructified. 

17 In 1900 this Society was again invited by the Institution of 
Civil Engineers and the Institution of Mechanical Engineers to visit 
them in England, and again invited by the French society to visit 
it in Paris. Thus the spirit of cooperation was still further advanced 
by these remarkable meetings. On both occasions the sister socie- 
ties abroad were untiring in the entertainment of the American engi- 
neers. 

18 The year 1904 was made memorable^by the acceptance of an 
invitation extended by this Society to the Institution of Mechanical 
Engineers of Great Britain to hold a joint meeting in Chicago. Thus 
the spirit of cooperation and good friendship was again strengthened 
and extended. 

19 Now the Institution of Mechanical Engineers of Great Britain 
has expressed the desire still further to promote this friendly spirit 
by inviting this Society to a joint meeting in England in July 1910. 
The Council of our Society has accepted this very cordial invitation 
of the Institution in the spirit of good will in which it was extended. 
It remains for the membership of The American Society of Mechani- 



432 THE PROFESSION OF ENGINEEEING 

cal Engineers to respond to this spirit and to go to England next 
year with its best talent and its best men. 

20 The helpful cooperation in professional work which has already 
been established with our sister societies over the seas is also be- 
coming manifest in our own country. The four national societies 
of civil, mining, mechanical and electrical engineers on March 24, 
1909, held in this auditorium a joint meeting on the Conservation 
of the National Resources, which did much to bring engineers close 
together and into cooperative relation. 

21 Our Society invited the Boston Society of Civil Engineers to join 
in the monthly meetings of the Society recently held in Boston. 
The Engineers' Club of St. Louis in like manner was asked to join 
with us in the Society's monthly meetings recently held in St. Louis. 
In both cases the invitations have been accepted in the best spirit of 
cooperation. 

22 The engineering societies of the country may be likened to 
the members of a large and harmonious family, each member inde- 
pendent to do its own special work in its own way, each member 
ready to help each of the others, each residing in its own home, but 
all ever ready to stand by each other, to work for the common good, 
to advance and dignify the profession of engineering. 

23 A striking example of the " getting together " of the engineer- 
ing societies is found in this building which is the home of our Society. 
It is also the home of our sister societies, the American Institute of 
Mining Engineers and the American Institute of Electrical Engi- 
neers. 

24 Under the same roof are grouped together fifteen other socie- 
ties of engineering and allied arts. Twenty-five thousand engineers 
practicing in all the specialties of engineering may call this building 
their professional home. We are hving together here in peace and 
harmony. We have brought our books together into a single library 
open to the profession and to the public, where every one is welcome. 

25 Our meetings are held in the same auditorium and lecture 
halls; the doors stand open that all who wish may enter. Our profes- 
sional brethren of every society of every country are welcome here. 
The large hall at the entrance to the building is a foyer where all 
engineers may come together on the same plane, where they may 
unite to strengthen each other and to sustain and advance the profes- 
sion of which they form a part. 

26 The spirit of cooperation which.' now exists must be fostered, 
strengthened, made enduring, to the end that as great solidarity will 



THE PROFESSION OF ENGINEERING 433 

exist in the engineering profession as exists in any of the other great 
learned professions. 

27 Numbers in membership are, of course, important m the 
societies which represent the engineering profession, but a high 
standard of membership is of much greater importance. 

28 With a considerable number of high-grade technical schools 
throughout the country all striving with each other to raise the stand- 
ards of engineering education ever higher and higher ; and with the 
graduates from these institutions taking, from year to year, a larger 
and more responsible part in the great activities of the country, there 
is no lack of material from which to form a membership in the en- 
gineering societies which will be worthy of the profession. 

29 In the Institution of Civil Engineers, as well as in the Institu- 
tion of Mechanical Engineers of Great Britain, we are informed, no 
person is admitted into the lower grade of membership unless he can 
pass a satisfactory examination as to the fundamental principles of 
engineering, conducted by an examining board of the Institution. 
The rules laid down by this examining board form the standard by 
which ihe applicants to membership are measured. If the technical 
schools in Great Britain maintain an equally high standard in grant- 
ing their degrees in engineering, then the degree may be accepted 
in lieu of an examination. In other words, the engineering institu- 
tions in Great Britain establish the standard for the degrees granted 
by the technical schools. A promotion from a lower to a higher 
grade of membership is made only upon a showing of sufficient 
experience in engineering to satisfy the rules laid down by the Insti- 
tution. 

30 In The American Society of Mechanical Engineers, a person 
may enter the Society as a Junior upon the presentation of a degree 
in engineering from a technical school. But this Society has not, 
up to the present, established a standard by which to measure that 
degree. I believe the standard for such a degree in engineering 
should be established by the Society, and that it should be as high 
as that of the best schools of engineering in this country. It will 
follow that the schools having a lower standard will soon be brought 
up to the higher standard. 

31 Promotion to higher grades of membership in our Society is 
only made upon a showing of engineering experience satisfactory to 
our Membership Committee. This committee is maintaining a high 
standard of membership, and I beb'eve that acting under the influence 
of the membership and the Council of the Society, it will not allow 
that standard to fall, but rather cause it to rise. 



434 THE PROFESSION OF ENGINEERING 

32 If we are to have a profession of engineering, as distinguished 
from the trade of engineer, wc must have a broad education befitting 
men of a learned profession, as distinguished from a narrower educa- 
tion sufficient for men of a trade. 

33 President Lowell of Harvard in his recent remarkable in- 
augural address, gave this as his conclusion: "The best type of 
liberal education in our complex modern world aims at producing 
men who know a little of everything and something well." If that 
conclusion be true of a liberal education leading to the learned pro- 
fession of the law or medicine or theology, why is it not also true of a 
scientific education leading to the learned profession of engineering? 

34 If preponderance be given to one part of President Lowell's 
conclusion over the other part, certainly knowing "a little of every- 
thing" leads to superficiality; while just as surely knowing but one 
thing well leads to narrowness. There would seem to be a happy 
mean between these two extremes in the education of the engineer. 

35 The engineer capable of being at the head of the larger engineer- 
ing works must know something of many things, several things well 
and one thing profoundly. 

36 The engineer, president of a great railway system, for example, 
must know something of the alignment and gradients of the perma- 
nent way, its construction and maintenance ; something of the proper 
location of sidings and stations; something of the system of signals. 
of the various kinds of cars, of the quality of water for the locomo- 
tives, of the heating and lighting of cars, and many other things. 
He must know well that the bridges have been designed for safety 
and endurance and that they have been properly constructed. He 
must know well that the tunnels are safely protected against 
external pressure and falling rocks. He must know well that the 
locomotives for drawing the high-speed trains, as well as those for 
the heavy freight trains, are of the very best design and capable of 
performing their duty with efficiency, economy and endurance. He 
must know well how to manage the traffic and keep the accounts. 
He must know profoundly how to coordinate all the different parts 
of this complex organization so that each part will perform its 
proper and full function, to the end that passengers and freight will 
be carried safely, surely, quickly and cheaply, and also that dividends 
will be paid to the shareholders. 

37 The engineer knowing something of many things, several 
things well and one thing profoundly, is still one-sided if all this 
knowledge is confined strictly to his profession. He will be a much 



THE PROFESSION OF ENGINEERING 435 

broader man and a better engineer, if in his leisure hours he can turn 
his thoughts entirely away from his professional work and toward those 
things in nature and art which give that rest and renewal of the pro- 
fessional mind necessary to continued work. 

38 Engineers have known for many years that tha profession 
of engineering is a learned profe'^sion ; the rest of the world is rapidly 
arriving at the same conclusion. 

39 When in April 1907, this building was dedicated "To the 
advancement of Engineering Arts and Sciences," President Hadley 
of Yale, where the learned professions have been taught for nearly 
200 yea'-s, said: 

The men who did more than anything else to make the nineteenth century 
different from the other centuries that went before it, were its engineers. 

Down to the close of the eighteenth century the thinking of the country was 
dominated by its theologians, its jurists, and its physicians. 

These were by tradition the learned professions, the callings in which pro- 
found thought was needed, the occupations where successful men were venerated 
for their brains. 

It was reserved for the nineteenth century to recognize the dominance of 
abstract thought in a new field — the field of constructive effort — and to revere 
the trained scientific expert for what he had done in these lines. 

Engineering, which a hundred years ago was but a subordinate branch of the 
military art, has become, in the years which have since elapsed, a dominant factor 
in the intelligent practice of every art where power is to be applied with economy 
and intelligence. 

It is encouraging to engineers to have their profession recognized as 
a " learned profession " by so great an authority as the president of 
Yale University. 

40 Enthusiasm and devotion to his profession are characteristic 
of the engineer, and from my observation these begin with the 
student in engineering and extend throughout his life. President 
Wilson of Princeton, in an address at Harvard not long since, dwelt 
upon " the chasm that has opened between college studies and college 
life. The instructors believe that the object of the college is study, 
many students fancy that it is mainly enjoyment, and the confusion 
of aims breeds irretrievable waste of opportunity." These conditions, 
I believe, exist to a much smaller extent in the technical schools 
where engineers are taught, than in the general colleges, where a 
liberal education is obtained. 

41 Enthusiastic love of work, for his profession's sake, resides in 
the heart of the engineer who becomes great. The man who merely 
works for wages, and without enthusiasm, does not rise; he remains 
a paid servant, and poorly paid at that. 



436 THE PROFESSION OF ENGINEERING 

42 Where enthusiasm exists, love of work exists; success follows. 
Our individual enthusiasm is quickened by the study of the work of 
our brother engineers. 

43 What engineer while being whisked through the tunnels which 
connect Manhattan Island with the lands surrounding it, can fail to 
rejoice in his profession as he contemplates the work of the civil engi- 
neers, the mining engineers, the mechanical engineers, the electrical 
engineers, which, joined together, supplemented each other to 
produce success in those marvelous undertakings? The highest 
knowledge and skill in each of the four branches of the engineering 
profession were called for. and were forthcoming, in the consumma- 
tion of this great work. It is not a question which engineers 
did the most toward the success of this problem in transportation; 
they all did their best; they all did well; each contributed a necessary 
part to the success ; they were all engineers working for the advance- 
ment of the profession of engineering. 

44 Will not every true engineer feel his enthusiasm in his pro- 
fession quicken, as he watches the great vessels of trade and the 
great vessels of war sweep out to sea, and stops to consider how- 
much brains, and long experience, and hard work of many men are 
concentrated in each one of them? 

45 We marvel still, our enthusiasm is inspired, as we see ponder- 
ous steam locomotives and mysterious electric locomotives compet- 
ing in the hauling of trains, ever heavier and heavier, ever faster and 
faster, and both succeeding. 

46 The automobile in its present highly developed and thoroughly 
practical form is the result of enthusiastic work of many engineers, 
principally within the last fifteen years. 

47 The enthusiasm of the engineer is never satisfied. Having 
conquered the highway with the automobile driven by the internal- 
combustion gas engine, he now proposes to conquer the air with the 
aeroplane driven by the same kind of an engine in improved form. 

48 The American Society of Mechanical Engineers has before it a 
future of usefulness to its members and influence in the profession, 
which is unUmited. It only requires that we stand by our tradition 
of increasing the membership with men of high quality as engineers; 
that the members maintain enthusiastic devotion to good professional 
work; that they cooperate with each other in the broadest and most 
friendly spirit to produce that solidarity of membership and devo- 
tion to high ideals, which will compel the world to class the profession 
of engineering with the other learned professions. 



No. 1251 

THE HIGH-PRESSURE FIRE-SERVICE PUMPS OF 
MANHATTAN BOROUGH, CITY OF NEW YORK 

DESCRIPTION OF PUMPS AND PUMPING SYSTEM WITH RESULTS 

OF TESTS 

By Prof. R. C. Carpenter, Ithaca, N. Y. 
Member of the Society 

The object of this paper is to present a concise description of the 
high-pressure pumping system installed for fire service in the city of 
New York and the results of a test of the pumping machinery. 

2 The system protects the district extending north from City Hall 
to Twenty-fifth Street, and east, approximately, from the North 
River to Second Avenue. It comprises about 55 miles of extra heavy 
cast-iron main, from 12-in. to 24-in. in diameter, with 8-in. hydrant 
branches; and two pumping stations so located that they never can 
be in the center of a conflagration. At the present time the pumping 
stations have a combined capacity of over 30,000 gal. per min. 
delivered at a pressure exceeding 300 lb. per sq. in. 

THE SOURCE OF WATER SUPPLY 

3 The supply of water is ordinarily obtained from the water 
mains of the city, which deliver Croton water to the stations at a 
pressure of from 14 lb. to 40 lb. per sq. in., depending upon the 
demand for water in that district. Both of the pumping stations are 
located close to tidal water and connections are made so that sea 
water can be obtained in case of difficulty with the Croton supply. 

4 The advantage of the Croton water over salt water is that it is 
less likely to injure goods, and as the amount required for fire purposes 
is only a small percentage of that consumed for the daily supply of the 
city its use for fire protection makes no material difference from 
financial or insurance standpoints. As this is a matter of consider- 
able importance data upon the quantity needed are given in the next 
paragraph. 

Presented at monthly meetings, New York and St. Louis (October 1909), 
of The American Societt of Mechanical Engineers. 



438 HIGH-PRESSURE FIRE-SERVICE PUMPS 

WATER REQUIRED FOR FIRE PURPOSES 

5 The general impression that an enormous quantity of water is 
required for fire purposes is erroneous as shown by figures furnished 
to Chief Engineer I. M. de Varona by the fire department for the 
Boroughs of Manhattan and Brooklyn, years 1900, 1901, 1902, 1903 and 
1904, These give the average quantity of water used for fire protec- 
tion during these years in the Borough of Manhattan as 74,010,803 gal. 
per year, of which 31,056,928 gal. was river water. The daily aver- 
age use of Croton water, therefore, for the above five years was 117,- 
000 gal. 

6 For the Borough of Brooklyn the average for five years was 
43,705,568 gal. of which 19,010,928 gal. was river water; daily aver- 
age, 67,000 gal. 

7 During these five years the greatest quantity used in the 
Borough of Manhattan was 99,000,000 gal. in 1901, which included 
69,500,000 gal. of river water, leaving 29,500,000 gal. for Croton water, 
and Mr. de Varona states (Report of the Department of Water 
Supply, Gas and Electricity): "Even if this quantity be made 100,- 
000,000 gal. per year, by comparing it with the average daily con- 
sumption of about 300,000,000 gal. it will be seen that the total 
amount used for fire purposes would be only about one-third of the 
amount used for all purposes in 24 hr., forming, therefore, an insignifi- 
cant percentage of the total consumption. The quantity needed for 
fire purposes (one-tenth of one per cent) may therefore be entirely 
neglected as a factor in determining the water supply of the city. 

8 "The capacity of each of the pumping stations will be for the 
present 15,000 gal. per min. or 43,000,000 gal. per day for the two sta- 
tions. By the installation of three additional units in each station, 
for which provision is made, this capacity can be increased in round 
numbers to 69,000,000 gal. per day. 

