Full text of "Armour engineer"

See other formats

Illinois Institute

of Technology

Libraries

Digitized by the Internet Archive

in 2010 with funding from

CARLI: Consortium of Academic and Research Libraries in Illinois

http://www.archive.org/details/armourengineer03armo

THE ARMOUR ENGINEER

THE SEMI-ANNUAL TECHNICAL PUBLICATION
OF THE STUDENT BODY OF

ARMOUR INSTITUTE OF TECHNOLOGY

CHICAGO, ILLINOIS

VOLUME 111 NUMBER 1

JANUARY, 1911

ILLINOIS INSTITUTE OF TECHNOLOGY
PAUL V. GALVIN LIBRARY
35 WEST 33RD STREET
CHICAGO, IL 60616

G. H. EMIN

THE ARMOUR ENGINEER

VOLUME 111 NUMBER 1

JANUARY, 1911

BRIQUETTED COAL AND ITS VALUE AS A RAILROAD
FUEL.

BY CHARLES T. MALCOLMSON, E. E.*

Member A. I. M. E.

The conservation of our fuel resources, which has become a
subject of active interest in past few years, is exemplified in
the history of the briquetting industry as we follow its de-
velopment in Europe and later in this country. We may nat-
urally expect to find its inception in the countries where a
thrifty people have learned to husband their resources and
turned to good account their poor or depleted fuel supply. A
country like ours, of such wonderful natural resources and so
profligate in their use, does not offer the proper stimulus to
an industry which depends upon trade conditions of high prices
where close profits have forced economy in the small detail
of saving.

Nomenclature.

The name "briquet," which is now universally used for
all forms of compressed fuel, was applied originally in Paris
to fuel made from peat with the addition of wet clay, similar
to our present day methods of making wet clay bricks. The
term was later made to include all fuel made by compression
without the use of a binder in contradistinction to that made
from bituminous and anthracite coal with pitch or other bind-
ers. We find numerous other names used, such as "boulet, "
"charbon agglomeres," or "houilles agglomeres," abbreviated
to "agglomeres" in France; "briquettes de charbon" in Bel-
gium; "patent fuel" and "compressed fuel" in England;
"kohlensteine" or "kohlenzeiglen" in Germany, applied gen-
erally to briquets made from true coals with binder: while
"artificial fuel" embraced all fuel manufactured from coal,
lignite, peat or other form of combustible.

In America the word "briquet" has been accepted as a
generic term for the product, while specific names such as
"pressed fuel," "coalette" and "carbonet" are found in the
trade. "Eggettes" are generally applied to briquets made on

*Class of 1897. Briquetting Engineer, Roberts & Schaefer Co., Chicago.

THE ARMOUR ENGINEER

[Jan., 1911

the so-called "Belgian roll" type of press, a name said to
have been invented by Mr. Ware B. Gay for the product of a
Loiseau press of this type. Fig. 1 shows samples of American
made briquets.
Historical.

The earliest record on the briquetting of coal was suggested
in a pamphlet by Sir Hugh Pratt in 1594. The first satisfactory
briquetting machine was built in France in 1842 by M. Marsais,
and since that time, the industry has gone steadily forward in
all the European countries.

FIG.

SAMPLES' OF BRIQUETS.

A. Briquet made by U. G. I. Co., Philadelphia.

B. Briquet made on Zwoyer press.

C. "Eggette" made by Solvay Process Co., Detroit.

D. "Carbonet" made at Hartshorne, Okla., plant.

E. Briquet made on Johnson press.

F. Briquet made by Briquette Coal Company.

The first briquetting plants were installed in England in
1846, Belgium in 1852 and Germany in 1861. About 1870, the
briquetting of brown coals was first successfully accomplished
in the last named country.

Prominence was given to the industry by the exhibits of
briquetting machinery at the Paris Exhibition of 1867,
and the following year we find the first recorded in-
terest for coal briquetting in America. In 1870 E.
F. Loiseau installed at Port Richmond, Philadelphia, the first
coal briquetting plant. The press used was of Belgian type
known as the "Loiseau rolls" and made eggettes weighing
about eight ounces, using 92% anthracite culm and 8% clay

Vol. Ill, No. 1] BRIQUETTED COAL: MALCOLMSON

as a bond. These briquets were water-proofed with a varnish
of shellac and benzine, but the cost was prohibitive. The plant
was never a success, either mechanically or commercially, and
was finally abandoned, but it marks the first step of the bri-
quetting industry in this country and had its influence on the
future, not without, we believe, beneficial results.

The Delaware and Hudson Canal Company built a similar
plant at Rondout, in 1876, which was later absorbed by the
Anthracite Fuel Company in 1878 and operated until 1880. This
plant also briquetted anthracite screenings using pitch made
from gas house tar as a binder. The third plant in the east
to use the Loiseau roll press was built at Mauch Chunk, Penn-
sylvania, and was short lived. The binder in these briquets
made a smoky fuel which disintegrated in the fire and was
otherwise unsatisfactory.

The next important plant established in the United States
was at Mahanoy City, Pennsylvania, in 1890, by the Anthra-
cite Pressed Fuel Company. The plant was designed by the
Uskside Engineering Company of Newport, England, using a
Stevens press. The briquets were rectilinear with an eagle on
one side and the word "Reading" on the other and weighed
eighteen pounds. The plant had a capacity of 400 tons per day
of ten hours. The dies were changed later to make two-pound
briquets and the capacity reduced to 300 tons. The binder was
pitch made from coke oven tar imported from England and
8% was used in making the briquets. The Philadelphia and
Reading Railroad expected to save $50,000.00 a year in their
fuel by means of this plant, but the briquets were not satis-
factory owing to the high ash content of the culm and the ex-
cessive cost of binder. The plant failed in 1892 owing to a
slump in the price of coal and inability to get sufficient quan-
tities of binder, but it is noteworthy as marking the first im-
portant attempt to make briquets for railroad purposes.

In 1892 Mr. Ware B. Gay built a plant at Gayton, near Rich-
mond, using one set of Loiseau rolls for the briquetting of
Virginia semi-anthracite slack and using coal tar pitch as a
binder. The capacity of this plant was doubled later. Similar
plants were installed at this time at Milwaukee and Chicago
for briquetting anthracite dust and bituminous slack made
at transfer plants in these cities. In the dull coal season the
Chicago plant made briquets of iron ore dust for the Illinois
Steel Company.

A more pretentious plant was built in the same year at
Huntington. Arkansas, under patents of M. Nirdlinger. con-
trolled by the National Eggette Coal Company of New Jersey.

THE ARMOUR ENGINEER [Jan., 1911

The Huntington plant made briquets of a mixture of Arkansas
semi-anthracite and bituminous coals, using hard pitch and coal
tar as a binder. These plants failed generally because of in-
experience in preparing the coal which, as a rule, was too
dirty; inability to get uniform pitch of the proper specifica-
tions ; the expense of briquetting ; and the cheapness of the coal
with which the briquets must compete. These observations
were made by Mr. Gay in referring to the Eichmond plant, to
which he added that "prismatic shape is less desirable than
one affording better combustion by forming interstices between
the pieces, especially when used for domestic purposes."
Recent American Plants.

We find the next development of the briquetting industry
in California, where more actual progress has been made than
in any other locality, although at the present time no plants
are operating. The first plant was built at Stockton, California,
to make briquets from lignite mined at Tesla. Bituminous
screenings were mixed with the lignite and asphaltum residuum
from the distillation of California petroleum used as a binder.
The press was designed by Mr. Robert Schorr of San Fran-
cisco, and combined the continuous operation of the rotary
type with the exactness and efficiency of the plunger press.
Two presses were installed having a capacity of 125 tons per
day of "boulets" weighing from 6 to 8 ounces. The Mammoth
Oil Refining Company, a subsidiary enterprise, spent consider-
able money in developing a distillation plant for making
briquetting pitch. The plant burned in 1005. These briquets
were used on the San Francisco and San Joaquin Railway
and made a satisfactory locomotive fuel, eliminating the ob-
jectionable features of raw lignite.

Another Schorr press was installed at Oakland, California,
by the "Western Fuel Company for briquetting the accumula-
tions of slack coal on its docks. The operations of this plant
were discontinued at the time of the earthquake when the price
of pitch became prohibitive.

A briquet press of novel design was built by Mr. C. R. Allen
and installed at Pittsburg, Calif., for the briquetting of lignite
mined at Somersville. This press was on the order of the roll
type, making a cylindrical briquet weighing 8 to 10 ounces.
The plant had a capacity of 5 tons per hour using asphaltic
pitch as a binder.

The Standard Coal Briquetting Company of Oakland and
the American Briquetting Company of San Francisco made un-
successful attempts to produce commercial briquets, which

Vol. Ill, No. 1] BKIQUETTED COAL: MALCOLMSON

failed on account of economic conditions already mentioned.
The latter plant experimented with Coos Bay lignite mixed
with coal yard screenings.

The Arizona Copper Company, Clifton, Arizona, is making
briquets for its own use, from the slack of sub-bituminous coal
mined at Gallup, N. M. A Yeadon press built in England is
used, making four-pound briquets of prismatic shape at the
rate of about 2% tons per hour. Asphaltic pitch is used as
a binder. The economic value is found in the storing qualities
of the fuel made from a slack that will either "fire" or at least
deteriorate rapidly when stored. Coke breeze, hitherto wasted,
has also been mixed with the slack coal.

The Washington Coal Briquetting Company of Seattle has
built a plant using a plunger type press designed by Mr. Henry
Mould of Pittsburg, constructed along the lines of his press for
briquetting flue dust and ores. This plant was completed in
1908, but up to date has not made a commercial product. It
was designed to utilize the slack from low-grade fuels sold for
domestic purposes in Seattle. A press of the Couffinhal type,
built by the Coal Briquette Machine Company of Oshkosh,
Wis., has been installed at Sheboygan to briquet anthracite
dust from the coal yards. This plant, installed in 1907, has not
yet been put in commercial operation. The briquets are cylin-
drical and weigh about 12 ounces. The press has a capacity of 4
tons per hour.

The National Pressed Fuel Company has installed a press
and plant designed by George W. Ladley and sold a limited
amount of briquets in Indianapolis last winter to domestic
trade. The press is an adaptation of the Brogneaux rotary
type, and belongs to the same class as the Schorr press, com-
bining the rotary and reciprocating types. It has a capacity
of 12 tons per hour of 6-ounce briquets, cylindrical in shape,
made from southern Indiana screenings and hard coal tar pitch.
About 8% of binder is used.

The National Fuel Briquette Machinery Company has a
small plant at the foot of Court St., Brooklyn, for demonstrating
the Devillers press. The press is of the Belgian rolls type, has a
capacity of 5 tons per hour, and makes an eggette weighing
about 2 ounces out of small-sized anthracite and coal tar pitch,
imported from Europe. The briquets are sold for domestic
purposes. One of these presses was purchased a few years
ago by the Consolidated Gas Company of New York to make
briquets from coke breeze, but the product has not yet been
marketed.

The Zwoyer Fuel Company of New York is one of the

THE ARMOUR ENGINEER [Jan., 1911

pioneers of briquetting in this country and has developed an
efficient press of the Loiseau type having a maximum capacity
of 15 tons per hour of 2-ounce briquets. The briquets are
pillow shaped, that is, rectangular in plan, but ovoid in both
cross sections. This shape is a development on the one advo-
cated by Hutteman and Spiecker and is designed to economize
the effective area of the rolls and reduce the amount of waste
in briquetting. Several plants have been built by this company
in and about New York, in the past ten years, marking the
perfection of their press and other equipment. At the present
time the only operating plant is at Perth Amboy, owned by
the New Jersey Briquetting Company. The product is loaded
mechanically in barges direct from the storage bins at the
plant, and sold in New York and Brooklyn in competition with
stove sizes of anthracite. The briquets are made from anthra-
cite screenings with 10% of hard coal tar pitch as a binder.
The most important plant using the Zwoyer press and process
is at Bankhead, Alberta, Canada, at the breaker of the Bank-
head Mines, Ltd. This plant has been operating since 1906
and has recently doubled its capacity, making during March,
1 909, over 15,000 tons of briquets. The coal used is a friable
semi-anthracite, and about 10% of coal tar pitch is used as a
binder. The output is sold principally for domestic purposes
and shipped as far east as Winnipeg.

A press of similar design is manufactured by the Mashek
Engineering Company of New York. One of these presses is
installed at a plant of the D. Grieme Coal Company, West 27th
Street, New York, making briquets of anthracite buckwheat
and coal tar pitch binder.

The United Gas Improvement Company of Philadelphia
and the Solvay Process Company of Detroit have done con-
siderable work in developing the briquet industry, as a means
of disposing of their by-products and not primarily to market
briquets. The United Gas Improvement Company purchased
and installed in 1905 a rotary press manufactured by the
Societe Nouvelle des Etablissements de L'Horme et de La
Buire, Lyon, France, and are making an eggette weighing 5
ounces. As the plant now stands they have the original press
and one of this type adapted to American conditions, making
a pillow-shaped briquet weighing 2 ounces. The presses have
a capacity of 5 tons per hour. Anthracite buckwheat and
smaller sizes are made with 10% water gas pitch into briquets
used exclusively for making water gas, and giving better re-
sults than the larger sizes of anthracite.

The Solvay plant has passed through a longer experimental

Vol. Ill, No. 1] BRIQUETTED COAL: MALCOLMSON

period beginning in 1904 with the installation of a Johnson
press similar to that used at the St. Louis plant of the govern-
ment. This press originally made 8-pound prismatic briquets;
the dies were changed to make briquets weighing 4 ounces, but
the troubles incident to feeding the dies and the reduced out-
put led the company to abandon that press and substitute one
built by Mr. Mashek, which did not prove satisfactory. The
company has recently installed a press similar to the one at
Point Breeze, made under the U. G. I. specifications, but mak-
ing 2-ounee eggettes. These briquets contain coke breeze,
Pocahontas slack and 8% hard pitch made from coke oven

Fig. 2. Johnson press at Government Coal Testing Plant,
St. Louis, Mo.

tar, and the company is now experimenting with a process
to eliminate the smoke by partially coking the briquets.

The Briquette Coal Company had an experimental plant
on Staten Island. A Couffinhal type of press, built by Schuch-
termann & Kremer, Dortmund, Germany, making 4 tons per
hour of 1%-pound rectilinear briquettes, was installed together
with a Belgian press made by H. Stevens of Charleroi, making
5-ounce eggettes at about 7 tons per hour. The plant was
never designed to operate commercially and has recently been
abandoned. The equipment is being installed in a plant near
the mines at Murphysboro, DL, to make briquets under con-

10 THE ARMOUR ENGINEER [Jan., 1911

tract with the St. Louis and Big Muddy Coal & Iron Company.

The work of the United States Geological Survey at St.
Louis is fully described in government bulletins and need only
be mentioned here. During the exposition period a Johnson
press made at Leeds, England, was installed, together with
the other equipment to make up a complete briquetting plant.
This press made 8-pound briquets and had a capacity of 7 to
8 tons per hour. A White press of the Belgian or Loiseau roll
type was also installed, but returned to the owners at the close
of the exposition. After the writer was placed in charge of
the plant, March, 1905, the die plate of the Johnson press, shown
in Fig. 2, was reduced to one-half its original thickness and
other improvements were made in order to briquet larger sam-
ples of coal, such as were subsequently used in the locomotive
road tests discussed further on. In rebuilding the plant in
February, 1906, the first operating press of the Renfrow
Briquette Machine Co. of St. Louis was installed, making a
briquet weighing approximately 8 ounces at the rate of 6 to 7
tons per hour. While this press embodied all the fundamental
principles of later presses, it could only be considered an ex-
perimental press, and briquets made were far from satisfac-
tory. The same difficulties were experienced as may be found
in the history of all briquetting presses in this country and
abroad. Insufficient pressure frequently made soft briquets
and required an excess of binder with a low melting point.

Profiting by this experience, the Renfrow Company built
a new press having a capacity of 8 to 9 tons per hour, and mak-
ing a briquet of the same shape weighing 13 ounces. This
press was installed at the Norfolk plant of the Survey and
made the briquets tested on the eastern railroads and for the
Navy Department. Upon concluding the Norfolk tests the
machine was sold to the Rock Island Coal Mining Company and
installed by the writer at Hartshorne, Okla. (See Fig. 3). The
plant has been operating since August, 1908. part of the time
on double shift, briquetting the bituminous slack mined by the
company and marketing the product for domestic purposes in
Oklahoma, Arkansas and Texas under the trade name of "Car-
bonets." This may be said to be the first plant in the Middle
West to be put on a successful commercial basis. The success
of the undertaking is largely due to the careful consideration
of the problems involved in the mechanical construction of the
plant, the binder used and the market conditions encountered.

The Western Coalette Fuel Company of Kansas City, who
used a Renfrow press, were unsuccessful because these prob-
lems were not given sufficient consideration. A still later press

Vol. Ill, No. 1] BRIQUETTED COAL: MALCOLMSON

of the Renfrow Company has been installed by the Detroit
Coalette Fuel Company to make briquets from Pocahontas coal
for domestic purposes. This plant was completed in the sum-
mer of 1909.

Kansas City is being supplied again this winter with bri-
quets by the Standard Briquette Fuel Company of St. Louis;
the plant was designed and built at Kansas City by the Roberts
and Schaefer Company of Chicago, using a Misner press. This
press is of the plunger type having a capacity of ten tons per

Fig. 3. Renfrow press making "Carbonets" at the briquetting
plant of the Rock Island Coal Mining Co., Hartshorne, Okla.

hour. Arkansas semi-anthracite and hard coke oven pitch will
be used. The briquets will be cylindrical with spherical ends
and average 14 ounces in weight.

Of the presses so far available, the maximum output has
been about ten tons — with the possible exception of the Zwoyer
press — in making briquets of 4 pounds and less in weight. If
we eliminate the binder, the cost of production varies directly
with the output. The speed of reciprocating press of the Couf-
finhal type is fixed by the time required to move the die plate.
Rotary presses, of the Loiseau type, do not make satisfactory
briquets larger than 5 to 6 ounces. Experience has shown that

THE ARMOUR ENGINEER [Jan., 1911

for railroad fuel made from bituminous coal, briquets of from
2 to 4 pounds each give best results. It is encouraging to learn
that there is being tested, at an Illinois mine near St. Louis, a
briquet machine which is a new departure from anything so
far exploited in this country, bearing some relation to the one
manufactured by Flaud et Cie of Paris. The dies are filled with
the same accuracy as in the Renfrow and other plunger presses,
and the compression is made by the positive action of plungers
with a straight line motion, but there are no reciprocating
parts> and in consequence no lost motion. This press has a
capacity of from 25 to 50 tons per hour, and is capable of
making briquets from 2 ounces to 20 pounds in weight by
changing the dies.

No effort has been made in this article to discuss plants
and briquets other than those using anthracite or bituminous
coal.

An excellent study of the treatment of Texas lignites has
been given by E. T. Dumble in a "Report on the Brown Coals
and Lignite of Texas," in which he states the earliest efforts
at briquetting were made by the Houston and Texas Central
Railway in 1877. The International Compress Company, the
American Lignite Briquette Company and the Eureka Bri-
quette Company of Texas, have been exploiting the briquetting
of lignite with binders, while the Washburn Lignite Coal Com-
pany and the Northwestern Briquet Manufacturing Company
of Minneapolis have been experimenting with the briquetting
of lignite without a binder.

Manufacturing Process and Binders.

The briquets which we shall consider are made by pul-
verizing the coal, already of the proper dryness, adding a
binder, mixing the mass thoroughly with the addition of suffi-
cient steam to melt or moisten the binder and moulding the
agglomerate in specially constructed presses.

In the briquetting process, the most expensive item of cost
is the binder, and every conceivable substance or mixture hav-
ing bonding properties has been proposed for this purpose.
Refuse containing starch and sugar, sulphite liquor, clay and
lime are among the best known. Binders soluble in water
must be water-proofed and dried before being handled, a
process, which is usually so expensive as to be prohibitive. The
inorganic binders are objectionable on account of the addi-
tional ash and clinker added to the fuel. Deodorants in the
form of compounds of chlorine are recommended to over-
come the odor from pitch and sulphur during combustion
and to reduce smoke, but their value is doubtful. Com-

Vol. Ill, No. 1] BRIQUETTED COAL : MALCOLMSON lb

pounds of manganese and other highly oxygenated com-
pounds are recommended as smoke preventives, where coal
tar pitch is the binder. But. except in special instances,
pitch alone is used which is made from tar recovered as a by-
product in the destructive distillation of coal, from by-product
coke ovens, or in carburetting water gas for illuminating pur-
poses.

Since the binder is of such importance, it is essential that
the amount be reduced to a minimum and that it be thoroughly
mixed with the coal. In American practice, the percentage
of pitch required varies from 5 to 10%, according to the
process used and the coal to be briquetted. An accuracy
within one per cent, more or less, seems reasonable from a me-
chanical standpoint, but should 8% be the amount of binder
used normally, it means 12%, more or less, in the cost, which
is of economic importance.

Briquetting Pitch.

In the fractional distillation of coal tar, a recovery of
65% pitch with 1.19 specific gravity is a fair average. On
account of the varying demands for by-product coke oven tar
in Europe, the quality is constantly changing at the different
works. In this country the lack of uniform methods and the
great variety of coals and oil used present the same difficulties
in obtaining a uniform product. Briquetting pitch should be
hard enough to be shipped in bulk in open cars and remain hard
on the hottest days. To effect this all of the lighter oils and
about 5% of* the anthracene should be extracted. In Europe
the pitch becomes soft at 75° and melts at from 100° to 120°
C. As pitch has no real melting point, the methods used in
fixing a melting point are arbitrary. Following the methods
established by the Government, and in use here, practice has
shown that a pitch with a melting point of 90° C. meets the
general requirements. Pitch should also contain as little free
carbon as possible, since this carbon or fine dust has not only
no binding property in itself but requires a bond to hold it
together in the briquet. In the distillation of coal, carried on
primarily for the manufacture of gas or coke, or both, the
time factor in the process determines the character of the tar
produced as a by-product. High heats "crack" the higher
hydrocarbons during the distillation, producing finely divided
free carbon which remains in suspension in the tar. This con-
dition is also maintained during the distillation of the tar in
making pitch. Briquets made with the hard pitch usually
sold in America today, are brittle and produce considerable
slack in handling. If a softer grade is used, the briquets are

THE ARMOUR ENGINEER

[Jan., 1911

Fig. 4. Loading briquets direct from machine to car, Government
Coal Testing Plant, Norfolk, Va. Briquets made from Poca-
hontas coal and tested on U. S. S. Connecticut.

Tig. 5. Another view of conveying belt shown in Fig. 4 showing
how briquets can be loaded mechanically without breakage.

Vol. Ill, No. 1] BRIQUBTTED COAL: MALCOLMSON 15

smoky, and have a disastrous effect on the faces and hands of
the workmen. To overcome these difficulties, experiments
were carried on at the Stockton plant and a binder produced
from California petroleum that was both hard and tough. The
Barrett Manufacturing Company has for some time recognized
the importance of a satisfactory "briquetting pitch" and is
now prepared to market a product following specifications
already suggested by the writer.

Handling and Breakage.

One of the principal problems which confronts producers
and users of coal, and particularly the railroads, is the deterior-
ation of its fuel in handling and storing. Bituminous coal
cannot be handled without breakage, which assumes a very
considerable percentage even in well designed coaling stations.
This is more noticeable in the friable, low volatile coals. Eng-
lish statistics show that with Welsh bunker coal the waste in
handling is 2 to 3%, and the breakage 20 to 30%, which aften
reaches 50% in rough weather. The cohesion of briquets made
in South Wales show 83%, against 40% for the same coal in
lump form for which the breakage was .88% for briquets and
2.13% for coal. It has been observed by a mechanical engineer
of a western railroad that the percentage of dust in handling
briquets three times should not exceed 8%.

Figs. 4 and 5 show the manner of handling briquets direct
from machine to ear, and the absence of slack in the car is
apparent. Fig. 6 shows a carload of briquets made from Po-
cahontas coal being delivered to a navy barge at Norfolk. Fig.
7 is from a photograph taken thirty seconds after the one shown
in Fig. 6, and is further evidence of the small amount of break-
age which we may expect from well-made briquets. The -drop
from bottom of car to deck of barge is about 15 feet. These
pictures further illustrate the rapidity with which this form
of fuel may be handled in coaling stations. It requires about
20 minutes* to unload a similar self-clearing car of coal at the
Norfolk and Western coaling piers at Lambert's Point, aij item
of considerable importance in bunkering a ship.

In all locomotive tests, referred to later, where the briquets
were reasonably well made, the breakage in handling was neg-
ligible; and the results of other tests made at the St. Louis
pfant bear out the European experience. The percentage of
slack in handling was approximated by a series of experiments
known as "drop tests," in which 50 pounds of briquets were
three times dropped a distance of 6V? feet on a cast-iron plate,
and the percentage of broken briquets recorded which was

THE ARMOUR ENGINEER

[Jan., 1911

Fig. 6. Briquets being loaded on Government barge at Norfolk
for test on U. S. S. Connecticut. This photograph was taken
as hoppers were opened.

Fig. 7. View taken 30 seconds later than that shown in Fig. 6,
showing rapidity with which this form of briquet can be dis-
charged from hoppered bottom cars and without breakage. It
takes 20 minutes to unload lump coal from same car.

Vol. Ill, No. 1] BRIQUETTED COAL: MALCOLMSON

retained on a 1-inch square mesh wire screen. These results
were used to check the tumbler tests, similar to those made
in Europe to determine the cohesion of the briquet. Fig. 8
shows the relative cohesion of briquets, coke and lump coal
after being subjected to the tumbler tests. It was found that
the constant jarring of the fuel on locomotive tanks created
considerable slack when the briquets were badly made. The
tumbler tests approximated the results obtained in practice.
If the briquets were well made, the cohesion was greater and
the erosion less than with the same coal in lump form. Break-
age not only produces a poorer locomotive fuel but increases
the losses due to wastage or otherwise unaccounted for.

Fig:.

8. Showing the relative cohesion of briquets, coke and lump coal,
having: been subjected to the tumbler test.

No. 1.

No. 2.

No. 3.

No. 4.

West Virginia coke.

Round briquet made of Arkansas coal.
Square briquet made of Arkansas coal.
Southern Illinois lump coal.

Storage.

In countries where labor is cheap, large prismatic briquets
are used because they can be easily stacked by hand, and
occupy less bunker space. For this reason the French Navy
specifies large briquets with an estimated bunker capacity of
51 pounds per cubic foot or 10% less than for coal. With the
increased calorific value this is of supreme importance by in-
creasing the steaming radius of the vessel. The British Ad-
miralty reports 20% increased steaming radius. The briquets
in this form, however, require twice as long to coal as with
the raw fuel. In India and the West Indies 28-pound briquets
are used because one constitutes a load for a native. It was
shown at the government tests at Norfolk that as much as

THE ARMOUR ENGINEER

[Jan., 1911

Fig. 9. This car was originally full of coal, approximately one-
half of which was taken out, briquetted with 6 per cent hinder
and returned to car. This illustration shows the reduced
hunker capacity required by briquets.

Fig 10. Carload of "Carbonets'' at Hartshorne, Okla.

Vol.111, No. 1] BRIQUETTED COAL: MALCOLMSON

20% increased space was required when prismatic briquets
were loaded without stacking. Fig. 9 illustrates the reduced
bunker capacity of briquets over that required for coal. This
car originally contained a maximum load of coal, approxi-
mately one-half of which was briquetted and returned to the
car. The briquets made at the Hartshorne plant loaded to
capacity on gondola or dump cars will weigh within 10 to

Fig. 11. Samples of briquets taken from open storage piles at the Government
Fuel Testing Plant, St. Louis, after 3 years' exposure. In each sample one
briquet was taken from surface and interior of pile to show effect of
weathering. In sample 108, 110. 118, 120 and 128 the briquets were made
with coal tar pitch binder. These briquets were made on the Johnson
press during 1904 and weigh 8 pounds.

