REINFORCED CONCRETE BUILDINGS

Published by the

McGraw-Hill JBooIk. Company

N<ev/ York.

\Succe.s.sons to theBookDepartments of the

McGra\v Publishing Company Hill Publishing Company

Publishers of Books for

Electrical World The Engineering" and Mining Journal

Engineering Record American Machinist

Electric Railway Journal Coal Age

Metallurgical and Chemical Engineering Power

REINFORCED
CONCRETE BUILDINGS

A TREATISE ON THE HISTORY, PATENTS
DESIGN AND ERECTION OF THE PRINCI-
PAL PARTS ENTERING INTO A MODERN
REINFORCED CONCRETE BUILDING

BY
ERNEST L. RANSOME

Assoc. Am. Soc. C. E., Charter Member, W. Soc. E., Hon. Corres.

Member, A. I. A., Member, Royal Society of Arts, President and

Consulting Engineer, The Ransome Engineering Company

AND

ALEXIS SAURBREY

Assoc. M. Am. Soc. C. E., Member, Dansk Ingenior forening, Mana-
ger and Chief Engineer, The Ransome Engineering Company

McGRAW-HILL BOOK COMPANY

239 WEST 39TH STREET, NEW YORK
6 BOUVERIE STREET, LONDON, E.G.

1912

THE -PLIMPTON -PRESS -NORWOOD -MASS -U-S -A

PREFACE

THIS little volume is presented to the engineering profession
for the purpose of showing what reinforced concrete is, and how
it came to be what it is. In deciding upon the scope of the book,
the authors have endeavored to select matters of interest to the
mature and experienced engineer, and for that reason the ultimate
result of the analysis has been treated in greater detail than the
derivation itself.

Many books on reinforced concrete have been written prin-
cipally for the practical man or even for the untrained man; those
who approve of that tendency will not approve of this book, in
which all references to weights and dimensions, earth pressure,
etc., have been avoided. The practicing engineer, the contractor,
or even the college student should look for such matters in special
"pocket books," and he must not expect the reinforced concrete
book to furnish a complete encyclopedia on civil and hydraulic
engineering.

On the other hand, much matter has been included in this
book which will be looked for in vain in other works, and this
is especially true of Part I, where an account of the history of
reinforced concrete has been given, with special reference to the
patents granted by the United States. Naturally, a selection of
the more important or interesting patents is a difficult matter,
and most likely some readers will find that too much has been
included, and others that not enough has been included. How-
ever this may be, the records of the patent office contain
on the whole more and better information than any other
source, and the engineer who is striving to attain perfection
cannot afford to relegate the most valuable thoughts and results
of his predecessors to the scrap heap. It is also hoped that
inventors and patent attorneys may find matters of interest in
the necessarily brief descriptions given; anybody wishing full
information in regard to any patents may obtain copies for

vii

241330

viii PREFACE

a nominal sum by addressing the Commissioner of Patents,
Washington, . B.C.

The theoretical analysis in Part II has been made as brief and
concise as seemed consistent with its purpose; the solutions of
equations, etc., have usually been given after the premises have
been stated, omitting all the intermediate steps. If this book
should find its way into the classroom, the teacher can easily
supply what is omitted; the practical engineer would not stop
to read a treatise on mathematics, in any case. The author be-
lieves that much of this is new and original, especially the use of
two constants in the bending problem; the further development
in Articles 34, 35, and 49, where the effect of an increase in depth
of beam is discussed, and the analysis of stresses in given beams.

The entire chapter on Transverse Stresses and U-bars is origi-
nal, and avoids the use (or mis-use) of the word " shear." In
reinforced concrete the steel is supposed to act in tension, and the
U-bars must follow this general rule.

Part III is devoted to the practical construction. Here again
more attention has been given to the useful facts not generally
known, than to those that are matters of common knowledge.
There is no necessity today for describing at great length the
various types of buildings, or their component parts; an excep-
tion has however been made in regard to "Unit Construction"
which appears to be coming rapidly to the front. No effort has
been made toward giving the details of form design, but the
general principles have been stated with great care. The chap-
ters on fireproofing and repairs should be of interest, and the
superintendents' specifications have proved their own value on
a number of large contracts. Immediately preceding this chap-
ter we have placed a short account of some bad failures; while
we have not been able to throw new light on the causes, we hope
that the perusal of the chapter on "accidents" may put the reader
in the proper frame of mind to not only read, but also follow, the
instructions given in the last chapter.

It has been customary with other writers to describe in more
or less detail the tests made on reinforced concrete beams. As
a general principle we have avoided such discussions, partly
because the plan of this book did not allow us to devote the
required large number of pages, and partly because the vast
majority of tests are of little value, not from want of ability or

PREFACE ix

care in the experimenters, but because the tests were not sys-
tematized, that is, every group should first demonstrate one
general fact, and then individual test specimens should be so
designed that they vary in only one feature from the standard,
so that the effect of the variation at once becomes evident. The
groups of tests so made are few indeed, and only during the last
few years have clear photographs of the broken specimens been
published. However, where a given problem requires the illus-
tration of a test, the best available source has been referred to.

In this book an earnest effort has been made toward stating
the truth when it was known, and to make it clear and evident
that the truth is not known in a number of cases. The chapters
relating to the mathematical design are so arranged that every-
one can readily assure himself of the correctness, but in regard to
such matters as cement testing, rolling of concrete floors while
setting, and numerous other practical or general propositions
where the authors have taken issue with prevailing ideas, and
gone contrary to accepted practice, our statements must either
be rejected as heresy or accepted as doctrine.

The authors desire to acknowledge their indebtedness to vari-
ous papers published in the Engineering Record, the Engineering
News, the American Machinist, and the Cement Age. Informa-
tion has also been gained from a paper read by Geo. W. Percy
before the San Francisco Chapter of the A.I.A. (Feb. 9, 1894);
from C. W. Pasley's "Observations on Limes" (1847), from
Hyatt's " Account of some Experiments" (1877); from papers by
Scott, Bernays, and Grant, edited by James Forest as a separate
volume under the name " Portland Cement" (1880); from " Re-
inforced Concrete in Factory Construction" by the Atlas Portland
Cement Co., and in regard to theoretical questions, from works
by Considere and Morsch. The author of Part II desires to em-
phasize the inspiration received from the study* of these two
authorities, who have contributed so much to the knowledge of
the subject. The discovery of the " water-marks " by Professor
Turneaure has perhaps influenced the U-bar theory here advanced
more than any other tests on record have.

The authors are greatly indebted to Professor L. J. Johnson,
M. Am. Soc. C. E., for certain data relating to reinforced concrete
beams tested at Harvard University. The results obtained
are extremely important and will undoubtedly revolutionize

x PREFACE

current practice in regard to the simultaneous manufacture of
beam and slab. Moreover, these tests confirm in a remark-
able manner the theories advanced in Chapter VII, which
were conceived and printed long before the test beams were
designed.

E. L. R.

A. S.
MARCH, 1912.

CONTENTS

PART I

CONTRIBUTION TO THE HISTORY OF REINFORCED
CONCRETE CONSTRUCTION

CHAPTER PAGE

I PERSONAL REMINISCENCE BY ERNEST L. RANSOME ... 1

The invention of "Ransome Stone" in England, 1844.
Introduction in America in 1870. Early work in Portland
Cement Concrete on the Pacific Coast. The invention of
twisted bars (1884). Reinforced Concrete Buildings in and
near San Francisco. Their behavior in the Earthquake.
The first "ribbed" floors (1889). Effect of retempering of
mortar and of continued mixing. Rolling the floors.
Development of Sidewalk Lights. Joining new and old con-
crete. Concrete made fireproof by adding salt, waterproof
by adding lime; effect of clay in the aggregates. The Re-
inforced Concrete Belt Course, its use and advantages. The
Coil Joint for joining Lapping Bars. Notes on falsework, the
importance of standardization; floor core boxes; unit con-
struction. "Wet" and "Dry" concrete. Injurious Agen-
cies. The growth of Reinforced Concrete Construction.

II BASIC PATENTS, AND A SHORT SURVEY OF THE EARLY HISTORY

OF THE ART BY ALEXIS SAURBREY 18

Elements of invention. Ancient use. Early ideas of the
properties of lime. Smeaton's chemical analysis. Discovery
of "Roman Cement" and "Portland Cement."

THE PERIOD OF DISCOVERY 19

Cement testing a necessity. Brunei's Reinforced Brick
Arches (1834-1838) and Beams. Pasley's Beams. Ranger's
patent and early works. Early English patents. Edward's
Analysis and Patents. Use of Reinforced Concrete in England,
in Germany, in Holland, in France. Early American patents.
- Hyatt's patent of 1878.

THE PERIOD OF IMPROVEMENT 35

RECENT PATENTS 40

xii CONTENTS

PART II

RATIONAL DESIGN OF REINFORCED CONCRETE BUILDINGS
BY ALEXIS SAURBREY

CHAPTER PAGE

III INTRODUCTION 51

Definition; homogeneity; mutual relation of steel and con-
crete; assumptions; principles; the necessity of " bond."

IV ADHESION 54

Laws; anchorage in beams; diameter of rod.

V COMPRESSION AND LATERAL EXPANSION 57

Laws; effect and necessity of hooping; calculation of plain
and hooped columns; least diameter; practical considerations.

VI BENDING 66

Notations. Assumptions. Design. Slab Formulas.
T-beams. Tile-Concrete-Construction. Simplified Formulas
for flat slabs. Discussion of the tables; numerical examples.
Analysis of given beams; numerical examples.

VII TRANSVERSE STRESSES 92

Is a concrete beam solid? -Effect of hair cracks. Effect
of large cracks. Adjustment of stresses to meet the new con-
dition in a cracked beam. The U-bar as cantilever reinforce-
ment. Calculation. Same result obtained in a simpler but
less convincing manner. The equilibrium curve must be
approximated. Moving and stationary loads. How many
bars should be bent up? Anchorage a necessity for rational
calculation. Spacing of U-bars. Shear must be considered
as a component stress resulting from tension and compression
acting simultaneously. Frictional stresses. Beams and
slabs not monolithic. Angle of friction. That the same rule
should be used whether beams are ca& in one piece with the
slab or not. Tensile stresses disregarded. The true assump-
tion. Details of reinforcement; the elements combined to one
beam.

VIII APPLICATIONS OF THE BENDING THEORY 109

Continuity. Moment of Inertia; double reinforcement;
Combined Bending and Compression ; a simple chimney formula ;
footings ; circular reinforcement in plates ; theory of plates.

IX INITIAL AND ALLOWABLE STRESSES 126

Setting in air; in water; wetting dry concrete; shrinkage;
temperature stresses; expansion joints. Assumptions; factor
of safety; allowable stresses fixed by custom. Columns, floors;
other structures; actual safety.

CONTENTS xiii

PART III

PRACTICAL CONSTRUCTION BY ERNEST L. RANSOME
AND ALEXIS SAURBREY

CHAPTER PAGE

X MATERIALS OF CONSTRUCTION 137

Cement : tests not reliable. Storage and use. Quick set-
ting cement. Samples; field tests of concrete. Sand: tests;
standard sand; specifications. Stone: dust; run of crusher;
separate piles; sizes used; furnace slag; boiler cinders, lime
stone; sandstone; brick; shale; conglomerate. Steel : plain
bars; rerolled bars; deformed bars; wire mesh. Tiles.
Concrete: Mixing; water used; deposited informs; supervision;
joints; cold and heat.

XI FLOOR SYSTEMS 152

Monolithic vs. Unit. Design. Monolithic Work. Forms;
reinforcement. Unit Work. Where used; the Ransome Sys-
tem.

' XII FOUNDATIONS 171

Brief description; piling.

XIII FINISHING OPERATIONS 176

Corners; flat surfaces; plastering; brick and terra cotta;
improved surfaces; tooling; rubbing; brushing; left unfinished.
Floor finish.

XIV FIREPROOFING AND FIRES 183

Smoke and water. Insurance. Behavior of concrete; thick-
ness required; protection of corners. Salt retards attack of
fire. Bayonne fire. Pittsburg, Baltimore, San Francisco.
Comparison between different methods.

XV REPAIRS TO EXISTING BUILDINGS 188

Cracks in floors; in beams; repairs of columns; cutting con-
crete; laitance; settling of footings; hardening soft concrete.

XVI ACCIDENTS 191

South Framingham; Long Beach; Rochester; Philadelphia;
Annapolis; Saybrook; Cleveland. Defective column design.
Three good rules.

XVII SUPERINTENDENTS SPECIFICATIONS 195

General; temporary offices and buildings, setting up plant,
etc.; excavating and grading; molds; concrete; steel; finish-
ing; acid joint.

xiv CONTENTS

XVIII THE ENGINEER 200

Reinforced Concrete manufactured in situ. Necessity for
skillful manipulation. Building Ordinances; in Cleveland; in
Boston. Uniform Regulations. Injurious influences. " Cost
plus profit" Contract. Lump Sum Contract. Monthly esti-
mates. Qualifications of the Engineer. Patents.

XIX THE THEORY OF BEAMS AS ILLUSTRATED BY TESTS . . . 207

Extensibility. Shear Resistance. Stirrups. German Tests
on T-Beams. Author's Tests at Case School. Professor
Johnson's Tests at Harvard.

INDEX 233

PART I

A CONTRIBUTION TO THE HISTORY OF
REINFORCED CONCRETE

REINFORCED CONCRETE BUILDINGS

CHAPTER I

PERSONAL REMINISCENCE

BY ERNEST L. RANSOME

WHEN, in 1859, I entered as an apprentice in my father's
factory in Ipswich, England, the concrete industry was in its
infancy, and was confined largely to the manufacture of arti-
ficial stone for ornamental purpose. One of the earliest appli-
cations of the new industry was invented in 1844 by my father
Frederick Ransome, who was then engaged as superintendent
of the well-known Iron Works of Ransomes and Sims at Ipswich.
Noticing one day the waste of good hard stone in the dressing
of mill-stones, he conceived the idea of cementing hard, selected
pieces together, and so to manufacture a superior grade of burr-
stones. The first difficulty was in finding a proper cementing
substance: plaster of paris, shellac, glue, isinglass, lime with
bullock's blood, mastic, etc., were tried and discarded. Among
the numberless ingredients tried were also common glass, but
it was not until experiments with soluble glass were made that
success became probable. It occurred to him that if he took
flint stones with a moderate amount of caustic alkali in solution,
and subjected them to heat in a Papin's digester under high
pressure, he might be able to concoct a soup from flint, as Papin
had done from bones. But the result was apparently a dis-
appointment, and in order to increase the heat, he finally tied
the safety valve with a piece of wire, and forced the fire until
the boiler became overheated. Fearing, however, that the boiler
would blow up, he threw it out into a cistern with cold water,
and the boiler, as might have been anticipated, was broken to
pieces and there, inside, was the glazy, syrupy mass of dis-
solved glass. The portions next to the walls of the boiler were
baked to a flinty hard stone; in one word, the problem was
solved.

2 REINFORCED CONCRETE BUILDINGS

Step by step, a process was now evolved whereby a cement-
ing substance was had, as above described; and based upon this
process, a large business was developed. Before long, the
parent Company, " Patent Concrete Stone Co.," was selling
its product in all parts of the world, especially after methods
had been invented whereby the stones were made not only hard
but also weather-proof. This process consisted originally in the
application of a solution of chloride of calcium to the silicate of
soda previously used, whereby insoluble silicate of lime, and
soluble chloride of sodium were formed by double decomposition.
The latter is common cooking salt and was easily removed by
washing.

A further experiment disclosed the fact that powdered mag-
nesian limestone, mixed with a small quantity of silicate of
soda, formed a very hard substance when submerged in a solu-
tion of chloride of calcium, in a very short time.

In America, the new process was introduced in 1870 by the
Pacific Stone Company of San Francisco, of which Company I
was the superintendent for four years. About this time, the
concrete industry was in slow development on the Coast, based
upon the use of imported Portland Cement; in 1874 I remember
to have paid as much as nine dollars per barrel of cement. But
even as late as 1882, the concrete construction was mainly
utilized in foundations and arches suspended between iron beams.
In the latter type of construction some trouble was experienced
with the cracking of the concrete over the beams, and to over-
come this tendency I patented, No. 263,579 (Figure 1), a con-

FIGURE 1.

struction in which an expansion-joint feature was introduced,
and several sidewalks have been built over cellar areas in this
manner.

Before long, I was called upon to devise a cheaper method
of self-supporting sidewalks for the Masonic Hall at Stockton,
Cal., and this I accomplished by using, instead of the I beams,
a 2" round tie-bolt to carry the tension, while the concrete

PERSONAL REMINISCENCE 3

carried the compression. The rods had upset ends and large
cast-iron washers at each end, and I soon found that the up-
setting and threading of the ends, the nuts and washers, etc.,
made the cost of the finished rod exactly twice that of the plain
rod. I looked around for means whereby a continuous tie or
bond could be developed along the length of the rod, and even
contemplated cutting a spiral groove in the rod, when suddenly
the idea of twisting a square or rectangular bar entered my head.
I happened to have a rubber band in my pocket, and the spiral
thread became at once evident when the rubber band was
twisted in the hand. My patent, No. 305,226 (Figure 2), was
granted in 1884, and the mills were soon turning out
twisted bars up to one inch square, at a cost of about
ten dollars per ton for twisting. Larger bars they
positively refused to tackle under the plea that the
common lathes used for the purpose did not have the
requisite strength. I had, however, in my yard an old
concrete mixer equipped with a worm and wheel, and
by modifying this arrangement I soon succeeded in
twisting 2" square rods, using hand power. The cost
did not exceed seventy-five cents per ton, and from
that date until a more recent period, all the twisting
was done in my own yards.

However, the introduction of the twisted iron was
no easy matter, and when I presented my new inven-
tion to the technical society in California, I was simply
laughed down, the concensus of opinion being that I
injured the iron. One gentleman kindly suggested that
if I did not twist my iron so much I might not injure it seri-
ously, in spite of all my references to the twisting of ropes and
similar devices. This argument I based upon the supposed
fibrous or laminated structure of the iron.

But all this criticism led to exhaustive tests, and when the
professors found that my samples stood up better than the plain
bars, one even went as far as to suggest that I had doctored my
samples. This led me to twist half of each test rod only, and the
superior strength of the cold twisted iron was finally admitted,
and in due time, when steel became common, even better
results were had with cold twisted steel. Even at this present
time, I do not believe that the increase in strength due to the

4 REINFORCED CONCRETE BUILDINGS

twisting has been accounted for. In this connection I call
attention to an interesting fact first discovered by Professor
Hesse, and that is, that bars tested at once after twisting do
not give as good results as those tested five or more days
after twisting, showing that a certain slow change takes place
in the structure of the iron.

From the earliest time of my career I have experimented
extensively with concrete mixers and other machinery, and the
Ransome mixer is now a standard article. However, a descrip-
tion of these experiments would carry us too far, and might
not interest the reader. Suffice it to say that my first patent
for a concrete mixer was granted in 1884, to be followed by
many more.

Up to about 1888 my work in reinforced concrete was
largely confined to what we now term small and unimportant
structures. The Bourn & Wise wine-cellar at St. Helena, Cal.,
was erected in 1888; the building is 75' X 400', three stories
high, with stone walls. The main floor only was of reinforced
concrete resting upon iron columns. The design is shown in
Figure 3. The next floors were erected for the Californian

FIGURE 3.

Academy of Sciences in San Francisco, Figure 4; during con-
struction these floors were subject to much adverse criticism
from many architects, builders, and members of the Society,
and efforts were made to have the fire wardens condemn the
work. However, inspectors from the fire department found
nothing about the construction that could be injured by fire,
and having sense enough to perceive its great strength, they
declined to take any action in the matter. To satisfy all skep-
tics in regard to the strength, a section of the second floor
15' X 22' was uniformly loaded with gravel to 415 Ibs. per
square foot; the deflection was J". For the further satisfaction
of the doubtful the load was left on for four weeks, but very few

PERSONAL REMINISCENCE 5

availed themselves of invitations to examine the work a second
time. The few who came were, however, convinced.

The Leland Stanford Jr. Museum, at Palo Alto, Cal., was

FIGURE 4. ACADEMY OF SCIENCES, SAN FRANCISCO

Reinforced Concrete Floors, Cast Iron Columns

Erected by Ernest L. Ransome

erected about this same time, and the entire wall and floor con-
struction was of concrete, the walls having superficial joint lines
as indicated in my patent, No. 405, 800, of 1889 (Figure 144A).
The outside surface was partly tooled, and the whole was built

REINFORCED CONCRETE BUILDINGS

in the classical design originally made for sand stone. The
greatest innovation was, however, the roof, and this was probably
the first instance on record where a finished and exposed roof
was made entirely of concrete. The roof was supported on iron
trusses 10 ft. on centers and the concrete construction rested
upon the iron rafters, as shown in Figure 5. The roof over the
central pavilion is quite flat and is 46' X 56' in plan, reinforced
with 2" twisted bars 60 ft. .long and with a flat reinforced con-
crete dome panelled with I" thick glass. In the same location
the Girls Dormitory of the Stanford University was erected soon

FIGURE 5.

afterwards, and its three stories were completed in ninety days
from the time the plans were ordered.

When, on April 18, 1906, San Francisco was destroyed by the
earthquake, the buildings at Palo Alto suffered severe damage,
in many cases beyond repairs. However, the old reinforced
concrete buildings referred to above stood the test with little
if any damage; see Bulletin No. 324 of the United States Geo-
logical Survey, pages 22, 23, 24, 75, 112-114.

It may be worth while to note that the addition to the Borax
Works at Alameda, Cal. (1889), was the first instance of the
ribbed floor construction erected; it will be seen from Figure 6
that the construction is identically the same as used for that
kind of floors today. The Columns were also of concrete, prob-
ably the first ever erected.

I desire to express here my sincere gratitude to the men who,
in those early times, had the confidence and foresight to realize

PERSONAL REMINISCENCE 7

the technical and commercial importance of the novel construc-
tion, often in the face of severe criticism and bitter attacks.
Chief amongst these are my associate for many years, Mr. Frank
M. Smith, Architect Percy and Governor Stanford, deceased.

FIGURE 6.

Additional information in regard to the preceding buildings
may be found in a paper " Concrete Construe tion," read by
George W. Percy before the San Francisco Chapter of the
American Institute of Architects, February 9, 1894.

In this paper reference is also made to my tests on delaying
the placing of 1:2 mortar; the results are really astonishing.
The tensile strength of briquettes made with Knight, Bevan,
and Sturgess cement was as follows:

Delay in hours 1 2 2^ 3 4

Tens. Strength, Ibs. 252 228 240 256 306 228

showing that a delay of 2J to 3 hours really gave the highest
results. Similar tests with White's cement gave the best results
with a delay of 1J hours, after which the strength fell rapidly.
In all these cases, the concrete was worked up again as soon as
it stiffened. Unfortunately, these tests have never been ex-
tended to modern cement. However, in my address to the
Society of American Architects, October 17, 1894, I called
attention to the truly remarkable results obtained by Mr.
Spencer Newberry, who found that a mixture 1 : 3 which, when
worked for one minute with a trowel, developed a tensile strength
of 87 Ibs. in seven days, developed a strength of 240 Ibs. in the
same period after being worked with a trowel for five minutes.
I also made experiments with continued mixing, keeping the

REINFORCED CONCRETE BUILDINGS

concrete in the mill for as many as 1000 revolutions. I found
that within this limit the strength of the concrete increased with
the number of revolutions, so that concrete given 1000 turns in
the mill was stronger than when it had had 700 revolutions only. 1
These observations led me to believe in the continued work-
ing of the concrete while it is setting; that is, when making a
slab, I put men on with rollers who were instructed to keep the
rollers going for several hours. The slab is laid on the forms in
the usual manner; as soon as the concrete is hard enough to
carry a man, the rolling begins, and is carried on with two or

FIGURE 7.

three sets of rollers of increasing weight until the rollers make
no impression. A more handy method is to use hollow iron drums
filled with increasing amounts of water. It is important that
the floors be not allowed to dry out too quickly, for which

FIGURE 8.

reason they may be sprinkled if necessary during rolling and kept
moderately wet for at least one week more.

The construction of " illuminating panels" in concrete floors,
or, as they are more commonly called, sidewalk lights, has cap-
tured the attention of inventors for many years. But owing to
the unequal expansion of glass and iron, the great majority of
such constructions embodying a combination of these two
elements have not been satisfactory. My patents, No. 448,993
(1891), Figure 7, and 518,045 (1894), Figure 8, aimed to avoid

1 It must here be noted that the mill used for this experiment was of a
different type from those used today, the modern machines having a much
more severe action. The danger in overmixing is that the aggregate is ground
very fine, thus giving a mortar with an excess of sand.

PERSONAL REMINISCENCE

the use of the iron plate by setting the glasses in a body of
reinforced concrete, and this was accomplished with great suc-
cess. The Ransome Sidewalk Lights may be seen in every
large city of the country, amongst other places the New York
Subway, and my patents formed during their terms the basis
for a large and prosperous industry.

One of the problems most troublesome to the reinforced con-
crete engineer is encountered in joining new concrete to old.
A more or less suitable joint may be had in a number of ways,
and from time to time I have given this problem much thought.
A purely mechanical bond is created by bedding an open coil
half way in the old concrete surface, so that the other half is
caught in the new concrete, subsequently molded against the old
surface, Patent No. 647,904 (Figure 9). This principle is

.A\\A\A\A\A\A\

wwwwwwfwwwwww

FIGURE 9.

utilized in the " unit " construction of reinforced concrete
buildings according to my patent No. 694,577 (1902), Figure 10,
and the tie thus made is so efficient that the subsequently

^^VTTTVx

FIGURE 10.

molded slab may serve as the compression flange for the beam
or girder erected ahead of the slab. The first application of this
principle was in connection with the office building erected for
the Foster-Armstrong Company at East Rochester, N. Y.
(1904-5), and it has since been extensively used. Honey-
comb slag may also be utilized in a similar manner, Patent
No. 694,578 (1902), especially for large surfaces, but here I

10 REINFORCED CONCRETE BUILDINGS

prefer my later invention, the removal of the surface skin with
hydrochloric acid. The surface is next washed with water,
and the finish coat is then placed in the usual manner, Patent
No. 800,942 (1905).

This latter method, " the acid joint," has been tried in prac-
tice with excellent results. Fresh bases were welded to old
concrete cylinders, the mass allowed to set, and in all cases the
concrete would split apart from the joint when tested. One of
my superintendents found himself unable to believe in the
superiority of the new joint, and I had therefore a slab, about
4' X 4', set aside, letting him finish one half by any method
desired, and reserving the other half for my acid joint. Try
as he would, he never succeeded in making an unbreakable *
joint. A reward of ten dollars was promised to any man on the
job who could separate the finish from the base on the half
treated with acid, and while many of the men availed themselves
of the opportunity, the reward is as yet unearned.

The Pacific Coast Borax Co/s building, at Bayonne, N. J.,
erected in 1897-98, in a measure marks the closing of the old-
time construction of reinforced concrete buildings, constructed
more or less in imitation of brick or stone buildings, with com-
paratively small windows set in walls (Figure 11). This build-
ing, however, occasioned the discovery of an important fact,
that of the greatly improved fire-resistance of concrete mixed
with salt. Before that time, salt had been known and used as a
frost preventative, and as this building was constructed in the
winter, I desired to use salt. I had some doubts as to the
strength of concrete so made, and I also anticipated some
trouble with efflorescence. A number of test cubes were made
with salt, some mixed by hand, others by mill, and to my sur-
prise I found that the hand-mixed specimens showed efflorescence
while the machine-mixed specimens did not. I am .unable to
explain this difference. As to the strength, I found it was not
impaired by the salt, when salt to the extent of one to five per
cent, of the weight of the cement was added; I also found that
the specimens without salt showed air or hair cracks, while those
with salt did not. It now occurred to me that salt might have
the same or similar effect on concrete that it has on clay; it
is well known how clay pipes, etc., are glazed by being burned
with salt. Test cubes with and without salt were heated in

Cfi 03 "S

6 J1

g o g
fe w S

l? fl

I I

1-H 'S

-H

g I

12 l tifi IN FORCED CONCRETE BUILDINGS

the boiler furnace to a red heat, and then plunged into cold
water; the specimens without salt had a soft surface easily
picked to pieces with the bare fingers, while those with salt were
intact, and the compressive strength appeared to be unchanged.
I had then the belief that different brands of cement were affected
in different ways by the addition of salt, but I have so far never
found a Portland Cement that was injured in the least by addi-
tion of the quantities indicated.

Of other additions to Portland Cement with which I have
experimented I must mention lime and clay. The former
addition is so liable to abuse that I have largely abandoned it,
except for the construction of waterproof tanks, and even then
it is not indispensable. The fact seems to be that an addition
of from 'three to five per cent, of slacked lime is beneficial when
added as " milk of lime," using the limey water for mixing
instead of plain water; the trouble arises as soon as lumps of
lime putty, however small, find their way into the concrete, or
when the amount exceeds five per cent. In cases where the
concrete for this or other reasons contains free lime, I have
sometimes hastened the setting by giving an artificial supply
of carbonic acid; usually supplying heat to the concrete at the
same time. For the details of this see my patent No. 652,732
(1900).

As to an addition of clay I have found that from two
to three per cent, in the aggregate are beneficial rather than
detrimental, and such moderate amounts help greatly to render
the concrete waterproof. My attention was first called to this
matter when conditions compelled me to use Niagara Gravel
for some works in Buffalo in about 1892, and I declined at first
to use this material as it evidently contained considerable quan-
tities of clay. Inspection of concrete work made by other
parties soon convinced me that the gravel in question was
excellent material, and earlier tests by Mr. Clarke of Boston
fully corroborated my own observations on this point. I do
not wish to go on record as stating that all clay is beneficial,
and if the grains of sand are coated with a film of clay, I am
convinced that it has a- most dangerous effect.

In the years between 1900 and 1902 I developed a radical
departure in the exterior construction of reinforced concrete
factory buildings, consisting mainly in the extension of the floor

PERSONAL REMINISCENCE

plate or slab over the exterior columns, forming a belt course
on the outside of the building. Between the exterior piers,
upward and downward extensions were added; the former to
be added after the next floor had been constructed, and the
latter forming an integral portion of the floor proper. This
innovation forms the subject matter of my patent No. 694,580
(1902), Figure 12, according to which a very large number of
buildings have been erected throughout the country.

The principal advantages of this construction as compared

FIGURE 12.

with the old solid concrete walls are, aside from the purely
technical ones, that large window areas are easily made possible,
that the curtain walls are utilized as carrying members, and the
shrinkage of the walls is taken care of by means of the expansion
joints existing at each end of each curtain wall, the same being
recessed into the sides of the piers. That this construction
affords great economy will be evident from the fact that all the
curtain walls may be cast with a few forms only, the several
forms being usually removed in twenty-four to forty-eight
hours after pouring.

The first building erected under this patent was the Kelly &
Jones Co.'s machine shop at Greenburg, Pa., 60' X 300', four
stories high (1903-4), Figure 13, built by the Ransome &

14* * REINFORCED CONCRETE BUILDINGS

PERSONAL REMINISCENCE

Smith Co., and followed by many more, chief amongst which I
mention the machine shop for the United Shoe Machinery Co.,
at Beverly, Mass.., aggregating about sixteen acres of floor space
(the recent additions comprising about four acres were built
under the Unit System), and the Foster-Armstrong Go's Piano
Factory at East Rochester, N. Y., including a dozen or more
large buildings.

From this time also dates my invention of the coil joint for
uniting reinforcing bars, consisting in an open metallic coil
surrounding the lapping ends of the bars to be joined; No.
694,576 (1902), Figure 14. It is particularly well adapted for
deformed and especially twisted bars, because the
initial sliding of the bars cannot take place with-
out driving out a wedge of the surrounding con-
crete, and this is effectively prevented by the
coil. Wherever beams have been built with bars
joined on this principle the results have been sat-
isfactory, and the much later tests by Professor
Morsch fully substantiate everything claimed for
the coil joint. (See Trautwine: Concrete, 1909,
p. 1174, where plain bars only have been used.)

During all this time I have given consider-
able study to the proper design of the falsework,
realizing in common with other concrete men that
the handling of the forms in many cases meant
the difference between loss and profit. The stand-
ardization of the forms and their repeated use
is one way of approaching this problem, and
with a standard layout there is no question but that invest-
ment in molds of a more permanent nature is a paying propo-
sition. More information in regard to this may be found in .a
later chapter on forms; suffice it here to say that I have
made " coreboxes " for one building and used them there
four times, shipped them to another building and used them
seven times, then again shipped them and used them four
times, and finally shipped them once more and used them,
but on the last job the repairs were so expensive that the profit
was doubtful. I am now convinced that the final solution of
the questions pertaining to economical construction must be
found along other lines, and I have had sufficient experience

FIGURE 14.

16 REINFORCED CONCRETE BUILDINGS

with " Unit Construction " to warrant the statement that great
economy and better workmanship, as well as quicker work, is
thus obtained. From a first attempt in 1905, and subsequent
experience, I have evolved a system adapted to buildings with
many stories, known as " The Ransome System of Unit Construc-
tion," which has been extensively used and is now being used
on work of considerable dimensions. (Patents No. 694,577,
1902, and 918,699, 1909.)

Not very long ago, a patent was granted for " wet mixed "
concrete to a Western gentleman, and this brings back to memory
the historical fact that wet mixture has been known from the
earliest days of the art. Thus, Coignet's early patents (1869)
speak of it and recommend it, but, nevertheless, dry concrete
rammed in was in general favor. Now, one of the products of
the old Ransome Stone Company was porous filter stones, made
under the old process, and I was very much interested in making
similar stones of Portland Cement concrete. Owing to the lack
of uniformity of the concrete stones, I never made a success of
this, but I did find that in order to make the concrete sufficiently
porous for the purpose, I had to use a dry mix. Reversely, in
making ornamental stones, I always had better results with wet
mixtures, especially for the facing.

It is believed that the argument in favor of the dry mix was
based upon the fact, known as early as 1890, that dry mixed
mortar rammed hard into the briquette molds gave higher
strength in the tensile tests than wet mixture. The arching
effect of the stone in the concrete was disregarded. Personally
I was confirmed in my observations by Bamber's tests which
ingeniously proved the fallacy of the arguments in favor of the
dry mix, and showed the greater density of a wet mix: A dry
batch was made, and rammed thoroughly into a mold 2' X 2' X 2'
so as to fill it level full. The mixing platform was now cleaned,
and the contents of the box dumped out on the mixing board,
thoroughly remixed with enough water to make a " wet " mix,
and then replaced in the box. But it now proved that the box
lacked 2" in being full, so that the greater compactness or density
of the wet mix was proved. Other tests have proved the superior
strength of wet concrete, and that the densest mixture is also as a
rule the strongest; as to the permanency I have had occasion
to compare dry and wet mixed concrete after they had been in

PERSONAL REMINISCENCE 17

place for many years, and found the wet mixture much harder
than the dry placed concrete.

From my long and varied experience with concrete I desire
to state that I have found no agency which actually injured old
well-made concrete properly proportioned, except acid. Such
items as sewage and oils have had no influence, neither have I
found that the gases from the salamanders injure the setting
concrete.

In closing this contribution to the history of Reinforced Con-
crete, I cannot help but marvel at the enormous growth of the
concrete industry during the last fifty years, and especially of
the reinforced concrete industry in its less than thirty years of
actual use. I venture to predict that the next thirty years will
see even greater advancements, but I would also ask the younger
men in the profession to remember that real knowledge and ever-
lasting care are necessary, so that the reinforced concrete indus-
try in the future may proceed without setbacks from accidents
caused by neglect or greed.

CHAPTER II

BASIC PATENTS FOR INVENTIONS RELATING TO REINFORCED
CONCRETE, AND A SHORT SURVEY OF THE EARLY HIS-
TORY OF THE ART

BY ALEXIS SAURBREY

REINFORCED concrete as used today may be said to have
arisen from the following basic inventions: (1) A combination
of a plastic material adapted to harden, with a metallic strength-
ening device, the word " plastic " being here used in its widest
sense so as to include masonry laid with a plastic mortar. (2)
That, in a combination of this kind, the concrete, or plastic
material, must carry whatever compressive stresses act in the
structure, and the metal the tensile stresses. (3) That there-
fore the concrete and the steel have a tendency to separate, so
that "bond" or "anchorage" must be provided, either locally
in certain parts of the structure, or continuously along the length
of the metal reinforcement. (4) That, in addition to a main,
or directly tensile reinforcement, a secondary, transverse rein-
forcement is desirable and beneficial, whether this transverse
reinforcement is made from separate bars, or some of the main
bars are arranged in a peculiar manner to gain the desired effect.
(5) That a compression member may be strengthened by longi-
tudinal as well as by transverse reinforcement, or both. (6)
That a multitude of various uses may be found for the compound
material, each requiring a special combination.

To trace back into remote antiquity the use of metal in
combination with brick work is beyond the scope of this book;
suffice it to say that the Romans are sometimes credited with
the first use of such constructions: it is said that a tomb has
been found in which the roof consisted of a concrete slab with
bronze rods embedded, crossing each other lattice-wise. This
construction dates a hundred years or more B.C. 1 It appears

1 "Reinforced Concrete," compiled by James Tozer & Son, Limited,
Birkenhead, England.

BASIC PATENTS FOR INVENTIONS 19

that on the authority of ancient writers, the prevailing method
of judging the quality of lime for setting purposes was by observ-
ing the hardness and color of the original stone, the harder and
whiter varieties being preferred, and that this method was in
general use for a score of centuries or more, until the more
modern method of learning by experiment and investigation of
the facts was first applied to .the subject by Smeaton in England,
in or soon after the year 1756. Smeaton is credited with the
discovery, as a result of actual chemical analysis, that the real
cause of the setting of limes and cements consisted in a combina-
tion of clay with lime.

The use of the English natural cement, commonly called
" Roman Cement," was discovered by Parker in 1796, who in
that year took out a British patent, No. 2120, for a cement
or tarrass to be used in aquatic or other buildings and stucco
work. The use of the word "Portland Cement" first occurs in
the specification of a patent granted in 1824 to Joseph Aspdin,
of Leeds, England, No. 5022, owing its name to its resemblance
to Portland Stone, and this discovery formed the basis of con-
siderable manufacturing operations after the establishment of a
factory at Wakefield in 1825.

The Period of Discovery. Although in 1847 three or four
cement mills manufacturing artificial cement were operating
in England, the use of the new product was limited, until definite
methods of determining the commercial value of the product were
developed. Amongst the engineers who made reliable and
scientific observations in this field, John Grant, subsequently
knighted for his eminent ability as engineer, must be mentioned
in the first line, as well as General C. W. Pasley, whose book on
" Limes and Calcareous Cements" was a standard work in its
day (first edition in 1838, second in 1847). From this work we
learn how the relative merits of the various cements were some-
times tested by building a row of bricks out from the face of a
wall (Figure 15), as many as twenty-nine or thirty bricks having
been stuck out in this manner in one day, and thirty-three bricks
in thirty-three days, before the bricks fell. The famous semi-
arches constructed by Sir M. J. Brunei, the builder of the Thames
Tunnel, were undoubtedly conceived in this same spirit, and
these arches are all the more interesting as they give the first
known rational application of the principle of strengthening

REINFORCED CONCRETE BUILDINGS

masonry by means of tension rods. The arches are shown in
Figure 16; the work was four years in building, having been
added to from time to time, and it fell on January 31, 1838, as a
result of a cave-in in an adjoining excavation. The long arch

flow of Bricks standing out flat

T?ow of Bricks standing out en edge

FIGURE 15.

was 60 ft., and the other about 37 ft. long, the latter being
loaded at its extremity with a weight of 62,700 Ibs., and this
remarkable result was obtained by the introduction of wooden
lath and hoop-iron bonding strips inserted in the joints of the
brick work. " This ingenious arrangement of Mr. Brunei will

FIGURE 16.

probably be found hereafter of great value in practical archi-
tecture," Pasley says, and time has shown that he was correct
in his prediction.

Brunei also built a brick beam 25' 1" long, reinforced with
strips of hoop-iron; this beam was broken in 1836 under a load
of 27,025 Ibs. A similar experiment was made at the Francis
Cement Factory at Vauxhall, the dimensions of the brick beam
being 4' 9" deep by 22J" wide, reinforced with fifteen pieces of
1J" hoop-iron in the bottom portion. The span was 21' 4" in
the clear, and the beam was broken under a load of 50,652 Ibs.

Pasley had now several such beams made, one laid in neat
cement without irons, one exactly similar, but reinforced with
five longitudinal irons, and a third beam similar to the second
one, but laid in 1 : 3 lime mortar. The result showed the su-

BASIC PATENTS FOR INVENTIONS

periority of the reinforced beam laid in cement, for the first beam
carried only 498 Ibs., while the second carried 4523 Ibs., and the
third failed by sliding of the irons under a load of only 742 Ibs.

As early as 1832, Ranger took out a British patent, No. 6341,
for making certain kinds of mortar with hot water, and this
invention was used in the first known case of modern engineering
work of any consequence executed entirely in concrete, viz., a
dock at Woolwich dockyard (1835) and sea-walls at Woolwich
and Chatham. The floor of the dock was a failure, but the sea-
wall at Woolwich was standing in perfect condition in 1879.
" Ranger's artificial stone" became well known in England, but
it was soon discovered that cold water was quite sufficient for
making good concrete.

A short list of some early English patents may be of interest:
Wilkinson, 1854, No. 2293 (Figure 17).

FIGURE 17.

"The wire rope H, H, is secured at its extremities at each
line of support by imbedding it in the mixture or concrete while
in a soft state, and forming the ends into loops, or by opening
out the strands and hirling them in various directions, which
renders it so secure as not to be drawn out under any force short
of the breaking weight of the rope. For ordinary dwelling
houses I propose placing such wire ropes about nine inches apart,
and to have a full depth of floor of one-sixteenth the span."
Dennet, 1857, No. 685 (Figure 18).

FIGURE 18.

Proposes to strengthen his arches with lamina of wood or
iron.
Bunnet, 1858, No. 1292 (Figure 19).

Uses iron tie rods and metal abutment plates for his arches
of hollow blocks.

REINFORCED CONCRETE BUILDINGS

Parkes, 1863, No. 317.

Proposes an iron bond consisting of a band or strip of iron
with transverse teeth, ridges, ribs or projections pressed out of
the solid, raised at intervals on each side of the strip for the whole
width thereof. He employs two rollers, with suitable indenta-
tions, for the manufacture.