9 "The two stations, with the motors and pumps as installed, 
have a total capacity in excess of that of all the fire engines in the 
Boroughs of Manhattan, the Bronx and Brooklyn working under 
normal conditions. This comparison assumes the engines to work 
on one line of 2^-in. hose, say 500 ft. long, under a pressure of, say 
200 lb., and with the capacities as printed in the official blank forms 
of the reports of the fire department. It should furthermore be 
remembered that provision is made for the installation of still another 
pumping station." 



HIGH-PRESSURE FIRE-SERVICE PUMPS 439 

MOTIVE POWER 

10 The power for driving the pumps is transmitted electrically 
from several of the electric power and lighting systems located on 
Manhattan Island. As the stations of systems are widely separated 
and any or all of them are available for motive power the system of 
electric transmission was considered more reliable in the case of a 
large and general conflagration than power plants maintained directly 
at the pumping stations. Each station is provided with two inde- 
pendent sets of transmission lines located as far as possible beyond 
danger or injury in case of a great conflagration. 

11 The cost of erecting and maintaining an independent power 
plant would have entailed a greater annual charge than the cost of 
the electric current; consequently the present arrangement is advan- 
tageous from a financial standpoint. 

12 In addition to the charge per kilowatt for the current delivered 
there is a charge aggregating $90,000 per year for reserving the first 
right of use for the necessary generating machinery for this purpose. 
The total cost of maintenance of the system is estimated at $170,000 
a year, which amount it is believed will be saved many times over by 
a reduction in insurance premiums now paid in the protected district. 

13 The electric current is supplied at a pressure of 6600 volts 
from the following stations of the New York Edison Company, hav- 
ing the capacity indicated: 53 Duane Street, 7600 kw.; 115 East 12th 
Street, 1700 kw.; 45 West 26th Street, 400 kw.; 140th Street and 
Ryder Ave., 4000 kw.; Waterside Stations No. 1 and No. 2, 196,700 
kw. In addition there are feeders extending to the Brooklyn Edison 
Company stations which can be called on in case of an emergency 
demand. 

14 The pumping stations are connected to 18 sub-stations, 
equipped with rotary converters and storage batteries, aggregating 
a capacity of 124,000 ampere hours at 135 volts, ah enormous reserve. 

15 Each station is connected with the main stations of the Edison 
Company by two 250,000 cm. three-phase cables laid in ducts, and 
two independent reserve feeders extend to the sub-station system 
of the Edison Company. With all these precautions, interruption 
of the power supply would seem a physical impossibility. 

THE DISTRIBUTION SYSTEM 

16 The following information upon the distribution system is taken 
largely from the department report of Chief Engineer de Varona 



440 



HIGH-PRESSURE FIRE-SERVICE PUMPS 




Hydra n-f yv/fh connection 
far ^f reef f /cashing hi/d9. 



Fig. 1 Showing Location of Stations and Areas Covered by High- 
Pressure Pumping System 
the area indicated is served by a system op mains ranging from 24 in. to 12 in. in 

DIAMETER "WITH 8-IN. HYDRANT CONNECTIONS 



HIGH-PRESSURE FIRE-SERVICE PUMPS 441 

for 1905. Fig. 1 shows the system to be bounded by mains laid on 
the north through Twenty-third Street; on the east, through Broad- 
way to Fourteenth Street, through Fourteenth Street to Third 
Avenue, down Third Avenue to the Bowery, down the Bowery to 
Chambers Street; through Chambers Street on the south to West 
Street; and on the west through West Street. 

17 The area actually protected is considerably greater than this 
as hose can be extended over a zone 600 ft. wide beyond the limits 
of the mains. 

18 This district was selected as that in which the fire losses were 
the greatest and which most urgently needed fire protection. Plans 
have been prepared for the extension of the system southerly to the 
Battery, easterly as far as the East "River, and, if necessary, northerly 
as far as Fifty-ninth Street, by the simple extension of the mains 
and probably the erection of a third pumping station. 

19 The pipes, castings and hydrants were tested at a pressure of 
450 lb. The specified allowance for leakage in a 10-min. test was 
at the rate of 4 gal. in 24 hr. for each lineal foot of pipe joint, equiva- 
lent to a leakage of 487,000 gal. for the whole system in 24 hr,, which is 
somewhat over one per cen I; of the total specified pumping capacity now 
installed. The actual leakage on test was at the rate of 264,000 gal. 
per day or about six-tenths of one per cent of the pumping capacitv. 
Considering the difficulties of construction and the high pressure, the 
results attained were remarkable and reflect great credit on the 
engineer in charge. 

20 There are sufficient hydrants so that if a block were on fire 60 
streams of 500 gal. per min. each, or the full capacity of both stations, 
could be concentrated on a block with a length of hose not exceeding 
400 ft. to 500 ft., assuming the use of 3-in. hose and l^-in, nozzles. 

21 The layout of the mains at the stations both for suction and 
delivery is on the loop system; that is, the supply can be taken from 
either one of two mains, and the discharge is through either one or 
both of two mains. With this system even the breakdown of one 
of the discharge mains at the station would only slightly reduce the 
pressure at the fire and would not affect the capacity of the station, 
as the pumps would be capable of forcing their full discharge through 
the short length of a single 24-in. main if made necessary by such an 
accident. 

22 The mains are of cast-iron, bell and spigot pipe, of the thick- 
nesses given in the following table: 



442 HIGH-PRESSURE FIRE-SERVICE PUMPS 

Unit Tensile Strain 



Size of Pipe 


Thickness 


with 300 lb. pres- 




Inches 


Inches 


sure 


Factor of Safety 


24 


n 


1920 


0.4 


20 


a 


2000 


10.0 


16 


li 


1920 


10.4 


12 


1 


1800 


11.1 


8* 


i 


1371 


14.6 



* Used only for hydrant branches. 

SUPPLY PIPING 

23 At the South Street Station the fresh water supply is derived 
from t^^o 30-in. lines, one connected at Chestnut Street to the 36-in, 
line on Madison Street, and the other connected at Pike Street to 
the 36-in. line on Division Street. These two main feeders, to which 
the two 30-in. lines are connected, increase to 48 in. in diameter and 
extend independently and directly to the Central Park Reservoir and 
are also reinforced by connections with the main feeders in this sec- 
tion of the city. 

24 An auxiliary salt-water supply, consisting of two 36-in. pipes 
about 140 ft. long, brings the salt water from the East River to a 
suction chamber located directly in front of the pumping station. 
This suction is so constructed that the pipes are always below mean 
low water, thus insuring a supply at all times and avoiding the possi- 
bilit}^ of a break in the suction caused by air getting into the suction 
lines. On the river end of this suction there are constructed heavy 
bulkhead screens and in the suction chamber are two sets of bronze 
screens which are readily accessible for cleaning. From the suction 
chamber there are taken two 30-in. flanged mains to the duplicate 
set of mains in the pumping station proper. The vacuum in these 
30-in. pipes is maintained by automatic electric vacuum pumps 
located on the pump room floor of the station. 

25 At the Gansevoort Street Station the fresh-water supply is 
derived from two 30-in. mains, one connected at Twelfth Street to 
the 4S-in. line on Fifth Avenue, which runs direct to Central Park 
Reservoir, and the other connected to the 36-in. line on Ninth Ave- 
nue at Little West 12th Street, which increases to a48-in. line and runs 
also direct to the Central Park Reservoir. These two main feeders, 
in addition to having their supplies direct from Central Park Reser- 
voir, are also reinforced by connections with the main feeders in this 
section of the city. 



HIGH-PRESSURE FIRE-SERVICE PUMPS 443 

26 The salt-water suction lines for this station are practically 
identical with those for the South Street Station except that the 
36-in. lines from the North River to the station are 650 ft. long. 

PUMPING STATIONS 

27 The two stations, known as the Gansevoort pumping station, 
located near Gansevoort Market on the North River, and the South 
Street station, located on the corner of Oliver and South Streets near 
the East River, are identical in construction and equipment. The 
buildings are of simple design, of steel fire-proof construction, with 
concrete foundations. The Gansevoort Street building, which is 
typical of both, is one story high with basement, 63 ft. 8 in. by 97 ft. 
4 in. Each station is large enough for eight pumping units. 

MACHINERY 

28 There are now five units in each station consisting of Allis- 
Chalmers five-stage centrifugal pumps driven by AUis-Chalmers 
induction motors and the necessary auxiliary machinery. The 
motors and pumps are alike and their parts are interchangeable. 

29 The pumps each have a specified capacity of 3000 gal. per min. 
of sea-water, working with a suction lift of 20 ft. and a delivery 
pressure of 300 lb. per sq. in. The actual capacity as indicated by a 
24-hr. test was about 30 per cent in excess of that specified. The 
original specifications contemplated the use of six-stage pumps, with 
the expectation that sea-water would be used at each fire. Because of 
the facts already referred to (Par. 4), that the relative amount of water 
required for fire purposes is insignificant and that sea-water may do 
considerably more damage to goods than fresh water, a change in the 
specifications was agreed to, whereby the pumps should work at best 
eflBciency when receiving water from the Croton mains at a pressure 
on the intake side varying from 15 lb. to 40 lb. per sq. in. 

30 To meet this new condition the pumps were all built with five 
stages. All the sea connections and priming machinery as originally 
contemplated were installed, so that sea-water can be pumped into 
the mains whenever desired. The effect of the change is merely to 
reduce the pressure head slightly in case sea-water is used. 

ARRANGEMENT OP MACHINERY 

31 The floor-plans of the buildings and general layout of machin- 
ery, piping, switchboards, etc., are shown in Fig 2. As will be seen 



444 



HIGH-PRESSURE FIRE-SERVICE PUMPS 




Fig. 2 Plan and Elevation Showing Arrangement of Hydraulic and 
Electrical Apparatus in Pumping Stations 



HIGH-PRESSURE FIRK-SERVICE PUMPS 



445 



space is provided for three additional units. Working detail plans of 
the machinery were furnished by the contractor. The arrangement 
shown in Fig. 2 is the same for both stations, the only difference being 
that the switchboard and office in the South Street station are on 
different sides of the building as compared with the Gansevoort 
Street station. 

32 The motors and pumps, with suction and delivery branches, 
are located on the main floor of the pump room. The switchboard 
and switchboard apparatus are placed in an enclosed two-story and 
basement gallery. 




Fig. .3 Interior View of Station 



33 The four high-tension feeders and all other wires entering the 
building are brought in through the gallery basement. All terminal 
work on the entering wires is located in the basement. On the first 
floor of the gallery, which is approximately on the same level as the 
pump-room floor, are placed the oil switches, with their controlling 
and protective devices, fire-proof cells and compartments. 

34 The operating switchboard is conveniently located in the 
enclosing wall of the gallery, and is so placed as to allow a man 
standing on the pump-room floor to perform all the operations neces- 
sary for controlling the apparatus in the station. The bus bars, 



446 HIGH-PRESSURE FIRE-SERVICE PUMPS 

with their fireproof compartments, are placed on the second floor of 
the gallery. 

MOTORS FOR CENTRIFUGAL PUMPS 

35 The motors are of the constant-speed, wound-rotor induction 
type, 3-phase, 25-cycle, 6300-volt to 6600-volt, designed to operate 
at about 740 r.p.m. Each pump is direct-connected to its motor 
by a flexible coupling which takes care of any variation from align- 
ment. In starting, an iron grid resistance is connected in the 
secondary circuit and gradually cut out by means of a handwheel 
on the motor switchboard panel. When the resistance is all cut 
out the rotor is automatically short-circuited and operated by 
specially constructed solenoids through a small switch mounted 
directly on the shaft of the handwheel above referred to. An 
interlocking arrangement prevents the operator from closing the 
switch connecting the motor to the line while the motor is short- 
circuited. 

36 The specifications required the motors to have sufficient 
starting torque to attain full speed between 30 sec. and 45 sec. after 
starting, with a current not exceeding 150 per cent of that used when 
the motor is working under full speed. Each motor was required to 
develop not less than 800 b.h.p. when using current of 6300 volts, 
25 cycles, and under these conditions to have an efficiency not less 
than 92 per cent, a power factor not less than 93 per cent, and a 
motor slip not in excess of 2 per cent. At three-quarters load the 
efficiency was not to be less than 92 per cent and the slip not to exceed 
1.5 per cent. It was specified that the temperature of the motors 
should not rise more than 40 per cent on a 24-hr. test at full load, 
when measured by a thermometer, the air in the room being 25 deg. 
cent. 

37 Prof. Geo. F. Sever of Columbia University tested two of the 
motors in the shops of the contractor and found them to meet the 
specifications and to have a full-load efficiency of 93.2 per cent. The 
other motors were inspected and found to be alike and were assumed 
to have the same efficiency. The motors were also tested for tempera- 
ture rise at the time of the official test to be described later. 

MOTORS FOR AUXILIARIES 

38 Direct-current motors of 240 volts are provided to operate the 
various gate valves in the station and the piston pumps employed foi- 
maintaining a vacuum on the salt-water "suction lines. 



HIGH-PRESSURE FIRE-SERVICE PUMPS 447 

PUMPS 

39 As previously stated the pumps were finally constructed with 
five stages, each to give a pressure of somewhat over 60 lb. per sq. 
in., making the combined pressure of the five stages about 300 lb. 
per sq. in. above the intake pressure, which is the maximum working 
pressure of the stations at normal speed of 740 r.p.m. This tj'pe of 
pump is the simplest now on the market for pumping water either 
against a high head or low head, and this simplicity was the deciding 
factor which led to the selection of this style of machinery. 

40 The pumps are water-balanced by a piston connected to the 
last impeller and upon which the water pressure acts, but should 
any additional end-thrust occur, it would be taken up by the ball 

• bearing provided in the outboard bearing. This ball bearing consists 
of two rings of l|-in. diameter steel balls and is water-cooled. The 
balancing piston is fitted very loosely in order to keep the friction 
losses small, and as a result a considerable amount of water leaks past 
it into a chamber at the end of the pump, which is provided with a 
discharge pipe and valve leading into the suction. By adjusting 
the valve in this pipe the difference of pressures on the piston can be 
regulated as desired. The bearings are of the ring-oiled type and are 
separated from the pump casing by packing glands which prevent 
foreign matter from entering the bearings. The impellers are of 
bronze and the shaft of forged steel. All parts of the runners and 
diffusion vanes are thoroughly lubricated by oil cups on the base of 
the pumps. A feature is the wide base, shown in Fig. 4, which allows 
the pump barrel to set low, giving stability. 

41 Each combined unit is equipped with automatic and hand 
control. The pumps are kept primed for instant service and the 
simple operation of a switch on the main switchboard starts the 
machine and gives full pressure in about 30 sec. 

PRESSURE -REGULATING VALVES 

42 A combined regulating and relief valve is interposed between 
the discharge pipe and the suction pipe of each pump, and set to regu- 
late the discharge of each pump to any predetermined pressure. 

43. When the^volume of the water discharged by the pump is in 
excess of that forced into^the system, this valve acts as a relief valve 
and by-passes this excess into the suction to the pump, the pres- 
sure on the main distribution system remaining at the predetermined 



448 



HIGH-PRESSURE FIRE-SERVICE PUMPS 



point. When no water is forced into the distribution system all of the 
water discharged from the pump is then by-passed into the suction. 
44 The pressure-regulating valves were made by the Ross Valve 
Mfg. Co., of Troy, N. Y., and much of the practical success of the 
station has been due to the accuracy with which they maintain anv 
desired pressure. 