15% as heavy as mine run coal, or about equal to egg size, as
shown in Fig 10. No difficulty was experienced in loading box
cars to capacity plus 10%.

Weathering.

It has been observed that carefully executed tests in
Europe show nearly 30% of the heating value of coal is lost
when stored in open piles, while English naval records have

THE ARMOUR ENGINEER

[Jan., 1911

Fig.

12. Briquets made from Pennsylvania coal after three months' open
storage. The broken briquets show the character of fracture.

Fig. 13. Briquets made from raw Kansas slack after three months' storage
in the open during the winter. This slack is high in sulphur and cannot
be stored without "firing."

Vol. Ill, No. 1 ] BRIQUETTED COAL : MALCOLMSON 21

mentioned that it required from 50 to 100% more stored coal
to operate vessels than when freshly mined, coal is used.

Experiments made in this country show that about 10%
depreciation may be expected from coal stored in the open
and that housing only helps the situation where the coals are
high in sulphur.

Fig. 11 represents samples of briquets taken from open
storage piles at the Government Fuel Testing Plant, St. Louis,
after three years' exposure. In each sample a briquet was
taken from surface and interior of pile to show effect of weath-
ering. In samples 108, 110, 118, 120 and 128 the briquets were
made with coal tar pitch binder. These briquets were made
on the Johnson press during 1904 and weigh 8 pounds. An-
alyses of these samples show practically no loss in calorific
value. In Fig. 12 one of the outer briquets was broken to
show the character of the fracture. These briquets show no
deterioration after three months' storage in open pile. The
briquets shown in Fig. 13 were made from a high sulphur coal
that cannot be stored without igniting from spontaneous com-
bustion, particularly if exposed to the weather.

Mr. W. H. V. Rosing, mechanical engineer of the Missouri
Pacific Railway, states that "it is our practice to store coal
during the summer months when the coal cars on the system
are not being fully utilized, and use coal from storage piles
later in the season when all the cars are required for com-
mercial use. In this manner several hundred thousand tons
are stored annually. During the summer of 1907 we lost. 14,-
400 tons of coal by spontaneous combustion alone, which
amounted to 8V2% of the total stored. In fact we can only
store coal from certain mines on the system, and this must be
stored in a certain manner to avoid loss by spontaneous com-
bustion. With the briquetted fuel we could store coal from
any of the mines without danger of spontaneous combustion,
without deterioration or loss of volatile combustibles which
occurs on the surface of the ordinary coal piles."

Mr. A. W. Gibbs, G. S. M. P., Pennsylvania Railroad,
makes the following statement in his report of briquet tests
at Altoona:

"To observe the effect on briquets of exposure to the
weather a number of the round and square briquets were
placed on the roof of the testing plant. After four months of
exposure for the round and three months for the square
briquets no change whatever from their original condition was
noticed. They appeared to be entirely impervious to moisture
and were still firm and hard.

22 THE ARMOUR ENGINEER [Jan., 1911

"The briquets were little affected by handling. They were
loaded at St. Louis in open gondola cars and shipped to Al-
toona, where they were unloaded by hand and stacked. They
were handled a third time in taking them to the firing plat-
form of the test locomotive. After these three handlings they
were still in good condition, very few were broken, and the
amount of dust and small particles was practically negligible."

Briquets Used on European Railroads.

Practically all of the European railroads use briquets and
the quantity varies from 15 to 40^ of the total coal consumed.
The briquets for railway and steamship use are prismatic in
shape. The French navy specifies 22-pound briquets. These
briquets are broken before firing, and if well made will break
into pieces without making dust. The railroads use briquets
not to exceed 11 pounds in weight, which are fired one or more
at a time by hand. Storage fuel is usually in the form of
briquets; they are carried on the tanks along with coal and
generally used to get up steam, to make up time, or over heavy
grades during the run.

The specifications to contractors furnishing briquets to the
state railroads on the continent are very rigid, particularly in
France. These specifications vary somewhat in the different
countries but are covered generally by the following items :

1st. Briquets shall be well made, sonorous, entire, with
sharp edges, breaking with a clean cut, brilliant and homoge-
neous fracture.

2d. Their cohesion shall not be less than 55% and they
shall not soften at 50° C.

3d. The briquets shall ignite easily without causing dense
black smoke, shall burn with a quick bright flame and be 'con-
sumed without disintegrating. The slag or clinker shall not
adhere to the grates or tube sheets.

4th. The briquets shall not be hygroscopic nor contain
more than 4% moisture. They shall contain between 15 and
22% volatile combustible, and not more than 11% ash. The
coal shall have been freshly mined and free from sulphur.

5th. Coal tar pitch is the only binder specified; it must
be practically odorless and limited to 10%.

6th. The briquets must be prismatic with a square base;
when specified they are from 3 to 11 pounds in weight, accord-
ing to kind of coal used, with a density of from 1.13 to 1.21.

Work of Government Plant at St. Louis.

During 1905, 1906 and 1907, over one hundred tests* were

♦Published by permission of Director, TJ. S. Geological Survey.

Vol. Ill, No. 1] BRIQUETTED COAL: MALCOLMSON 23

conducted by the government on eastern and western rail-
roads to establish the relative value of briquetted and raw
coal for locomotive use. Seventy road tests were made on the
Burlington, Rock Island. Missouri Pacific and Chicago &
Eastern Illinois Railways, and twenty tests at the Altoona
laboratory of the Pennsylvania Railroad, under the direction
of the writer, assisted by G. E. Ryder and Ralph Gait. The
co-operation of all the railroad officials was secured so that
these tests would be of value to them in comparison with other
locomotive fuel tests. An abridgment of Mr. E. D. Nelson's
report on the Pennsylvania laboratory tests has been publisbed
in Bulletin No. 363 of the Survey.

The briquets were made at the Fuel Testing plant at St,
Louis, and the details of manufacture have already been re-
ported in Bulletin No. 332 of the U. S. Geological Survey. The
object of the road tests was to discover, if possible, the prob-
lems to be encountered in the use of briquets in actual practice
and wherein this practice was affected by good or faulty manu-
facture of the fuel. It must be remembered that the best
efforts at the St. Louis plant could not produce uniformly sat-
isfactory briquets. The English machine was designed to meet
European requirements and the American machine was in an
experimental stage and made at best a product of varying
quality. The problems involved in the manufacture of briquets
from our coals have been already reported. It is one object
of this article to show their value for railroad use.

Locomotive Road Tests.

The first locomotive tests were made during the autumn
of 1905 on locomotives of the Missouri Pacific Railroad between
St. Louis and Sedalia, after having tested the burning quality
of the briquets in stationary locomotive boilers. The briquets
were made on the Johnson machine from Arkansas semi-an-
thracite slack, and were tested in comparison with the Illinois
lump coal regularly furnished for that division. The results
given in a report by W. H. V. Rosing indicate an increased
evaporation of 23% and a decreased consumption of fuel per
1,000 ton miles of 37% in favor of the briquets. The briquets
were broken in halves on this test, which created about 20%
slack.

About this time, the New York Central Lines became in-
terested in the use of briquetted coke breeze, and burned some
briquets made at the plant on the same machine. These
bricpiets were tested in freight and switching service by the
Lake Shore Railroad near Cleveland. The report of Mr. H.
F. Ball, Superintendent of Motive Power, indicates that the

THE ARMOUR ENGINEER

[Jan., 1911

Fig. 14. Shows two suburban trains on the Rock Island passing:
each other in the yards north of Englewood. The locomotive
to the right is burning briquets.

Fig.

15. Illustrates the characteristic puff of smoke lasting: 3 to 5
seconds which appears directly after firing-.

Vol.111, No. 1] BRIQUETTED COAL: MALCOLMSON 25

briquets were not satisfactory for heavy service, but had some
advantages for switching, owing to the entire absence of black
or gray smoke and very few cinders. The briquets were hard
to ignite, which, added to the high ash content of the coke,
and the size of the briquets, made it difficult to maintain a
good fire.

Briquets made from mixtures of gas-house coke and Illinois
screenings were tested in switching service by the Missouri
Pacific Railroad at St. Louis. The briquets gave better results
than the ones made from coke oven breeze. Even with those
containing 50% coal, however, there was always an interval
of time directly after firing when the steam pressure would
fall. If the engine was working this was objectionable. No
smoke was discernible even when the blower was shut off.

In June, 1906, the Rock Island Railroad became interested
in the use of briquets and 100 tons of Hartshorne, I. T., slack
were briquetted and shipped to Chicago for test. The report of
C. A. Seley compared these results with Illinois lump coal
used in freight service and indicated an average in the coal
consumption of 26.2% in favor of briquets. The observer's
notes on these tests state that "the briquets did not 'honey-
comb' the tube sheets sufficiently to give any trouble and this
slag was not as hard to remove as with Illinois coal. The
ashes from the briquets did not clinker. The nozzle could be
increased *4 incn an(* still produce a sufficient draft. This
fuel burns with an intense heat, much like coke, and the depth
of the fire is easily regulated. On arriving at Joliet, an in-
spection showed fire next to the grate which would not be the
case with coal. A slight puff of black smoke appeared only
when briquets were fired ; this almost immediately disappeared
— a desirable feature for suburban service."

Fig. 14 shows two suburban trains on the Chicago, Rock
Island and Pacific Railway passing each other in the yards
north of Englewood. They are both running at full speed.
The approaching train is burning coal.

Fig. 15 illustrates the characteristic puff of smoke which
appears directly after firing briquets and lasts three to five
seconds ; engine is on Chicago, Rock Island & Pacific suburban
service approaching Morgan Park at full speed. Fig. 16 shows
engine standing at Walden, blower off; and Fig. 17 shows
engine approaching Tracy at full speed.

The general interest awakened by these preliminary tests
warranted the government in making more extensive records,
and the co-operation of the Rock Island, Missouri Pacific, Bur-
lington and Chicago & Eastern Illinois Railroads was sought to

26 THE ARMOUR ENGINEER [Jan., 1911

Fig. 16. Shows Rook Island suburban engine standing at Walden
with blower shut off.

Fig. 17. Shows engine of same train approaching Tracy at full speed.

Vol. Ill, No. 11 BRIQUETTED COAL: MALCOLMSON 27

this end. The samples, varying in weight from 100 to 400 tons,
were shipped to St. Louis and briquetted on the English and
Renfrow machine in about equal amounts, using from 6 to 9%
of water gas pitch binder. The briquets were consigned to the
railroad and loaded through its coal chutes in the usual man-
ner in which coal is handled. Where the briquets were soft, care
was taken to handle them with coke forks in weighing onto the
tender. The water tanks were calibrated and the water meas-
ured as used. Flue gas analyses and front end and furnace
temperatures were taken during the run. Careful selected
samples of fuel, ash and cinders were shipped to St. Louis
and anaylzed. Steam pressure, feed water temperature, leak-
age, smoke, condition and thickness of fuel bed, method of
firing and draft were recorded. The same engine, and, as
nearly as possible, the same crew were furnished by the rail-
road for all tests on that road. At the end of the run the
condition of the engine was noted, and the test written up.

Comparative tests were made on lump coal from the same
mines as the slack shipped from Oklahoma, Kansas and Mis-
souri. A condensed summary of the results of these tests is
given below. A record of tests on Carterville lump and mine
run coal made by the Burlington Railroad is included for com-
parison with the tests of briquetted slack from the same district.

The foregoing report is condensed from a data sheet in
which a total of 122 observed and calculated items made up the
record of each test. Averages of all tests made on each kind
of fuel are given, as for obvious reasons it is desirable to
abridge the report. Noting the equivalent evaporation per
pound of fuel as fired, it will be observed that in nearly all
eases the rate is in favor of briquets.

In the case of the tests of Illinois coal on the Burlington
Railroad, all of the fuels are from different mines and are
therefore of comparative value only when their cost is taken
into account. Tests of Illinois coal on the Missouri Pacific
Railroad are omitted because of no comparative tests on lump
coal. The Burlington tests with Missouri coal show practically
the same results with briquets and lump coal, while the Indi-
ana coals offer the same problems in briquetting and show
the same characteristics in burning as Illinois coals.

The most representative tests, and therefore the most
accurate expression of what may be accomplished with well
made briquets, are the tests made on the Rock Island and Mis-
souri Pacific Railroad with Oklahoma and Kansas coals. The
Rock Island tests show an increase equivalent evaporation of
8% and increased boiler efficiency of about 15%, while the

©•otrcitooo

-,.

ooos

* E

— CO <M CO lO

tOr>^

tc.

OS COIN CO©

OOO

ffl
6

on a
w 0>

6
2

7.6
32.8
30.5
19.9

4 5

«o

©m

2 S

.03
■am

fflJOOtOW

oo-.oso*f

-»

•pi

o <

fiS

£ra

2c?^J— n

cor-

A g

■doa
■5 ■"

7 14

6 02
4.18
2.66
3 63

1X1 o

t, 1 r-iro — r-

COT*

tOOJOiift^f

tjO

PL.

li-ia

cN — r-osn.

00 OS

■oS

^. — lO i>-

„

^..o

1-<M"J<OSCO

g e

noiMOn

t»

ooos

£W

sssss.

■wee

E ■

z <

Nonto-

£ s

O o

I- J

•ana

.«,»._-, CO

1°

— .cor-"3to

oa g,

OT<l

-o

O -»■ -W <M

c c
Sol

"■^^S

ooco

o -
O

osr-©^r ■

* £

►J

"■n

•~-

00 35

•am

, <"oo

to to

O 8 ■

eoio-H

00 oo

OONfrt
■©-■ON"

to OS

too

eoio^n

£h

3 ■

o ■

a '.

J ■

:<Q

in ■

■

o1 ;

1.3

o.2u

6 :

cr ■
i- ;

j a ■

c
c
<

- '3

- O"

I :

•2 3
rt °
0 i_

C <b

rt a

-5

S>fe<cfl

■a

3 £ 3

1 3 ■£

c "1

3
!
3
1

ft, -a

{ft

3 0

S 1

i 1
P^5

'S
pq

Vol. Ill, No. 1] BRIQUETTED COAL: MALCOLMSON 29

Kansas briquets show 25% increased evaporation and boiler
efficiency over lump coal. The causes to which may be attributed
the variation in results of the various fuels tested are discussed
further on.
Altoona Laboratory Tests.

In a testing plant, such as is maintained by the Pennsyl-
vania Railroad at Altoona, careful regulation and accurate
comparative data are obtainable under varying conditions of
engine and boiler performance. This data is valuable in afford-
ing results which may be at least approximated under the best
road conditions in practice. The fact that they compare fav-
orably with the road tests is encouraging. The briquets used
were manufactured at the St. Louis plant from a low volatile
high grade friable coal in the form of mine run, mined in
Cambria county, Pennsylvania. About equal proportions of
square and round briquets were made with 5, 6, 7 and 8% water
gas pitch binder. The same coal in mine run form was
shipped to the Altoona plant for comparative tests. As this
coal is used by the Pennsylvania Railroad, its characteristics
as a locomotive fuel were well known. The principal objection
to its use was the large percentage of fuel lost through the
stack as fine coke, which amounted to as much as 23% when
operating under heavy load. To this may be added the front
end cinder, which greatly obstructed the draft. The coal pro-
duces less smoke than other coals used on the system and this
feature made it a valuable fuel for fast passenger and terminal
service. The object of these tests was to determine what effect
briquetting would have on these characteristics, and in addi-
tion, on boiler efficiency and capacity. Tests were made at 4
or 5 rates of combustion for each kind of briquet and the mine
run coal, starting with 30 pounds of coal per square foot of
grate surface per hour and running to the maximum capacity
of the boiler. The maximum rate with coal was 102 pounds,
against 127 pounds for briquets, an increase of 25%.

The comparison of coal and briquets at equal rates of com-
bustion, shows an average increase in boiler efficiency of about
15% and an increased equivalent evaporation of 20% in favor
of briquets for the different rates compared:
Evaporation per

sq. ft. of Heating

Equivalent Evaporation

per pound of

Surface per hour.

Raw Coal.

Briquets.

8 pounds

9.5 pounds

10.7 pounds

10 "

8.8 "

10.2 "

12 "

8.0 "

9.7 "

14 "

7.3 "

9.2 "

16 "

6.6 "

8.7 "

THE ARMOUR ENGINEER [Jan., 1911

NalsMS« |,m 40—M. NO.O • |—»J» No0 i- -,o

(Mo o 1-3S

No.Tb 7»o-«^

No.o -2-os.

No 1 -: ' o

No O -2-15

No O 2-2.©

NoO 2--2B

NpO 2-3Q No O 2-35

V .

No.o 7-40

No.o a- so

No.l 2-55

No.O 3«o^,«a

NO.O AT WOP

No.l s^o^y^

Figr. 18. Smoke observations taken every 5 minutes during test of briquets
at testing laboratory of Pennsylvania R. R. at Altoona.

Vol. Ill, No. 1] BRIQUETTED COAL : MALCOLMSON 31

In comparing the coal consumed per dynamometer horse-
power per hour, with the total power developed, the record
showed a difference of nearly 35% in favor of the briquets,
when the engine was working most efficiently. This figure
compares favorably with data obtained from road tests on the
coal consumption per ton mile.

Smoke readings were taken at stated periods during each
test and the density stated in terms of Ringlemann's charts.
Based on 5 as representing very black smoke, the average of
all readings for coal was 1.5, for round briquets 0.9 and for
square briquets 0.6, or in other words the coal made twice
as much smoke as the briquets. Fig. 18 shows a series of
photographs of the stack at Altoona, taken every 5 minutes
during the test. The average smoke record for this test was
0.3.

Importance of Physical Characteristics of Fuel.

When a sample of coal is burned in a calorimeter, all of the
combustible is consumed and the total heat value of the fuel
is given in British Thermal Units. In practice this result can
only be approximated, since there will always be a loss in the
stack gases, by radiation and in the fuel left unburned in the
refuse. The efficiency of the furnace as a heat producer and
of the boilers as a heat absorber play an important part. The
real value to the consumer is the evaporation possible in actual
practice. The purchase of coal on a B. T. U. basis is an im-
provement over the old method that "coal was coal" and the
lowest prrce made the cheapest fuel. The physical character
of the coal as delivered on the locomotive tank and its be-
havior in the fire during all conditions met with on the road is
often of more importance than its theoretical heat value. An
official of the Pennsylvania Railroad once told the writer that
his company could afford to pay the same price for a certain
coal of higher ash content and consequent lower heating value
than for a much cleaner and theoretically superior coal, because
the poorer coal made an ash which did not clinker and was
easily shaken through the grates.

Combustion.

Coal will burn only where there is sufficient air in the
presence of an ignition temperature; and the rate of combus-
tion is usually limited by the air supply and the ability to mix
it with the gases from the coal. When a lump of coal burns,
the tendency is for the gases to pass off through the lines of
least resistance, that is, from the crevices made in the coal as
it breaks up in the fire. In the ease of briquets there is no

THE ARMOUR ENGINEER

[Jan., 1911

tendency to do this, owing to their homogeneous and porous
structure.

If we examine a briquet in the process of burning, as in
Fig. 19, we find that it burns entirely from the outside. As the
volatile combustible is driven off, a layer of coke is formed
which burns to ash and falls off or is carried away by the draft.
Thus we find successive layers showing partial combustion of
the fuel while the inner part is unaffected, and the briquet
retains its identity as such until entirely consumed.

The density of the briquet is of prime importance. Harder
briquets do not break up so easily and they burn more slowly

Fig. 19. Briquets made from Hartshorne, Okla., coal showing various stages
of combustion. The smaller briquets were made on Renfrow press and
weigh 8 oz. each. The larger briquets were made on Johnson press and
weigh 4 lbs. each. No. 8 shows the interior of the briquet intact, and
outside layer of coke. The depth of the coking is shown in No. 7. In
No. 9 the briquet has been reduced nearly to ash. The briquets swell
slightly in burning and their efficiency is largely due to the uniformity
with which the gases are delivered from the surface of the briquet, and
mix with the air.

in the fire. By this means the volatile combustible is driven
off more nearly at the rate at which it can be burned with
greatest economy, and the briquets form coke during the
process of combustion even though made with an otherwise
non-coking coal. This is more essential with high volatile than
with high carbon coals.

With the harder volatile coals of Illinois, the tendency of
the lump coal to break up in the fire is less than with the more
friable low volatile coals of the Appalachian field. The east-
ern coals also produce much more slack in handling so that the
main objection to their use as a railroad fuel is the percentage
lost through the stack and the coking of the coal in a mass in
the firebox. The Arkansas and Oklahoma coals have similar

Vol. Ill, No. 1] BRIQUETTED COAL : MALCOLMSON 33

characteristics. With well made briquets from these good
coking coals, the briquets coke separately and do not run to-
gether in the fire. It was thought necessary to break the
Arkansas four-pound briquets before firing, but the same
briquets made from Hartshorne and Loydell coals were fired
whole and burned with good results. The eight-ounce briquets
did not give as loose a fire and could not be fired with such a
heavy bed when made from Illinois coal. It must be observed
in this connection, however, that the more friable coals can lie
made into better briquets with less pressure than the other
coals tested.

The whole value of the briquet is clue to its uniform
size and freedom from slack in handling. These statements
are borne out in the tests of Illinois and Missouri coals,
where the lump coal did not break up badly in handling, while
the briquets used on these tests produced at least 15% slack,
which was naturally very fine; a considerable portion being
lost through the stack. The fuel could not be "wet down" as
uniform conditions had to be maintained for test purposes.
These coals do not readily coke so that while a poorly made
briquet would hold together, if made from the eastern coal, it
would tend to disintegrate when made from these coals. It
was therefore necessary to fire with a thin bed as the fine coal
would cause heavy clink to form and cut off this draft when
a thick fire was carried. It was the usual experience that the
briquets made no objectionable clinker and the ash was finely
divided and easily shaken through the grates. This is to be
expected from the manner in which the coal is prepared before
briquetting, and from the uniform distribution of the slag
producing elements of the ash, such as iron pyrites, throughout
the briquet.
Smoke and Cinders.

The reduction in cinders and sparks by briquetting depends
on the quality of the coal as well as the density of the briquets.
Certain coals, like the Loydell coal, produce a fine scale of coke
in burning which is often loosened from the surface of the
briquet by the action of the draft and carried partially burnt
through the stack. With Hartshorne and Arkansas coals the
coking is much different in character, probably due to the
higher ash content, and these coke scales are scarcely notice-
able. The same difference was noticed in burning briquets
made from Pocahontas coal and "bone coal" picked from the
mine run coal. The latter was high in ash and the scales were
greatly reduced.

The results at Altoona -show no appreciable reduction in

;;4

THE ARMOUR ENGINEER [Jan., 1911

Fig. 20. Burning briquets made from Pocahontas coal on tugboat running
under full speed in harbor at Norfolk. Photograph taken at time of firing.

Fig. 21. Another view of tug shown in Fig. 20 burning Pocahontas coal,
under similar conditions. Photographs were taken every 15 seconds for
five minutes to cover a "firing period," and these photographs illustrate
the densest smoke observed.

Vol. Ill, No. 11 BRIQUETTED COAL: MALCOLMSON 35

the weight of cinders from briquets, but a decided reduction in
their calorific value.

During December, 1907. a test of briquets was made on the
TJ. S. S. Connecticut between Xew York and Hampton Roads.
The results were so encouraging' that more briquets were or-
dered by the Navy Department for test during the trip of the
fleet around the world. Figs. 22 and 23 are from photographs
taken by the writer as the ships passed out the Capes and illus-
trate the relative smoke producing qualities of raw and
briquetted coal. Various samples of Pocahontas and New
River coals were briquetted and tested for burning qualities
on tug boats in Hampton Roads. The boilers were fired at
intervals of five minutes, known as "firing periods," and as
nearly as possible the same furnace conditions and service
were maintained throughout the test. Photographs were taken
every fifteen seconds covering a firing period and one series
taken an hour while smoke readings were taken every fifteen
seconds throughout the test. Figs. 20 and 21 illustrate the
densest smoke observed for similar tests of briquetted and raw
coal.
Firing.

The work of the fireman is reduced by the use of briquets.
Their uniform size makes the handling easier ; it is easier to
keep up steam and only necessary to fill up the holes in the fire
without leveling. No slicing is necessary as is usual with east-
ern coals. The comparative absence of clinker, when briquets
are properly fired, is a big advantage in forcing the boiler for
heavy grades or higher speed.

Summary of Advantages of Briquetted over Raw Coal.

In general the following advantages may be claimed for
briquets made from bituminous coal over the same coal not
briquetted :

1. Comparative absence of smoke.

2. Uniformity of size and quality.

3. Less loss of fuel in ash.

4. Increased furnace and boiler efficiency.

5. Reduced consumption of fuel per ton mile.

6. More fuel can be burned per square foot of heating

surface, hence greater capacity.

7. Less slack in handling fuel.

8. Less clinker and cinders.

9. Longer life of grates.

10. Fires can be kept up for longer period without clean-
ing.

THE ARMOUR ENGINEER

[Jan., 1911

Fleet passing out the Capes, December 16, 1907. U.
Connecticut burning briquets.

Fig. 23. Remainder of fleet burning coal. All of the ships were
well under way and boilers working under similar conditions.

Vol. Ill, No. 1] BRIQUBTTBD COAL: MALCOLMSON

11. Less cleaning of tubes.

12. Less labor in firing, hence

13. Greater efficiency of fireman.

14. Less draft needed.

15. More uniform steam pressure.

16. Steam pressure can be increased more rapidly.

17. No liability to spontaneous combustion.

18. Availability for storage without deterioration.

CATALYSIS, OR CONTACT ACTION.t

BY B. B. FREUD.*

Hydrogen and oxygen can exist side by side in the same
containing vessel without any perceptible combination or re-
action taking place. Introducing a bit of platinum, in the
proper state of division, into such a mixture, however, causes
an immediate reaction, an explosively violent one. Such a
reaction is a catalytic one. The platinum is termed the cataly-
zer.

On considering the illustration just given, it is seen that
thus increasing the speed of a chemical action requires, ap-
parently, the expenditure of no energy whatever. The em-
ployment of the platinum adds nothing to the energy the hy-
drogen and oxygen originally possessed. The platinum itself
is .recovered, without being in any way altered, in its entirety
and is just as efficient after the action as before. Theoretically,
the use of the platinum costs nothing. The increased speed in
the reaction of the hydrogen and the oxygen is obtained with-
out cost. In this connection Ostwald says, "When we consider
that the acceleration of a reaction by catalysis is achieved
without consumption of energy, and so proceeds in this sense
gratis, and that in chemical industry as in all other, time is
money, we perceive that the systematic utilization of catalytic
appliances is likely to lead to the most thorough going changes
in manufacturing processes."

The phenomenon of catalysis was first recognized as such
by Berzelius in 1835. The role of the sulphuric acid in the
formation of ether from alcohol, the action of dilute acids and
of malt extract on starch, the decomposition of hydrogen perox-
ide under the influence of metals and oxides, the action of fine-
ly divided platinum in mixtures of certain gases, all led him to
the conclusion "that substances by their mere presence and not
by their affinity have the power to rouse latent affinities so that
compound substances undergo reaction and a greater electro-
chemical neutralization occurs." And this conclusion, after

|In the preparation of this report, free use has been made of Smith's General
Inorganic Chemistry ; the chapter on Catalysis in Duncan's Chemistry of
Commerce; Ostwald's article on Catalysis in Nature, Vol. 65; Stieglitz'
article in the Congress of Arts and Science. St. Louis, 1904, Vol. IV, and
Stieglitz' various articles in the American Chemical Journal, 1008, et seq.
To all of these due acknowledgment is made.