FIGURE 19.

Ransome (Fk.) 1865, No. 1337.

Molds slabs of artificial stone around pieces of hoop-iron on
edge running from end to end, so that the hard concrete prevents
the irons from buckling under load.
Scott, 1867, No. 452 (Figure 20).

v////////////s///////////////j/}?/7y/////////////tf ///>//'/.>*/////>///////.

IT TT

FIGURE 20.

Proposes to dispense with the use of ordinary joists and to
make use of wrought-iron tie-rods extending from wall to wall.
"The floor becomes one solid beam, having the tie-rods and hoop-
iron in combination with the concrete to take the tensile strain,
and the concrete to take the compressive action resulting from
the weight of the floor."
Lythgoe & Thornton, 1868, No. 640 (Figure 21).

FIGURE 21.

" The method of constructing floors with bars of J_-iron and
concrete as shown."
Johnson (Coignet) 1869, No. 884 (Figure 22).

An invention relating to the facing of concrete blocks with
cast-iron or steel protecting plates, to be used as street curbs,
etc.
Gedge (Monier) 1870, No. 1999 (Figure 23).

BASIC PATENTS FOR INVENTIONS

" In short the iron is the skeleton and the cement its
covering."
Tall, 1871, No. 1001 (Figure 24).

Iron hooping, wirework, or netting are interlocked between

FIGURE 22.

FIGURE 23.

the lateral cross bars, and form a close lattice or basketwork.
Portland Cement stucco is applied.
Brannon, 1871, No. 2703 (Figure 25).

FIGURE 24.

" Wirework embedded in concrete, to give cohesive strength
against transverse and tensile strains."
Hyatt, 1871, No. 3124 (Figure 26).

REINFORCED CONCRETE BUILDINGS

" The peculiar construction of floor which I designate an
'all-beam' floor, composed of a number of separate tubes laid
side by side."
Turner, 1872, No. 1396.

On the iron beams "I strain my wire from the plates in the
walls; these wires are intended to supersede the use of floor
joists of wood, and will form beds for my concrete floors, and also
answer on the underside instead of laths for the plastered ceil-
ings, which work of plastering may be carried on at the same
time as the laying on of the floors in concrete."
Emmens, 1872, No. 2451 (Figure 27).

FIGURE 25.

FIGURE 27.

" The employment of sheets of corrugated iron as founda-
tion for roadways, paths, steps, and flooring."
Lish, 1873, No. 1621 (Figures 28, 29).

The drawing shows a sectional view of a floor and girder
of concrete with tension .rods embedded therein, as indicated by
the dotted lines.
Hyatt, 1873, No. 3684.

Asbestos combined with perforated, corrugated sheet metal
or with crimped sheet metal or upon a hollow grate bar system.
Coddington, 1873, No. 1004 (Figure 30).

BASIC PATENTS FOR INVENTIONS

The figure shows a water pipe or tube, C being the cemented
material, E the interwoven metal.

FIGURE 28.

FIGURE 29.

Hyatt, 1873, No. 3381.

" The system or mode of forming cellular or honeycomb
structures by connecting together single cell blocks by means

FIGURE 30.

of tie-rods or crimped blades of metal, with or without addi-
tional straight tie-rods."

FIGURE 31.

Hyatt, 1874, No. 2550 (Figure 31).

"I form the tie in a way which gives it power to grip and
hold the foreign material in a manner and by a method which

REINFORCED CONCRETE BUILDINGS

brings the load and consequent strain upon the tie at the same
instant it is felt by the concrete or foreign material, by which
means the tensile and compressive forces act in harmony with
each other."
Hyatt, 1874, No. 1715 (Figure 32).

FIGURE 32.

" Making hollow metal beams of interlaced lattice or open-
work, as the holder of a tie-rod, to connect the same with con-
crete or equivalent material."
Edwards, 1891, No. 2941 (Figure 33).

1892, No. 1415 (Figure 34).

1894, No. 15,466 (Figure 35).

FIGURE 33.

i.--.

FIGURE 34.

Edwards' patents show a remarkable insight into the nature
of reinforced concrete construction. It is proposed to cast

BASIC PATENTS FOR INVENTIONS

the slabs separately and set them when hard, owing to the great
cost of the centering; the bending up of the principal tension
rods is described at great length, and stress is laid upon the
benefit of many small rather than fewer but larger rods. The

FIGURE 35.

importance of preventing sliding of the reinforcement is shown,
and it is described how the beams may be pierced by openings
in much the same manner as done under the Visintini System.
The benefits as well as troubles arising from the fixing of the
ends of the beams into the walls are perfectly understood, and the
entire argument advanced is illustrated by tests (by Kirkaldy).

While in England the new construction made but scant
headway, a considerable activity took place in Germany, where
the Monier Patents were bought and exploited by G. A. Wayss,
and where M. Koenen advanced the first rational method of
calculation in 1886. The " straight line formula" was fully
discussed by Koenen in " Centralblatt der Bauvervaltung,"
May 14, 1902, and is to this day the commonly accepted
standard.

In Holland, the first ribbed floors were erected in 1886 in
connection with the Public Library in Amsterdam.

In France, it seems that Monier's first patent was taken out
in 1867, but it has been intimated that he had knowledge of
the earlier patent granted to Lambot, who had made a reinforced
concrete boat of small dimensions in 1855. This boat is said
to be in existence today. Monier 's efforts toward the intro-
duction of his inventions were not very successful, partly per-
haps because he failed to realize the necessity of placing the
reinforcement near the bottom; it is told that when Wayss
showed him slabs so reinforced, Monier severely criticized this
arrangement, and abruptly ended the argument by exclaiming,
"Who is the inventor, you or I?" 1 As a matter of fact, little

1 Suenson: Jaernbeton, P. 5.

REINFORCED CONCRETE BUILDINGS

was done in building construction until 1892, when Henne-
bique and Coignet took the reinforced concrete construction
up with great success, each introducing his own system.

In the United States, the first indication of anything ap-
proaching reinforced concrete may be found in a patent granted
to P. Summer, 1844, No. 3566 (Figure 36), for a metal lathing,
which was still further improved by J. B. Cornell in 1859, No.
22,939 (Figure 37). At this early date, a number of patents

FIGURE 36.

for cement pipes were granted, as to R. B. Stevenson, 1854,
No. 11,814, for a combination of a pipe of sheet-metal and an
exterior coating of hydraulic-cement mortar of " requisite thick-
ness for strength." In the Wyckoff patent, No. 32,100, of 1861,
the interior pipe is of wood wound with wire of iron or other
metal; in the Knight Patent, No. 32,298, of the same year, a metal
tube is disposed " intermediate between the inner and outer
surfaces" of a cement pipe. In 1868, A. P. Stephens took out a
patent, No. 78,336, on a similar pipe, in which the strengthening
tube was made of corrugated iron; in 1872, Patent No. 127,438,
the tube was changed to a spirally formed sheet metal tube,
and in the same year J. A. Middleton, Patent No. 133,875, pro-
posed to strengthen his cement pipes by a layer of wirecloth
embedded in the cement, thus combining what we now consider
the essential elements of a reinforced concrete pipe (Figure 38).
The first reinforced concrete wall-patent appears to be one

BASIC PATENTS FOR INVENTIONS

granted to S. T. Fowler, in 1860, No. 28,069, where the concrete
wall is to be strengthened with vertical and horizontal timbers,
to be buried in the concrete; a more rational construction is

FIGURE 38.

proposed in 1862, No. 37,134, by G. H. Johnson, for grain-bins:
" a new construction formed of brick- work tied together by
plates and rods of iron." In 1869, No. 87,569, G. H. Johnson

FIGURE

improved this construction, using " horizontal annular tension-
bars . . . the ends of each bar being so united as that it shall
form an endless, unbroken band ... in the combination . . .
with . . . vertical connecting-rods so as to form a metallic

REINFORCED CONCRETE BUILDINGS

frame within the walls of the structure." This invention (Fig-
ure 39) was not the only important improvement of that day;
in 1868, C. Williams, No. 75,098, invented the metal lattice-
reinforcement for concrete walls. The lattice-work was built
up by riveting the slats together (Figure 40).

The first use of concrete in columns must be conceded to
W. H. Wood who, in 1862, No. 36,747, patented an improvement
in piers and bridges. The invention consists in the use of hol-
low cast-iron columns filled with concrete or cement, and sup-
ported on wooden spiles below the surface of the bed of the
river. The first ceiling was proposed by J. Gilbert, 1867, No.
64,659. This patent shows corrugated iron plates filled with

FIGURE 40.

concrete, the concrete to extend an inch or so above the top of
the corrugation (Figure 41). His solution of the problem
" self-centering reinforcement" is not very inferior to those
proposed by more recent inventors. Thus we see that around
the year 1870 the combination of masonry of various kinds
with a strengthening metal work was quite well known. The
patent, No. 88,547, granted to F. Coignet, a Frenchman, in
1869, states the general principles very clearly: " In the body of
artificial stones": "skeletons or metallic framework, linked or
arranged so as to strengthen the same." This is the whole
science of reinforced concrete construction in few words. As
an example, he proposes to use a cylindrical web of small rod-
iron or wire in combination with a cement envelope, for the pur-
pose of resisting the interior pressure in pipes, as well as T- or

BASIC PATENTS FOR INVENTIONS

L-irons for other purposes. The series of patents granted to
Coignet in 1869 deserve more than usual attention, as they
contain much good advice of value to engineers; they are
No. 88,545, 88,546, 88,547, 88,548, and 88,549.

FIGURE 41.

The brick arch with abutment-shoe and tension bar between
abutments was invented by C. Henderson, in 1871, No. 113,881
(Figure 42); the brick arch reinforced on the cantilever prin-
ciple was invented by F. Alsip, No. 120,608, in the same year.
It is not clear from the description whether Alsip really con-

FIGURE 42.

sidered his invention as a cantilever construction, but the fact
remains that all the essential elements of a cantilever are pres-
ent in this patent. A very interesting patent No. 122,498 is the
one granted to W. H. Smith, in 1872, for a concrete pavement.
On soft ground, the arched pavement is intended to be self-
supporting. Tie-rods are then carried under the pavement from

curb to curb, or "chords may be embedded in the composition
to operate in lieu of abutments to the arch." In the drawing,
the tension rod is shown provided with a large button on the
end, evidently for the purpose of preventing slipping of the
bar (Figure 43).

REINFORCED CONCRETE BUILDINGS

The patent issued to Sisson and Wetmore, in 1872, No.
124,453 (Figure 44), shows "a combination of trussed and un-
trussed frames of light bar-iron to form skeleton wall-posts,
girders, etc., in combination with a filling of beton or other
suitable concrete, to be poured in a state more or less liquid.
Our object is to have the beton and iron frames furnish mutual
support and protection to each other." Considered as a beam,

FIGURE 44.

the wall-post of this patent exhibits many of the essential fea-
tures of present day practice: the top bar extending from one
span to another, the trussed bar bent up over the support, the
horizontal lacing of the verticals and the vertical lacing of the
horizontals, etc.

But generally speaking, the reinforced brick-arch continues
to hold the interest of the inventors. In 1872, P. H. Jackson
received a patent, No. 126,396 (Figure 45), for a peculiar con-
struction of abutment-casting to be used in connection with
reinforced arches, and in 1873, No. 137,345, N. Cheney proposes

BASIC PATENTS FOR INVENTIONS

to make the tension reinforcement of light wires placed close
together and interwoven with cross-wires, to serve the addi-
tional purpose of a metallic lathing. The earthquake-proof
house invented by D. L. Emerson, in 1873, No. 137,833, calls

FIGURE 45.

for vertical rods or plates in the walls, and anchors passing
through them, the plates and anchors being connected with
strap iron. In the same, year J. W. Basset, No. 138,118 (Fig-
ure 46) shows a construction of individual plaster slabs with a
metallic trellis work within, the ends of which extend beyond
the block, for the purpose of locking the various blocks together.
While not strictly within the scope of this paper, attention
is called to the patent, No. 172,641, granted to O. C. Matthews,

FIGURE 46.

in 1876, for a foundation, in which piles are driven and again
withdrawn and the holes filled with concrete (Figure 47).

In 1878, T. Hyatt, No. 206,112, ended the " period of dis-
covery " and put the theory of reinforced concrete construction
on a rational basis, and at the same time received a patent of
remarkably broad scope, covering practically the entire field
of reinforced concrete and masonry construction. The general
purport of this invention is set forth in a volume entitled "An
account of some experiments with Portland Cement concrete,
combined with iron," of which a copy was deposited in the
library of the Patent Office, but which was otherwise designed
for private circulation. Hyatt appears to be the first to state
specifically that the steel must be able to resist sufficient tensile

REINFORCED CONCRETE BUILDINGS

stress to balance the compressive stresses on the concrete, that
all metal may be dispensed with save the tension rod only, that
both baked bricks and concrete possess in themselves cohesive
power and strength sufficient to perform the functions ordina-
rily performed by the metallic web. He realizes the value of
deformed bars and says: " I prefer to use metal specially rolled
for the purposes, with bosses or raised portions formed upon

FIGURE 47.

FIGURE 48.

the flat faces of the metal. When I make use of common bar
or hoop iron, I stud the slips with pins; or I make use of several
blades threaded upon wires, as represented by Figure 1." In
the book mentioned above, he laid down the results of his ex-
periments which led him to bend some of the bars up, and also to
use a rigidly attached separate " shear member." The analysis
is very complete, both in his book and in his patent speci-
fication. He reinforces his columns with longitudinals or hor-
izontal hoops, as the case may require, or both. He says: "In

BASIC PATENTS FOR INVENTIONS

constructing the columns or piers wholly of concrete, I make the
structure solid, the concrete then bearing the load, and, giving
way under compression, would naturally incline to yield in the
first place, not from absolute crush of the materials, but from
want of sufficient tensile resistance at the circumference of the
column. But this tendency being resisted by the circular
ties, such a concrete could give way only by the crush of its
particles." In short, the whole theory of hooped columns.
The only difference is that Considere prefers the use of spirally
wound reinforcement, while Hyatt uses the individual bands
(Figure 48).

To what extent Hyatt was familiar with Pasley's tests, if
at all, we do not know; in his book of 1877 he gives a brief ac-
count of the history of fireproof construction, but gives no ref-
erence whatever to the tests just mentioned. It appears that
he had a test made in September, 1855, in New York, under
the general supervision of Mr. R. G. Hatfield; the beam was
about 9" square, and had a tie-rod passing through holes made
for the purpose in the bottoms of the bricks. More important
tests were made by Kirkaldy in London, from 1874 to 1877, on
beams made by Hyatt.

The Period of Improvement. Broadly speaking, the Hyatt
Patent, No. 206,112, shows and describes everything necessary

FIGURE 49.

for the practical use of reinforced concrete, and the patents
of the following period are therefore mainly for improvements,
many of which are due to Hyatt. Most interesting is the one
granted in 1883, No. 290,886, for a concrete floor, showing not
only transverse arches between the ribs, but also the use of web
reinforcement in a continuous sheet along the center of the
beam. In 1881 a patent, No. 237,471, was granted to S. Bis-
sell (Figure 49) for an arch-bridge, showing diagonal straight
reinforcement within the masonry, the object being to construct

36 REINFORCED CONCRETE BUILDINGS

"an arch of limited span without causing any horizontal thrust
upon' the abutments." The Cubbins patent of 1883, No.
285,801, shows a circular cistern cover "of artificial stone,
having a metallic band or tire" (Figure 50) or "consisting of

FIGURE 50.

a concavo-convex or arched disk . . . inclosed by a metallic
band or tire." This appears to be the first slab with " cir-
cular reinforcement." "Expanded metal" was patented,
No. 297,382, in 1884, by J. F. Golding: "metallic screening
formed of slashed and stretched metal." The particular use
to which the invention was to be put is not specified, and at
first it was used exclusively as a metal lath. Its use as
reinforcement for structural concrete is of much later date
(Figure 51).

FIGURE 51.

A number of interesting patents are granted at various
dates to P. H. Jackson. The first, No. 302,338, in 1884, is not
of interest in this connection; it shows principally the usual
tie-rod construction in a brick arch. But the following year,
1885, he took out a patent, No. 314,677, showing, for the first
time, the bent-up or "trussed" arrangement of reinforcement
(Figure 52); the bars are carried to the support where they are
anchored by means of nuts. The concrete and its reinforcement
rest upon corrugated iron plates, and the bars may be secured

BASIC PATENTS FOR INVENTIONS

or not at intervals to the bottom of the corrugated plates.
Another patent, No. 320,066, of the same year, shows the rein-
forcement continued into the adjacent bay and there hooked

FIGURE 52.

over the tops of the I-beams (Figure 53), which here have the
function of the main girders. The patent, No. 339,296, of 1886,
specifies an expansion joint in the construction of a reinforced
concrete arch; evidently the troubles caused by expansion and
shrinkage were well known at this early date. Two patents,
Nos. 366,839 and 366,840, were taken out in 1887 for " series

FIGURE 53.

of arches composed of concrete, and a longitudinal tie on which
the footings of the said arches are supported and to which they
are fastened," and a construction of arches with longitudinal
reinforcement near the bottom; these arches rest on one side
on the front girder of the building, on the other side upon the
area wall. Also from that year date the following patents:
two, Nos. 367,343 and 370,625, showing the application of dove-
tailed corrugated plates filled with concrete (Figure 54), and

FIGURE 54.

three, Nos. 371,843, 371,844, and ?71,845, showing the use of
I-beam reinforcement in the bottom of the beam, as well as

REINFORCED CONCRETE BUILDINGS

compression-reinforcement in the top (Figure 55). The reis-
sued, RIO, 921, and the original patent, No. 375,999, issued in
1888, may be noted in passing.

When we consider the state of the art as it appears from
the patents mentioned above, the Monier patent of 1884, No.
302,664 (Figure 56), cannot be called much of an improvement.

FIGURE 55.

FIGURE 56.

Nevertheless, the name Monier was for many years synonymous
with " reinforced concrete," at least in Europe, where the Mon-
ier patents were bought and greatly developed by German
engineers. "My invention," he says, "relates to the use and
sale of integral elements of construction of metal and concrete
or mortar combined, the mortar forming the covering for a
metal skeleton. This skeleton is composed of longitudinal
bars or rods and transverse ribs, secured together by metal
ligatures." The Monier patent, No. 486,535, of 1892, is practi-

FlGURE 57.

cally nothing but a series of special designs based upon this same
principle, and contains little new material. Yet a great industry
was based both here and in Europe upon the Monier patents.

The Ransome patents have been described in an earlier
chapter and are not referred to here.

The "trussed" arrangement of the steel was, as stated
above, invented by Jackson in 1885. The Gustavino patent,
No. 336,048, of 1886 (Figure 57) shows the same feature, as

BASIC PATENTS FOR INVENTIONS

well as the rod with a continuous curve between supports. In
addition to this tie-rod which extends from wall to wall, " I
may in practice use a straight tie-rod extending between wall
and wall above the arch." The same year, 1886, saw the origin
of another new type of construction which stands on the border
between reinforced concrete and plaster work. The Rabitz
construction, No. 339,211 (Figure 58), calls for a metallic skele-

FIGURE 58.

ton frame of vertical rods and a reticulated metallic netting,
in combination with a suitable coating of cement mortar or sim-
ilar material. In the patent issued to P. M. Bruner, No. 356,703,
in 1887, something approaching the U-bar (Figure 59) is shown;

FIGURE 59.

although the construction would not be classed as reinforced
concrete at this present time, the rods being disposed princi-
pally on the compression side, from which rods transverse ties
hang down in the beam. A telegraph pole was invented by
D. Wilson, No. 374,103, in 1887; it was to be composed of a skel-
eton frame having rods and horizontal hoops, and a coating or
body of cement inclosing the frame. The same idea was pat-
ented, No. 411,360, in 1889, by O. A. Stempel, who claims a
post, rail-tie, or beam, composed of "a metal frame, the filling
and inclosure of imperishable material that protects said frame

40 REINFORCED CONCRETE BUILDINGS

from the inroads of moisture and rust, and said frame arranged
to protect said structure from breakage." The drawing looks
somewhat like what an engineer would prepare for a column at
this time (Figure 60).

The patents granted to M. F. McCarthy show
again "the combination (with an I-beam supporting
the slab) of the wire strands extending over and
drooped between the same, and the concrete filling
wherein said beams and strands are embedded." This
quotation is from the patent issued in 1891, No.
455,687 (Figure 61); the four patents, Nos. 520,489,
520,490, 520,491, and 520,492, issued in 1894, show
various combinations and variations of the same prin-
ciple. The patent issued to P. Cottancin, No. 459,944,
in 1891, is for a strengthening web " characterized
by the union in a reticulated fabric of a warp and a
weft, each composed of a wire, band, or bar bent on
itself into a sinuous or like shape." This patent
forms the base for a large industry especially in
France. The J. Melan patent, No. 505,054, of 1893
(Figure 62), claims "a vault or arch consisting of
abutments, beams, or girders, arched ribs rigidly con-
nected with said abutments, beams, or girders, and
a filling of concrete or the like between said ribs."
A number of arch-bridges have been constructed
under this patent. A. L. Johnson patented, No.
550,177 (1895), a construction of floors much used
at one time in the West, comprising mainly I-beams
with suspension straps fastened at the tops of the
beams and drooping between the beams; the straps
are flat and support the concrete rib of the beam
(Figure 63), upon which in turn rests the concrete
slab. Another important arch-patent, No. 583,464,
was granted to F. von Emperger, in 1897, for an im-
provement in the Melan patent described above; it

FIGURE 60. consists mainly in using two ribs instead of one, each
rib being placed near one surface of the concrete.
Secondary members connect the top and bottom ribs (Fig-
ure 64).

Recent Patents. The idea of molding reinforced concrete

BASIC PATENTS FOR INVENTIONS

members separately and afterwards erecting them in place
appears to be almost as old as the art itself, and a number of

FIGURE 61.

the patents mentioned above refer to this possibility without
going much into the details. In 1898, a patent, No. 606,696,

FIGURE 62.

was issued to G. B. Waite for a beam construction (Figure 65),
the sole object of which is to provide members adapted to be

FIGURE 63.

molded in advance and erected in place after hardening. The
individual sections are made of I-shape and reinforced in top

FIGURE 64.

and bottom, or in the bottom only; " shear" members of vari-
ous forms are used in the beam-webs. The De Man twisted

FIGURE 65.
bar was patented, No. 606,988, in the same year; it consists

REINFORCED CONCRETE BUILDINGS

of "a thin flat bar having twists formed therein at intervals"
(Figure 66).

The patents granted to F. Hennebique, in 1898, are three in
number. The first, No. 611,907, shows the now almost univer-

FIGURE 66.

sally used combination of open, U-formed shear members with
horizontal and trussed main reinforcement, with the main bars
extending into the adjacent span (Figure 67). While the

FIGURE 67.

authorities seem to disagree in regard to the value of the pro-
tection afforded by this patent, there is not the slightest reason
to doubt that this construction has been of the greatest benefit
to the art. The second Hennebique patent, No. 611,908, is
for a system of separately molded members, claiming in sub-
stance a combination of joists and " a plurality of slabs having
projecting cores embedded in said joists " (Figure 68); the word

FIGURE 68.

core means here the reinforcing bar, and the slabs are placed
with their ends resting upon the side-forms for the joists, so

BASIC PATENTS FOR INVENTIONS

that, when concrete is poured in the joist-molds, the project-
ing ends are embedded in the concrete. The third patent,
No. 611,909, is for a pile of reinforced concrete having grooves in
two faces, so that a tight cofferdam may be made by using the
piles for sheet piling, and filling in the grooves with grout. The
structures erected under the Hennebique patents are numbered
by the hundreds in any one of the several civilized countries.

The patent, No. 617,615, issued to E. Thacher, in 1899 (Fig-
ure 69), for an arch construction, claims the combination of the

FIGURE 69.

concrete arch with its abutments, and reinforcing bars in pairs,
one bar near the intrados, and one near the extrados, the two
bars of each pair to be above one another, either both or only
one of these bars to extend well into the abutment, and, in par-
ticular, "each bar of a pair to be independent of the other."
A comparison with the Melan and v. Emperger patents is of
interest, as the bars in the v. Emperger patent extend into the
abutments and are placed one above the other. In the same
year, 1899, a patent (Figure 70), No. 634,986, was granted to

FIGURE 70.

A. Matrai for a system of wire reinforcement embodying many
interesting features. One object of the construction is to unload
as far as possible the middle of the supporting beam or girder,
and these again are reinforced with a number of suspension
cables or wires. This construction is in considerable favor in
Europe. In 1900, a patent, No. 654,683, was issued to I. A.
Shaler, for a construction embodying the use of longitudinal and
transverse rods, the latter welded to the main bars at intervals,

REINFORCED CONCRETE BUILDINGS

and in the same year, L. G. Hallberg had a patent, No. 659,967,
issued for a foundation built on the principle of " circular rein-
forcement " (Figure 71) in combination with radial bars. The

Wayss patent, No. 673,310 - 72) of 1901, is of interest, on

account of the rigidly attach ear members and other fea-

tures, the purpose being to ^otain similar advantages as out-
lined for the Hennebique patent without infringing the same;

FIGURE 72.

the construction is dissimilar to Hennebique in the particular
arrangement of the parts. The well-known Thacher bar was
patented in 1902, No. 691,416 (Figure 73), and in the same year

BASIC PATENTS FOR INVENTIONS

a patent, No. 709,794 (Figure 74), was granted to W. C. Farm-
ley for a concrete arch construction, in which the steel is so
arranged as to make the same bar pass from the tension region

FIGURE 73.

near the intrados to the tension region near the extrados, etc.
The Visintini patent, No. 735,920, of 1903, shows the peculiar
type of construction known under that name; instead of the

FIGURE 74.

ordinary solid beam, a lattice-girder of reinforced concrete is
used. The top and bottom flanges are reinforced with longi-
tudinal bars, and the cross-bars are embedded in the concrete

FIGURE 75.

work of the lattices (Figure 75). The Visintini beam has
been used but little in this country, but abroad a large number
of structures have been erected under this patent. In 1903,

FIGURE 76.

the first Kahn patent, No. 736,602, was issued, to be followed
by many more (Figure 76). The principal features are well

REINFORCED CONCRETE BUILDINGS

known: The rigidly attached secondary members are manu-
factured in one piece with the main tension rod, then sheared
loose from the main body along the greater part of the length

FIGURE 77.

of the rod and bent up as desired. The Weber chimney-con-
struction was patented in 1903, No. 748,242; the lower portion
of the chimney is provided with a circumferential air-space

open at its base to the outer air and leading at its upper end into

the chimney flue at the base of the upper single flue (Figure 77).

A. Considere took out a patent, No. 752,523, for his well-

BASIC PATENTS FOR INVENTIONS 47

known column construction, claiming "a solid concrete core
with independent helicoidal coils of metal surrounding said
core, and arranged very close together," and also the combina-
tion of these elements with separate longitudinal rods, in 1904
(Figure 78). With this patent we may consider the period of
invention as coming to an end. A very large number of patents
have been granted since, mostly for slight improvements, and
an enumeration of all these details would be very tedious and
without serious importance, although several patents of the
greatest interest may be found in this great mass of dead
material.

PART II

RATIONAL DESIGN OF REINFORCED
CONCRETE BUILDINGS

BY ALEXIS SAURBREY

CHAPTER III
INTRODUCTION

1. EXPERIENCE teaches that concrete beams may be greatly
strengthened by introducing a comparatively small amount of
steel within the concrete, according to certain principles of which
the following is a discussion. This combination of concrete and
steel is called Reinforced Concrete; the essential peculiarity of
reinforced concrete structures is that both the concrete and the
steel, if alone, would be grossly inadequate for the load which
they will carry when combined; the load carrying capacity is
not the sum of the individual capacities of the concrete and the
steel. This general rule is not without exception, if structures
like the ordinary reinforced concrete column are included;
strictly speaking, only the hooped column is entitled to be clas-
sified as reinforced concrete, because in that case a small amount
of steel added to the concrete changes the structural properties
of the column entirely.

2. The stresses in a reinforced concrete structure are neces-
sarily complicated. Not only is the steel entirely dissimilar
in nature to the concrete which it reinforces, but the concrete
itself is not homogeneous in the strictest sense of the word.
Yet two cubes of large size, cut from different parts of the beam,
must be assumed to be theoretically alike; we make there-
fore the necessary and justified assumption that the lack of homo-
geneity of the concrete is of second order as compared with that
of the structure as a totality: necessary, because otherwise we
cannot advance any theory; justified, because the differences
between the nature of steel and concrete are sufficiently large
to overshadow completely the small differences which un-
doubtedly exist within the concrete itself.

3. Generally, the properties of reinforced concrete are known
when the properties of the two materials are known; there is

52 REINFORCED CONCRETE BUILDINGS

no reason for believing that the properties of either material
are changed in any way by the presence of the other. It is,
however, necessary to expand the limits of our research when
dealing with a combination of two materials, because the prop-
erties of the combination depend primarily upon the ability of
the two materials to co-operate, and only in second line upon
their individual properties such as strength, elasticity, etc.
This co-operative ability is of a somewhat obscure nature;
without making any attempt of explaining it, we must admit
its existence. In the following it is referred to as the "bond"
or the "adhesion." When this bond is broken the structure
fails.

4. The purpose of design is to produce not only a structure
of adequate strength, but one of equal strength in its several
parts. With consistent formulas for the various elements, the
allowable stresses should therefore be the same for all elements
of the structure considered. Experience shows, however, that
the difficulties to be overcome in the erection are different for
different parts; we can readily see that a local deposit of bad
concrete as large as a hand will affect a 10" column and a 6"
floor slab in dissimilar ways. This is the reason for variable
allowable stresses in any case the purpose of fixing certain
maximum stresses is to insure an ample factor of safety. For-
tunately the investigation of stresses in a given beam is very
much simpler under moderate loads than near ultimate failure;
the coefficient of elasticity for steel E a is a constant, and that
for concrete E c varies but slightly. For practical purposes
the ratio E S /E C r is assumed to be a constant up to the limit
of the allowable stresses. Within this same limit we assume
sections plane before the load was put on to remain plane under
load, and we assume proportionality between stress and deform-
ation. The tensile strength of the concrete is entirely disre-
garded. None of these assumptions can be called absolutely
correct; they are, however, no more inaccurate than any other
set of assumptions which we would be able to suggest in our
present state of knowledge; moreover, they are the simplest
possible.

5. As the tensile strength of concrete is much less than its
compressive strength, the principle is to utilize the available
compressive resistance and use steel bars to carry the tension.

INTRODUCTION 53

Sometimes steel is also used in compression, although with less
success, the object being to limit the size of the columns and to
fortify them against excentric loads. We shall see later that
it is possible to construct a column in which the steel is stressed
in tension (Article 12).

6. In any kind of concrete structure the embedded steel
has a tendency to displacement in its own longitudinal direc-
tion under load. The value of the steel as reinforcement de-
pends upon its ability to withstand any forces tending to either
push or pull it out; reinforced concrete is an impossibility
without adhesion between steel and concrete, and destruction
of the bond or adhesion means failure. The law governing
adhesion is therefore the foundation of all theoretical study of
reinforced concrete.

CHAPTER IV

ADHESION

7. THE adhesion is measured in Ibs. per square inch of em-
bedded surface of the rod; its value is different for pulling and
pushing tests. As the latter is somewhat higher it is sufficient to
investigate the laws governing the pulling resistance and apply
these laws to the pushing resistance also. The mathematical
analysis of the bond stresses is impossible with the material on
hand; even the test-data are meager and often contradictory.
We know, however, that the following statements are approxi-
mately true, so that an embedded rod pulls out of the concrete
block :

(a) when the stress in the steel reaches the elastic limit of
the steel.

(6) when the tensile resistance of the concrete, in a lateral
direction, is reached, because the block splits.

expands sufficiently to let the irregularities of the rod
pass through.

(d) when the adhesion is destroyed.

Obviously, then, the designer must keep the steel stress well below
the elastic limit, allowing for this and other reasons an ample
factor of safety, while, at the same time, the concrete must be
strong enough to meet the demands made upon it. Hence the
diameter of the concrete block, or the thickness of the piece, is a
very important factor, but unfortunately nothing is known in
regard to the minimum allowable diameter, except that it is
greater for deformed bars than for plain round or square rods.
We can readily see that both the tensile strength and the coeffi-
cient of elasticity of the concrete has great influence upon the
minimum allowable diameter; with a well-proportioned mortar
and a mixture of say 1 :2:3J, we may perhaps suggest a diameter
of concrete equal to ten diameters of the embedded steel as

' 54

ADHESION 55

reasonably safe. In floor construction the bars usually find
their ultimate anchorage in much larger bodies, the slab bars
passing through the beams, the beam bars through the girders,
and finally the girder bars through the columns. In all these
cases the concrete is reinforced in a direction transverse to the
direction of the pull, and the expansion in a lateral direction is
thus partially or entirely prevented.

8. In the beam theory to be outlined below, great importance
is attached to the length of embedment beyond the supports of the
beam, in fact, this length- represents the ultimate reserve of
strength of the beam. It is usually considered good practice
to imbed the bent bars from twenty-four to thirty-six diameters

6 Diameters

J, Diam..

V 1

FIGURE

-f

79.

beyond the support, using the lower figure for deformed bars
stressed to 16,000 Ibs./sq. inch, and the higher figure for deformed
bars stressed to 20,000 Ibs./ sq. inch. For plain bars, an addi-
tional hook is made on the end of the bar, equal in length to six
diameters. In many cases the length of embedment here recom-
mended cannot be obtained for the reason that there is no adjacent
concrete into which the bars may be extended, as, for instance,
in the case of a beam finding its bearing in an outside brick wall.
The bars are then hooked, and the length of embedment calcu-
lated from the center of the seat to the end of the bar, including
the curved end of the hook. Square hooks must be avoided, a
gentle curve of, say, six times the radius of the bar is much more
effective, and the more so the greater the radius of the bent
(Figure 79).

9. The diameter referred to in the preceding paragraph is
not the diameter of each individual rod or bar, unless the rods be
spaced so far apart that each will pull out individually, leaving
the concrete intact between. The diameter is that of a circle or
other curved line in which all the rods may be enclosed if laid
closely together. It follows that it is good practice to spread the
rods out as much as possible; in a beam this is easily obtained by

56 REINFORCED CONCRETE BUILDINGS

bending some of the bars up over the support, as is also done for
other important reasons. It is a common but inexcusable
mistake to use a number of small diameter rods bunched together;
it is almost impossible to concrete such beams properly, and
the fallacy of the argument leading to such construction should
be evident.

CHAPTER V
COMPRESSION AND LATERAL EXPANSION

10. WITH few exceptions, materials submitted to deformation
in one direction undergo deformations in all other directions.
If the principal deformation is a shortening, the lateral deforma-
tion is a swelling, which must be taken as evidence of certain
interior stresses in the body in a direction normal to that of the
principal stress. These transverse stresses are of the greatest
importance for materials like reinforced concrete, because, if not
restrained, they bring about the premature failure of the con-
crete, while, if restrained, they may be used to increase the
strength of the structure. Thus, as pointed out above, the
transverse swelling affects the bond of an embedded rod; if
restrained (by surrounding the bar with a coil of large diameter) ,
the value of the bond may be increased as much as fifty per cent,
or more. Even a loose stirrup circling the tension rod at the
bottom of a beam increases the sliding resistance of the rod, so that
a rod, covered at the most with two inches of concrete, may have
the same sliding resistance as one embedded in a large body of
concrete. Similarly, the Ransome Coil Coupling may be used
with good results when splicing rods, although the rods should
always be made in one continuous piece whenever it is practically
possible. The coupling is made simply of a coiled piece of very
heavy wire or a light bar surrounding the splice for its entire
length, which should be equal to at least fifty diameters of the
rods to be spliced (Figure 14).

11. In figure 80 a short block is shown loaded and compressed
in one direction, thereby shortening the length of the vertical
side from aa to bb. We notice now that the block expands in a
horizontal direction, the diameter increasing from cc to dd.
It requires very careful observation to discover this swelling in a
concrete block, which usually fails along a diagonal such as ae,
but, in any case, experiments with greased surfaces have shown

58 REINFORCED CONCRETE BUILDINGS

that when the friction is eliminated, the block fails along vertical
planes such as ff. 1 It is therefore clear that longitudinal rein-
forcement in the direction of the compressive force is not very
efficient, because the longitudinal rods simply add their own

strength to that of the concrete. The rods act as slender columns
and have a tendency to buckle, so that if no other provisions are
made, the strength of the rods is practically nil. To prevent
buckling, horizontal ties or " hoops " are introduced, but it is
evident that unless closely spaced the hoops are of little value.
If therefore the column or block is to have vertical reinforce-
ment, it must have closely spaced horizontal hoops, and these in
turn prevent the concrete from breaking apart along the vertical
planes ff described. In this way the hoops become a very
efficient means of reinforcing.

12. In order to understand this fully, let us consider a cylinder
filled with water, one end being equipped with a water-tight but
frictionless piston. This piston will carry an immense weight
on its upper surface ; in fact, the entire system cannot fail before
the water pressure within the cylinder exceeds the capacity of
the cylinder walls, so that the cylinder bursts. The pressure
within the cylinder is the same in all directions per unit of area;
more particularly there is a horizontal (lateral) pressure on each
and every square inch equal to the vertical pressure produced by
the load on the piston. If now the cylinder is filled with sand in-
stead of water, the conditions are only changed to this extent
that the lateral pressure against the walls is less than before, so
that it takes a greater load on the piston to burst the walls.
Finally, if the cylinder is filled with liquid concrete, and the con-
crete is allowed to set hard, the pressure on the walls will be even

1 According to tests by Foeppel and Mesnager. See, for instance, Con-
sidere, "Reinforced Concrete," page 120.

COMPRESSION AND LATERAL EXPANSION 59

less than before, but the concrete will stand much higher pressure
when enclosed in the cylinder than when free. This, then, is the
principle of the " hooped column," that the horizontal metal
jacket prevents the concrete from spreading and thereby in-
creases its carrying capacity. 1

For practical reasons it has been found impossible to use a
continuous sheet of iron around the concrete; the horizontal
reinforcement is always in the shape of hoops encircling the body
of the concrete. Under pressure the concrete is sometimes seen to
ooze out between the hoops, indicating the failure of the column,
but usually the column fails by the bursting of the hoops or the
complete disintegration of the concrete. In practical construc-
tion this need not concern us, as the stresses naturally always are
low; more 'important is the relatively great shortening of the
hooped column under working loads. This objection is overcome
by the rational use of vertical rods, so that the true " hooped
column " contains both hoops and verticals (Figure 81)

FIGURE 81.

13. The computation of a hooped column naturally centers
around the calculation of the lateral pressure against the hoops.
With a given concrete area F and a given load X the unit stress
on the concrete becomes

^ IK / K 5 wnere -X" is m Ibs.

F lbs ' /Sq - mch (and F in square inches.

If we were dealing with water, the horizontal unit pressure would
be the same. For concrete this is not the case; according to

1 Attention is called to some very interesting tests by Prof. Ira H. Wool-
son, Eng. News, 1905, Nov. 2, Steel tubes, 4" in diameter and 12" long,
| " thick walls, were filled with concrete. When seventeen days old, the tubes
were tested in compression under loads as high as 120,000 to 150,000 Ibs.
The tubes bent out of shape, and shortened 3", while the diameter increased
from the original 4" to 5". When the tubes were removed, the concrete was
found unbroken, solid, and perfect.

See also Trautwine, 1909, p. 1160.

60 REINFORCED CONCRETE BUILDINGS

experiment, the ratio between intensity of vertical stress and
transverse stress is as 1 to 1/4.8. In other words, if the load
produces a direct compressive stress of 4801bs./sq. inch, the lateral
pressure would at the same time be 100 Ibs./sq. inch. It is now
a simple matter to write up an expression giving the resistance
due to the hoops, in a granular material having this same coeffi-
cient 4.8. Let us denote by u the ratio between this resistance
and the volume of metal in the hoops, and let us denote by U a
similar ratio obtained between the resistance due to vertical
reinforcement, and the volume of the material in the verticals.
The expressions u and U will then give the effect produced by a
unit of material, used as hoops and as verticals. We find,
assuming the same stress in hoops and verticals:

^-M_24
U 2

which' shows that pound for pound, the steel employed in the
hoops is 2.4 times as effective as steel employed for longitudinal
reinforcement.

14. The question is now to find the effect of the verticals.
Assuming that they are well tied so as to prevent buckling of the
individual rods, the unit stress on the verticals must be r times
the stress on the concrete, if the sections are to remain plane as
assumed in Article 4. It is easy to see that this assumption is
on the safe side, because, if the sections curved, the stress in the
steel might be very much more than r times that on the con-
crete, which latter forms the starting-point for our investigation.
The value of r can only be indicated in a general way, as the
properties of concrete vary greatly with the circumstances; let
us assume r = 20. Then, if the unit stress on the concrete is
500 Ibs./sq. inch, the stress on the steel becomes 10,000 Ibs./sq.
inch. Let F be the area of the concrete inside the hoops, and the
allowable stress on this concrete C Ibs./sq. inch. Let p denote
the percentage of the verticals with reference to the volume of

concrete, then the effective concrete area is F- -

1UU
7?

and the area of the longitudinals F T^TT ;

hence the load carried by the concrete is C - F fnf) ^s.

COMPRESSION AND LATERAL EXPANSION 61

Let the stress on the longitudinals be S Ibs./sq. inch, then their

share of the load becomes S F - Ibs.

Disregarding for the moment the influence of the hoops, the total
carrying capacity of the reinforced column is

while if allowance be made for the hoops, the percentage of which
is q with reference to the concrete section, we have an additional

strength due to the hoops equal to 2.4 S F -^
and the total carrying capacity of the column becomes

. (2)

15. The formula (2) above is the true formula for a reinforced
concrete column and should always be used except in localities
where the building code prevents its use, in which case formula
(1) may be used. In any case, hoops must be used, otherwise
the column steel is of no value as reinforcement. For the hooped
column, Considere, the inventor and first experimenter, recom-
mends p = q = 2, which, with C = 600 Ibs./sq. inch, and r = 20,
gives

X 2 = 1400 F.

The hoops are spaced as closely as possible, leaving 1" to 2"
clear space between the hoops to facilitate concreting. The
spacing should under no circumstances exceed 1/6 of the diam-
eter of the core. Finally the core is protected with a suffi-
cient thickness of concrete to prevent rust and fire danger, about
1 " to 2 " of protection being required according to location and
exposure.