■ ■ ' i 1 

V - 




^* ' — 




.1 » « 

liJI" . ^,M^ 




WB^'^A ^^^^ 


k\ mWir'>Y^'.'« 


f^:-^l 


\mkM^& S 


^ "Wt^ 


M 


k^ 




^j 



Fig. 4 Multistage Pump, Capacity 3000 Gal. per Min.; Maximum Head, 

300 lb. per sq. in. 



PRIMING APPARATUS FOR SALT-WATER SUCTION LINES 

45 The priming apparatus In each station consists of three motor- 
driven vacuum pumps, each arranged to maintain automatically a 
vacuum of 26 in. in the suction lines. These pumps are of the piston 
single-action type, one having a displacement capacity of 300 cu. ft. 
per min. for a piston speed of 200 ft. per min. and each of the others 
a displacement capacity of 50 cu. ft. with a piston speed of 160 ft. 
per min. 

46 An air-collecting chamber is connected to each of the salt- 
water suction lines and equipped with water-gage glass and vacuum 
gage. The air-suction piping Ijetween the air chambers and the air 
pumps is provided with a veitical loop sufficiently high to prevent 



HIGH-PRESSURE FIRE-SERVICB PUMPS 



449 




u-^ 



450 HIGH-PRESSURE FIRE-SERVICE PUMPS 

water being carried over to the pumps. The air pumps are inter- 
connected to each air chamber. 

VENTURI METERS 

47 Venturi meters for measuring the discharge of water from the 
station and from one main to the other were set by the contractor on 
each discharge main and on the cross-connecting main. The meters 
of the discharge main are 24 in. in diameter and on the cross-over 
main 12 in. in diameter. These meters were standardized under the 
direction of F. N. Connet, Manager of the Venturi Meter Sales De- 
partment of the Builders Iron Foundry, Providence, R. I., and were 
provided with dial-indicating gages and also chart-recorders gradu- 
ated to indicate the flow in gallons per minute; and in addition 
with an integrating meter which registers the total flow in gallons. 

48 The readings during the test were taken by a mercury mano- 
meter, graduated to show the capacity in thousands of gallons per 
minute. For this purpose a Venturi manometer was attached with a 
temporary connection to each of the 24-in. Venturi tubes. The 
manometer gave essentially the same reading as the indicating dial 
on the main register. 

49 The Venturi manometer is practically a tube partly filled with 
mercury, one side of which communicates with the upstream pressure 
chamber of the meter tube, while the other communicates with the 
throat-pressure chamber. The connections with the manometer 
are indicated in the diagram, Fig. 6. 

50 The sketch shows a 24-in. high-pressure meter tube, its register- 
indicator-recorder and manometer. The instruments and meter tube 
are drawn to scale, but in the pumping station the meter tube is about 
75 ft. distant from the instruments. 

TESTS OF MACHINERY 

51 The specifications for the pumping system provided for an 
endurance test of each motor and pump lasting 24 hr. without stop, 
during which time the capacity and eSiciency of the pumps and 
motors were to be determined. The tests were to be in charge of an 
expert appointed by the commission. 

52 The specifications provided for making the test with sea water, 
but this was later changed to a test with Croton water under the con- 
ditions of actual use. In view of this change the contractor increased 
the efficiency guarantee from 70 to 71 per cent. 



HIGH-PRESSURE FIRE-SERVICE PUMPS 



451 




452 HIGH-PRESSURE FIRE-SERVICE PUMPS 

53 The original specifications called for a capacity of 3000 gal. of 
sea water per minute against a discharge pressure of 300 lb. per sq. 
in. and a suction lift not exceeding 20 ft. The total increment of 
pressure is equivalent to 308.66 lb. from the intake to the delivery 
side. The Croton pressure varies at the stations in different parts 
of the day from about 40 lb. to 13 lb. per sq. in. and is affected by the 
amount of water being drawn from the mains. Consequently, to 
meet the requirements, the delivery pressure would need to be 
308.66 lb. in excess of the intake pressure. There is also a further 
correction from the fact that sea water is heavier than fresh water 
and this correction under maximum conditions might amount to 2.5 
per cent. 

54 The specifications further provided that the brake horse- 
power developed by the motors under test should be computed from 
the electrical energy supplied to them, corrected for the efficiency 
of the motors as determined by the test. They further provided that 
if the aggregate of all stops exceeded one hour for any motor the test 
for capacity for such motor was to be run over again for a period 
of 24 hr. 

55 The specifications also provided that the pumping capacity of 
the apparatus and the efficiency of the pumps should be based on 
the minimum rate of pumping during any eight consecutive hours of 
the endurance test, during which none of the motors were stopped. 

56 The discharge of the pumps was determined by the reading 
of the Venturi meters, one of which was located in each discharge 
fine. These readings were under the direction of F. N. Connet, and 
were checked by observers representing the contractors and also the 
city. 

57 The modified specifications also required that the efficiency of 
each pump should be not less than 70 per cent and its capacity not 
less than 3000 gal. of sea water when lifted to a pressure equivalent 
to 308.66 lb. To determine whether the requirement was met, a 
separate test of each pump was required. 

58 The efficiency of the pumps was computed by dividing the 
horse-power output of the pumps by the horse-power input as 
received from the motors. The horse-power input was computed 
as follows: 

1 • X total wattt^ ^ . f X /r^o o i.N 

n.p. input = X efnciencv of motors (93.2 per cent) 

746 



HIGH-PRESSURE FIRE-SERVICE PUMPS 453 

The horse-power output was computed as follows: h.p. output 
_ wt. per gal. (8,34) X 2.31 head in pounds X no. of gal. per min. 
. 33000 

SOUTH STREET STATION TEST 

59 The test of the South Street Station was begun at 12:30 p.m. 
on September 2, 1908, after about 2 hr. of preliminary running for 
the purpose of adjusting the delivery pressure; it was continued with- 
out interruption for 24 hr. With the exception of a short stop of 
motor No. 2 which was shut down from 2:11 to 2:41 a.m., September 
3, to remedy a slight defect in the^^insulation of the field coils, no pump 
was stopped. During the time No. 2 was stopped the pressure on the 
delivery mains fell to about 300 lb. ; during the remainder of the test 
the pressure was maintained at or above the contract requirement, 
as will be noted by consulting the last column of Table 1. 

60 The average results for each hour for the 24-hr. test of all four 
motors are given in Table 2. The smallest delivery for eight con- 
secutive hours occurred at the last part of the test, when the 
average capacity, as shown by the readings, was 18,447 gal. per 
min., and the average efficiency was 72.2 per cent. During this 
time the average pressure pumped against was 314.5 lb., or an excess 
of about 6 lb. over contract requirement. 

61 It will be noted from the last column of Table 2 that there is 
considerable variation in the efficiency; that during the first hour the 
efficiency was less than 70^ per cent, whereas during the third and 
fourth hours the efficiency ^^exceeded 75 per cent. This variation 
in efficiency was doubtless caused by variation in the amount of water 
by-passed from the pressure to the suction side of the pump over the 
balanced piston and through the bearings, and possibly during the 
first hour by the discharge of some water through the relief valve 
which was pumped but not metered. The valves for regulating the 
differential pressure on the balance pistons were nearly closed during 
the third, fourth and fifth hours of the South Street Station test, but 
were opened the normal amount for the remaining portion of the 
test. The amount of water which for maximum difference of pres- 
sure may leak around the balance piston of any pump without passing 
through the meter could not be accurately determined but was esti- 
mated to be in excess of 4 per cent. Hence it appears that slight 
changes in the opening of the valve controlling the differences of pres- 
sure at this piston must materially affect the efficiency. The normal 



454 



HIGH-PRESSURE FIRE-SERVICE PUMPS 



opening of this valve appears to correspond to an efficiency of about 
72.5 per cent. 

62 During the test of the South Street Station all the bearings 
ran cool with the exception of those on No. 6 pump, which heated up 
during the third and fourth hours but were brought to a normal con- 
dition without stopping the pump or reducing its load by the appli- 
cation of lubricants and cooling water. 

TABLE 1 HOURLY AVERAGE OF READINGS OF DISCHARGE AND INJECTION 

GAGES ON PUMPS 

SoDTH Strekt Pumping Station, September 2 and 3, 1908 



Hour 


Pump No. 6 


Pump No. 4 


Pump No. 2 


1 

Pump No. 1 Pump No. 3 

1 ^ 


Average 


Net 
Pres- 


















1 








sure 




Disc. 


Inj. 


Disc. 


Inj. 


Disc. 


Inj. 


Disc. 


Inj. 


Disc. 


Inj. 


Disc. 


Inj. 


Lb. 


i 
12:30- 1:15 335.0 


21.8 


332.3 


22.6 


329.8 


22.5 


331. C 


24.0 332.9 


23.3 


332.2 


22.8 309.4 


1:30- 2:15 347.2 


20.9 347.0 


22.2 


345.5 


22.1 


343.8 


22.6 346.2 


22.1 


345.9! 22.0 323.0 


2:30- 3:15 345.0 


20.9 343.8 


21.8 


342.8 


22.1 


341.0 


22.5 343.4 


21.8 


343.2 21.8 321.4 


3:30- 4:00 344.2 


20.9 343.8 


21.9 


340.3 


22.3 


339.8 


22.6 342.2 


21.9 


342. li 21.9 320.2 


4:30- 5:00 341.7 


21.1 341.8 


21.9 


340.8 


22.6 


339.3 


23.1 342.2 


22.4 


341.2 22.2 319.0 


5:30- 6:00 336.2 


22.4 336.3 


22.9 


335.3 


23.8 


333.3 


25.3 335.7 


23.6 


335.41 23.6 311.8 


6:30- 7:00 337.2 


24.4 336.8 


24.9 


334.8 


25.3 


333.8 


26.3 335.7 


24.9 


335.7 25.2 


310.5 


7:30- 8:00! 339.7 


25.1 337.8 


25.9 


335.8 


26.6 


334.8 


27.3 337.7 


25.9 


337.2 26.2 


311.0 


8:30- 9:00' 341.2126.6 339.3 


26.9 


338.3 


27.3 


337.3 


28.8 338.7 


27.1 


339.0 27.3 


311.7 


9:30-10:00 344.7 27.6 342.8 


27.9 


341.8 


28.6 


343.3 


29.3 344.7 


27.9 


343.5 28.3 


315.2 


10:30-11:00 342.228.9 341.3 


29.1 


340.3 


29.3 


340.8 


30.8 342.7 


29.6 


341.5 29.5 


312.0 


11:30-12:00 343.730.4 344.3 


29.9 


343.3 


30.3 


342.3 


32.1 343.7 


30.6 


343.5 30.7 


312.8 


12:30- 1:00 345.730.6 347.3 


30.1 


347.3 


30.8 


345.8 


32.6 347.9 


31.4 


346.8 31.1 


315.7 


1:30- 2:15 334.0 


30.9 334.5 


30.6 


* 




332.6 


33.1 333.9 


31.4 


333.7 31.5 


302.2 


2:30- 3:00, 332.4 


31.1 331.0 


30. 9| 






330.8 


32.8 332.2 


31.6 


331.6 31.6 


300.0 


3:30- 4:00 349.2 


31.4 348.3 


31. 1* 


346.3 


31.6 


347.3 


33.3 349.2 


31.6 


348.1 


31.8 


316.3 


4:30- 5:00 347.2 


31.1 346.8 


30.9 


346.3 


31. -3, 


345.8 


32.8 349.4 


31.4 


347.1 


31.5 


315.6 


5:30- 6:00 346.7 


28.6 346.3 


28.4 


345.3 


29.3 


345.3 


30.6 348.2 


29.4 


346.4 


29.3 


317.1 


6:30- 7:00 342.2 


27.6 342.3 


25.4 


340.3 


26.1 


340.3 


27.1 341.2 


26.1 


341.3 


26 5 


314.8 


7:30- 8:00 332.7 


21.6 332.3 


22.1 


330.3 


22.6 


329.3 


24.1 331.2 


22.6 


331.2 22.6 


308.6 


8:30- 9:00 332.7 


20.9 332.3 


21.4 


331.3 


22.1 


329.3 


22.6 331.2 


21.6 


331.4: 21.7 


309.7 


9:30-10:00 331.7 


20.6 332.3 


20.9 


328.8 


21.8 


328.3 


22.8 331.2 


22.1 


330.51 21.6 


308.9 


10:30-11:00 334.2 


21.4 336.8 


21.6 


332.8 


22.3 


331.8 


23.3 336.2 


22.4 


334.4! 22.2 


312.2 


11:30-12:30 336.021.9 338.0 


21.9 


333.6 


22.6 


334.0 


23.8 337.7 


22.6 


335.9 22. R 

1 


313.3 



Readings corrected for error of gage and to center of pumps. 

* Pump No. 2 shut down from 2:11 to 2:41 on account of motor. 



63 It will be noted from Table 2 that the average results of the 
24-hr. test of the South Street Station exceeded the contract require- 
ments in capacity, pressure head and efficiency. 

64 The horsepower delivered by the motors during the test aver- 
aged for the 24 hr. about 920 or about 15 per cent above rating, with- 
out excessive heating. 



HIGH-PRESSURE FIRE-SERVICE PUMPS 



455 



TABLE 2 COMPUTATION OF PUMP EFFICIENCIES 
South Street Pumping Station, September 2 and 3, 1908 









Total h.p. 




Net 






Hour 




Total kw. 


from 


Gal. per 


h.p. 


Efficiency 


Beginning 


r.p.m. 


per hr. 


motors 


rain. 


pressure 
lb. 


delivered 


per cent 








(input) 








12:30 p.m. 




3875 


4841.4 


18334 


309.4 


3311.6 


68.6 


1:30 


757.0 


3851 


4811.4 


18634 


323.9 


3523.6 


73.2 


2:30 


755.0 


3829 


4784.0 


19217 


321.4 


3605.7 


75.4 


3:30 




3819 


4771.5 


19220 


320.2 


3592.8 


75.3 


4:30 


755.0 


3811 


4761.6 


19145 


319.0 


3565.4 


75.0 


6:30 




3837 


4794.0 


18995 


311.8 


3457.6 


72,1 


6:30 


765.0 


3818 


4770.3 


18970 


310.5 


3438.7 


72.2 


7:30 


755.0 


3815 


4767.5 


18980 


311.0 


3446.0 


72.3 


8:30 


757.0 


3863 


4826.4 


19020 


311.7 


3461 . 1 


71.6 


9:30 


756.0 


3868 


4830.2 


19120 


315.2 


3518.3 


72.8 


10:30 


757.0 


3873 


4838.9 


19095 


312.0 


3478.1 


72.1 


11:30 


756.5 


3859 


4821.4 


19120 


312.8 


3491.6 


72.4 


12:30 a.m. 