•Associate Professor of Analytical and Organic Chemistry, Armour Institute
of Technology.

Vol. III. No. 1] CATALYSIS: FREUD

making allowances for the development of the science since
1835. is correct in the light of present knowledge.

Tt must not he supposed that catalytic reactions and cata-
lytic suhstances are uncommon. Quite the contrary is the case.
In fact there seems to he no kind of chemical renction which
cannot he catalytically influenced and no chemical suhstances.
whether elements or compounds, which cannot act catalytically.
It must not be supposed, however, that a catalyst can he found
to inaugurate any possible chemical change. No reactions are
possible under the influence of catalysts that could not take
place in their absence without a breach of one of the laws of
energy. The original change may proceed, no doubt, very
slowly, as in the action of hydrogen and oxygen mentioned in
the first paragraph of this paper. So slowly may this change
proceed, that without careful quantitative measurement spe-
cially directed to the point no change at all can be observed.
The laws of energy demand that the reaction must take place
without the presence of the catalyser. They prescribe no nu-
merical value to the velocity of the change, they prescribe only
that it shall not be zero, that it shall have some finite value,
however small.

Since, then, any chemical action that can take place at all
can be catalytically influenced, this influence must have vast
consequence in technical application. And before taking up
further our inquiry into the nature and cause of catalysis. I
will mention some of the technical applications of the phenom-
enon. All of the ferments and enzymes are catalysts. Hence
all of the fermentation industries are catalytic in their nature.
Chlorine is made by the Deacon process, in which hydrochloric
acid is oxidized by air in the presence of copper chloride, the
resulting chlorine being used in various substances of com-
merce. Not only is this particular application of catalysis valu-
able in itself, but the utilization of the hydrochloric acid, a by-
product of the Le Blanc soda process, makes it the more valu-
able because thereby it saves the entire Le Blanc process from
commercial annihilation. Another catalytic application in the
soda industry is the Claus-Chance process, by which the "tank
waste" is used as a source of hydrogen sulphide, which, when
mixed with air is passed over iron oxide and changed into wa-
ter and the commercially valuable sulphur. The manufacture
of 'salt-cake" by the Hargreaves-Robinson process is also a
catalytic application. Here sulphur dioxide and air react with
common salt in the presence of copper chloride and the valuable
"salt-cake" and chlorine are produced. The greatest of all
applications of catalysis is the manufacture of sulphuric acid
by the "contact" process. Of course the original "chamber"

THE ARMOUR ENGINEER [Jan., 1911

process is also a catalytic one, but the newer "contact" pro-
cess has more of the characteristics of a catalytic action ap-
parent in it. This process stands as one of the greatest achieve-
ments of industrial application of the catalytic idea. Another
great triumph of technical chemistry, the synthesis of indigo,
is also based on a catalytic action, the oxidation of naphthalene
by sulphuric acid in the presence of mercury. It may be men-
tioned that this successful preparation of indigo on a commer-
cial scale has resulted in an agricultural revolution in India.
Then there is the oxidation of ammonia to nitric acid under the
influence of platinum black, and the oxidation of methyl alcohol
to formaldehyde and "formalin" under the influence of the
same agent. Platinum is also the catalyst in one of the re-
actions in the synthesis of vanillin, "artificial vanilla." Cop-
per compounds are used as catalysts in the manufacture of
various dyes, such as aniline black and methyl violet. Man-
ganese and lead compounds used as "dryers" in the oxidation
of linseed oil, act catalytically. Iodine, in the manufacture of
that universal organic solvent, carbon tetrachloride ; barium
carbonate, in the manufacture of acetone from acetic acid;
nickel, in the manufacture of stearic from oleic acid; lime, in
the metallurgy of lead; zinc, in the manufacture of aldehyde
from alcohol; these all are technical applications of the phe-
nomenon of catalysis.

Thus we see what a catalyst can do. How it accomplishes
these remarkable results, the mechanism of its action, this has
been the inspiration of many hypothetical assumptions, which
did little service other than to delay experimental work and so
postpone a scientific explanation of the phenomenon. In recent
years much experimental evidence, particularly of a quantita-
tive nature, has been obtained. This of course is vital to a sci-
entific explanation of the nature and mechanism of the phe-
nomenon. And since we have seen that the applications of
catalysis in the industries are so important and so extensive,
whatever will be discovered in regard to the nature of catalysis
will find immediate applications in the industries. Hence, I
may say with Ostwald that the subject has not only a chemical
interest, but that the scientific knowledge and investigation of
catalysis must have vast consequences in technical application.

To recall the specific problem to our minds, let us examine
the example given in the first paragraph. It will be remem-
bered that the hydrogen and the oxygen both remained peace-
fully inactive in each others' presence so far as we could see.
On introducing the platinum, however, the two gases immedi-
ately and explosively reacted. It is evident that the addition
of the catalyzer cannot add to the intrinsic energy contained in

Vol. Ill, No. 11 CATALYSIS: FREUD 41

the original substances, and therefore, it cannot increase their
intrinsic energies to unite. It increases the speed of the reac-
tion merely. The platinum itself is recovered unchanged, and
as efficient as ever. How this increase in speed is effected, that
is the problem. Of course the "increase in speed" is to be
taken in a negative as well as in a positive sense. For while
by far the greater percentage of catalytic reactions show a tre-
mendously increased velocity, nevertheless a few reactions are
on record which are catalytically retarded.

It may be that there is no general answer to the question,
why and how catalyzers exert their marvelous accelerating in-
fluences. Such a generalization would be possible only after a
study of a large number of individual cases. One of the most
important contributions to the subject has been made by Stieg-
litz and his co-workers, whose experimental evidence shows
exactly how the particular catalyzers in the particular reactions
they studied, accomplish their results. And there is no doubt
but that the conclusions of Stieglitz in this study can be very
largely generalized. In the endeavor to answer the question
as to how and why a catalyzer works, he studied the catalysis
of methyl acetate under the influence of water and acids. The
following are the facts. The saponification of methyl acetate
by water proceeds very slowly, according to the following
equation :

CHsCOOCH3 -f H20 -> CH,COOH + CH3OH. (1)

Acids greatly accelerate the saponification proportionately to
the concentration of the hydrogen ion used. It has been
shown, also, that the final condition of equilibrium of the
reversible reaction,

CH3COOH + CTI,OH -» CH3COOCH3 + H=0, (2)

is not appreciably altered by the catalyzer. In other words,
that the acid accelerates the velocity in either direction to the
same extent. It has also been shown that the catalyzer, the
acid or hydrogen ion, appears to act by its presence, simply;
that it appears to remain unchanged throughout the course of
the reaction. These three properties have been assumed to be
typical of all catalytic actions.

Stieglitz took the first vital and decisive step when he de-
parted from the old idea that catalytic action may be studied
only in reactions which complied with these three fundamental
principles, which for emphasis I will repeat; first, that the ac-

42 THE ARMOUR ENGINEER [Jan., 1911

celeration is proportional to the concentration of the catalytic
agent ; second, that the condition of equilibrium in a reversible
reaction must not be measurably changed by the presence of the
catalytic agent ; and third, that the catalytic agent must appear
to act by its presence simply, and not to form a compound in
quantity with any other components. These three properties
have, in the past, been assumed to be necessary and typical of
catalytic action, but, in this study, the vital fact of acceleration
(in a positive or negative sense), alone was considered charac-
teristic.

In regard to the acceleration of the velocity of saponifica-
tion of methyl acetate (equation No. 1) by acids, the most fun-
damental fact concerning acids, namely, their ability to form
salts with bases and oxides, suggested itself. The logical con-
clusion was that the methyl acetate has basic properties, and
the salt formation with acids is intimately connected with, if
not the cause of, the catalysis. Of course, the basic properties
of methyl acetate must be far too weak to permit of quantita-
tive measurement of its constants, so the class of closely related
bodies, the imido esters, whose quantitative measurement of all
necessary factors could be had, were studied. The imido esters
are esters in which the imide group (=N — H) replaces the oxy-
gen atom of the ester, as in imido methyl acetate CH3C(=NH)
OCH3. These are distinctly basic substances which form well
defined salts. The free bases are very slowly decomposed by
water, chiefly according to

CeH5C (=NH) OCH3+H20 -> C,,H5CONH2+CH3OH (3)

and yet more slowly according to

C6H5C (=NH) OCII3+H20 -^ C0Hr,COOCH3+NH3, (4)

both reactions being practically non-reversible. The addition
of hydrochloric acid greatly increases the velocity of the second
reaction (equation No. 4), which becomes almost the exclusive
one. How does the acid accelerate the reaction? That is the
question. The acid forms the hydrochloride, but as the imido
esters are weak bases, partial hydrolysis takes place and a con-
dition of equilibrium results, thus

C6H5C (NH2C1) OCH3 : H20->C6H5C (NH2OH) OCH3 • IIC1 (5)

The reaction presents, therefore, at least three possibilities,
the velocity may be proportionate to the concentration at any

Vol.III.Xo.il CATALYSIS: FREUD 43

moment of the salt, to that of the free base, or to the total sub-
stance,

(G)

dx
dt

= Ksalt • (salt),

dx
dt

= Kbase • (base),

= Ksllhstanc.e • (substance).

(7)

(8)

In order to decide between these three, it was necessary to de-
termine experimentally the actual change (X) in time (t), and
also the proportions of salt, free base, and acid present at any
moment (t). This latter is determined according to Arrhenius'
e< [nation for the solution of a hydrolyzed salt

Positive ion Kbase

= == Khydrolysls, (0)

base • H KwatPr

The constant K was determined by conductivity measurements.
and with its aid the concentrations of salt. base, and acid for
the differential equations (6, 7. and 8) were calculated. The
results show that the true course of the reaction is given by
equation (6). which alone leads to a true constant. It is there-
fore certain that hydrochloric acid which enormously increases
the velocity of saponification of the imido ester according to
equation (4), does so simply and quantitatively on account of
salt formation.

The accelerating or catalytic action of the acid is here
surely due then to salt or ion formation of a different, less sta-
ble, more reactive molecule.

This increase in the velocity of a chemical action is the
main characteristic of the phenomenon of catalysis, and we
have seen a simple explanation of it based on rigorous experi-
mental proof. Now it remained to ascertain whether the two
other important characteristics for many catalytic reactions,
namely; the fact that the catalyzing agent need not appear to
combine with anv of the reacting substances, and the fact that

44 THE ARMOUR ENGINEER [Jan., 1911

in a reversible reaction it need not measurably change the final
condition of equilibrium, are also in agreement with this con-
ception of the course of a catalytic reaction.

The following experimental evidence shows why the cata-
lyzing acid need not appear to combine with any of the react-
ing substances. It was shown that the saponification of imido
esters takes place according to

= Ksalt ■ (salt), (6)

According to Arrhenius and Walker

(salt) =K • (base) • (H), (10)

substituting,

= K' • (base) • (H). (11)

Now this is exactly the velocity for ester catalysis, substituting
"(ester)" for "(base)." This shows the connection between
the catalysis of the imido esters and that of the ordinary esters.
For if the latter could be considered to be a base, and could
form salts, its saponification could undoubtedly be due to the
saponification only of its salt or positive ion. This link in the
proof was supplied by Baeyer, who showed that esters form
well defined salts (oxonium or quadrivalent oxygen salts) with
acids, very unstable ones, but nevertheless salts: and Coehn
proved that they are electrolytes. According to this idea we
can write for the reaction,

CH3COOCH3 + H20 -> CH3COOH + CH3OH (1)

the following:

— Kgap0nincation ■ (positive ester ion) • H.,0 (12)

as was proved experimentally for the imido esters.

Vol. Ill, No. 1] CATALYSIS: FREUD 45

For the combination of methyl acetate with water to form
an oxonium base and for its ionization, we have,

OH + —

CH3COOCH3+H20 *± CH3COOCH3 ?=> CH3COOCH3+OH, (13)
H H

consequently,

-ft-base

(Positive ester ion) = • (ester) • (II) (14)

Substituting in equation (12) we have for the saponification
of methyl acetate by water,

Velocity of saponification

Kbase

(ester) • (H) • (H20). (15)

If we saponify with hydrochloric acid the reaction similarly
will lie,

CH3COOCH3 + HC1 <=±

CI OH

CH3COOCH3 + H20 <=* CH3COOCH3 + HC1 (16)
H H

According to Arrhenras, the equation for hydrolyzed solutions
of salts of weak bases with strong acids, is

-"-base

Positive ester ion = ■ (ester — y) • (H'). (17)

For an almost completely hydrolyzed salt "y" is negligible.

46 THE ARMOUR ENGINEER [Jan., 1911

Hence the velocity of saponification of methyl acetate in the
presence of hydrochloric acid becomes

-ft-base

• (ester) • (H') ■ (H.,0) (18)

Comparing the two velocity equations (18 and 15) for the
saponification in the presence of water alone, and in the pres-
ence of acid, it is seen that the velocity must in fact increase
directly proportionate to the concentration of the hydrogen ion,
since all other factors remain constant. Hence the experi-
mental results do not dispute the consequences of the theory.
And now, the last important fact, namely, that in a reversi-
ble reaction, the catalyzer need not measurably change the
final condition of equilibrium. The saponification of methyl-
acetate is a reversible reaction.

CH3COOH + CH3OH -> CH3C00CH3 + H20 (19)

The velocity of this reaction is also accelerated by the addition
of hydrochloric acid. This increased velocity under the influ-
ence of an acid must be due to minimal basic properties of
acetic acid, or methyl alcohol. Surprising as it may seem, it is
to the basic properties of acetic acid that we must look in this
instance. Other workers, Rosenheim and Euler, have shown
that acetic acid must form oxonium salts and have some basic
functions. Applying this conception to the study of the veloc-
ity of esterification, the velocity in the absence of water is,

Vesterification = Kest ' (pos. acetate ion) • (CH3OH), (20)

K base

= Kest • • (acetic acid^ • (H) • (CH3OH). (21)

In the presence of acid, the equation becomes,

xrr

Vest . hci = Kebt - - "" — ■ (acetic acid) • (H') • (CH3OH).(22)
K"

The change in the velocity of esterification is seen to be pro-
portionate to the change in concentration of the hydrogen
ion. This is in accord with the theory proposed, and is with

Vol. Ill, No. 11 CATALYSIS: FREUD 47

the experimental parts. When equilibrium is established, in
absence of acid,

"saponification *eSterif icat ion; Or \^)

-ft-base

Ksap •— - ■ (ester) • (H20) ■ (H) =
K'

-ft- base

Kest- • (acetic acid) • (CII3OII) ■ (H) (24)

and m the presence of acid,

* saponification HC1 * esterif icat ion HC1, OF {^)

-ftbase

Ksap • - — • (ester) • (H') • (H.O) =
K'

Kest : — ■ (acetic acid) • (CH3OH) • (H') (26)

In other words the addition of the catalyzing acid will not
effect the ultimate condition of equilibrium between the com-
ponents of the reaction.

In accordance with these results of Stieglitz, and his co-
workers, our views concerning catalytic action must be modi-
fied in regard to all three of the commonly assumed funda-
mental characteristics of catalytic action. These three charac-
teristics are practically true only for limiting cases, where the
amount of salt formation is too small to measure. None of
them are absolutely true under any condition.

The one vital fact, OF AN ACCELERATION DUE TO AN
INCREASE IN THE ACTIVE MASS OR CONCENTRATION
OF A REACTING COMPONENT IN A CATALYTIC ACTION
is the ONLY fundamental fact common to ALL catalytic action.

GAS CALORIMETRY.
BY C. E. BECK.*

Up to about ten years ago little was known in the United
States about gas calorimetry and its commercial possibilities.
A number of the technical institutions had gas calorimeters,
as did also several of the large gas manufacturing companies,
but its use by the latter was not intended to be of any great
commercial value. As a matter of fact there was no object in
knowing the heating value of a gas, because a candle-power
criterion was the acknowledged standard and a gas having the
required illuminating qualities usually ran high enough in heat
value to cope with all practical requirements. In Germany,
however, greater stress had been laid upon the thermal quali-
ties of a gas, it, perhaps, being due to their more advanced
commercial means of using gas. It is very evident that Ger-
many is without a peer as regards the development of the in-
ternal combustion engine, and from this it is natural to assume
that she is not inferior to anyone in gas manufacturing.. Things
progressed rather slowly in the United States until the gas
engine manifested its commercial possibilities. The gas calori-
meter at once fell into demand as gas engine manufacturers
were incited to make guarantees based on the heat value of the
gas to be used in their engines. It was entirely a matter of
heat value that regulated the rating of internal combustion
engines, and with a few exceptions gas constituents were not
considered.

It is now estimated that about 90% of all the gas manu-
factured is used for power, heating and incandescent lighting,
and it is very evident that these three factors depend entirely
upon heat value for their output. With this in view the civic
authorities in some of our states enacted laws whereby all gas
companies having a certain minimum output of gas per year
were compelled to maintain a given standard heat value of
their gas. Wisconsin was the first state to enact these meas-
ures, being followed by New York and quite a number of
other large cities. Consequently, gas calorimetry as well as
the gas calorimeter now received quite an impetus, but the
question that troubled the minds of authorities was what
calorimeter would be the most satisfactory to use. As a result
the American Gas Institute appointed a committee on calori-
metry to investigate and conduct exhaustive tests on all avail-
able instruments.

♦Class of 1911, Armour Institute of Technology. Manager, Sargent Steam
Meter Co., Chicago.

Vol. Ill, No. 1]

GAS CALORIMETRY

BECK

Gas calorimetry may be defined as the quantity of heat
generated by the complete combustion of a unit volume of gas.
The apparatus used to determine this quantity is called a
"Calorimeter," which when complete consists of a calorimeter
proper, a gas meter, governor, thermometers, weighing buckets

Jankers Calorimeter. Sectional Views and Elevation.

and scales. In this article only the water heater type of in-
strument will be dwelt upon, as it has proven to be the most
satisfactory. In its performance the heat generated by the
complete combustion of a unit quantity of gas is absorbed by
a given weight of water, thereby causing a rise in tempera-
ture. The unit in which this heat is measured is called a British
Thermal Unit in the English system, and is defined as the

THE ARMOUR ENGINEER

[Jan., 1911

quantity of heat required to raise the temperature of one pound
of water one degree Fahrenheit at 62° F.

There are but four gas calorimeters manufactured that are
worthy of mention, and only two of these have wide applica-
tion in commercial practice. The first instrument to be described
is the Junkers, a sectional and external elevation of which is

Fig. 2. Sargent Automatic Gas Calorimeter. Sectional Elevation.

shown in Fig. 1. This instrument was designed by Hugo
Junkers of Germany, and is perhaps the best known instru-
ment of its kind, although it is being supplanted by the Sar-
gent, an American made instrument, to be described later.

In Fig. 1, water at approximately room temperature enters

Vol. Ill, No. 11

GAS CALORIMETRY: BECK

the weir "A," flowing down the inlet pipe "B" to the thermom-
eter at " C, " where the temperature is taken. A quadrant valve
"D" is used to regulate the rate of flow. The water on enter-
ing the instrument flows upward against the direction of flow
of the products of combustion, which come down through the
thin copper tubes "E," being discharged at room temperature

through the flue "J," where a damper regulates the velocity
of discharge. Combustion of the gas takes place at "K, " a
Bunsen burner being used for the purpose. The water passing
upward absorbs the sensible heat liberated by the products of
combustion as well as the latent heat given up by the conden-

THE ARMOUR ENGINEER

[Jan., 1911

sation formed in the combustion of hydrogen and other hydro-
carbons. At "I" a series of baffle plates thoroughly mix the
water, which, after passing over the outlet thermometer "G,"
where the outlet temperature is taken, is discharged through
the weir "H." About the only criticism to be offered on this
instrument is that the inlet and outlet thermometers not being
on the same level make it a tiresome operation for an observer
to perform continuous runs.

Fig. 4. Complete Apparatus for Gas Calorimetry.

The Sargent Automatic Gas Calorimeter, section elevation
of which is shown in Fig. 2, was designed by C. E. Sargent
of Chicago, who, through his experience with internal com-
bustion engines, realized the application and value of a gas cal-
orimeter in commercial practice. The Sargent instrument, since
its inception, has been built in several different models, each
one better than the preceding. In the figure is shown the latest
model which is now recognized as a standard and which em-
bodies all the suggestions set forth by the Committee on Gas
Calorimetry of the American Gas Institute. In construction the
Sargent instrument does not differ widely from the Junkers.
The inlet and outlet thermometers are on the same level and
the device is equipped with an electrical automatic attach-

Vol. Ill, No. 1]

GAS CALORIMETRY: BECK

ment so that at each revolution of the gas meter needle, the
discharge water is automatically switched from one receptacle
to another. By this means continuous operations can be
performed and the personal error in switching a hose elim-
inated. The efficiency of the calorimeter is about 99.5%. As
a complete outfit the Sargent apparatus is recommended be-

Fig. 5. Boys Calorimeter. Sectional View.

cause the heat value of a gas is computed in English units
direct, by the use of Fahrenheit thermometers and decimal
scales which weigh the discharge water to .01 of a pound.
The gas meter measures 1/10 of a cubic foot per revolution
and with its integrating train has a range of from .001 to
100 cubic feet. The meter is equipped with a single "which-
way" level, has three leveling screws, a drain, filler, ther-
mometer and a U-gauge. The entire outfit is made of brass
and copper. Fig. 3 is a cut of the meter and Fig. 4 the com-
plete apparatus.

THE ARMOUR ENGINEER

[Jan.. 1911

The Boys Calorimeter, shown in Fig. 5, was designed by
C. V. Boys of London, England, and has been adopted by the
London Referees for determining the heat value of London gas.
This instrument is not very satisfactory on account of the
whole body having to be removed to light the burner "B." ' It
also has a burner that produces a luminous flame, and when
burning a gas containing a considerable quantity of hydro-

rig. 6. Simmance-Abady Calorimeter. Sectional View.

carbon, a carbon deposit is effected, indicating incomplete com-
bustion. It will be noted that the products of combustion pass
up a central flue and then down and up over a coil provided
with considerable heating surface, and through which the
water passes.

The Simmance-Abady Calorimeter, shown in Fig. 6, is
another English instrument, and differs in construction from
any yet shown. The water and products of combustion pass

Vol. Ill, No. 1] GAS CALORIMETRY: BECK 55

up and down through a series of annular chambers. The device
is lagged with wood, which has been found inferior to a metal
jacket, and the outlet thermometer comes in direct contact
with irregularly heated water, which is not churned by baffle
plates and which causes wide variations in the outlet tempera-
tures. The annular chambers and their connections cause air
trapping, and at times a very irregular flow of water. The
excessive weight of the Simmance-Abady also renders it im-
possible to detect slight changes in the heat value of a gas on
account of the heat inertia set up by the excess metal.

As a standard the Boys and the Simmance-Abady instru-
ments are not considered in the United States. The Sargent
and the Junkers are recommended, there being about 150 of
the former now in use, with the demand rapidly increasing.

Of course, there are many things about gas calorimetry and
the design of a gas calorimeter that have not been mentioned
in this article. Take for instance the effect of using water be-
low and above room temperature, the effect of humidity, rate
of combustion and of varying the temperature of the discharge
products of combustion. All of these factors are responsible
for the errors manifested in commercial results, but as their
combined effect under the most unfavorable conditions does not
cause excessive errors, corrections are usually neglected. In
calorimeter design it is advisable to use light material in order
to reduce the effect of heat inertia to a minimum. The water
of contents should be as small as possible in order to detect
slight changes in heat value quickly and accurately. The inlet
and outlet thermometer, should have no heat communication
with each other, the outlet water must be thoroughly mixed
and a uniform flow should be maintained by the use of weirs.
By all means a metal surrounded air jacket is always advisable
as the evaporation of water spilled on a wood jacket will lower
the heat value of the gas being tested.

The above suggestions as well as many others of minor
consideration have been incorporated in the Sargent and Junk-
ers instruments, and when we consider convenience of opera-
tion, accuracy and efficiency, these instruments have no equal.

ARTESIAN WATER IN THE ORIENT.
BY TENNEY S. FORD.*

The water supply for old Sidon, a town of some 10,000
people built on or very near the ruins of the ancient Phoenician
city of the same name, is piped about two miles from a river
flowing by the city on the north. The built-up portion rises
gently back from the sea to perhaps seventy-five feet above sea
level, and is crowned by an old Crusader castle, which over-
looks the city and the surrounding fields and gardens. Distri-
bution of water is made in the lower town from a standpipe
which will send water to an elevation of about sixty-five feet ;
so that the upper town is not furnished with running water.
Rising up thru the center of the stone tower is the main pipe,
which overflows inside of an open stone basin some four feet
in diameter. Holes in the rim of this basin, of equal size and
under equal head, gauge the units of supply, and these units
are sold outright at the tower for a market price, just as land
would be. From here an owner may do as he pleases with his
supply — pipe it to any place as a whole or in part, or sell any
part of it.

Such a system of course involves the use of many long
strings of small pipe and this causes the principal drawback
to its use, for in the rainy season much fine silt goes the full
length of the system, finally to settle in the small pipes and
eventually to close them. Partly to avoid this, the water is
cut off toward the end of the dry season for a week or so, while
the mains are cleaned, and at such times the people have to use
the brackish water of old wells under the houses or else pack
spring water in jars from some distance. Of course, the fact
that the river Water is largely surface drainage adds to the
danger of pollution, but modern orientals seem to have paid
less attention to that, until very recently, than even the an-
cients did.

Under these conditions the Americans in charge of the
Mission Schools sank two drilled wells to about 900 feet, and
the water in them rose from about 750 feet below sea level
to 20 feet below the surface of the school yard, which is there
at an elevation of +50 feet. This was in 1901, and for some
years a rather unique pumping system was in use. The first
45 feet was drilled thru the earth filling of the old city moat
over which the school stands, and this had been dug out around
the casing down to an elevation of +5.0 feet, where a coarse

•Class of 1906, Civil Engineer, Board of Local Improvements, City of Chicago.

Vol.111, No. 1] ARTESIAN WATER IN ORIENT : FORD

brown sandstone that underlies the whole region was found.
It was noticed that considerable amounts of water, which over-
flowed the casing during drilling, escaped quite freely thru
this sandstone. Using this fact, the idea was hit upon of join-
ing the two wells (about 90 feet apart) with the drive pipes of
two hydraulic rams, these pipes taking the water from one
well and delivering it to the rams set on a platform in the
other well, this well receiving the overflow from the waste
valves.

For a long while the porous sandstone carried off this
waste, but finally dust and dirt sifted in to such an extent as
to clog the pores so that the larger ram, and then the smaller
one, had to be stopped. This was no small hardship to both
the school and the neighbors, for many had made good use
of the little stream kept flowing outside the school compound
in a sort of public fountain. Hand pumps were put in, but
their use by townspeople during school hours was a great an-
noyance, and the forcing of water from the boys' school, where
the wells were, thru the town some thousand or eleven hundred
feet to the girls' school, to supply about one hundred people,
was a tiresome job.

During a year spent in 1909-1910 assisting in various proj-
ects connected with the schools and the Industrial Farm be-
longing to them, the writer helped to remedy the conditions
outlined above. An underground pump room was dug out
around the well, walled up, and arched over so as to leave the
boys' yard undisturbed except for an entrance way. In this
was installed a ten-inch Rider-Ericsson hot air engine, chosen
because of its simple build and action as perhaps the best
adapted to such a distance from repair facilities, and to opera-
tion by a native workman. A system of distribution pipes and
tanks among the various school buildings was arranged, not
without some difficulty owing to the native architecture, for

Sidon as Viewed from Seminary Compound.