The plain column has a vertical reinforcement varying from
one to ten per cent, of the concrete area, although reinforcement
in excess of say five per cent, should be avoided on account of
the uncertainty of the strength of columns reinforced with large
amounts of steel. It is evident that hoops are indispensable also
in these columns; it is quite common to see the hoops spaced
one or even two feet apart; such hoops are of no use. The
steel cannot be depended upon to carry its load unless securely

62 REINFORCED CONCRETE BUILDINGS

tied, say, 1/3 to 1/2 column diameters apart. With p = 4,
8 = 12,000, C = 600, we have, for r = 20:

X l = 1060 F.

16. Owing to difficulties in filling columns of small diameter,
the diameter should not be much less than 10 " in any case,
although there are many 8" columns on record. On account of
the danger of " column failure " the length should not exceed
15 diameters. It is possible to advance a theory for " long "
columns, but experience shows that columns exceeding 15 diam-
eters in length are rare indeed except in roof stories where the calcu-
lations often give very light sections. Moreover, all such theories
depend alone upon theoretical considerations and have never been
conclusively tested in the laboratory, so that in the rare cases
where " long " columns are required it is better to make the col-
umn a little larger and avoid the uncertainties of the theory.

17. In tall buildings, or in warehouses, the column bars become
quite heavy, and it is necessary to join the bars of the column
above with those of the column below in a substantial manner.
The most satisfactory way is to square the ends of the bars
carefully and join them in rather closely fitting sleeves, taking
care that each bar has a full bearing on the bar below. Absolute
certainty is had by cutting threads on both bars and sleeves, and
drawing the bars together tight with the sleeve, but this must
be done with great care and under strict supervision in order to
be at all effective; unless carefully made this joint is worse
than useless. When light bars are used they may be spliced by
lapping by the required number of diameters, say about thirty,
but this method is hardly to be recommended.

Each bar of the story above should find bearing on a bar
below; the number of bars therefore increases downward in the
building. The number of bars in each story should be such that
the bars can be symmetrically arranged in the column, unless
there is some extraordinary reason for arranging them otherwise
(excentric loads). The proper arrangement of the column bars
may sometimes cause the designer to spend a good deal of time in
working out the correct solution, but he may feel assured that
this time is well spent.

The hoops may be made from round or flat stock; the round
stock may be obtained in long lengths and lends itself more readily
to the requirements of the hooped column, especially where the

COMPRESSION AND LATERAL EXPANSION

reinforcement is manufactured in the shop, with permanent
devices for coiling and fastening the hoops to the longitudinals.
The hooped reinforcement may also be bought ready-made;
quite frequently the manufacturer overlooks the importance of
having the spiral hoops in one continuous piece from top to

FIGURE 82. COLUMN REINFORCEMENT

Loomis Building, Cleveland, Ohio. Alexis Saurbrey, Consulting
Engineer

bottom, or, where the wire is joined, he makes a flimsy joint.
It must be remembered that the hoops are tension-reinforcement
and subject to all the rules governing the design of such bars.
The best joint is made by simply bending the ends of the wire
to the center of the column, making the loose end long enough
to secure the requisite grip.

The hoops may also be made in individual pieces, slipped
over the previously erected verticals and wired in place. If
the hoops are neatly made an excellent job may be had in this
way (Figure 82).

64 REINFORCED CONCRETE BUILDINGS

18. It follows from what is said above that a hooped column
should preferably be made of a circular cross-section, because
in that case the hoops are subject to direct tension only. In
many cases the expense incidental to the use of circular forms is
prohibitive; the concrete may then be made square or octagon
in section while the circular form is retained for the hoops. In
either case only the concrete within the hoops can be taken into
account in the calculations. Sometimes the hoops are made
square or rectangular, in which case they are less effective, but
we do not know how much.

19. The top and bottom of each column deserves special
attention as the tests made so far seem to indicate that these are
the weakest parts of the column, although there are many ex-
ceptions to this rule. Suitable caps and bases are inexpensive,
improve the appearance and increase the strength. Special
investigation is always necessary at points where the concrete
column finds a bearing on another material; the weight carried
by the reinforcing rods must be distributed over such an area
that the concrete in the column is not over-stressed. This is
particularly true where the column rests on the footing; a steel
base plate must be used to distribute the load on the rods, and
the concrete must be enlarged so as to bring the average pressure
within the allowable. This will be considered in detail under
"footings."

20. Before leaving the subject of hooped columns, attention
is called to the possibility of strengthening existing concrete
columns with hoops wound around the outside of the column.
In many cases it would be impossible to obtain satisfactory results
in this manner, but when the concrete is of good quality, and the
existing reinforcement is such as to give a sufficient amount of
longitudinal reinforcement in the finished column, there should
be no theoretical objections to this procedure. In practice it
would of course be difficult to wrap the core tightly, but this is
not absolutely necessary, as grout rich in cement may be forced
between the hoops and the old concrete. Great care would be
required in this operation, but it is not at all impossible, as has
been shown by actual experiments on a small scale. 1

1 Considere: "Reinforced Concrete," page 175. The prism tested in
this manner was allowed to set for three months, then wrapped with hoops
and covered with cement, and tested after ten more days. The crushing

COMPRESSION AND LATERAL EXPANSION - 65

21. In many cases columns are subject to excentric loads, so
that, in addition to the direct compressive force, a bending
moment exists and must be taken care of. This will be con-
sidered in detail in Article 81.

strength was 10,500 Ibs. per square inch. There were no longitudinals in
this prism.

CL AFTER VI
BENDING

22. THE theory of bending used for reinforced concrete beams
is different from the ordinary " theory of flexure " as used for
homogeneous beams in a few particulars only, and this difference
is more apparent than real. We consider here only the point of
maximum bending moment; this is also the point of maximum
depth, and we may assume both the compressive and tensile
resultant to be normal to a vertical section through this particular
point, under the particular loading described below.

The notations used are as follows (Figures 83, 84) :

d or D = depth from top of concrete to center of steel,

inches.
xd = depth from top of concrete to neutral axis, inches.

xd
x = ~j~ = ratio between the two preceding items.

di = distance center of compression to center of ten-
sion, inches.
E c and E s = coefficients of elasticity for concrete and steel.

Tjl

r = -pr-= ratio between these coefficients.

tic

t or T = thickness of a flange, inches.

5 = width of flange considered, inches.

B = ^iS = width of flange considered, feet.
\2i

n = thickness of stem of beam, inches.
r c and r s = deformations of concrete and steel, at extreme

fiber.

C = unit stress on concrete in outside fiber, compres-
sion, Ibs. per square inch.

S = unit stress in steel, tension, Ibs. per square inch.
a = area of steel, square inches.
St. = total pull in steel in tons.
Clj c 2 , c 3 = coefficients relating to balanced design of the

section.

a, = coefficients relating to T-beams with greater than
minimum depth.
66

BENDING

w = dead plus live load on. slab, Ibs. per square foot.
/ = span in feet.
q = factor of continuity.
M = bending moment in tons-inches.
m = 2000 M = bending moment in Ibs.-inches.

23. In regard to the load, we wi^ let all loads act in the same
vertical plane along the center line of the beam as is usually the
case in practical construction. This excludes at once all loads
which would cause the beam to rotate around its longitudinal
axis and all loads which would cause the beam to slide in its own
direction.

24. In regard to the deformations, we will consider these as
very small in comparison with the dimensions of the beam, so
that the stresses are considered as acting upon the original cross-
sections, not upon the deformed cross-sections or upon the
deflected beam.

25. This does not mean that the change of shape of the section
is of no importance. In figure 83 a vertical section is shown

\ Shortening
;= < of Top Fibre,
* Concrete

Elongatio
| of~Steel

FIGURE 83.

with the deformations produced by the bending of the beam;
we assume sections plane before bending to remain plane after
bending. Inspection of the diagram shows that the upper fibers
are shortened, the lower fibers extended under the load; the
neutral axis forms the division line between shortened and ex-
tended fibers. The assumption of plane sections is evidently
equivalent to assuming that the deformation of any fiber is in

68 REINFORCED CONCRETE BUILDINGS

direct proportion to its distance from the neutral axis, and thus
we get the equation:

r -< = ** = _^_ , 3)

r a (l-x)d l-x

26. We further assume that the stress on any small unit is
directly proportional with the deformation; this gives the equa-
tions :

for concrete C = r c E c or r c = ^r

for steel S = r 8 E s or r s = 7

___

T S S' EC

(4)

27. We shall later 1 have occasion to use the moment of inertia
of the section. It is therefore necessary to note that the assump-
tions made in the preceding paragraphs are identically the same
as those used in the " common theory of flexure " which leads
to the well-known expression

a _ ^ . e where v = stress per unit (5)

I M = bending moment

/ = moment of inertia
e = distance from neutral axis to fiber
considered.

The new feature in a reinforced concrete beam is now that in
writing up the moment of inertia we have to disregard the con-
crete below the neutral axis entirely, and instead consider the
steel area. To this we shall return later.

28. Combining now equations 3 and 4 we find

S E c

hence x

which expression determines the location of the neutral axis.

29. If now a vertical section is laid across the beam and
stresses added on and in the section to represent the removed
portion of the beam, the beam will remain in equilibrium. Let
us project all forces and stresses on a horizontal line: then the

1 Article 79.

BENDING 69

loads, being vertical, give no projections, and similarly the
stresses acting in the vertical section itself disappear. There
remain only the normal stresses acting against the section; as
equilibrium presupposes that the sum of all the projected forces
and stresses is zero, we have

horizontal component

of stresses
on tension side

horizontal component

of stresses
on compression side.

Referring now to Figure 84, the area stressed in compression is
xd inches high, b inches wide, and the average stress J C Ibs. per
square inch. Hence

total compression = \ C xd b Ibs.

Denoting by s t the total pull in the steel in tons, we have, neglect-
ing the tension in the concrete,

total tension = s t 2000 Ibs.

Hence s t 2000 = i Cxdb

, . , . Cxdb c 2 , ,

which gives st = TTTT or s t = Jo"

where c 2 = - (6)

30. Two more conditions must be fulfilled in order to create
equilibrium: (1) the sum of all stresses and forces must be zero
when projected upon a vertical line (when the loads are vertical,
Article 23); this condition we will consider later under " U-bars."
(2) The sum of all moments around any arbitrary point must be
zero. Select for this point the point of application of the com-
pressive stresses; the moment of the loads is then the " bending
moment " m inch-lbs. The moment of the stresses is 2000 s t d\
inch-lbs. We must then have

= m - 2000 s t . di

but according to the diagram (Figure 84)

d, = (1 - J x) d
hence = m - 2000 s t ' (1 - J x) d.

Eliminating st we find

m = -I Cxb (1 - J x) d*

hence d = - V inches where Ci = V I Cx (1 -| x) (7)

c\ o

Finally the steel area: a = s t square inches.

REINFORCED CONCRETE BUILDINGS

31. The formulas apply to all rectangular beams and therefore
also to slabs. As we disregard the tensile resistance of the con-
crete, the concrete below the neutral axis does not in any way
enter into the calculations at this point, and the formulas are
therefore also correct for T-beams where the bottom of the flange*
coincides with the neutral axis. In this case the thickness of
flange simply becomes

t = xd inches.

32. We have now everything required to proceed with the
design:

S t Tons

FIGURE 84.

The depth in inches:

The pull in the steel, tons:

*~^T

a =

The thickness of flange, inches: t = xd

2000
Ihe steel area, square inches: a =

(8)

(9)
(10)
(11)

Simple as these formulas are they can only be used when the
values of the coefficients x, ci and c 2 are known, and these values
in turn depend upon the allowable stresses and the factor r.
The Tables I, II, and III give full information in regard to the
values of the coefficients; it will be noticed that the same tables
may be used for any value of r, by simply shifting the position
of the S-column in relation to the values of the coefficients. On
the left the ordinarily used ^-column is indicated, corresponding
to r = 15; while on the right, the ^-columns corresponding to
r = 12 and r = 20 are show r n. Usually existing building codes and
engineers' specifications call forr = 15 in bending-problems, but

BENDING 71

this selection is arbitrary, and other values of r may very well
be used. It is impossible to predict the coefficient of elasticity
of concrete beforehand, and even if determined by careful ex-
periment there is no reason to believe that it would remain the
same on the building to be erected as in the laboratory, while it
is quite certain that it changes materially from day to day as
temperature and moisture affect the mixture used for the
concrete.

In Table IV values of the coefficient c 3 = 1 i x are indicated;
the use of this table will be clear from the analysis above. In
Table V the percentage of steel in a rectangular beam is indicated
corresponding to r = 15; when the allowable stresses are decided
upon, the percentage of steel in the section is a fixed quantity.

33. In the formulas above all dimensions are in inches, the
moment in inch-lbs., the pull s t in tons. In practical design it is
usually convenient to have the bending moment in inch-tons, M,
and the width in feet, B. The formulas then become :

The depth in inches : d = y -^- (8a)

The pull in the steel, tons: s t = c%Bd ; (9a)

The thickness of flange, inches: t = xd (10a)

2000
The steel area, square inches: a = ~ Si (11)

These formulas are different from those given above in this
respect only, that the figures handled are much smaller and there-
fore it becomes easier to avoid mistakes, as figures of two or three
places may be multiplied and divided, etc., approximately, without
the use of paper and pencil, so that all calculations are easily
verified.

34. The formulas given above apply, as stated, to slabs, to
rectangular beams, and to T-beams in which the neutral axis
coincides with the bottom line of the flange. Usually these
two lines do not coincide, so that it becomes necessary to make
further investigation in order to derive a general formula. The
formulas given above have this peculiarity, that, for a given width
of beam, the dimensions derived are minimum dimensions which
cannot be decreased without adding to the stress on the material,
thus exceeding the allowable stresses on which the design was
based. Briefly stated, the problem before us consists in finding

REINFORCED CONCRETE BUILDINGS

TABLE I. DEPTH OF NEUTRAL Axis = xd

r = 15

TABLE I x

r = 12

r =20

S = 24,000

.158

.200

.238

.272

.304

.333

.360

.385

S = 19,200

22,000

.170

.214

.254

.290

.322

.352

.380

.405

17,600

20,000

.184

.231

.272

.310

.344

.375

.404

.429

16,000

18,000

.200

.250

.294

.333

.369

.400

.429

.454

14,400

16,000

.219

.272

.318

.360

.397

.429

.458

.483

12,800

S = 24,000

14,000

.244

.300

.349

.392

.429

.463

.491

.519

11,200

21,300

12,000

.272

.333

.385

.429

.468

.500

.529

.556

9,600

18,600

10,000

.310

.376

.429

.474

.513

.546

.574

.602

16,000

C =

300

400

500

600

700

800

900

1000

r = 15

r = 12

r= 20

TABLE II. EFFECTIVE DEPTH

12.9 /"M 1 /m

d = 4/ or d = ->. -

ci y B Cl y 6

r = 15

TABLE II ci

r = 12

r = 20

S = 24,000

4.7

6.1

7.4

8.6

9.8

11.0

12.0

13.0

-S = 19,200

22,000

4.9

6.3

7.6

8.8

10.0

11.2

12.2

13.2

17,600

20,000

5.1

6.5

7.9

9.1

10.3

11.5

12.5

13.5

16,000

18,000

5.3

6.8

8.1

9.4

10.6

11.8

12.8

13.9

14,400

16,000

5.5

7.0

8.4

9.7

11.0

12.1

13.2

14.2

12,800

S = 24,000

14,000

5.8

7.3

8.8

10.1

11.3

12.5

13.6

14.7

11,200

21,300

12,000

6.1

7.7

9.2

10.5

11.7

12.9

14.0

15.1

9,600

18,600

10,000

6.5

8.2

9.6

11.0

12.2

13.4

14.5

15.5

16,000

C =

300

400

500

600

700

800

900

1000

r = 15

r = 12

r =20

TABLE III. TOTAL PULL IN STEEL

s. = 02 Bd or s f = 02 6d
* 12

r = 15

TABLE III c 2

r = 12

r = 20

S = 24 : 000

.14

.24

.36

.49

.64

.80

.97

1.16

S = 19,200

22,000

.15

.26

.38

.52

.68

.85

1.03

1.22

17,600

20,000

.17

.28

.41

.56

.72

.90

1.10

1.29

16,000

18,000

.IS

.30

.44

.60

.78

.96

1.16

1.36

14,400

16,000

.20

.33

.48

.65

.83

1.03

1.24

1.45

12,800

S = 24,000

14,000

.22

.36

.52

.71

.90

1.12

1.33

1.56

11,200

21,300

12000

.25

.40

.58

.77

.99

1.20

1.43

1.67

9,600

18.600

10,000

.28

.45

.64

.86

1.08

1.31

1.55

1.80

16,000

C =

300

400

500

600

700

800

900

1000

r = 15

r = 12

r = 20

BENDING

TABLE IV. ARM OF "COUPLE OF STRESSES."

r = 15

TABLE IV c 3 = 1 - \x

r = 12

r = 20

S = 24,000

.95

.93

.92

.91

.90

.89

.88

.87

5 = 19,200

22,000

.94

.93

.92

.90

.89

.88

.87

17,600

20,000

.94

.92

.91

.90

.89

.88

.87

.86

16,000

18,000

.93

.92

.90

.89

.88

.87

.86

.85

14,400

16,000

.93

.91

.89

.88

.87

.86

.85

.84

12,800

S = 24,000

14,000

.92

.90

.88

.87

.86

.85

.84

.83

11,200

21,300

12,000

.91

.89

.87

.86

.84

.83

.82

9,600

18,600

10,000

.90

.88

.86

.84

.83

.82

.81

.80

16,000

c =

300

400

500

600

700

800

900

1000

r = 15

r = 12

r = 20

Amount of Steel In Section

P i*T * 100
-b =12

TABLE V

r = 15

S = 24,000

.098

.167

.247

.339

.442

.553

.672

.801

22,000

.115

.196

.288

.393

.510

.636

.771

.920

20,000

.138

.231

.339

.465

.602

.750

.907

1.07

18,000

.166

.278

.406

.553

.714

.884

1.07

1.26

16,000

.204

.339

.493

.672

.865

1.07

1.27

1.50

14,000

.261

.428

.621

.839

1.07

1.33

1.58

1.85

12,000

.339

.554

.799

1.07

1.36

1.66

1.98

2.31

10,000

.463

.750

1.07

1.42

1.79

2.17

2.57

2.99

C =

300

400

500

600

700

800

900

1000

the effect on the T-beam of an increase in depth, which must, in
order to balance the design, be accompanied by a corresponding
decrease in thickness of flange and amount of steel.

REINFORCED CONCRETE BUILDINGS

35. Let, then, Figure 85a represent a section of T-shape of
minimum dimensions, having a depth d, a thickness of flange
t a , and a total pull in the steel of s a tons. Let, further, Figure
856 represent a new section with a new, larger depth D = ad.

b=12B

= xd

/Neutral Axis

<: n >

* r

ad=

/Neutral Axis |

FIGURE 85a.

FIGURE 856.

The given M and B remain the same; we wish to determine the
new values s b and T pertaining to Figure 856.

We observe, then, that the proportionate depth x of the neutral
axis is the same in the two beams, because the allowable stresses
are the same, so that the depth of neutral axis is calculated as
t a = xd in the first beam and as t b = xD in the second. The
" effective depth" in the first beam is

di = (1 | x) d
and in the second approximately

Di = (1 - %X) D = (1 - J X) ad

The approximation consists in disregarding the tendency of the
center of compression to rise on account of the removal of the
concrete near the neutral axis; the discrepancy is negligible in
most cases and on the safe side. Since now

M ,

-- and

we get the equation s a = t*s b where s a =
and, by reference to Figure 86

T-6 +

4000

C T

4000

Introducing these values in s a
an equation

(at - T) - n.

4000

sb we get after some reduction

BENDING

ST\

Solving for ( ) and denoting by

\t a J

p b ~H i t

a 2 1 r ] ' ta r

\ a ta/

ft the value of this ratio we find

n (L ^ }
b'

1 3FC |- K"~

*>*^

Neutral Axis

-Jn <&-

C =r at a-T j

T C ~^T- |

FIGURE 86.
Showing the stresses in the beam of Figure 856.

The thickness of flange in our new beam is now
T = pt a = Pxd = --x-D

and the new total pull in the steel is

a. a a*

corresponding to the new depth D = ad.

36. We have now the following general formulas for any re-
inforced concrete T-section:

12.9 I'M

The depth, in inches

D = a.

The total pull in the steel, in tons s< = -|

The thickness of flange, in inches
The steel area

a =

2000

(13)

(14)
(15)

s (

where M is the bending moment in tons-inches, a an arbitrary
coefficient larger than unit, while the width B feet may be given
or selected. The coefficient ft is derived from by formula 12
but to facilitate calculations, Table VI has been prepared giving
the values of fi for various combinations of a and n/b. This
latter ratio has little influence on the result within the ordinary
limits, and Table IX may also be used in cases where n/b is

REINFORCED CONCRETE BUILDINGS

different from 1/4, if the variation is not too large, although
prepared especially for n/b = 1/4.

37. The theory of T-beams is of great importance as all the
floor systems in common use involve this principle. Lately,
beamless floors have come into use, and to these we shall return
later; the beam and slab floors may be divided into two groups,
the first including solid concrete floors, the second what is known
as " tile-concrete " floors. The first of these two is by far the
oldest, but the " tile-concrete " is gaining in favor with every
day, and justly so, as its cost is less for light buildings owing
primarily to the simplicity of the form work. The long flat
ceilings are well adapted to modern store building and office-
structures, especially where the loads are light and distributed.
These floors have a flat portion supported on main girders A
(Figure 87), the flat portion consisting of ribs B built between

I c I cTc I I I

Section A-A

FIGURE 87.

rows of hollow tiles C and a top covering of two or more inches of
concrete, thus forming series of comparatively light T-beams
side by side. The main girders are also of T-shape, the flanges
being formed by leaving out the requisite number of tiles next
to the stem of the girder. Sometimes lighter tiles D are used
near the stem, in which case the flange becomes thinner than when
the tiles are omitted entirely, Figure 88. The commercial sizes

FIGURE 88.

of tiles are usually 12" x 12" in plan, the depth ranging from 4"
to 12" or even 16". When designing, it becomes necessary to
proportion the depth of floor so as to allow for these commercial

BENDING 77

sizes; the function of the tiles is simply to create a void in the
concrete, and they do not enter into the calculated strength of
the floor. The calculations require considerable time if exact,
and tables VII and VIII have therefore been prepared for C = 700
and 5 = 20,000 and 16,000 respectively. These tables show at
a glance the depth of tile and thickness of concrete required for
any given bending moment, together with the corresponding
pull in the steel. Note, however, that the bending moment must
be calculated for a width b of slab equal to the distance between
centers of ribs. If other allowable stresses are assumed than
those for which the tables have been prepared, we may easily
prepare new tables. We have in all the preceding formulas,
that the bending moment is directly proportional to the square
of the coefficient ci, while the total pull in the steel is directly
proportional to the coefficient c 2 . But we have

- = a coefficient times c 3

A glance at Table IV shows that c 3 itself is practically a constant
within fairly wide limits, so that, for the allowable stresses in
ordinary use, we may make

= constant.
c 2

It follows that the new tables are prepared from the tables here
given by multiplying both the bending moment and the pull in
the steel of the old table with a factor; this factor is the same
for both items and is

the new value of c 2

the old value of c 2

A completed floor of this kind is shown in Figure 89.

38. Flat Slabs. If, in formulas Sa and 9a, we make B = l,
we have the slab formulas

d = ^?i IM and s t

But the load on the slab is usually given and in Ibs./sq. foot;
denoting by w the total dead and live load in Ibs./sq. foot, the
bending moment per foot width becomes

M = - - I 2 - 12 tons-inches

REINFORCED CONCRETE BUILDINGS

TABLE VI.
INCREASING THE DEPTH FROM d TO D = ad

See Table IX for special case .- = *

a = 1.0

1.0

1.1

.64

.62

.59

.56

.50

.44

.38

.25

.08

1.2

.54

.50

.46

.41

.34

.26

.15

1.3

.47

.42

.37

.31

.23

.12

1.4

.42

.37

.30

.23

.14

.01

1.5

.38

.32

.25

.16

1.6

.35

.28

.21

.11

1.7

.32

.25

.16

.06

1.8

.30

.22

.13

.01

1.9

.28

.20

.09

2.0

.27

.18

.06

n/b =

0.0

0.1

0.2

0.3

0.4

0.5

0.6

0.7

0.8

TABLES OF =

S t TONS

TABLE VII S = 20,000

700

T " =7.0

393/20.6

6.5

372/20.0

6.0

350/19.3

5.5

276/17.1

329/18.6

5.0

204/14.9

255/16.5

308/17.8

4.5

189/14.3

236/15.7

284/16.9

4.0

62/8.3

93/10.1

132/12.1

174/13.6

216/14.8

259/15.9

3.5

55/7.7

85/ 9.6

120/11.4

158/12.7

196/13.8

235/14.8

3.0

26/5.3

48/7.2

76/ 9.1

108/10.6

141/11.7

175/12.2

210/13.6

2.5

21/4.8

42/6.8

67/ 8.4

94/ 9.7

123/10.6

153/11.5

184/12.3

2.0

17/4.4

37/6.2

57 / 7.6

80/ 8.6

104/ 9.4

130/10.2

158/10.9

1.5

14/3.9

29/5.4

46/ 6.5

65/ 7.3

85/ 8.0

107/ 8.7

132/ 9.3

1.0

10/3.2

21/4.4

35/ 5.0

50/ 5.7

66/ 6.4

84/ 6.9

105/ 7.5

0.5

6/2.3

13/3.0

23/ 3.5

33/ 4.0

46/ 4.5

60/ 5.0

76/ 5.5

0.0

1/0.6

4/1.1

9/ 1.6

15/ 2.0

23/ 2.5

33/ 3.0

45/ 3.5

H " =

BENDING

q 2000
TABLE VIII S = 16,000

C = 700

T " = 7.0

443/23.9

6.5

420/23.0

6.0

396/22.2

5.5

312/19.7

372/21.4

5.0

231/17.1

288/19.0

348/20.5

4.5

214/16.4

266/18.0

320/19.4

4.0

70/9.5

105/11.6

149/13.9

197/15.6

244/17.0

292/18.3

3.5

62/8.9

98/11.0

136/13.1

178/14.6

221/15.9

266/17.0

3.0

29/6.1

55/8.3

86/10.5

122/12.2

159/13.5

198/14.6

238/15.7

2.5

24/5.5

47/7.8

75/ 9.7

106/11.1

138/12.2

173/13.2

208/14.2

2.0

19/5.1

41/7.1

64/ 8.7

90/ 9.9

118/10.8

147/11.7

179/12.5

1.5

16/4.5

33/6.2

52/ 7.5

73/ 8.4

98/ 9.2

121/10.0

149/10.7

1.0

11/3.7

24/5.1

40/ 5.8

56/ 6.6

74/ 7.4

94/ 7.9

119/ 8.6

0.5

7/2.6

15/3.5

26/ 4.0

38/ 4.6

52/ 5.2

68/ 5.8

86/ 6.3

0.0

1/0.7

5/1.3

10/ 1.8

17/ 2.3

26/ 2.9

38/ 3.5

51/ 4.0

H" =

which gives the very convenient formulas (16)

d =-.J inches
Ci V q

and s t = c%d tons. (17)

Here the span I is expressed in feet, and the factor q is equal to
8 for non-continuous construction, and from 10 to 16 for con-
tinuous construction.

39. If reinforced in both directions, and supported on all
four sides, the slab is calculated by the formulas above, dividing
the total load w into two portions w\ and w 2 where wi -f- w 2 = w.
If we denote by LI and L 2 the span in each direction, we have the
arbitrary formulas for the division of the load:

and

= W

w<> = w

Li 4 + I* 4

The heaviest of these is now assigned to the shortest span, and
determines the depth and the reinforcement running the short
way, the cross reinforcement is designed in a similar manner
using the other load. In case of square panels the two loads
become equal, each one-half of the total.

The formula is entirely irrational and the only reason it is

REINFORCED CONCRETE BUILDINGS

TABLE IX - =|

/S = a - \/M 2 - 1)

a 2

/3/a

1.00

.000

1.000

1.00

1.01

.005

0.890

0.89

1.02

.099

0.847

0.84

.03

.015

0.815

0.80

.04

.020

0.789

0.77

.05

.025

0.767

0.75

.06

.030

0.747

0.73

.08

1.039

0.712

0.69

1.10

1.049

0.684

0.65

1.12

1.058

0.658

0.62

1.14

1.068

0.636

0.60

1.16

1.077

0.616

'0.57

1.18

1.086

0.596

0.55

1.20

1.096

0.579

0.53

1.25

1.118

0.541

0.49

.30

1.140

0.508

0.44

.35

1.162

0.479

0.41

.40

1.183

0.452

0.38

.45

1.204

0.430

0.36

.50

1.225

0.409

0.33

.60

1.265

0.371

0.29

.70

1.304

0.338

0.26

.80

1.342

0.308

0.23

.90

1.378

0.283

0.21

2.00

1.414

0.258

0.18

2.25

1.500

0.209

0.14

2.50

1.581

0.167

0.11

2.75

1.658

0.130

0.078

3.00

1.732

0.099

0.058

3.50

1.870

0.045

0.024

4.00

2.000

0.000

a 2

/3/a

given here is that it is on the safe side and better than other
existing formulas.

The supporting girders are designed with reference to the
load brought upon them by the particular direction which they
support, and the bending moment is often increased above the

BENDING 81

calculated because the load seems rather more concentrated
towards the center.

FIGURE 89. TILE CONCRETE CONSTRUCTION, READY FOR
PLASTERER

Wise Building, Cleveland, Ohio. Alexis Saurbrey, Consulting
Engineer

Discussion of Tables I to IX.

40. TABLE I. x =

-3

The value of x determines the location of the neutral axis,
xd being the distance from compression face to neutral axis.

82 REINFORCED CONCRETE BUILDINGS

We have seen that the position of the neutral axis within the
section of a T-shaped beam leads to the division of T-beams into
two groups, according to whether the neutral axis falls, above
or below the bottom line of the flange. In this latter case we
introduce the coefficients a and /?, and the problems contain an
arbitrary element which is absent in beams of the first type,
where the dimensions depend mutually upon one another as in
formula (8). The table shows that the neutral axis only in
exceptional cases approaches the middle of the beam where it is
located in all symmetrical beams following Hooke's Law (steel
for example). It is obvious that a greater amount of steel is
required for low steel stresses than for high; we therefore see
that the neutral axis is lowered by increasing the amount of steel.

41. TABLE II. Cl = \J\Cx (\ - x

, 12.9 I'M
We have d = -

The smallest possible value of d is obtained when a large value
of Ci is used, or, in other words, when high concrete stresses are
combined with low steel stresses. The influence of the concrete
stresses is much more pronounced than that of the steel stresses;
it is, therefore, not economy to increase the amount of steel in
order to save on the concrete. It is not impossible to analyze
this problem mathematically, but owing to variation in unit
prices it seems hardly worth while. The possibility of decreas-
ing the depth of construction by using high concrete stresses
and low steel stresses may, however, be of importance in special
cases where the head room is limited.

Cx
42. TABLE III. C2 = 333

The total pull in the steel is st = 02. ^ d.

The total amount of steel is a = s t

2000
S

2000 bd fa\fS

so that -- ' C2 ' 12 r C2 = (bd) VT66

The coefficient c 2 , then, is a measure of the amount of steel used

BENDING 83

for a given cross-section, bd being the area of the cross-section
in square inches. We note that

^-, times 100
bd

is the percentage of steel required for the beam; if we denote
the percentage by p we have

16600

This expression has been used in calculating Table V.

43. TABLE IV. c 3 = 1 - I x

In general the total pull in the steel is obtained by dividing
the bending moment by a certain lever arm di, equal in length
to the distance between the centers of compression and tension.
Reference to Figure 84 gives at once

di = | xd + (1 - x) d = (1 - \x) d = c,d

When d is known we get d\ as c$d; Table IV gives the values
of c 3 for various combinations of stresses.

44. TABLE V. p = ^ X 100 = c 2

The percentage of steel has but little interest for the prac-
tical designer as the problems usually present themselves.
The table is added for the convenience of those who are in the
habit of selecting the percentage of steel rather than determine
the allowable stresses. The table is correct only for such beams
where a = 1 and r = 15.

45. TABLE VI will be found useful when designing T-beams
of larger than minimum depth. When we have selected , as
explained in connection with formula 12, etc., the correspond-
ing ft is found by Table VI for any value of n/b. The method
of design will be clearly evident from the example in Article 47.
Table IX is a more extensive table for the special case where
n/b = J, which is a common value in practice. The variation
of n/b does not affect the values very much, so that for small
values of a Table IX may be used for other values of n/b than J.

46. TABLES VII and VIII. Tile-concrete floors.

The use of these tables is best explained by an example.
The span of the flat portion is 20 feet; the total dead and live

84 REINFORCED CONCRETE BUILDINGS

load is assumed to be 250 Ibs. per square foot. With 4" ribs
we get the width of beam

b = 4 + 12 = 16"
and the corresponding bending moment in inch-tons

250
M = * x ^ x 2 2 X 16 = 100 inch-tons.

If the allowable stresses are S = 20,000 and C = 700, we must
use Table VII, and we see at once that we can use either a 10 "
tile with about 2f" concrete, or a 12" tile with 2" concrete.
As we do not wish to have less than 2" of concrete over the tiles,
we cannot use the larger tiles economically. If we select 12"
tiles and 2" concrete, the corresponding pull in the steel is 9.4
tons according to the table, requiring .94 square inches of steel,
for instance, one f " square and one f " square bar.

It should not cause surprise that the moments tabulated in
Table VIII are larger than the corresponding values of Table
VII, although the allowable stress on the steel is smallest in
Table VIII. The explanation is given in the remarks under
Table II, Article 41, and, in accordance with the statements
made there, it will be seen that the larger moments of Table
VIII are obtained only by increasing the steel areas.

TABLE IX. See Table VI, Article 45.

47. EXAMPLE 1. T-Sections. Continuing the example given
above under the discussion of Tables VII and VIII, we proceed
as follows to design the girder:

The load on the floor is 250 Ibs. per square foot, the span of
the flat portion on each side of the girder is 20' 0", and the
girder therefore carries a load of 250 X 20 = 5,000 Ibs. per
lineal foot, to which should be added the weight of the girder
itself. Assuming this item to be included in the 5,000 Ibs., and
assuming a span of 24' 0" for the girder, the bending moment
on the girder becomes

M = J X x 24 2 X 12 = 2160 inch-tons.

We decide to use high tension steel, for which S = 20,000, and
we allow C = 700 Ibs. per square inch on the concrete. We get
then from Table II : Ci = 10.3; from Table III : c 2 = .72; and from
Table I: x = .344, and we may now proceed with the design,
using formulas (8a), (9a), (lOa), and (11). The width of flange,

BENDING 85

B, may be selected arbitrarily. Let us make B = 4' 0".
Then, by

(9a) ......... s = c 2 BD = .72 X 4 X 29.1 = 84 tons.

(11) ......... a = ? - s t = X 84 = 8.4 square inches.

(10a) t = xd = .344 X 29.1 = 10"

We have to make the stem of the beam wide enough to accom-
modate 8.4 square inches of steel, say n = 12", and the girder
is then designed as far as concerns the bending moment. Ques-
tions pertaining to shear, etc., will be considered later.

We can, if we desire, reduce the thickness t of the flange by
increasing the depth d. While this operation is not always neces-
sary, or even desirable, we will nevertheless continue the ex-
ample to show the method of procedure.

If, then, we increase the depth from 29.1" to, say, 35", we get

a = J^_ = a6a>2

the coefficient a indicating the proportionate increase in the
depth. The value of ft is next obtained by Table IX, remem-
bering that

i _ OK
6 " *~

the stem being 12" wide and the flange 48".

For - = .25 and a = 1.2, Table IX gives ft = .43; then,
6

by the theory outlined for this case, Article 36, we have
The new depth D = ad = 1.2 X 29.1 = 35".

s 84
The new pull in steel s& = = y-x = 70 tons.

The new thickness of flange T = ftt = .43 X 10 = 4.3".

It is of course unnecessary to calculate the dimensions of
the " minimum " beam first, as done here, unless we expressly
desire to have these dimensions. Let us, for instance, again
consider a given bending moment of 2,160 inch-tons; let us

REINFORCED CONCRETE BUILDINGS

further select arbitrarily the width B = 4' 0", and let us
finally choose the coefficient a = 1.2, then, by Table IX, we

get ft = .43 for an estimated value of = .25; and we also

have a2 = 1.44. We may now find the dimensions directly, by

12.9 /M 12.9 2l60

'^'V- =

12.9
<X

(14)

8t = -* . B - D =

1.44

03
X 4 X 35

70 tons.

........ T== a" :r ' /)== r x - 344 x 35 = 4 - 3 "

Figure 90 shows the resulting construction. The space Z

FIGURE 90.

under the flange may be taken up of tiles with less depth than
those used for the balance of the floor, as long as the thickness
of concrete over these lighter tiles is made equal to the thick-
ness of flange just found, or thicker. In this way, the construc-
tion of the girder is lightened somewhat, and the form work
may be made of the same light construction up to the face of
the stem. A floor constructed in this manner also presents a
tile surface to receive the plaster in all places except on the
stem of the girder.

If we now wish to check our calculations, we may proceed
as follows:

From Table IV, we get at once c 3 = .89, and the lever arm
of stresses becomes

D! = c 3 D = .89 X 35 = 31.2"
and we get the total pull in the steel

M 2160
S < = Si = 3L2 = 69 ' 3

BENDING 87

The calculations above gave 70 tons. We now have to find the
compressive resistance of the beam, and this we get as the dif-
ference between two items : A = the total resistance of the entire
area above the neutral axis, and B = the section cut out be-
tween the neutral axis and the bottom of the flange. First we
find the location of the neutral axis

xD = .344 X 35 = 12.25"
and we get then

A = \ X 48" X 12.25" X 700 = 205,800 Ibs.
To find B, we must first find the concrete stress at the bottom
of the flange, and by reference to Figure 86, we get

C T = 700 x 12 ^2 ^ 5 4 ' 3 = 70 x i^nl = 454 lbs - p er s <i- inch -

When determining B, we must remember that the " width of
beam " is not 48", but 48" - 12" = 36", the 12" being the
width of the stem, and we get now

B = \ X 36 X 7.95 X 454 = 65,000 lbs.
Then

A - B = 205,800 - 65,000 - 140,800 lbs. = 70.4 tons.
The total compression must of course equal the total resist-
ance, and we see that our design is correct as this is the case.
The slight difference between the 70 tons of the design and
the 70.4 tons of the check is due to the inaccuracies of the slide
rule and the various interpolations made, and is entirely too
small to warrant further investigation.

48. EXAMPLE 2. FLAT SLABS.

Given: Live plus dead load 500 lbs. per square foot. Span
12' 0" between centers of support. Allowable stresses,
concrete, 600 lbs., steel, 14,000 lbs. Continuous construction
U=10. - ).

We get from Table II: ci = 10.1; Table III: c 2 = .71; and
we can now proceed with the design using formulas (16) and
(17), and we get

I /w 12.0 /500
~c 1 '\- q == Wl'\^'-

and s t = c 2 d = .71 X 8.4 = 5.96 tons per lin. ft. of width.

To the depth must now be added the amount of concrete required
to properly protect the rods.

REINFORCED CONCRETE BUILDINGS

49. To Find the Stresses in a Given Beam, when the bending
moment is known, requires a knowledge of the ratio S/C and of
the ratio r. A simple mathematical analysis of the beam gives
the first ratio, as we shall see presently; the value of r we cannot
determine, so that it will have to be assumed in the same manner
as when designing. The value 15 may very well be used.

We found above the expression (Article 35, Formula 12)

-n/b

while Article 36 gives the means for eliminating in this expres-
sion first (3 and a; by means of (6) and (11), c 2 and s t are elim-
inated. The resulting equation may be solved for S/C, and
gives :

T'H'ii-'^}-

(18)

The quantities entering on the right side of the equation mark
are all known except r; in Figure 91 the several dimensions are

FIGURE 91.

shown; the value V is written as an abbreviation of ra/6, so
that

represents the thickness of an imaginary strip of concrete hav-
ing the same width as the beam considered, and the same resist-
ance as the tension steel; a and b are taken directly from the
drawing, same as the other dimensions.

In special cases this formula is greatly simplified, although
there is no difficulty whatever in using the formula given.

BENDING 89

(1) For rectangular beams, or slabs, we have n = b and
T = and we get

J'-i + J^ 1 ? M

(2) Disregarding the influence of the stem on the com-
pression, as is sometimes done, we have n = and get

S = 2D - T

C~ ?T\ 2 Da (20)

\rj Tb

(3) If, in Formula 18, T 2 = 2V H, we find after some re-
duction

S_ = Tb
C ~ 2a

The- value of n/b disappears entirely, which evidently means
that the neutral axis coincides with the bottom of the flange.

The ratio ^-r r is therefore the criterion of the section. If
2 V H

= 1, the neutral axis coincides with the bottom of the flange,
if < 1, it falls below, if > 1, above the bottom of the flange.

If we now wish to determine the stresses in a given beam, we
begin by selecting r, next we determine the value of the cri-
terion, so that, if equal to unit, we use formula (21), while if
larger than unit, we use formula (19), and if smaller, the orig-
inal formula. Then the location of the neutral axis is calcu-
lated from

and the coefficient c 3 = 1 f x is determined from the value
of x just found. The effective depth is then

where d is the depth from ultimate compression fiber to the cen-
ter of the steel. We have then

s =-

and, by (11)

2000

REINFORCED CONCRETE BUILDINGS

We know now S and S/C, and it is a simple matter to determine
C.

50. EXAMPLE 3. T-SECTION.

Given the beam shown in Figure 92, find the stresses, when
the bending moment is 2,160 inch-tons. We have

T 2

4.5 X 4.5

= 20.25

2 VH ~ 2 X 2.19 X 30.5 " 113.5

V 7 sq. inches

FIGURE 92.

which is -evidently < 1. Using therefore formula (18) we find

325.4

_ 15X
C ~ 5 X 168.4

Now

x =

^ = .342

and c 3 = 1 - J a; = 0.886.

Hence di = c 3 d = 0.886 X 35 = 31"

Bending moment 2,160 inch-tons, then
2160

= 69.6 tons.