757.0 


3870 


4835.2 


19175 


315.7 


3534.1 


73.0 


1:30 


756.5 


3672 


4587.8 


18790 


302.0 


3315.0 


72.5 


2:30 


757.0 


3667 


4581.6 


18776 


300.0 


3288.4 


71.5 


3:30 


757.0 


3890 


4860.2 


19190 


316.3 


3543.5 


72.8 


4:30 


757.5 


3891 


4861.4 


19160 


315.6 


3530.2 


72.6 


5:30 


754.7 


3861 


4823.9 


19110 


317.1 


3337.7 


73.2 


6:30 


757.0 


3865 


4828.9 


19005 


314.8 


3492.7 


72.4 


7:30 


754.7 


3706 


4630.4 


18710 


308.6 


3370.8 


73.0 


8:30 


745.6 


3659 


4571.5 


18100 


309.7 


3272.5 


71.5 


9:30 


745.6 


3651 


4536.5 


17890 


308.9 


3226.2 


71.. '^ 


10:30 


745.0 


3619 


4521.6 


17795 


312.2 


3243.4 


71.8 


11:30 


747.0 


3618 


4519.1 


17806 


313.3 


3256.8 


71.8 


Average 


756.1 












72.5 



Average eflBciency, 1st period of 8 hr. = 73 . per cent. 
Average efficiency 2nd period of 8 hr. = 72 . 3 per cent. 
Average efficiency 3rd period of 8 hr. — 72 . 5 per cent. 

No. of cycles per 8ec. 12:30 p.m. to 6:30 a.m. » 25.6 
No. of cycles per sec. 6:30 a.m. to 12:30 p.m. =» 25.0 



TABLE 3 TEST OF INDIVIDUAL PUMPS 
South Street Station, September 3, 1908 



I'ime 


No. of 
pump 


Gal. per 
min. 


Pressure 
delivery 


Lb. per 


Sq. In. 


h.p. 
output 


Efficiency 




Inj. 


Net 


of pump 


12:58- 1:14 


1 


3372 


344.4 


29.3 


315.1 


620 


74.6 


1:22- 1:37 


2 


3809 


336.0 


27.9 


308.1 


683 


70.1 


1:43- 1:58 


3 


3495 


334.0 


28.7 


305.3 


623 


73.2 


2:03- 2:18 


4 


3705 


334.5 


27.8 


306.7 


662 


76.0 


2:24- 2:38 


6 


3740.7 

1 


344.5 


28.8 


315.7 


689 


77.0 



Immediately following the 24-hr. test for capacity. 



456 HIGH-PRESSURE FIRE-SERVICE PUMPS 

65 Immediately after the close of the endurance test of 24 hours, 
a short test was run on each motor separately, which was continued 
long enough after uniform results were shown to obtain 12 to 15 
readings. This test was run for the purpose of ascertaining whether 
there were deficiencies in any of the individual motors, and to meet the 
requirements specified in the printed specifications for the work, viz: 
that each pump should be free from defects, should have a capacity 
of 3,000 gal. per min. and an efficiency not less than 70 per cent. 
The results of these tests. Table 3, show that the individual pumps 
had an efficiency from 4 per cent to 6 per cent in excess of 
the average when operated together, and that the capacity for the 
specified discharge pressure was considerably in excess of the require- 
ment of the specification. It is, I believe, generally the case that 
individual centrifugal pumps delivering water into a main singly 
show a greater efficiency by from 4 per cent to 6 per cent than the 
same pumps delivering together into a single main, due probably to 
less loss in eddy currents and friction head, etc. 

GANSEVOORT STREET STATION TEST 

66 The endurance test of the Gansevoort Street Station with all the 
pumps in operation was begun at 9:45 a.m., September 5, after the 
pumps had been operated for about 20 min. giving uniform results. 
The test was continued for 24 hr. The method of testing and the 
various observers were the same as for the tests at the South Street 
Station and the results are given in Tables 4 to 6. 

67 For the Gansevoort Street Station the efficiency average for 24 
hr. was 72.9 per cent, with a variation (excepting the first hour) of less 
than one-half of 1 per cent. It fell below 70 per cent during the first 
hour, which was due to the opening of an automatic relief valve on 
pump No. 2, which discharged some of the water into the suction 
before it had been metered. For that reason the efiiciency during the 
first hour has not been considered in determining the performance 
of the pumps. 

68 The least capacity during the eight consecutive hours when 
all the water pumped passed through the meters occurs from 10.45 
a.m. to 6.45 p.m. The average capacity during this time is 17,419 
gal. The average net pressure in pounds is 324.3 which is nearly 
16 lb. in excess of the contract requirements. The average efficiency 
for the period above is 72.90 per cent 



HIGH-PRESSURE FIRE-SERVICE PUMPS 



457 



TABLE 4 HOURLY AVERAGE OF READINGS OF DISCHARGE AND INJECTION 

GAGES ON PUMPS 
Gansetoort Stkbbt Pumping Station, September 5 and 6, 1908 





Pump _, 

XT a Pump ] 
No. 6 


"^0.4 


Pump No. 2 


Pump No. 1 Pump No. 3 


Average 


Net 


Hour 
















Pres- 






















sure 
Lb. 




Disc. 


Inj. 


Disc. 


Inj. 


Disc. 


Inj. 


Disc. 


Inj. 


Disc. 


Inj. 


Disc. 


Inj. 


9:45-10:30 


342.4 


24.7 


342.6 


25.6 


344.8 


24.9 


346.9 


25.2 


345.1 


25.3 


338.4 


25.1 


319.3 


10:45-11:30 


347.4 


24.7 347.1 


25.7 


345.4 


25.2 


348.4 


25.4 346.4 


25.2 


346.9 


25.2 


321.7 


11:45-12:15 


348.924.9 348.9 


26.2 


347.9 


25.7 


349.4 


25.9 347.9 


25.9 


348.6 


25.7 


322.9 


12:45- 1:15 


351.425.7 350.9 


27.7 


350.1 


26.7 


352.4 


26. 9| 350.4 


26.9 


351.0 


26.8 


324.2 


1:45- 2:15 


352.4 25.7 352.4 


27.4 


350.9 


26.7 


353.4 


26.7 351.4 


26.9 


352.1 


26.7 


325.4 


2:45- 3:15 


352.925.7 352.9 


27.7 


351.4 


26.9 


354.9 


27.2 351.4 


27.2 


352.7 


26.9 


32,5.8 


3:45- 4:15 


353.9i26.4 353.9 


28.2 


353.4 


27.7 


355.4 


27.4) 353.4 


27.9 


354.0 


27.5 


326.5 


4:45- 5:15 


354.426.9 354.4 


28.7 


353.4 


27.9 


355.9 


28.2 353.4 


28.7 


354.3 


28.1 


326.2 


5:45- 6:15 


349.9 27.4 349.9 


29.7 


350.4 


28.7 


351.9 


28.9 350.9 


29.2 


350.6 


28.8 


321.8 


6:45- 7:15 


349.928.4 348.9 


30.2 


349.4 


29.2 


351.4 


29.2' 349.4 


29.4 


349.8 


29.3 


320.5 


7:45- 8:15 


351.928.9 350.9 


30.7 


350.4 


29.7 


352.4 


29.2 349.9 


29.9 


351.1 


29.7 


321.4 


8:45- 9:15 


351.929.7 351.9 


30.9 


350.6 


29.7 


353.4 


29.9 350.9 


30.7 


351.7 


30.2 


321.6 


9:45-10:15 


352.4 29.7 354.4 


30.9 


351.4 


30.2 


353.4 


30.4* 352.9 


30.9 


352.9 


30.4 


322.5 


10:45-11:15 


354.430.4 353.9 


31.2 


353.4 


30.2 


354.9 


30.91 353.4 


31.2 


354.0 


30.8 


323.2 


11:45-12:15 


354.431.2 


353.4 


31.7 


352.9 


31.2 


353.9 


31.7 352.9 


31.2 


353.5 


31.4 


322.1 


12:45- 1:15 


353.931.2 


352.4 


31.7 


352.9 


31.2 


354.9 


31.7 353.4 


31.2 


353.5 


31.4 


322.1 


1:45- 2:15 


352.431.2 


351.9 


32.2 


352.9 


31.2 


354.4 


31.9 


353.4 


31.2 


353.0 


31.5 


321.5 


2:45- 3:15 


350.431.7 349.4 


32.2 


349.4 


31.4 


351.9 


32.2 


350.4 


31.7 


350.3 


31.8 


318.5 


3:45- 4:15 


350.9'32.2 350.4 


32.7 


348.9 


31.2 


351.4 


32.7 348.4 


31.4 


350.0 


32.0 


318.0 


4:45- 5:15 


350.9'31.9 350.4 


32.4 


350.4 


31.2 


352.9 


32.2 349.9 


31.2 


350.9 


31.8 


319.1 


5:45- 6:15 


352.431.2 351.9 


32.2 


352.4 


30.7 


353.4 


31.7^351.9 


31.4 


352.4 


31.4 


321.0 


6:45- 7:15 


350.929.9 349.9 


31.2 


349.4 


29.9 


351.9 


30.7 '349.4 


30.9 


350.3 


30.5 


319.8 


7:45- 8:15 


349.428.7 349.4 


29.7 


347.4 


29.4 


349.9 


29.2 ''347.9 


29.4 


348.8 


29.3 


319.5 


8:45- 9:45 


348.127.4 347.7 

1 


28.9 


346.4 


28.0 


348.4 


28.0 346.4 


28.6 


347.4 


28.2 


319.2 



Readings corrected for error of gage and to center of pumps. 

69 The average capacity for the entire test is 17,867 gal. which 
was obtained with an average speed of 753.6 r.p.m. 

70 Immediately after the completion of the endurance test of 24 
hours duration, each pump was tested when operating alone for a 
period sufficiently long to obtain 12 to 15 readings after they had 
become practically uniform. These tests gave in every case an 
efficiency several per cent greater than that obtained when the pumps 
were all discharging into the same main. 



CONCLUSIONS 



71 It appears from the endurance test in each station that the 
capacity, efficiency and pressure exceeded the contract requirements 
by a large margin, and that during the endurance test no mechanical 



458 



HIGH-PRESSURE FIRE-SERVICE PUMPS 



TABLE 5 COMPUTATION OF PUMP EFFICIENCIES 
Gansetoobt Stbeet Pumpinq Station, SEPTSMBsa 5 and 6, 1908 







1 


Total h.p. 




Net 

pressure 

lb. 


1 ' 




Hour 
Beginning 


r.p.m. 


Total kw. 
1 per hr. 


from 
motors 


Gal. per 
min. 


h.p. 
delivered 


Efficiency 
per cent 








(input) 






1 


9:45 a.m. 


740 


3671 


4586.5 


17107 


319.3 


3188.9 


69.5 


10:45 


749 


3589 


4484.2 


17310 


321.7 


3251.0 


72.5 


11:45 


750 


3591 


4486.7 


17290 


322.9 


3259.3 


72.9 


12:45 


752 


3591 


4486.7 


17280 


324.2 


3270.5 


72.9 


1:45 p.m. 


752 


3604 


4502.9 


17285 


325.4 


3283.6 


72.9 


2:45 


753 


3604 


4502.9 


17315 


325.8 


3293.3 


73.3 


3:45 


753 


3630 


4535.3 


17345 


326.5 


3306.1 


72.9 


4:45 


753 


3685 


4604.0 


17670 


326.2 


3365.0 


73.1 


5:45 


756 


3696 


4617.9 


17855 


321.8 


3354.4 


72.6 


6:45 


756 


3661 


4574.0 


17825 


320.5 


3335.2 


73.4 


7:45 


754 


3676 


4592.9 


17775 


321.4 


3335.2 


73.3 


8:45 


753 


3685 


4604.0 


17755 


321.5 


3332.5 


72.8 


9:45 


755 


3657 


4569.0 


17720 


322.5 


3336.2 


72.9 


10:45 


755 


3693 


4614.2 


17755 


323.2 


3350.1 


72.7 


11:45 


756 


3704 


4627.9 


17830 


322.1 


3352.8 


72.6 


12:45 a.m. 


756 


3753 


4689.0 


18195 


322.1 


3421.4 


73.0 


1:45 


756 


3760 


4697.8 


18310 


321.5 


3436.6 


73.3 


2:45 


756 


3735 


4665.5 


18315 


318.5 


3405.5 


73.0 


3:45 


755 


3725 


4654.0 


18290 


318.0 


3395.5 


73.0 


4:45 


756 


374:5 


4677.6 


18315 


319.1 


3411.9 


73.0 


5:45 


756 


3784 


4727.9 


18330 


321.0 


3435.0 


72.7 


6:45 


755 


3747 


4681.6 


18315 


319.8 


3419.4 


73.0 


7:45 


755 


3723 


4656.5 


18255 


319.5 


3405.0 


73.1 


8:45 


755 


3722 


4655.3 


18189 


319.2 


, 3389.5 


72.8 


Average 

1 


753.6 


i 


1 






1 1 


72.9 



Average efficiency, 1st period of 8 hr. = 72 . 9 per cent. 

Average efficiency, 2d period of 8 hr. = 73 . per cent. 
Average efficiency, 3d period of 8 hr. = 72 . 9 per cent. 

No. of cycles per sec. 9:45 a.m. to 2:45 p.m. = 25.00 
No. of cycles per sec. 2.45 p.m. to 4.45 p.m. = 26.25 
No. of cycles per sec. 4:45 p.m. to 6:45 p.m. = 25.50 
No. of cycles per sec. 6:45 p.m. to 7:45 p.m. = 25.00 
No. of cycles per sec. 7:45 p.m. to 9:45 p.m. = 25.25 
No. of cycles per sec. 9:45 p.m. to 9:46 a.m. = 25.50 



or electrical defects were observed. During the test of the South 
Street Station one of the pumps was stopped for half an hour to 
repair the motor insulation, while during the test of the Gansevoort 
Street Station no stop was made. The bearings in both stations 
were in perfect condition at the end of the test and the temperature 
of the motors not suflficiently high to interfere with the continuous 
operation for a longer period. Apparently the endurance test could 
have been continued indefinitely without injuriously overworking 
or overloading the pumps and motors. 



HIGH-PRESSURE FIRE-SERVICE PUMPS 



459 




460 



HIGH-PRESSURE FIRE-SERVICE PUMPS 



72 The specifications call for pumping sea water, which most 
authorities consider to be approximately 2,5 per cent heavier than 
fresh water. The effect of substituting sea water for fresh water 
would have been to reduce the capacity of the pump by about 2^ 
per cent for the same horse-power delivered by the motor, without 
sensibly affecting the efficiency. Because of the large capacity 




150 300 250 300 350 400 
Net Pressure on Pumi) Lb. per Sq. In. 

Fig. 8 Characteristic Curves of the Pump for Varying 
Discharge-Pressures 



RESULTS OF TESTS 

shown by the pump, this does not materially affect the results in 
relation to the contract requirements. 

73 The data and results of the tests at the two stations are given 
concisely in the tables. The efficiency is given as computed 
for each hour, and shows a slight variation which probably can be 
accounted for by changes in the amount of water leaking past the 
balancing piston. The individual pump tests at the South Street 



HIGH-PRESSURE FIRE-SERVICE PUMPS 



461 



Station show a variation in efficiency from 70 per cent to 77 per cent, 
and at the Gansevoort Street Station from 70 per cent to 79 per cent. 
This variation may have been due to the structure of the pumps but in 
my opinion is more probably due to variable leakage past the bal- 
ancing piston or through the relief valves. 

74 Pump No. 6 at the Gansevoort Street Station was tested with 
varying openings of the valve in the discharge pipe. The results are 
shown in the latter half of Table 6. 