THE ARMOUR ENGINEER

[Jan., 1911

when those innumerable corners had to be turned and the
heavy walls pierced for such modern innovations as bath-room
plumbing, it was no easy matter. However, the real experience
of the job, amusing enough in the narration, but far. from
amusing at the time, was the tearing up of the old piping
between the two schools and the laying of new and larger
pipe, by another route.

In the accompanying sketch of a typical cross section of
a Sidon street can be seen the general arrangement of paving,
drains and sewers, and water pipes when there are any. It may

Bc^ild

Typical Cross

The F\rmo<it
!•<•( ion of Street.

be apparent that, except for crowding, the system is essentially
that in any ordinary city, but it was this crowding which
caused all our troubles, or most of them. One of the old lines
was of 1^4** pipe which had been in use for the past twenty-five
years to carry one-half of the city water belonging to both
schools over to the girls' school. It was found about three-
fourths full of fine river silt, but, considering its age, not badly
corroded. The other, a 1-inch pipe used with the pump and ram
for artesian water, was so badly eaten on the inside after only
eight or nine years, that a man's little finger would scarcely en-

Vol. Ill, No. 1] ARTESIAN WATER IN ORIENT: FORD 59

ter the end. In the recent purchase of some property, the Mis-
sion acquired another unit of city water for use at the girls'
school, and this was fed from a sort of sub-station along the line
of our work, to which we made connections. These smaller
standpipes are scattered all over the town and are fed from the
main tower in larger pipes than individuals would use, thus
saving expense and friction, both of which facts even the
oriental has found out.

For the supply of artesian water to the other school we
used a lot of 3-inch pipe, already on hand, and this we connect-
ed to the discharge main from the air engine. Right here our
troubles began. Often we came to some old house drain almost
falling to pieces and built so high that between the cover flags
of the drain and the paving flags there was no room for the
pipe ; then if the drain fell apart, it had to be rebuilt. At some
other time a slight bend would occur in a narrow part of the
street, too small to use any angle fitting we could possibly pro-
cure, and since our pipe supply was limited, and even standard
fittings were scarce articles, we were forced to bend our pipes.
A pipe sixteen feet long, by the way, with a bend near the
fixed end, was not the easiest thing in the world to connect in
cramped quarters. Add to this the fact that the paver was
replacing the flags only two or three joints back, so as to keep
as little of the street open as possible, and that every lifting of
the end of the pipe jarred his work and scraped off the tar
that had been used to protect the pipes as the only means of
waterproofing at hand. Then, perhaps, in getting over or
under some other man's line of pipe we either broke it and had
to stop to repair it or else we worked too close to some loose
sewer stones and broke through them for another job of re-
pairing.

These were some of the mechanical difficulties, but what-
ever fate it was that was amusing itself at our expense, it had
other methods than these. Imagine the streets shown in the
photographs (the weather prevented taking shots of the actual
scene of the work) to be lined with the little eight-by-ten shops
of an oriental market quarter, with their little display stands
on the pavement beside them and with the buyers grouped
around, joined perhaps by curious children and not a few
curious older people, commenting pleasantly or sarcastically
on "the way these Americans do things." Imagine, too. the
weather to be the Mediterranean winter with the sort of cold
rain that Ave have here in October or March, gradually making
mud out of your little 15-inch spoil bank for the people and
the donkeys to spread in a slippery layer over the cobblestones !

THE ARMOUR ENGINEER

[Jan., 1911

At one time, when the rain was heavy and had blocked work
for five or six days, with quite a stretch of pavement in bad
shape, a young Syrian workman remarked that we'd better
be getting something done pretty soon, "For," said he, in his
expressive Arabic, "those shop-keepers are already cursing us
with curses each one so long," and he held his arm out-
stretched.

View of a Sidon Street.

It was with considerable relief that we made the final con-
nection to the tanks in the other school, opened everything
wide and watched a good stream slip out the end of the pipe.
There was some anxiety even then, however, for the new string
of pipe for city water wouldn't drip a drop! We found that
the water had been shut off for a day or so because of the

Vol. Ill, No. 1] ARTESIAN WATER IN ORIENT : FORD 61

heavy rains that muddied the river. So we waited till it was
turned on and still no flow. By opening a plug at a low point
we found that the pipe carried water but could not get it at
the usual height of delivery. After trying various means we
made a long plunger with a light rod and two or three leathers
and some nuts, and worked it up and down in the inlet at the
stand-pipe till we had pumped the air bubble, which was evi-
dently causing the trouble, out of the way and got a flow.

Of course the rams had been at work some years before
this and there were perhaps one or two others in the country,
but except for the regular English water-works plant in the
large city of Beirut, some twenty-five miles away, this little air-
engine constituted the only intra-urban, artificial power water
works in all that part of the country. The ram, too, described
at the outset, though in very common use, would seem, so far
as the writer's knowledge goes, to be quite unique in having
its source, its active parts, and its waste water run-off all below
the surface of the ground from twenty to forty feet.

Innumerable incidents of both technical and general inter-
est could be told out of the year's experience and of the experi-
ence of three previous years, but those given will afford a
glimpse of one feature of this curious, new-old country, where
American educational effort has wrought a peculiar combina-
tion of the habits and speech of two thousand years ago with
the newer thought and industrial methods of the West.

THE HIGH-TEMPERATURE ELECTRIC RESISTANCE
FURNACE.

BY H. RALPH BADGER.*

In three important factors the electric resistance furnaces
are fundamentally superior for most industrial operations that
come within their scope ; first, in the quality of heat they pro-
duce ; second, in the distribution of heat throughout the work-
ing chamber that is possible with them ; and third, in the means
of temperature regulation they possess.

As to the first, "electric" heat, as developed by electrical
resistance, is in effect quite different from the heat produced
from such other sources of energy as coal, oil or gas, in that
it is not the result of the process of combustion. With all of
these latter sources another element, the "supporter of combus-
tion," is of course absolutely essential. The very presence of
this oxidizing agent is for most purposes detrimental to the
quality of the finished product in hand. Such a depreciating
condition is entirely unnecessary to the development of even
high temperatures by electrical resistance, where no chemical
changes are required whatever.

That a more desirable distribution of heat, whether uni-
form or definitely varied, is possible throughout a working
chamber heated by means of electrical resistance than in those
heated by combustion, of any fuel, is largely due to a me-
chanical advantage. In the former, the heat is generated in
the actual material of the resistor, while in the latter it depends
upon a process. The material of an electrical resistor can in
general be arranged to better advantage in the structure of a
furnace, for producing the desired heat distribution, than can
the mechanical equipment necessary to carrying out the pro-
cess of combustion (regardless of the fuel).

The regulation of temperature in the electric furnace is
dependent practically upon but one thing, i. e.^the energy sup-
plied. While this is mechanically not as simple to control as
in "combustion" furnaces, it at the same time can be more
uniformly varied. Again, these latter furnaces are- compli-
cated by further conditions, viz., the 'supply of the "supporter
of combustion" must be controlled, as this too influences the
temperature, and, disposal must be made of the "products of
combustion" resulting from their operation.

Recognizing from these fundamental principles, the ad-
vantages of electrical operation, a line of Electric Resistance

"Class of 19(17.

With the Hosklns Manufacturing Company, Detroit, Mich.

Vol. Ill, No. 1] ELECTRIC FURNACES: BADGER 63

Furnaces has been developed for application chiefly to the
industrial operations involving the heating of metals and
metallic parts to high temperatures.

Obviously the material of the "resistor" adapted for such
furnaces must successfully resist the action of the temperatures
it is desired to produce. Graphite or carbon was chosen, hav-
ing the desirable properties of reasonable mechanical strength
and comparative low cost, in addition to its very high refrac-
tory powers. As heat is produced in these furnaces by supply-
ing electric energy to their resistors, their temperatures may
be altered by varying the quantity of energy so supplied. With
a constant voltage this is accomplished naturally by varying
the resistance of the working circuit, which is carried out by
having the resistor made up of a number of carbon strips.
Increasing the mechanical pressure on these, that is, by forcing
them closer together, their resistance as a circuit is lessened
and they draw more current, with resulting rise in tempera-
ture. Lessening the pressure on the strips, the reverse takes
place. This is the familiar principle of the carbon-plate rheo-
stat, thouerh in the furnace carried to a degree hardly sug-
gested by the other.

A point to be noted is that in the furnace itself the means
of temperature variation is completely within the working
chamber, so that the change comes exactly where the resulting
effect is desired. This is an important efficiency consideration
in each of the following designs of the Electric Resistance Fur-
nace produced under the patents of Albert L. Marsh by the
Hoskins Manufacturing Company.
Crucible Design Furnace

The chamber walls constitute the heating unit and consist
of two series of carbon plates in contact ("a" and "a," Fig. 1)
and the graphite end-blocks "b." These are under a slight
longitudinal pressure from the electrodes "c," which are also
of graphite. Under even "ten-hour-a-day" operation a set of
carbon plates will last over a week, the end-blocks two weeks
to 15 days, and the electrodes a month or more. Renewal of
these parts is obviously very simple.

At the terminals "e" the low-voltage current is supplied,
this being conducted directly to the electrodes through water-
cooled blocks "f :" "g" indicating the inlet and outlet for the
running water. The hand screws "d" serve as a fine means
of regulation of the energy supplied by varying the resistance
— and hence the resulting temperature in the chamber.

A thick, highly refractory insulation surrounds the heat-
ing-unit walls and separates them from the outer steel casing.

THE ARMOUR ENGINEER

[Jan., 1911

This insulation is both to withstand the high temperatures pro-
duced and to conserve within the chamber the maximum
amount of heat generated.

The various sizes made of this design operate on from 10
to 30 volts, alternating current being preferable so that these
potentials may be obtained by transformers. In a furnace of

Fig. 1. Crucible Design of Electric Resistance Furnace.

10"xl0"xl0" chamber dimensions, a pressure of 30 volts is used.
In this size a maximum of 45 k. w. of energy will run the cham-
ber temperature from cold to about 2700° F. in an hour's time.
At this point the carbons are of course incandescent. Once at-
tained, such a temperature may be held constant while drawing
but 50% of its initial energy. For such a furnace the regula-

Vol. Ill, No. 1] ELECTRIC FURNACES: BADGER

tion at any time may be considered as about 60% to 75% of the
maximum power consumption.

Temperatures up to 3600° F. may be obtained in these fur-
naces and due to this high range they are used for melting
platinum. In one of the government mints they are applied
to the melting down of gold and silver. Other metals such as
nickel, iron, copper, brass, steels and various alloys are, in

Fig. 2. Crucible Furnace Showing Transformer.

crucible charges, melted in them. They may be used as well
for heating barium chloride in steel treating, for heating
cyanide and lead baths and for "fire" assay work.

Muffle Design Furnace

Exactly the same principles applied to the "Muffle" form
of furnace are illustrated by Fig. 3. In this the carbon plates
are in a horizontal position, on either side of the chamber, a
graphite block connecting the two series across the top or roof
of the chamber. A vertical regulating pressure is applied by

r;«i

THE ARMOUR ENGINEER

[Jan., 1911

the handscrews which may be seen underneath the body of the
furnace.

In this furnace, which is considerably more enclosed than
the crucible form, the safe operating temperatures are prac-
tically limited to 2500° F., though this is quite high enough for
the principal use for which it has been designed, namely, the
hardening of steel tools and parts.

Fig.

Muffle Design of Electric Resistance Furnace.

Almost an entirely new understanding is at present work-
ing out in a practical way on the subject of heat treatment of
various steels. Scientific methods, in place of the customary
shop-room guess work, are fast being developed and applied to
this very important branch of machine and tool production.

Vol. Ill, No. 1] ELECTRIC FURNACES: BADGER

Foremost in this movement is the use of electricity, both to
produce the required heat by means of resistance, and to meas-
ure its intensity through application of the thermo-couple.

Besides the fundamental advantages of electrical operation
already outlined, the electric furnace, shown in Fig. 3 — and
again as part of the installation in Fig. 4 — possesses inherently
the desirable quality of an actually reducing chamber atmos-
phere while operating. This is brought about by the great
affinity the incandescent carbon walls exert for the slight
amount of oxygen that is allowed to enter while raising the
door.

Fig. 4.

Electric Furnace for Tool Hardening in the Plant of the Link-Belt Co.
Chicago.

A working installation is shown in Fig. 4. In this corner
of the hardening room of the Link-Belt Company's Plant, the
electric furnace is seen to the left of the center; a pre-heating
furnace being on its left and an, oil bath to the right. The
switchboard contains, besides the necessary controlling appa-
ratus for the motor-generator set supplying the current, a
pyrometer meter indicating the temperatures of the furnaces,
in each of which a thermo-couple is installed.

THE ARMOUR ENGINEER

[Jan., 1011

Tube Design Furnace

A third design of this type of furnace is shown by Figs. 5
and 6. In this the heating unit is composed of* carbon plates in
the form of rings "a," which terminate with graphite blocks,
"b. " The pressure-regulating hand screws are indicated by
"d;" "e" being the electrical terminals and "g" the water-
cooling connections.

Tube Design of Electric Resistance Furnace.

Again, due to the enclosed structure, the maximum safe
operating temperature of this furnace is about 2500° F. It is
used particularly for heat-treating special steel parts such as

Fig. 6. Tube Furnace Showing Chamber Opening and Adjusting Hand-Screws.

Vol. ITT, No. lj ELECTRIC FURNACES: BADGER

drills, taps and dies; and also for annealing metal tools and
rods.

The objective production of heat, as a branch of the elec-
trical industry, while later in line of commercial development
than cither that of light or dynamic power, nevertheless is
growing rapidly in importance. The fundamental advantages
of electricity as a form of working energy are commanding for
it the same considerations that are now so widely recognized in
the other branches of the art.

THE MEASUREMENT OF TEMPERATURE.
BY JAMES CLINTON PEEBLES, E. E *

It is the purpose of this paper to discuss some of the de-
vices which are in use for the measurement of temperature,
with special reference to the measurement of comparatively
high temperatures, from about 500° F. up. Also, the errors
to which such instruments are subject will be pointed out, and
so far as possible, methods of avoiding these errors will be
indicated.

The Mercurial Thermometer

The instrument in most general use for the measurement
of temperature is the mercury-in-glass thermometer. Other
liquids than mercury are sometimes used in a glass thermom-
eter, but the principle is the same. When such an instrument
is used at a temperature of 500° F., or higher, it becomes sub-
ject to errors which are often considerable, and which the
simplicity of the instrument tends to conceal.

The most common source of error in such a thermometer
is boiling of the mercury, which may occur at as low a tem-
perature as 300° F. This produces a vaporization of a portion
of the mercury, with a subsequent recondensation in the upper,
cooler part of the capillary tube. This trouble can be over-
come by introducing an inert gas, such as nitrogen, into the
capillary tube above the mercury. As the mercury rises in
the tube upon an increase in temperature, the gas pressure
is increased and the boiling of the mercury thus prevented.
This precaution is observed in the higher grade thermometers,
but instruments are on the market which are subject to a
considerable error from the boiling of the mercury. When
using a thermometer which has not been filled with nitrogen,
a good precaution to observe is to keep the top of the mercury
column as cool as possible, and so prevent boiling, or to keep
the whole stem hot, which prevents condensation of the mer-
cury. This precaution will be effective up to 350° to 400° F.,
but for higher temperatures a nitrogen rilled thermometer
should always be used.

After a thermometer has been made and calibrated it may
undergo certain changes which will very seriously affect its
accuracy. The most important changes, and the only one
which need be considered here, is a permanent contraction of

♦Class of 1904, Instructor in Experimental Engineering, Armour Institute of
Technology.

Vol. Ill, No. 1] TEMPERATURE MEASUREMENT : PEEBLES 71

the glass which renders all the readings of the instrument high.
This is caused by a slow readjustment of the internal stresses
in the glass which were produced when the stem was made.
Tests made by the writer have shown thermometers to read
as much as 60° high at a temperature of 700° F., due to the
contraction of the glass. Errors of this magnitude were found
in supposedly high grade instruments, much above the average
thermometer to be found in the market. It is quite probable
that many thermometers in general use, the indications of
which are accepted as correct by the users, would be found
on test to be subject to errors even greater than this.

In the manufacture of the best thermometers the glass
stem is annealed before the scale is etched on. This is done
by heating the stem at a temperature somewhat above the
highest at which it is to be used, and maintaining it at that
temperature from 5 to 10 days. It should then be cooled very
slowly, the cooling lasting from 4 to 6 days. This removes all
the strains from the glass and produces a thermometer which
will not change with age. In buying a thermometer for use
in work where reliable indications are essential, only one
which has been properly annealed in the making should be
considered. Many thermometers which have not been thus
artificially aged are in the market.

In perhaps the majority of cases where a thermometer is
used, the bulb is placed in contact with the object of which
the temperature is desired, while a considerable portion of
the stem emerges into a very different temperature. For
example: Suppose an experimenter wishes to determine the
burning point of a sample of gas cylinder oil. The bulb
and perhaps a small portion of the stem are immersed in the
oil at a temperature of say, 650° F. The greater portion of
the stem emerges into the air above the oil bath, the average
temperature of which may not be much above 100° P. Most
high grade thermometers are calibrated under a condition of
total immersion, and are correct for that condition only. When
only the bulb is immersed and the stem emerges into the air
at a much lower temperature, the indications of the instru-
ment will be considerably in error. The amount of the error
will depend upon the difference in temperature between
immersed and emergent parts, the number of degrees on the
emergent stem, and the glass of which the stem is made.

Stem correction = Kn (T°— 1°), where K is a constant
depending on the glass, n is the number of degrees on emergent
stem; T° is the temperature of the immersed part, and t° the
temperature of the emergent part. The value of K must be

72 THE ARMOUR ENGINEER [Jan., 1911

determined experimentally for each thermometer, as it differs
for different glass. This can be done by comparing the read-
ing of the thermometer when exposed to a given temperature
under a condition of total immersion with the reading under
a condition of partial immersion. This gives all the factors
in the above equation except K, which may therefore be calcu-
lated.

In the thermometer mentioned above in connection with
the test of cylinder oil, the value of K is .00009. The oil is
at a temperature of 650° F., and the air above the bath at
100° F. Hence T°— 1°=650°— 100°=550°. Assume that the
stem is immersed in the oil to the 50° point. The stem correc-
tion is

0.00009 (650—50) (650—100) = 29.7°.

When the stem is colder than the bulb the stem correc-
tion must be added to the observed reading. Hence, the correct
burning point of the oil is 650° + 29.7° = 679.7°. When the
stem is hotter than the bulb the stem correction should be
subtracted.

A reliable mercury-in-glass thermometer should be well
annealed to prevent slow contraction of bulb and stem with
age; the upper part of the capillary tube should be filled with
nitrogen to prevent boiling of the mercury; and the stem cor-
rection should be known.

Resistance Thermometer.

The practical limit of a mercurial thermometer is from
800° F. to 900° F. Above this it is very difficult to obtain
reliable results with a mercurial thermometer, and so some
other method becomes necessary. The electrical resistance of
the metals is known to change with temperature, and since
electrical resistance can be measured with considerable ac-
curacy this furnishes one of the most reliable and accurate
methods for the measurement of temperature.

Platinum is practically the only metal which has come
into general use for this purpose on account of its high melting
point and resistance to the attack of gases at high tempera-
tures. The first electrical resistance thermometer was de-
signed by Sir William Siemens, and was later improved and
perfected by Callender and Griffiths. Siemens' instrument was
made by winding fine platinum wire on a fire clay cylinder
and surrounding it with a protecting tube of porcelain or
quartz. This thermometer was found to be sluggish in its

Vol. III. No. 11 TEMPERATURE MEASUREMENT: PEEBLES 73

action, requiring considerable time to come to the temperature
of the surrounding medium. It gave very good results, how-
ever, where the temperature was nearly constant, as would
he the case in an annealing furnace.

In Callender's instrument the platinum wire was wound
on a strip of mica and surrounded with a steel tube. The
porcelain tube is very fragile and was found to break with
the slightest blow when hot. It is also very likely to crack
when exposed to sudden changes in temperature, and hence
cannot be used with success in a metal bath. The Callender
instrument, with steel protecting tube, was found to be quite
sensitive and extremely accurate. In fact the resistance ther-
mometer is probably without doubt the most accurate instru-
ment that we have for the measurement of temperature. It is
a matter of record that a thermometer of this type, designed
and constructed by the United States Bureau of Standards,
reached the temperature of the surrounding medium to within
1/1000 of a degree Centigrade in three seconds.

Fig. 1. Whipple Indicator with Resistance Thermometer.

74 THE ARMOUR ENGINEER [Jan., 1911

The measurement of the resistance of the platinum coil
is an important point in resistance thermometry. The Wheat-
stone bridge method is the one usually employed, involving
the use of a galvanometer, an adjustable resistance and a
battery. The operation consists in adjusting the variable
resistance in one branch of the bridge until a balance is ob-
tained, indicated by a zero reading of the galvanometer when
its circuit is closed. A certain amount of manipulation is
therefore always necessary to secure a temperature reading
with a thermometer of this kind. Hence a resistance ther-
mometer is not a directly indicating instrument, unless the
necessary manipulation is done automatically.

A well known form of instrument for use with a resistance
thermometer is a Whipple Indicator. This instrument is shown
in Fig. 1, connected to the thermometer and ready for use.
It consists of a Wheatstone bridge, battery and galvanometer,
contained in a single case as shown. The adjustable resistance
consists of a coil of wire wound upon a drum which is revolved
by hand until a balance is obtained. Since the temperature
sought depends upon the resistance of the platinum coil in the
thermometer, and since this in turn is equal to or proportional
to the adjustable resistance on the drum, at the time a balance
is obtained, it follows that the temperature scale can be placed
directly on this drum. Thus it is that instead of reading the
resistance which has been wound upon the drum we read the
temperature directly.

Of course certain known points must be located on this
temperature scale, in order to make possible the step from
resistance to temperature. The freezing points of certain of
the metals are known with a considerable degree of accuracy,
and this supplies a convenient and reliable method for cali-
brating an instrument of this kind.

Resistance thermometers are made in various sizes and
lengths, up to one inch in diameter and about thirty inches in
length. The platinum resistance coil usually occupies not
more than four inches in the lower end of the protecting tube,
from which platinum leads are run to binding posts on the
boxwood head at the other end. It is important that no
change in the resistance of these leads, due to temperature
changes, should affect the measurement of the resistance of
the thermometer coil. For this reason two compensating leads
of the same size and length as the true leads, are placed side
by side with the latter, and connected to two separate binding
posts in the boxwood head. "Any change in the resistance of
the true leads is balanced by an equal change in the com-

Vol. Ill, No. 1] TEMPERATURE MEASUREMENT : PEEBLES

pensating leads. All that remains to be done is to connect the
compensating leads to the adjustable resistance of the Wheat-
stone bridge. Thus any change in the resistance of these leads
simply adds to or subtracts from the adjustable resistance in
exactly the same magnitude as a change in the resistance of

Callender Recorder.

the true leads affects the resistance of the thermometer coil.
Since the adjustable resistance must be balanced against
the thermometer coil when a reading is obtained, it follows
that the changes in lead resistance are eliminated. Thus an
instrument of this kind becomes independent of the depth of
immersion, a very important point in high temperature ther-
mometry.

76 THE ARMOUR ENGINEER [Jan., 1911

A very excellent instrument for use with a resistance
thermometer is the Callender Recorder, shown in Fig. 2. This
instrument gives a continuous record of temperature, the
chart covering a period of twenty-four hours. In this recorder
the adjustments of the Wheatstone bridge are made auto-
matically by means of magnets and clock work. Two magnets
are made use of, one operating when the current through the
bridge is in one direction and the other when the current is
in the opposite direction. The adjustable resistance is operated
by the clock work, the magnets simply serving to release a
brake which holds the clock in check. The clock operates the
adjustable resistance in one direction or the other, according
to which magnet has operated. As soon as a balance is ob-
tained the current ceases to flow through the magnet, which
immediately lets go of the brake and stops the clock motion.

Thus, all the manipulation necessary for a measurement
of resistance is 'done automatically and the instrument will
give a continuous record of temperature.

Up to about 2200° F., the platinum resistance thermometer
gives the most accurate measurement of temperature that we
have. The only objections to it are a slight change in the
resistance of the platinum with time, and the fragile character
of a porcelain or quartz protecting sheath.

Thermoelectric Thermometer

When two dissimilar metals are fused together and the
junction heated, the latter becomes a source of electromotive
force. The magnitude of this electromotive force is propor-
tional to the temperature to which the junction is raised. This
fact offers a simple and convenient method for the measurement
of comparatively high temperatures.

Such a junction of two different metals is known as a
thermo-electric couple, and much study and investigation have
been devoted by physicists to the thermo-electric measurement
of temperature. The credit for finally placing thermo-electric
pyrometry on a satisfactory basis belongs to LeChatelier. He
made a couple consisting of one wire of pure platinum and the
other an alloy of 90% platinum and 10% rhodium. This is
known as the LeChatelier couple and is the one in general use
at the present time.

In the commercial application of this principle, the two
wires forming the couple are first fused together, and then
are insulated from each other throughout their length by
winding asbestos thread upon them. Each wire is then run

Vol. Ill, No. 1] TEMPERATURE MEASUREMENT: PEEBLES 77

through a small porcelain tube, the tubes extending almost to
the junction. The whole is then covered by a large porcelain
or quartz tube, and the two free ends of the wires led to
binding posts on the wooden handle or socket to which the
enclosing tube is fastened. A millivoltmeter graduated to read
temperature in degrees completes the apparatus.

The magnitude of the electromotive force produced by
such a couple depends, not upon the absolute temperature of
the junction, but rather upon the difference in temperature
between the junction and the other ends of the wires, where
they are connected to the external leads. Hence, we have the
terms "hot junction" and "cold junction" to designate the
different ends of the wires forming the couple or "element."
It is important, therefore, that the cold junction be kept at
a constant temperature while the thermo-couple is in use.
Neglect of this precaution may lead to considerable error in
the indications of the instrument.

In addition to its simplicity and ease in handling, the
thermo-electric pyrometer has the advantage of a very small
time lag. It comes quickly to the temperature of the medium
in which it is placed, and hence is suitable for measuring
changing temperatures. In this particular it is superior to the
resistance thermometer, but is not capable of such great
accuracy as the latter instrument.

The sensibility of a platinum-rhodium thermo-couple
diminishes rapidly below 500° F., and hence, it is not suited
for measuring comparatively low temperatures. In the range
between 300° F. and 900° F., the best results are obtained from
the use of a couple of copper and constantan or iron and Con-
stanta n. Such a couple gives a much greater electromotive
force in this range than can be obtained from a platinum-rho-
dium couple, and hence is more satisfactory.

All metals disintegrate more or less when exposed to high
temperatures for a considerable length of time. This dis-
integration of the metal forming a thermo-couple is also accom-
panied by a loss in electromotive force, and hence after long
exposure to a high temperature, the indications of such a
couple are likely to be somewhat in error. If platinum be kept
at a dull red heat (about 1800° F.) for eight hours it will
suffer a loss of about Yi% in electromotive force. Continued
heating will not increase this loss materially, and when it is
considered that all other metals suffer a much greater loss, it
is easily seen that platinum is by far the best metal for a
thermo-couple.