69.6 = 20,100

20100 ,

c = W = 695 '

or about 20,000 and 700 Ibs. /square inch for steel and concrete,
respectively.

51. EXAMPLE 4. T-SECTION. Special case.

Given the beam shown in Figure 93.

We have V = = 15 X ~ = 2.62.

10 X 10

= 100
2 VH ~ 2 X 2.62 X 19.1 == 100

= 1.

BENDING

Use formula (21) which gives

S 10 X48
C =

= 28.6

2 X8.4

The balance of the calculations may now be continued ex-
actly as in the preceding example.

j-8.4 sq. inches

FIGURE 93.

52. EXAMPLE 5. RECTANGULAR BEAM. Slabs.
Given the rectangular beam shown in Figure 94; we use
formula (19) which gives

C 2 + 2\

V_0.85 sq. inches

FIGURE 94.
225 + ^ '

0.85

CHAPTER VII

TRANSVERSE STRESSES. U-BARS

53. IN addition to the longitudinal stresses examined in
the preceding articles, transverse stresses exist in reinforced
concrete beams as well as in beams of other materials. But
the transverse stresses are different in trusses and in solid beams :
in the truss, each individual member is stressed in its longi-
tudinal direction only, and there is no shear. In the solid beam,
longitudinal stresses exist in the top and bottom chords or
fibers, and the web is then subject to shear stresses both longi-
tudinally and transversely. In special cases these shear stresses
may vanish, as for instance in the I-shaped steel beam of vari-
able depth, when the ratio

bending moment at any point
depth at the same point

is a constant. This is the case in a parabolic girder loaded
over its entire length with a uniformly distributed load.

54. In view of this difference between trussed beams and
solid beams, it becomes necessary to decide whether to treat
the reinforced concrete beam as the one or the other. To the
eye a reinforced concrete beam certainly appears solid enough,
and such is indeed the case when the beam is first made and the
load is being put on. But when the load reaches a certain in-
tensity the " solidity" of the beam is destroyed. Slight cracks
soon become evident, at least when arrangement has been made
to observe them, and that under loads corresponding to a steel
stress of from 4,000 to 6,000 Ibs./square inch, or a concrete
stress of 350 Ibs./square inch. It follows that under the ordi-
nary working load our reinforced concrete beam is perforated
with cracks extending from the bottom fiber up toward the
neutral axis, without quite reaching the neutral axis, so that,
under any circumstances, the beam is certainly not a "solid"
beam. These hair cracks have been noted by all who have

TRANSVERSE STRESSES. U-BARS 93

taken the trouble to look for them with but one exception
(Considere) ; they are not an occasional occurrence, but a uni-
versally recognized phenomenon of the greatest importance for
our understanding of the stresses within a reinforced concrete
beam. The presence of these cracks is accounted for by the
simple fact that concrete is unable to stretch as much as steel
before cracking, so that, under a certain load, the concrete refuses
to follow the steel in its elongation and goes to pieces. The
cracks of this class appear throughout the length of the beam,
fairly uniformly spaced, and increase in size with increasing
load.

55. The crack of course is an open space existing between
surfaces which at some earlier time were in close contact and
united. We must now understand as a fundamental principle
that stresses cannot be transmitted through open cracks. Com-
pression may be transmitted through a contact only, and fric-
tion may exist on surfaces pressed together, but no kind of
stress will jump across an open space. It follows that shear
in the ordinary sense of the word cannot exist in a reinforced
concrete beam loaded above a certain limit, because the nature
of shear requires equal intensity on a horizontal and a vertical
plane, and this is of course impossible when the beam has ver-
tical cracks. Or, we may simply say that the vertical shear
cannot exist in the crack itself. Where a crack occurs there is
therefore nothing but the compression flange and the tension
steel to carry the shear, a distribution of the shear which is,
to say the least, not easily reconciled with current ideas of shear
in solid beams.

56. Entirely different from these hair cracks are the much
larger, pronounced failure cracks which predict the approach-

.1 t

FIGURE 95.

ing collapse of the test beam. If located at or near the point
of maximum bending moment, they are undoubtedly due to
excessive elongation of the steel disclosing a failure by tension
in the steel; if near the end, the crack usually takes the shape
shown in Figure 95, either with or without the horizontal crack

94 REINFORCED CONCRETE BUILDINGS

D. The vertical crack E is wide open, especially at the bottom,
decreasing in width as it approaches the top of the beam. As
the steel stress at this point certainly cannot exceed the steel
stress at the point of maximum bending moment, this crack is
not due to excessive tensile stresses in the steel. It must be
due to sliding of the reinforcement: the steel is pulling out of
the end of the beam at the same time bursting its concrete
envelope, and causing the horizontal crack. Let it be under-
stood that no amount of shear will cause a gaping crack, but once
sliding sets in and causes the vertical crack, it is clear that the
one end of the beam will be compelled to revolve around the
other end, causing in the first place the double-curved line
of cleavage, and, secondly, great friction on the surfaces of
contact.

57. The above remarks lead to the conclusion that a concrete
beam is a solid beam up to a certain load at which point the
tensile resistance of the concrete is exhausted, and a readjust-

FIGURE 96.

ment of stresses takes place within the beam. This readjustment
is different for different types of beams.' In a rectangular
beam (Figure 96) we may well assume that the com-
pression follows lines as AC and BC when the load is placed

FIGURE 97.

at C; if the load moves to D, the lines change to AD and DB.
Under these circumstances there is no shear, at least not in the
ordinary sense of the word. We may compare a system of this

TRANSVERSE STRESSES. U-BARS

kind to a triangular frame with hinged corners (Figure 97).
The chords AC and BC will be in compression, and the chord
A B in tension, hence at A and B the hinges are subject to severe
stresses. The same is the case at A and B in the reinforced
concrete beam, so that the " length of embedment " AE and BF
in Figure 96 must be made long enough to prevent sliding of
the rod. The shear existing in a system of this kind is that
negligible quantity caused by the stiffness of the system as a
whole, a kind of friction caused by the lack of flexibility at the
supposed hinges.

The system A BC is an equilibrium curve for the load C
and the reactions A and B', this same argument would of course
hold true for any number of forces, or even for uniformly dis-
tributed loads, in which latter case the compression curve would
be a continuously arched curve from A to B (when the load
covers the entire span). But if the beam under consideration
is a T-beam instead of a rectangular beam it becomes impos-
sible to make the compression line curve down to the support-
ing points, except for a width equal to that of the stem. A
T-beam (Figure 98a) may be considered as consisting of two

FIGURE 98a.

FIGURE 986.

FIGURE 98c.

beams side by side; a T-beam proper (Figure 986) and a rect-
angular beam (Figure 98c). In this rectangular portion it is
quite possible for the compression lines to dip down at the sup-
ports, but not so for the T-beam portion, there being no con-
crete left to carry the stresses down to the steel. This leads
to the idea of bending the steel up over the support to meet
the compression flange, reversing the conditions shown in
Figure 96.

58. Let us consider a portion of a reinforced concrete beam
between two points a and b (Figure 99). The bending moment
at a is denoted by M a , the distance between center of com-

REINFORCED CONCRETE BUILDINGS

pression and center of tension by d a , then the total pull in the
steel at a is

M a
Sa = -T-

and at b

s b =

d b

FIGURE 99.

The difference between these two is

_M a _ M b
' Sa Sb - T a ~d b

The prefix A simply denotes the difference in the item con-
sidered so that As means the variation of s between the points
in question.

It is now evident that the portion abed is subject to two
pulling forces acting near its lower end cd: one force s a pulling
toward the left, another pulling toward the right, s b . If s a is
larger than s b , the end cd must have a tendency to move toward
the left in precisely the same manner as if pulled that way by
a force equal to the difference of the two pulling forces; we
may therefore consider the force As = s a s b as acting alone.
This condition is represented in Figure 100, and this diagram

f 1

f )

As^

^j ^ Al

FIGURE 100.

shows at once that cdef is a cantilever fixed at its base ef, and
loaded near its end with a load As. The depth ce we do not
know at the present time; let us indicate this unknown quan-

TRANSVERSE STRESSES. U-BARS

tity by h. The bending moment on the cantilever is then h As;
the arm of " the couple of stresses " in the cantilever is c 3 AZ;
hence if a vertical reinforcing rod is disposed near bd the pull on
this rod becomes

But this pull k can exist only when counterbalanced by a cor-
responding compression, so that the beam becomes a trussed
beam as shown in Figure 101. The vertical reinforcement

J /]

xx K |>

h K

XI Fl

X X 1 ^ 1

I N N

* 1 x 1^
\t \| \j

A t

R a

C!R C

FIGURE 101.

designed in this manner is usually made of a bar bent to U-
shape and circling the main tension rod (Figure 102a, 6); they
are therefore called U-bars or stirrups. The U-bar is unneces-

FIGURE 102a.

FIGURE 1026.

sary when k = 0, which is always the case when As = 0: i.e.,
when either the tension chord, or the compression chord, or
both together, follow the equilibrium curve. As shown above,
this is always the case in a rectangular beam with well-anchored
reinforcement, and it is also the case for such parts of a T-beam
in which the reinforcement is bent up to follow the equilibrium
curve. In all other cases k has a definite value. For straight
reinforcement and straight top chord, we have (Figure 103) :

d a = db = Cj,d

M a - M b

hence

As =

REINFORCED CONCRETE BUILDINGS

where

and

hence

M a = Ra -
M b = R (a - AQ - 2P (p - AZ)
M a - M b = AZ (R -

A S = - (R -

and k = ^ (R -

It is now easy to understand that the length h cannot exceed

1r.

r sr

a-Al

FIGURE 103.

the distance from the center of the steel to the neutral axis.
This gives

h (1 - x)d 1 - x

The value of this expression cannot exceed unit; for ordinary
cases its value is about three-fourths. Hence the maximum
possible value of k, for the conditions named, is:

kmax == K Zr. (23)

A ,

mi:

FIGURE 104.

59. It is now interesting to note that this same expression
may be obtained directly in the simplest manner. Let, in
Figure 104, the section A B remove the right end of the beam

TRANSVERSE STRESSES. U-BARS 99

leaving the main tension bar and the U-bar projecting. The
stress-resultants acting upon A B are then, when the chords are
parallel: the horizontal compression X, the horizontal pull
Y'j and the vertical force k in the U-bar. The loads are PI, P 2 ,
etc., and the reaction R. If we now project on a vertical line
MN, the horizontal stresses vanish and we have
R- (P! + P 2 + P 3 + ...... ) =k

or k = R - 2P.

60. The beam with straight top- and bottom-chords is an
exception. Usually the stem of a T-beam may be considered
as in equilibrium, and in addition some of the bars are bent
up in the T-beam portion to approximate the equilibrium curve,
so that a material reduction in the value of k takes place in all
practical beams. With the notations of Figure 98 we have,
on account of the stem, a reduction equal to

b n , j b n ro vrn

r hence k = r [R 2P]

If out of the total number x of bars in the T-beam, a certain
number y follow the equilibrium curve, we have a further re-
duction equal to

fc = ?LH . *r_ . [R _ 2P] (24)

x x b

Thus, if b = 48" and n = 12", we have

b - n _ 48 - 12 _ 3
~b~ 48 ~ 4

so that, out of a total number of say eight bars, the six belong to
the T-beam. If out of these six, two are bent as required, we
have x = 6 and y = 2, hence

k = i X i X [R - 2P] = } - [R - SP]

61. Thus, in order to calculate the stress on the U-bars,
it becomes necessary to know the properties of the curve of equi-
librium for the system. When the loads are stationary, the
curve is drawn as a force polygon to the actual loads and reac-
tions. For a uniform load, covering the entire span, this curve
is a parabola; it is not practical to bend the bars to this shape,
but it may be closely approximated by a system of bars with
straight portions between the several bents. A uniformly

100 REINFORCED CONCRETE BUILDINGS

distributed, moving load has no definite curve of equilibrium,
so that in that case the most dangerous position of the load
must be found and the U-bars proportioned according to For-
mula 23 above, while the bent bars are arranged to meet the
requirements of some particular type of loading, for instance,
the total load. Similarly, concentrated loads may be either
stationary or moving. In buildings the concentrated loads
are usually stationary. The given load is a uniform load, so
that the beams are loaded as explained above; these beams
in turn frame into the girders, one, two, or three beams to each
span, and these concentrated beam loads are stationary. It
is a simple matter to bend the main tension bars to conform to
this type of loading; examples are given in Figures 105 and 106.

I I

FIGURE 105. FIGURE 106.

The moving concentrated load is usually found only in structures
like highway bridges, subject to steam-roller traffic, in crane-
track girders, etc. In such cases, the live load is large in pro-
portion to the dead weight of structure and covering, so that
the T-beams are usually not economical structures for this class
of girders. They may be constructed by using the adequate
number of U-bars; or rectangular beams may be used of the
required cross-section.

62. The problem of designing a T-beam under a uniform
load confronts the reinforced concrete designer every day. It
is customary to consider the load as covering the entire span,
except in cases where it is expressly stipulated that the most
dangerous position of the load shall form the basis for the cal-
culation of the U-bars. Arguments may be advanced pro et
con., usually the load specified is a maximum load which
seldom, if ever, covers the entire beam, and the designer will
have to use his best judgment as to what constitutes proper
practice in each individual case. It is hardly necessary to say
that in other lines of engineering the most dangerous condition
is always considered in making the calculations as a matter of
course, and there is no reason why other professional ethics
should prevail when dealing with reinforced concrete.

TRANSVERSE STRESSES. U-BARS

101

In Figure 107 the moment-curve is shown corresponding to
a uniformly distributed load covering the entire span. The
maximum moment is taken as unit, and the several ordinates
of the curve are given under the assumption that there is no
continuity. The reinforcement must be made to conform to

n 12

.75

.89

1X0

.97

FIGURE 107.

this curve as closely as possible, hence we see that at points
3 and 9, only f of the total number of bars is required, at 2 and
10, slightly more than J, and less than J is required at points
1 and 11. The quota of bars not required may and should be
bent up at the points specified, provided that no other kinds of
loading can occur. In Figure 108 the corresponding curve is

FIGURE 108.

shown when the beam is considered as continuous, with q = 10.
63. The entire theory outlined for the calculation of the
U-bars is based upon the assumption that sliding of the steel
cannot take place. In such cases where the anchorage beyond
or at the supports is insufficient to prevent sliding of the main
tension bars the factor of reduction must be decreased, so that a
correspondingly larger amount of vertical reinforcement is used
for the U-bars. In the present state of our knowledge this must
be taken care of by judgment alone, there being no way of cal-
culating a beam with inefficient anchorage. It must here be
sufficient to point to the fact that the U-bars retard the sliding
of the reinforcement, and that, for that reason, light U-bars should
always be used even in cases where the theoretical considerations
show that they may be dispensed with. This applies particularly
to rectangular beams.

102 REINFORCED CONCRETE BUILDINGS

64. Spacing of the U-Bars. It will be noted that the entire
line of argument advanced in the preceding paragraph is based
principally upon the inability of the concrete to resist tensile
stresses, and that the entire problem finally resolves itself into
one of tension carried entirely on the steel, and compression
carried entirely on the concrete. The word " shear " is referred
to incidentally only, and this is a natural consequence of the
fundamental principle of disregarding the tensile stresses in
the concrete. As this development leads to rather important
results, it may be well to consider these matters a little more
in detail.

65. Figure 109 shows the simplest conceivable system of
material units, i.e., three particles, A, B, and C. Whatever

the nature of the force uniting these par-
tides, if the particle C is moved to the
position D through the influence of some
\ external force, the displacement CD rep-
. \ resents in all cases the result of the influ-

__ _\; ence of that force and is called the "shear

A B deformation " if parallel with the line

FIGURE 109. . T , . ,., , ,,

AD. It is readily seen, however, that

the more direct and more readily understood deformations
are (1) the lengthening of AC to AD, and (2) the shortening of
BC to BD. Hence this shear deformation CD is nothing but
the resultant of the deformations along the original lines AC
and BC, and we perceive that even in the most complicated
system of particles any deformation may be reduced to a sys-
tem of lengthenings and shortenings, that is, tension and com-
pression, if we speak of stresses instead of deformations. The
word " shear," therefore, has no real or material meaning,
except as a pure figure of speech to express in one short word
a rather intricate condition of tensile and compressive rela-
tions, in precisely the same manner as the word " bending
moment " is used to indicate a mathematical conception of the
mutual condition of a number of forces acting upon a beam.
Needless to say that nobody has ever seen, or will ever see, a
bending moment in the realm of things as they are, and that
whoever undertakes to explain the so-called " shear stresses "
in a solid body will ultimately have to account for pure tensional
and compressional stresses.

TRANSVERSE STRESSES. U-BARS 103

66. If two material bodies are in contact, the stresses act-
ing in the contact surface are termed frictional stresses which,
as far as the materials themselves are concerned, are compres-
sive stresses with no possibility of accompanying tensile stresses
in the direction perpendicular to the contact surface.

67. Of the nature and extent of frictional stresses we know
next to nothing. A force acting parallel with the contact sur-
face will cause sliding of one body in relation to the other; if
the force is inclined, the sliding becomes increasingly difficult
as the angle of the force increases, and the sliding becomes
impossible when the angle at which the force acts exceeds the
" angle of friction," which has a definite value for each material,
depending in part upon the character of the surface. For con-
crete upon concrete, this angle appears to be near 41.

68. In certain types of reinforced concrete construction the
floor beams are not made in one continuous operation with the
floor slab resting upon the beams, and U-bars or similar mechani-
cal devices are then resorted to in order to tie the slab and stem
together, and to so unite them that they may be considered
as acting as one piece. In this case, the slab would form the
upper flange of a T-beam, and in order to insure this action,
sliding between flange and stem must be prevented. Figure
110 represents a portion of a beam, the lines AC, CD, and CB

c P

'Mf ;|i *" |, \M

FIGURE 110.

indicate the directions of the principal stresses. If now the
line of diagonal compression BC is inclined so that the angle
BCD is less than the angle of friction, the flange would slide
in relation to the stem, on account of the joint along the line
MM; the U-bars AC and BD would resist this tendency by
virtue of their " shear " resistance (and this resistance we know
is very small, and cannot exceed the compressive edge resist-
ance of the concrete; see Figures 111, 112, where the black areas
indicate the crushed concrete). If, on the other hand, the angle
BCD is larger than the angle of friction, then there can be no

104

REINFORCED CONCRETE BUILDINGS

sliding, and therefore no shear stresses on the U-bars, which
will act directly in tension as described above. The rule derived
from this argument may be briefly expressed thus: The spacing
of the U-bars must not exceed the depth of the beam, in which
case the angle of forces would be about 45.

FIGURE 111.

FIGURE 112.

69. If we now turn to the T-beam manufactured in one
continuous operation, where no separation exists between stem
and slab, we note that, theoretically at least, this beam is in
the same condition as the one just considered, owing to the orig-
inal assumption whereby the tensile stresses in the concrete
are considered as non-existing. Each and every horizontal
stratum must be considered as isolated and influenced by its
neighbor through the medium of frictional resistance only,
and the direction of the diagonal compression must be such
that no sliding can take place. The same rule must therefore
be imposed in this case.

But this rule gives the maximum spacing possible: owing to
the usual considerations of a margin of safety, the spacing must be
made smaller, and we would therefore recommend that the spacing
of the U-bars must in no case exceed one-half the effective depth of
the beam.

70. Tensile Stresses in Concrete Disregarded. The ready-
made reinforced concrete beam formulas now in common use
are derived under the apparent assumption that the steel rein-
forcement takes all the tensile stresses, and this is also the case
in this book. In reality, we cannot wholly disregard these ten-
sile stresses in the concrete, or, at least, we cannot deny their
existence, because if we did, we would also rob the concrete
of its cohesion, and we would have a granular mass such as sand
or crushed stone, wholly unsuitable for our purpose. The true
statement is that we disregard the tensile stresses in certain
directions and for certain purposes. In this book, we have
considered the concrete as fractured vertically along the planes

TRANSVERSE STRESSES. U-BARS 105

of the U-bars (1) because the cracks in probability will appear
in the weakest plane, there being less concrete to resist the ten-
sion where the concrete is displaced by the steel of the U-bar;
(2) because the U-bar encircling the main tension rod in a meas-
ure acts as a washer on the rod, causing the somewhat resilient
concrete to crack immediately behind this point of gripping;
and (3) because such tests as throw any light upon the location
of the cracks indicate that they occur very largely at just these
points.

71. NOTE: For the gripping action of a loose U-bar encir-
cling the tension rod, see Morsch, page 47.

For the location of the cracks, see the same book, page 155. *

72. We have also considered the stem of our T-beam ' as
composed of horizontal layers acting upon one another by con-
tact only, and thereby determined the spacing of the U-bars.
But between these vertical and horizontal lines of weakness,
we have assumed the concrete to be solid. Hence, we have
assigned to the concrete a certain amount of tensile resistance
in certain locations and directions.

73. It follows that with increasing loads the compressive
stresses in the beam do not increase as rapidly as the load,
especially not in beams where the slab and the stem are sep-
arately manufactured. In such beams, the compression at
rupture must in many cases be uniformly distributed over the
entire compressive zone, and we find here the explanation of
the fact sometimes observed that the compressive strength of
concrete is much higher in a beam test than in a cube test. An
analysis of these conditions would be interesting and of great
value practically.

74. Details of Reinforcement. The various arguments
advanced above will lead to rational design of the steel if con-
sistently applied, and there is but little new to add. The great
principle in all beam construction is that there is a compres-

1 Morsc^i: Concrete Steel Construction, 1909. While the cracks do not
all occur at the U-bars, the tendency is fairly pronounced, especially in the
beams with U-bars in one half only, see Figure 149, Beam V; Figure 153,
Beam VIII; Fig. 154, Beam IX; Fig. 157, Beam X; and compare the cracks
in the U-bar end with those of the other end of the same beam. The draw-
ings of all these beams show them just before final collapse, while our calcu-
lations have reference to a much earlier stage, viz., under the working load,
or at the most a load not more than twice the working load.

106 REINFORCED CONCRETE BUILDINGS

sion and a tension, separate from one another, but with hori-
zontal projections of equal intensity or magnitude, provided
the loads are vertical. Whatever the arrangement, the com-
pression and tension must ultimately meet one another and
annihilate one another, whether this takes place gradually by
increments, as in the plate girder of constant depth; or in one
operation, as in the King truss, where the tension chord meets
the compression chord at the ends of the beam; or in a number
of places, all well defined, as in the Howe truss. We have seen
that the rectangular beam is somewhat similar to the King
truss, and that the T-beam is very similar to the Howe truss;
we have also pointed out that the theory of stress-transmission
by gradual increments is not tenable under high loads owing
to the slight tensile resistance of the concrete. We must
assume that the sooner the compression and the tension are
brought to annihilate one another the better will our beams
withstand the loads, hence the necessity of bending the rods
up as soon as possible, and the desirability of closely spaced
U-bars. A simple and effective way of bending the bars is
shown in Figure 113. The point of bending should be deter-

FIGURE 113.

mined by the bending moment, so that there is steel enough to
meet the requirements at all points. In this beam and the fol-
lowing we must suppose that there are some straight bars, but
these are not shown in the figures. Hence the principal stresses
in Figure 113, disregarding the straight bars, are: a constant
compression along the slab, a constant tension in the rod, and
certain vertical resultants. The rod has a curve under the load
A, against which the concrete is pressing. The resultant of
all these pressures should go through the point of application
of A, hence the rod should be bent to a circle with center in the
point of application. The same applies to the reaction, B, and
in addition the rod should be extended beyond the support to
develop the full adhesive resistance.

A somewhat more complicated method is shown in Figure 114

TRANSVERSE STRESSES. U-BARS 107

where there are two systems of bent rods (aside from some
straight ones). The " first " rod, AC, is curved under the load
P for the reasons explained above; in addition, the resultants
C and D must be made to meet one another in the same point
and with the same direction and same force. Hence the num-
ber of rods in each chord should be the same. The length of
rod in compression flange ab should be sufficient to develop the
full strength of the bond, in the same way as for the " second "

\ \

FIGURE 114.

chord over the point of support. The slope or angle of the bent
bars would seem to be of no importance; but many authorities
are of another opinion and recommend an angle of about 45
degrees. (In practice the bars are seldom bent to such large
radii as shown in Figure 114, this diagram being purposely
exaggerated.)

75. The shape of the U-bars should be as shown in Figure
102a, 6, with curved top and bottom, and hooked over. The
downward projection of the end makes it easy to support the
U-bar on the form work, and the entire U-bar is firmly anchored
against sliding, both top and bottom : the top on account of the
curves, the bottom because it passes around the reinforcement.
The direction of the U-bar should be vertical. The sloping or
slanting U-bar is said to strip the concrete away from the ten-
sion rod, as we might expect if our theory is correct, and it does
not give as efficient reinforcement in the small cantilevers as
the vertical U-bar. Round U-bars appear to be better than
flat bars; but there is a great amount of information along this
and similar lines which will have to be furnished before rein-
forced concrete design can be perfected. But our lack of in-
formation in this and similar cases is not different from that
existing in other lines of engineering.

76. When we now finally combine all these elements to one
beam, Figure 115, we have a structure of a very complicated
nature, and we must ask ourselves if all these stresses can travel
through and between one another as here assumed without

108

REINFORCED CONCRETE BUILDINGS

upsetting our calculations and assumptions entirely. To this
we must answer that we do not know, but if we compare our
problem with those met in other lines of engineering we must
admit that there is no fundamental difference between the diffi-
culties. Thus a combination of two simple Pratt trusses is
treated as if the two trusses were really present individually
instead of combined into one structure, and many other instances

FIGURE 115.

could be cited to show that we often have to dissolve a struc-
ture into its apparent elements in order to solve its problems.
Assuming the reinforced concrete beam to be similar to a Howe
truss, as here proposed, seems to be no more of a mistake than to
assume the connections in a riveted truss to be frictionless,
movable joints. But the approximations made in steel con-
struction are so old that they seem almost part and parcel of
the art, while the comparatively new assumptions made for
reinforced concrete have hardly had time to solidify, and they
are therefore supposed to be of a more questionable nature
than the older ones, which have indeed had the profit of the
test of time. Yet there is a number of reinforced concrete
buildings about thirty years old which stand up as well as any-
body could wish, and the modern steel sky-scraper is of no older
date.

CHAPTER VIII

APPLICATIONS OF THE BENDING THEORY

77. Continuity of Reinforced Concrete Beams. The dif-
ference between the beam with simple supports and the
continuous beam is that the continuous beam is subject to a
" reverse" bending moment over the support, while in the simple
beam there is no such reverse moment. The cantilever beam
is an example of the beam in which only reverse moments exist,
and as we have found it feasible to construct reinforced con-
crete cantilevers we cannot deny that continuity may exist in
reinforced concrete beams. In fact, unless special precautions
are taken to eliminate reverse moments over the supports, we
know that continuity must exist and should be taken into ac-
count. The question is then: to what extent are the ends of
a reinforced concrete beam restrained? When this question
is answered we must make the beam strong enough to resist
the bending moment at the column, and then it is a matter for
further investigation to decide in how far the beam is actually
benefited by the restraint to such extent, that the moment at
the middle of the beam may be reduced.

In Figure 116 a beam is shown in which the ends are per-

p Ibs. lin. foot

FIGURE 116.

fectly restrained, and where the uniform load covers the entire
span. The bending moments are over the supports:

MA = MB = & pi 2
at the center: Me = aV pi 2 -

Hence M A + M c = M B + M c = ( T V + A) P? = i P?;

109

110 REINFORCED CONCRETE BUILDINGS

or: the total amount of bending moment to be taken care of in
the beam with " built in " ends is the same as in a simply sup-
ported beam. The bending moment carried by a reinforced
concrete beam is

m a constant X bd 2 (Formula 8);

hence for constant depth the allowable bending moment is
directly proportional to the width 6. At C, Figure 116, the
width is = 6, but at the support where the reverse moment
must be taken care of, the width of beam is only that of the
stem = n. Hence if we assign a moment M c to the middle of
the beam, the end will only carry a moment

M A =

so that M A + M c = \M C + M c = \ pi 2 ',

u o

b 4

we have MC = % i pi 2 = iV pi 2

and M A = MB = I- T V pi 2 =

T, n 1

For 6 = 6

we have M c = f 1 pi 2 = - pi 2

and M A = MB = J - -& pi 2 = A pi 2 -

The moment at the center of the span, in the case of a T-beam,
will therefore be about

and at the end ^ pi 2 .

i 4U

If greater depth is provided near the support the reverse moment
may be increased and the moment at the center of the span may
be decreased a corresponding amount. 1 In Europe it is quite

Attention is called to the obvious fact that no degree of "restraint"
can be allowed at wall ends; this is especially true for beams resting in brick
work.

APPLICATIONS OF THE BENDING THEORY

111

common to make the beams and girders deeper at the columns;
in America the beams and girders are usually of the same depth
throughout. The American practice is to be preferred, because
the continuous effect depends entirely upon the stiffness of the
supports: the slightest yielding of the footings, or even the com-
pressibility of the columns may destroy the continuity entirely,
and too much dependence upon the continuous effect may lead
to serious trouble.

In a slab, the depth and the " width of beam "is the same
at the middle of the span and at the supports. If the supports
are unyielding there may be some excuse for allowing a higher
degree of continuity for slabs than for beams; the more so
because tests on reinforced concrete buildings point distinctly
to such effects. Let us assume a degree of continuity leading
to the following bending moment:

M e = - p I 2 .
In Figure 117 the equivalent system of construction is shown

FlGURE 117.

in which the center portion is considered as a simple beam
resting upon cantilevers of span R. We have then

(L - 2 R)* = - L\

hence R = % L 1 -

(25)

It is hardly necessary to say that we have no absolute certainty
that the slab will adjust itself to conform to this arbitrary divi-
sion of the bending moment. Yet if the cantilever is made
strong enough to carry its load, and the central portion strong
enough to carry its share, it is difficult to see why such a system
should not be perfectly safe. Other assumptions may be made
and carried through in the same manner; this analysis will be
used later for the calculation of the " mushroom " system as
invented by Mr. Turner.

112 REINFORCED CONCRETE BUILDINGS

The formulas usually given for continuous beams depend
upon the factor EL The value of / for a reinforced concrete
beam is not a constant; in Article 79 we shall consider this
in detail. We will find that the moment of inertia depends
upon the maximum unit stresses in the point considered, and
we cannot expect these stresses to be uniform throughout the
length of the beam. The usual application of the formulas
for continuous beams presupposes that the moment of inertia
is constant throughout the length of beam, and we cannot there-
fore apply the formulas used for homogeneous beams to the
reinforced concrete beams with any degree of certainty.

78. While, then, the exact degree of continuity cannot be
determined, continuity does nevertheless exist in many cases
if not in all, and the stresses thus created must be taken care
of. These are, primarily, tensile stresses over the supports,
requiring reinforcement in the top of the girders over the col-
umns, in the beams over the girders, in the slabs over the beams.
The top bars may be loose bars, but it is rather difficult to main-
tain such bars in their proper position; the bent-up bars may
be utilized as top reinforcement with good results, especially
as they extend a distance into the next bay in any case. It is
evident from the remarks made above that the top reinforce-
ment over the support should not be less than 25 per cent, of
the bottom reinforcement; usually more bars are bent up, but
they need not all extend as far beyond the support as the bars
designed to resist the reverse moment. For a uniform load
covering the entire span, the point of inflexion is evidently deter-
mined by the Formula 25:

R = JLfl

v/S

so that 25 per cent, of steel mentioned above should be carried
at least that distance out from the center of the support. The
bars must be embedded in a sufficient amount of concrete to
develop the bond, not less than four diameters from the face of
the concrete, or, if closer to the face of the concrete, they should
be provided with inverted U-bars. The stress on these U-bars
cannot be calculated, it is their presence rather than their
strength which benefits the beam.

79. Moment of Inertia. The moment of inertia in a rein-

APPLICATIONS OF THE BENDING THEORY

113

forced concrete beam is of interest only because certain prob-
lems connected with continuity of the beam, deflection, etc.,
cannot be solved except through a knowledge of its value. The
expression given below is of indirect value only, showing that
the ordinary formulas for continuity do not apply to reinforced
concrete beams, because the moment of inertia is not a con-
stant for the length of the beam, as is usually assumed in the
solution of such problems.

The moment of inertia with reference to the neutral axis
may be found as the sum of two moments: /i, referring to the
concrete above the neutral axis, and J 2 , referring to the steel
below the neutral axis, the concrete below this line being dis-
regarded as usual. We have then (Figure 118)

"a sq. inches

FIGURE 118.

and

Jl=l

7 2 = rad 2 (1 - x) 2

the steel being considered as equal to ra square inches of con-
crete. But according to the formulas given in Articles 25 ff. we
have

i Cxdb = aS

a _xdbC_ Mb .
2 S 2(l-x) r'
hence (after some reduction) :

7 = /i + 7 2 = J fed 8 (1 - J a) x 2 (26)

By means of Formula 5 in Article 27 the expressions derived
above for d and s t , etc., may now be verified. The real import-
ance of Expression 26 is, however, that it shows that the moment

114

REINFORCED CONCRETE BUILDINGS

of inertia depends upon the location of the neutral axis which
again changes with the stresses in the various points of the
beam.

80. Beams with Reinforcement in the Compression Side.
Sometimes it is found impossible to make the compression
flange of the beam wide enough to bring the concrete stress down
to the allowable maximum. In that case some engineers use
compression reinforcement, but as a matter of fact, our knowl-
edge of the properties of such beams is very slight, and there
is grave doubt as to the advisability of using this method of
construction in important cases. The calculations are simple:
to the bending moment sustained by the beam with its ordi-
nary amount of reinforcement is added another bending moment
due to extra reinforcement in top and bottom, this latter cal-
culated as for an ordinary steel beam, but with quite low stresses
(not to exceed 10,000 Ibs./square inch). The compression bars
must be laced carefully to r the tension bars, but under any cir-
cumstances it seems hardly possible to provide properly for the
excessive shear stresses set up in this kind of beams. A steel
I-beam is cheaper and better in places where this kind of con-
struction is actually necessary.

81. Combined Bending and Compression. The section
is best designed by trial. In the case of an arch ring, the sec-
tion is rectangular, and the symmetrical reinforcement is of

small area compared with the concrete sec-
tion. The bars on the compression side
must therefore be disregarded, as it would
require too many hoops to make this rein-
forcement effective in compression. We must
select the depth and the reinforcement by
judgment; the stresses due to the bending
moment alone are then easily found by For-
mula 19 in Article 49. Let Figure 119 rep-
resent the section; let C m and S m denote the
stresses just found due to the moment alone.
If now in Figure 119 gh is made equal to
S m /r and ef is made equal to C m , then the line gf will represent
the distribution of stresses on the section due to the bending
moment alone. The stress due to the pressure P is now P/bd
Ibs./square inch; this is represented by the line ik parallel with

FIGURE 119.

APPLICATIONS OF THE BENDING THEORY 115

fg. The total pull in the steel is then equal to the area of
the triangle khl times the width 6 of the section. For slabs or
arches the width is usually taken as 12 ". The final concrete
stress ei must not exceed the allowable stress; we can therefore
arrive at a preliminary estimate of the dimensions required by
Formula 16, assuming a materially lower " allowable stress "
for the concrete, and a higher stress for the steel, when making
the first trial.

If the section is one in a column the calculations are essen-
tially different. The eccentrically loaded column is of fre-
quent occurrence; in fact, few columns are always loaded
centrally. In practical cases it is almost always impossible to
calculate the eccentricity of the load, and elaborate formulas are
therefore of little or no use. Tension should never occur in the
column; if there is tension with the selected arrangement it is
better to change the lay-out. The percentage of steel will
always be much greater than in the case considered above,
and, as there is no tension, we may perhaps calculate our col-
umn as a homogeneous section, using, however, for the moment
of inertia the expression

/ = I e + r I s (27)

where ] T c ~

i. l s =

I c = mom. of inertia of concrete alone,
mom. of inertia of steel alone.

The cases where the condition of loading can be ascertained
with any degree of certainty are very few indeed, and when
they do occur the bending moment is likely to be very small.
If such is the case it is simpler and probably as correct to cal-
culate the column as a pure column, using a correspondingly
higher factor of safety, and then, if necessary, finally investi-
gate the problem assuming the neutral axis to be disposed at
the center of the section, and take the moment of inertia with
reference to the center line.

82. Chimneys. As an example of approximate methods
of calculating a piece subject to bending and compression, let
us consider a single shell chimney of uniform thickness. The
diameter d (in feet) of the flue is given, and so also the height H
(in feet). Let the outside diameter be D (in feet); the area
presented to the wind pressure (w Ibs. per sq. ft.) is then DH

116

REINFORCED CONCRETE BUILDINGS

square feet, and the total pressure DHw Ibs. Hence the bend-
ing moment at the base (the overturning moment) becomes
DHw times J H = \ wDH 2 Ibs. X feet.

If now the total allowable compressive stress on the concrete
is C Ibs./sq. in. and the compressive stress due to the weight
of a column of concrete 1" square and H feet high is (approxi-
mately) H Ibs./sq. in., then the compressive stress due to
the overturning moment must not exceed C -- H Ibs./sq. in.
Assuming the neutral axis to go through the center of the sec-
tion, which indeed is not true, and disregarding further the bene-
fit derived from the steel in the compressive side (which is on
the safe side), the moment of inertia of the ring is

hence

(C - H) 144 =

(D 4 - d 4 )

D
2

which, when solved, gives the outside diameter

V -

187T (C - H) ' V V187T C - H,
and the tension per inch of circumference becomes

\ (C - 2 H) (D - d) Ibs.

83. Footings. In Figure 120, 2R (inches) denotes the side
of the footing, 2r (inches) the side of the column. The bending

FIGURE 120.

moment on side ab (considering the footing as a cantilever-
slab) corresponds to the loaded area dabc. We have, for a
load p Ibs./sq. in. :

Load dcef = (R r)2Rp; arm of bending moment around
ef - i (R - r).

APPLICATIONS OF THE BENDING THEORY

117

Hence bending moment

A = pR (R - r) 2 ;
load aed plus bcf = (R r) 2 p ; arm of bending moment around

ef = 402-0.

Hence bending moment

B = I p (R - r) 3

The total bending moment due to the area abed is then the
difference between A and B;

and the depth of footing becomes, according to Formula 8, for
a width of beam 2r = b

' *--'-v*/ +1I

-R-rJp

ci VG

to which corresponds a pull s, in the steel, for the distance ab
s t = 77: 2 r

It is, however, quite necessary to provide reinforcement for
the portions ae and bf ; for this reason the amount found above
may be multiplied by a factor estimated at about 2, which
gives :

s = C ^rd (29)

for each layer of steel (Figure 121). The radius of the column

314

FIGURE 121.

should be made as large as possible, because a material saving
in depth of footing is obtained thereby; usually the column
must have an enlarged base for other reasons as well. In Ar-
ticles 14 and 15 we found the cross-sectional area of column:
X = 1400 F for a hooped column
X = 1060 F for a plain reinforced column,

118

REINFORCED CONCRETE BUILDINGS

so that the average pressure, under the conditions assumed, is
1400 and 1060 Ibs./sq. in., respectively. With higher per-
centages of reinforcement these pressures may become mate-
rially higher; the column base is. therefore enlarged so that the
pressure on top of the footing does not exceed the allowable
unit pressure, and a steel plate is put under the bars in order
to distribute the pressure over the requisite area. According
to tests by Bach this allowable pressure may be somewhat
increased, owing to the reinforcing effect of the surrounding
concrete of the footing, but it does not seem wise to exceed say
1000 Ibs. per square inch. The thickness of the plate may be
approximately determined by means of a formula by Grashof:

t = ^Vp (30)

where

t = thickness of plate, in inches,
r = radius of reinforcement (= | of diameter of column, less 2"),

in inches,
p = pressure on plate, in Ibs. per square inch.

The dimension Q in Figure 129 may be found by Formula

FIGURE 122.

28 above, using for p the allowable pressure on the concrete.

84. Circular Reinforcement in Plates. The circular plate
in Figure 122 is supported on a central column. The load is

APPLICATIONS OF THE BENDING THEORY 119

uniformly distributed over its surface, or symmetrically and
continuously disposed along the circular circumference. A
segment, Oab, will then be subject to a certain bending moment,
which moment determines the depth D at the circumference of
the column. It is now easy to show that when the load is uni-
formly distributed over the entire surface, the same formula
applies in regard to depth as was derived above for a square
footing; the calculations are practically the same and need not
be repeated here. When the load is distributed along the
edge, the Expression 32 in the following article may be used.

In any case, we will assume that the depth is known in the
thickest part of the plate (at the edge of the column). If now
the distance dc is one inch long, we have by Formula 9

s t = T V c 2 D,

which expression leads to the amount of steel required along
the radii, the bars being I" apart on the circumference of the
column. Imagine now that all these radial bars be cut asunder
over the top of the column disregarding the tensile strength
of the concrete, each bar will then have a tendency to move
outward, so that if a steel ring surrounded the entire plate,
each bar would exert a pressure s t against the inner face of the
ring. If now dc equals one inch, then db equals one inch times
R/r , hence the pressure on the ring, measured in pounds per
lineal inch of circumference, equals s t X r /R. The tension
on the ring is then

SR = s t ' r ^.R = ^czr D. (31)

It is now evident that the ring with radius R and designed to
resist the tension S R is, mathematically, sufficient reinforce-
ment, so that the radial bars may be dispensed with. In actual
practice this is somewhat modified owing to the fact that con-
crete shrinks when setting, so that it would pull away from the
ring; the ring would therefore exert no pressure against the
concrete until a substantial, and perhaps dangerous, deforma-
tion had taken place. But when the ring is used in combina-
tion with a radial reinforcement and when at the same time the
depth D is not too small compared with the radius, say D larger
than ^ R, then the ring would seem to be a very efficient rein-
forcement. Direct proof of this statement is indeed missing,

120 REINFORCED CONCRETE BUILDINGS

but the " Mushroom " floors furnish at least some indirect
information in this respect, as they probably owe their strength
in a great measure to the intelligent use of circular reinforce-
ment. That this type of reinforcement is successful in other
types of structures may be seen from the remarks made under
" columns," where hoops are extensively used to take care of
stresses somewhat similar to those existing in a plate, although
the plate at the same time acts as a beam. Exact analysis is
of course difficult in these structures which border upon the
class where reinforcement may sometimes be omitted entirely : it
is well known that tapering footings are often constructed with-
out steel, and the same may be true of columns in special cases.