TABLE 6 TEST OF INDIVIDUAL PUMPS 
Gansevoort Stkeet Station, September 6, 1908 



Time 


No. of 
pump 


Elect. 

h.p. 

input 


Gal. per 

min. Hg. 

Col. 


Pressure 
delivery 


Lb. Pbb Sq. In. 


h.p. 
output 


EfiSciency 


Inj. 


Net 


of pump 


10:05-10:31 


1 1 


916 


3800 


356.8 


35.4 


321.4 


711 


77.6 


10:36-10:51 


2 


877 


3800 


350.8 


35.1 


315.7 


700 


70.8 


10:54-11:12 


3 


920.5 


3820 


350.4 


34.1 


316.3 


703 


78.0 


11:17-11:30 


4 


892 


3751.4 


352.5 


35.6 


316.9 


695 


77.7 


11:37-11:53 


6 


899 


3880 


350.9 


35.2 


315.7 


714 


79.4 


11:55-12:03 


6 


880.3 


3457 


376.1 


36.0 


340.1 


686 


77.9 


12:03-12:07 


6 


929 


4500 


304.4 


34.6 


269.8 


708 


76.1 


12:09-12:13 


6 


946 


5070 


255.6 


33.6 


222.0 


654 


69.4 


12:24-12:28 


6 


952 


5500 


207.4 , 


33.2 


174.2 


559 


58.7 


12:32-12:36 


6 


927 


5588 


155.2 


33.2 


122.0 


397 


42.8 



Immediately following the 24-hr. test for capacity. 



PRACTICAL RESULTS FROM THE NEW SYSTEM 

75 The high-pressure fire system in New York, which was put 
officially into service on July 6, 1908, has been successfully operated 
at many fires, but it had a crucial test on January 7, 8 and 9, 1909, 
when it was brought into service for five simultaneous fires, three of 
them of more than the usual extent and activity, and one particu- 
larly so. Information upon the results attained with the system and 
the amount of water consumed was given by Chief Engineer I. M. 
de Varona and published in the Engineering News of February 11, 
1909. 

7G The fires occurred at Hudson and Franklin Streets, Hester' 
Street and the Bower}'-, Houston Street and Broadway, Sixth Ave- 
nue and 17th Street, and Houston Street and the Bowery. The 
situation became so dangerous that every engine south of 37th Street, 
or 40 engines, were summoned, as well as a force consisting of 12 
battalion chiefs and more than 600 men, but there was no need to 
use a single one of the engines. 



462 



DISCUSSION 



77 As the violence of the fires increased, additional pumps were 
brought into service, so that at one time four pumps and motors were 
in commission at the South Street Station and three pumps at the 



TABLE 7 SPECIFIED CHEMICAL ANALYSIS FOR PUMP MATERIALS 



Nickel steel 



Parts of 1 per cent 

Phosphorus not to exceed 0.04 

Sulphur not to exceed 0.04 

Tensile strength at rupture, pounds 100,000 
Tensile strength at elastic limit, 

pounds I 65,000 

Per cent elongation in 8 in 2 

Per cent elongation in 2 in 22 

Contraction of area per cent 32 

Carbon not less than 20 parts of 1% 

Nickel percentage I 21 to 24 



Medium 
steel 



0.10 

650,000 

32,500 
22 



Steel 
forging 



0.04 
0.04 
75,000 

38,000 

22 
32 



Steel 
casting 



0.05 
0.05 
65,000 

32,000 

18 
24 



Gansevoort Street Station, delivering 35,500 gal. per min. against 
an average pressure of 225 lb. at the pumps and 205 lb. at the hydrants. 
During the operation of the pumps 14,095,000 gal. were pumped as 
recorded by the meters, and the current used was 81,450 kw-hr., 
the cost of which was $1222, 



DISCUSSION AT NEW YORK 

Prof. George F. Sever.* The electrical features of this installa- 
tion are of much interest but the reasons for selecting that system 
which is now in operation should be given. In the discussion of this 
problem both alternating and direct-current power were considered for 
the operation of the motor-driven pumps, and alternating-current 
power was decided upon. The reasons for such selection I have 
noted herewith: 

a Absolute simplicity, which is the key-note of the electrical 
end of this power installation. 
' Professor of Electrical Engineering, Columbia University. 

Note. — The high-pressure system was designed by I. M. de Varona, Chief 
Engineer of the Department of Water Supply, Gas and Electricity of New York. 
It was also constructed under his supervision. The construction of the electrical 
machinery was supervised by Prof. Geo. F. Sever as Consulting Engineer. The 
details of construction were in charge of Thomas J. Gannon, John P. Reynolds 
and Henry B. Machen, assistant engineers of the department. The machinery 
of each station was designed a,nd erected by the Allis-Chalmers Co. of Milwaukee. 



HIGH-PRESSURE FIRE-SERVICE PUMPS 463 

b Commutating apparatus and brushes are entirely absent- 

c Induction motors provide very quick starting when it is 
necessary to operate the station on a fire signal. 

d There is less expense for copper in the distribution system 
to insure continuity of service, 

e The induction motor is a less expensive apparatus than the 
direct-current motor. 

/ With the induction motor there are absolutely, no exposed 
live circuits in the station, as there might be with a 
direct-current apparatus. The final decision was for 
3-phase service at 6600 volts and 25 cycles. It was 
decided that it would not be desirable to establish a 
power house to be operated by the city because it would 
be a municipal plant. 

2 In order to insure continuity of service there is brought to each 
pumping station an independent feeder from each of the two Water- 
side stations of the New York Edison Company. There is also brought 
to each pumping station an independent feeder from the nearest sub- 
station of the New York Edison Company, as follows: to the Ganse- 
voort Street station two feeders from the Horatio Street sub-station, 
and to the South Street station two feeders from the Duane Street 
station of the company. Hence there are really four independent 
sources of power supply for each pumping station, assuring practi- 
cally no possibility of shutdown. 

3 The contract for electric power for the Manhattan stations was 
let to the New York Edison Company. This contract provides for 
two payments, the first for a reservation of 3250 kw. capacity, of gen- 
erating, distributing and controlling apparatus, available at either 
pumping station at an instant's notice, or practically without any 
notice at all. Thus four pumps can be thrown on with absolutely no 
notice to the New York Edison Company that they are to be used. 
For that reservation, and care and maintenance of the whole distribu- 
ting system, the city pays about $63,000 per year, and the city also 
pa5''s one and one-half cents per kw-hr. for all high-tension power 
used in each station. 

4 Another stipulation in the contract may be of interest to 
engineers as it provides for the protection of the city. This stipu- 
lation is as follows: "If the contractor, under the terms of this 
contract, shall fail to maintain and deliver a continuous and uninter- 
rupted supply of electric power when required, the contractors shall and 
will pay to the city the sum of five hundred dollars per minute for 



464 DISCUSSIOM 

each minute's interruption or delay of electric power supply after 
the power has been interrupted or delayed for three consecutive 
minutes. " So, if they cannot deliver power after an interruption of 
three minutes, immediately a charge of $500 per min. is imposed and 
is deducted from the bills which the New York Edison Company 
renders. 

5 The operation of both these stations is extremely simple. The 
handle of the oil switch is turned, throwing the 6600 volts directly on 
the stator of the motor. By turning a hand wheel, the motor is 
brought up to speed in less than 33 sec, and in starting the current is 
not supposed to exceed 150 per cent of the full-load current, which 
is 64 amperes. As far as I have observed the operation of the stations, 
there has been absolutely no trouble from the electrical end, no trouble 
with the feeder system, and none with the motors, and I think the 
City of New York has two plants which will give it for many years 
to come absolutely no trouble whatsoever. 

Wm. M. White. The paper deals with questions in which I am 
directly interested. The methods employed in making the tests 
were probably the best that coula have been selected. There is 
probably no more accurate method of determining the quantity of 
water delivered by a pump than by the venturi meter, especially 
when in the hands of an expert who is familiar with its workings. 
The venturi meter, as Professor Carpenter says, has been used for 
a number of years; it has been tested in various ways and proved to 
give accurate results. The power deUvered to the pumps can be 
most carefully measured by electrical instruments. 

2 The writer accepts without question the various efficiencies 
obtained and^presented by^the author, who states, calling attention to 
the variation in efficiencies obtained, that the individual observations 
do not agree asj^closely as he would like. I do not think Professor 
Carpenter should offer any apology as the results seem to agree very 
closely, and certainly are as accurate as are generally obtained on work 
of this kind. The efficiencies obtained on^these pumps, though not 
the highest that have been obtained, are as high as is usual for 
similar conditions of head, capacity and speed. The designers of 
the pumps deserve credit for the performance shown by the pumps. 

3 I am at a loss to find a reason for the variation in efficiencies of 
the pumps, as mentioned in Par. 65,jwhere it is stated that individual 
pumps delivering water into a main singly show greater efficiency 



HIGH-PRESSURE FIRE-SERVICE PUMPS 465 

than the same pumps delivering together into a single main. I 
assmne, of course, that the variation in efficiency refers to the pumps 
when they are deUvering exactly the same quantity against the 
same head at the same speed, whether working singly or in parallel. 
In the normal operation of pumps, it would be a fact that when 
one pump was operating from a suction main to a discharge main, 
the efficiency of that pump would be different from what it would 
be when working with another 3ump from the same suction main 
and discharging into the same discharge main, because the two 
pumps would usually be working against a higher head than when a 
pump was working singly. The increased head on the pumps would 
mean a decrease of capacity, and the increase of power demanded by 
two motors instead of one would mean a shght increase in line loss, 
which would again sUghtly decrease the speed and slightly change 
the conditions of operation for two pumps over that which would 
exist when one pump only was in operation. Of course, under 
these conditions, the two pumps would show different efficiencies, 
because the efficiency curve of a pump varies as its capacity and 
head. 

4 I do not believe, however, that this is the condition to which 
Professor Carpenter refers. I assume that he has corrected for 
this difference, and has obtained from two pumps working in parallel 
the same capacities, heads and speeds as though one pump were in 
operation, and that under this latter condition he finds the differ- 
ence in efficiency in the two pumps. If this be a fact, it is the most 
important point brought out from a designer's point of view. 

5 I am at this time attempting to duplicate the conditions, to see 
whether the efficiencies are different under the same conditions of 
capacity, head and speed, as mentioned by Professor Carpenter. 

George L. Fowler. A number of years ago I was associated 
with Joseph Edwards, who at that time had the contract for exca- 
vating the ship channel in New York Harbor, probably one of the 
first, if not the first, very large hydraulic engineering projects suc- 
cessfully accomplished by the contractor and to the satisfaction of 
the Government. 

2 The ship channel leading from the Narrows down to Sandy 
Hook and out to sea, is about 15 miles long, and runs almost due 
south first, turning to nearly due east before reaching Sandy Hook, 
and passing through Gedney Channel to the sea. Cutting across it is 



466 



DISCUSSION 



the Swash Channel, not used by any deep-draft boats. When the 
work was undertaken New York Harbor was shoal at two points on 
the Gedney Channel and the ship channel, where the water depth 
was a little less than 24 ft. The Government had a survey made and 
an estimate of costs based on material actually removed by the ordi- 
nary methods of dredging. Through the open space from Sandy 
Hook to Coney Island the whole lower bay is subject to all the winds 
coming in from the Atlantic on the east and across Raritan Bay, so 
that the water is nearly always rough. Two contractors had at- 
tempted the work by ordinary bucket dredging and both had failed. 




Fig. 1 Hydrattlic Dredger for Deepening Ship Channels 



3 In the ship channel the material was sand and sedimentary 
clay, lying over hard sand ; in the Gedney Channel it was gravel, shell 
and sand, for two feet overlying hard shingle. Hydraulic dredging 
was specially suited 'for this kind'pf work, and many kinds of material 
were removed Jrom^the channel besides the ordinary silt. 

4 Three sea-going vessels were built for this work by the Joseph 
Edwards Company: the ReHance, the Advance, and the Mt. Waldo. 
Fig. 1 shows the general arrangement of the ships. At A is the long 
drag aft, where the pipe goes into the vessel and where the pumps are 
located, each driven by a 192-h.p. engine at 178 r.p.m. The suction 
and delivery pipes were 15 in. in diameter, with a shell of 40 in. The 



HIGH-PRESSURE FIRE-SERVICE PUMPS 



467 



pumps delivered 10,000 gal. per min. at a velocity of 1 100 ft. The 
efficiency was thus between 65 and 70 per cent, although in later 
tests made by the Government, when nothing but water passed 
through the pipes, the efficiency rose to as high as 80 per cent. 

5 The shoe used is a hook that drags along the bottom, chains 
being fastened to the vessel for this purpose. The vessel never 
stopped from morning to night, simply running out to sea, dumping, 
and coming'^back again to work. 




Fig. 2 Detail of End of Suction Line 



6 At the point L, Fig. 2, was the heavy shoe that served to dig 
into the mud and gravel. At was a butterfly valve, kept open all 
the time to admit water above the drag to mix with the material 
raised. At the bottom K was another valve which could be opened 
in an emergency, in case not enough water was admitted at 0. 



468 



DISCUSSION 



7 The pump itself was of a plain centrifugal type, 40 in. in diam- 
eter, with vanes cut away at the center, as shown in Fig. 3. Because 
of this arrangement, the material would come in at C and out of the 




Fig. 3 Sectional View of Centrifugal Pump for Dredging 



vanes at the discharge, without damaging the pump when heavy 
substances were drawn in. The three vanes were made with wings 



HIGH-PRESSUKE FIRE-SEKVICE PUMPS 



469 



bolted on, and accessible from both sides. The thrust was taken up 
by the bearing at T (Fig. 4) , the nuts marked m being screwed into a head 
carried by the bars 0, bringing the thrust plates at the point i. The 
reason for threading the nut m was to adjust it to the vanes in proper 
relative position to the sides of the pump. That is a simple construc- 
tion maintained ever since, with the exception that ball bearings are 
now used. 

8 Although the pumps were originally intended to take water and 
other loose material, such as sand and gravel, they proved capable of 
lifting practically anything that came in their way. The three fol- 
lowing specimens are interesting as showing the pumps' lifting power: 




Fig. 4 Detail of Thrust Beaking of Pump 



o A piece of shaft weighing 70 lb. raised and passed by a 15-in. 
dredging pump; improvement of New York Harbor, 
Steamer Reliance. 

6 A piece of tree root raised and passed by a 12-in. pump from 
14 ft. of water at Miami, Fla. ; Florida East Coast Railway 
Company improvements. 

c A piece of pig iron measuring 11^ in. by 4| in. by 3^ in. and 
weighing 35 lb. raised and passed by an 8-in. special cata- 
ract wrecking-pump from 15 ft. of water from the wreck 
of a canal boat sunk at Puas Dock, Yonkers, N. Y.; by 
Baxter Wrecking Company, New York. 

9 For hydraulic dredging, the Government pays by the scow load 
and gets what is excavated. In ordinary hydraulic dredging, like 
that in the ship channel, about 15 per cent of the pump discharge was 
solid matter. About 40 per cent in excess of the amount deposited 



470 DISCUSSION 

in the bins went overboard with the overflow, and was carried out to 
the flats at the sides by the cross currents, which also carried the loose 
material stirred up b}'- the drag. The result was that the Government 
obtained an excavation about 70 per cent in excess of what would 
have been obtained had all of the material removed from the bottom 
been caught in the bins. This, of course, greatly reduced the actual 
cost of the excavation. For example: the last contract made on the 
ship channel was at the rate of 16|- cents per yard, while with the 
allowance indicated, above the actual cost per yard — channel meas- 
urement — it was about 11 cents. 