78 THE ARMOUR ENGINEER [Jan., 1911

Optical Pyrometers

There are many industrial processes carried on at tempera-
tures where platinum either disintegrates rapidly or fuses.
Such temperatures are to be found in the electric furnace,
which now has a wide commercial application. Manifestly
none of the temperature measuring devices discussed thus far,
as electric resistance and thermo-electric pyrometer, are suit-
able for use with such high temperatures.

For some time it had been the custom to estimate these
temperatures by the trained eye of the experienced workman.
But this method was only approximately correct at best for
the same eye may vary considerably in the estimation of color.
This crude method, however, furnished the clue to the discovery
of a much more accurate and scientific method, whereby the
temperature of a hot body is measured by the intensity of the
light which it radiates. An instrument for measuring tempera-
ture by means of light radiation is known as an optical pyro-
meter.

The principle upon which optical pyrometry is based is
known as the Stefan-Boltzmann radiation law. These two
physicists, after much study and research, succeeded in estab-
lishing the physical law that the total energy of radiation from
a hot body is proportional to the fourth power of its absolute
temperature. The research from which this law was deduced
is discussed by Waidner and Burgess in Bulletin No. 2 of the
United States Bureau of Standards.

It will be evident that if it is possible to measure the total
energy of radiation with a fair degree of accuracy, we imme-
diately have a very accurate measure of temperature, because
the latter is proportional to the fourth root of the energy of
radiation. Thus a considerable error in the measurement of the
total energy of radiation will give a very small error in the
determination of the temperature.

Photometric methods have been made use of for the pur-
pose of measuring the intensity of the light radiated from an
incandescent body, and along this line the optical pyrometer
has been worked out. The method consists in comparing the
intensity of the light from the hot body with that from a
standard lamp, by ordinary photometric methods.

One of the best optical pyrometers is the invention of
LeChatelier, the man who did much in the development of the
thermo-electric pyrometer. Inasmuch as the principle used in
this instrument is typical of all others, it will be described
rather carefully, from which it should be possible to obtain a
fair idea of the optical pyrometer in general.

Vol. III. No. 1] TEMPERATURE MEASUREMENT : PEEBLES

The construction of the instrument may be seen from Fig.
3. A small gasoline lamp is placed at A, so> that light from
its central portion passes through the lense B, is reflected from
the 45° mirror, brought to a focus by the eye-piece, and ob-
served through a red glass.

This provides a red comparison field of constant intensity.
The lamp A is mounted eccentrically, and may be turned so
that the image of the flame is exactly bisected by the edge of

"i ! r

H- r -W ~Rbsorbiop Gloss

[~ , id

Fijf

L,eChatelier Optical Pyrometer.

the mirror C. Light from the incandescent body under ob-
servation is focused by the objective, passes by the edge of
the 45° mirror, and forms a red field immediately beside and
touching the first.

A measure of temperature is made by bringing the two
red fields to the same brightness. This is done by opening or
closing the iris diaphragm D in front of the objective, thus

80 THE ARMOUR ENGINEER [Jan., 1911

admitting more or less light from the body whose temperature
is sought. For very high temperatures additional absorbing
glasses of known coefficients of absorption, are placed below
the objective, and for lower temperatures before the com-
parison lamp. The opening of the iris diaphragm, when equal
intensity has been established, is read upon a scale, the square
of whose reading is a measure of the intensity of the light
from the incandescent source.

Since, according to the Stefan-Boltzman law, the tempera-
ture is proportional to the intensity of the radiation, we have
immediately a measure of the temperature when we have
measured the intensity of the radiation. All that is necessary
is to have two sources of light of known temperatures, molten
metal for example. Note the reading of the pyrometer when
focused upon each of these bodies and plot two points having
for their co-ordinates temperatures and scale reading on the
iris diaphragm. Draw a straight line through these two
points, produced in both directions, and the pyrometer is cali-
brated for all temperatures.

There is one important point to be kept in mind in con-
nection with the Stefan-Boltzman law quoted above. The law
is true only for what is technically known as a "black body."
The conception of such a body is due to Kirchhoff, who defined
it as a body which would absorb all radiations falling on u
and would neither reflect nor transmit any. Kirchhoff pointed
out that the radiation from such a body is a function of the
temperature alone, and hence may be used to measure the
temperature. The first experimental realization of such a
"black body" was made by Wein and Lummer. who heated
the walls of a hollow opaque inclosure as uniformly as pos-
sible and observed the radiations coining from tin1 inside
through a very small opening in the walls of the inclosure.

It is evident that such a body will absorb all the radia-
tions incident through the small opening, no matter what the
material of the walls may be, for unless the walls are totally
reflecting, all radiations must sooner or later be absorbed,
except that portion which may again escape through the small
opening. The presence of this small opening makes a slight
departure from a theoretical black body.

No body is known whose surface radiation is that of a
black body. The radiations from carbon and iron are very close
to black body radiation, while the radiation from polished
platinum and the white oxides departs very evidently from it.
It follows, therefor, that a number of different bodies all heated
to the same temperature will radiate different amounts of

Vol.111, No. 1] TEMPERATURE MEASUREMENT : PEEBLES

energy, and hence the optical pyrometer would show tempera-
tures for them all. In this connection the term "black body
temperature" has come into use. Two bodies are said to be at
the same black body temperature when they radiate the same
amount of energy. Manifestly they are not at the same tem-
perature, and hence the term, black body temperature violates
the conception of equal temperatures which is based upon
thermal equilibrium between the two bodies if brought into
contact.

Nevertheless, when pyrometers are calibrated in terms of
black body temperature, as all instruments based upon the
Stefan-Boltzman law must necessarily be, the conception of
equal black body temperatures is of great practical value.

It will be evident from the foregoing that the optical

pyrometer cannot be depended upon to give the correct ab-
solute temperature of all bodies. It will, however, repeat its
indications upon the same body with unerring accuracy, and
in a large number of industrial operations this is all that is
required. After the proper temperature for any operation
has been discovered, the optical pyrometer will make it pos-
sible to duplicate this temperature day after day with great
accuracy.

There are a large number of operations where the radia-
tion differs but slightly from that of a black body and hence
the optical pyrometer will read the absolute temperature cor-
rectly. A boiler furnace, a steel furnace, a porcelain kiln, a
pot of molten glass, an electric furnace, hot fire brick, etc.,
are examples. Thus the optical pyrometer has a wide field
for usefulness in the mechanic arts.

One of the most convenient instruments, based upon the
energy of total radiation, is Fery's Thermo-electric Telescope.

82 THE ARMOUR ENGINEER [Jan., 1911

This instrument combines the thermo-electric and optical prin-
ciples in the measurement of temperature in that its indications
depend upon the energy or radiation and are obtained from a
thermo-couple and galvanometer. The construction is shown
in Fig. 4. Radiation from the incandescent body passes
through the lens C and falls upon a very small and sensitive
thermo-couple shown at F in the sketch. A diaphragm D D
fixed in size and position, gives a cone of rays of constant
angular aperture, independent of the distance from the incan-
descent body. These rays, falling on the thermo-couple, pro-
duce an increase in temperature proportional to the total
energy of radiation, which in turn induces an electromotive
force proportional to the energy of radiation. Thus a direct
reading is obtained upon millivoltmeter which, according to
the Stefan-Boltzman law, may be read in terms of temperature.
The leads from the thermo-couple are led to the binding posts
shown at P P in the sketch, to which the millivoltmeter leads
are also connected. A and B in the sketch are screens placed
on each side of the thermo-couple to exclude all light except
that which comes from the incandescent source under obser-
vation. E is the eye-piece by means of which the image of
the light source is focused on the thermo-couple.

An instrument of this kind is independent of the distance
of the incandescent body, within certain limits, as will appear
from the following considerations. Tf the instrument is sighted
on an incandescent body of limited dimensions, the amount of
radiation passing through the opening in the diaphragm D D
will vary with the distance from the hot body, being inversely
proportional to the square of the distance. If the thermo-
couple were of such size as to receive all of the radiation con-
verged upon it by the lens C, then the indications of the gal-
vanometer would decrease as the distance from the incan-
descent source increases. But the thermo-couple, however, is
not large enough to receive all of the radiation converged
towards it. The image of the source of light, formed by the
lens C is large enough to overlap the thermo-couple on all
sides, so that when the observer sights the instrument the
thermo-couple appears as a dark disc in the center of a bright
field of light. When the instrument is brought nearer to the
source of light, thus increasing the size of the image produced,
the only effect is to increase the amount of this overlapping,
while the thermo-couple receives no more radiation than be-
fore. On the other hand, however, if the instrument be with-
drawn to such a distance from the source of light that the
image formed is not large enough to completely cover the

Vol. Ill, No. 1] TEMPERATURE MEASUREMENT : PEEBLES 83

thermo-couple, the readings obtained will he too low, and will
become less as the distance from the source of light is in-
creased.

From the foregoing it will be evident that the instrument
is independent of distance only within certain limits. The
image of the incandescent source must always be large enough
to completely cover the thermo-couple. In general, the diam-
eter of the hot body should measure as many inches as the
distance from the instrument to the hot body measures yards.

The Fery pyrometer is best adapted for use in the range
from 1300° F.. to 2800° F., where its indications are sufficiently
accurate to answer all requirements of industrial work. It
should not be forgotten, however, that an optical pyrometer
depending for its reading upon a measurement of the energy
of radiation reads black body temperatures, which in some
cases may vary considerably from true temperatures. But
when the same temperature is to be repeated time after time
in the same process, the Fery pyrometer will repeat its read-
ings with a degree of accuracy sufficient for all practical pur-
poses.

THE ELECTRIC DRIVING OF ROLLING MILLS.t

BY WILLIAM T. DEAN, E. E.*

The recent successful development of internal combustion
engines in large sizes suitable for use with blast furnace gas
has directed the attention of steel works engineers and man-
agers to the possibility of electrically driving all the machinery
in such plants.

So few years have elapsed since the electrical department
of most steel plants consisted of a chief electrician and one arc
lamp trimmer, that the growth of the electric drive has been
almost incredible. Today contracts are being carried out in-
volving the complete electrical operation of mills to produce
100,000 tons of rails per month, where the motor units reach
the enormous output of 10,000 horse-power each.

In this country, the Edgar Thomson works of the Carnegie
Steel Co. has the honor of using the first heavy rolling mill
drive by electric motors. The system used in this mill (250
volts direct current) is probably the most expensive in first
cost and least economical in operation that could have been
selected, nevertheless, the installation bas been a notable suc-
cess from the beginning. It is quite probable that the trans-
mission line for the light rail mill at the Edgar Thomson works
cost as much as the two 1,500-horse-power direct-current motors
used, and the building for housing the starting and speed con-
trolling rheostats would accommodate a very fair boiler plant.

The first consideration in any particular case involving
electric drive is — will it pay? Can more steel be turned out for
a given cost, or the same steel for a lower cost than with a
steam driven mill ? The next question is — will the electric mo-
tor meet the severe requirements of steel mill practice such as
continuous operation 24 hours per day and 30 days per month,
will it withstand severe overloads even to the point of stalling,
will the serious mechanical shocks incident to rolling, destroy
bearings and deteriorate insulation to such an extent as to ren-
der the maintenance cost of such machines prohibitive? The
comparative costs will be taken up later. All the questions
that arise affecting the adaptability of the motor for rolling
mill operation have been asked and successfully answered in
the past as applying to less important machinery. The solution
of the problem from the electrical manufacturer 's standpoint is

fAn article published in THE IRON TRADE REVIEW.

♦Class of 1900. District Manager, Power and Mining Dept., General Electric
Co., Chicago.

Vol.111, No. 1] ROLLING MILL ELECTRIC DRIVE : DEAN 85

only one of degree and therefore rests with the designer. That
the many problems entering into the design of the successful
mill motor can be solved is evidenced by the mills now being
operated electrically and by those undertaken on so great a
scale by the United States Steel Corporation.

The ability of a motor to operate continuously at a given
load is only limited by its ability to radiate the heat in which
the relatively small energy losses appear.

Constants of electrical design, such as safe amperes per
square inch in copper conductors, and flux densities in lami-
nated steel, are well known to the electrical designer by years
of experience. Generators of as great capacity as the largest
motors contemplated, have already been designed, built, and are
in successful operation. It may be conceded then with refer-
ence to continuous operation, that no serious difficulties will be
encountered. The electric motor has a great advantage over
the steam engine in the matter of performance under overload.
Speaking of the induction motor particularly, it may have an
overload capacity as great as 2y2 times its continuous output
and the motor may be brought to a complete standstill by an
unusual overload and the current flowing in the motor wind-
ings under these conditions may be precisely calculated before
the motor is built, and provision made to limit the maximum
current flowing to a predetermined value. What is of equally
great importance, however, is the fact that the motor current
may be automatically controlled so that excessive strains cannot
occur.

The only uncommon problem in the design of large mill
motors aside from that of mere size is that of mechanical pro-
portions to withstand shock and ordinary wear. It is in this
particular that the electrical manufacturer has been obliged
to revolutionize all his previous ideas. How well he has profited
by the experience of the engine builder may be gathered from
the massive construction shown in the illustrations.

Many of the mechanical shocks occurring in rolling opera-
tions with steam engines are due to the reciprocating motion
of the engine and not to the mill and gears. All such shocks
disappear when a motor is used, for one of the motor's most
valuable characteristics is its uniform turning moment.

When a mill is driven by a cross-compound or twin-tan-
dem-compound engine the shaft receives its turning moment
in four impulses; if the cranks are quartered there are four
points in each revolution when only one cylinder, or one en-
gine, if twin engines are used, is effective. Very heavy fly
wheels must be provided to overcome this defect, or if the

THE ARMOUR ENGINEER [Jan., 1911

mill is of the two-high reversing type, each cylinder or engine
must be made large enough to provide for the maximum torque
and there is great probability of entering a slightly cold bloom
or ingot at one of the low torque points in the cycle, entailing
backing out and loss of time. A motor having uniform turning
moment and a maximum torque of 2^ times normal full load
torque could experience no such difficulties. Indeed, it would
be a very poorly heated bloom that would not pass through
the rolls. As a matter of fact, it is necessary to provide auto-
matic torque limiting controllers not to protect the motor but
to protect the rolls and gearing between the motor and the
steel.

Having outlined the manifest advantages of the motor for
mill driving, the cost of operating a steam engine and an elec-
tric motor must be compared. Consider a mill requiring an
engine or a motor of a given rated brake-horsepower, and
assume a non-reversing three-high mill operating practically
continuously, conditions most favorable to the steam engine.
If steam must be generated on the premises the engine-driven
mill will be most economical since the motor must be charged
with the cost of transformation from mechanical into electrical
energy at the engine and generator, and the cost of transfor-
mation from electrical into mechanical energy at the motor as
well as the transmission losses. Even the superior economies
of large steam turbine generator units will not overcome such
double transformation losses. Assume, however, that there is
a distant source of power, natural gas, blast furnace gas, coke
oven gas, cheap coal or water power, it will be conceded with-
out argument that power may be transmitted more econom-
ically electrically than by any other means. It remains to show
the relative cost of transmission and the adaptability of the
possible sources of power to rolling mill drive.

The internal combustion engine is not adapted to the direct
driving of mills on account of its inability to sustain severe
overloads and its somewhat instability under Widely varying
loads. With gas power available the question narrows to the
cost of the transmission of gas and the consumption of the
same under boilers at the mill, as compared with the utiliza-
tion of the gas at the source of supply to produce electrical
power for transmission to motor driven mills.

In recent blast furnace practice, it has been found that
approximately 123,000 cubic feet of gas is produced per 24
hours per ton of pig iron. Two-thirds of this is available for
power for the operation of blowing engines and other pur-
poses, the remaining third being used to heat the air blast. A
500-ton furnace will produce therefore 41,000,000 cubic feet

Vol. Ill, No. 1] ROLLING MILL ELECTRIC DRIVE : DEAN

■ST

of gas per 24 hours. Numerous tests have shown this gas to
have a heat of combustion of 100 B. T. U., or more, per cubic
foot. Assuming 90 B. T. IT. per cubic foot as a conservative
figure the total heat available per 24 hours is 3,690,000 B. T. U.
The heat equivalent of one horse-power is 2,545 B. T. U. There-
fore, the theoretical power available from the gas is 1,450,000
horsepower-hours, or 60,417 horsepower. Assuming the net
efficiency of the gas engine at 22.5 per cent which, if in error,
is too high, the total available power from a 500-ton. furnace
is 13,600 horsepower.

2000 H. P. Three-phase Induction Motor Geared to Two-high Blooming Mills.

On account of the lean quality of blast furnace gas, cylin-
der dimensions must be large and this has kept the size of
single units down to about 3,000 horsepower and generators
rated at 2,000 kilowatts are generally used. Such generators
have an efficiency of about 95 per cent, making the total elec-
trical energy available per 500-ton furnace, 9,360 kilowatts.
Of this power about 600 kilowatts is required to operate gas
washing machinery, to pump jacket water, provide exciting
current for alternating current generators and minor purposes,
leaving a net available power of 9,000 kilowatts. From the
figures above we have :

THE ARMOUR ENGINEER [Jan., 1911

153,750,000 = total B. T. U. per hour.

2,545 = B. T. U. equivalent to 1 horsepower, theoretical

2,545

3,420 B. T. U. equivalent to 1 kilowatt, theo-

0.746 retical

153,750,000

45,000 kilowatts, theoretical

0.2

or the net efficiency of the entire plant will be 20 per cent.

In a large power plant using steam turbine driven genera-
tors and every known method of obtaining high efficiency, it
has been found that 27,000 B. T. U. in the coal produce one kilo-
watt at the switchboard. This is under regular commercial
conditions and includes all losses such as banking fires, opera-
tion of boiler feed pumps, circulating pumps, air pumps, coal
and ash handling machinery, etc., and the plant in question is
subject to heavy day loads and light night loads. Careful
tests at this plant indicate that if the plant could be operated
with a constant 24-hour load, such as obtains in steel mill prac-
tice, the economy would be 23,000 B. T. U. per kilowatt at
the switchboard. In a similar plant, gas fired, but subject to
a steady 24-hour load, a still higher economy could no doubt
be secured by the use of furnaces, especially designed to burn
the gas. Assuming, however, a fuel economy from gas of
23,000 B. T. U. per kilowatt, we have :

153,750,000 total B. T. U.

= 6,685 K. W.

23,000 B. T. IT. per kilowatt

or with the highest type of plant burning the gas to produce
steam and using turbo-generators of large size the power avail-
able from a 500-ton blast furnace is 6,685 kilowatts.

As before the theoretical power available is 45,000 kilo-
watts, giving a net efficiency of 14.8 per cent, or approximately
three-fourths the efficiency of the gas engine plant. It should
be noted that the higher output of the gas engine plant only

Vol. Ill, No. 1] ROLLING MILL ELECTRIC DRIVE : DEAN

applies to cases where blast furnace gas is available, since in
producer gas plants the engine must be charged with the heat
losses in the gas producer ; moreover, the quality of coal for a
gas producer must be much higher than is used in the steam
plant, on the economy of which the above calculations are
based.

The relative reliability of the gas engine and steam tur-
bine plants must be given serious consideration. A steam tur-
bine plant may be operated at its maximum rating indefinitely
with almost absolute freedom from shut-downs or necessity of
repairs. The gas engine, on the other hand, has not yet reached
a perfection of development where it may be depended upon
to operate 24 hours consecutively. In one plant of this nature
containing four large engines where in emergency one unit
could carry the entire load, an entire shut-down occurred, all
four of the gas engines requiring simultaneous repairs. The
enormous weights of the reciprocating parts, the great cylinder
dimensions, the rapid and wide variation in temperatures com-
bine to make the large gas engine somewhat unreliable. That
wider experience will teach engine designers methods of in-
creasing reliability cannot be doubted, and the manufacturers
of this class of prime movers are to be congratulated that the
great steel interests of the country are willing to invest their
capital so as to promote so important a development.

Gas engine builders have made many claims as to the re-
liability of their apparatus, most of which are based on Euro-
pean practice. To those in possession of reliable data on
European practice, such claims are taken with a degree of
allowance, as the following quotation from Engineering (Lon-
don) testifies:

"In this connection we may note that we recently heard
of a large continental power station where it has been deemed
advisable to install a reserve plant of 200 per cent of the nomi-
nal capacity of the station. As opposed to this, some English
builders of turbo-generators are advocating the absolute aboli-
tion of all reserve whatever, apart from that provided by the
by-pass valve, enabling the unit to take, if necessary, an over-
load of 50 to 100 per cent.

"Certainly the large continental gas-engine may be a suc-
cess if judged from a non-commercial standpoint. So long as
it works, it undoubtedly produces power very economically,
but those who have had. most experience with them are the
very ones who have the longest catalog of their defects."

Early in 1904, contracts were let for gas engines and elec-
tric generators of a total capacity of 8,775 kilowatts for Jo-

THE ARMOUR ENGINEER

[Jan., 1911

Vol. III. No. 11 ROLLING MILL ELECTRIC DRIVE : DEAN 91

hannesburg. South Africa. It was not until 1906 that power
was first delivered and the supply was extremely unreliable.

Quoting from The Engineer :

"On one occasion when five engines of 7,000 horsepower
were running, in the space of a quarter of an hour, every man
in the engine room but one was rendered unconscious from gas
poisoning. * *

At a coroner's inquest held in May (1907) on a gas fatality,
it was stated that 63 cases of poisoning had occurred in six
months. * *

At last on May 15 the plant was finally shut down. The
engine contractors threw up the contract and admitted that
they could not get the plant into shape to pass the specified
tests. The town council then rejected the plant. * *

In a report dated March 1 last, the general manager stated
that owing to continued break-down the whole of the generator
sets had never been available for use at one time. * *

The direct current plant of 5,400 kilowatts had had the
greatest difficulty in dealing with the maximum load of 2,750
kilowatts, and it had not been found possible to run the alter-
nating current generators in parallel on account of the un-
steady running of one of them. The generator sets could not
be got to give more than 80 per cent of their normal full load,
and it was seldom that two-thirds of the plant was fit to run.
For the three and one-half months ended Feb. 28 (1907) the
total works cost had been 5.1 cents per kilowatt-hour as com-
pared with the original estimate of 1.72 cents."

A complete turbine plant of large size, including the best
machinery, boilers, auxiliaries, etc., and the highest type of sta-
tion construction can be built for at least 60 per cent of the
cost of a blast furnace gas engine plant of the same capacity
with its auxiliaries. Assuming as arbitrary figures that a steam
turbine plant can be built for $60 per kilowatt, and that a gas
engine plant can be built for $100 per kilowatt, and that a plant
of 40,000 kilowatts average capacity is required, the investment
in the turbine plant will be $2,400*000.

It is to be remembered that the figure for the gas engine
plant is undoubtedly low when all the elements are considered
and that a turbine plant can be built for $45.00 per kilowatt
where no coal storage plant is required, and further that a tur-
bine plant can be built for $40.00 per kilowatt where cheap real
estate is available. My assumed figure of $60.00 per kilowatt
is based on a complete coal burning plant with coal and ash
handling machinery and includes real estate in a great city
where land values form a considerable portion of the total in-

92 THE ARMOUR ENGINEER [Jan., 1911

vestment. These modifications make a very material difference
in the total investment in a large plant, and on this point the
showing in favor of a turbine plant is much greater than indi-
cated in this article.

The gas engine plant, however, will require at least 25 per
cent excess in capacity in order to maintain the average out-
put given above and many engineers think that in the present
stage of the art, 50 per cent excess capacity should be installed.
The investment in the gas engine plant will therefore be $5,000,-
000, or 108.2 per cent in excess of the investment in the steam
turbine plant. The interest on $2,600,000 (the difference in
investment) at 5 per cent is $130,000 per year, or sufficient to
buy 86,750 tons of coal. If this coal were burned under boilers
in addition to the gas obtained from the blast furnaces, it
would generate 86,750,000 kilowatt-hours or 1,156 kilowatts 24
hours per day, 26 days per month, throughout the year. This
figure makes a very respectable addition to the power given
above, which may be legitimately expected to be generated by
the steam turbine plant, and leaves a relatively small margin of
total power in favor of the gas engine plant.

From this showing, the plants depending entirely on gas
engines must face very unfavorable conditions. They will have
on their hands enormously expensive plants, requiring four or
five times as much labor as the equivalent steam turbine plant
with constant danger of delay, due to break-downs and with
maintenance expenses reaching large figures.

In addition to all inherent disadvantages of the gas engine
plant previously noted, the fact must not be overlooked that in
the space occupied by one 2,000-kilowatt gas engine generator,
a steam turbine unit having a rating of 14,000 kilowatts can be
installed with all its accessories, and that a gas washing plant
used with the gas engines requires more space than a boiler
plant for an equivalent turbine installation. In fact, a gas
engine plant with its accessories, requires so much space, both
in the engine room and out, that considerable difficulty must be
experienced to properly operate all sections as a unit plant.

Earlier in this discussion it was proposed to pipe the waste
gases to the rolling mills and there generate steam for the
operation of mill engines. Neglecting losses in piping which
will exceed the losses in electrical transmission, this system
would show a lower economy than the turbo-generator system
by the amount that the reciprocating engine falls below the
turbine in efficiency. Non-reversing mill engines of large size
under favorable conditions, operating condensing, require ap-
proximately 14 pounds of steam per brake-horsepower-hour.

Vol. Ill, No. 1] ROLLING MILL ELECTRIC DRIVE : DEAN

Large size turbines only require about 9.4 pounds of steam per
brake-horsepower-hour. The total efficiency of such a system
would be

9.4

X 11.15 — 7.5 per cent.

14
For purposes of comparison on the mill basis the efficiency
figures on the turbo-generator system and the gas engine plant
must be decreased by the losses in the mill motor. Such a mo-
tor will have a full load efficiency of about 93 per cent.

Primary Control for Three 6000 H. P. Induction Motors.

The relative efficiency of the three systems of utilizing the
waste gases are at follows :

Per cent
Efficiency

(1) Gas transmitted to mill to produce steam

for mill engine 7.5

(2) Gas burned at source producing steam for

turbo-generators, energy transmitted elec-
trically to mill motor 15 . 17

(3) Gas used in internal combustion engines

driving generators, energy transmitted
electrically to mill motor 18 . 6

94 THE ARMOUR ENGINEER [Jan., 1911

The above figures should not be taken as absolute but give
fairly correct relative values, and clearly indicate the wisdom
of motor driving rolling mills. The writer believes that the
steam turbine system, while not the highest in efficiency, will
provide ample power for all rolling mill purposes. If such is
the case, the greater reliability of the steam turbine system far
outweighs the lower fuel economy, and the extra investment in
a gas engine plant will only pay in case there is a profitable
market for the excess power outside of the steel mill proper.

There is another source of electrical power for existing
steel plants which is extensively used in Europe but has only
been utilized in one case in this country and that on a relatively
small scale. I refer to the steam regenerator and low pressure
turbo-generator receiving an intermittent steam supply from
reversing mills or from non-reversing mills subject to wide
variations in load or speed, or both, and delivering a constant
supply of electrical energy.

A large reversing engine requires 60 pounds of steam (or
more) per horsepower. A steam turbine operating between 15
pounds absolute pressure and 28 inches vacuum requires 45
pounds of steam per kilowatt-hour. Thus, there is available
1.33 kilowatts in electrical energy for each horsepower of such
engines, more than enough to operate a duplicate mill electric-
ally. The gain by such an installation is all "velvet" and re-
quires a relatively small outlay of capital.

In choosing a system of transmission and utilization of the
electric drive the steel works engineer is at first inclined toward
direct current, owing to his greater familiarity with that sys-
tem and to its apparent simplicity. A direct current system
has some advantages such as the extension of an existing plant
to care for the heavier requirements of rolling mill drive. How-
ever, unless the centers of distribution of the power are very
close to the generators, the transmission line will be so expen-
sive as to be prohibitive. If an alternating current system is
selected the cost of the transmission line may be reduced to a
relatively small proportion of the plant equipment.