85. Theory of plates. The " Mushroom " System.

A reinforced concrete floor without beams or girders is first
indicated and patented by Mr. C. A. P. Turner of Minneapolis.
As far as known there is no perfectly satisfactory way of finding
the stresses in constructions of this kind, although buildings
actually constructed on this principle have given good satis-
faction, according to the published records. The stresses must
necessarily be of a very complicated nature, especially under
concentrated or unsymmetrical loads; the following analysis
does not pretend to solve the problem in anything approaching
a general way, and the formulas apply only in case the entire
building is loaded with a uniformly distributed load. The
formulas are not inconsistent with the assumptions made for
reinforced concrete construction, and they are therefore pre-
sumably a step in the right direction. It is well known that
most of the proposed formulas are based upon the theoretical
strength of the plates with equal tensile and compressive resist-
ance, and reinforced concrete does not possess any such qualities.

Figure 123 shows the general scheme for a floor of this kind:
the floor slab is simply a flat plate resting upon columns, the tops
of which are enlarged. Let the uniformly distributed load be w
Ibs./sq. foot and the span I feet. The slab is divided into six
strips: two diagonal strips AD and BC, and four strips along the
sides AB, BD, DC, and AC. If we now suppose the panel to be
square, the load on each of the crossing diagonals may be taken
as \w, while the span AD = BC = I V2. Then, by Formula 30

for AB: d = lJ and for BC: d= ^ . t/SL 1 - 1/?
Ci V q a V q a V q

APPLICATIONS OF THE BENDING THEORY

121

so that the depth is uniform, and our problem centers around the
design of a side strip like AB. The notations are shown in Figure
124, where

FIGURE 123.

FIGURE 124.

L inches is the span between column centers.

p the load in Ibs./sq. inch.

R the radius in inches of a certain circular plate.

r the radius in inches of the support under the plate, here
referred to as the " cap."

p the radius of the column in inches.

d the depth of the slab in inches.

D the depth of the cap in inches.

We will now proceed as follows : We consider the floor slab as
supported on the edge of the circular plate with radius R; this
plate will then have a uniform load on its surface and a concen-
trated load along its circumference. Finally the " cap " with
radius r will be designed for a load concentrated on its circum-
ference, disregarding the uniform load on its surface.

The total area of the floor panel between the four column
centers is L 2 square inches; the total weight corresponding to
this area is pL 2 Ibs. The area of the circular plate with radius
R is irR 2 , the total weight on same ptrR 2 . Hence the weight
of the portion outside the circular plate becomes p (L 2 irR 2 ) Ibs.

122

REINFORCED CONCRETE BUILDINGS

In the following computations, L 2 is always large compared with
TrR 2 so that this quantity may be neglected, which is also on the
safe side. The load is therefore pL 2 , and as the circumference
of the circular plate is 2 TrR, the load per lineal inch of circum-

ference becomes

producing a bending moment equal to

27TR

(R r ). If measured per lineal inch of the circumference

of the cap with radius r it becomes, by multiplication with R/r c
(Figure 125),

L 2 -7T R 2

27TK

FIGURE 125.

FIGURE 126.

and the corresponding depth, for 6=1"

R - r ) (32)

for the circular plate. For the slab portion we have, by
Formula 16:

7 LT, T. ' l~

V < 33 )

q c,\ q

'The value of r must now be such that the two depths become
alike, which gives

r n

(34)

at the same time, the value of R is determined by the selected
value of q by formula

APPLICATIONS OF THE BENDING THEORY 123

see Article 67, Formula 25. The depth of the cap. is found by
the formula above, as the load again is pL 2 , substituting only r
for R and p for r ot hence

D--J^-(r.-P). (35)

Ci V 2-n-p

According to Article 84 the reinforcement may be disposed in a
ring with radius r ; the tension in this ring becomes:

S T = ^ c,r D (36)

The arrangement is shown in Figure 126, where the thickness t
should be about 4" so as to cover the ring thoroughly. The cap
should be cast in one piece with the column, but there is no
reason why a joint may not be made between the top of the cap
and the bottom of the flat portion along line a a in Figure 126.
The reinforcement for the flat portion is designed as for any
other slab. We have the depth

- (33)

and the corresponding pull in the steel $/ tons, for a band one
foot wide, is therefore, according to (17)

S f = c z d (37)

This reinforcement should be disposed near the bottom of the
slab at the center of the span, and near the top over the columns.
It will be seen that this leaves a considerable space around the
column without reinforcement near the bottom, which should
be avoided. We may therefore follow the prevailing practice
and bend every alternate bar up, leaving the balance of the
steel straight near the bottom. The reinforcement over the
column is then inadequate, and we will have to introduce addi-
tional steel at that point; if we decide to use rings we may
use one ring with radius R, the strength of which is determined
according to Article 84 by the formula

(38)

We have now (Figure 127)

Thickness of slab, in inches d = y - (33)

Pull in slab steel per foot wic
Radius of upper ring, inches

Tension in upper ring, tons
Radius of cap-ring, inches
Tension in cap-ring

Depth of cap, inches
q is the factor of continuity,

L- r - -

ith, tons Sf = c<id (37)
R = \L (\ - J^\ (25)

S R =^Rc 2 d (38)

r ^ -7? (\

'o r> . , ft (<J*J
TT ( O

S r =-^c.sr D (36)

D = \ V^2^ (r s) (35)
q = 10 to 16.

n -[

H ^

^^^^M^M^M^MM^^^^^.

teusiobsSt. *

FIGURE 127.

All dimensions are in inches; the load p is in Ibs. per square inch
and includes both the given live load and the weight of the con-
struction, itself.

V protection

FIGURE 128.

FIGURE 129.

The plan, Figure 128, shows the disposition of the strips. If
we draw the four circles with radii n, and if n = 1/5 of L, the

APPLICATIONS OF THE BENDING THEORY 125

outlines of the strips will be determined as tangents to the circles,
and all portions of the slab will be covered with reinforcement.

It is evident that this entire treatment cannot lay claim to
great exactness, as the stresses probably are very much more
complicated than here assumed. In any case, the solution here
given is correct under the assumption only that the entire floor
is covered with the same load at all points. The results are
somewhat in accordance with current practice for q = 16, so that
the tests made on actual structures of this kind may to some
extent be taken as circumstantial evidence of the soundness of
the formulas, if not of the argument on which they are based.

It must be noted that these formulas do not apply to wall
panels, because continuity of construction does not obtain at
those points, while also the arrangement is unsymmetrical.
The outside bays should therefore always be carried on girders
and beams in the usual way; it seems, however, that in some
applications of this type of floor the flat construction has been
carried entirely out to the walls.

CHAPTER IX

INITIAL AND ALLOWABLE STRESSES

86. CALCULATION of the " initial stresses "in reinforced concrete
structures is an impossibility with the data on hand at the present
time, hence it becomes impossible to combine these stresses with
those considered in the preceding articles, the " static " stresses.
What we know about the initial stresses is due chiefly to the
careful investigations of Considere; the results may briefly be
described thus:

87. Concrete Setting in Air Shrinks, the more so the richer
the concrete is in cement. If this shrinkage can take place unre-
strainedly no stresses are set up ; in reinforced concrete the steel
will naturally counteract the shrinkage so that the steel is com-
pressed and the concrete put in tension. This rather benefits
the beams, as the tension below the neutral axis is disregarded in
any case, while the compression in the steel decreases the tension
stresses produced under load. On the compression side, the
tensile stresses in concrete due to shrinkage counteract the com-
pression produced by the load, so that on the whole the initial
shrinkage stresses are beneficial in beams. In the case of a
column this is entirely reversed, as the initial compression in the
steel must be added to the compression due to the load, while on
the other hand the compression in the concrete is less than calcu-
lated. In the hooped column the shrinkage of the concrete has
some influence on the stresses in the hoops : a higher intensity
of lateral pressure is required in order to bring tension on the
hoops.

88. Concrete Setting in Water Swells, the more so the richer
the concrete is in cement. In un-reinforced concrete these
stresses, if restrained, are beneficial; in reinforced concrete the
swelling puts the concrete in compression and the steel in tension.
Usually reinforced concrete members set in the air, except
footings, sea-walls, and such .structures; it must therefore be

126

INITIAL AND ALLOWABLE STRESSES 127

possible to keep the concrete in such a moderate state of moisture
that no initial stresses of importance are set up during the
hardening period. Hence the necessity of sprinkling the concrete
freely for the first two or three weeks; this should also be done
liberally in order to furnish the setting concrete with the necessary
water to make the chemical action in the cement take place as
required.

89. When once the Concrete is Hard and Dry, addition of
water makes the concrete swell; these variations in volume are
the more dangerous the less cement the concrete contains, because
the stresses produced are nearly the same if not higher, while the
resistance against tensile stresses is, of course, less in the
leaner concrete. It is important to keep the concrete equally
moist all the time, and water should therefore be put on regularly
and frequently; on concrete three or four weeks old, and dry,
a sudden addition of large quantities of water may prove inju-
rious. If, however, the concrete has been test-loaded so that
larger stresses have existed in the concrete prior to the wetting,
no change in volume seems to take place.

90. It is a question of great importance to settle the exact
intensity of stress in a reinforced concrete member before the
load is put on. Only then will it be possible to design concrete
structures with absolute economy : the better informed we are
in regard to the distribution of stresses in any given case the
smaller can we make the factor of safety. It seems indeed that
the shrinkage stresses are large enough to crack reinforced con-
crete slabs, even sometimes beams; the best remedy is to keep
the concrete moist, so as to avoid excessive stresses, and to
extend the bars well beyond the supports, or otherwise anchoring
the bars, to prevent sliding. While cracks certainly look bad,
this kind of cracks cannot have any great effect upon the initial
stability of the structure, because we assign no tensile resistance
to the concrete. There is no definite or conclusive information
available to the writer giving data on the durability of members
cracked in this way; we are perhaps justified in concluding that
no bad action takes place. The only danger seems to be from
corrosion of the exposed steel, but the efflorescence comes to
our help in this case, frequently filling the cracks completely.

91. Temperature stresses do not usually exist in unrestrained
reinforced concrete members, because the concrete is a fairly

128 REINFORCED CONCRETE BUILDINGS

good conductor of heat, and the coefficient of expansion is nearly
the same for steel and for concrete. In large buildings, tempera-
ture expansion may indeed cause some trouble, because the
concrete, in expanding, throws the columns out of plumb and
causes the walls to turn. Expansion joints are frequently made
in long buildings; in later years, however, expansion joints are
not used as much as they were, except of course in retaining
walls and similar structures where frost and heat have unchecked
sway. In all structures such as arches, continuous bridge
girders, etc., a serious error may be committed by disregarding
the temperature stresses.

92. The expansion joint as used for plain concrete work does
not interest us in this connection. In the reinforced concrete
wall it is indeed questionable whether more is not lost by giving
up the continuity than by having a few fine cracks at intervals.
It must be remembered that usually the expansion joint is a
point of discontinuity in the steel as well as in the concrete, and
a section exposed to accidentally higher loads or stresses than
planned loses the support of adjacent sections which perhaps
are not quite so overloaded. In any case, an expansion joint
must be a clean joint through the entire body of the concrete;
simply marking off of the surface does not constitute a joint.

93. Allowable Stresses. The analysis given in the preceding
articles applies, in so far the mathematics are concerned, to any
composite material having properties consistent with the assump-
tions made. These are, as will be remembered, that (1) the
concrete has no tensile resistance, (2) that sections plane before
loading remain plane after loading, and (3) that the coefficient
of elasticity remains constant up to the point of loading investi-
gated. In addition, a number of assertions have been made, for
instance that a bond exists between steel and concrete, that
concrete has a lateral expansion, that continuity exists in rein-
forced concrete beams, and other similar statements. As a
matter of fact, the concrete has a definite tensile resistance, the
coefficient of elasticity is not a constant as usually determined,
and it is doubtful whether or not the sections remain plane.
That the bond exists cannot be doubted, and the lateral expan-
sion as well as the continuity are well-established facts. The
question here is their numerical value, without which we can-
not design consistently and economically.

INITIAL AND ALLOWABLE STRESSES 129

94. The allowable stresses are used in the design for the
purpose of obtaining an ample " factor of safety." At the
present time, the relation between allowable stress and factor
of safety is in doubt, and it will probably always remain so.
The reason for this is that, even with all the materials stored
before our eyes and open to investigation, the strength of the
concrete cannot be predicted with certainty, much less the
bonding strength. Even the strength of a given cement, mixed
with a given quantity of water in a room of given temperature,
is a variable quantity as reported by different testing laboratories.
In addition, the more recent experiments made on reinforced
concrete specimens show that their ultimate strength is not an
absolute quantity, but depends within wide limits upon the
number of repetitions of the load. It follows that a very large
number of experiments made in the usual way loading the
specimen once only are misleading in their results and cannot
be of much value unless compared with tests in which a large
number of repetitions of the load have taken place.

95. In spite of these apparently unsurmountable difficulties,
reinforced concrete design is at present established on a fairly
firm basis, especially if compared with design involving the
use of other engineering materials. The strength of wood, of
natural stone, and even of steel, is subject to doubt in many cases.
It is only necessary to point to such questions as those connected
with the strength of steel columns with more or less " fixed " ends
to show that not everything is settled beyond doubt, and many
other instances could be cited. But we have to bear in mind
that the allowable stresses assigned to reinforced concrete are
principally established through practice and have but little to
do with the laboratory experiments. The actual proof of the
stability of reinforced concrete construction is furnished by the
many splendid structures erected of this material, and only to
slight degree by the many haphazard experiments made, out of
which most any kind of a theory could be construed. The
difference between concrete as actually used and as tested in
the laboratory is, that in the building all steel rods are securely
anchored in the adjacent span (or ought to be), while in the
laboratory individual beams are tested without any arrangement
to secure the proper sliding resistance. When these results are
analyzed by means of an erroneous shear-theory, it is no wonder

130 REINFORCED CONCRETE BUILDINGS

that the " shear " resistance today, after 25 years of experi-
menting, is as much in dispute as ever. The same applies to
bonding tests : the diameter of the concrete specimen is entirely
disregarded, yet this dimension is at least as important as the
length of embedment.

96. For these reasons, the allowable stresses are not taken as
a certain fraction of the ultimate strengths, or, if they were,
that fraction would not necessarily be the " factor of safety."
The allowable stresses are fixed by practice grown out of the
accumulated experience of many years, as a compromise between
the conservative designer on one side and the economist on the
other side. They have no meaning whatever unless accompanied
by an extensive set of specifications calling for certain materials
prepared in a certain way, and they are therefore largely a local
issue to be determined by the quality of obtainable and prevailing
materials in each locality, coupled with and correlated by the
obtainable engineering supervision prevailing in that same
locality. This supervision is always necessary; not only because
the temptation to " save " may be too great, but much more
because the intelligent and efficient handling of the concrete is,
in reality, the factor which determines the final factor of safety.
It is the duty of the inspector to enforce the specifications in
letter and spirit; it is not less the duty of the engineer to draw
up specifications which can be enforced, and at the same time
compel the use of first-class materials. It must not be understood
that good inspection alone will save the reinforced concrete job
from all dangers: willing cooperation between contractor and
engineer, inclination on the owner's side to pay a fair price for
the work, and, most of all, a full understanding of the why's
and wherefore's are indispensable.

97. With first-class materials, it is customary in the United
States to use the following allowable stresses:

(a) Columns. The allowable stress on the steel is dependent
upon that used for the concrete; we have S rC. It is usual
practice to take r = 15 or 20 for columns; the latter figure pre-
vailing. It is common practice to allow 500 Ibs./sq. inch for
concrete when the columns are centrally loaded; when a small
eccentricity exists which cannot be estimated in figures 400 Ibs./
square inch may be allowed. This is for a concrete mixture
having a mortar base of one part of cement to two parts of sand;

INITIAL AND ALLOWABLE STRESSES 131

according to the size of the column and of the aggregate the
proportion of stone may be from two to three times the volume
of the cement. However, when first-class materials are used for
a first-class job, these stresses are very conservative, and 600
to 700 Ibs./square inch are frequently used when the percentage
of steel is low. It stands to reason that large amounts of steel
should be avoided, the more so the higher the stresses on either
steel or concrete.

(b) Floors. While it is questionable practice to use different
mixtures for different parts of the same building, it is frequently
done because the richer mixture makes it possible to decrease the
size of the columns, while economy calls for as lean a mixture in
the floor as may consistently be used, while at the same time
the leaner mixture is less liable to cracks. Some engineers use
therefore as lean a mixture as 1 : 3 : 6, corresponding to an allowable
stress of 450 or 500 Ibs./sq. inch, while 600 Ibs./sq. inch is used
for a 1:2^:5 mixture and 700 for a 1:2:4 mixture. It is recom-
mended to use the higher stress and the richer mixture, and avoid
the cracking as far as possible by liberal sprinkling; the author
does not consider the usual mixture of 1 : 2 : 4 as being very good for
thin reinforced concrete floors and would prefer 1 : 2 : 3J as allowing
a greater latitude in the manipulation. While the steel stress is
fixed in its relation to the concrete stress in columns, there is no
such relation for the floors. It is customary to use 16,000 Ibs./
square inch for low-tension steel, and 20,000 Ibs./square inch for
high-tension steel. At the same time, proper anchorage is
provided for by extending the steel bars into the next bay, or
by hooking the bars with an open hook at the end; see Article 8.
The higher the stress, the longer should be the embedment, hence
for low-tension deformed bars 24 diameters should be used, and
36 diameters for high-tension deformed bars. Plain bars should,
as said, have an additional hook on the ends equal in length to 6
diameters. It is prevailing custom to make r = 15; for continuous
construction q is usually taken as 10, while for non-continuous
beams 8 is used.

(c) Other Structures. Arches are usually designed by calcula-
tion of the bending moments and thrusts due to the live load,
the weight of the structure itself and the fill on same, and the
changes in temperature existing in the locality where the arch
is to be built. The allowable stresses are then usually taken as

132 REINFORCED CONCRETE BUILDINGS

for columns; the tension in the steel (if any) should not exceed
10,000 or 12,000 Ibs./square inch, but much depends upon the
care and method of analysis. Some engineers consider arbitrarily
the condition where one-half of the arch is loaded over its entire
area; it is by no means certain that this load is the most danger-
ous. Where the analysis is based upon the actual maximum
stresses, the allowable stresses may be somewhat increased,
especially where the foundation is of unyielding nature, as bed
rock, and the abutments are of sufficient area and weight.

Retaining walls may be considered as pieces subject to bend-
ing in case of the modern, ribbed construction. Special attention
must be paid to the proper anchoring of the tension reinforcement,
as walls of this kind usually are built on the cantilever principle.

Stand-pipes and water pipes subject to internal pressure re-
quire low stresses in order to become water-tight under pressure;
10,000 Ibs./square inch in tension on steelshould be the upper limit.
The stability of the finished structure depends upon the integrity
of the joints between the hoops belonging to the same circle
or spiral, hence the thickness of the concrete and the length of
" lap " are most important factors. The concrete should be at least
equal in thickness to 10 diameters of the embedded steel bars
if these are comparatively heavy and spaced comparatively far
apart, say 8 to 12 inches; if light, closely spaced bars are used
the thickness should be increased rather than decreased, and
special pains taken with each joint. The best way is to break
joints whenever possible so that no two adjacent joints come in
the same plane, to keep the bars to be joined a distance apart
equal to two steel diameters, and to surround the joint for its
entire length with a coil. It is not often that all these precau-
tions are taken at once. It has been found, however, that stand-
pipes may fail by the concrete inside the hoops separating from
the concrete outside the hoops, thus destroying whatever bond
may have existed between the steel and the concrete. This
danger is best avoided by having no vertical, concentric planes
of weakness in the concrete.

98. The factor of safety of a reinforced concrete structure
depends therefore upon the selected stresses for the two materials
and the bond stress selected for anchoring the ends of the bars;
also upon the selected value of the % ratio r between the coefficients
of elasticity of the two materials, and finally upon the degree q

INITIAL AND ALLOWABLE STRESSES 133

of continuity. All of these influence the safety of the building,
each in a different manner, and some in a different manner at
different times. It has therefore been found impossible to estab-
lish a definite " factor of safety " for reinforced concrete build-
ings, but we know that the carrying capacity of a building
designed as here described, and erected in a first-class manner,
will easily carry three times the calculated dead and live load
under one or a few test loads. We also know that it will carry a
very great number of repetitions of twice the calculated total load,
but it will probably not carry an unlimited number of repetitions
of three times the calculated total load. It follows that a
material increase in the allowable stresses suggested is dangerous
at the present time.

PART III

PRACTICAL CONSTRUCTION OF REIN-
FORCED CONCRETE BUILDINGS

BY ERNEST L. RANSOME AND ALEXIS SAURBREY

CHAPTER X

MATERIALS OF CONSTRUCTION
REQUIREMENTS AND TESTS

Cement. The strength of concrete depends principally upon
the quantity and quality of cement used. In order to insure a
satisfactory and uniform grade of cement, the shipments are
tested according to rules laid down by the " American Society
of Civil Engineers," and the results must conform to the require-
ments of the standard specifications prepared by the " American
Society for Testing Materials." Copies of these publications
may be obtained from the societies named, but, as the rules are
subject to variation, the specifications are not printed here.

It is an open question to what extent test reports may be
depended upon. The specifications call for certain numerical
values to be obtained, but it is now well established that no two
individuals can obtain the same numerical result from identical
samples of cement submitted to them. The reasons for this are
many; for instance, atmospheric conditions may be different
and many other uncontrollable factors may and do enter. The
greatest trouble is, however, that the manipulation influences
the strength of the test piece, and no two experimenters will
handle the cement mortar in identical manner.

Whatever the reasons the facts remain. On the following
page, Table A gives the results of cement reports made in various
laboratories on identical samples of cement: Case I is a cement
which, while somewhat deficient in some respects, was neverthe-
less of fair quality, yet proved to be extremely quick setting on
the job. The letters A, B, C, etc., refer to the parties making
the test; A, B, and E are testing laboratories in Cleveland,
Ohio, all of good reputation; C is instructor in cement testing
at a large engineering college; D is a concern of national reputa-
tion. The reader is asked to compare the results in each group
and then draw his own conclusions. Table B gives the results

137

138

REINFORCED CONCRETE BUILDINGS

TABLE A, SHOWING VARIOUS TESTS ON CEMENT BY DIFFERENT LABORATORIES

-Lbs. per sq. in.-

Volume Am.
Soc. C. E.

Strength, neat Strength, 1:3 * Accelerated

Tested by Initial. Final. 100m. 200m. 24hrs. 7d. 28 d. 7 d. 28 d. Air. water. Steam. Boil.

< Fineness ^

Standard
CASE I:
A
CASE II:
A

1.0 to
+0.30 10.0

3.30 6.0
30 50

92 75
94.4 82.6
96 6 85 8

175 500 600
165 324 653
371 679 771

175 250
159 307
274 351

OK OK
OK OK

OK
OK"

1 55 30

93 5 82 8

179

1.35 4 30

97.5 86

273 613

181

OK OK"

CASE III:
A

4.45 8.10

94.9 81 7

207 614 749

195 293

OK OK

0.45 1.50

93.0 81.5

168 492 585

134 391

Soft

CASE IV:
A

3 50 5 10

94 6 82

267 602 697

252 333

OK OK

C
CASE V:
A

1.15 3.10
3.15 5.25

96.0 82.0
93 7 81 4

273 641
233 614 427

190
261 370

OK OK

OK OK
OK

1.40 4.35

98.0 80

117 507

121

OK OK

CASE. VI:
-A

4 10 6 25

95 3 82 4

231 602

050

[Soft
1 Warped

C .

2 25 4 55

96 2 82

90 587

150

[ Cracked
OK OK

4.30 7 30

94 6 79 2

359 833

259

OK OK

( Distorted

CASE VII:
A

3.35 6.5

93.0 77.2

343 gel

243

( Soft
OK

1.5 6.20

93.7 76.4

179 601

154

OK OK

CASE VIII:

3 20 60

95 2 81 2

325 707 799

315 436

OK OK

2.15 5.20

92.5 75 5

233 708

267

OK OK

CASE IX:
A
B

3.30 6.25
1 55 40

93.4 84.8
94 9 79

36S 677 817
203 519 686

383 446
212 477

OK OK
OK OK

OK
OK

CASE X:
A

3.30 5 30

94 2 83 8

420 670 687

363 428

OK OK

2.15 4.35

95.1 80 4

226 539 696

220 483

OK OK

CASE XI:
B
B .

2.20 4.45
2 35 4 50

95.0 80.0
94 7 80 3

208 507
202 485 597

218
207 475

OK OK

OK OK
OK OK

2.15 5 10

95 79 7

213 495 603

207 485

OK OK

CASE XII:
Mill Report . .
B

3.0 5.58
2 25 4 55

79
92 7 75 1

290 627
211 494 611

224
221 484

OK
OK OK

2 15 55

93 1 77

476 607

215 484

OK OK

2.10 5.25

92 9 76 5

207 493 612

223 478

OK OK

A
E
CASE XIII:
Mill Report . .
A

2.30 4.50
2.43 5.08

3.3 6.0
2.50 4.30

94.4 81.4
95.0 77.9

79.0
95.2 82 5

365 694 699
421 802 846

284 690
429 748 710

284 345
325 389

255
302 343

OK OK
OK OK

OK
OK OK

OK
OK

B
B

1.55 4.25
2.15 5.0

94.9 75.4
94.0 77.0

482 608
195 482 609

177 416
225 479

OK OK
OK OK

2.30 4 50

94 8 81 8

325 691 696

272 380

OK OK

E ...

2.43 5.5

96.5 77 5

404 786 877

306 417

OK OK

The first line of this table shows the requirements of the American Society for Testing Materials in force
when this was printed, in December, 1909.

MATERIALS OF CONSTRUCTION

139

TABLE B

TABLE SHOWING TIME OF SETTING OF SIXTEEN SAMPLES FROM ONE CAR LOAD

Bag
No.

Accelerated Pat Test

Time
Initial,
bra. min.

of set .
Final,
hrs. min.

Hard sound

3 5

Soft left glass slightly warped

3 5

Hard sound

1 10

3 20

Hard sound

2 30

Hard sound

2 45

Hard sound

3 5

Hard broke glass, slightly warped

1 5

2 50

Hard, sound

1 45

Hard sound

2 50

Hard left glass

3 15

Hard glass broke

Hard glass cracked .

Hard sound

2 55

14
15

Hard, glass cracked
Hard sound

1 50

2 50
2 35

Hard, left glass

2 45

Tables from a paper in the Eng. News, Dec. 9, 1909, by Alexis Saurbrey:
"Comparison of Reports on Tests of the Same Cement by Various Labo-
ratories."

of simultaneous tests on 16 samples taken from the same car,
and show either that the tester was incompetent, which is hardly
to be believed, or else that the shipment varied very considerably
from bag to bag, which of course was emphatically denied by
the manufacturer.

Undoubtedly, there is a vast amount of dissatisfaction with
present methods of cement testing. This is not the place to
discuss whether the tests now in common use are too difficult to
make correctly, or the present staff engaged in cement testing is
deficient in skill, or both. Professor Waterbury is the author of
the following statement:

" (1) It is nearly impossible for two persons to obtain the same numer-
ical results for tests upon a given sample of cement. (2) The results ob-
tained by any one person, who has had some experience in testing cement,
are generally in accordance with other results obtained by the same observer.
(3) There is likely to be a greater variation in the results of the 24-hour neat

140 REINFORCED CONCRETE BUILDINGS

tensile tests than in the result of neat tensile tests for longer periods of time.
(4) With the exception of the 24-hour neat tests, there is likely to be a
greater variation in the results of 1 : 3 mortar tests than in the results of
neat tensile tests."

Mr. W. P. Taylor, one of the leading cement experts in this
country, when addressing the National Association of Cement
Users, said in part:

" Cement testing is a difficult process requiring experience, care, pre-
cision, and knowledge, and hence should only be entrusted to well qualified
men, but too often this important work is relegated to utterly untrained and
careless operators and the results obtained by such methods are really worse
than nothing, as they often are positively misleading. Many tests made
at the present time by supposedly responsible parties are ridiculous in their
inaccuracy, as any one having knowledge of this subject will admit. In-
stances might be cited without number. In one case a cement was rejected
as being quick setting, but an investigation showed that the test had been made
in a hot room in a temperature of over 80 F. and the specimen besides placed
directly over a radiator the cement itself was entirely normal. Strength
tests are often made by inexperienced boys committing every possible error
of manipulation. In one case a cement reported as breaking at 125 Ibs. was
found to give a strength of over 250 Ibs. when accurately tested. Cases of
unjust rejection on the accelerated test for soundness through improper
manipulation and interpretation of the results are by no means uncommon.
Of the sieves used for testing fineness, not one in four has been properly
standardized. These inaccuracies, it must be remembered, are not only
found in the small laboratories, but only too often in those of some reputa-
tion, and the cause may be found to be entirely due to the desire to cheapen
the cost.

It should be recognized at once that if cement tests are made it is worth
while to make them well, even at possibly a somewhat increased expense." l

Granting, however, that satisfactory and reliable cement has
been received, it becomes necessary to so store and use the
cement that it will not deteriorate. For this reason cement
must be stored in a house of substantial design, where water or
even dampness will not penetrate. The temperature must be
kept as low as possible in the summer, as a temperature of from
80 to 100 may seriously interfere with the setting qualities of
the cement, changing normal cement into extremely quick setting
cement. This knowledge should always be imparted to every one
on the job, so that close watch is kept of all batches deposited
in the forms. If the concrete hardens in the wheelbarrows, it
must not be used; it is playing with fate to retemper such con-
Quoted from an editorial in the Engineering News, Dec. 9, 1909.

MATERIALS OF CONSTRUCTION 141

crete with more water, as it hardens very slowly and probably
never reaches the calculated strength. Without doubt, many
accidents may be traced back to neglect of this one point.

On a certain large foundation job this very thing happened.
When the pier forms were removed the concrete was quite wet
and soft, and fell entirely apart. Samples were taken, and owing
to the very plastic condition of the concrete it was possible to
mold test cubes which were allowed to harden. As expected,
the 7 and 28 day specimens had barely cohesion enough to
stand the handling of placing them in the machine. One wall,
16 inches thick and 4 feet high, was allowed to remain in place,
as it carried no loads. When six weeks old it was still so soft
that impressions were readily made with the thumb alone, but
in the course of a few months the concrete seemed to get quite
hard, and it was decided to leave the wall in place. At the
present time, the wall is about five years old and apparently
in a satisfactory condition. Similar cases have come under
observation from time to time, so that the conditions just
described are by no means exceptional.

Dampness is best avoided by careful attention to ventilation
on clear and dry days. While opinions are divided on this
subject, it seems the best, and certainly the safest, practice to
reject cement with lumps or cakes.

Under any circumstances, when cement is received it should
be well seasoned and ready for immediate use, and it should be
used at once. Hence, " warehouse cement " must always be
regarded with suspicion. On large work, the most careful cement
inspection and the most scientific testing may easily be had at
a negligible cost, but on very small work the problem becomes
quite difficult. The parties most likely to suffer are the small
manufacturer of cement blocks and the sidewalk man.

The cement used in reinforced concrete work is always Port-
land Cement; the exceptions are so few that they may be said
not to exist. Each car load received is sampled individually,
a small amount from each bag out of every 30 bags received,
or, in case of delivery in barrels, 1 barrel out of every 10 is
customarily sampled. These samples are mixed to an average
sample representing the car load, taken to the testing laboratory,
and there submitted to the standard tests. In order that the
average sample may fairly represent the car load, the individual

142 REINFORCED CONCRETE BUILDINGS

samples from the several bags or barrels must be taken from
various parts of the car. The tester marks each and every
package with the name of the testing laboratory and the number
of the test, each car load receiving its own " test number."
The aggregate amount of cement taken out of each car load is
about 16 Ibs., one-half of which is sent to the laboratory, and the
other half is stored for reference by the engineer. Eight pounds
are usually sufficient for the purpose of the tests in common use.

On work of any importance it is good practice to make field
tests of the cement (setting time and soundness principally) to
guard against changes, and also compression tests on cubes
made from concrete taken from the mixer or the wheelbarrows.
These tests simply supplement the laboratory tests and cannot
replace them; however, the compression test on the concrete as
actually used is in itself a very excellent check on the efficiency
of the mixture used, and gives also important information as
to the proper time for removal of the forms. As a matter of
fact, the engineer is not greatly interested in the strength of
cement as tested in the laboratory under conditions which never
obtain in the field, and he is probably relying upon the compli-
cated and difficult laboratory test because nothing better is
available. The compression test in the field might well be used
more extensively, although, as far as rejection or acceptance of
a given cement shipment is concerned, it is of no importance
whatever.

The consulting engineer is sometimes called upon to examine
an existing structure, and inasmuch as such examination
usually has its cause in troubles of various kinds, he might be
interested in knowing what the original proportions of the
concrete were. Unfortunately, it seems that there is no very
satisfactory method whereby such information can be obtained,
and if obtainable, testimony along these lines would probably
have little weight in court unless a large number of samples were
analyzed. 1 This is also true of compression tests made on

1 See also paper in Eng. News, 1908, p. 46, by Prof. R. L. Walls, who made
a successful analysis of this kind.

One case of this kind happened in Oakland, Cal., where skimping of
the cement was proved to the satisfaction of the jury by careful chemical
analysis, based upon the amount of lime in the analyzed concrete. Many
years afterward, I came across evidence of other nature that showed the
chemical analysis to be correct. ERNEST L. RANSOME.

MATERIALS OF CONSTRUCTION 143

cubes or cylinders cut from the concrete, because it would be
difficult, but not impossible, to show that the pieces were not
injured in the process of cutting.

Perhaps it is well here to call attention to the fact that the
size and shape of the test specimen has some influence upon its
strength, so that results obtained from 4" and 6" cubes cannot
be directly compared without reference to the laws governing
such cases.

In tests on concrete and mortar, the relative size of the aggre-
gate and the test specimen might also have some influence.

Sand. Next to the cement, the sand is the important factor
in determining the strength of the concrete. Various elab-
orate theories exist whereby the proper composition of the sand
may be determined; where a large concrete job is to be sup-
plied from a uniform, local supply of large capacity such inves-
tigations may, of course, be of great value, as it is possible to
determine just what should be added or deducted in order to
get the best results. But ordinarily such investigations are of
little value, as the character of the sand may vary from day to
day, if indeed it does not vary from shovelful to shovelful.
Hence a quick and cheap test is required for daily or occasional
use, and such a test we have. It consists in simply comparing
the strength of briquettes made from " standard " sand and
cement with that of similar briquettes made at the same time
from the proposed sand and the same cement. Obviously, this
method of testing is free from nearly all the objections made
against the usual cement test, as only comparison is wanted and
not absolute figures. The standard sand is not particularly
strong sand if used for building purposes, so that for good
results the proposed sand should give a tensile strength 25 to 50
per cent, in excess of that obtained with standard sand; when
the strength is about equal, the sand may be termed " pass-
able " if only low stresses are used in the design, while sand
well below standard may be rejected without error. Speci-
fications drawn along this line avoid entirely all questionable
and unfair regulations. Where water-tight work is required,
the more elaborate granularmetric analysis may be used if
the supply is uniform in character. Frequently, an addition
of a small amount of fine sand, or preferably stone dust, greatly
improves the strength of the concrete.

144 REINFORCED CONCRETE BUILDINGS

It is well understood by skilled concrete men that the best
grade of sand is clean, sharp, and well graded from fine to
coarse, and these words are therefore usually inserted in the
specifications. It is believed that only in exceptional cases
such specifications are enforced, and the policy of writing speci-
fications which nobody can or will enforce is not to be recom-
mended for obvious reasons. Up to three per cent, of impurities
are not usually injurious, but sometimes even a much smaller
quantity of clay is detrimental, especially if the several grains
are covered with a thin film of clay. The test recommended
above settles such questions at once, provided that the mix-
ture made in the laboratory is not such that the film is removed;
too much mixing is as bad as not enough. On the job, however,
the mixing must of course be greatly prolonged if the grains
are coated so as to wash the film off the sand.

The specifications must state the maximum size of grain
allowed; usually the sand is required to pass a screen with four
meshes per lineal inch.

Stone. The stone should be clean and hard, two require-
ments easily complied with. Loose dust should not be allowed
when the concrete mixture is specified in proportions as 1:2:4
or similar ratios, because the dust acts as so much sand and
decreases the strength of the mortar base. Some dust always
clings to the stone, hence the word " loose " should be used.
On the other side, if it should be found desirable to use " run
of crusher " with some additional sand there is no reason why
good results cannot be obtained in that way with continuous
and intelligent supervision. As a general thing, the engineer will
save himself a large amount of trouble by insisting upon separate
stock piles for sand and stone, and specify his materials by
definite proportions. There can then be no room for dispute.

If the specification suggested above is used, the stone should
be required to pass a ring f" in diameter for thin reinforced
concrete pieces as used for floors, beams, and columns; for heavy
work, the size may go up to 2" ring or larger. The stone should
be retained on the \" mesh screen, perhaps with a small allow-
ance, so that, for instance, 5 or 10 per cent, may pass through
the screen, the balance to be retained. Certain kinds of rock
give oblong stones, and a maximum length should be specified,
for instance 1".

MATERIALS OF CONSTRUCTION 145

Attention is called to the ever-increasing use of furnace
slag, a by-product from the manufacture of pig-iron. Slag makes
excellent concrete if used with the proper proportions of mor-
tar, for instance, 1:2:3 or 1 : 2 : 3J. The slag is very dry and
absorbs water in large quantities; the stock pile should there-
fore be kept soaking wet at all times. Otherwise the concrete
may not set up well.

Boiler cinders should not be allowed in reinforced concrete
work, as little as soft limestone or soft sandstone, or any kind
of stone disintegrating under the influence of the atmosphere.
Fair concrete may be made from soft or friable aggregate by
limiting the time of mixing to a minimum; good limestone makes
excellent and very hard concrete, and crushed brick if not very
soft makes a very fair concrete for many purposes. Brick
dust is a good substitute for natural sand.

Certain kinds of shale have great toughness when in the
natural deposit, but fall to a powder when exposed to the influ-
ence of the air. Conglomerate cemented together from a large
number of small pieces must be prohibited, even if apparently
hard. The authors recall one or two instances where this kind
of stone led to very serious trouble. Certain kinds of slag con-
tain very injurious chemicals, such as sulphate of lime, etc.
Chemical analysis should be insisted upon before the use of an
unknown slag.

Steel. The selection of the steel is rather embarrassing,
each " system " claiming special advantages of its own. In
most cases these advantages exist largely on paper only, the
fact being that almost any kind of steel may be used with suc-
cess. The form of the particular bar proposed is a compara-
tively small matter; the real importance is in the material from
which the bar is manufactured, and the method of manufac-
ture. In their own work, the authors prefer the cold twisted
square bar. The specifications should call for minimum
ultimate strength, minimum limit of elasticity, minimum per-
centage of elongation, and a bending test. The engineer is
interested in the kind of steel furnished, not in the method of
manufacture.

Plain Bars. Round, square, or flat bars are used, but the
round bar should be favored as easier to handle, and flat bars
are generally considered as giving smaller bonding strength hi

146 REINFORCED CONCRETE BUILDINGS

the concrete. High tension or low tension bars may be used if
a proper length of anchoring is had in each case. The high
tension bars are frequently rerolled from old railroad rails
in itself a very good idea. But this steel is rather high in
carbon and requires extra care in manufacture; rerolling at
too high or too low temperatures may be the cause of brittle-
ness and other trouble. Many engineers decline to use rerolled
or hot twisted bars for this reason, and it cannot be denied that
unless properly tested, such steel may not be what is expected
and required. Figure 130 shows a bad piece of rerolled steel.

FIGURE 130. UPPER BAR: REROLLED STEEL, IMPROPERLY MANU-
FACTURED. LOWER BAR: GOOD SPECIMEN.

Both bars from the collapse of the Henke Building (Column Rods).
Photo by Alexis Saurbrey, who examined ruins for the owner.

Deformed Bars. The great class of bars distinguished by
projections and recesses on the surfaces have this in common,
that if sufficient concrete surrounds the bar, it is harder to pull
out than the plain bar of same cross-section. It is almost
impossible to discriminate between all these bars which belong
to types merging one in the other by gradual steps.

Patent royalties are collected on most of these bars; the
twisted bars, both hot and cold twisted, are exceptions. The
value of the steel as reinforcement is not greatly affected by

MATERIALS OF CONSTRUCTION 147

the various forms of projections in use; bars with deep grooves
should not be used, as water instead of concrete is likely to
collect in the pockets, especially on the under side of the bar.
If deformed bars are used it should be of a type having the same
cross-sectional area at all points of the length.

Wire Mesh. Reinforcement with ready-made wire mesh
is adapted only for short span slabs, and even then additional
bars of larger diameter are often used. The cost is high, and
attempts are therefore frequently made of talking the buyer
into allowing much higher stresses on wire fabrics than on plain
steel. It is of course possible that our present methods of cal-
culating stresses in slabs are in error, but the proof has yet to
be furnished. In the meantime we cannot consistently allow
stresses on drawn wire as high as 30,000 or 40,000 Ibs. /square
inch, even if this material has a tensile strength of 100,000 to
120,000 Ibs. /square inch.

Requirements and Tests for Steel. The ultimate strength
of bars used for reinforcing purposes should be at leas't four
times the allowable stress, and the elastic limit should be at
least twice the allowable stress. Of these two, the latter is
by far the most important; a high elastic limit increases the
factor of safety of the entire structure, although not in direct
ratio. Both of these figures are easily determined by a ten-
sile test, except in the case of bars without a definite elastic
limit, such as cold twisted bars. In this case, the strength and
elastic limit may be determined either before or after twisting;
generally speaking, the ultimate strength is raised about 33
per cent, by twisting soft stock, while the elastic limit, if it
can be at all determined, will be found to be about 75 per cent,
of the ultimate strength. This applies to cold twisted bars
only; hot twisting does not change the strength or elastic
properties of the bar.

The bending test is very important, as practically all the
bars are bent on the job; all bending should of course be done
cold. For high-carbon steel it is usually specified that the bar
must bend cold around a pin four times the diameter of the
bar without showing signs of distress. Good cold twisted bars
will easily bend around a pin twice the diameter of the bar.
Three or four diameters are however more commonly spe-
cified. Soft stock should fold flat upon itself without showing

148 REINFORCED CONCRETE BUILDINGS

signs of cracking or checking. It follows that the kind of
steel to specify may depend largely upon the bending problems
encountered.