10 As for the time of loading, some records indicate that this 
ship, 157 ft. long and with a capacity of 650 cu. yd., was loaded in 
48 min. ; there are also records of its being loaded at the rate of 16 cu. 
yd. per min., of solid matter placed in the bins; and records of its 
taking out to sea nearly 4000 cu. 3'd. per day. The vessel was 
worked in all kinds of weather, even when tackles had to be used to 
board her; and yet the ship was taking her load steadily. Except 
in the case of an actual breakdown the work could be carried on for 
16 hr. per day. 

John H. Norris. In a pumping plant of the character described, 
this type of equipment seems in the present state of the art the most 
suitable that could have been selected. I would like, in this connec- 
tion, to call attention to another type of installation for service of 
this kind, though not on so large a scale, which appeals to me as 
being more desirable than the electric-driven centrifugal pumping 
plant taking its power from a public utilities company. 

2 At Coney Island was installed the first plant operated by the 
City of New York for fire protection by means of water delivered 
into mains under high pressure, with the idea of taking care of a 
restricted area where there was great danger from fire. 

3 This plant consists of three 150-h.p. three-cylinder, vertical 
gas engines direct-connected to triplex pumps, each unit capable of 
pumping 1500 gal. per min. against a pressure of 150 lb. These 
engines take their fuel from the mains of the local gas company and 
can be arranged if necessary to run on gasolene. They are installed 
in a building on city property and are arranged to take their water 
supply from the city mains or from Coney Island Creek, within 50 ft. 
of the pumping station. The engines are started with compressed air, 
and the three units can be started up in less than three minutes. 
On every occasion they have been found ready for service whenever 



HIGH-PRESSURE FIRE-SERVICE PUMPS 471 

the demand was made upon them. The cost of this pumping station 
was as follows: 

Building $10,000 

Equipment 37,000 



$47,000 



The annual operating expenses are: 

Labor $13,140.00 

Supplies and Repairs 897 . 27 

Fuel 150.00 



$14,187.27 

4 By comparing the foregoing figures it will be evident that for 
service smaller than is required in the City of New York, the gas- 
engine-operated triplex pump gives an economical equipment that 
can be allowed to stand idle for any length of time and yet be ready 
for instant service. 

5 New York City pays the New York Edison Company an annual 
charge of $90,000 for the privilege of calling for sufficient current to 
operate the equipment at any time. This item capitalized at 5 per 
cent would pay for a good-sized gas-engine plant. 

6 The following data were taken from the capacity tests of the 
Coney Island units: 

Duration of test 14 hr. 

Average piston speed of pump 90.3 ft. per. min. 

Total head pumped against 156 . 5 lb. 

Average pump horsepower for each unit 142.2 h.p. 

.Average gas consumed per hour for the 3 units 8914.0 ft. 

Average capacity 4512 . gal. per min. 

Slip of pump 3 . 45 per cent 

Average efficiency of pumps 82 . 00 per cent 

J. R. BiBBiNs. Although Professor Carpenter's paper deals pri 
marily with multistage pumps, I wish to direct attention to the ques- 
tion of motive power, upon which the success or failure of the system 
practically depends. We have seen excellent examples of two systems 
diametrically opposed in regard to power supply — the electrical and 
the gas-driven system. Under certain conditions, both are extremely 
serviceable. The first high-pressure installation on a large scale, in 
this country, was the gas-driven system at Philadelphia. Although 
I have not had an opportunity to follow the results of that station for 
the past two or three years, the results obtained and pul)lished for 



472 DISCUSSION 

the first year or so showed that such a system of gas-driven pumps 
merits every consideration, 

2 First as to the security of power supply: In Philadelphia the 
Delaware Avenue station receives its gas supply directly from a 
24-in. trunk main running between two very large gas holders, located 
in different parts of the city. Roughly, the pipe line measures four 
miles in length, its capacity constituting a considerable reserve in 
itself, if both the holders were unavailable. There is no intermedi- 
ary apparatus whatever between the pipe line and the engine ; that is, 
the plant may draw directly on these two large holders of several 
million cubic feet capacity. This constitutes a very safe and reliable 
source of motive power which can hardly be paralleled except, per- 
haps, by the situation in the New York electric service, where there 
are so many stations to draw from. 

3 In this connection, I would like to ask whether it is at present 
possible to utilize the storage battery capacity in the various sub- 
stations for reserve service at the high-pressure pumping station. 
It is stated that the storage batteries are available for reserve in 
emergencies, such as discontinuance of the main high-tension current 
supply. I am under the impression that an inverted rotary requires 
a direct-driven exciter to maintain a definite frequency and prevent 
racing. Without special controlling apparatus, this inversion would 
be impossible in the ordinary sub-station equipment. Possibly special 
provision has been made in the New York systems, in which case, 
the security of power supply is certainly beyond criticism. In other 
words, would it be possible to invert the synchronous converters on 
short notice? 

4 Second, quick starting: It seems to be a fact that a large part 
of the minimum time required for the starting of a fire-service station 
is consumed in the operation of the motor-driven by-pass valves. In 
Philadelphia these valves are operated from an independent supply, 
as in New York, and at least fifteen seconds are required to close them; 
whereas the engines are brought up to speed within half a minute 
from the time the signal is given, the remaining time being usually 
consumed in closing this motor-driven valve. 

5 The various tests of the Philadelphia plant showed that each of 
the units could be readily put on the fine in well under one minute. 
It is an interesting fact that the original underwriters* tests specified 
the time limit as twelve minutes for the starting of the first three units, 
whereas the whole station can be started in that time, and has been 
started in seven minutes. 



HIGH-PRESSURE FIRE-SERVICE PUMPS 473 

6 During the 36 days of preliminary service trials of the Phila- 
delphia station, out of one hundred alarms given, onl3'-four misses were 
made in getting any of the eleven units started. In not a single 
instance has the station, as a whole, failed to respond to the service, at 
least during the period over which my observation extended. This 
has been accomplished with the regular operating force of three men. 

7 Third, in regard to the cost of service at Philadelphia; The 
only data on a large fire available, are those of the fire in the Coates 
Publishing House, which lasted about nineteen hours. The average 
cost for pumping was about six cents per thousand gallons, including 
gas, wages and supplies. The cost of the large East Side service, 
cited .in the paper, is about nine cents for power alone, and I think 
this does not include the readiness-to-serve factor. On the other 
hand, it is patent that the cost of service in either the gas or the 
electrical station is relatively unimportant. The main desideratum 
is reliability. 

8 Finally, I desire to advance an argument for the development of 
a new type of pump unit, namely, a high-speed gas-driven centrifugal 
pump. Some time ago, in connection with water-works service, I 
found great difficulty, even with the present high-speed single-acting 
gas engine, in matching engine speeds with those required in centrifugal 
pump work However, for the pressure necessary in water-works 
practice, about 125 lb., one or two sizes of engines were found to be 
directly applicable to multistage pumps, with fair proportion of parts 
and good efficiencies. It seems possible to adopt a modified type 
of gas engine which would permit the direct connection mentioned. 

9 This modification would naturally follow along lines of short 
stroke and high piston speeds with perhaps four cylinders. The 
engines at Philadelphia were designed with a piston speed of but 730 
ft. per min. with a 22-in. stroke. This might be increased to 1000 ft. 
per min. without exceeding present-day limits, especially for units 
designed for occasional service. Such a unit would find immediate 
application in many industries and would combine the high economy 
of the gas engine with the simplicity of the centrifugal pump. The 
efficiencies shown by Professor Carpenter place the centrifugal pump 
in a position of closest competition with reciprocating pumping units. 

J. J. Brown. I recently made a series of tests on three 6-in., 8-stage 
centrifugal pumps, each designed for 1000 gal. per min. and 560 lb. 
pressure at 1200 r.p.m. One of these pumps gave an efficiency from 
wire to water of 71 per cent, or a pump efficiency of 76 per cent. 



474 DISCUSSION 

regret that Professor Carpenter did not give the results of his tests 
on the New York fire-service pumps at lower capacities. All of the 
tests were made at capacities considerably in excess of that for which 
the pumps were designed and they apparently show their best effi- 
ciency at approximately 25 per cent over the normal rating. This 
increased efficiency at excess capacity seems to be apparent in several 
recent tests made on high-lift centrifugal pumps. The 8-stage 
machines previously referred to give their best efficiency at 1300 gal., 
or about 30 per cent over rating. 

2 Mr. White has raised a question as to the difference in efficiency 
between the New York fire-service pumps working in multiple and 
as separate units. I think this is occasioned by the variation in 
capacity of the pumps when working together on a common suction 
and discharge line. I have found it rather difficult to balance two 
centrifugal pumps on a common discharge, and pitot tube tests indi- 
cate in almost every case a considerable difference between the amounts 
of water handled by the individual units under these conditions. 

3 I have in mind one installation on fire service, where the pumps 
were called upon to deliver against the maximum pressure for which 
they were designed and it was only with considerable difficulty that 
we were able to cut in additional units. I think that if venturi meters 
or pitot tubes had been placed on the discharge of each of the five 
pumps when they were working in multiple, a difference in capacity 
of the several units would have been shown, which would account 
for the difference in eflaiciency observed when the pumps were working 
individually and not in multiple. 

George A. Orrok. At the time of the award of contract for these 
fire pumps, the New York Edison Company was obtaining proposals 
for centrifugal feed pumps — a somewhat similar service — and eight 
1000-gal. 300-lb. pressure five-stage pumps were purchased. There 
was no attempt to obtain a high guarantee for efficiency, but the 
builders did state that under the above conditions an efiiciency of 
65 to 68 per cent would be obtained. These pumps were of the Jager 
type and under test showed an efficiency of about 68 per cent. 

2 Fig. 5 shows that the high-pressure fire-service pumps are of the 
Kugel-Gelpke type and should be a trifle more efficient because of 
smaller friction and leakage. Seventy-one per cent seemed a very 
high efficiency and many doubts were expressed regarding the ful- 
fillment of the guarantees. The extreme figure of 79 per cent 
obtained is probably the result of careful design and extra good shop 



H1GH-1'1{ESSURE FIUE-SERVICE PUMPS 475 

work and I believe has not been excelled. That this figure came as 
a surprise may be explained by the fact that most centrifugal pumps 
are stock pumps and not specially designed for the work they have to 
do. Pump manufacturers have been more concerned in getting a 
line of patterns that will suit standard conditions than in developing 
a line of pumps and system of patterns capable of doing the best work. 
3 As a centrifugal pump is a mixed-flow or Francis reaction turbuie 
reversed, similar care in design and construction would probably 
give efficiencies similar to those of the best makes of reaction turbines, 
which approximate 90 per cent. 

Frederick Ray. The difference in efficiency of the units oper- 
ated individually from that obtained when several were operated in 
parallel might be due to the different rates of flow through the 
ventuii meters under the two conditions. With one pump operating, 
this flow would be low and the mercury column reading would be but 
slightly over an inch, so that with a given error of observation the per- 
centage of error would be much greater than with two or three pumps 
discharging through the same meter. 

2 Professor Carpenter here replying that the pipe connecting the 
two meters was open all the time, Mr. Ray continued: 

3 This would equalize the flow in the meters, so that the mercury 
column reading when the whole station was running would be 
about 6^ times the reading with one pump. It has not been my 
experience that parallel operation of a number of pumps has any 
tendency to decrease or otherwise change the efficiency obtained 
when operated individually. The efficiency should be the same, and 
in this case, as the pressures were taken at each pump, any losses in 
the piping system due to parallel operation would be external to the 
gages and would not show in the calculations. If the pressure had 
been taken at the discharge of the whole system, losses in the piping 
would affect the results. 

4 Many pumps are running under similar conditions, at the 
efficiencies given. I have myself obtained efficiencies of 80 per 
cent and higher, but I do not rely as much on them as on some a 
little lower. I am now testing a 6-in., 2-stage underwriter pump, 
having a normal capacity of 500 gal. per min. against 100 lb. pres- 
sure, which has developed a maximum efficiency of 73 per cent. . 

5 I think the centrifugal pump is the ideal one for fire service, 
not only on account of its simplicity and reliability, but also on 
account of its characteristic increase in capacity as the pressure is 



476 DISCUSSION 

reduced. Thus, the 500-gal. underwriter pump referred to will dis- 
charge 870 gal. per min. at 60 lb., or enough for four streams at this 
pressure. It will give three streams at 90 lb., two streams at 110 
lb. and one at 117 Ib.-^all at constant speed without any regulation 
whatever. 

6 The City of Toronto has recently issued specifications for cen- 
trifugal pumps for their general municipal water supply, among which 
are several fire pumps capable of discharging against 300 lb. pressure. 
These pumps, however, are to be equipped with variable-speed induc- 
tion motors, the pressure regulation being obtained by speed variation. 
This is superior to throttling regulation from the standpoint of cur- 
rent economy and in the case of the New York installation a con- 
siderable saving could be made by this means, as most of the fires can 
be handled with 200 lb. pressure or less. 

H. Y. Haden. a somewhat unusual result obtained from this 
type of pump is that as the total head continues to increase beyond a 
certain point, the capacity falls off, with the result that the capacity 
curve, as given in Fig. 8, shows a backward tendency. It will be 
interesting to get the explanation of this. 

2 There is unquestionably a large field in fire protection for steam- 
turbine-driven centrifugal pumps, and it is to be hoped that the Fire 
Underwriters will officially accept this type of fire protection unit. 
I believe that a properly designed centrifugal pump, for high speeds and 
of few stages, can be used to great advantage when direct-connected 
to high-speed turbines. 

Thomas J. Gannon/ It was decided to use electricity as power 
for the pumping stations, because [the first cost of installation, 
yearly cost of operation and maintenance and ^fixed charges 
were estimated to be lower, taking into account the intermittent 
service. The construction and operation of a steam plant were 
entirely out of consideration and the choice lay between gas-engine- 
driven and electric-driven pumps receiving power from outside 
sources. 

2 It was estimated that gas operation of plants equal in capacity 
to the present electrically driven plants, would involve a fixed 
charge of $50,000 a year, in addition to the cost of the gas actually 
consumed. The question as to who should build and maintain 

* Engineer, Dept. Water Supply, Electricity and Gas, Manhattan Borough 
New York. 



HIGH-PRESSURE FIRE-SERVICE PUMP8 477 

the necessary large gas mains, the cost of which would approximate 
a million dollars, was not definitely settled. That the cost of a 
gas-engine-driven pumping plant would have been approximately 
double, both for machinery, building and area of land to be pur- 
chased, is borne out by the actual costs of the installations in Man- 
hattan and at Coney Island. 

3 The capacity of the gas-operated Coney Island plant is 4500 
gal. of water per min. against a head of 150 lb. per sq. in. The com- 
bined capacity of the two pumping plants in the Borough of Man- 
hattan, as originally laid out, was 30,000 gal. per min. against a head 
of 300 lb., with provision in each station for three additional pumping 
units of a capacity of 3000 gal. each, making a total combined capacity 
of 48,000 gal. per min. agauist 300 lb. pressure. On actual test, 
however, the capacity of the pumps was approximately 20 per cent 
greater than the designed capacity. 