A universal formula for copper in a transmission line of
whatever system, voltage or frequency is the following :

D X W X C

P X E2

where A = area per conductor in circular-mils,

D = distance of transmission (one way) in feet,

Vol.111, No. 1] ROLLING MILL ELECTRIC DRIVE: DEAN 95

W = total watts delivered at end of line,
C = constant depending on the system and power

factor if alternating current,
p = percentage of loss in power delivered,
E = voltage at end of transmission line.

The constant C has the following values for the various
svstems employed :

Value of "C"

Per cent power factor 100 95 90 85 80

Direct current 2160

Alternating current, single

phase 2160 2400 2660 3000 3380

Alternating current, 2 phase.. 1080 1200 1330 1500 1690
Alternating current, 3 phase.. 1080 1200 1330 1500 1690

As an example in point let us consider the relative cost of
transmitting 5,000 kilowatts a distance of 2,000 feet by direct
current at 220 volts and by alternating current at the same
voltage.

From the above formula :
For direct current

2,000 X 5,000,000 X 2,160

A = ■ = 44,628,100

10 X 220 X 220

Assuming a loss of 10 per cent, two conductors of this cross
section each 2,000 feet long will be required, and since the
weight of one foot of copper having an area of one circular-mil
is 0.00000302 pound, the weight per foot of the above conductor
would be 134.77 pounds, and the total weight of copper required
would be 539,107 pounds or over 269 tons of copper, not includ-
ing insulation. At 20 cents per pound the copper alone for such
a line would cost $107,821.40.

For alternating current,

2,000 X 5,000,000 X 1,690

A = = 34,917,024.

10 X 220 X 220

Assuming the same energy loss as before and a power fac-
tor of 80 per cent, three conductors (for three-phase transmis-
sion) of this cross section each 2,000 feet long make a total
weight of copper of 632,702 pounds, which at 20 cents per
pound would cost $126,540.40.

!m;

THE ARMOUR ENGINEER

[Jan., 1911

Vol. Ill, No. 1] ROLLING MILL ELECTRIC DRIVE : DEAN 97

From the above it is evident that at low power factor and
equal voltages the alternating current transmission system
would be more expensive than the direct current. By reason
of commutation and insulation difficulties, direct current volt-
ages cannot be greatly increased ; on the other hand there is no
reason why alternating current generators and motors cannot
be built to operate at 6,600 volts or even higher and by the in-
terposition of transformers the transmission voltage may even
be raised to 100,000 or 150,000 volts.

Assuming a transmission at 6,600 volts the area of a con-
ductor becomes :

2,000 X 5,000,000 X 1,690

A = = 38,800.

10 X 6,600 X 6,600

The current per phase for the above 6,600-volt circuit is
545 amperes. It is not safe to allow less than 1,000 circular
mils per ampere, hence 54,500 circular mils or more per con-
ductor must be used. The next larger commercial size of wire
is No. 2 B. & S. gage, which has an area of 66,000 circular-mils.
Substituting this area in the above equation and solving for
the value of P we find the power loss to be 5.82 per cent. The
three conductors of No. 2 wire would have a weight of 1,205
pounds and would cost $241.80. Thus by using alternating
current at 6,600 volts, the cost of copper has been reduced to an
insignificant sum and the power loss cut in two. If the power
loss were further reduced 50 per cent ,the cost of the copper
would only be $323.20. It is true that insulators for 6,600 volts
cost more than for 220 volts, still in this case the fewer number
required and the greatly reduced labor cost overwhelmingly
favor the alternating current high voltage system.

The saving is proportional in all cases. A few summers
ago, the writer had occasion to figure on a system involving
three 500-kilowatt generators with a transmission of 2,000 feet.
The purchaser was offered three 500-kilowatt, 2,300-volt alter-
nating current generators, three 500-kilowatt rotary converters
with all necessary switchboards and the copper for the trans-
mission line for a sum about $45,000 lower than he paid for the
250-volt apparatus and line decided upon. The consulting engi-
neer in charge, while an excellent blast furnace man, knew
nothing of electrical matters and allowed his client to pay the
extra cost for a plant which must sooner or later be remodeled
to an economical basis.

Alternating current motors are now offered, which success-

THE ARMOUR ENGINEER

[Jan., 1911

Vol. Ill, No. 1] ROLLING MILL ELECTRIC DRIVE : DEAN

fully perform all the functions of direct current motors and
have many superiorities, such as absence of commutator and
ability to handle extreme overloads. These motors are built in
all sizes for the operation of roller tables as well as for the oper-
ation of the main rolls, hence, there is no longer an excuse for
the direct current system where large powers are used.

The particular voltage to be selected depends on tbe dis-
tance of transmission and the amount of power transmitted as
shown by the above examples. Consideration should also be
given to the possible expansion of the plant. For a plant gen-
erating 10.000 kilowatts or more. 6.000 volts should be the mini-
mum for transmissions of a mile or so. The choice of frequency
is limited to 25 or 60 cycles. The lower frequency is prefer-
able, owing to the lower motor speeds obtainable with reason-
able cost. The speed of an induction motor is inversely propor-
tional to the number of its poles and directly proportional to
the frequency of the source of supply. If the frequency be
stated in alternations per minute, the speed will be

of the alternations, where N represents the number of poles.
Thus, a six-pole 25-cycle (3.000 alternation) motor will have a
synchronous speed of 500 revolutions per minute, while a six-
pole 60-eycle (7,200 alternation) motor will have a synchronous
speed of 1,200 revolutions per minute. The difference in fre-
quency becomes very apparent in large slow speed motors for
direct connection to rolling mills.

Such a motor operating at 75 revolutions per minute would
have 40 poles if 25 cycle, and 96 poles if 60 cycle. A large num-
ber of poles makes a difficult design unless very great diameters
are used and in any case has a very bad effect on the constants
of the machine, particularly the power factor. Unquestionably
a frequency of 25 cycles is preferable for power purposes.

The question of suitable speed for a motor driving a rolling
mill should be solved by the mill engineer rather than the motor
manufacturer. The former should, however, bear in mind the
limitations of speed within which the latter must work. Prob-
ably the lowest speed for direct connection to rolls which would
be considered would be 50 revolutions per minute for 25-cycle
motors. The accompanying table gives the synchronous speeds
possible and the probable full load speeds for large 25-cycle
motors.

100 THE ARMOUR ENGINEER [Jan., 1911

Synchronous and Full Load Speeds for 25-Cycle Motors.

Synchronous Full Load

Poles.

Speed.
50.00

Speed,
48.50

51.70

53.50

55.50

50.20

51.80

53.80

57.70

60.00

55.90

58.20

62.50

60.70

65.20

63.30

68.20

66.10

42 .

71.40

69.30

75.00

78.90

88.20

72.80

76.50

85.40

93.75

100.00

107.10

91.00

97.00

103.30

. 115.40. .

111.40

20 '.'..'..'.

125.00

136.30

150.00

166.60

120.60

131.50

144.80

160.80

16 .

187.50

181.00

214,30

207.00

250.00

300.00

241.00

290.00

375 00

362.00

500 00 . .

482.00

750.00

720.00

1500.00

...... 1440.00

Whenever the speed of the rolls exceeds 55 revolutions per
minute it is preferable and cheaper to direct-connect the motor
than to employ a great reduction. This refers to mills requiring
motors of 3,000 horsepower and larger for a single roll stand,
and to many cases where smaller motors would be employed.
Where it is necessary to drive more than one roll stand from a
single motor, thus entailing gearing which would in any case
be charged against the mill, a higher speed motor should be
used, but in no case should the gear ratio be greater than 3:1.
For motors of 2,500 horsepower or larger, a lower gear ratio
should be selected.

Vol. Ill, No. 1] ROLLING MILL ELECTRIC DRIVE : DEAN

101

For a close group of small mills, each of which in succes-
sion does a portion of the total work of reducing a bloom to a
commercial shape, the most economical drive is by a single
motor. This is due to the higher efficiency of a single motor of
large size over several small motors and to the fact that such a
motor may be operated at more nearly its full load continuously.
The first cost of such a plant will be appreciably lower when a
single large motor is installed.

The method of driving a group of small mills from a single
motor must be very carefully considered in each case. If bevel
gears are used the driving shaft must operate at a relatively
low speed in order to avoid the use of too high a gear ratio

75 100 «25

REVOLUTIONS PER MINUTE

Curves showing efficiency and power factor of a 5000-Horsepower Mot
designed for various synchronous speeds.

102 THE ARMOUR ENGINEER [Jan., 1911

between the shaft and rolls. This means a low speed and con-
sequently an expensive motor. If the conditions will allow the
use of a rope drive the motor may have a fairly high speed and
the total friction loss may be reduced. It is probable that the
maintenance cost of a rope drive for a group of small mills will
be lower than for the equivalent bevel gear drive. Any flexible
connection, such as a rope drive between the mills and the mo-
tor, will be favorable to the operation of the latter. In general
it may be said that the higher the motor speed adopted the bet-
ter will be the efficiency and power factor. This is shown by the
curves giving approximate value of power factor and efficiency
for a 5,000-horsepower motor designed for various synchronous
speeds.

Summarizing the various conditions which must be consid-
ered in any proposed rolling mill drive, it has been shown that :

1 — The electric drive is absolutely reliable.

2 — Alternating current motors and transmission system
should be used.

3 — A frequency of 25 cycles per second is preferable.

4 — With blast furnace gas available the greatest amount
of power may be obtained from a gas engine plant.

5 — A boiler plant with steam turbines will produce three-
fourths the power obtainable with the same fuel used in a gas
engine plant.

6 — The great reliability of a steam turbine plant outweighs
the excessive power obtainable from a gas engine plant.

7 — The saving in investment in a steam turbine plant over
a gas engine plant is very considerable.

8 — It is more economical to generate electric power at the
source of gas supply and to transmit same to motor driven mills
than to burn the gas under boilers at the mill.

9 — The electric drive is the most economical system for
every case excepting where coal must be burned under boilers
at the mill and in this case approximately double the power
can be obtained from a given amount of steam by using low
pressure turbines in the exhaust from mill engines.

The writer believes the foregoing conclusions to be correct
and that a careful investigation on the part of any steel works
engineer will prove their correctness.

THE SAFETY FACTOR.

BY WILLIAM F. DIETZSCH, M. E.*

Paramount to all other considerations in the solution of
the innumerable and ever changing problems of the construct-
ing engineer of the present day must be his unstinted regard
for maximum safety, reliability, and economy in each and
every one of his constructions. Not one cubic inch of excess
material can he afford to have his mechanism or structure
carry as an unnecessary ballast, in order to conform with the
principles of economy; and yet, on the other hand, it should
not lack a jot when the reliability of the device and the safety
of the users, into whose hands it is placed, is involved. Every
element and feature entering into the design of his engineering
proposition must be based upon thorough, practical, and sci-
entific study and investigation. He must feel secure in his
claim that every dimension in his design has its good and
logical "raisou d'etre."

Any novel type of engine, boiler, dynamo, bridge, etc.,
representing a successful design, must primarily be strong
enough and sufficiently resistant to withstand the repeated
strain of the forces acting upon it. The elastic deformations
of all or any of its composite members when stressed to their
greatest loads must remain within the confines of safety — i. e. :
within the elastic limit of the materials of which they are
constructed. The choice of the proper allowable factor of
safety, in each and every case, is an extremely important and
vital question for the engineer to decide. The rule of
the thumb will not do when it comes to the proper di-
mensioning of an engine or bridge, and the calculations
for the correct proportions, dimensions and the proper
and economic distribution of the different materials that com-
pose the various parts of his structure must be based upon
established scientific and practical facts. Experience, science,
and theory must be the guiding factors in the effective elabo-
ration of all bis engineering work, each reinforcing the other.
The safe allowable working stress in one specific case
may be one-half to one-fourth of the ultimate strength of the
material, and yet in another instance conditions may be such
that twenty to forty would not represent an excessive
value for the safety factor. Then again there are times
when the allowable working stresses mav not be of sneh
a significant moment as the allowable working strains, which
then shonld form the basis in the ealcnlations for the deter-

♦Instructor in Experimental Ensnneerins\ Armour Institute of Technology.

104

THE ARMOUR ENGINEER

[Jan., 1911

mination of dimensions and selection of proper materials.

We frequently meet conditions where the cross-sec-
tional area of a certain member of a structure can be kept
at a low figure, when the working stress only is kept in view,
but when we consider the corresponding strain induced by
this stress, we may find that the deformation of this single
clement in the mechanism or structure may have such an
influence, in its relationship to immediate adjoining members,
that it may not only nullify its own function, but that of the
entire design.

From an illustration in the (London) Engineer.
View of Wreckage of Quebec Bridge.

To illustrate the responsibility of the engineering pro-
fession and to emphasize this fact to the young graduate
about to enter the field of practical endeavor, let it suffice to
point to such examples as the collapse of the Quebec bridge
over the St. Lawrence and numerous boiler explosions.

An instance that came to the writer's personal notice
some time ago where he was called upon to render expert tes-
timony regarding the probable cause of the explosion of a
steel tank that had been designed for the storage of com-
pressed air under pressure of several hundred pounds per

Vol. Ill, No. 1] THE SAFETY FACTOR: DIETZSCH 105

square inch, may serve to illustrate the importance of
the proper regard for correct calculations in the strength of
the design, and for the choice of the right materials. The
failure of the tank could he directly attributed to the weak-
ness of a cast iron flange, which, instead of measuring 21/2"
to 3" in thickness, was erroneously considered heavy enough
with IV' of metal. The cost of this blunder on the part of
the designing engineer was the serious injury of several work-
men engaged in testing the tank.

There is a growing tendency among many engineers of
today to base the factor of safety upon the actual elastic limit
and not upon the ultimate strength of the material. In many
cases it is really incorrect, in the writer's estimation, where
it is necessary to deal with alternating stresses at frequent
and rapid applications, to base the factor of safety upon the
ultimate strength, because it is an established fact that when
stresses exceed the elastic limit of any material it only requires
a definite number of these applications until the actual failure
is reached.

IT IS THEREFORE NOT ONLY MISLEADING, BUT
ERRONEOUS TO USE THE ULTIMATE STRENGTH AS
THE BASIS FOR THE FACTOR OF SAFETY. THE TRUE
FACTOR SHOULD REPRESENT THE RATIO OF THE
WORKING STRESS TO THE ELASTIC LIMIT OF THE
MATERIAL IN SERVICE.

PNEUMATIC ASH HANDLING SYSTEMS.
BY R. B. HARRIS, M. E *

It has happened that several minds working under similar
conditions and with like surroundings, but with entirely dif-
ferent objects in view, have almost simultaneously reached the
same conclusion. Such a situation seems well illustrated in
the development of vacuum cleaning apparatus so com-
monly known and quite universally applied during the period
in which the same principle was being perfected for the hand-
ling of ashes. The success of vacuum cleaning by applying air
suction with provision for free flow of air to carry dust was
paralleled by the success of conveying ashes in similar manner.

The difference, however, in the actual design of the appa-
ratus for two so widely different substances, both to be handled
by air, required ingenuity along different lines. In vacuum
cleaning the dirt and dust to be carried is relatively insignifi-
cant in weight or volume and the necessity of making the
apparatus light and portable is pre-eminent, while wear of its
parts, in conveying such material, is inappreciable. On the
other hand, in handling ashes, the weight and volume to be
carried is the first consideration, and the wear and tear while
conveying such material demands design of parts of unusually
heavy construction.

It at once appears necessary to make the apparatus for
handling ashes so heavy that possibility of portable form is
out of consideration and such apparatus, therefore, becomes
stationary. Ashes also are ordinarily produced under condi-
tions of fixed or stationary character, and the systems, there-
fore, may readily be designed to serve as collectors of ashes
from various points to one place of final accumulation and
disposition. This makes the design of pneumatic ash handling
machinery to be adapted to fixed— but different — conditions in '
every plant an engineering problem, and not merely a piping
layout.

The GrECO Pneumatic Ash Handling System, as exclusively
manufactured by the Green Engineering Company, is typical of
the development of applying air for conveying purposes. In
these Systems a conveyor pipe is located convenient to ash
pits, where ashes from furnaces are deposited. In this pipe,
ash intakes are provided, into which ashes may be conveniently

*Class of 1902. Superintendent of Construction, Green Engineering Com-
pany, Chicago.

Vol. Ill, No. 1] ASH HANDLING SYSTEMS : HARRIS 107

raked or shoveled. A continuous air suction with high velocity
air current entering ash intakes is maintained, and the ashes
are thereby readily fed into such openings. The conveyor
pipe is continued to a separator and accumulating ash tank, in
which the high velocity of the air current, bearing the ashes
in suspension, is suddenly reduced to almost no velocity by
expansion and the ashes at once deposited in the tank. To
further facilitate such deposit, the ashes are subjected to a
water-spray just before entering the tank, the water serving to
wet the ash particles, increasing their weight, as well as to
attach the dust of suspension to the larger ash. At an angle
radically opposed to the angle of entry of ashes, an exhaust
pipe serves to withdraw the air entering the tank by way of
its connection to a powerful exhauster ; in fact, the exhauster
produces the air current through the entire system and its
suction on the tank is sufficient to produce the high velocity
in the conveyor pipe connected thereto.

Between the tank and the exhauster a dust collector is
placed through which the air, separated from ashes, must pass
in a somewhat helicoidal path. More particularly, the ar-
rangement provides for imminent contact of air current with
water surfaces for the purpose of still further extracting any
particles of dust which may not have been deposited in the ash
tank.

The exhausters for Pneumatic Ash Handling Systems vary
widely in capacity and pressure, depending upon the amount
of ashes to be handled, and length of conveyor pipe through
which the requisite air must be drawn. The exhausters are
driven by either engines, turbines or motors, depending upon
the most convenient motive power available in the plant to be
served.

The Systems are built in various sizes, rated by the amount
of ashes to be conveyed per minute. Thus, for example, a 6"
System has a capacity of 150 pounds per minute ; an 8" System,
300 pounds ; and a 10" System. 500 pounds. Such rates of con-
veying ashes are, relatively speaking, enormously greater than
possible capacities of any other form of ash conveying appa-
ratus. In fact, conveying capacity of Pneumatic Ash Handling
Systems exceeds the ordinary ability of one man shoveling
ashes under usual circumstances. It is therefore necessary to
arrange the intakes convenient to the pits where the ashes
accumulate, in order that the possible rate of handling ashes
into intakes may be such as to permit one man to feed the full
capacity of the System.

It is to be borne in mind that but one ash intake should
be open at any one time, and that at this point only can ashes

Vol. Ill, No. 1] ASH HANDLING SYSTEMS : HARRIS

be fed to conveyor pipe. This condition is made apparent by
considering two ash intakes open at the same time; the one
farthest removed will be without air suction, as suction will
naturally be spent at the opening nearest the separator tank.
To more readily insure free air supply to the System at all
times, the farthest extension of conveyor pipe is left open by
attachment of a fitting known as an "Air Intake." This does
not interfere with the full suction at any other opening nearer
the tank, and at the same time permits free operation of ex-
hauster set during the interval while operative is passing from
pit to pit and all ash intakes are closed.

The necessity of restricting the ash intakes to but one
opening at any one time is really a decided advantage, from
the standpoint of labor required, as the relatively enormous
carrying capacity of these Systems makes it readily possible
to handle as high as fifteen tons of ashes per hour with one
operator, and there are few power plants in which the ash
accumulation cannot be handled by one man in ten working
hours. It is customary to arrange the size of ash pits suffi-
ciently large that the ash cleaning periods are limited to the
work of one man. On account of the convenience, and as the
labor involved is not objectionable, ashes are usually cleaned
from the pits successively in very short time, and often by the
same man firing the furnaces.

The nature of ash from coal differs widely. It may not
be generally known that ashes from different coals may vary
in weight from as low as 30 to as high as 60 pounds per cubic
foot. Inasmuch as the weight of ash to be handled by air in-
fluences the amount of air required to float the ashes in the
air current, it is apparent that for ash of varying weight dif-
ferent air currents, or air velocities, or really air density must
be provided. In this respect, the Systems require other calcu-
lations for heavy ash than for light ash. Further, the air cur-
rent actually established throughout the System is dependent
upon the suction maintained at the opening farthest away
from the separator tank. As the suction to be maintained in
any pipe line with air current moving there through is depend-
ent upon the friction of the air current through the pipe, the
length of conveyor pipe, as well as the number and nature of
obstructions (such as bends or turns) causing additional fric-
tion loss, will enter the calculation of air suction to be fur-
nished by the exhauster set. Thus, the weight of ashes to be
handled will determine the relative volume of air to be ex-
hausted, and the friction loss through the System will deter-
mine the amount of suction required.

The power consumption of engine, turbine, or motor, driv-

110 THE ARMOUR ENGINEER [Jan., 1911

ing the exhauster will then depend upon the volume and suc-
tion produced by the exhauster as illustrated above. In some
6" Systems 15 H. P. is sufficient, whereas the same size System
may require 60 H. P., if the latter should employ an extremely
long conveyor pipe, with probably several elbows or bends.
Similarly, either of these Systems of the same size conveyor
pipe may require 15 to 40 more H. P., if intended to handle
heavier ash. Corresponding figures for 8" Svstems may involve
from 30 to 100 H. P., and 10" Systems from 50 to 150 H. P.

As the necessary air currents for floating or carrying the
ash in suspension are relatively of high velocity, the effects of
turns in the conveyor pipe at once become of considerable
consequence. The ash entering the System immediately trav-
els toward the center of the pipe where the greatest air veloc-
ity is maintained, on account of the least retardation by skin
friction of the conveyor pipe. In other words, the ash really
does not touch the pipe after entering the air current and while
continuing at its maximum velocity. That such is the case is
readily understood by considering the ash particles as having
relatively large surface on which the air may impinge, and
therefore at all times ash particles are affected most by the
air current of greatest velocity.

However, when such ash laden air current reaches a bend
in the conveyor pipe the heavier ash particles are projected
forward to the outer surface or back of such bend, and. at their
high velocity ashes at once attack this surface in the same
manner as a sand blast would attack any surface towards which
it is directed. In these Systems suitable provisions are made
for replacement of such backs and to conveniently and cheaply
repair such abrasions of the fittings occurring at each bend in
the conveyor pipe. These fittings may readily be opened by
removing hand-holes on the inside curvature, giving access to
the back. The life of wearing backs is dependent on the veloc-
ity required in the System, on the nature of the ash, and the
amount of ashes handled, and may vary from ten days' life to
two years', depending upon such conditions. In any event,
the cost of replacing and maintaining such wearing backs is
insignificant when compared with the savings possible with
these Systems over any other method of handling ashes and
maintenance of apparatus therefor.

Inasmuch as the disturbance at elbows may, under some
conditions, be continued by one or more rebounds of ash from
wearing-back to pipe immediately beyond the fitting, it is cus-
tomary to provide short lengths of pipe directly beyond such
fitting for convenient turning of pipe and eventual replacement.

Inasmuch as the conveying capacity of these Systems de-

Vol. Ill, No. 1] ASH HANDLING SYSTEMS : HARRIS 111

pends on the velocity established, and the velocities attained
depend on the suction provided, it is apparent that air-tight
apparatus, fittings, and connections should be provided and
maintained throughout, as any leakage between an ash intake
opening and the separator tank will at once decrease the suc-
tion and thereby the velocity of air current, and also decrease
the suspending property of the air in motion at point where
ashes are introduced and elsewhere within the System. Should
decreased velocity result for some reason, viz., should ashes be
fed at excessive rates (which rates then would be beyond the
suspending capacity of the air current), a tendency for ashes to
lag would result; that is, the effect of gravitv would exceed
the effect of velocity contributed by the air current, and then
the ashes tend to, and would, eventually, lodge in the bottom of
the conveyor pipe. Establishing the proper air currents for
such excessive feeding, or decrease in the rate of feeding the
ashes, would again restore the velocity needed and such parti-
cles would be picked up and assume the necessary velocity,
passing on as originally designed.

Wet ashes, having much greater weight per unit of vol-
ume, would enter the System under the same circumstances as
ashes heavier than those capable of suspension by the air cur-
rent provided in the System. Therefore, wetting ashes and
then introducing them into the System usually deposits them in
the pipes, and, as wet ashes will stick to any surface even after
dried, the introduction of wet ashes is impractical. However,
as ashes may be fed hot, or even on fire, into the System, and as
all ashes are quenched by the spray before entering the tank,
there is really no occasion for wetting prior to feeding into
the System. Further, as ashes do not require reshovelmg or
rehandling, causing the stirring of dust, and as dust from first
handling is at once drawn into the System, there is really no
desire on the part of operators to expend the additional labor
of wetting ashes down.

The operation of these Systems depends only on an engine
or motor and its exhauster as the entire moving machinery.
These are both conveniently located and away from the ashes
to be handled, and may be housed to protect them from the
machinery deteriorating conditions usually existing in boiler
rooms, or incident to the handling of ashes. The simplicity of
the moving machinery of these Systems, and particularly its
remoteness from the point where ashes are accumulated, greatly
reduces the maintenance cost thereof, and always readily per-
mits its inspection under most favorable surroundings.

As the ashes are drawn into the System, any dust occa-
sioned is at once drawn away, and no opportunity exists

Vol. Ill, No. 1] ASH HANDLING SYSTEMS: HARRIS 113

for gases, steam or dust to arise and contaminate the surround-
ings. The work of the operator, therefore, is relatively pleas-
ant as compared to usual methods of handling ashes. Under
these conditions, conspicuously clean boiler rooms and sur-
roundings are some of the attractive features contributed by
the System.

The absence of moving machinery in boiler rooms and ash
pits has made this System most desirable when compared to
ash conveying devices of any other type. No possible danger
can confront the operator while feeding ashes into the conveyor,
nor is it necessary to risk life or limb in lubricating any mov-
ing parts, so common in other systems usually involving
parts located in dark or inaccessible places. As ashes are
finally stored in sealed tanks, danger from fire is eliminated.

In the description of the System I have mentioned the
provisions for angles or bends in the conveyor pipe. These
possibilities adapt the System to almost any construction of
boiler rooms, as the conveyor pipe line can be arranged to
avoid conflict with other apparatus and either pass around,
above, or below, where no other system is possible. The ash
storage tank may be located either inside or outside of the
building, wherever most convenient to final discharge by grav-
ity to either car or carts. Suitable valve for this purpose is
attached to the cone-shaped bottom of the separator tanks.

The final discharge from the exhauster set is ordinarily
conducted into the chimney or breeching serving the furnaces
or, directly into the atmosphere.

The piping connections to water-spray are usually located
convenient to controller operating motor, or to the engine throt-
tle, as the starting of the entire apparatus involves only the
turning on of the engine, or motor, and opening the water-
spray.

The sizes of ash storage tanks vary, depending on disposal
arrangements peculiar to any particular plant. Tanks of five
tons capacity would be considered small, while 75 tons capacity
for railroad car discharge would not be unusual.

The GECO Pneumatic Ash Handling Systems above de-
scribed are manufactured and installed exclusively by the
Green Engineering Company, Chicago, who employ a staff of
engineers for the calculations and design of this apparatus. It
is not inopportune to mention that several A. I. T. boys are
among these engineers.

STRESSES IN AEROPLANES DURING QUICK TURNS.
BY M. B. WELLS.*

The evolutions performed by aviators include, among the
most daring, the spiral glides and quick turns. In making a
spiral glide the aviator rises to a height of probably several
thousand feet and descends in a spiral path either with the
motor running or with the power shut off. The speed attained
is sometimes very great, the time of making one circuit is
short, and the banking of the machine is at a steep angle.
No definite data is obtainable in regard to the speed or the
time of a circuit in these glides, but in quick turns made in
ordinary exhibition flights, time as short as five and one-fifth
seconds has been reported. A speed of fifty miles per hour is
not unusual in ordinary straight flights, and this speed is
doubtless often exceeded in the downward circular flights.