The elongation serves practically the same purpose as the
bending test. The minimum amount specified varies from
10 per cent, for high-tension steel to 20 per cent, or 25 per cent,
for soft steel. These figures must not lead us to believe that
the smaller per cent, of elongation at fracture is a point in
favor of the high-tension steel; in fact, the coefficient of elasticity
is practically the same for all kinds of steel in common use, and
the elongation under a working load depends upon the coefficient
of elasticity.

The following quotation from the Cleveland Building Code
may be of interest:

"Steel reinforcement shall be divided into two classes, Medium and
High Tension. Medium steel shall have an ultimate strength of 60,000 to
70,000 Ibs. /square inch, and shall conform to the Manufacturers' Standard
Specifications as revised Feb. 6, 1903. High-tension steel shall have an ulti-
mate strength of not less than 80,000 Ibs. /square inch, and an elastic limit
of not less than 45,000 Ibs. /square inch. The elongation shall be at least
ten per cent, in eight inches. Bars shall bend cold around a pin of diameter
equal to 4 times the least dimension of the bar without showing signs of
cracking."

Tiles. The tiles are usually made 12"X12" in plan; the
width usually cannot be changed, but tiles 12 " X 6" or 12 " X 18 "
may sometimes be obtained. The thickness of walls and webs
is frequently about \" , subject to variation. Dense or semi-
porous tiles can be obtained; it makes little difference in
the results which is selected. The surface should be deeply
scored so that the plaster may be firmly bound to the tile;
only in the roughest kind of work are the tiles left uncovered
on the exposed bottom side. In burning, the tiles shrink; a
well-burned tile is often as much as f " smaller each way than
specified. This should be made up in concrete, so that the
plans should show the full thickness of floor, not the thickness
of concrete topping. At the same time, tiles may be too large,
in which case the minimum amount of concrete to be placed
over the top of the tiles should be specified. Broken, badly
warped, or otherwise defective tiles should not be allowed.
Before the concrete is run, the tiles should be made soaking wet,
as they will otherwise absorb the moisture from the concrete.

MATERIALS OF CONSTRUCTION 149

Concrete. Hand-mixing is used only in exceptional cases;
the engineer should reserve the right to permit hand-mixing if
practically unavoidable. Machine-mixing on continuous mixers
is not desirable; the mixer should preferably be a revolving
batch-mixer of approved design. The aggregate and cement
is measured by volume; it is very convenient to take the cubic
foot as unit and consider the bag as containing one cubic foot
of loose cement. When the job is started the wheelbarrows
or receiving bins are checked by means of the standard unit,
and the required depth of filling marked. All wheelbarrows
and bins must be brought to a level when filled; a small top on
the wheelbarrow load looks like a small matter, but may in
fact mean a material decrease in the proportional amount of
cement. Wheelbarrows containing the required amount when
level full can easily be obtained.

Sufficient water should be added so as to make the mixed
concrete into a flowing paste that will pour from the wheel-
barrow. The concrete is mixed with an excess of water if pools
are immediately formed on top of the concrete when deposited
in the forms: the pools increase in size, the water finds an out-
let to a lower point, and the cement is washed away from the
mortar. Inclined " sand-streaks " on the sides of girders or
beams are usually due to this cause. Years ago " dry " concrete
was specified, but the manipulation becomes so difficult with
dry concrete that " wet " concrete soon became universally
used, and at present there is a tendency to exaggerate the
amount of water. Where the concreting of large girders pro-
ceeds from one end, and the working face of the concrete body
is sloping, the excess of water naturally seeks the lower level
in the part of the girder box not yet concreted. The water
carries " laitance " with it, and this sets without hardening,
forming a white plaster-like film on the bottom of the girder,
often |" thick or more. Such conditions should of course be
avoided.

The mixing must continue until all parts are thoroughly
incorporated in the mixture and all stones covered with mortar:
the concrete will then be of uniform consistency and color, and
if sufficient water only is used there will be no precipitation in
the bin or in the wheelbarrows of the heavier particles. This
separation is much more likely to take place if perfectly clean

150 REINFORCED CONCRETE BUILDINGS

sand is used, especially if the grains are round, and lake or
river sand does therefore require less water and more care than
bank sand containing a slight amount of clay. As the aggre-
gate contains varying amounts of moisture in the different parts
of the stock-pile, and as the stock pile seldom is perfectly uni-
form throughout, no hard and fast rule can be laid down for
the requisite amount of water. The amount of water used,
and the number of turns given the mixer before dumping should
be left to a competent man in charge of the mixer. The mixer-
man is an important person on the job.

The concrete must be deposited in the forms before it stif-
fens perceptibly; concrete held longer than that time in either
the receiving hopper or the wheelbarrows should be wasted.
The concrete must be well spaded and churned in the forms
to insure dense concrete and fine surfaces. The entire con-
crete work should be under the personal and direct supervision
of an intelligent and careful man; this man should be present
at all times when concreting is going on, and should have no
other duties during concreting. If the engineer or owner de-
sires to have an inspector of his own on the job at the same
time, so much better, but it should be made clear to the con-
tractor and his men that such inspection is for the purpose of
enforcing good work only, not for the purpose of waiving the
specifications. The concrete must be conveyed to and depos-
ited in the forms in such a manner that the steel reinforcement
is not disturbed and so that older concrete is not injured. The
mechanical plant must not be braced to the form work or the
building, or have any solid connection therewith. All these
rules are dictated by common sense, but nevertheless daily vio-
lated, and it is therefore well to put them in the specifications.

Joints in the concrete work should not be allowed except
at the natural end of each day's run; it is therefore essential
that the plant be in such shape that breakdowns are avoided.
The necessary joints must be well-defined, straight lines, prefer-
ably through the center of the span of all slabs and beams.
The joint should be made perpendicular, not sloping; the ver-
tical joint is easily repaired in case of trouble. New concrete
should be joined to old concrete only after the surface of the
old concrete has been removed by mechanical or chemical
means, and the rough surface made in this way thoroughly

MATERIALS OF CONSTRUCTION 151

cleaned with scrubbing-brushes and water. Neat cement
paste is then rubbed into the clean surface, and concreting
proceeded with at once. The care taken is of no avail if the
cement paste is allowed to dry out or set before the new con-
crete is put on. The slab and the beam supporting it are usu-
ally run in one continuous operation, without any joint between
them; but the design can easily be so arranged that a joint can
be made if desired.

The concrete gang, and especially its foreman, should be
made to understand that it is their duty to make concrete which
will not require after treatment to make up for their careless-
ness or haste. No pointing-up should be allowed before the
engineer has seen and approved the concrete; but when so
directed the contractor must at once proceed with the pointing.

In monolithic construction, the columns should be run a
sufficient length of time ahead of the floor, to allow the con-
crete in the columns to settle and shrink; the interval may
conveniently be utilized in putting the floor-steel in place.

The setting of the concrete is greatly influenced by atmos-
pheric conditions. Hot weather accelerates the action, and
cold weather retards it. Otherwise, neither heat nor cold need
have any injurious action on the concrete if proper precautions
are taken. In hot weather, it may be necessary to cover the
green concrete against the direct rays of the sun, and in any
case the concrete should be sprinkled liberally to make up for
the loss of water by evaporation, as concrete cannot gain its
full strength without water. Much more serious is the action
of frost, and especially of repeated freezing and thawing; the
precautions to be taken in the summer are simple and cheap
compared with those required in winter, where the weather
may suddenly change from mild to bitter cold. The concrete
is made with heated materials and heated water; the green
concrete is covered with a tent or boards with straw on top,
but not manure, which is said to injure the concrete; in fact,
one or perhaps two accidents have been ascribed to the use of
manure. Moist heat is supplied to the space below the green
concrete, and between the concrete and the covering.

CHAPTER XI

FLOOR SYSTEMS

BROADLY speaking, reinforced concrete may be used in one
of two ways : cast in place, or cast in the yard and erected after-
wards when hard. In the first case, the construction becomes
more or less continuous by virtue of the method of erection,
and the continuity is then usually emphasized in the design,
so that all parts are thoroughly tied and united together. This
type of construction is therefore referred to as " Monolithic."
In the second case, the structure is divided into separate pieces
or " units," and both the designer and the erector must then so
articulate the building that the weights and dimensions of the
several pieces come within reasonable limits. In distinction
from the monolithic work, this type of construction is referred
to as " Unit." A large number of " systems " exist within
either of these broad divisions, several of which claim protec-
tion under United States patents, and all of which claim either
superior strength or greater economy. It would, however, be
outside the scope of this book to enter into a discussion of
these points; moreover, there is no accepted standard of com-
parison, so that, in all probability, the contractor's bid or
proposal on any given building gives the only safe way of
determining the relative cost in each case.

Some types of construction are essentially monolithic, such
as the flat-plate and column construction, and the tile con-
crete construction. Others are essentially of the unit type,
such as the Visintini System, where each unit is a small truss
in itself. The common " ribbed floor," with beams and gir-
ders supporting the floor plate proper, is usually built in a
strictly monolithic way, but recently considerable efforts have
been made toward perfecting the " unit " method for floors of
this kind. Two distinct methods have been followed in the
unit construction: (1) Each beam or girder is a complete carry-

152

FLOOR SYSTEMS 153

ing member in itself, and in that case the slab portion is usually
molded on the ground, and set in place when hard. (2) The
beams and girders are not complete carrying members, the
floor slab proper forming the upper flange of the T-beam, and in
this case the beams, girders, and columns are cast on the ground
and set in place when hard, while the slab proper is cast in place
over the top of the beams, and serves the dual purpose of tying
the building together, and of forming the compression flange
of the beams and girders. 1

Design. After the mathematical design has been perfected,
additional steel must be introduced to take care of shrinkage
stresses, especially in the slabs. These bars are disposed cross-
wise over the tension rods, and also serve as "distributing rods 11 ;
the structures erected under the so-called "Monier System"
always had large quantities of such bars which greatly strengthen
the building as a whole. The exterior belt courses should have
ample additional reinforcement, and these bars should run con-
tinuously around the entire building at each floor, with sufficient
lap at each joint. In the " tile-concrete " construction, the
value of such additional steel is too frequently overlooked,
although it is here of particular importance owing to the absence
of secondary beams. It seems to be an open question to what
extent the slab can be considered as active in compression, and
the laws governing the influence of the width of the top flange
are practically unknown except for a few sporadic tests. It can,
however, be stated that the active width of slab depends upon
the thickness of the slab, and upon the intensity of compres-
sion stress, and the stiffness of the system as a totality probably
enters to some degree. Common rules are: One-third or one-
sixth of the span; two-thirds of the distance between beams;
six or ten times the width of the stem, etc. None of these rules
is derived from either test or convincing analysis.

In monolithic construction, the questions of bearing for
beam or girder ends do not usually arise, and when they
do, the strength of the supporting material is usually the govern-
ing factor. In many brick buildings the pressure on the wall
bearing has been limited to 200 or 250 Ibs. per square inch.

The connections between the several elementary parts are

1 This type of construction is covered by my U. S. Patents of March 4,
1902, No. 694,577, and April 20, 1909, No. 918,699. ERNEST L. RANSOME.

154 REINFORCED CONCRETE BUILDINGS

easily taken care of; however, it is a common error to have the
re-entrant angles square instead of chamfered, and this is of
particular importance where the slabs rest on the beams or
girders. The removal of the forms is greatly facilitated by
having chamfered or beveled corners, and the finished struc-
ture is less likely to crack.

A special problem arises in connection with the exterior
construction. In modern practice, the columns and floors are
usually erected first, and separate curtain walls are next placed
between the columns. These curtain walls may very well be
utilized as deep beams at the same time by uniting the lintel
below to the curtain wall in a substantial manner. 1

The expansion and contraction of these walls is readily
taken care of by setting their ends into recesses in the columns
and the windows may be set into similar recesses above the cur-
tain walls.

The cornice is tied to the roof structure by means of iron
stubs projecting from the concrete.

A good design should show not only the location of all pro-
jecting stubs, ledges, recesses, etc., but also all the minor open-
ings for heat and sewer pipes, and the location of pipes for gas
and electricity. It is a surprising fact that those details are
so much neglected; it is certainly much cheaper and better in
every way to set proper sleeves, etc., for all such openings. Some-
times, the pipe-risers come up through the columns, but this
practice is hardly to be recommended, as it is both unsanitary
and makes repairs and alterations difficult or impossible. Spe-
cial pipe shafts may be arranged for; or in some cases, the
exterior columns are made large enough to accommodate the
piping in cored flues of ample size. Heat flues or ventilation
may be arranged for by having the exterior columns hollow
with register openings leading to the several stories.

While the floors are usually covered with cement finish,
wooden floors are in favor in many places. The floor is nailed
to sleepers laid on top of the rough concrete base, and cinder
concrete or stone concrete poor in cement is run between the
sleepers. In many cases the sleepers apparently give good
satisfaction, in others they are soon destroyed by dry rot. The

1 This construction forms the subject-matter of my U. S. Patent, No.
694,580, March 4, 1902. ERNEST L. RANSOME.

FLOOR SYSTEMS 155

keeping qualities cf wood embedded in concrete are not well
known, but where the concrete is in direct contact with the
wood, it must certainly act as a preservative if it has any action
at all.

In hotels and similar establishments, linoleum or carpeting
over a fairly smooth cement base should be very satisfactory.

As a general principle we must maintain that pipes should
not be put inside the structural concrete, as they are practically
inaccessible; the electric conduits may perhaps form an excep-
tion to this rule. After the steel has been placed the electri-
cian places his outlet boxes and connects them up, so that the
conduits rest immediately upon the steel. The slab must then
be so thick that the entire pipe is buried below the neutral axis
of the slab, as otherwise the strength of the slab is jeopardized.
Note, however, that when the lights are suspended from the
bottom of a beam, the outlet box and perhaps a short riser
must be placed before the main tension steel is put in. Sewer
or gas pipes should always be left exposed; sleeves are placed
where they go through the floor, so that no cutting is required.
Attention to all such detail goes a long way toward success in
reinforced concrete work; if the plumber is turned loose in a
building to cut whatever holes he may see fit he is almost cer-
tain to go through one of the main girders, steel and all. In
fact, such a case came under the author's observation once. 1

Where a wood-floor finish is placed over the concrete, many
of the pipes may of course be concealed in the space occupied
by the sleepers. Only risers and outlet boxes are then placed
before the concrete is run. The specifications should state in
detail who will furnish and set the various sleeves required,
as there will otherwise be considerable friction between the
several contractors.

A number of devices are on the market by means of which
shafting may be attached at any place in the building. All
such devices must be decided upon in advance and placed be-
fore the concrete is run. If a plain factory ceiling is all that is
wanted, it is convenient to place suitable bolts at intervals,
with their threaded ends projecting from the concrete; timbers

1 In the Academy of Science Bldg., San Francisco, I once caught a plumber
in the act of cutting off the brick corbeling on which the floor rested. Many
such cases have come to my attention from time to time. E. L. RANSOME.

156 REINFORCED CONCRETE BUILDINGS

are then bolted to the ceiling wherever wanted, and the shaft-
hangers attached to the timbers. All threads must be pro-
tected against concrete and rust. Sometimes the operation is
reversed and tapped sleeves provided in the concrete, into which
the necessary bolts are screwed. The head of the bolt project-
ing into the concrete should be enlarged so that the bolt will
not tear out; the pressure may be distributed over a larger
area by means of bars or plates underneath the head of the
bolt.

Monolithic Construction. The monolithic building is usu-
ally erected a story at a time. First: the forms are set up,
forming a complete wooden shell for the concrete to be depos-
ited; next, the steel is put in place, and the concrete run around
the steel and within the forms. Simple as this series of opera-
tions may seem, there are, nevertheless, a great many details
to be attended to. This is particularly true with reference to
the form work, which in itself absorbs a large proportion of the
total cost of the building.

Forms. The simplest, but in the long run the most expen-
sive method, is to cut the boards as needed and put them to-
gether box-fashion, nailing all the joints securely. Such forms
cannot be removed without breaking the lumber to pieces and
destroying a great deal of the concrete, particularly the corners,
and at the present time no experienced worker in reinforced
concrete would consider using such rough methods.

The first improvement consisted in making the slab-panel
forms each in one piece, resting upon the form-panels for the
beam-sides. All these panels had cleats nailed to the side
facing away from the concrete, so that each panel remained a
unit in itself throughout the erection of the building, and each
panel could then be used over and over again. For long flat
slabs, the panels rested upon joists, and sometimes it would
even be necessary to shore the joists midway between the
beams. For the shorter spans, up to five or six feet, the cleats
used under the panels to hold them together would usually be
sufficient. Figure 131 shows schematically the most essential
parts of this arrangement, of which there is a very large num-
ber of variations. Usually, however, the parts are so arranged
that the beam-bottom with the shores under same can be left
in place while the panels are being removed, for the purpose

FLOOR SYSTEMS

157

of keeping the beams supported for a longer period than the
slab.

Even when nicely adjusted, a falsework of this kind is
soon destroyed by the continuous prying and pounding required
to get it loose from the concrete, and the jarring and knocking

FIGURE 131.

about while shifting from floor to floor. An improved method
of centering was therefore devised, whereby some of these
objections would be overcome. This centering is shown in
Figure 132, where the long box is split centrally down the middle,
each half being held together by the triangular cleats, while
the two halves are hinged together. The beam-bottoms rest

FIGURE 132.

upon cleats along the lower edge of the box, and these cleats
also strengthen the bottom of the boxes where they rest upon
the supporting horses. At each end, the boxes are closed by
means of removable heads. When the forms are to be removed,
the start is made with the horses; next the removable heads
are taken off, and finally, the boxes are collapsed and removed.
Of course, the sketch shows the essential outlines only; such
portions as the stays for holding the boxes expanded, etc., have
been omitted.

158 REINFORCED CONCRETE BUILDINGS

The main advantage of this arrangement rests in the fact
that the benches or horses used for supporting the boxes form
at once a safe and even foundation for the form work, and that
the forms themselves are taken down, again put together, and
erected by ordinary labor, there being no cutting or adjustment
of any kind. Their use presupposes a standardized layout,
and this, by the way, is a point to which altogether too scant
attention has been given in the past. It is believed that, if a
number of typical or standardized buildings are to be erected,
the simple attention to duplication of parts may reduce the
cost from 10 to 15 per cent., even if the buildings are at
widely distant points. On the rougher and simpler -forms
of falsework it is frequently estimated that 60 per cent,
of all lumber purchased is used on the job for which it was
bought.

The great secret of success lies in attention to one funda-
mental point: that all parts must come easily apart when the
forms are stripped from the concrete. Hence all shores must
rest on wedges, and all joists, etc., must be keyed in place with
wedges; wherever possible, bolts must engage in slotted holes
from which they can be removed by simply loosening the nut
without taking it off. Thus, for many purposes, the arrange-
ment shown in Figure 133a is greatly superior to that shown in

FIGURE 133a. FIGURE 1336.

FIGURE 134.

Figure 133&, because a slight motion sidewise releases the bolt
in A, while the bolt in B must be drawn through the hole.

The posts or shores should have one side of the bottom cut
away at an angle, as shown in Figure 134, to facilitate removal.

Similar lines of argument lead to the result that all forms
should be made with sufficient " slip " to leave the concrete
readily, and that the re-entrant corners should be beveled;
in short, what is good practice in making patterns for cast iron
is also good practice in making molds for reinforced concrete.

FLOOR SYSTEMS 159

Amongst the more common errors in the construction of
forms, attention is called to the following:

Insufficient stiffness, so that forms sag or bulge.

Untight forms, so that cement is lost by leakage.

Irregular thickness or width of boards, so that bad-looking
board-marks result.

Too tight fitting, necessitating crowbars and sledge-hammers
when forms are removed.

Much time must be devoted in the office to the preparation
of details of the forms and making up of lumber schedules; in
fact, it would pay in many cases to have the forms made in a
well-equipped carpenter shop and haul the forms to the job.
Much time should also be devoted on the job to the inspec-
tion of the forms, both during erection, concreting, and removal,
to insure against costly errors. But most of all, cleanliness must
be enforced at all cost, so that no shavings or ends of boards
find their way into the concrete.

In order to preserve the forms, the woodwork is frequently
covered with crude oil, soap, or similar materials, and the re-
sults undoubtedly justify the expense. However, if the ceil-
ings are to be plastered, no oil must be put on, as it prevents
the adhesion of the plaster. In that case, the forms are simply
given a good soaking with soapy water some little time before
the concrete is run, and it must be admitted that the forms
usually come away from the concrete as readily as when they
are greased.

While the entire reinforced concrete floor in many cases
may be stripped of all form work in a week's time after the con-
crete is poured, it is not always good practice to do so. In the
summer, slab-panels and beam-sides may be removed in about
one week, but the beams and girders should be shored up for
at least three weeks. In the winter, the time must be extended
considerably. Altogether, the removal of the forms calls for
careful work when it is being done, and for discrimination as
to the proper time. The strength of concrete depends greatly
upon the nature of the materials entering into its makeup;
hence what is safe practice in one place may be dangerous in
another.

Reinforcement. We have considered the amount and re-
quirements of the steel above; we shall here consider briefly the

160 REINFORCED CONCRETE BUILDINGS

placing of the steel on the forms, and the preparatory work done
on it.

Several devices of merit are on the market which facilitate
the bending and shaping of the steel; the bending should be
done cold and with so large radii in the curves that no injury
results. While the difficulties incidental to bending heavy steel
bars to sharp corners usually prevent such practice, it is differ-
ent with the U-bars and other light steel, and many half broken
bars of the lighter sections have without doubt found their way
into important work.

Quite frequently the steel bars are assembled to suitable
units, and from every point of view this practice must be rec-
ommended. It has, however, been argued that the assembling
of the bars prevents each bar from sagging to its natural level,
so that some bars are bound to be stressed higher and earlier
than others. There is of course some truth in this, and per-
haps some otherwise unaccountable cracks may be explained
in this manner. The remedy is obvious perfectly straight
bars should be used only, but this follows from numerous other
reasons as well.

The steel may also be bought ready-made, assembled in
units. Owing to the cheapness of factory labor as compared
with field labor, and to the better facilities found in a well-
equipped factory, ready-made steel ought in many cases to
be used with a considerable saving in money and time. How-
ever, the steel yard affords an outlet for the surplus labor, and
for this reason it is often desirable to do the bending, etc., on
the job.

In the early days of reinforced concrete, the beam steel
was placed after a small amount of concrete had been run in
the bottoms of the beams; similarly for the floor, the steel was
placed during concreting. At the present time, the prevailing
and better practice is to place all the steel in the beams and on
the floor, and not to concrete before the steel has been inspected.
Many errors and much poor workmanship are thus eliminated
(Figure 135).

Means should be used for keeping the steel bars the proper
distance away from the face of the form. Metal clips or
cement blocks may be used, and are much to be preferred.
The ordinary way is to have a laborer raise the rods from the

FLOOR SYSTEMS 161

forms with his shovel-blade or a hook made for the purpose,
but such methods rarely result in satisfactory work. There are
many and strong reasons for keeping the slab bars one inch
from the face of the panel forms, and all other steel 1J" to 2"
from the face of the concrete. The specified dimensions should
be adhered to.

FIGURE 135. PLACING STEEL.

Morley Chemical Laboratory, Western Reserve University, Cleveland, O.
C. F. Schweinfurtli, Architect; Alexis Saurbrey, Engineer.

Unit Construction. If we consider the list of patents given
in a preceding chapter, we see that from the earliest days of
the art, the method of casting the pieces in a yard and setting
them when hard has been engaging the attention of inventors.
Nor is this strange when we remember that our present " rein-
forced concrete construction " is a direct off-shoot of the arti-
ficial stone industry, and was originally introduced by men
engaged in that kind of work not less than by men occupied
in the manufacture of monolithic walls.

In the United States, the actual use of " Units " was not
in much use before 1904, and the Textile Machine Works,
erected in the winter of 1904-5 at Reading, Pa., was probably
one of the first serious attempts. The Visintini System was
used for the floors and girders, but the columns were appar-
ently molded in place as in monolithic work; this building is

162 REINFORCED CONCRETE BUILDINGS

50' X 200', four stories high, and it is stated that the 2900 units
were put in place for a total cost of $586.35 (labor only), which
is only about 20 cents each. The completed building cost
7.7 cents per cubic foot.

In 1906 a one-story building was erected for the Edison
Portland Cement Co. at New Village, N. J., the flat roof slabs
were cast on the ground, on top of one another, separated by
paper. After trying bare, oiled, waxed, and soaped paper,
soaping just before casting was found best and most economical.
The roof girders, each 50 feet long, were also cast in the yard.

The one-story building erected for the Central Pennsylvania
Traction Co. at Harrisburg, Pa., in 1909, had roof girders about
37 feet long; these as well as the slabs were cast in the yard and
set when hard. Another building of exactly the same dimen-
sions and similar design had been erected close by several years
before, by the monolithic method; it is stated that the saving
in favor of the unit type was 15 per cent, in this case.

Recently the Unit Construction Co. of St. Louis has erected
a number of buildings up to five stories high under the Unit
System; the general design will be apparent from Figure 136.

In all the buildings just described, each member has been
designed as an individual carrying element without assistance
from the superimposed slab. This necessitates the use of T-
beam sections in order to get the required compressive strength,
or the use of extra deep beams or girders. In the system to
be described below, the slabs are utilized in compression, and
also used as an extra means of tying the entire building together,
while in the cases just described, the pieces are tied together
by virtue of bars projecting into pockets or open spaces in which
concrete is poured.

In the Ransome Unit System, the beams, girders, and col-
umns are made in the yard, but the -slab is cast in situ. The
first building so erected was the three-story office building of
the Foster Armstrong Plant at East Rochester, N. Y., 1904-5,
and the same method was subsequently used extensively in the
United Shoe Machinery Co.'s plant at Beverly, Mass., for a
group of four-story buildings, 60' X 300' in plan, and elsewhere.
(See American Machinist, Sept. 7, 1911; E. L. Ransome: An
Innovation in Concrete Building, from which the following is
taken in part.)

FLOOR SYSTEMS

163

Part "Elevation.^

FIGURE 136.

Unit constructed building with separate slab-section reinforced with marginal
and central beams. (From the Engineering News.)

164 REINFORCED CONCRETE BUILDINGS

Generally speaking, the beams and girders are cast of a depth
equal to the distance from the bottom to the neutral axis only,
and are provided with projecting iron ties. The slab forms
are erected between the beams, which are usually spaced about
4 feet on centers, and rest upon 3X6 inch stringers bolted to
the sides of the beams (Figure 137). The beveled corners of

FIGURE 137.

the slab mold bring the concrete down to the tops of the beams
or girders. In the design, the U-bars must be made with a
view toward creating the required tie between slab and beam.
This is easily and economically taken care of. In addition,
the beams and girders must be calculated to permit a working
load equal to the weight of the forms, the wet slabs, the impact
from concreting apparatus, and the like. When this is properly
attended to there is no necessity for shoring of any kind; in
fact, none is used. It is evident that the slab panels may be
removed in a much shorter time when so constructed than would
be allowable with the monolithic construction, because the
old and properly seasoned beams take care of the entire load,
up to the time when the full " live " load is brought on the
floors.

The beams are mutually connected by means of tie bars
placed in grooves in the tops of the several beams, and have
vertical holes so that the hooked ends of the bars may engage
in holes in the body of the beam. In the more recent develop-
ments of the system, the reinforcing rods project above the tops
of the beams, and the union between the several pieces is effected
simply by a loose rod inserted alongside the tops of the rein-
forcing rods and concreted in with them when the slab is run.
In neither case is the beam considered as part of a continuous
system, although with proper design the continuity might
probably be taken advantage of to some extent.

The column rods are made discontinuous and the tops and
bottoms of the column are enlarged so that the concrete alone
will be sufficient to carry the weights at these points. That
this method of construction is adequate, both for columns and

FLOOR SYSTEMS 165

beams, is amply demonstrated by the absence of vibration under
heavy loads and high-speed machinery.

The column details are of considerable interest. Aside
from the usual reinforcement near the sides of the columns, a
longitudinal rod is inserted in a central cored hole extending
lengthwise through the column. This hole runs from footing
to roof slab. These rods, therefore, tie the columns of each
story to those of the stories below and above. In order to
unite the members, a thick, cream-like grout of 1:1 cement
and sand is poured down this central hole, which is enlarged
and flares out at the bottom, so that a secure and even bed
is insured when the grout flows into this larger space.

One of the most important features is that of setting the
majority of the pieces dry and grouting the joints afterward.
That all joints really are filled is readily ascertained by the
inspector who is instructed to see that a small surplus of mortar
is forced out at the bottom of the joints. For this purpose
small holes are left in the mortar with which the joints are
calked. This mortar is a fairly dry and stiff mixture applied
in the ordinary way with a trowel.

The buildings at Beverly are of rather complicated exterior
design and a number of details, therefore, have been introduced,
which would not be found in ordinary plain factory work. Thus,
the very large flue columns in one of the courts are cast in place,
as the weight of each exceeds the capacity of the derricks.
Pockets and recesses are left in which the beams and girders
are set. Otherwise, all the members are made in advance and
set in place, with the exception of certain of the curtain walls
which are cast in their final position and keyed to the columns
with the ordinary recesses.

Figure 138 shows the principal details of a unit-constructed
building erected under this system. Figure 139 shows some of
the beams in the process of being set by the derrick. Figure
140 shows one of the stairs being set in place. One flight was
built in place, another set of stairs was erected by the unit
method with a saving of about 50 per cent.; the more compli-
cated the required forms are, the greater is the saving.

The contractor's plant used at Beverly comprises an auto-
matic mixing plant whence the concrete is discharged into an
overhead hopper straddling an industrial track. This track

166

REINFORCED CONCRETE BUILDINGS

parallels the building under erection and serves the purpose
of bringing the concrete buckets of about one yard capacity
to either the stiff-leg derrick used on the building proper, or to

the locomotive crane used in the casting yard. The columns
are cast in gangs of four and other pieces in corresponding num-
bers as required. The side forms are removed in one to two
days when the weather is warm, and the pieces are left undis-

FLOOR SYSTEMS

167

FIGURE 139. SETTING THE BEAMS.

United Shoe Machinery Co., Beverly, Mass. Ernest L. Ransome,
Managing Engineer.

FIGURE 140. SETTING A FLIGHT OF STAIRS.
United Shoe Machinery Co., Beverly, Mass. Ernest L. Ransome,
Managing Engineer.

168

REINFORCED CONCRETE BUILDINGS

turbed on their molding bed for about ten days. The periods
are somewhat longer if the weather is cold.

If the pieces come without the reach of the stiff-leg derrick
they are picked up and brought to the building by the loco-
motive crane. The latter serves a multitude of purposes in

Elevation

FIGURE 141. ARRANGEMENT OF PLANT, RANSOMS UNIT
SYSTEM. (FROM CEMENT AGE.)

stripping and moving the forms, concreting in the yard, and the
like. One of the most interesting operations is the removal
of the column cores. These are slightly tapering to facilitate
drawing; they are six inches in diameter at the top, four inches
at the bottom; they are made of wood and covered with sheet
iron. In removing, the crane simply gives a slight pull on the
core, which comes out easily if the concrete is fairly green. No

FLOOR SYSTEMS

169

instances have been recorded where a column was in the least
injured by this treatment.

The concrete is usually handled in one-yard bottom-dump
buckets. The molding of the units presents very few difficul-
ties, and the workmanship is greatly superior to that obtained
with the old method. This high standard is evidenced in all
the work; the lines are straighter and the work generally truer,
than can possibly be obtained commercially by the older methods.

FIGURE 142. SETTING A SLAB.

United Shoe Machinery Co., Beverly, Mass. Ernest L. "Ransome,
Managing Engineer.

Ii is practically impossible to produce a large monolithic rein-
forced-concrete building commercially without some indications
of bulging forms, or of supporting shores having been carelessly
wedged up, or corners fractured in prying the forms loose.
There are no such troubles on work of this kind, because there
are no shores to give way, no f^rms to bulge.

It is quite common to erect a space of floor (for a height of
one story) 60 feet wide by 40 feet long in one working day,
including the setting of slab forms and pouring the concrete.
This also allows time for calking the joints and pouring the
grouting into the cored holes and recesses.

170 REINFORCED CONCRETE BUILDINGS

In one of the one-story buildings erected at Beverly, the
exterior walls were made of 3" concrete panels, reinforced with
\* twisted bars vertically and horizontally spaced about two
feet apart. These panels were set when eight days old, some of
them with door or window openings, or even with the windows
concreted in place (Figure 142). The wall panels were cast
on top of one another in stacks, and the layers separated by
means of a heavy coat of common lime-whitewash.

In all unit work, a proper margin must be allowed for in
the design; all horizontal pieces are made from |" to J" short
and all pockets or recesses are extra large; the openings are
filled with grout afterwards. Two transits were generally used
when setting and plumbing the columns.

CHAPTER XII
FOUNDATIONS AND PILING

Foundations. The reinforced concrete footing in common
use is simply of pyramidic shape with the top removed. The
discussion given in Part II will suffice for the design of ordinary
footings; where two or more adjacent footings are merged, the
design may be of either the flat slab type or it may embody the
slab and beam principle.

Quite commonly, the hole or trench is excavated to the
approximate size of the footing, and the concrete dumped in
the hole without forms of any kind. Such practice is not to
be recommended unless the ground is very stiff. On the con-
trary, as the stability of the entire building depends upon the
integrity of the foundation, the greatest care should be taken
both with the forms for the outside, with the banks of the exca-
vation so that dirt* will not fall into the footing, and with the
proper placing of the steel. The latter should be protected
with not less than 4" of concrete, preferably more, and the con-
crete around the rods should be rich in cement and dense, so
as to exclude water.

If metal base plates are used under the columns, special care
must be taken to prevent hollow places under the plates. This
is best accomplished by setting the plates in a thin grout 1 : 2,
examining afterwards each plate by pounding with a hammer.

Reinforced concrete footings cannot conveniently be put in
under water, so in a wet excavation it is advisable to use plain
concrete footings of ample dimensions to meet all emergencies.

Piling. A number of patents exist covering the various
methods of manufacturing concrete piles, and several of these
are operated by companies making a specialty of concrete piling.
To name a few examples:

The Chenoweth pile, made by rolling a sheet of fresh concrete,
with fire-fabric reinforcement, around a central reinforcement
or tube.

171

172 REINFORCED CONCRETE BUILDINGS

The Raymond pile, cast in a thin shell of steel, which lat-
ter is driven by means of a collapsible pile-core. The shell
remains in the ground.

The Simplex pile, cast in a cylindrical shell strong enough
to stand driving and withdrawing, leaving, however, the point
or shoe behind.

It seems, however, that two methods are open to the public:
(1) To drive a pile, withdrawing it, and then fill the hole with
concrete; and (2) the use of concrete piles molded in advance
and driven as wooden piles. The first of these methods is open
to several objections, so we shall here give an account of a pile-
driving job according to the second method. 1

The piles support a one-story building with 45-foot roof
spans resting upon exterior piers. Underneath each pier three
piles were used; in addition, the chimneys and other founda-
tions rest on piles. A total number of 484 piles 10" X 10" and
13 feet long were required, driven through fill and bog into
tenacious blue clay. The penetration into the clay was from
2 to 3 feet. The piles were cast in the yard (Figure 143); the
ground was levelled and tamped, then covered with a layer of
sand, one inch thick, and re-tamped. On T this bed the piles
were molded, the sand forming one side of the mold. Two
sides were formed by surfaced boards ; the 45 pointed end was
made by simply filling in with molding sand to the required slope,
by two If" thick pieces for the sides, and by trowelling the top
down to the required angle. Otherwise, the surface of the pile
was smoothed with the back of a shovel. The forms were re-
moved in sixteen hours and immediately set again; thirty piles
were cast in one operation in a gang mold.

The concrete was mixed in the proportion 1:1:2, using
clean bank sand and crushed trap rock, pea size. This concrete
had an average compressive strength of 97 tons per square
foot at seven days when tested in the compression machine.
It was found, however, that a 1:2:4 mixture gave an average
strength of 104 tons per square foot in seven days, using 2"
rock. This concrete could have been used successfully and
would have saved $1.03 per pile, reducing the cost from $6.63
to $5.60.

1 From a report submitted by Mr. B. C. Gerwick, who acted as resi-
dent engineer on the job referred to.

FOUNDATIONS AND PILING

173

The reinforcement consisted of 4 \" square twisted rods,
and a spiral reinforcement of \" X \" hoop iron, 4" pitch.
Experiments were made with No. 6 and No. 8 wire of same
pitch, but such piles did not seem to stand the driving as well.
In either case, a \" twisted steel bar was used as an extra col-
lar reinforcement near the head, and the point also had an extra
reinforcing bar of same section.

FIGURE 143. MANUFACTURING REINFORCED CONCRETE PILES.

United Shoe Machinery Co., Beverly, Mass. Ernest L. Ransome,
Managing Engineer.

The sand in the bottom of the form was first tamped and
sprinkled; two inches of concrete were carefully placed, the
reinforcing cage put in, and the balance of the concrete placed
at once. As the pitch of the hoops was 4", they offered little
obstruction to the concreting. When the piles were lifted the
bottoms proved to be smooth and the sand did not adhere to
the concrete.

The fresh concrete was covered with old sacks and kept
damp for three or four days. In loading, the pointed end was
raised with a bar, a rope sling slipped underneath, and the pile
put on the stone wagon by the locomotive crane available in con-
nection with other work. Four piles, each weighing about 1350
pounds, constitute a load, and the haul is about one-half mile.

174 REINFORCED CONCRETE BUILDINGS

A drop-hammer weighing 1600 Ibs. was used in driving.
On top of the pile a cushion of several layers of old fire-hose
rubber and felt was placed, and over this again a 5" cast-iron
block. An oak follower four feet long rests on this block, and
takes the blow of the hammer. The follower must be of good
quality with both ends banded. The fall of the hammer under
the last blow was from 10 to 20 feet, with a penetration under
the last blow of about f ". On an average, it required about
seventy blows to drive a pile which would penetrate from three
to four feet into the clay. One pile was pulled by means of a
lever and found to be in perfect condition. The piles are driven
when eight days old.

The crew consisted of foreman, engineer, four pile driver-
men and two laborers. This crew, including the use of the pile
driver, was hired for $30.00 per day at eight hours. The piles
were, as stated above, usually in groups of three, the distance
between two adjacent groups being 16 feet, so that when three
piles were driven, the driver had to be shifted 16 feet. The
average time of the shift was 23 minutes. The repairs, etc.,
totaled 15 minutes per day, and the average time consumed in
actual driving was 12 minutes. The total average time per
pile was about 20 minutes, or 24 piles put in every eight
hours.

COST OF PILING

484 piles, each 13'-0" long 10"xlO" in section

per pile

Grading casting yard for bottoms $25.45 $ .053

Cost of gang mold for 30 piles

900' B.M. spruce @ $21.00 18.90

Nails, lOd 20

Labor, carpenters @ 47|c per hour 26.65 $45.75 .095

Setting forms, per gang of 30 7.00 .233

Stripping forms, per gang of 30 72 .024

Cleaning and greasing, per gang of 30 1.16 .038

Placing concrete, per cu. ft 0235 .20

Mixing concrete, per cu. ft 022 .187

Labor on reinforcement .66

Cement, stone, sand, and steel, cost 3.517

Hauling mile: Loading 4 with crane 26

Hauling 4 37

Unloading 4 .11 $ .74 .185

FOUNDATIONS AND PILING 175

(Cost of Piling Continued)
Driving: Crew under contract at $30 per day,

Average day's work, 24 piles' $1.25

Coal, oil, and grease .125

Cushion cap, $31.40 .065

Total cost per pile, in place $6.632

(Overhead charges not included.)

Where a large number of piles can be made in a centrally
located yard, it sometimes pays to cast the piles vertically.
Round forms can then be used as well as square.

A permanent plant of this kind exists at Cleveland, Ohio,
as perhaps elsewhere.

CHAPTER XIII
FINISHING OPERATIONS

Corners. Square corners are contrary to the nature of con-
crete. Projecting corners are difficult to make in the first
place, as the concrete seldom penetrates to the very apex of
the angle; in the second place they are liable to injury both
when the forms are removed and while the concrete is green.
Once broken, they cannot be repaired so that the patch looks
like the balance of the work. The re-entrant corner is easily
made, but objectionable for the reason that a sharp dent in the
concrete very often forms the starting-point for a crack which
might otherwise have been avoided. This is explained by the
same observations made in regard to cast-iron; in addition the
form is often locked to the concrete by a sharp corner so that
the workmen use too much force in removing the forms. Broken
corners are the great drawbacks in concrete construction; they
may easily be avoided by chamfering the forms so that all
sharp angles are excluded.

Flat Surfaces. All the defects in the form work will show
on a flat concrete surface; in addition, all defects in the con-
crete will show. It is difficult if not impossible to make per-
fect forms; it is practically impossible to maintain the forms
in perfect condition, because the water in the concrete is ab-
sorbed by the wood in the forms, causing swelling and warping.
The marks left by the forms are called " board marks"; if a
finished piece of work is desired the board marks must be either
concealed or erased. In the first place, the concrete work is
faced with various materials, such as brick, terra cotta, plaster,
etc.; in the second case, the surface itself is improved by tool-
ing, rubbing, or brushing.

Plastering. Plaster usually comes off again sooner or later,
especially on outside work. It should be used for indoor work
only, and then only in emergencies; if plaster is insisted upon,

176

FINISHING OPERATIONS 177

the ceilings should be burned with acid, and the form work
should be made as rough as possible. In that case, very fair
results may be obtained, but plastering always remains an art
more than a science, so that skilled labor is a most essential
feature. There are various plasters on the market, made espe-
cially for concrete surfaces. A great deal of satisfactory work
has been done with such material, but its general use is still
too recent to warrant absolute confidence.

Tile-concrete construction is well adapted for plastering.