4 Furthermore, the flexibility of this type of pump permits of an 
increased discharge at lower pressures, which gives a capacity of 
approximately 5500 to 5600 gal. per min. for pressures between 150 
and 200 lb., or a combined total capacity of 55,000 gal. per min. 
against 200 lb. pressure. This corresponds to the pressure at which 
the station is operated for most fires. In other words, the water 
horsepower of the electric-driven as compared Avith the gas-engine- 
driven riant is approximately in the ratio of 20 to 1. 

5 The cost of the machinery in the Coney Island plant was 
approximately S^37^000, and the cost of the building approximately 
SI 0,000. The cost of each of the two Manhattan pumping stations 
complete, exclusive of land, was practically S240,000. The first cost 
of installation of the gas-engine-driven plant is therefore more than 
double the first cost of installation of an equivalent electrically-driven 
plant, in the city of New York. 

6 The high-pressure fire-service pumping stations went into 
official operation on July 6, 1908. It was at first decided to put the 
stations in service only when called on by the fire department, and 
up to and including November 20, 1908, the pumping stations were 
called upon to go into actual service for but 17 fires. On that date, 
the method of operation was amended so that the pumping stations 
are put in service in response to every alarm in the high-pressure 
district, and continue in operation awaiting instructions from the 
fire department. Under this system, from November 20 to December 
31, 1908, the pumps responded to 116 first alarms. From the best 
available information, water was used in 55 instances, making a 



478 DISCUSSION 

total of 72 fires for which the high-pressure service had been used 
up to that date. 

7 To insure readiness for service at all times, daily tests are made, 
of at least half an hour's duration, unless the station has been in 
actual operation during the preceding 24 hours. 

8 During the first quarter of 1909 the number of aiarms received 
was 239, and water was taken from the station for 125 actual fires. 
The total amount of water pumped was 17,840,000 gal., and 145,900 
kw-hr. was consumed. It was on January 7, 8 and 9 of this quarter 
that the three large simultaneous fires mentioned in Par. 75, occurred, 
for which over 14,000,000 gal. of water was pumped, leaving about 
3,800,000 gal. for the balance of actual fires occurring dm'ing the 
quarter. For these three simultaneous fires more than 81,000 kw- 
hr. was consumed while the total consumption of power for the 
quarter for all fires and testing purposes was but 145,900 kw-hr. 

9 As to why a pump running singly develops a higher eSiciency 
than when running in conjunction with several others, it is observed 
that pumps of the same type do not necessarily develop their best 
efficiency at the same speed and pressure. The pump running 
singly will naturally develop a pressure which corresponds to its 
own design, but when working in multiple, it will have to adjust 
itself to the common pressure. 

10 As to reliability I have neither seen nor heard of any time 
when any one of the ten pumps installed in the Borough of Man- 
hattan has failed to respond instantly when called on for service 
and to develop the full pressure on the system within one minute's 
time. At no time in service have the pumps shut down of their 
own accord. 

Henry B. Machen.' Among the many difficulties encountered 
during the construction of the distribution system, perhaps the 
greatest was that due to the congested sub-surface of the street, 
which was a source of continual extra expense to the contractor, 
and of worry to the man in charge of selecting the location for the 
excavation of the trench. 

2 The intersection of Sixth Avenue and Fourteenth Street may 
be cited as an example, since complete notes are available, due to the 
station excavation for the Hudson Tunnels. Here there were nine 
gas mains east and west, and nine north and south, belonging to 

' [engineer, Dept. Water Supply, Electricity and Gas, Manhattan Borough, 
Nnw York. 



IIKiH-PRESSURE FIKE-SERVICE PUMPS 470 

four different companies; two water mains in each direction; sewers 
and their connections on each side of the street; five Edison duct 
lines, and five duct lines with large manholes belonging to the Con- 
soUdated Telegraph and Electric Subway Company or the Empire City 
Subway Company; the conduits and banks of ducts of the Fourteenth 
Street and the Sixth Avenue trolleys; and lastly, the columns of 
the elevated railroad with their deep foundations. 

3 Through this network the high-pressure main had to be so 
laid that the construction of the Sixth Avenue tunnel would not 
require it to be relaid. The excavation was carried on by tunneling, 
with here and there an opening through which the earth could be 
hoisted, using a pail let down by a rope. The pipe was lowered 
into the trench some distance up the street and pulled through, 
piece by piece, inspection of the running of the joint and caulking 
being almost impossible, since the space admitted but one man 
at a time after the pipe had been hauled in. 

4 This condition existed at nearly all intersections of the main 
thoroughfares, such as Broadway, Sixth Avenue, Fifth Avenue, 
the Bowery, etc., and accounts for the high cost of la5dng the mains, 
averaging about $11 per ft. complete. 

5 The second great difficulty encountered was in obtaining the 
prescribed test, which called for 450 lb. pressure per sq. in. to be 
held for 10 min., during which time the leakage was measured. 

6 The system contained about 40,000 castings, 30,000 being 
straight pipe, tested at the foundry to 650 lb. The specials were not 
tested. All these castings, as already stated, were tested in the 
ground to 450 lb., the mains being under pressure in sections about one 
block long, between gates. 

7 During the eighteen months the system has been^^in service, 
there have been but three breaks in the mains, all three in castings 
which had been subjected to the foundry test of 650 lb., two breaking 
at 150 lb. and the third at 300 lb. pressure. 

8 To overcome the danger should a break occur [during a fire, 
the proposed extensionSjto the distribution system now under contract, 
amounting to about $1,500,000, are laid out on what the department 
calls the [duplex system. This method of overcoming the difficulty 
was first suggested by Mr. Blatt, assistant engineer of the high- 
pressure bureau. It consists of laying two entirely independent 
systems of mains^and hydrants in alternate streets, the hydrants 
of one system being painted red and the other green. The mains are 
so laid out that at nearly all intersections of streets hydrants of 
l)<>th colors are available. 



480 DISCUSSION 

9 Should a break occur in either system, the operator at the 
pumping station would at once know in which system the trouble 
was located by looking at the venturi meters, and by throwing a 
switch he would start the closing of two electrically driven valves, 
separating one system from the other. Hydrants would then be 
available and in service pending the location and isolation of the 
damaged section. 

10 The section now in operation was designed to give 20,000 
gal. per min. on any one block with a loss due to friction from pumps 
to hydrant not to exceed 40 lb. The duplex extension will give 
the same results, and should either half be out of service by an acci- 
dent, there will still be available at the same location 10,000 gal. per 
min., with a loss from the pumps to the hydrant in the most unfavor- 
able location not exceeding 50 lb. 

Richard H. Rice. This paper shows that the installation de- 
scribed was made after the most careful study and a very intelligent 
choice of the types of apparatus to be used. The choice of the 
centrifugal pump for the work described is thoroughly justified by 
its simplicity and by the efficiencies obtained. The centrifugal pump 
is today the popular means of producing pressure for emergency fire 
purposes, as in the fire boats of New York, Chicago, Duluth and San 
Francisco, and the new high-pressure service of San Francisco . In San 
Francisco twelve of these pumps are now being installed, four on fire 
boats and eight for an auxiliary fire installation. On the fire boats 
centrifugal pumps are particularly adaptable as they can be run in 
series or in parallel. In parallel they give 150 lb. pressure, and in 
series the pressure is doubled. This pressure is particularly valuable 
where walls have to be battered down, or streams thrown long 
distances. 

2 The choice of alternating current as the source of power, in view 
of the unlimited supply of current existing and the duplicate means of 
conducting it into the station, is also justified. In cases where 
electricity is not so available as it is in New York, steam turbines 
are being installed, and they offer advantages over the gas engine, 
where maximum reliability is considered. 

3 As an emergency installation pure and simple, I think the 
installation mentioned in the paper can be still further simplified. 
I believe the speeds chosen for operating the pumps are too low, 
and that the pumps contain too many stages. I have had occasion 
to make extensive researches in centrifugal pump design with special 



HIGH-PRESSURE KIRE-SERVICE PUMPS 481 

reference to operation at steam-turbine speeds, and have found that 
they can be operated at high speeds with a smaller number of 
stages, giving efficiencies comparable with those obtained here, 
although the question of efficiency is subsidiary to reliabihty for 
this service. Pumps for this service should be designed with two or 
three stages at the most, and with considerably higher speed. 

4 Pumps can also be designed without balancing pistons, which 
are undesirable from the viewpoint of possible interruption of service. 
An inspection of Fig. 5, illustrating the construction of the pumps, 
will show that the balancing pistons used are quite liable to damage 
if water containing sand or other impurities is used, and this damage 
would very probably result in stoppage of the pump when it is 
badly needed. The use of balancing pistons is unnecessary in such 
emergency apparatus and should be avoided. 

C. A. Hague. A question has been asked several times with 
reference to the results of tests of efficiency on centrifugal pumps 
operating singly and in multiple or group. Professor Carpenter 
has given the very plausible explanation that the difference in effi- 
ciency in favor of the pumps running singly is probably due to the 
presence of eddies and disturbances in the pipes when the pumps 
are operating together and the absence of such eddies and disturb- 
ances when only one pump is at work. In my experience in installing 
pumps and condensers singly and in groups I have found them 
extremely sensitive to each other in operation, both in taking in 
and discharging the water, when more than one pump is working on 
a line. 

2 In the Manhattan stations, it seems to me that the suction or 
inlet pipes and the discharge pipes are coupled too closely for best 
efficiency; and also that the inlet pipe close to the pumps is not large 
enough for operation in multiple, although perhaps ample for a 
single pump when the water is undisturbed by the draft and dis- 
charge of several pumps. I have experimented considerably in 
that line, and have found that a comparative!}^ large body of water 
next to the pumps on the suction side will materially ease the machines 
in their performance. The idea is to come up to the building with a 
normal supply pipe, and then enlarge it very considerably just where 
it enters the building, providing the inlet pipe with a good-sized air 
chamber wherever possible. I have tried this several times with 
excellent results. 

3 Mr. Brown mentioned the difficulty of cutting in with a second 



482 DISCUSSION 

pump where the first pump was akeady running, a difficulty which 
I think is also due to too close connections along the inlet and outlet 
lines and a cramped conditior generally. Of course, a disturbance [in 
the water column and in the hydraulic horsepower would unbalance 
the electric power to a certain extent, perhaps not much, but the 
total disturbance may very easily result ^in the loss of several points 
in the efficiency. 

4 Considering the fact that the city pays by the kilowatt-hour 
for its electric current as per switchboard reading, it would be no 
more than proper to state the efficiency of ^the machine as a whole, 
and not exclusively upon the basis of motor efficiency obtained in 
the shop of the makers a thousand miles or so away. In this case 
when 100 h.p. in current is supplied to the switchboard, the motor 
has shown an output by a competent test of 93.2 h.p. (Par. 37) , the 
balance of 6.8 h.p., charged against the city in the power bills, being 
lost in heat and friction. Then, all that is charged against the 
pump is 93.2 h.p. The 67.57 h.p. shown by the pump for each 100 
h.p. at the switchboard indicates only 67.57 per cent total efficiency, 
although the 67.57 h.p. indicates 72.5 per cent efficiency of the power 
delivered by the motor. I have tested several centrifugal pumping 
plants of various sizes and powers, and the total efficiency generally 
shows from 64.5 per cent to about 68 per cent and very seldom above 
the latter figure. 

5 Mr. Bibbins touched upon ;^the possibihties of utiUzing the 
centrifugal pump for waterworks service, but uponj investigation 
he would find a vast difference between emergency service, where 
operating economy counts for little in the face of great danger from 
fire, and the steady and necessarily economical service required for 
the continual pumping in waterworks stations. To show how decep- 
tive a portion of the truth may be, a case is cited where a pumpage 
of a capacity of 10,000,000 gal. per day against 110 lb. load could 
easily be accomplished with displacement steam machinery by an 
expenditure of $10,000 per annum for coal. But an attempt to 
drive centrifugal pumps by electricity resulted in a cost for electrical 
power, at $6.50 per 1,000,000 gal., of $23,725 per annum; showing a 
difference in favor of displacement steam machinery equal to 
5 per cent per annum on $275,940. There is no conceivable 
difference in cost of machinery, buildings, maintenance, attendance, 
or anything else, that would justify such a preference for electricity 
and centrifugal pumps over steam and displacement pumps. Note 
the following figures: 



HIGH-PRESSURE FIRE-SERVICE PUMPS 



483 



10,000,000 gal. daily, against 110 lb 440 pump-h.p. 

120,000,000 steam duty with S lb. evaporation in the 

boilers, coal at $2.50 per net ton delivered §9928 per annum 

Electric power at S6.50 per 1,000,000 gal. against 110 lb. 

means 3,650,000,000 gal. per annum at S6.50 §23,725 per annum 

The difference in cost for the element of power is S13,797 

per annum, which at 5 per cent would capitalize at $275,940 

6 The steam-driven, reciprocating, displacement pumping engine 
can show a mechanical efficiency from the power put in through the 
throttle to the water-horsepower of the pumps, as high as 96 per 
cent, never as low as 90 per cent, under the above conditions. The 
centrifugal pump when steam-driven has a corresponding efficiency 
of about 65 per cent, and when electrically driven of about 67 per 
cent. A comparison of tests is given in Tables 1 and 2 in which it 
will be seen that the steam plant saves enough to pay 8.6 per cent 
on its entire cost. 



TABLE 1 COST OF OWNING AND PUMPING WITH HIGHEST TYPE 
AND CLASS OF STEAM PUMPING MACHINERY 

One Unit, Steam-Driven, Reciprocating, Displacement Machinery, 
Capacity of 25,000,000 Gal. Against 87 Lb. 

Pump horsepower 870 

Boiler horsepower for triple-expansion vertical pumping engine 450 

Engine house and foundations and engine foundations ^ 

Boiler house and foundation, boiler foundations, chimney, etc 

Vertical triple-expansion pumping engine \ $150,000 

450 h.p. of boilers 

Building for coal supply 



CHARGES against PLANT PUMPING ENGINE 

Interest 4 per cent 

Sinking fund 5 per cent 

Depreciation 2 per cent 

Oil waste, etc 1 per cent 



Total 12 per cent 

CHARGES against PLAN'J — BOILERS 

Interest 4 per cent 

Sinking fund 5 per cent 

Depreciation 5 per cent 



Total 



3 enginecs. 6 firemen. 3 oilers. 
Ooal it $2.10 aer net ton 



14 per cent 



484 DISCUSSION 

StTMMARY FOR StEAM RECIPROCATING MACHINERY 

Coal per annum $11,957.40 

Wages per annum 9,900.00 

Capital charges on engine 13,920.00 

Capital charges on boilers 1,260.00 

Capital charges on buildings 1,548.00 

Total charges per annum $38,585 . 40 

Cost per 1,000,000 gal $4.11 

Cost per horsepower 43 . 16 



TABLE 2 COST OF OWNING AND PUMPING WITH HIGHEST TYPE 
ELECTRO-TURBINE PUMPING MACHINERY 

One Unit, Electric-Driven, Centrifugal Machinery, Capacity 25,000,000 

Gal. against 87 Lb. 