The following is a discussion of the principal stresses ex-
isting in a biplane making a complete circuit in five seconds at
the assumed speed of fifty miles per hour : The circumference

5280 X 50

of the circle swept bv the machine will be X 5

60 X 60

367

— 367 feet, and the radius of the circle will be

3.14 X 2
== 58 feet.

From mechanics we have the centrifugal force of a body of
mass M

47r*a
= M

where a is the radius of the circle in feet and T the time
in seconds during which the mass moves around the circle.
The total weight of the machine and operator is about 1,200
pounds. Substituting the known quantities in the above
formula and solving for the centrifugal force it is found to be

"Associate Professor of Bridge and Structural Engineering, Armour Institute
of Technology.

Vol. III. No. 1] AEROPLANE STRESSES: WELLS 115

1200 X 4 X 3.14 X 3.14 X 58

= 3420 pounds.

32.2 X 5 X 5

This acts horizontally and outward, while the weight of the
machine acts vertically downward. The resultant of the two is

[ (3420) * + (1200) '] »./» = 3624 pounds.

The cosine of the angle that the line of action of this resultant
force makes with the horizontal is 3420/3624 = 0.94, and the
angle is 20 degrees.

With the pressure of the air perpendicular to the surfaces,
and sufficient to halance the above resultant force, the angle
that the machine must make with the horizontal is 90° — 20°=
70°. Observations and photographs of machines in circular

r—i

6-.o" 1 .S'-o' !»,-°*[. *°" ,1- «'-«' r|, «'■"' J

» <t OK FT 4 »

"'o

TT-tc Rrmour £V

Outline of Biplane Truss.

flight indicate that under extreme conditions they approach
such an angle.

The total downward and outward resultant force being
3624 pounds, each pound of weight of the machine exerts a
force of 3624/1200=3.02 pounds along lines parallel to this
resultant ; that is, perpendicular to the chords of the main
trusses of the machine.

The accompanying sketch is an outline of one of the two
main trusses of a machine, the plane of the truss being perpen-
dicular to the line of flight. For convenience it is placed with
its longer dimension horizontal. It is assumed that one-half the
weight of the machine is distributed over this truss as follows :
At each of the points A and a, 7V2 pounds; at each of the
points B, b, C, c, and D, 15 pounds; and at d 210 pounds. The
right half is loaded the same. Multiplying each of these
weights by the constant 3.02 gives the pressure exerted at the
respective points, the truss being now turned so that these

116 THE ARMOUR ENGINEER [Jan., 1911

pressures are vertical. They are as follows : At each of the
points A and a, 7.5X3.02 = 22.65 pounds; at each of the
points B, b, C, c and D, 15 X 3.02 = 45.3 pounds; and at d,
210 X 3.02 = 634.2 pounds. The total downward pressure on
this -truss is then, (634.2 + 5 X 45.3 -f 2 X 22.65) X 2 = 1812
pounds, and this is balanced by the pressure of the air in the
opposite direction.

The surfaces are 5% feet wide, and the total area support-
ing this truss and its load is 5.5 X 2 X 39 / 2 = 214.5 square
feet. The pressure per square foot on the surfaces is 1812/
214.5 == 8.45 pounds. The number of square feet at A is
(5.5/2) X 3 = 8.25, and the number of pounds of upward
pressure is 8.25 X 8.45 = 69.7. The upward pressure at a is
the same ; the upward pressure at each of the points B, b, C, and
c is two times the above or 69.7 X 2 = 139.4 pounds; and the
upward pressure at D, also at d, is (5.5/2) X (3 + 1.5) X
8.45 = 104.5 pounds ; the corresponding upward pressures on
the right half being the same.

The differences between the upward and downward pres-
sures at the respective panel points of the truss give the re-
sultant loads at these points which are to be used in determin-
ing the stresses in the truss members. At A the resultant load
is 69.7 — 22.65 = 47.05 pounds upward, and at a it is the
same. At B, b, C, and c the resultant is 139.4 — 45.3 = 94.1
pounds upward, at D it is 104.5 — 45.3 = 59.2 pounds upward,
and at d it is 634.2 — 104.5 = 529.7 pounds, downward. The
algebraic sum of these upward and downward loads is zero.

All diagonals are designed to take tension only. Passing
a section through CD. Cd, and cd, and taking moments at d
gives the following equation :

94 X 18 -f 188 X (12 -f 6) + CD X 6 = 0;
Solving,

CD = —846 pounds.

This is also the stress in DE and EF. The stress in de is
the same but of opposite sign. With the same section and the
center of moment at C gives the equation 94 X 12 + 188 X
6 — cd X 6 = 0, which when solved gives cd = +376 pounds.
The stress in BC is the same, but with opposite sign. With
a section across the panel ab and center of moments at b

94X6
the stress in AB = = 94 pounds. The stress in be

Vol. III. No. 1] AEROPLANE STRESSES: WELLS 117

is +94 pounds, and the stress in ab is zero. »

The upper portion of the post Dd has a tension equal to
the upward resultant at D, or 59 pounds. The portion of the
post below the engine connection has a compressive stress of
529 — 59 = 470 pounds. Passing a section cutting BC, Cc, and
cd, the forces on the left are all upward and are equal to
2 X 47 + 3 X 94 = 376 pounds. 376 + Cc = 0. and Cc =
— 376 pounds. Similarly the stress in the post Bb is — 188
pounds, and in the post Aa it is — 47 pounds.

The vertical component of the stress in Cd is (2 X 47) +
(4 X 94) = +470 pounds, and this multiplied by the secant of
45 degrees = +470 X 1.41 = +663 pounds. The stress in Be
is (2 X 47 + 2 X 94) X 1.41 = +398 pounds, and the stress
in Ab is 2 X 47 X 1.41 = +133 pounds. The stress in the
diagonals of the panel de is zero.

In these results the plus sign has been used for tension and
the minus sign for compression stresses.

The above given stresses will be modified when the planes
of the main trusses are not approximately perpendicular to the
chords of the curved supporting surface.

FOUNDATION CONSTRUCTION OF PROPOSED COM-
MERCIAL BUILDING.

BY H. W. CLAUSEN, C. E.*

During the years 1872-1874 a circular, brick-lined water
tunnel of 7' internal diameter was constructed by the city of
Chicago from the present old two-mile crib to a point near
22nd Street and Ashland Avenue, where it connects with the
old West Pumping Station, now called the 22nd Street Pump-
ing Station. The line of this tunnel runs straight from the
two-mile crib to a shaft located at Chicago Avenue and Lincoln
Parkway, just outside the present Chicago Avenue Pumping
Station, and straight from this shaft to a shaft located just
outside the pumping station at 22nd Street and Ashland Ave-
nue. The depth of the tunnel below the surface varies, but
it is not less than 60 feet at its highest point. This straight
tunnel line was adopted on the principle that the hypothenuse
of a triangle is shorter than the sum of the other two sides, and
of course on this proposition the cost of the tunnel was re-
duced. It was also believed justifiable to adopt this plan of
crossing under private property, because it was supposed that
no foundation would ever be carried to such a depth. This
reasoning, which was sound enough at the time, is now in-
correct, because foundations for our large buildings are now
often carried down to rock, which may be as far as 120 feet
below the surface.

During the construction of a pile foundation for a building
in another part of the city, it was discovered that a pile had
penetrated one of the city's tunnels which luckily happened
to be only a connection for equalizing purposes, and it cost
the city approximately $120,000 to have the damage repaired.
After this occurrence it was made a rule of the Building De-
partment, before issuing a permit, to require the O. K. of the
City Engineer on all plans showing a foundation of any con-
sequent depth. In 1904 an application was made for a permit
to construct a commercial office building on a pile foundation
on the south side of Madison Stret and on the east bank of the
Chicago River. Now, when this was referred to the City En-
gineer it was discovered that the tunnel herein described
crossed under the property at a depth of about sixty feet below
the surface. The permit was therefore held up pending an
understanding between the owner and the city. After con-

"Class of 1904. Assistant Engineer, City Chief Engineer's Office, Chicago.

Vol. III. No. 1] FOUNDATION CONSTRUCTION: CLAUSEN 119

ferences it was finally agreed between the parties that the city
should build a caisson foundation for the building and pay the
difference in cost between the pile and caisson foundations,
respectively. Since the tunnel in question was the only source
of water supply to the southwest side, it was imperative that
no damage should result to the tunnel as a consequence of

_L7ackso^ Blvd

\ 1/

^ ze- o." Y 3'shcft

these operations; therefore, the city reserved the right to
let the contract directly and supervise the construction of the
foundations.

Before the plans could finally be revised to show the lo-
cation of the concrete caissons it was necessary to know ex-
actly where this tunnel crossed the property. Accord-
ingly a survey had to be made connecting various working
and fire shafts along the line of the tunnel. This survey proved

£ !

co
o

■* b- OS

GO 00 ©
OSHr-

O CO
O r-
CO CO

kffl

mi>c

O CM
TfH O"

CO
Ol

o
cm

ISO
GO

5©

o
o

r—

cr.

CO
GO
t-

CO
CO

TjH CO
CM

oi
CO

-t-

OS •

CO •

os •

1 — 1

©*

co :

tr-

ee
CM

wm^.m^.^.^^.^^.mK

C><S> f^rH^M^r^ c^

»(oocom©MOOO»0

iO-Ht^tJH CO ^ ^ CM

CqH COcMCMtHCM cm

OOi— ICOO0OOOO0O

0"^Oi— llr-t^cMCOi-OOSitr-CM
-e CO OS TjH OS CO tJH i-j CO CO OS CO fH
"l C^ CM © CO O 00 CO CO O ** o °o

(iN C_5 <^S <— ' OO vi CTO i — i ^r O

CMOCOCOCOcoOiCCMCMo
os co -^ -^ co tr- ~*

OS CO ^ CO

pqWHfcfcaj^cc^GQpqjz;

^ o g § 3 « o Q § W m m

CUCUOCPCDa>CUCUCUCUCPCD

oooooooooooo

Vol. III. No. 1] FOUNDATION CONSTRUCTION: CLAUSEN 121

to be rather a difficult one to make because the traffic of the
downtown district made progress slow and frequently dis-
turbed the set up of instruments, while the smoky atmosphere
made sighting difficult. The survey, as shown in Fig. 1, was
finally completed, however, and the latitudes and departures
calculated. The error of closure, as may be seen from the
accompanying table, was found to be 0.08', or one inch, east
and west; and 0.06', or three-quarters of an inch, north and
south . Under the existing conditions this was considered to

- Buildinc

/ &1 / / />? '"" "— Lf-\: x J _

O- -O

rvladisori 5t

Fig. 2. Plan of Foundation Caissons for Proposed Commercial Building.

be a good closure, and so the line of the tunnel with reference
to the building lot was then figured . The foundation plans
were next completed, as shown in Fig. 2— the caissons along
the tunnel being planned to clear the outside of the tunnel
brick-work by at least three feet. The soil through which the
tunnel is built is medium stiff blue clay, and, as the internal
hydrostatic pressure in the tunnel was about 20 pounds per
square inch, it was considered necessary to make the excavation
of the tunnel caissons from a point 8 feet above the top of the
tunnel using 20 pounds air pressure. The other caissons were
to be excavated by ordinary means, i. e., under atmospheric
conditions. The column loads over the tunnel were to be

122

THE ARMOUR ENGINEER

[Jan., 1911

supported by steel box girders spanning the tunnel and were
to be completely surrounded by concrete.

Work was commenced on a number of the common cais-
sons, the tunnel caissons being left for the installation of the
air locks. Caisson No. 22, the first tunnel caisson attempted,
was excavated by ordinary means to a depth of about 8 feet
above the top of the tunnel. At this point excavation ceased

and an air lock, constructed about as shown in Fig. 3, and
located as indicated in Fig. 4, was installed. After the air
lock was ready for service work was again resumed and the
excavation carried to rock, the caisson then being filled with
concrete up to the air lock under air pressure, and finally to
the surface, after the removal of the air lock. In order to ex-
pedite the work, another air lock was constructed for use in
the other tunnel caissons, this lock being different from the

Vol. III. No. 1] FOUNDATION CONSTRUCTION: CLAUSEN 123

one shown, in that instead of bolting the top hood, as it were,
to the lower plate securely bedded into the clay, the lock con-
sisted simply of plates similar to the lower plate of the one
shown, with the space between the plates being lined with
the usual wooden lagging of the caisson. This latter air lock
was installed at about the same depth in Caisson No. 15 on the
opposite side of the tunnel. When the excavation under air

Fig. 4. Typical Construction (with Air Pressure) of Caisson Near Tunnel.

pressure in this caisson had proceeded to an elevation corre-
sponding to the springing line of the tunnel, the well digger
exposed the brick-work of the tunnel at the southeast edge of
the caisson and the water slowly began to come in. He im-
mediately came up, and the well, or caisson, was allowed to
fill up with water. After an investigation, during which
boring holes were sunk to locate the tunnel, it was found that
the tunnel was located 3 feet northwest of the supposed or

124 THE ARMOUR ENGINEER [Jan., 1911

surveyed location, and that the caisson in question had ex-
actly struck the tunnel tangentially.

It may here be stated that the clay soil, through which
the caissons were excavated, was full of sand seams and sand
pockets so that often the air pressure would suddenly drop
from 20 pounds to 5 pounds, this creating consternation among
the engineers in charge. This unfavorable circumstance,
coupled with the slow and cumbersome method of progress,
made the excavation under air pressure anything but popular.
The plans were accordingly modified so that the caissons on
the northwest side of the tunnel were moved over 3 feet, this
necessitating much heavier and longer box girders for span-
ning the tunnel. Due to the fact that the caissons on the
southeast side of the tunnel would now be 6 feet instead of
3 feet away from the tunnel, it was decided to sink these
caissons by ordinary means, except that in passing the tunnel
only one-half of a section (or 21/£ feet) should be excavated
before being "lagged up," the usual section being 5 feet
deep. This method was successfully employed on all the cais-
sons on the southeast side of the tunnel.

It may be of interest to digress for a moment from the
subject in hand and relate an experience in one of these cais-
sons which for a while gave the writer no small scare. It
was about 1 a. m. — the writer being stationed on the work
for twenty-four hours or more at times when any one caisson
was being excavated past the tunnel, and the excavation in
Caisson No. 30 was down to a point about opposite the top
of the tunnel. It had taken the well diggers a whole shift of
8 hours to set in place the last set of lagging and steel rings,
due to the extreme swelling of the clay, and of course this
caused a good deal of apprehension on account of the proxi-
mity of the tunnel. Now, when the new shift of well diggers
had excavated about a foot below the lagging, a stream of
water suddenly broke through the clay with great force,
striking one of the diggers in the face. The men were
quickly hoisted up away from danger, for it was supposed
that the water from the tunnel had broken into the caisson.
By means of a float attached to a steel tape, the writer
discovered that the influx of water diminished instead
of increased, finally ceasing when there was about 10 feet
of water in the caisson ; this proved of course that the water
had not come from the tunnel. After an investigatioL it was
found to have come from Caisson No. 29, which was only 4
feet east and which had been excavated and concreted up

Vol. III. No. 1] FOUNDATION CONSTRUCTION: CLAUSEN 125

some two or three months previous. The top of the concrete
in this caisson was about 14 feet below the surface, and, being
a low point, rain water had accumulated in it to a depth of
about 10 feet. When Caisson No. 30 had reached the depth
stated, the hydrostatic pressure from this water was sufficient
to break through four feet of clay and drain the water from
the higher to the lower level.

For the caissons on the northwest side of the tunnel it
was decided to excavate in 2% foot sections as before, but
to employ a steel shield about 3 feet long with a cutting edge,
thus leaving no exposed excavation at all. The cutting edge
was jacked down 2% feet when a set of lagging and steel
rings would be inserted inside of it before proceeding with the
next section. This method was carried out in Caisson No. 35
but was found to be so slow and cumbersome that it was
abandoned in favor of the others. It happened in Caisson No.
35 after hard pan had been reached, and the full depth of the
caisson excavated, that while it was being belled out to give
greater bearing area, water broke in from the tunnel and
filled it. thus making it necessary to concrete it under water.
This leak was attributed to the clay drying up and cracking
as a result of the long exposure to the air, this long exposure
being necessary when using the shield. The shield was there-
fore discarded and the work satisfactorily completed, except
in Caisson No. 1, by means of 2% foot sections as used on
the southeast side of the tunnel. In Caisson No. 1 water came
in, as in Caisson No. 35, while the belling-out was in progress.
Due to this technical violation of the contract, the owner re-
fused to accept the foundation as built, so the city was com-
pelled to purchase the lot, which it still owns and uses as a
storage yard for the Bridge Department. Caisson No. 15,
where the tunnel was actually struck, was filled up for 8 feet
with sand and cement, in bags, and then with clay to the
surface.

This unfortunate circumstance cost the city a good deal
of money and so it was decided that it would be economy to
build under the streets belonging to the city another tunnel,
this to replace this old cross-town tunnel, and thus avoid any
further difficulties with foundation construction. This was
accordingly done and the Blue Island Avenue tunnel now
has replaced it.

It was the good fortune of the writer, about a year ago,
to have the opportunity of walking through this old cross-
town tunnel from Van Buren and Jefferson Streets to Chicago
Avenue and Lincoln Parkway. An examination showed no

126 THE ARMOUR ENGINEER [Jan., 1911

cracks and no traces whatsoever of any damage done to the
tunnel at the point where it was encountered in Caisson No. 15
several years before. This inspection also revealed the fact
that the fire shaft at Market and Madison Streets was off the
center line of the tunnel by about 2% feet to the southeast,
which accounted for the error in tunnel alignment as obtained
from the survey. It seemed a pity to be compelled to abandon
this old tunnel, because the examination showed a perfect
piece of work in an excellent state of preservation after 40
years continuous service. The line and grade of the tunnel
was also perfect, and the depreciation from any cause what-
soever, except in the top of the working shafts, was, I should
say, nothing.

It is now the intention of the city to use the property
purchased and herein discussed as the site for a large central
police station and fire engine house.

THE ARMOUR ENGINEER

The Semi-Annual Technical Publication of the Student Body of
ARMOUR INSTITUTE OF TECHNOLOGY.

VOL. Ill CHICAGO, JANUARY, 1911 NO. I

Publishing Staff for the year 1911:

C. W. Binder, Editor.

G. H. Emin, Business Manager. L. H. Roller, Assistant Editor.

M. A. Peiser, Associate Business Manager.

Board of Associate Editors:

H. M. Raymond, Dean of the Engineering Studies.
L. C. Monin, Dean of the Cultural Studies.
G. F. Gebhardt, Professor of Mechanical Engineering.
E. H. Freeman, Professor of Electrical Engineering.

Published twice each year, in January and in May.

Publication Office: Thirty-third St. and Armour Ave., Chicago, 111.

TERMS OF SUBSCRIPTION:

The Armour Engineer, two issues, postage prepaid $1.00 per annum

Single Copies 50 cents

EDITORIAL.

Realizing that in the past much has been said and written
in an effort to bring to the attention of both our student
and alumni bodies the aims and expectations of THE ARMOUR
ENGINEER, we hardly believe that any further discussion
or information regarding the scope of this publication will
be particularly interesting to our readers at this time. How-
ever, we now wish to express to our contributors our apprecia-
tion of their painstaking efforts in the preparation of these
articles, for we feel that their substantial support will go far
toward increasing the general interest in our magazine. Also,
for the benefit of those who still expect that a special solicita-
tion is necessary, we want to repeat the now standing invita-
tion to all Armour men for contributions to our columns on
any of the engineering topics in which they are particularly
interested.

128 THE ARMOUR ENGINEER [Jan., 1911

In the article appearing in this issue on "Briquetted Coal
and Its Value As a Railroad Fuel," Mr. Malcolmson discusses
a subject which we believe will prove of exceptional interest
to our readers, it being so closely allied to the present and
familiar subject of conservation and development of our
national resources. A very instructive account of the coal
briquetting industry from its earliest days is given, together
with present conditions in its development; the article conciud
ing with a summary of the many advantages of briquetted over
raw coal.

The electric-resistance furnace, which several years ago
gave the graphite industry such an impetus, is now rapidly
being extended in its application to the numerous industries
in which exceedingly high temperatures are required, and we
have from Mr. Badger one of the first of a number of articles
on the various forms of these furnaces, together with an ac-
count of their applications to the heating of different metals
and other refractory substances.

The many advantages of the electric-motor over the steam-
engine operation of rolling mills are carefully brought out in
the article appearing in these pages on "The Electric Driving
of Rolling Mills." In this discussion is also contained an answer
to the question so frequently asked by those not very familiar
with electrical power transmission, as to the advantages of
alternating current in the transmission of large amounts of
energy over long distances. Another item of particular in-
terest is the strong argument in favor of steam turbines over
the gas engines as the prime movers for the generators, al-
though at first glance it would seem that the direct conversion
of blast furnace gas into mechanical and then into electrical
energy would be more economical if the intermediate genera-
tion of steam were not involved. However, a careful con-
sideration of the points brought out in Mr. Dean's article will
show that, everything considered, the steam turbine has a
great many "talking points,"

Vol.. 111. No. 1] EDITORIALS 129

The Pneumatic Ash Handling System as discussed in this
issue by Mr. Harris offers a very simple, economical, and
comparatively new solution to the problem of ash handling in
present day boiler rooms.

Owing to the rather recent development of this system,
the calculations entering into the various designs have not
been gone into until they shall have been confirmed by data
from actual tests ; so the author has written principally on
the general features encountered in design and installation of
these svstems.

With the extremely rapid application of water poAver to
the production of electrical energy, there has grown up in
recent years a new field of engineering activity in which the
demand for trained and experienced men is far greater than
the supply. Realizing this, and the need of giving to the tech-
nical student desirous of fitting himself for work in this sphere
of engineering a course of thorough instruction in the funda-
mentals, as well as in many of the details in the successful
design and construction of hydro-electric plants, there is
now being offered at A. I. T. a course in hydro-electric
engineering. As this name indicates, and as a glance at the
subjects taught confirms, no really new course has been
created — simply a combination of subjects from the hydraulic
and electrical courses effected.

It is doubtful if many of the hydro-electric engineers of
today ever had the opportunity of pursuing courses of study
containing a combination of these two branches, for the reason
that up to within a comparatively few years very distinct lines
have separated the hydraulic and electrical branches. Now
however, we see in the hydro-electric course where much has
been done toward closing the gap existing between the two
separate courses just mentioned, and in view of this fact be-
lieve that graduates now entering this field of work are
equipped with a foundation which will soon enable them to
hold their own in matters of design and construction.

130 THE ARMOUR ENGINEER [Jan., 1911

That a course in hydro-electric engineering should be es-
tablished in all large technical schools is evidenced by the
prejudice and suspicious attitude shown toward many of the
proposed hydro-electric developments of the present day ; and
the reason for this existing suspicious attitude may be attrib-
uted largely to the results of unintelligent engineering and
management on the part of men who really are not quali-
fied by their previous training and experience to be put in
charge of work requiring such a broad training. While
there are comparatively few instances of absolute failures
which can be laid directly to the miscalculations of the en-
gineers in charge, yet there are many instances of where in-
sufficient preliminary investigations and calculations have per-
manently reduced the efficiency of what might otherwise have
been verjr successful power developments.

The necessity of a broad insight into the many phases of
a successful development of our water power resources has
been mentioned, and in an attempt to show this we might add
that it is just as essential for an engineer connected with
hydro-electric work to have a knowledge of the legal, financial,
and commercial aspects of the problem, as it is of the engineer-
ing features. In fact many of the problems of design and con-
struction admit of much easier solutions than do those in
financial and commercial matters, and the ability to grasp them
will often settle in a week what might otherwise take the
strictly technical man years of investigation to determine in
regard to the feasibility of a given project.

This, then, would also seem to present a good argument in
favor of even more instruction than is now given the technical
student along the lines of work not usually supposed to enter
into the work of the engineer and yet so often connected with
it ; and doubtless when engineers become more familiar
with those subjects outside of their own sphere much will have
been done toward giving to the engineering profession the
rank among other professions to which its achievements show
it is entitled.

Vol. III. No. 1] ENGINEERING SOCIETIES. 131

CIVIL ENGINEERING SOCIETY.

The Civil Engineering Society is today the most pros-
perous of the several engineering societies here at school, this
being due in part to the interest taken by the upper classmen
of the Department of Civil Engineering, and. in part, to the
co-operation of the department's faculty, all of whom are
members — Prof. Phillips, Associate-Prof. Wells and Assistant
Prof. Armstrong, as honorary members, and Messrs. Dean and
Penn as Senior members. The active membership of the society
at the present time is about fifty.

Tbe first meeting of the year was held on Tuesday, October
2o, 1910 in the Engineering Rooms, Chapin Hall, the speaker
of the evening being Dean L. C. Monin, whose subject was
"Standards of Professional Conduct." Dean Monin especially
emphasized that an engineer has duties to four parties — (1)
his client, (2) himself, (3) his fellow engineers, and (4) the
public. The several codes of professional ethics now adopted
by several of the older professions were mentioned, and the
hope expressed that the engineering profession soon adopt
such a code. The necessity of membership in engineering so-
cieties was discussed. Dean Monin emphasizing the fact that
the engineering student should start by joining the society
here at school of whatever branch of engineering he was par-
ticularly interested.

On the evening of November 8, 1910, the Society was ad-
dressed by Mr. Henry W. Clausen. Class of '04. Assistant En-
gineer in charge of pumping stations and tunnels for the city
of Chicago. Mr. Clausen gave some "Hints to Young En-
gineers," and in his talk, as well as later by answering ques-
tions put to him. made his audience acquainted with many
practical suggestions and short cuts on the job, especially as
relating to street, tunnel, and road improvement work.

"Irrigation in the West" was the broad title of Mr. Frank
A. Coy's talk on Tuesday evening, November 22, 1910. Mr.
Coy. also an Armour graduate of the Class of 1904. has re-
cently been engaged in work on several irrigation projects
in several of the western states. The reasons for these proj-
ects, preliminary proceedings necessary to get them under
way. the methods of design, and actual construction, were all
taken up in detail and discussed by Mr. Coy.

The last meeting of the calendar year was held on De-
cember 6, 1910, with Mr. George H. Bremner. Engineer of the
Illinois District, Chicago, Burlington & Quincy Railroad Com-

!32 THE ARMOUR ENGINEER [Jan., 1911

pany, as the speaker of the evening. His subject was : "The
District Engineer on a Eailroad: His Duties and How He
Performs Them." Mr. Bremner is well qualified to speak on
that topic, and many phases of a District Engineer's work were
taken up. Organization and standardization of work were
especially emphasized as being essential to success in such a
position. Blue-prints of various C. B. & Q. R. R. Co.'s stand-
ard constructions in track work, construction plans for their
large freight yard at Galesburg, 111., and plans for several
bridge sites, were shown. '

A new feature of the work this year is the attempt to
interest the alumni of the Department of Civil Engineering
in the society and its work by keeping them posted regarding
its meetings. Many of them, of course, cannot be reached, but
a large number of those who have been communicated with
have accepted the invitation to attend the meetings. This is
indeed gratifying and it is to be hoped that this interest of the
alumni will continue to increase.

O. R. ERICKSON.

MECHANICAL ENGINEERING SOCIETY.