Brick and Terra Cotta Facing. The concrete must be true
to line and level, as it is otherwise difficult to put on the brick
facing, and impossible to put on the terra cotta facing. For
brick, galvanized wire wall ties are left projecting from the
concrete, say 12" apart diagonally, and bolts are placed in the
concrete to receive the angle irons which carry the brick work
over door and window openings. For terra cotta the arrange-
ment of the ties and supports varies greatly with the design.
Usually a hollow space is left between the concrete and the
terra cotta facing; cement mortar deposited in this space ties
the facing to the concrete behind, as the facing blocks have
projecting ribs on the back, while iron anchors project from the
concrete, so that the whole is locked securely together. At
intervals supporting ledges must be arranged to transmit the
weight of the facing (which is considerable) to the structural
concrete. It is advisable to make all ties of soft iron so that
they will not break when adjusted. It is very important that
suitable play be provided for, as neither brick nor terra cotta
can be made to exact dimensions, while the concrete construc-
tion is very apt to vary slightly from the specified dimensions.

Improved Surfaces. The removal of the board marks is
possible under one condition only, and that is, that all joints
between boards must be tight. The joints fill with the finer
parts of the mixture, especially with cement, so that the small
ridges between the boards are rich in cement. It follows that
the concrete immediately behind the ridge is leaner in cement
than other parts of the surface, and it is therefore softer than
the surface generally, so that any mechanical treatment of
the surface removes too much at the ridges, forming small
grooves looking almost as bad as the original ridge. Hence
the joints must be tight, so that no cement can ooze out, and

178

REINFORCED CONCRETE BUILDINGS

fairly smooth, so that few and small ridges only are formed;
the carpenter work must be good, the forms must be made of
good lumber and nailed securely to the cleats to prevent spring.
The completed form must be coated with grease, vaseline,
crude oil, or, best of all, a cheap grade of black japan. This
kind of work is expensive when undertaken on a large scale;
to secure the forms which retain the concrete so that there will
be absolutely no deflection, is not easy. When now the forms
have been completed to the satisfaction of the engineer the
concrete is deposited at once, as sun and wind will destroy the
best-made piece of form work. The concrete must be placed
with the greatest care, as every defect will show in the finished
work. The concrete mixture must be so gauged that there
is a surplus of mortar; usually 1: 2: 3J or 1:2:3 will be found
suitable. A mortar facing run in the form with the body of the
concrete may be used on work of very large dimensions only,
as it is otherwise practically impossible to deposit the mortar.
But a mortar without stone looks very dull when tooled. .

Tooling. The surface is bush-hammered either by hand,
or, preferably, with a pneumatic tool. An ordinary chisel may
be used, but special tools are sold for this purpose. The defects

FIGURE 144a.

in the concrete are brought out strongly by this method, and
repair work looks very bad, especially in rainy weather. But if
the concrete was put in right in the first place the effect is very
pleasing, and for large surfaces tooling must be considered as the
most satisfactory and most pleasing finish (Figures 144a and 6).

FINISHING OPERATIONS

179

Rubbing with carborundum blocks (or similar hard material)
is very expensive. The form work must be absolutely first-
class, and the concrete must be very hard before rubbing is
attempted, but the results justify the expense. The surface
is removed to a depth of J" to f"; the grain of the concrete is

FIGURE 1446. GIRLS' DORMITORY, LELAND STANFORD JR. UNIVERSITY,

PALO ALTO, CAL.

The exterior construction and tl;e floors are of reinforced concrete.
Ernest L. Ransome, Engineer.

thereby exposed with a smooth and glossy surface. It has beer*
found most satisfactory to rub the concrete down dry, in the
cases of ceilings and similar surfaces where large quantities of
water cannot be applied; certain kinds of cement floors are
manufactured in the same way, but by a wet rubbing. When
the concrete is to be painted, a very good surface may be had
by this method by taking only the board marks off, leaving a
practically, but not entirely, smooth surface (Figure 145).

Brushing. The forms are removed as soon as feasible and
the green surface brushed hard with wire brushes so that the
mortar between the stone aggregate is removed. Plenty of
water is used in this process, and the stones finally show in re-
lief on the dull gray mortar backing. A sparkling, many-
colored aggregate is used, and the effect is very good, although
perhaps a little artificial.

Unfinished surfaces are sometimes used for factory build-
ings, stables, etc., and all degrees of work, from very good to
very poor, may be found. Occasionally an effort is made to
improve such surface by rubbing cement mortar into the pores
at once upon removal of the forms, and then rubbing the entire

180

REINFORCED CONCRETE BUILDINGS

surface down with cement bricks, or sometimes hardwood blocks.
If the purpose is to fill the pores only there can be no criticism
of this method, provided all the surplus mortar is again removed;

FIGURE 145. MONOLITHIC CONCRETE STAIRS AND RAIL, CAST IN
ONE PIECE THREE STORIES HIGH. CONCRETE WAS RUBBED WITH
CARBORUNDUM AND PAINTED.

Morley Chemical Laboratory, Western Reserve University. C. F.
Schweinfurth, Architect; Alexis Saurbrey, Engineer.

often the mortar is allowed to remain on the surface as a thin
film, in which case more or less pealing is bound to follow.
Many arches and abutments throughout the country have
been provided with a coat of this kind, and there is hardly any

FINISHING OPERATIONS 181

locality where samples of this work may not be found, showing
the disgraceful results obtained.

Cement Finish. Only in exceptional cases may the rough
concrete floor be used, partly because the surface is too coarse,
partly because it wears out too rapidly, partly because it forms
part of the structure itself and therefore needs protection against
wear. The rough floor is therefore covered with a sheet of
cement mortar, called " cement finish." It may be applied to
the green concrete surface as soon as it is hard enough to allow
walking on it, or at any time afterwards. In the first case a
good bond is assured by simple precautions, such as removal
of the " slam," a white scum forming on top of concrete laid
with an excess of water. If the concrete is hard and old, the
surface must be cleaned with muriatic acid, water, and scrub-
bing brushes, as no cement finish will stick to a dirty surface.
The finish is put on in a rather dry condition, like soft dough;
about |" or 1" thick as a minimum, and up to 2" thick as a
maximum. It is trowelled to a hard surface in order to ensure
good wearing qualities, and it may be divided into panels or
not, according to circumstances. The object in dividing the
surface into panels is purely ornamental; cracks may be avoided
by dividing the base as well as the surface into suitable blocks.
A structural reinforced concrete floor is not readily divided in
this manner; in fact, one of the objects of good design is to
make the floor continuous as far as possible. It is therefore
proper to divide basement floors, sidewalks, and similar pieces
into blocks by deep and wide separations, while the finish on
a reinforced floor may as well be laid in one continuous sheet.
Thereby is also avoided the breaking of the edges of the individ-
ual blocks, so likely to take place under heavy trucks in ware-
houses. A surface laid in this manner will show all the cracks
in the base below, and for this reason as well as on general
principles, all care should be taken to avoid cracking. These
are, as stated above: proper arrangement of principal and sec-
ondary reinforcement, bevelling of all re-entrant corners between
beams and slabs, protection against wind and sun, liberal
sprinkling, and avoidance of premature loading and jarring.
Great care should be taken when the forms are removed; in
fact, a large number of " unaccountable " cracks are due to
carelessness in removing the forms.

182 REINFORCED CONCRETE BUILDINGS

To avoid cracks entirely is hardly possible, especially in
tile-concrete floors, where top cracks are very frequent, due
to the unyielding nature of the tiles. Such floors are better
provided with a wood floor on top of the concrete.

The mixture used for the finish should be one part of cement
to two parts of selected sand, whereby is meant a sand with
particles well graded from fine to coarse, clean, and sharp.
The largest particles should not exceed the 1/4" to 3/8 " ring.
The finer the sand, the easier it works under the trowel, so that
fine sand is the preference of the cement finisher, to the injury
of the work. Aside from proper materials, skilled cement fin-
ishers are indispensable to good results. The engineers' super-
vision of the workmanship is usually confined to the results, as
few engineers are sufficiently well posted on cement finish to
supervise the details of the workmanship.

CHAPTER XIV
FIREPROOFING AND FIRES

No building is absolutely " fireproof," and the most that can
be accomplished is to retard the spread of the fire to such an
extent that the fire can be brought under control before the
barriers are destroyed. But this is only one side of the question,
for in many cases more damage is caused by smoke or water than,
by the fire itself. To prevent the smoke from penetrating to
portions of the building not affected directly by the fire, is usually
impossible, but much may be done to prevent the water from
leaking down into the stories below the fire. We encounter
here a much neglected problem: In most cases, pipes for heat,
sewerage, etc., are carried through the floors by means of open
sleeves, and the water naturally finds its way out through all
these holes in the floors. However, if we consider a perfectly
waterproof floor without means of escape for the water, the load
on the flooded floor might easily exceed the capacity of the struc-
ture to a dangerous degree.

While therefore many owners of reinforced concrete buildings
carry no insurance on the building itself, it is not advisable to
neglect the insurance on the contents, except where they are of
such a nature that they are not easily injured by water, smoke,
or heat. Much will also depend upon the character of windows,
partitions, stair- and elevator- wells. In the majority of cases,
the so-called " fireproof " building is equipped with wood
trimmings, plain glass in wooden casings, and has a wood floor
over the concrete base. While in such cases the reinforced
concrete escapes injury, the contents are usually a total loss,
frequently with loss of human lives. There is, without doubt,
room for great improvement along these lines. The arrange-
ment of these matters, as well as those pertaining to stairwells
and elevator openings is, however, beyond the scope of this book.

Turning now to the concrete itself, it is admitted that no

183

184 REINFORCED CONCRETE BUILDINGS

material is absolutely fireproof, and concrete as well as other
materials must finally fail under a long and severe fire test.
But each particle in the crystallized concrete contains chemically
bound water which is given off under high temperature, and
the temperature of the concrete itself is thereby prevented from
reaching a high intensity throughout the mass. Concrete itself
is a good conductor of heat compared with the true insulating
materials.

The concrete surrounding the steel must be made so thick
that a large part of it may lose its water (and thereby its strength)
without injuring the strength of the concrete touching the steel,
as otherwise failure would result. But large amounts of concrete
are expensive: practice has therefore settled upon 2" of protec-
tion on columns and girders or beams, and I" on slabs. These
thicknesses are entirely arbitrary and may be varied according
to location and exposure, but it will be seen that they are also
structural minima, and that with less concrete around the steel
there can be absolutely no bond, and a small enough factor of
safety as far as the workmanship is concerned. It happens
quite often that these specified minima are still further decreased
by carelessness on the part of the concreting gang, so that rust
spots show through the concrete, or the steel is even exposed
to view. Such conditions are always indications of workmanship
of the poorest class.

The parts most exposed to the attack of fire are the projecting
corners. Experience has shown that rounded or chamfered
corners are much less liable to attack than a plain square corner ;
in addition, square corners are almost always more or less
fractured when the forms are removed. In warehouses or other
buildings where heavy stuff is handled, the lower part of the
columns should have extra protection, such as angle iron guards
for the corners, or even an iron mantel surrounding the concrete
entirely. The same is true of thresholds and stairways; the
steps of the latter are often protected with some patented metal
covering, of which the nosing piece is the most essential part.
The elevator hatches should also have a proper protection;
angle iron guards are easily fastened and effective.

In addition to these obvious safeguards, we have an excellent
method of increasing the fire resistance of concrete. It simply
consists in adding a small amount of salt to the water with which

FIREPROOFING AND FIRES 185

the concrete is mixed. This fact was proved in a peculiar
manner: During the erection of the Bayonne, New Jersey,
warehouse for the Pacific Coast Borax Co. in the winter 1897-
1898, experiments were made with salt as a frost preventive,
the work being carried on in very severe weather, with tempera-
tures sometimes below zero. In 1902, the building went through
an exceptionally hot fire, started from a burst oil main in
the basement, which soon was flooded with burning oil. On
the upper floors, combustible materials of all kinds, including
heavy barrels and boxes, added to the fire, yet the concrete came
out of the fire with hardly any damage, and the concrete work
of the entire building, 200' X 240', and partly four stories high,
was repaired for less than $1000. Quantities of fused cast-iron
from the machinery and copper from the dynamos and motors
were in evidence after the fire (See Iron Age, May 28, 1902),
showing that the fire must have been unusually hot. (Figure
146 shows a block of fused cast-iron from this fire).

The increased fire-resistance due to an admixture of salt has
also been demonstrated on test specimens made for the purpose.

The general behavior of reinforced concrete in conflagrations
has been very satisfactory in the Pittsburgh and Baltimore
fires, where but few reinforced concrete buildings were within
the fire-swept area; the most convincing proofs were however
furnished in the San Francisco earthquake and conflagration,
where buildings of all kinds suffered, but those of reinforced
concrete less than any others. Tests without number have been
made to determine the conductivity and fire-resistance of con-
crete. As a result, it may be stated that the better the concrete
is made originally, the better it will be adapted for fireproofing
purposes, and a four-inch concrete wall may be exposed to the
hottest fire for hours, on one side, while the other side remains so
cool that the hand may be placed against it without fear. The
use of reinforced concrete for heat flues and chimneys is justified
from this fact.

At the present time, a " fireproof " floor may be built in one
of three ways: (1) By combining steel and concrete, whether
the steel be in the form of a reinforcement, or as an independent
skeleton. (2) By combining steel and tile, in which case the
steel forms the well-known skeleton so commonly used in the
modern skyscraper. (3) By combining steel with both tile and

186

REINFORCED CONCRETE BUILDINGS

concrete in various ways. (There is indeed a fourth method,
by suspending brick arches between steel beams, but this is in
little use at present except for special structures.)

FIGURE 146. FUSED CAST-IRON PULLEY FROM THE
BAYONNE FIRE.

The competition is therefore between concrete and hollow
tile, both or either in combination with steel. Wherever put
to the test, concrete appears to have carried the day. Two
reasons suggest themselves for this fact: (1) The expansion of
concrete and of steel is practically the same, while the expansion
of tile is different from that of steel, so that there is a tendency
to readjustment under fire in the latter case, and none in the

FIREPROOFING AND FIRES 187

former; and (2) the hollow tiles in common use are very poor
conductors of heat, so 'that, while the steel is protected in an
excellent manner while the tiles stand up, the unequal expansion
of the several parts of the same tile causes the lower flange to
break off, especially when suddenly cooled, thus exposing the
steel. The proofs of this statement are ample and convincing
and may be seen by reference to the photographs in the Govern-
mental report on the San Francisco fire.

It must be understood that lath and plaster construction is
not included as a fireproof material, and has no value as such.

In closing this paragraph, we must call attention to the ever
present danger attending the use of reinforced concrete buildings
veneered with brick or similar material, where the horizontal
supports are exposed over the window openings as is nearly
always the case. A hot flame through the window would
probably injure the supports and wall-ties sufficiently to cause
parts of the veneer to fall, although no such accidents have ever
come to our attention.

Other considerations also lead us to doubt the continued
stability of thin veneer walls, and there seems to be no good
reasons for their extensive use.

CHAPTER XV

REPAIRS TO EXISTING BUILDINGS

THE general wear and tear on a well-constructed reinforced
concrete building is insignificant and confined to the finish coat
of the floor. The repairs consist in careful removal of the
worn surface, thorough cleaning of the floor, eventually with
weak muriatic acid, the application of a bonding substance
such as Livingstone, Ransomite, or similar, and the placing cf
a new surface coat.

It happens, however, occasionally that carelessness when the
work was made causes trouble, and a brief description of some
cases of this kind may be of interest.

Cracking of the floor slab may be due to a number of causes:
concrete poorly proportioned with accompanying excessive
contraction, too rapid drying out of the concrete, etc. All cracks
may be repaired that are caused by a natural adjustment when
a stable condition has been reached, by simply cutting out the
cracks, dove-tailing the bonding surface on both sides, and filling
in with fresh concrete. Many slabs are broken when the forms
are removed, although the crack does not appear for some time.
Cracking of beams or girders is usually due to careless or pre-
mature removal of the forms. A gaping crack is an indisputable
sign that the reinforcement has slipped, and it is then a question
of removing the entire beam and putting in a new one. In that
case, pockets are left for the new beam at each end, and the
new concrete tied as well as possible to the old work.

If upon examination the steel is found too high in the beam,
as sometimes happens, it is possible to cut away the bottom por-
tion of the beam for its entire length, and to put in a new bottom
with proper reinforcement, leaving the old steel in place. The
new bottom is tied to the old beam by means of frequent U-bars
which are concealed in vertical grooves cut for the purpose in
the sides of the beam ; the upper ends of the U-bars are carefully
anchored in new portions of the slab inserted in spaces made to

188

REPAIRS TO EXISTING BUILDINGS 189

receive them. The new reinforcement must extend well over the
supports, and firm anchorage must be provided for it. Hollow
spaces in the columns are repaired by cutting the poor concrete
entirely away, cleaning the surfaces, and pouring new concrete
in. It must here be observed that the concrete surfaces are
made to slope to such an extent that all air can escape, preferably
through a vent on the opposite side of the funnel through which
the new concrete is poured. The purpose of the vent and funnel
is to make certain that the concrete fills all cavities by putting
the fresh concrete under pressure. The surplus stuff is dressed
off and the surfaces smoothed down.

Under no circumstances should cutting in concrete be done except
in the presence of a reliable engineer who understands the structural
importance of each member, and proper shoring must be put under
the beams, etc., before the cutting is proceeded with.

When cutting holes in structural concrete of any kind, the
lighter hammers and chisels should be used in preference to the
heavy tools, and many light blows rather than a few heavy ones
should be insisted upon for the reason that heavy blows have
considerable shattering effect on the concrete, especially in the
first few weeks after pouring. Hence the pneumatic drill is to
be preferred where obtainable, even if at much greater cost,
especially when putting in new bolt holes, etc., in great number.
But drilling into the bottoms of beams and girders, or into the
sides of hooped columns, should not be allowed when avoidable,
and it is quite often possible to confine the drilling to the slabs
and the sides of the beams.

Sometimes, an annoying and troublesome condition arises
from the fact that the laitance, or dead cement, accumulates
in the beam-bottoms. This can happen only where the concrete
has been made with a surplus of water, and the water has been
allowed to run ahead of the concrete into the bottoms of the
beams, carrying considerable amounts of cement with it. The
water and cement form a soapy, white substance which never
sets up, and, after a while, large cakes drop from the beam-bot-
toms, sometimes an inch thick. The only efficient manner of
repairing is by putting in a new beam-bottom, tying the new con-
crete to the old, and this procedure is very expensive, although
cheaper in the long run and far better than repairs with cement
mortar troweled on.

190 REINFORCED CONCRETE BUILDINGS

In exceptional cases, poor foundations cause unequal settle-
ment and cracks. If the footings finally adjust themselves to
a permanent level, the cracks may be repaired as described above;
but even in that case, the building has lost considerably in carry-
ing capacity, especially if constructed with continuous beams
and girders. In one case, it became necessary to install new
footings, columns, and girders alongside the old work, but in
that case, the original footings had not been brought down to
the proper level, and the girders had been erroneously designed.

After-treatment of the surface of cement finish may be desirable
in exceptional cases to further the hardening, in which case a
wash of equal parts of water and the commercial solution of
Silicate of Soda is applied. Silicate of Potash may be substituted
for the Soda, but the best results are obtained by using a wash
of two parts of water with one part of Silicate of Soda and one
part of Silicate of Potash ; the two latter in the ordinary commer-
cial solution.

A final wash with Chloride of Calcium is very desirable,
especially if there is no free lime in the cement.

For repairs to small cracks, a mixture of Silicate of Soda and
Chloride of Calcium may be poured into the cracks, but as this
solution sets very rapidly, Alum may be used instead of the
Chloride of Calcium as this mixture sets much slower.

CHAPTER XVI

ACCIDENTS

A GOOD reinforced concrete building is as permanent as any
type of construction known today, and where a building of this
kind has been taken in use, it has never been known to fail,
with one or two exceptions where the design was faulty, or
where the foundations were entirely inadequate. In the- very
few cases where reinforced concrete buildings have been pur-
posely demolished the task has proved an arduous one, as
for instance the seven-story building of the Baltimore News
which was taken down in the spring of 1911 to give room for a
larger structure.

It would be possible to enumerate a number of minor mishaps,
serious enough to those whom they affected, but similar to those
which do occur in all lines of building construction, whether
brick, steel, or concrete. Here we will limit ourselves to the
few disasters which attracted universal attention, and give a
brief account of the cause in each case, in so far as the cause is
known. It will be appreciated that the tangled mass of debris,
and the more or less colored account of the actual conditions
given by the parties directly affected, furnish but poor material
upon which to base an unbiased opinion.

The collapse of a portion of the Amsden Block at South Fram-
ingham, Massachusetts, in July, 1906, has been traced to the
settling of the foundations, and inasmuch as the interior con-
struction consisted of reinforced concrete only for the slabs and
fireproofing, the beams being of steel, and the columns of cast-
iron, there is no reason to believe that the reinforced concrete
was to blame for the failure.

The Bixby Hotel, at Long Beach, California, was a building
H-shaped in plan, with reinforced concrete construction of the
tile-and-concrete variety for the interior, and the usual concrete
skeleton for the exterior construction. A large portion of the

191

192 REINFORCED CONCRETE BUILDINGS

bar of the " H " fell while the roof was being concreted, on
November 9, 1906. Questions as to design and unit stresses
assigned to the columns have been raised, and it seems probable
that some of the columns failed in one of the upper stories.
However this may be, premature removal of the falsework
undoubtedly entered into the causes of the collapse.

The Eastman Kodak Company's building at Rochester, New
York, was partly demolished by a sudden failure on November
21, 1906, while the waterproofing was being put on the roof,
which at that time was seventeen to eighteen days old. The
initial failure seems to have been traced without doubt to column
No. 47 which at that time was about three weeks old, but there
also seems to have been more or less neglect on the job with
reference to the proper placing of the column reinforcement and
some of the columns had considerable amounts of saw-dust and
chips of wood embedded in the concrete. The shores were
probably being removed in some portions of the building during
the time preceding the collapse.

The Bridgman Bros, building in Philadelphia was partly
wrecked on July 9, 1907, when some foreign laborers removed
all the shores under the roof which at that time was only 5| days
old, owing to a misunderstanding of orders. The falling por-
tions of the roof carried with it all the floors directly below,
except a portion of the first floor, which partly withstood the
shock of the falling concrete.

Failure under test load took place in the roof of the reservoir,
at the United States Naval Academy, Annapolis, Maryland, in
December, 1908. The footings apparently had been put in on
very wet clay, and these footings were without reinforcement
except for the wirecloth which was used to reinforce the floor,
and which was run into the footings, being depressed under the
columns to near the bottom of the footings as well as possible.

On April 7, 1910, a car barn then nearly completed, and
belonging to the Shore Line Electric Company, at Saybrook,
Connecticut, partly collapsed owing to the premature removal
of the forms under the roof.

Finally, the Henke Building in Cleveland, Ohio, was entirely
destroyed by collapse on November 22, 1910. (Figure 147).
The building was four stories high, and, with the exception of a
few of the old brickwalls used for the outside, the entire building

ACCIDENTS

193

fell so as to fill the basement level with the sidewalk of the
street. The roof was just completed, and it has been suggested
that work was going on in the building on the removal of the
few remaining shores in the second story from the top; there
were several indications of column failures in that story. 1

FIGURE 147. WRECK OF THE HENKE BUILDING IN CLEVELAND.
Photo by Alexis Saurbrey, who examined ruins for owner.

In nearly every one of these cases, serious errors in regard to
supervision and workmanship have been proved, but it has

1 While this was in the press, the current issues of engineering papers
reported the failure, on Dec. 6, 1911, of a three-story building under erection
for the Prest-O-Lite Company, at Indianapolis, with considerable loss of
life. The building was of the beam and girderless type, but the details
are not available.

194 REINFORCED CONCRETE BUILDINGS

not been possible, as far as known, to connect the neglect with
the actual causes of the collapse. Nearly all of these buildings
fell in the spring or in the fall, when the setting of the concrete
is greatly retarded by the cold weather, and even if the days may
be quite warm, the nights are cool, and the water used for the
concrete very likely quite cold. It is easy to say that if the shores
had been left in a few weeks longer, the failures would not have
occurred, but it is no easy matter to prove such assertions.

Attention is called to the circumstance that the failures have
frequently suggested weakness in certain columns, and in all
such cases, the horizontal column reinforcement has been found
grossly inadequate or even missing. It is a positive necessity
to provide proper ties or hoops in the columns, not one or two
column Diameters apart, but two to three inches apart, thoroughly
binding the loose ends of the hoops together so that they cannot
slip. In addition, no shores should be removed before the
column sides have been opened and a careful and thorough
inspection of all the columns made; not of a few isolated spots
on a column here and there, but of the entire height of all four
sides of each column.

Undoubtedly, there are grades of efficiency in concrete work
as elsewhere, although the best is none too good in most cases.
It is however confidently believed that serious failures of rein-
forced concrete buildings will not occur, if the following simple
precautions are taken:

Tie all steel bars into the next span. Use closely spaced hoops
in all columns. And see that the concrete is hard before the shores
are removed.

CHAPTER XVII

SUPERINTENDENT'S SPECIFICATIONS

THE following instructions have been used by the Ransome &
Smith Company, as a standard of daily practice for their Super-
intendents.

General. Order and close attention to details is essential.
Want of due care in proportioning, in mixing, or in the placing
of the steel may lead to destructive results. Reinforced con-
crete construction requires close, continuous, intelligent super-
vision. If this is not given, disaster is not far off. A
superintendent places a severe handicap upon himself unless
he so organizes his men that from the lowest up to the high-
est each clearly understands his duties and limitations and
knows what he has to do, and unless he so arranges his own
time that he can, as a usual thing, devote sufficient time to the
unexpected demands that will frequently be made upon him for
his attention.

All accounts must be kept up to date and promptly passed
upon. All orders must pass through the New York Office
except in emergencies, then use emergency orders and forward
copies to the New York Office for confirmation.

Temporary Offices and Buildings, Setting up Plant, etc. In
setting up the plant see that the mixer is in good line and securely
placed upon a level bed. Keep all running gear and wearing
parts free from dirt and well oiled and greased. Keep both
inside and outside of the mixer, hoisting tub, hoppers, gates,
barrows, etc., free from accumulated dirt or concrete of over a
day old. Thoroughly cleanse off every night the day's accumu-
lation of concrete &nd dirt upon tools -and machinery. Protect
scaffolds and all openings in floors with suitable hand-rails and
use every reasonable precaution against accident.

Excavating and Grading. Make these of the dimensions and
depth that shall be determined, upon the final examination of the

195

196 REINFORCED CONCRETE BUILDINGS

ground. In excavating, give sufficient slope to the sides of the
hole or trench to prevent caving in, or protect with sheet piling,
and excavate the final depth, corresponding to the depth of the
footing, of the exact size required. Do not excavate this final
depth much before the time for filling in the concrete. Re-fill
as rapidly as the work permits and thoroughly compact all re-
filling that subsequently becomes floor-bearing. In grading,
follow specifications of the Contract.

Molds. Molds shall be made in strict accordance with draw-
ings, which will be furnished from headquarters.

All molds to be thoroughly fastened together. They must
not only be set true and plumb to line, but must be so rigidly
held in place that they will resist successfully all tendency to
move them that the placing of the concrete may give. All
interior, or molding, faces to be thoroughly greased with crude
oil before using, and thoroughly cleaned and re-greased at every
re-use. All open joints, broken off corners, knot holes, to be
properly puttied up with ordinary or improved putty immediately
before placing the concrete.

Concrete. Every car load of cement must be tested.

Aggregates will be finally determined upon, at which time the
proportions of cement with these will be given. Salt shall
be used at the rate of four pounds to a barrel of cement.
It shall first be dissolved (in a tank placed above the level of
the top of the mixer) to a saturated solution. Then for every
bag of cement used in the batch, add three pints of this saturated
solution.

In mixing, put the water in first, then the rock, then cement
and sand; mix thoroughly and in placing see that the mixed
concrete is of such consistency and character that it will pour
from the wheelbarrows.

In starting the piers, use a very wet concrete, into each batch
of which an additional bag of cement has been placed, for the
first foot of height of the piers. Fill each pier in a continuous
operation until it is full ; short intermissions of time not sufficient
to permit the concrete to stiffen may be disregarded and con-
sidered as continuous filling. Keep the column work at least
twelve hours ahead of the floor work.

All floors must be thoroughly rolled with the first, second, and
third roller, beginning with the lightest; continue rolling until

SUPERINTENDENT'S SPECIFICATIONS 197

the effect thereof is not apparent. 1 The concrete shall be com-
pleted in any one unit part before the initial set appears on its
surface.

In concreting strike the bars in preference to the concrete
between the bars, with the tampers. Great care must be taken
to see that the bars are thoroughly embedded in the concrete.
Wherever there is a nest of cross-bars that the concrete will not
readily penetrate, pour into same sufficient cement grout 1 : 2 to
thoroughly fill all spaces. Special care must be taken (especially
in hot weather) to follow up this grout with the body of the
concrete before the grout has stiffened. If the circumstances are
such that the grout stiffens too quickly for convenient working,
time may be gained by throwing on the face of the grout sufficient
fresh concrete to cover it, and in turn should this fresh concrete
stiffen before it is covered with the main body of concrete, it
may be renewed from time to time as above by further small
additions of concrete. It is, however, important that neither
the original surface nor any of the renewed surfaces be allowed
to stiffen before the next layer is applied.

The natural slope of the concrete may be used to terminate
any days' work or the work of any period provided the following
precautions are taken:

The surface of this slope must be finished with a drier mixture
than usual into which an extra batch of cement has been added.
Care must be taken also that this sloping surface is thoroughly
tamped down into a compact surface, no loose porous lumps or
portions being left anywhere. Before starting the work anew,
if this concrete is sufficiently soft to permit of the cement on its
surface being thoroughly brushed off with wire brushes, brush
it off thus and top off the surface with a liberal coat of pure
cement grout well brushed in. If it is too hard for this oper-
ation use acid joint. 2 For the concrete needed to cover the
sloping surfaces of the previous work throw into each batch an

1 The use of rollers on concrete floors is not in accordance with usual or
current practice. However, it might well be used with beneficial results as
shown by my own practice of many years. Note however the necessity of
good strong centering that will not yield the least under the heaviest roller
used. E. L. R.

2 In more recent practice, the vertical joint has been used, as the sloping
joint is rather difficult to make and not so easily repaired in case of trouble.
A. S.

198 REINFORCED CONCRETE BUILDINGS

extra bag of cement, then proceed with the work as previously
described.

Care must be taken to leave the surface of the concrete at the
proper level. A variation of more than 1/4 " in the finished level
will not be considered as good work.

The concrete must be kept wet for at least ten days. During
concreting, a surveyor must be kept constantly at the work to
determine whether or not there is any settlement in the falsework,
and, in case there should be in exceptional cases, the defect should
be rectified before the concrete sets. This also applies to the
alignment of the exterior surfaces of the work.

Steel. All steel shall be kept as free from rust as prac-
ticable. All bars must be placed as shown on the drawings.
No variation in height of over 1/2 " is allowable, or in other
dimensions of over 3/4 ". No steel must appear on the surface
of the work. Steel that would otherwise reach the surface
must be wrapped with one or more turns of protected wire or
stout marlin.

All the steel must be placed ahead of the concrete except where
instructions are given to the contrary (in very exceptional cases
only).

Finishing. All floors shall be treated with acid joint
and finish, except where the finish is put on before the floor is
thoroughly set. This latter shall be of the proportion given,
mixed quite stiff and thoroughly well troweled down and
worked to a true smooth finish. Extreme care must be taken
to follow closely the instructions given here below relative to
the acid joint. 1

Acid Joint. (1) Thoroughly sweep the floor, removing all
loose concrete dust and debris, etc.

(2) Wash floor thoroughly with water.

(3) Wash floor with acid mixture (1 acid 18% to 1 water)

pouring it on the floor freely and slowly sweeping it
forward. Follow this washing with a second and third
in like manner.

(4) Give the floor a final and thorough washing of water.

Immediately before laying the finish:

(5) Thoroughly wet the floor.

iThis method is covered by my U. S. Patent No. 860,942, Oct. 3, 1905,
Ernest L. Ransome.

SUPERINTENDENTS SPECIFICATIONS 199

(6) Rub in a pure cement cream with wire brushes, sweep-

ing forth and back, going over the same ground seven
times.

(7) Before this shows signs of setting, sweep over it more of

the cement cream so as to leave on the surface a thick-
ness of about 1 /8 ". This cream should be thicker than
the first.

(8) Before the above layer shows signs of setting, put on the

finish.

CHAPTER XVIII
THE ENGINEER

As compared with other methods of construction, reinforced
concrete is essentially a manufacture. From the earliest days
of the art, this was recognized by the makers, who called them-
selves artificial stone manufacturers and concrete manufacturers.
The contractor receives the raw materials in the form of cement,
sand, stone, and steel bars, and from these he manufactures
the structure, while in other types of building contracting, the
finished product is received at the building, and then simply
erected in place. Hence the reinforced concrete contractor is
charged with two duties, namely, manufacture and erection,
where the other contractor has only one, namely, erection.

It follows that expert knowledge, similar to that possessed,
for instance, by the steel mill organization, must in some manner
be supplied on the reinforced concrete job. According to cir-
cumstances, the expert services are provided by either the
owner, the architect, the contractor, or the local building depart-
ment, if indeed they are not wholly absent, which appears to
happen occasionally. The latter case is entirely too frequent,
due to the prevailing lack of understanding of the difficulties
incidental to reinforced concrete work. It is the duty of those
who know, to emphasize this fact, each in his own locality, so
that the general public may at last appreciate the absolute neces-
sity of expert skill on all reinforced concrete work.

It is not believed that building ordinances or regulations
can cope with this problem successfully. In Cleveland, Ohio,
the owner is required by law to provide an inspector who shall
be present at all times when concrete is being placed on reinforced
concrete buildings; the inspector must pass an examination
before the building authorities. But this examination is so
elementary that nothing even remotely approaching expert
supervision is obtained. In many respects, the ordinance is
objectionable to the owner, who cannot always command the

200

THE ENGINEER 201

services of an examined inspector at the proper moment, as
well as to the contractor, who may sometimes have to
wait for the inspector. In spite of these minor objections,
the system is undoubtedly beneficial in Cleveland at the pres-
ent time, although the possibilities for misuse are great and
always present. The chief objection would seem to be in the
fact that the owner's conscience is lulled to sleep in the hope
that a paternal city department will see him through all troubles,
while as a matter of fact the inspection is barely sufficient to
guard against gross and continuous blunders.

In Boston, Mass., 1 the law provides: "When the struc-
tural use of concrete is proposed, a specification stating the
quality and proportions of materials and the methods of mixing
the same shall be submitted to the Building Commissioner,
who may issue a permit at his discretion and under such further
conditions in addition to those stated below as he sees fit to
impose." The " conditions stated below" give the allowable
unit stresses and other provisions foreign to our present purpose;
the discretionary conditions which the Commissioner imposes
at the present time are :

(1) That the plans before being submitted to him shall have
been approved by an expert engineer satisfactory to himself.

(2) That during the placing of concrete an inspector shall
be employed at the expense of the owner; the inspector must
be satisfactory to the Commissioner, and must report to the
Department of Buildings.

In regard to the expert engineers, the Commissioner reserves
to himself the right to pass upon them at any time. Objec-
tions have been raised to this arrangement on the ground that
the expense of examining the plans should be borne by the Build-
ing Department, and not by the owner (although it seems
proper that each owner should pay the expenses of his own plans).
The advisability of employing an expert in the department has
been considered, but so far without result.

In regard to the compulsory inspection, the same objec-
tions may be raised as in Cleveland, that really efficient inspec-
tion is not obtained in that manner, and that the owner meanwhile
is brought to believe that his work is efficiently inspected.

1 The authors are indebted to Mr. J. R. Worcester, M. Am. Soc. C. E.,
for information in regard to the Building Regulations in force in Boston.

202 REINFORCED CONCRETE BUILDINGS

The Building Regulations of Boston and Cleveland, just
cited, throw a very remarkable light upon prevailing conditions.
It is almost unbelievable that it should be necessary to actu-
ally force the owner into engaging adequately trained men to
plan and supervise the structure in which the owner, more than
any one else, is vitally interested. Undoubtedly, the efforts
of local building departments have succeeded in keeping the
standards of workmanship and design above a certain level,
even if that be low, but it must not be forgotten that the final
decision rests with the public generally and the building owners
in particular, and for that reason, the real problem before the
concrete engineer today is to reach and educate the public so
that better work is not only insisted upon but also paid for.

It would be very desirable if uniform regulations could
be made for methods of design and calculation, eventually in
the form of State Laws. Efforts toward standardization of
calculations have been made by the American Society of Civil
Engineers and others, and that such recommendations or regu-
lations are not impractical may be seen from their successful
operation in Prussia, Austria, France, etc. Owing, however,
to the great variation in available supplies of aggregate, the
allowable stresses must always remain a local issue.

Various influences are at work which greatly retard the
development of sound engineering. Certain concerns engaged
in the selling of reinforcement will furnish free plans showing
designs calculated to land the job rather than to give efficient
service. The method is objectionable when worked through
the medium of a small contractor, but much more so when a
so-called " architect" is made to act as a cat's-paw. The
architect (or engineer) who holds himself out as qualified to
design reinforced concrete work, and either has not, or does not
provide for the requisite skill, is guilty of deception, and obtains
his money under false pretenses. As a matter of fact, all our
best architects have competent engineers on their staff, or
engage the necessary talent when required, and the owner can
always obtain such services by simply insisting upon having
them.

Another objectionable practice has sprung from the indis-
criminate use of ''tables of design." The modern steel indus-
try would certainly be an impossibility without standard shapes,

THE ENGINEER 203

and here the structural steel tables in common use are the only
proper thing. But while a certain degree of standardization
in reinforced concrete construction is urgently important, it
is practically impossible to provide for the innumerable pos-
sibilities of design, at least at present. Moreover, while the
table itself may be a labor-saving device, it is likely to be used
most by those who are least conversant with the underlying
principles, leading to disastrous results.

Again in certain sections, particularly in the Middle West,
a class of contractors has been created whose slogan appears
to be: "get the job at any cost." No contractor can afford in
a lump sum contract to take work at less than actual cost to
him plus a reasonable profit, the cost to include overhead ex-
penses, depreciation, idleness of plant and staff, contingencies,
etc. For a while, a contracting business may be run in viola-
tion of these principles, but not for long. Of course, every job
on which the specifications are honestly and consistently en-
forced hastens the day of judgment for such concerns, and they
strongly resent anything that looks like supervision. These
concerns have injured not only themselves, but have succeeded
in lowering the general standard of workmanship by training
foremen and young engineers in sloppy and slovenly work.
When these abuses become too great emergency provisions
are in order, and the compulsory inspection paid for by the
owner under the supervision of the building department is one
way.

At the present time, there are reasons for believing that
reinforced concrete contracts should be let on the "cost plus
profit" basis. Such contracts protect the owner against pooled
bids and against extortionate charges for contingencies or profits.
It is evident that the structural steel contractor has no "con-
tingencies of manufacture," but only "contingencies of erec-
tion," while the reinforced concrete contractor has both.

In fact, if the contract be not awarded to the lowest bidder,
there is no good reason for taking bids, and if the owner has
so much confidence in any one bidder that he prefers him in
spite of his higher bid, he might as well trust him to the extent
of giving him the contract on the cost plus profit basis. We
are not here concerned in discussing the various types of con-
tracts possible under this system, as to whether the maximum

204 REINFORCED CONCRETE BUILDINGS

cost 'ought to be guaranteed or not, or whether the profit should
be a percentage of the actual cost or a certain stipulated sum.

It is believed that such contracts are usually given to con-
tractors having an engineering department in their organiza-
tion, and who are, as a matter of fact, "contracting engineers"
whether so called or not. We must remember that reinforced
concrete construction was first introduced, and has since mainly
been developed, by just such men or concerns. It is safe to
say that a very large portion, if not the largest portion, of all
reinforced concrete buildings of any consequence is erected
by " constructing engineers," who plan, design, and erect the
work from start to finish, frequently on the cost plus profit
basis. The owner should have these plans checked by a con-
sulting engineer and provide for adequate inspection of the work.
The position of the inspecting engineer is one that calls for
considerable tact, because he, as well as the contractor, are
virtually members of the same organization, viz., of the owners'
building staff.

Under the lump sum contract, the engineer's position is
radically different. He, and he alone, should prepare the gen-
eral and detail plans, with adequate specifications, and once the
contract is let, it becomes his duty to enforce the specifications
in letter and spirit, making himself as disagreeable as condi-
tions demand. Even if the specifications (or contract) give the
engineer the right to make necessary alterations, he should be
exceedingly careful not to waive any of the requirements by
commission or omission. Inspection of this kind is efficient
only when explicit and full specifications have been prepared,
but this does not mean that the specifications should be burden-
some or unfair to the contractor. It is no easy matter to write
good specifications for reinforced concrete work, and it requires
first of all full acquaintance with local conditions. There are
many places, for instance, where " clean," " sharp" sand can-
not be obtained locally, and the engineer must so word his
specifications that suitable sand is called for, and he must then
see that good sand is really used; not, as some engineers do,
call for clean, sharp sand, and then allow the use of sand that is
neither the one nor the other.

Attention is called to the difficulties encountered in making
monthly estimates for reinforced concrete buildings. The false-

THE ENGINEER 205

work enters only as machinery, tools, or other appliances, and
its full value should not at any time enter into the estimate,
but only a certain proportion of its value. This leaves consider-
able room for argument as to just what proportion to include;
the better and safer way is to have a clause in the contract
stating that a certain reasonable proportional amount of the
contract price must be paid: (1) when the footings are in; (2)
when the first floor has been concreted, etc., etc. It is very
much easier to arrange the amounts to be paid before the con-
tract is signed than after the work is under way.

It is now evident that whatever the position of the engineer,
- whether connected with owner, architect, or contractor,
he must possess certain qualifications of his own. First of all,
he must make himself felt as a useful factor in the community,
and not be satisfied with remaining a subordinate, and apparently
superfluous appendix. He alone has it in his hands to make the
industry advance or decline, and his essential function is, not
only to economize in the proper place, but to make the owners
see the folly of parsimony. He will have to overcome criticisms
of impracticability and extravagance, and this will be the more
difficult as he will rarely be brought face to face with the charges.

He must be an expert designer, not only of the usual ribbed
floors, arranged in the conventional cigar-box type of factory
building, but also of the more complicated types of flat floors,
of ribbed arches and other unusual forms for which reinforced
concrete is so well adapted and as yet so little used. Never-
theless his ability as a mathematician must not kill his ability
as a business man, for if he cannot get the work to exercise his
mathematics on, he will have scant use for them. He must be
fully posted on methods of erection not only how to do things,
but also how not to do them yet his knowledge must not make
him overbearing with the common foreman who "knows every-
thing about it," yet whose main asset is his ignorance.