Pump horsepower 870 

Two-stage, electric-driven centrifugal pump 

Engine house and foundations and pump foundations 

Transformer house and foundations \ $43,750 

Transformers, lightning arresters, conductors, controllers and auxil- 
.aries 



charges against plant — PUMPING MACHINERY, ETC 

interesi 4 per cent 

Sinking fund 5 per cent 

Oil, waste, etc 1 per cent 

Depreciation 2 per cent 



Total 12 per cent 

3 Engineers. 3 Extra men 
Electric current, $4.50 per 1,000,000 gal. 

Summary for Electro-Turbine Machinery 

Electric current per annum $41,062.50 

Wages per annum 5,700 . 00 

Capital charges on machinery 4,314 . 00 

Capital charges on buildings 468 . 00 

Total charges per annum $51,544 . 50 

Cost per 1,000,000 gal $5.64 

Cost per horse power 59 . 24 



HIGH-PRESSUBE FIRE-SERVICE PUMPS 485 

Thos. J. Gannon. In reply to Mr. Hague I will read the condi- 
ditions which occurred on the evening of January 7, when both 
pumping stations were put to a crucial test: 

7.22 First alarm, Hudson and Franklin Sts. 

7.28 Second alarm, Hudson and Franklin Sts. 

7.29 Third alarm, Hudson and Franklin Sts. 
7.46 Fourth alarm, Hudson and FrankUn Sts. 
7.54 First alarm, Bowery and Hester Sts. 
8.17 Automatic, Mercer and Houston Sts. 
8.19 Second alarm, Bowery and Hester Sts. 
8.29 Second alarm, Mercer and Houston Sts. 
8.32 Third alarm, Bowery and Hester Sts. 
8.40 Third alarm, Mercer and Houston Sts. 
8.43 Fourth alarm, Mercer and Houston Sts. 
8.45 Fifth alarm, Mercer and Houston Sts. 

2 In due time seven pumps were put into operation, with a dis- 
charge which reached at times over 35,000 gal. per min., and it was 
estimated that over 52 fire streams were in service at the same time. 
Each pump responded instantly and remained in service until ordered 
shut down. The pressure was ordered gradually increased from 125 
lb. to 230 lb., where it was maintained throughout the greater part 
of the time that the fires raged. The operating force at each pump- 
ing station consisted of but one engineman, one oiler, one telephone 
operator and one laborer. 

Prof. George F. Sever. A question was asked as to the feasi- 
bihty of using the storage battery capacity to invert the rotaries 
and provide alternating current, to be spread through the alternating- 
current system to the sub-stations, and from those to provide alter- 
nating current to the pumping stations. In our preliminary investi- 
gation, if I recall the facts correctly, we were assured that this could 
be done; giving us another feature of reliability in the operation 
of the system. If the Waterside station should go out of business, 
we could still get current from the sub-station. 

A. C. Paulsmeier.' While the reasons given in the paper for 
the selection of electric-driven turbine pumps do not coincide with 
the conclusions as to reliability that have been reached in the West, 
there can be no question about the careful study given by the engi- 
neers who planned the high-pressure fire system described. 

1 Chief Enginef r, Byron Jackson Iron Works, San Franciso, Cal. 



486 DISCUSSION 

2 The pumps show a remarkable efficiency, and one of the principal 
points that should commend them to those interested is their great 
flexibiUty as to capacity, a characteristic that every fire pump should 
possess. 

3 The eight fire pumps now being built for the City of San 
Francisco are of a combined capacity of 216,000 gal. per min., 
under a working pressure of 300 lb. Each of these pumps is driven 
by a 750-h.p. Curtis steam turbine, operating at a normal speed of 
1800 r.p.m. 

4 In addition there are now being completed four fire pumps 
for the boats Dennis Sulhvan and David Scannel, of an aggregate 
capacity of 9000 gal. per min. under 300 lb. working pressure, or 
18,000 gal. per min. under 150 lb. working pressure, the pumps 
being so arranged that they work either in series or in parallel. 
The pumps have all been subjected to 24-hr. tests, and while the 
data on these tests are not sufficiently complete for pubhcation, 
they show that the pumps are not as flexible as to capacity, or 
are not as capable of pumping an excess quantity of water, as are the 
Manhattan pumps. The reason for this is that the impellers in 
the San Francisco pumps are only 13 1 in. in diameter, while the 
inlet to the impellers is less than 10 in. in diameter, this opening 
being further restricted by the pump shaft, so that it is impossible 
to obtain much excess water, no matter how much below the normal 
the discharge pressure is carried. 

5 In the station pumps now being built the velocities at the 
entrance to the impellers have been somewhat decreased, although 
it is impossible to make anything like the excess capacity shown by 
the Manhattan pumps, which have impellers of such a size that 
the inlets may be made anything consistent with good practice. 

Prof. W. B. Gregory. It is gratifying to know that efficiencies 
ranging from 70 to 80 per cent may be obtained with well-designed 
five-stage turbine pumps. The high-pressure fire-service pumps in 
New York represent one extreme of conditions, while at the other 
extreme is the centrifugal pump used in the rice irrigation territory 
of Louisiana and Texas for raising large quantities of water through 
comparatively small lifts. 

2 The improvement in design of pumps of the latter class in 
the last ten years, and especially in the last five years, has made it 
possible to specify an efficiency' uf 75 per cent, even with heads as 
low as 10 ft. Pmchasers of pumping plants in this section are no 



HIGH-PRESSURE FIRE-SERVICE PUMPS 



487 



longer satisfied with pumping outfits having efficiencies ranging 
from 50 to 60 per cent. 

3 As examples of the results obtained with pumps of the class 
that deals with large volumes of water, the tables are quoted from 
recent acceptance tests conducted by the writer, of pumping plants 
used for rice irrigation. 



TABLE 1 ACCEPTANCE TESTS 
Tanrem-Compouxd Condensing Engines, Direct-Connected 
Cane and Rice Belt Irrigating Company, Fulshear, Texas, August 12 and 14, 1908 



WORTHXNGTON Pt7MP8 



Size of pump (diameter discharge pipe), in 

Water pumped, gal. permin 

Head on pump, ft 

Efficiency of engine and pump, % 

Efficiency of pump(engine93 %) 



First 
Lift 



I = 



Second 
Lift 



45 ^ 45 

47,620 / 46,430 

33.90 13.95 

69 . 5 73 . 6 

74.7 /'9.2 



Cross-Compound Condensing Corliss Engine, Direct-Connected 
Sabine Canal Company, Vinton, La., May 22, 1909 



Worthington Pump 

Size of pump (diameter discliarge pipe), in 

Water pumped, Ral. per min 

Head on pump, ft 

Efficiency of engine and pump, % 

Efficiency of pump (engine 90 % ) 



45 

44.010 

23.2(5 
^" 69.5 

77.3 



Tandem-Compound Condensing Corliss Engine, Direct-Connected'" 
Second Lift, Neches Canal, July 16, 1909 



Morris Machine Works Pump 



Size of pump (diameter of discharge pipe), in 

Water pumped, gal. permin 

Head on pump, ft 

Efficiency of engine and pump (maximum), % 

Efficiency of pump (engine efficiency 93.2 %max.). 



60,300 
10. 1 i' 
69.0 
75 



Charles B. Rearick. Electrically driven fire pumping-stations 
for large cities are dependent upon current from an outside source, 
usually a large central power plant. It would seem quite practicable 
in many cases to locate new fire pumping stations adjacent to some 
large power plant having considerable boiler capacity. In such 
cases it would be possible to drive the centrifugal or turbine pumps 
with steam turbines, and thus eliminate the necessity of large over- 



488 DISCUSSION 

load capacity in electric generating units for the central station, and 
also the liability of derangement of the lines between the power 
stations and the pumping stations. The charge for standby service 
per annum should be less than for similar electric service. 

2 The steam turbines have the advantage of being operative at 
any speed, and in this manner will maintain in the discharge mains 
any pressure desired. Furthermore, automatic regulating valves can 
be used in connection with the turbine to maintain constant pressure 
irrespective of demand or flow. 

3 It is probable that the cost of installation would be less than 
for electric-driven units. The turbine could run non-condensing, as 
the question of steam consumption is of small moment for fire service. 

Henry E. Longwell. The last paragraph of the paper furnishes 
a striking illustration of how purely academic is the ordinary official 
efficiency test, and of how little value as a basis on which to predicate 
the results that may be expected when the plant is operated under 
service conditions. 

2 This paragraph gives general figures on the performance of the 
pumps during the fire run. There were 14,095,000 gal. pumped, 
with a current consumption of 81,450 kw-hr. The average net pres- 
sure against which the pumps operated is not stated, but assuming 
it was 300 lb. per sq. in., the pump efficiency, after allowing for 
the losses in the motor, would be only 40 per cent. However, we 
know that for part of the time the pressure did not exceed 225 lb., 
or, considering the pressure in the suction mains, about 200 lb. 
net. If the entire quantity of water had been pumped against this 
lower pressure, the [efficiency would be well under 30 per cent. 
It is therefore perhaps fair to assume that the actual average effi- 
ciency was not far from 35 or 36 per cent, or say, in round numbers, 
only one-half that shown on the official test,, when the load and other 
conditions of operation were more favorable. 

W. M. Fleming. With the rapidly increasing size and height of 
office buildings, the annual fire loss in the business districts of the 
cities of the United States is increasing to an alarming extent. The 
installation of these tremendously effective fire-fighting systems has 
already proved of definite value in the reduction of city fire losses, 
and consequently of insurance costs. 

2 What was probably the pioneer large and independent so- 
called high-pressure fire system in this country was installed at 



rnCH-PRESSURE FIRE-SERVICE PUMPS 



489 




490 DISCUSSION 

Philadelphia in 1903-1904. This plant differs in almost every 
important detail from the New York system more recently installed ; 
yet the general results in each case have been excellent. In Phila- 
delphia the plant has so many times proved of great value in actual 
service that a much larger fire-fighting system, consisting of pump- 
ing units identical with those originally selected, is now being installed 
to protect what is known as the Kensington mill district. 

3 From the original Philadelphia station at Delaware Ave. 
and Race St., a location unlikely to be seriously injured by con- 
flagration, Delaware River water is supplied to independent high- 
pressure fire-service mains which effectually cover more than 425 
acres at the center of the business district. The pumping units 
consist of vertical double-acting triplex power pumps built by the 
Deane Steam Pump Company, direct-connected to Westinghouse 
vertical 3-cylinder 4-cycle gas engines each of 280 h.p. The seven 
large pumping units have each a nominal capacity of 1200 U. S. 
gal. per min., at 300-lb. pressure, and two small units have a capacity 
of 350 U. S. gal. at the same pressure. 

4 The general arrangement of the Philadelphia pumping station 
is similar to that of the large NtiW York installations (Fig. 1). 
Two rows of pumping units occupy the main floor of the station. 
The pumps are nearest the center, and the gas engines are located 
in the same relative positions thereto as the motors in the New York 
pump houses. A platform extending along the sides of the building, 
about ten feet above the floor, serves as a working gallery for the 
operation of the engine throttles. Space is provided for the installa- 
tion of three additional pumping units, and all mains are propor- 
tioned with the ultimate probable capacity of the plant in view. 
Suitable connections are provided to the mains so that the capacity 
of the pumping station may be supplemented by the use of the 
city's powerful fire boats, should occasion require. 

5 The internal -combustion engines are of the well-known standard 
Westinghouse type and require little explanation. Speed regulation 
with varying loads is accompHshed by the action of a centrifugal 
governor controlling the quantity of combustible admitted to the 
cylinders. Ignition is by a very neat type of make-and-break mecha- 
nism contained in a cyhndrical plug. Two independent igniters are 
provided in each cylinder, and three independent sources of ignition 
current are available at all times. The engines are started by the 
use of compressed air, which is admitted to one of the cylinders at 
the proper time to secure rotation in the direction required until the 



HIGH-PRESSURE FIRE-SERVICE PUMPS 



491 



^ 



regular cycle of operation is established. The pumps are started 
under no-load. 

6 The pumps are of the vertical, double-acting piston, triplex 
power type, requiring comparatively small floor space and giving a 
rate of discharge so smooth and uniform as to make imperceptible at 
the hose nozzles any pulsation in pressure. 

7 In Fig. 2 is a sectional view of one of the pumps, indicating quite 
clearly the extreme simphcity and accessibility of the machine, 
and its general construction. All valves are of the poppet type, 
readily accessible through handhole openings. Valve areas and 
waterways naturally are comparatively large, so that friction losses 




Fig. 2 Side and Sectional. End Elevation of Triplex Pumps ton thk 
Philadelphia ITigh-Pressure Fire-Pumping Station 

are reduced to a minimum. The water ends are thoroughly brass- 
fitted in order that the pumps may be readily started after a long 
period of disuse. ; 

8 There is a connection through a 12-in. check valve, from the 
city mains to the high-pressure system, so that the mains and pumps 
are constantly primed with a pressure of 60 lb. and are ready for 
service at all times. A complete system of fire-alarm boxes and tele- 
phones, with underground wires, permits direct communication 
between the vicinity of any fire and the pumping station. On the 
sounding of the alarm, the station force, consisting of an engineer 
and his assistant, can bring the total plant of seven large units 



492 DISCUSSION 

into service in seven minutes, and have repeatedly done so. Work- 
ing pressure is invariably available at the hydrants one minute 
from the time of the alarm. Such a result would be impossible 
with ordinary movable apparatus. 

9 The pumping units are started up under no-load, by the 
use of a motor-driven by-pass valve, through which the pump dis- 
charges into an overflow, until the normal cycle of operations has been 
set up in the gas engine, when the switch is closed, causing the by- 
pass valve to close and the discharge to be directed into the fire mains. 

10 Experience has indicated that the maximum pressure of 300 
lb. is required only for the most extensive fires, and for fires in the 
higher parts of tall buildings. The pressure records show that 
probably 75 per cent of the water pumped is required at not more 
than 150 lb. to 175 lb. pressure. The pressure desired in each case, 
is dictated over the telephone by the fire chief, the required pressure 
regulation being obtained by proportioning the number of units in 
operation to the requirements. 

11 The practical results of the use of the Philadelphia fire system 
have been: material reduction in fire losses in the protected district, 
large decrease in fire insurance rates, and a greater willingness on 
the part of property owners in the protected section to erect pre- 
tentious office buildings. 

12 Though the writer is unable to present a statement as to 
the annual saving to property owners by the installation, yet in 
view of the low cost of operation of the plant, there can be no question 
but that it presents a considerable yearly saving to the city. During 
the year 1907, which is perhaps typical, water was deUvered to 16 
fires, the longest one lasting 44 hr. The plant responded to 1 16 alarms 
at which no service was required. The operating expenses for the 
year were as follows: 

Gas, 839,488 cu. ft. at $1.00 $839.49 

Electric lighting 343.99 

Electric power 7 . 98 

65 tons pea coal at $3. 50 227.50 

Supplies furnished the pumping station for the entire year 1907 1,500 . 00 

Total fixed chargesfor 1907 ." $2,918. 96 

Summary I 

Salaries (Total for entire staff) $8,389.72 

Total cost materials 2,918.96 

Total operating expenses $11,308 • 68 

Total daily maintenance charge, salaries and operation . S31 . 12 



HIGH-PRESSURE FIR