The Armour Institute Student Branch of the American So-
ciety of Mechanical Engineers has progressed thus far this
year with pronounced success, meetings being held on the first
Wednesday of each month. While it is desired to select some
talent from the engineering profession at large to lecture at
these meetings, the main purpose of the society is to call upon
its members for papers and lectures on engineering subjects,
in order that they may subdue the reluctant attitude often
manifested when called upon to speak, and to acquire the
ability to lecture in public without hesitancy.

The "Annual Smoker" of the Society was held in October,
and the first lecture given in November by Mr. J. C. Peebles,
whose subject was "Simultaneous and Automatic Control of
Coal and Air Supply to a Boiler Furnace." At the December
meeting Mr. R. B. Ambrose (member) lectured on "Producer
Gas and Some of Its Methods of Manufacture." Each meeting
was attended by about thirty-five persons, constituting stu-
dents and members of faculty. The men who lectured gave
very interesting discussions, and the society feels grateful to
them for their efforts.

On January 4th, 1911, Mr. C. E. Sargent, M. E., of Chi-
cago, delivered an illustrated lecture on "Gas Engines." The

Vol. III. No. 1] ENGINEERING SOCIETIES. 133

meeting was held in Science Hall, and the anticipation of a
large attendance was fully realized. Mr. Sargent easily indi-
cated that he is a pioneer authority on this subject, and for
his kindness the society feels very thankful.

The present membership of the society numbers about
twenty-five, and any Junior or Senior Mechanical student is
eligible to membership. The local society, being affiliated with
the American Society of Mechanical Engineers, renders it pos-
sible for the members to procure proceedings of the above
society and to become members of its student branch.

C. E. BECK.

CHEMICAL ENGINEERING SOCIETY.

Students of the Chemical Engineering Society are very
active in the affairs of their society this year and have
taken exceptional interest in all the meetings held so far. The
objects of this organization are. of course, substantially the
same as those of all engineering societies; namely, to create
a general atmosphere of sociability among its members, and
to become familiar with the application to modern chemical
engineering practice of those principles studied in the class
room.

The first smoker of the society was held Wednesday, No-
vember 9th. with the three upper classes well represented and
as many faculty members present as could possibly attend.
During the evening Professor Tibbals gave a short talk in
place of Professor McCormack, who was unable to be present;
the remainder of the evening being spent in the conventional
smoker style.

The first regular talk before the society was given by
Mr. F. M. De Beers, an Armour graduate of 1905. now presi-
dent of the Swenson Evaporator Company, on Thursday. De-
cember 8th. Mr. De Beers discussed and explained the many
factors entering into the design of evaporators for various
substances, and clearly showed that this work requires the
services of men who. as it were, are a combination of chemist
and mechanical engineer, a combination to be looked for in
the chemical engineer.

H. SIECK.

134 THE ARMOUR ENGINEER [Jan., 1911

ARMOUR BRANCH OF THE AMERICAN INSTITUTE OF
ELECTRICAL ENGINEERS.

Recognizing the advantages to the student body of meet-
ings for the reading and discussion of professional subjects,
the Armour Branch of the A. I. E. E. was organized for the
reading and discussion of papers as published in the "Trans-
actions of the American Institute of Electrical Engineers,"
and for the preparation, presentation and discussion of orig-
inal papers by members of the organization and other in-
dividuals.

The policy pursued so far this year has been to have
original papers presented by members of the society, believing
that the new members would feel more disposed to take part
in the discussion of the subject if presented by a fellow student.
The remainder of the school year will be devoted to papers
presented by members of the Faculty and practicing engineers,
thus giving those of the society who complete their course this
year an opportunity of discussing the practical side of en-
gineering work with men of experience.

The Armour Branch held its first meeting of the school
year on October 27, 1910, at which Mr. L. L. Williams read a
paper on "Car Lighting," in which he pointed out the ad-
vantages of electric illumination in cars. Following this came
a description of several present day methods of car lighting,
after which the advantages of the "head end," the "axle-
light" and storage battery systems were discussed. The main
part of the paper, however, took up and explained the various
parts of the Bliss system, which uses a generator driven from a
car axle.

On November 17, 1?..,, Mr. G. E. Williams read a paper
before the society on the "Otis Electric Elevator Control."
He discussed the duties and early development of electric
elevatjrs, and by means of slides and blue prints illustrated
the form and principle of the Otis control. The wiring dia-
grams were traced, the operation of each explained, and the
merits of the safety appliances discussed.

Following the plan of the society, the third meeting, held
Dec. 15, 1910, was given up to the discussion of the "Inter-
poles in Synchronous Converters," as published in the "Trans-
actions of the A. I. E. E." The subject was divided amongst
the members and each assigned the preparation of a portion
of the paper.

Vol. III. No. 11 ENGINEERING SOCIETIES.

Mr. James H. Jacobson, Engineer Inspector. Hoard of
Supervising Engineers, City of Chicago, addressed the society
January 5, 1911. on "Railway Converter Sub-stations." By
means of photographs projected on a screen, Mr. Jacobson
explained the construction and equipping of sub-stations from
the time earth is turned for the foundation until the last ma-
chine has been installed.

.7. II. FLETCHER.

Of,

'THE CAR WITH THE GUARANTEE'

$1700

You select this car because there is more real, honest value in exchange for your
money than you will find in any other car on the American market today —

The Armour Institute of Technology one of the greatest schools of Engineering
in the world, purchased a 19..1 Model Halladay 40 H. P. chassis after an exhaustive
investigation, because in their judgment the Halladay is the best.

If they don't know — who does?

The Halladay is best on account of its simplicity in construction, accessibility for
repairs, workmanship, quality of material and flexibility. What more — besides a
guarantee insuring you of the protection you are entitled to? Our guarantee is the
original — real — bonafide guarantee— no ifs — buts — providings or legal jokers.

GUARANTEE— READ IT Prof. Gebhardt's Letter-READ IT

GUARANTEE
Halladay Motor Company

1421 Michigan Avenue. Chicago

Mr Address

We guarantee for the season of 1911 Halladay

Car No Engine No

to be free from defective material and imperfect
workmanship. We also agree to keep this car in
repair and proper adjustment without cost to the
owner, providing the car is kept properly oiled
and supplied with a sufficient amount of water; also
that the owner will bring to us this car for our in-
spection at least once each month during the
season.

This guarantee does not cover leaky radi-
ators, caused by water freezing, nor any
damage resulting from accidents or abuse.
It is. however, understood that we make
no warranty whatever regarding pneumatic
tires, coils, magnetos or batteries, com-
plaints regarding which must be taken up
with their respective makers, who fully
cover the same with a sufficient guarantee.
Signed the day of 19

Armour Institute of Technology

Chicago
F, W. Gunsaulus, President

Oct. 19, 1910
Halladav Motor Company.

1421 Michigan Ave.. Chicago. 111.

Gentlemen: — In reply to vour favor of the 15th
inst. beg to state that the Halladay 1911 "40"
chassis recently purchased by us is to be used in
our Mechanical Engineering Laboratory for ex-
perimental purposes.

A number ot types of machines were examined,
but the Halladay 1911 "40" proved to be the best
suited for our purpose on account of its simplicity
in construction, accessibility for repairs, flexibil-
ity for experimental purposes and all-around good
workmanship.

The machine is to be equipped with power,
transmission, absorption and traction dynamo-
meters, optical power indicator, speed indicating
and recording devises, special apparatus for re-
cording the shock-absorbing properties of the
tires and springs — in fact, a full complement of
special appliances for studying the action of all
working parts under various conditions of opera-
tions, Yours very truly.

G. F, Gebhardt,

Agents— If you expect to add to or change your line, do it now. Nine models, $\ 100
to $2650. Manufactured by STREATOR MOTOR CAK CO., Streator, Illinois

HALLADAY MOTOR CO.

CHAS. M. HAYES, President

1431 Michigan Ave., Chicago

Calumet 2574

—17—

THE ARMOUR ENGINEER

THE SEMI-ANNUAL TECHNICAL PUBLICATION

OF THE STUDENT BODY OF

ARMOUR INSTITUTE OF TECHNOLOGY

CHICAGO, ILLINOIS

VOLUME III. NUMBER 2

MAY 1911

G. H. EMIN

THE ARMOUR ENGINEER

VOLUME III. NUMBER 2

MAY 1911

EFFICIENCY TESTS OF SCHENECTADY POWER CO.'S
HYDRAULIC TURBINE UNITS NOS. 2 & 3.

By STANLEY DEAN, C. E.*

The Schaghticoke Hydro-Electric Development of the
Schenectady Power Co. is located on the Hoosic River at
Schaghticoke, N. Y., twelve miles X. E. of Troy. X. Y., and 22
miles east of Scheneetady. X. Y. At this point the Hoosic
River bends in the form of a letter ilS" in a distance of about
two miles, measured along the stream, and in this length it has
a fall of approximately 150 feet. The stream flow and head
are used to develop 20,000 H. P.. a small part of which is used
locally for lighting and motor service, but the greater part of
which is transmitted to Schenectady.

In brief, the development consists of a solid concrete spill-
way dam built diagonally across the river at the head of a
series of falls and rapids, and an intake at the down stream end
of the dam leading to a canal 2,300 feet long cut through earth
and rock, ending in a forebay from which the water is led by
means of a circular steel pipe across a bend in the river to a
circular steel surge-tank on the opposite bank.

From the surge-tank four main unit steel penstocks each
six feet in diameter and one exciter penstock two feet in di-
ameter, lead to four 5000 H. P. vertical-shaft water-turbines
and to two horizontal exciter turbines, respectively, located in
the power house, from which the water is discharged into the
river below the foot of the last rapids.

Each of the four main units consists of a "Pelton-Francis"
inward and downward flow type of reaction turbine, designed
for a full load output of 5000 H. P. at 300 R. P. M. under a
head of 146 feet, and mounted on a vertical shaft directly be-
neath its alternator. Each alternator has a full load output of
3000 K. W. at full load at 4100 volts. For transmission to
Schenectady this is stepped up to 32000 volts.

It is the purpose of this paper to describe the apparatus
and methods used in tests to determine the efficiency of two
of the main water turbines.

Class of 100r>. Instructor in Civil Engineering, Armour Institute of Tech-
nology.

THE ARMOUR ENGINEER

[May, 1911

Fig. 1. General Plan of Schaghticoke Plant of the Schenectady Power Co.

Vol. III. No. 12| TURBINE EFFICIENCY TESTS : DEAN

Previous to these tests an attempt had been made to meas-
ure the water discharged from the wheels, by means of a sharp
crested weir built across the race. Owing to the shortness of
the tail race however it was found that an accurate result
could not be obtained in this manner on account of the eddies
from the draft tube causing a varying head over the weir.
After various methods bad been tried to obtain an adjustment
of the water surface to eliminate the sources of error, the con-
clusion was reached that some other method must be used. At
this time an accomodating flood came along and carried off
weir and gauges, leaving only the abutments standing, and
thus settling that method of testing. It was then decided to
install a pitot tube apparatus in the penstock feeding the
wheel to be tested, which was accordingly done.

Description of Pitot Tube Apparatus for Measuring the Velocity in
the Penstock

If a straight tube be bent at the end to form a right angle
and the tip submerged in a flowing stream and so pointed that
the mouth of the tip is directly opposed to the current, the wa-
ter will rise in the upright part of the tube to a height above
the water surface which is theoretically equal to v2/2g, the ve-
locity head of the stream. If, noAv. a pitot tube be inserted in a
pipe containing water flowing under pressure, the mouth of the
tip being parallel to the axis of the penstock and opposed to the
direction of flow, and a straight pipe be inserted in the edge
of the penstock so that its mouth is normal to the axis of the
penstock and direction of flow, water will rise i?i the stem of
the pitot tube to a height of h -\- v2/2g equal to the sum of the
pressure and velocity heads in the penstock, while in the stem
of the straight tube the water will rise to a height of "h"
equal to the pressure head in the penstock. The difference
in height of the water columns in the two stems will be
eqval to v2/2g and will represent the velocity head in the pen-
stock at the tip of the pitot tube. To measure a head of
approximately 150 feet it would be necessary to have a ver-
tical stem 150 feet high, but we may confine the heights
within reasonable limits by forcing the water column down,
by connecting a source of compressed air to the top of the
tube. This was done in the test (see Fig. 2) where the distance
A-B equal v2/2g, the velocity head at point of pitot tube.

At a point on the penstock about thirty feet uphill from
the elbow at rear of powerhouse, shown on Fig. 5. the pitot
tube apparatus was installed to measure the velocity at stated
points in the cross section of the penstock. The apparatus

142

THE ARMOUR ENGINEER

[May, 1911

(see Fig. 2 and Fig. 3) consisted of two duplicate pitot tubes,
shown in detail in Fig. 6, arranged to slide horizontally and
vertically through stuffing boxes screwed into the shell of
the penstock, and so graduated that the point of the pitot
tube could be set at any desired position on the horizontal
and vertical diameters (see Fig. 7) of the penstock. The
stem of each pitot tube was of brass and about seven feet

Fig. 2. Elevation of Gauge Board, Penstock, Pressure Tubes and Pitot Tubes.

long. To the end of the stem was securely clamped one end
of a flexible hose about twelve feet long, the other end of
which was passed over and clamped to the lower end of one
of the exterior vertical glass tubes shown on gauge board in
Fig. 2, the vertical pitot tube being connected to the glass
tube on extreme left and the horizontal pitot tube to the
one on extreme right. About one foot uphill from the pitot
tube cross-section, eight small iron pipes were tapped in to the

Vol. Ill, No. 21 TURBINE EFFICIENCY TESTS: DEAN

14?,

penstock at points equidistant from each other around the
circumference, i. e., separated from each other by 45 degree
angles, and bent around in easy curves to connect by short
lengths of rubber hose to the eight vertical glass tubes on
gauge board, numbered 1 to 8 on Fig. 2, between the pitot
tube gauge glasses. Each of these eight pressure gauge glasses
and two pitot tube glasses were connected by rubber hose and
iron "tee" sections of pipe to form a horizontal header at

Figr. 3. Section and Elevation of Gauge Board, Penstock,
Pitot Tubes.

ressure Tubes and

top of gauge board. One end of this header was connected by
rubber pressure hose to the compressed air storage tank located
in the power house. At the other end of the header was placed
a blowoff cock to let out excess compressed air. Pet cocks
were placed at the top of each pitot tube, at the connection
of each pressure pipe to penstock, and at the base of each
glass gauge tube on the gauge board. One pet cock was also

144

THE ARMOUR ENGINEER

[May, 1911

connected in between the header and compressed air hose
to admit or cut off the air pressure. All gauge glasses were
securely fastened in position on gauge board by clamps and
screws. Immediately behind the gauge glasses and forming
the background of the board was pasted a sheet of cross sec-
tion paper graduated in inches and tenths. The gauge board
was mounted on a timber platform and securely braced into
position as indicated in Fig. 3. To insure true position of the
pitot tubes and to support them, planks were placed parallel
to and immediately beneath same. These planks were gradu-
ated to correspond to the numbered points shown in Fig. 7. In
order to keep the point of pitot tube from being bent down-
stream by the flow of water in the penstock, 2"x%" iron guides

"TTv f\rmaur Fnoineer:
Fig. 4. Plan of Penstocks, Surge Tank and Power House.

at right angles to each other were securely bolted together
and to the penstock immediately behind the line of travel of
the pitot tubes, to support same. The detail of these cross
braces and guides is shown in Fig. 8.

The detail of the tip of pitot tube shown in Fig. 6 is
worthy of note. The tip proper was of brass, three inches
long, accurately bored, and smoothly finished to the dimen-
sions shown, and was screwed on to the stem of the pitot tube.
The mouth of the tip was 14", which dimension decreased
to 3/32" at the throat and then enlarged to %" again, the
inside diameter of the stem and glass gauge rods. The object
of this contraction at the throat was to reduce the surging
of the water columns in the glass gauge tubes due to small

Vol. Ill, No.

!] TURBINE EFFICIENCY TESTS: DEAN

145

changes in the velocity of water in the penstock during the
test periods and to render the time of surge the same for hoth
up and down motions, thus enabling the observer to read the
gauge with greater accuracy.

Description of Apparatus for Measuring the Effective Pressure Head
on the Turbine

As stated in the brief description of the plant at the be-
ginning of this article, the water was led from the impounding
reservoir by means of a canal and single conduit to a surge
tank in which it was allowed to rise to the level of the hydrau-
lic gradient. From the surge tank the feeder pipes led di-
rectly to each unit at the poAver house. The effective head

FI.-33I

The f^fmowr ^nairteer*.
Fig. 5. Vertical Section, showing Penstocks, Surge Tank and Power House.

on the turbines when running was therefore the difference in
level between the water surfaces of the surge tank and tail
race, minus the loss in head between the surge tank and the
entrance to the scroll case, due to entrance loss at connection
of feeder penstock to surge tank, and to bends and friction in
penstock, plus the velocity head of the water at the entrance
to scroll ease. In such a case all losses in the scroll case,
guides, vanes, and draft tube are considered as the hydraulic
losses in the turbine unit itself, and therefore are not to be
deducted from the effective head as stated above.

If we measure the actual pressure head at the entrance
to the scroll case and add to this the difference in level be-
tween the point of measurement and tail race, we have the

THE ARMOUR ENGINEER

[May, 1!)11

total pressure head. Adding to this the velocity head at this
point — i. e., at the point of measurement of pressure head,
we obtain the total effective head on the turbine. This method
was accordingly pursued. In Pig. 9 is shown diagramatically
the location and relation of gauges for determining the pres-
sure head on the wheel. At two points in the tail race were
located ordinary hook gauges with their zeros set accurately
at elevation 150.0. The gauge reading added to 150.00, thus
gave the exact water elevation of the tail race above datum.

Flexible tt-ose

T3r-3S5 Clamp

"TKe ^rmour Fnoineer,

Fig. 6. Detail of Pitot Tube.

To measure the pressure head on the scroll case, a small
iron pipe was tapped into same, near the center, and led
to a vertical "U" mercury gauge and connected to the upper
end of one of the vertical tubes by a rubber hose clamped
over same.

Between the upright columns of the gauge a steel tape
graduated in feet and tenths, was stretched with its 150.0-
foot mark accurately placed at elevation 165.0. A stop cock

21 TTKBINE EFFICIENCY TESTS: DEAN

shown in Fig. 9 near point of connection of pipe to scroll
case controlled the admittance of water under pressure to the
gauge tubes. Before opening the stop cock, mercury of spe-
cific gravity of 13.54 was poured into the funnel at the top
of the gauge, and of course rose to equal heights in the paral-
lel columns. To measure the pressure head, the stop cock
was then cautiously opened, the full pressure gradually al-
lowed to depress the right hand column and correspondingly
force upward the left hand column. By reference to Fig. 9
it will be seen that the difference between the tops of mercury
columns "Z" read on the tape represented the height of a
column of mercury that just balanced the transmitted water

^11 connections with g botts

The F\

rmour Trr> ameer.

Figr. 1. Cross Seetior
showing Points «i
Traverses.

of Penstock,
i Pitot Tube

Fig. 8. Details of Cross Braces
and Guides fow Pitot Tubes.

pressure from the scroll case at the elevation of the top of
right hand mercury column. Multiplying this difference of
level "Z" by the specific gravity of the mercury (13.54)
gave the height of the hydraulic gradient above the top of the
right hand mercury column. Add to this height (13.54 Z)
the reading "Y" and we have the height of the hydraulic
gradient above elevation 165.00. Adding this reading to
165.00 we have the elevation above datum of the hydraulic
gradient. Subtracting from this latter elevation the elevation
of tail water we thus have the total pressure head on the
turbine. From our observations with the pitot tube on the
penstock we obtain the mean velocity and quantity <>i' water

148

THE ARMOUR ENGINEER

[May, 1911

flowing through the penstock to the wheel. Then q = a^ =
a 2v2, where q= total quantity of water in cubic feet per
second, a,1 = area of penstock at point of velocity measure-
ment, a2 = area of cross section of scroll case at point of
pressure tap, v, = mean velocity at same point, v2 — mean
velocity thru scroll case at point of pressure tap. The velocity
head at this point equals v2/2g, which, added to the total pres-

y(gX+r Svr\ac e m 5*->ra*Tanl*. ^

<L ^Scroll Cns*
(S.o on oaone

^ fli-mour Eoorii

Fig. 9. Diagraniatio Sketch showing Position of Gauge* for Measurement of
Pressure Head on Turbine.

sure head, gives the total effective head on the turbine. The
mercury used in the test was tested carefully by the refiners
and guaranteed to be of a specific gravity of 13.54. In order
to make sure of this figure the mercury column was checked
at the beginning and end of each day's test from the known
static head on the wheel when same was shut down and no

Vol. III. No. 2] TURBINE EFFICIENCY TESTS: DEAN 149

water passing through. Id this case there was no velocity
head, and when the other turbines were shut off the difference
in height between mercury columns multiplied by the specific
gravity of the mercury should give the difference in elevation
between the top of the lower mercury column on the right
hand and the surface of the water in surge tank. Prom care-
ful levels this was checked and the specific gravity of the
mercury found to be correct as given by the refiners/

READINGS OF PRESSURE GAUGE TUBES AND PITOT TUBES, AND
REDUCTION OF VELOCITIES.

Unit No. 2. Test No. 4.

Traverse A (Horizontal)

Calibration of Pressure Gauge 8 in respect of the average readings of gauges

No. 1 2 3 4 5 6 7 S
—1 —4 —1 -4 0—6 +1 0
8)— 1.5

—.2 = C
Remarks. — Iu calibrating pressure gauges we find that gauge 8 reads
.2 high ; therefore we add .2 to the difference, or velocity head, which gives
us corrected difference, or corrected velocity head.

(A— B)

(A— B)±C
Corrected

Velocity

Difference

difference

feet per

Position

Average

second

piezometer

Average

velocity

corrected

2g[(A— B)±C]

reading in

point in

head in

head iu W

pitot

tube Time

inches

19 1 :37

41.50

. 51.40

9.90

10.10

7.35

41.40

52.50

11.10

11.30

7.77

41.20

55.30

14.10

14.30

8.75.

41.10

56. SO

15.70

15.90

9.23

40.90

57.50

16.60

16.S0

9.48

41.15

58.00

16.85

17.05

9.54

41.30

57.00

15.70

15.90

9.23

41.00

59.50

18.50

18.70

10.00

41.15

60.50

19.35

19.55

10.22

36.70

58.70

22.00

22.20

10.90

36.60

59.80

23.20

23.40

11.20

36.90

60.00

23.10

23.30

11.19

36.80

58.40

21.60

21.80

10.82

36.90

57.10

20.20

20.40

10.46

37.40

57.00

19.60

19.80

10.80

37.10

53.00

15.90

16.10

9.29

37.90

53.90

16.00

16.20

9.31

37.50

50.30

12.80

13.00

8.34

1 1:51

38.00

48.50

10.50

10.70

7.56

Table 1.

Readings and Reductions for Horizontal Traverse 4-A.

150 THE ARMOUR ENGINEER [May. 1911

Method r-|

In making the test it was determined to divide the cross
sectional area of the penstock into ten parts of equal area
made up of nine concentric rings or bands and one central
circular area. At the center of each one of these rings read-
ings were taken as shown in Fig. 7, thus giving nineteen read-
ings on each horizontal and vertical traverse. In making
readings with the pitot tube apparatus the eight pressure
gauge tubes on the gauge board were first calibrated with

READINGS OF PRESSURE GAUGE AND PITOT TUBES. AND REDUCTION
OF VELOCITIES.

lnit No- ~- Test No. 4.

Traverse B (Vertical)

Calibration of Pressure Osiim 1 in r..<.,..,.t t,. ti ..Q.,.i!„ , *

readings or gauges

:ure

Gauge 1 in respect t<

> the average

No.

1 2 3 4 5

0 7 8

0 —1 +1 —4 —1

-6 +2 +1

81— .8

— .1 = c
Remarks.— In calibrating pressure gauges we find that gauge 1 reads
.1 high; therefore we add .1 to the difference, or velocity head, which gives
us corrected difference, or corrected velocity head.

(A— B)

(A— B)±C
Corrected

Velocity

Difference

difference

feet per

Position

Average

second

piezometer

Average

velocity

corrected

2g[(A— B)±C]

pitot

head in yj-
inches

tuhe Time

inches

1 1:52

40.10

5G.G0

10.50

10.60

7.54

38.80

51.70

12.90

13.00

8.34

38.70

53.20

14.50

14.60

8.84

38.50

54.20

15.70

15.80

9.20

38.30

56.30

18.00

18.10

9.85

38.30

56.50

18.20

18.30

9. B0

38.10

59.20

21.10

21.20

10.66

37.00

59.60

21.70

21.80

10.80

38.10

59.80

21.70

21.80

10.80

37.80

58.90

21.10

21.20

10.66

38.10

58. (X)

19.90

20.00

10.35

38.00

59.10

21.10

21.20

10.66

38.20

58.70

20.50

20.60

10.50

38.40

57.20

18.80

18.90

10.07

38.90

54.70

15.80

15.90

9.23

38.40

55.60

17.20

17.30

9.02

38.70

53.00

14.30

14.40

8.78

39.20

51.S0

12.60

12.70

8.25

10 2:04

39.40

48.00

8.60

8.70

6.83

Table 2. Readings and Reductions for Vertical Traverse 4-

Vol. Ill, No. 2] TUKBIXE EFFICIENT "V TESTS : I >EAN

respect to the pressure tube adjacent to the pitot tube being
used, and by applying a correction to the readings of this
adjacent pressure tube the average reading of the eight pres-
sure gauges was obtained, it being found by observation that
the pressure gauge columns rose and fell together; thus the
average bore a constant relation to this adjacent tube column.
The observer then read simultaneously the columns Ph and S.
if the horizontal pitot tube was being operated, or Pv and 1 if
the vertical pitot tube was being operated. Two observers
working together checked each other's observations, these

M?i^iir?

?•■>•• ;: ■•■?•■•*•'; ■; f^.f~-fni:

Fig. 10. Plot of Discharge Curve.

being called to two recorders who cheeked each other's fig-
ures at the end of each run. The recorded readings (A — B^
-|-C gave the velocity head in inches. These were reduced to
feet, and the velocity, v= V 2gH = V 2g [(A — B) + C] in
feet per second, calculated for each of the nineteen points
on the horizontal and vertical traverses. Values of the veloci-
ties for thirty-eight separate positions of the pitot tubes were

calculated from the observations of each run at a given load
and gate opening. Tables 1 and 2 show in detail the readiugs
taken and reduction of "v" for runs No. 4- A and No. 4-B,

152

THE ARMOUR ENGINEER

[May, 1911

April 29-30, gate — open, load 2700 k.w. In connection with the

80
foregoing and figures No. 2 and No. 7 these tables should be
self explanatory.

Calculation of Penstock Discharge "Q"

In order to obtain the quantity of water "Q" from the
pitot tube readings, a plot (Fig. 10 )of velocities at the several
points on the horizontal and vertical traverses was made, using
velocities as ordinates and areas as abscissae. A curve was
passed through the nineteen points on the horizontal and ver-
tical traverses and the quantity "Q" determined by using
a planimeter to measure the area below the curve, which evi-
dently is equal to 2 av between the limits of the penstock
sides, or the total quantity of water flowing in the penstock.

Measurement of Head

Throughout the test runs readings were taken on the
tail race gauges every fifteen minutes, and on the long and
short columns of the mercury gauge every three minutes, the
readings in both cases being nearly constant throughout each
run. Two observers checked each other and recorded sep-
arately at each point. The readings for test runs 4-A and 4-B
are shown in detail in Fig. 10.

(a)

(b)

(c)

(b— a)

(a-c)

Time

Short

Long

Tail Race