Granting now that our engineer approaches to some extent
the ideal just outlined, he must also possess a certain amount
of skepticism in regard to precedents. Without question, there
are wide fields for investigation as yet open. We have referred
in an earlier chapter to the fallacy of too implicit faith in cement
testing. We have considered the impossibility of current ideas
of shear in reinforced concrete beams. There may be, and prob-

206 REINFORCED CONCRETE BUILDINGS

ably are, many others. Criticism of this kind is beneficial,
not only professionally, but sometimes financially as well, be-
cause sound criticism leads to improvements, and good improve-
ments are well worth while.

In order to gain material benefit from an improvement or
invention, it must be patented. Reinforced concrete men are
too prone to decry the value of patents generally, but this atti-
tude appears to be founded in ignorance. In order to avoid
infringement, the engineer must certainly be familiar with
patents and patent law; only in that way he can save the client
undue expense and trouble, and judge for himself of the value
of a new invention. The only circumstance saving many a
man from patent suits is that the patentee cannot afford the
expenses of court trial, which may run anywhere from $5,000
to $20,000 or more, and extend over many years. If there are
any good and substantial reasons for granting patents, the
engineering profession should recognize the existing conditions
and inform themselves, treating patent rights in the same man-
ner as they do other property; if no such reasons exist, the
engineers should use their influence in having the patent office
abolished. There is little likelihood that the latter alternative
will be followed, and patents should therefore be respected.
One way of ensuring the rights of the patentee would be to have
an injunction issued at once when proper evidence was pre-
sented to the court, and leave it for the infringer to prove the
patent invalid; as it is, the patentee practically has to prove the
validity of his patent before any injunction will be issued. It
would well pay the owner to see that his engineer is well posted
on the question of patent rights, for if infringement should
occur, the patentee will certainly look to the owner for reparation.

CHAPTER XIX
THE THEORY OF BEAMS AS ILLUSTRATED BY TESTS

The Extensibility of Concrete is not changed by the presence
of reinforcement. It was discovered in tests made at the Uni-
versity of Wisconsin in 1901-1903 that beams cured in water
and partially dried showed " watermarks" or fine dark lines
on the tension side under loads which would have fractured
non-reinforced pieces, and it was proved that these watermarks
indicate cracks. 1

In reinforced concrete beams, these cracks appear under
tension stresses in the steel of about 5,000 Ibs. per square inch,
and we are therefore not justified in calculating on any tensile
resistance in the concrete.

The Shear Resistance of Concrete is not affected by the pres-
ence of reinforcement. Prof. Morsch 2 made shear experiments
with cement mortar prisms 1" x 7" in section and found:

for mixture 1:3; 2 years old: 879,835, 1098 Ibs./sq. inch

average 937 Ibs./sq. inch

for mixture 1:4; 6 weeks old: 549,593, 441 Ibs./sq. inch

average 528 Ibs./sq. inch

Reinforced prisms of same mixture, size, and age as the last
series sheared under the following stresses: 550, 495, 528, 451,
473 Ibs. per square inch. It made little difference whether the
reinforcement was straight or bent. The final carrying capac-
ity of the. reinforced prisms was, however, much greater than
their apparent shear resistance, for after the concrete had sheared
it was still possible to increase the loads considerably. Professor
Morsch considers that this increase was due to the shear resist-
ance of the steel, which, mathematically speaking, was stressed
in shear as follows, when the final collapse took place:

1 Turneaure and Maurer: Principles of Reinforced Concrete Construc-
tion, 2d Edition, p. 42.

Subsequent tests by Bach (Zeitschrift des Vereines deutscher Inge-
nieure, Band 51, Nr. 26) have fully supported the Wisconsin tests.

2 Morsch: Concrete Steel Construction, p. 33 ff.

207

208 ' REINFORCED CONCRETE BUILDINGS

Specimen 1: 47,650 lbs./sq. in. straight bars only
" 2: 45,230 " " " " " "

3:55,050 " " " straight and bent bars

4: 47,080 " " " " " "

" 5: 50,350 " " " " " " "

The ultimate shear strength of the steel was only 47,790 Ibs. per
square inch, so where specimen 3 acquired its additional 17 per
cent, strength does not seem clear. Specimen 2 failed under a
load of 40 tons, "at which point a horizontal crack appeared at
the left end." This, we know, is an indication that the steel
is pulling out of the concrete, and it seems altogether likely
that the resistance really measured in these specimens was the
tensile resistance of the reinforcement, in accordance with
the theories advanced in Part II, Art. 57, of this book.

Various other tests have been made to determine the resist-
ance of concrete to pure shear. They generally confirm the
figures given directly above, but the results vary greatly owing
to the great difficulty in eliminating tensional stresses. In
practical construction, pure shear is rarely encountered in rein-
forced concrete beams.

The Function of the U-Bars. With the foregoing remarks in
mind we must admit that the U-bars cannot in any way influ-
ence the shear resistance of the concrete. If we consider the
U-bars as active in shear, their action cannot take place before
the shear resistance of the concrete is exhausted, and whatever
view we take of the stresses, the total shear resistance of the
beam is not the sum of that of the concrete, and that of the U-
bars (or other "shear" reinforcement). In this book, the
U-bars have been considered as (1) retarding the sliding of
the main tension reinforcement and (2) supplying the vertical
tension resistance caused by deviation from the equilibrium
curve of either the compression or the tension "chords." The
first proposition is easily investigated by test; the second is
closely related to problems connected with trussed rods and
kindred matters, and will be considered in that connection here
below.

The Stirrups Retard the Sliding of the main tension rods.
The "Commission du Ciment Arme" (1907) tested specimens
as shown in Figure 148 a, 6, and c. The specimens gave the fol-
lowing average sliding resistance per sq. inch of embedded sur-

THE THEORY OF BEAMS AS ILLUSTRATED BY TESTS 209

face, the first group having stirrups of flat iron, the second of
round iron:

6 months old

3 months old

a
b
c

159 Ib./sq. inch
214 Ib./sq. inch
281 Ib./sq. inch

125 Ib./sq. inch
252 Ib./sq. inch
284 Ib./sq. inch

or about the same values for the specimens with stirrups as for
rods centrally embedded in a block of concrete. This shows
the increasing importance of U-bars in beams with thin con-
crete covering on the rods, even in the case where U-bars are
not theoretically required.

FIGURE 148.

These tests show the gripping action exerted by the U-bar
on the rod, and explain in part the tendency of the beams to
crack at the U-bars, because the U-bars act as washers on the
rod, so that the concrete naturally would split immediately
behind such points.

Straight Reinforcement in T-Beams German Tests. In
the famous series of T-beams tested by Prof. Morsch, beams
I, II, and III were equipped with U-bars for one-half the length
only. The beams all failed at the end without U-bars under
loads fairly proportional with the width of the stem, showing
that the resistance of the stem was the deciding factor. It
makes little difference for our present purpose whether the fail-
ure was caused by actual shear or by the pulling out of the rein-
forcement ; Prof. Morsch appears to favor the former of these
alternatives, although in the two beams with narrow stems, the
concrete surrounding the reinforcement was split off, indicat-

210 REINFORCED CONCRETE BUILDINGS

ing that the concrete was not sufficient to resist the lateral expan-
sion, thus allowing the rods to slip. The resistance called into
action in this manner would be proportional with the thickness
of the stem.

These tests show conclusively that T-beams with straight
reinforcement only and without U-bars are not economical struc-
tures. As to the U-bars themselves, the tests show they are
beneficial, and Prof. Morsch further states:

" If the cause and the formation of the cracks in these three
beams are examined, it is established that the cracks first became
visible where the moment was greatest, and that with increase
of load more distant cracks appeared. On the end supplied with
stirrups, the cracks appeared to occur at the sections in which
the stirrups were located, since the concrete section was weak-
ened at those points." The same observation has been made
in other investigations.

These beams were tested with a uniformly distributed load
covering the entire span.

Bent Reinforcement in T-Beams German Tests. In con-
tinuation of the tests just described, Prof. Morsch investigated

Beam of the TRAJECTORY Type

FIGURE 149.

several beams with a combination of bent and straight bars.
Two distinct types were used, the bent bars being of either the
" trajectory" type (Figure 149) or of the "suspension" type
(Figure 150). In the table herewith, the principal details of

Beam of the SUSPENSION Type

FIGURE 150.

the arrangements are given (the letters T and S indicating the
type of bent reinforcement), as well as the ultimate load.

THE THEORY OF BEAMS AS ILLUSTRATED BY TESTS 211
TABULATED RESULTS OF PROF. MORSCH'S BEAM TESTS

Tension Rods

Beam

Type

B
No.

ent

Straight

fl o

U-bars

Type of
Loading

"S-S u 03

111 I

P>3 3

Diam.
in

No.

Diam.
in

m/m

IS -2

15 and 1

none

Uniform

42.0

15 and 1

full supply

load

37.8

-2

15 and 2 16

one end only

covering

31.0

entire span

Two

VII

16 and 1

full supply

concentrated

34.0

VIII

16 and 2 16

one end only

loads

23.4

16 and 2 16

one end only

at third

25.6

points

16 and 1

one end only

One

27.0

16 and 2 16

none

concentrated

26.0

XII

16 and 1

none

load

26.0

at center

The tests naturally divide themselves into three groups,
according to the manner of loading:

(1) Uniform load, beams IV, VI, and V. It is at once
apparent, by comparing beam IV (without U-bars) with beam
VI (with U-bars), that the influence of the U-bars is very slight,
if any, the difference in ultimate load being accounted for by
the fact that the ends of the straight bar in beam IV were hooked,
while those in beam VI had no hooks. Both of these beams were
of the trajectory type, and if we compare them with beam V of
the suspension type, the superiority of the trajectory type seems
clearly established. But we must not lose sight of the fact that
in the two first beams three of the four rods were bent up, while
in the latter, only two of the four rods were bent up. This
beam failed in the end without U-bars, and while therefore
this group does not prove the author's theories, as outlined in a
preceding chapter, it does not disprove them, and still leaves
the question open whether or not the bending of one additional
rod, or the proper use of U-bars, would not have changed the
results materially. It will be remembered that in a T-beam,
the straight reinforcement is effective only as reinforcement of
the stem, while the bent bars correspond to the flange; if the rods
are not so arranged, stirrups must be introduced to again bal-

212 REINFORCED CONCRETE BUILDINGS

ance the design, the size of the U-bars being in direct ratio to
the violation of the principle outlined.

(2) Two concentrated loads, beams VII, VIII, and IX.
Here again, beam VII of the trajectory type, with a full

supply of U-bars, is compared with two beams of the suspension
type, the two latter being without U-bars. Again the traject-
ory type seems superior to the suspension type, and again we
find the reason to be that in the trajectory beam, the proper
amount of rods have been bent up, while in the suspension type,
only two of the four rods have been bent, and no U-bars have
been introduced to overcome the deficiency.

It is rather interesting to note that the difference in width
of stem between beam VIII (10 cm.) and beam IX (14 cm.) affects
the ultimate strength but slightly, both beams being of the
suspension type.

In his discussion of these tests, Prof. Morsch has taken
occasion to criticize the suspension, or Hennebique, type. It
is to be regretted that the tests were not carried out so as to
have the same number of bent-up bars in both of the types
considered, in which case the suspension type would probably
have stood up as well as the trajectory beams. It is only fair
to note that the U-bars or stirrups have always been considered
as an essential part of the Hennebique system, and that such
tests as these, however valuable otherwise, give no indication
whatever as to the merits of this system.

(3) One concentrated load at center, beams X, XI, XII.

In this group, the two systems give the same carrying
capacity, owing undoubtedly to the fact that in no one of these
beams the reinforcement is arranged according to the equilib-
rium curve, while in no case U-bars have been introduced to
compensate for the deviation.

Bent Bars in T-Sections Author's Tests. The beam tests
just referred to were published by Prof. Morsch in "Deutsche
Bauzeitung," April 13, 1907. It occurred to the author of the
theory of this present volume that the description of the action
of the suspension rods was subject to doubt, for the reasons
outlined above, and that additional information might possibly
be gained by tests on beams with trussed rods only. The author
designed a series of nine test beams which were tested in the
winter 1907-1908 at Case School in Cleveland, in co-operation

THEORY OF BEAMS AS ILLUSTRATED BY TESTS 213

Type A
ams 1,4,7

TypeB-
ams 2,8.

>> c

r U

omag jo JlH 9U I 5"

^ ^H^

214 REINFORCED CONCRETE BUILDINGS

with Prof. F. H. Neff. It will be seen from Figure 151 that
these beams had no straight reinforcement, and that the sloping
stem terminated at the supports, so as to make the system one
of equilibrium under two concentrated loads. The results were
first published in the Engineering Record, August 22, 1908, from
which the following is an extract:

" Three different molds were made, types A, B, and C,
respectively, each one of which was used three times with a
different percentage of steel for reinforcement of the beam.
In this way, three beams of type A were made, one of which was
reinforced with 0.5 per cent., one with 0.75 per cent., and one
with 1.0 per cent. In the same way three beams B and three
beams C were made, reinforced as described, so that of the total
number of nine beams no two were alike in all respects, but any
one beam would have a corresponding one which was different in
one detail only. In this way, it would be possible to compare the
beams and find the exact effect of a certain change, which is a
safer way than to try to obtain absolute results from so few tests.

"All the beam swere provided with U-bars in one end only,
the object being to show that the stirrups were of no conse-
quence at all. The stirrups made no difference in the results
obtained, four of the nine beams failing in the end equipped
with U-bars.

"Two short cross bars were placed in the slab at the points
where the loads were applied, and three similar bars were placed
in the slab near the support. These bars were i-inch square
twisted bars. The main tension bars were 1-inch square twisted
Ransome bars. It was found that the elastic limit of these
bars averaged about 56,000 Ibs. per square inch, and their ulti-
mate breaking strength was 73,600 Ibs. per square inch.-

" The concrete was made quite wet and very carefully placed.
The mixture used was 1:2:3^, Lake Erie sand and Euclid bluestone
being used for the aggregates. The strength of the cubes was low,
as might be .expected with the aggregates used, and the average
of the 6-inch cubes in pounds per square inch was as follows:

Age, days 7 14 28 60

Strength, pounds 660 1,065 1,440 1,787

"The beams were all tested when sixty days old. In the
table here below the results are given, and this table, together

THE THEORY OF BEAMS AS ILLUSTRATED BY TESTS 215

with the diagrams of the beams, should give all the information
needed. Attention is called to the ways of supporting beams 5
and 6. While all the other beams are supported at the point
where the sloping stem begins, these two beams are supported
further out from the stem, making the overhang shorter for
them than for the similar beams of same type.

"As to the column headings used in the table, the percen-
tage of reinforcement is calculated with reference to the ' enclos-
ing rectangle' proposed by Professor Talbot. Under 'lever' the
distance from point of support to point of application of the
load is given, while 'overhang' means the length of the pro-
jecting end beyond the support.

"The bending moment given in this table is found by mul-
tiplying the 'lever' by one-half of the ultimate load, disre-
garding entirely the weight of the beam itself. The lever arm
of the internal stresses is assumed to be 0.85 times the distance
from the top fiber to the center of the steel, which distance is
approximately 9 in., giving a lever arm of 7.65 in. This, of
course, is not quite correct, as the position of the neutral axis
varies with the percentage of steel and the coefficient of elas-
ticity, which latter again depends upon the stress on the
concrete. It is, however, sufficiently accurate considering the un-
avoidable variations in the position of the steel bars and in the
elastic properties of the concrete, and the 'total stress in the
steel' may therefore be found by dividing the bending moment
by 7.65, giving the values shown in the table as well as the
stress in the steel per square inch of its cross-section.

RESULTS OF TESTS AT CASE SCHOOL.

Beam. Type. Per cent. Lever. Overhang. Ultimate

load.

0.5

0.75

1.00

30 in
26 in
22 in
30 in
30 in
30 in
30 in
26 in
22 in

10 in.
14 in.
18 in.
10 in.
10 in.
10 in.
10 in.
14 in.
18 in.

12,500
16,000
27,800
11,900
16,200
16,000
13,950
22,000
28,900

Bending

- Stress in steel

moment.

Total.

Per sq. in.

187,500

24,500

49,000

208,000

27,200

54,400

305,800

39,900

79,800

178,500

23,300

31,000

243,000

31,800

42,400

240,000

31,400

41,800

209,250

27,300

286,000

37,400

317,900

41,600

"Eef erring now to the several photographs of the beams
after failure, it will be noticed that the failures are of uniform
nature. Comparing the figures given in the table above, it
will be seen that the ultimate load varies greatly as well as the
total stress and the stress per square inch. If the failure has

216

REINFORCED CONCRETE BUILDINGS

Beam 1

Beam 2

Beam 4

FIGURE 152.

FIGURE 153.

FIGURE 154.

Beam 5 FIGURE 155.

THE CASE SCHOOL BEAMS AFTER TESTING.

TtiE THEORY OF BEAMS AS ILLUSTRATED BY TESTS 217

a common cause in all these beams it cannot be due to either
tension or compression in the usual sense of the word. It may
also be assumed that shear had little to do with the failure.

Beam 6

FIGURE 156.

Beam 8

FIGURE 157.

Beam 9 FIGURE 158.

THE CASE SCHOOL BEAMS AFTER TESTING.

On account of the trussed form of the beams, the steel follows
the curve of equilibrium of the external forces acting upon the
beam, and the only stresses possible are tension in the steel and
compression in the concrete.

218 REINFORCED CONCRETE BUILDINGS

"This is also evident from the behavior of the beams under
load. The cracks started on the tension side and opened slowly
with increasing load, at the same time becoming longer, until
finally the compressive area left above the top of the crack became
too small to carry the stress on it and crushed. A shear crack
cannot grow in this manner. It is well known that the maxi-
mum shear stress does not occur at any fiber near the extreme
top or bottom of a beam. Therefore, when the crack extends
up into the stem and reaches the neutral axis, the shear resist-
ance of the beam is practically exhausted.

" The beams also made it evident in other ways that no ver-
tical shear was active. In some cases the beams had received
a vertical crack in handling, the crack being located about 3 in.
inside the support, and extending clear through the concrete.
At first, it was believed that these beams would not give a fair
test, and it was taken under consideration to leave these beams
out. It proved, however, that the crack closed up as soon as
the load was put on, and after the load was increased to a cer-
tain amount, the cracks were hardly visible, while the final
failure took place some distance from the injured section. If
there had been any vertical shear acting on the beam, the ulti-
mate load would have reached a comparatively small value only,
and in all probability the injured section would have sheared
off at once.

"The tension in the steel must be constant from end to end
of the beam between the supports. The steel would have a
tendency to pull out of the overhanging ends with a force equal
to the total pull in the steel, which is the same near the supports
as at the center of the beam. The overhanging ends furnish
the necessary anchorage for the bars on account of the grip
of the concrete around the bars, which increases with the com-
pression in the concrete, and, therefore, also with the load, the
horizontal cross-bars giving the required horizontal restraint of
the concrete to produce the desired effect. The numerical
value of the length of the anchorage may therefore be expressed
in figures by simply dividing the length of the overhang into
the total pull on the steel, the quotient giving the value of the
bond in pounds per lineal inch of embedment, regardless of the
amount of steel. This figure is given in the accompanying
table, the length of the anchorage being the length of the over-

THE THEORY OF BEAMS AS ILLUSTRATED BY TESTS 219

hang, and disregarding the extra length of the hook at the
ends of the bars. Beams 5 and 6 are not included in the table,
as these beams had an overhang of only 10 in., leaving a hori-
zontal space inside the support, and this, of course, makes it
impossible to compare these two beams directly with the rest.

VALUES OF BOND OBTAINED

Total pull Bond

Beam Type Overhang in steel per lin. in.

1 A 10 in. 24,500 2,450

2 B 14 in. 27,200 1,940

3 C 18 in. 39,900 2,230

4 A 10 in. 23,300 2,330

7 A 10 in. 27,300 2,730

8 B 14 in. 37,400 2,670

9 C 18 in. 41,600 2,310

"This table, it is believed, is remarkable when the uniform-
ity of the results is considered. The beams tested here had
reinforcement varying from ^ of 1 per cent, to 1 per cent.,
spans varying from 74 to 80 in., and tension stresses in steel
varying from 27,300 to 79,800 Ibs. per square inch. It seems safe
to say that these beams all failed by sliding of the steel.

" So far, no attention has been paid to beams 5 and 6. The
overhang for these beams was 10 in. in each case, the slab con-
tinuing for a distance inside the supports. The bond stress
developed in the overhang, if figured as for beams above, be-
comes 3,180 and 3,140 Ibs. per linear inch, or quite high when
compared with the results of the table above. Remembering,
however, that the straight portion of the bar is continued inside
the supports for a distance of 4 and 8 in., respectively, the bond,
if distributed over the total distance of 14 in. for No. 5 and 18
in. for No. 6, becomes:

Total stress Bond
Beam Type Overhang in steel per lin. in.

5 B 10" + 4" = 14" 31,800 2,270

6 C 10" + 8" = 18" 31,400 1,745

" If any importance can be given these two isolated results,
they would show that the bond inside the support is quite as
effective as that outside the support, but for a short distance
only, and that its value decreases rapidly with the distance
inside the support."

220 REINFORCED CONCRETE BUILDINGS

The lessons to be drawn from these tests are:

(1) That, with the arrangement used, the presence or ab-
sence of U-bars does not influence the strength of the beam.

(2) That "shear," properly understood, does not exist in
beams of this kind.

(3) That, with proper arrangement of the end supports
and of the anchorage, such beams will not fail until the com-
pressive strength of the concrete, or the tensile strength of the
steel, is exhausted.

(4) That such beams are rational structures capable of prac-
tical and economical use.

(5) That the sliding resistance of the steel does not depend
upon the number or size of the individual rods, but only upon
the anchorage of the group of rods, the length of embedment
being much more important than the diameter of either each
rod or of the group of rods.

Effect of Joint between Slab and Stem, Tests by Professor
Johnson. In connection with the introduction of the Ransome
Unit System (p. 162 ff.) in Boston, a series of very interesting
tests were made on T-beams of both the monolithic and unit
types, reinforced with straight bars only, and with both straight
and bent bars. All the beams had U-bars. In the "Unit"
beams, the slab was cast from four to nine days later than the
stem. A total of twenty-eight beams were prepared, of which
eleven have so far been tested, the balance being held for a
longer-time test. 1 The beams were all reinforced with Ran-
some steel, that is, cold-twisted squares. The U-bars were
round except in Type C, where square twisted U-bars had been
used.

Type A, Beams 1, 2, 4, 5, 9, 10, and 12. See Figure 159.

1 The authors are indebted to Prof. L. J. Johnson, M. Am. Soc. C. E., for
the following data, and for permission to publish the same. The beams were
designed by Prof. Johnson, by Mr. J. R. Worcester, M. Am. Soc. C. E., Consult-
ing Engineer, by Mr. J. R. Nichols, Jun., Am. Soc. C. E., by the Concrete
Engineering Co. of Boston, and the Ransome Engineering Co. of New York,
each having designed one series of beams or contributed to the design by
suggestions. Professor Johnson, who made the tests on the testing machine
in the Harvard University laboratory, expects to publish in due season a
complete report of both this series and of the long-time tests. The authors
of this present volume, eye-witnesses of these tests, are solely responsible for
conclusions reached herein.

THE THEORY OF BEAMS AS ILLUSTRATED BY TESTS 221

In the Unit beams, the top of the stem was either left fairly
smooth, as it would be in usual every-day practice, or corru-

Toggle for lifting beam

8-8

Rods

"13-"^

Bars y \ t 1 4^ ^ 6 * Stirrups 5^

Straight'

FIGURE 159.

gated as shown in Figure 160. The ends of the stem rested in
previously prepared seats (Figure 161), and the joints were

FIGURE 160.

sealed with grout to ensure a similar action as obtained in actual
construction, where the Unit beam rests in a pocket in the gir-
der. Nine days after the casting of the stem, the slab
was put on, while in the monolithic beam, the entire
amount of concrete was, of course, deposited in the
forms in one operation.

Beam No. 5 was a Unit beam, with the top of the
stem corrugated. The age of the stem was forty-
five days, that of the slab thirty-five days. At a
total load of 12,000 Ibs. the first tension crack ap-
peared near the middle of the span. Inclined cracks
became evident near the ends under a load of 23,000
Ibs.; the ultimate load was 41,000 Ibs., when failure
occurred, through compression of the slab between
the loads, and slipping of the straight tension bars
(see beam No. 12 below).

Beam No. 4 was a Unit beam, the top of the
stem being fairly smooth; that is, no attempt had
been made toward getting a particularly .rough sur-
face. The age of the stem was forty-five days, of the slab

FIGURE 161.

222

REINFORCED CONCRETE BUILDINGS

thirty-six days; the first crack was observed under a load of
20,000 Ibs.; ultimate failure took place under 48,000 Ibs. in
precisely the same manner as in No. 5.

Beam No. 1 was exactly similar, except that slab and stem
were both one day older than in No. 4; the ultimate load was
49,400 Ibs., and the beam failed in the same manner as the fore-
going.

Beam No. 2 was of the same age and detail as No. 1 ; the first
crack was observed at 10,000 Ibs. loading; the ultimate failure
occurred in the same manner as above under 54,300 Ibs. total.
The higher load on this beam is perhaps due in some measure
to the fact that the rocker-supports for the beam came to a
bearing, making possible some horizontal thrust on the beam.

Beam No. 12 was monolithic, forty-one days old; of same
design as the foregoing Unit beams, except that the fillet between
stem and slab was slightly reduced (see Figure 162). The first

FIGURE 162.

crack occurred at 4,000 Ibs., the ultimate load was 45,800 Ibs.,
and failure occurred through a slip of the straight reinforce-
ment, causing the sudden collapse of the left end.

Beam No. 9 was of the same general design, cast in one
piece, and forty-one days old. The first crack occurred at
3,000 Ibs.; the beam failed suddenly at 50,000 Ibs. by slipping
of the rods at the right end.

Beam No. 10 was also a monolith forty-one days old, show-
ing a tension crack at 6,000 Ibs., with ultimate failure at
52,400 Ibs. from a combination of initial sliding of the tension
rods with compression at the center.

It will be seen from these data that the Unit beams stood up
as well under the load as the monolithic beam, so that the joint
between slab and stem was perfectly adequate, whether cor-
rugated or plain. The general behavior of all these beams up
to the point of failure was so much the same that no one, from

THE THEORY OF BEAMS AS ILLUSTRATED BY TESTS 223

observation of the beams in the machine, could have pointed out
which beams were unit and which monolithic. In fact, they all
failed in the customary manner, exhibiting the usual inclined
and vertical cracks, and no sliding was noticeable between slab
and stem, although carefully looked for.

Type B, Beams 25 and 27. See Figure 163.

2- 3 /- Bars'T t I Ho Stirrups

2-K" Straight

Ho Stirrups 5K
FIGURE 163.

The beams of Type B were built exactly as the beams of
Type A, except that the tension rods had been reversed, being
2|" bars bent and 2^" bars straight. This reinforcement
would, under the theories- advanced in this book, be more effi-
cient, and the U-bars were therefore reduced from T V' round
stock in Type A to i 3 s " round stock in Type B, thus having about
one-third of the area of the former.

Beam No. 25 of Unit construction had a stem twenty-nine
days old and a slab twenty-five days old; the first crack was
observed under a load of 9,000 Ibs., and ultimate failure took
place under simultaneous compression of the slab and of the
side of the stem, at the point where the tension rod was bent,
under a load of 42,500 Ibs. (See Figure 164.)

2 - Straight

FIGURE 164.

Beam 27

Beam 25

Beam No. 27 was monolithic, of same design, and twenty-
seven days old. The first crack was seen at 10,000 Ibs. ; while
the ultimate load was 47,500 Ibs. Also in this case was com-
pression in both slab and stem evident as shown in Figure 164.

224

REINFORCED CONCRETE BUILDINGS

It is interesting that these two beams carried practically
as much load as the older beams of Type A, in spite of the great
reduction in the weight of the U-bars. The explanation is to
be found in the theory set forth in Chapter VII of this book,
where the relation between the bent bars and the U-bars has
been considered at length; in fact, the design of beams 25
and 27 was made to prove, or disprove, these theories as far as
possible.

Type C, Beams 13 and 20. See Figure 165.

i /i

1 ,,0-10 ,,-J

'oggle f orjif ting beam k-8 >r< 8 >i

9_ S/'

1
-f

/8-X Rods

_^.

r T"~T~7~r~7~ 1

i ! i ! , ! ! !

V f7

J.-L_L_L_L

_o_L 1 X i J J J J

._ 1

-L/

V/ ///\

V 2-1"

y///k

" Stirrups *'Y

*Y///A

FIGURE 165.

In designing these beams, Professor Johnson had endeavored
to secure a high strength in compression and tension. The
special feature was the absence of bent bars so that the stresses in
the stem and in the joint between slab and stem were especially
severe under the common theory of shear. The beams rested in
concrete supports shown in Figure 166.

Beam No. 20 was of the Unit type with corrugated top
(Figure 167). The stem was forty-two days old, the slab thirty-

1 i "^ i 1 1

_/ \ / \ / ' \ / j

i Wj

[ ! !

FIGURE 166.

FIGURE 167.

two days old. Tension cracks developed in the usual manner,
beginning under a load of 10,000 Ibs. ; at 22,000 Ibs. small diag-
onal cracks appeared. At 50,000 Ibs. it was noticed that the
visible end of the curved tension rods began to slide, and the

THE THEORY OF BEAMS AS ILLUSTRATED BY TESTS 225

ultimate failure occurred under a load of 55,600 Ibs., when
the compression area was crushed.

Beam No. 13 was again of the Unit type, with smooth top of
stem, which was forty-three days old; the slab was thirty-four
days old. The first tension craok occurred at 9,000 Ibs., and
the cracks then developed in the usual manner. Failure took
place at 50,000 Ibs., when the adhesion between concrete and
steel was broken; the rods began to pull through, and the slab
was crushed at the center.

The analysis of the stresses follows:
Types A and B

By reference to formula (18), page 38, we have

2 = 1; T = 4"; H = 9"; D = 13"; V = ~ X 1.62 = 2

hence

S 82.0

and

^ = 7^M = ^ = - 436;1 -? = - 8 ^

Now, the bending moment is f L.42 = 21 L, and the arm of
internal stresses approximately

.855 X 13 = 11.1 inches, hence the pull in the steel
s = jj-L = 1.89Llbs.

The beams had 1.62 square inches of tension steel, hence the
unit tension on steel:

S L89 T 1 17 T
= ' L = 1 ' 17 ' L

and the unit compression on the concrete

Type C

| = i ; T = 5"; H = 6"; D = 11"; V = 15 - ^ = 2.5;

226 REINFORCED CONCRETE BUILDINGS

hence

~ = 15.9
and

x =

1 +

15.9
15

- .485; 1 - l -x = .838

Again, the bending moment is \ -L-30 = 15 L, and the arm of
internal stresses approximately

11 X .838 = 9.2", hence the pull in the steel is

The beams had 2.0 square inches of steel, hence the unit tension
on steel

and the unit compression on the concrete

It is evident that these calculations do not give the true
stresses existing at rupture, because r is not equal to 15 at that
time, and the assumption of plane sections probably does not
hold good. For the sake of comparison, however, they may be
useful. The results are indicated in the table. The testing

Corresponding

Age (days)

calculated stresses

"SI

Ultimate

165 sq. in.

How made

Load

Ibs.

Stem

Slab

Unit, Smooth Top

49,400

3060

57,800

Unit, Smooth Top

54,300

3370

63,500

Unit, Smooth Top

48,000

2980

56,200

Unit, Corrugated Top

41,000

2540

48,000

Monolith

50,000

3100

58,500

Monolith

52,400

3240

61,200

Monolith

45,800

2840

53,600

B
B

25
27

Unit, Smooth Top
Monolith

29
27

42,500
47,500

2640
2950

49,700
55,600

Unit, Smooth Top

50,000

2560

40,700

Unit, Corrugated Top

55,600

2840

45,300

Harvard Test Beams. Summary of Results obtained

THE THEORY OF BEAMS AS ILLUSTRATED BY TESTS 227

machine was equipped with means for registering the deflec-
tions automatically; the diagrams are shown in Figures 168,

FIGURE 168.

169, and 170. Generally speaking, there is little difference in
the deflection of the Unit and monolithic beams.

A number of interesting observations were made during
these tests. First, the feasibility of the Unit beam was estab-
lished beyond doubt, contrary to what many engineers would

Deflection

40000

^25

^<^

30000

20000

1.0*

0.9"

0.8 "

0.7"

0.6"

0.5"

0.4"

0.3"

0.2"

0.1" \

FIGURE 169.

probably have expected. In fact, many building regulations
throughout the country specify positively that the beam and its
superimposed slab must be concreted in one continuous opera-

228

REINFORCED CONCRETE BUILDINGS

tion. Where improperly designed, or otherwise inadequate,
U-bars are used, this rule is undoubtedly highly beneficial, but
where proper U-bars are used, the rule is wholly unnecessary.
The progress report of the special committee of the American
Society of Civil Engineers recommends that the slab be con-
sidered effective in compression when " proper bond" is pro-
vided between slab and stem; it will be appreciated that this
is a much more consistent requirement, although somewhat
indefinite. The beams tested so far have shown that the bond
provided was adequate, whether the more elaborate method of

Deflection 0.9" 0.8"

50000

30000

0.6'

0.5'

FIGURE 170.

corrugating the top of the stem was used, or whether the top of
the stem was simply left as it was upon completion. It would
be very interesting to learn what would happen when the bond
was " inadequate," and just where the limit may be found, and
in this particular the present tests furnish no information, as
the bond remained intact in all cases. See Figures 171 and 172,
showing the Unit Beam No. 25.

In the second place, these tests confirm in a remarkable
degree the theories set forth by the author in Chapter VII in
regard to the action of U-bars.

Compression failures of the stem were observed in beams 25
and 27; these are shown in Figure 164 and in Figures 171-174.
It was observed that the compression failure of the stem was
on the same side as the corresponding bent bar, the two bent
bars being each near the opposite face of the beam ; in beam 27

THE THEORY OF BEAMS AS ILLUSTRATED BY TESTS 229

FIGURE 171. HARVARD BEAM No. 25.

The black lines are ink marks indicating the principal cracks

Photo by Mr. J. R. Nichols, Jr., Am. Soc. C. E.

FIGURE 172. A CLOSER VIEW OF BEAM No. 25, SHOWING CRUSHING
OF THE CONCRETE AT THE ROD.

This beam was a Unit beam, it will be noticed that there was no indication
of slipping between stem and slab.

Photo by Mr. J. R. Nichols, Jr., Am. Soc. C. E.

230

REINFORCED CONCRETE BUILDINGS

FIGURE 173. BEAM No. 27, SHOWING PRINCIPAL CRACKS AT LEFT END,
AND THE CRUSHING OF THE CONCRETE AT THE ROD.

Photo by Mr. J. R. Nichols, Jr., Am. Soc. C. E.

FIGURE 174. CLOSER VIEW OF BEAM No. 27.

Photo by Mr. J. R. Nichols, Jr., Am. Soc. C. E.

THE THEORY OF BEAMS AS ILLUSTRATED BY TESTS 231

crushing took place at both bent bars, one spot on each side,
but in different locations, corresponding to the position of the
curves in the bars. It is self-evident that this upward pres-
sure of the rod must be resisted by an equal downward pres-
sure (from the load) thus dissolving the beam into a number of
well-defined compressive zones in a manner very different from
what takes place in a " solid" homogeneous beam. The same
observation was made by Prof. Morsch in regard to his test
beam No. VI.

Third, a deep, gaping crack was observed in the top of beam
No. 20 (Figure 175), near the support. The explanation of this

FIGURE 175.

crack may be found in the distribution of internal stresses
indicated in the drawing, the horizontal arrow at the steel indi-
cating the pulling of the steel, the inclined arrow indicating the
sum of the compressive forces in the concrete. It will be noted
that if these two do not intersect on the vertical line of the
reaction, a " re verse" bending moment is created at the end
which would cause just such a crack. Here again we have a
fact showing that a reinforced concrete beam cannot be consid-
ered as a " solid" beam, in which such stresses' are impossible.
Considering the beam as a truss, we see at once that the crack
comes outside the "end panel," and so would have no influence
on the load-carrying capacity.

In addition, it must be admitted that "shear," so called,
would have caused the instantaneous collapse of a beam with
such a crack. As an actual matter of fact, this beam, with the
gaping crack in the top, carried a total load of 55,600 Ibs., or
more than any other beam of the entire series. The stem was
perforated with inclined and vertical cracks so that the only
portions of the beam which could actually carry some shear
were the main tension rods. This proposition has been consid-
ered above and cannot be maintained. The truth is that there

232

REINFORCED CONCRETE BUILDINGS

was no active shear in this beam, the system consisting approx-
imately of members as shown in Figure 176. 1

Fourth, it was established that the quarter turn given the
straight tension rods at the ends was not sufficient to develop
the desired amount of sliding resistance. Thus, beams 9 and 12
failed suddenly by the entire separation of the lower rods from
the concrete, while the beams of Type C (13 and 20) showed a
sliding of from i" to f " (in these beams, the ends of the rods

I I

FIGURE 176.

could easily be observed by breaking away a thin shell of con-
crete). The behavior of the balance of the beams, and especially
inspection of the deflection diagrams, makes it, however, appar-
ent that only a very small additional margin of sliding resistance
was required in order to prevent the sudden collapse. Without
doubt, the large turn of the upper bar might profitably have
terminated at its lowest point, as the last fourth of the circle
materially weakened the concrete along the lines of cleavage

1 The authors are aware of the fact that other observations were made
during the testing of the Harvard series which strongly support the theory
advanced in Chapter VII of this volume. We are, however, requested to
withhold this matter from publication at the present time, and we must
refer to the later report to be published by Prof. Johnson.

INDEX

Accidents, 191

Acid, carbonic, for hardening con-
crete, 12

hydrochloric, for joining con-
crete, 10

joint, 10

joint, specifications for, 198
Adhesion, 54

Allowable stresses, 52, 128
Alum, 190

Arches, allowable stresses in, 131
Assumptions, homogeneity, 51

in general, 52

tensile resistance of concrete
disregarded, 104

Basic inventions, 18

Beam formulas, 70, 71, 75, 88

Belt course, Ransome patent, 13

reinforcement of, 153
Bending, 66

combined with compression, 114
Board marks, 176
Brushing, 179

Cement, 137

Chimneys, approximate formula for,
115

Chloride of calcium, 2, 190
of sodium, 2, 190

Clay, effect on concrete, 12

Coil joint, concrete to concrete, 9
for joining rods, 15

Columns, allowable stresses, 130
hooped, 58
least diameter, 62
repairs of defective, 189
strengthening existing, 64

Concrete, dry or wet mixture, 16
mixing and placing, 149
shrinkage and swelling, 126
specifications, 196

Conflagrations, 185

Continuity of beams, 109

Core boxes, permanent, 15

Corners chamfered, 176

Cracks, in cement finish, 181
in slabs and beams, 127
in tile-concrete floors, 182
repairs of, 188
structural significance, 92

Delayed placing of concrete, 7
Design, general remarks, 153

Earthquake, San Francisco, 6
Embedment, required length of, 55
Expansion joint, 2, 128

Facing of concrete, 177

see also Veneer
Factor of safety, 129, 132
Falsework, general design, 156

instructions, 196

standardization, 15
Finish, cement, general, 181

instructions, 198
Fireproofing, 183

effect of salt, 16
Floor coverings, cement finish, 181

wood, 154
Footings, formulas, 116

foundations, 171
Forms. See Falsework
Frost protection, 151

effect of salt, 10

233

234

INDEX

Hooked ends of reinforcement, 55
Hooped columns, 58

Inertia, moment of, 112
Injurious agencies, 17
Illuminating panels, 8

Joining new concrete to old, 9, 198

Laitance, 149, 181, 189

Lateral expansion, 57

Lime, slacked lime in concrete, 12

Mixing and placing of concrete, 149
Molds. See Falsework
Monolithic construction, 156

Notations used in bending theory, 66
Overmixing of concrete, 7, 8

Patentees, Alsip, 31
Aspdin, 19
Basset, 33
Bissell, 35
Brannon, 23
Bruner, 39
Bunnet, 21
Cheney, 32
Coddington, 24
Coignet, 22, 30, 31
Considere, 46
Cornell, 28
Cottancin, 40
Cubbins, 36
De Man, 41
Dennet, 21
Edwards, 26
Emerson, 33
Emmens, 24
Fowler, 29
Gedge, 22
Gilbert, 30
Golding, 36
Gustavino, 38
Hallberg, 44
Henderson, 31
Hennebique, 42

Patentees, Hyatt, 23, 24, 25, 26, 33,
34, 35

Jackson, 32, 36, 37, 38

Johnson, 22, 29, 40

Kahn, 45

Knight, 28

Lambot, 27

Lish, 24

Lythgoe, 22

McCarthy, 40

Matrai, 43

Matthews, 33

Melan, 40

Middleton, 28

Monier, 22, 27, 38

Parker, 19

Parkes, 22

Parmley, 45

Rabitz, 39

Ranger, 21

Ransome, E. L., 2, 3, 5, 8, 9,
10, 13, 15, 16

Ransome, Fk., 22

Shaler, 43

Sisson, 32

Smith, 31

Stempel, 39

Stephens, 28

Stevenson, 28

Summer, 28

Tall, 23

Thacher, 43, 44

Thornton, 22

Turner, 24

Visintini, 45

von Emperger, 40

Waite, 41

Wayss, 44

Weber, 46

Wetmore, 32

Wilkinson, 21

Williams, 30

Wilson, 39

Wood, 30
Wyckoff, 28
Piling, 171

cost of, 174
Plastering, 176

INDEX

235

Plates, concrete, 118, 120
steel base, 118, 171

Reinforced concrete, defined, 51

elements of invention, 18
Reinforcement, details of, 105

double, 114

circular, 118

kinds of, 145, 147

requirements, 147

placing, 159

Repairs to buildings, 188
Rolling of floors, 8

instructions, 196
Rubbing of surfaces, 179

Salt, effect on concrete, 10, 184

instructions for using, 196
Sand, 143
Sidewalk lights, 8
Silicate of lime, 2

of potash, 190

of soda, 2, 190
Slab formulas, 79
Slag, aggregate, 145

for making joints, 9
Sliding, of concrete upon concrete,
103

of reinforcement, 101