Concrete Structures
Concrete Structures
Concrete Structures
Concrete
Structures
Second Edition
Concrete Structures
Mehdi Setareh Robert Darvas
Concrete Structures
Second Edition
Mehdi Setareh Robert Darvas
College of Architecture College of Architecture
and Urban Studies and Urban Planning
Virginia Tech University of Michigan
Blacksburg, VA, USA Ann Arbor, MI, USA
Since the publication of the rst edition, the requirements for the design of concrete
structures have gone through a number of changes as reected in the American
Concrete Institute, ACI 318 publication. The latest revision of the Code was
published in 2014 and represents the state-of-the-art of the current knowledge in
concrete and reinforced concrete design. This revision, ACI 318-14, forms the basis
of the second edition of this textbook.
The book retains the features that made it well received by students, instructors,
and practitioners alike. The popular step-by-step approach of problem solving,
augmented by owcharts and supported by numerical solutions, clearly describes
the processes that need to be followed to provide safe and economical designs of
reinforced concrete components. The self-experiments included at the end of the
chapters help students better understand the behavior of concrete structures through
the construction and testing of scaled models.
To make the book more useful to students in Construction Engineering pro-
grams, a new chapter (Chapter 8) on formworks for monolithic concrete construc-
tion has been added. This chapter covers the fundamentals of formwork and shoring
design, and detailed step-by-step solutions of numerical problems along with
mathematical formulae and tables to help students and practitioners to design
these temporary structures. In addition, to provide more visual clarications of
the topics discussed in the book, a new appendix (Appendix B) is added, which
includes color images of various stages of concrete construction and completed
buildings.
We gratefully acknowledge the support of the following individuals and orga-
nizations by providing images that are used in the book: Professor Jack Davis, Dean
of Virginia Tech College of Architecture and Urban Studies, Ms. Kathe Hooper
from the American Society for Testing and Materials, Mr. Charles James from the
National Information Service for Earthquake Engineering, Ms. Angela Matthews
from the American Concrete Institute, Ms. Gwen Wang from the Portland Cement
Association, and Mr. Doug Peters, PE, President of Christman Constructors, Inc.
vii
viii Preface to the Second Edition
We also owe special thanks to the following individuals: Mr. Nadir Makhlouf, a
partner of Robert Darvas Associates, PC, for his great help in preparing the graphics
work for Chapter 8, Mr. Xiaoyao Wang, Mr. Hasheem Halim, and Mr. Sriram
Sankaranarayanan, students at Virginia Tech schools of architecture and engineer-
ing, for their efforts on updating the book and the solutions manual.
The authors would like to express their thanks to Springer International Pub-
lishing AG Switzerland, in particular Mr. Michael Luby, senior publishing editor,
and Mr. Brian Halm, project coordinator for helping us bring this second edition to
publication.
ix
x Preface to the First Edition
Finally, we wish to thank the Pearson Education editorial and production staff
for their support and assistance. Many thanks to Bret Workman, who did a great job
with text editing. In particular, the assistance of Penny Walker from Techbooks is
greatly appreciated.
xiii
xiv Contents
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 663
Chapter 1
Reinforced Concrete Technology
1.1 Introduction
Design
architect/structural
engineer
Contractor for Other trades whose
concrete work items must be in place
prior to concrete
Design drawings, placement (conduits,
layout and size of pipe-sleeves, etc.)
all elements, reinforcing Shoring and formwork
requirements (schedules),
and specifications
Supplier of concrete
Reinforcing is placed
and secured in place.
Manufacturer of
reinforcing Proposed mix for each
class of concrete to
Reinforcing is checked and satisfy specifications
approved (designer or
Shop drawings; all bars testing agency).
are detailed as to length,
bends, and location.
Review and approval
by designer
Concrete is delivered and
conveyed to the location
Review and approval of placement.
by designer (last line
Approved mix is
of defense for designers)
produced and
delivered in quantities
Concrete is placed, requested by concrete
consolidated and finished. contractor.
Bars are cut and bent
per approved shop
drawings and tagged.
Curing Testing agency
Finish of hardened
concrete per design
requirements (leaving as
cast, paint, cover, sand-
blast, bush hammer, etc.)
The rules and regulations governing any construction in the United States are set by
different model codes. The rules of a model code become the law in a given
municipality when the legislative body (state legislature, city council, etc.) adopts
them as the governing standard for building construction under its jurisdiction.
There are many model codes, among them the Uniform Building Code (UBC) and
the International Building Code (IBC). The UBC is traditionally used in the western
states of the United States. The IBC rst appeared in the year 2000 and is a joint
effort of the BOCA (Building Ofcials and Code Administrators), the ICBO
1.3 Concrete Ingredients 3
Thus the four ingredients of concrete (see Figure 1.2) are (1) cement (i.e., the
binder); (2) ne and coarse aggregates, which ll the bulk of the volume; (3) water
and air; and (4) admixtures, which are used to impart certain desirable properties.
These ingredients, carefully proportioned, are combined in a mixer.
Mankind has used natural cements since ancient times. The magnicent stone
structures built by the Romans all used nely ground cementitious materials
(pozzolans) in the mortar. Other materials may be used to bind aggregate particles
together (asphalts are used in making asphaltic concrete for road construction), but
in the making of structural concrete, hydraulic cements are used without exception.
Hydraulic cements harden by reacting with water. The most important hydraulic
cement is the one rst made by an English mason, Joseph Aspdin, who received an
English patent in 1824 for the composition and the process. He called it portland
cement because the concrete made with the cement had the color of natural
limestone quarried on the Portland peninsula. Portland cement is a ne powdery
material, composed mainly of calcium silicates and aluminum silicates. The mate-
rials needed to make cement are found in virtually every part of the world:
1. Limestone, which provides calcium oxide (CaO)
2. Clays, shales, and so on, which supply silicon dioxide (SiO2) and aluminum
trioxide (Al2O3)
1.3 Concrete Ingredients 5
The materials are pulverized, mixed in the right proportions, then baked in a
rotary kiln at very high (about 2,300 F) temperatures. The product from the kiln, a
glassy-looking ceramic, is called clinker. During the baking, chemical changes
occur in the original materials, which form four important compounds (among
others of somewhat lesser importance). These are dicalcium silicate, tricalcium
silicate, tricalcium aluminate, and tetracalcium aluminoferrite. The relative pro-
portions of these compounds inuence the characteristics of different cements.
The clinker is then ground into a ne powder. The average size of the particles is
only 0.0004 in. This is just an average; there are many smaller size particles.
Actually there are about seven trillion particles per pound of cement. The particles
combined surface area is about 2,000 ft2/lb. Usually, small amounts of gypsum and
various other minerals are mixed with the ground clinker to adjust the setting time
of the cement or to impart some desirable properties to the nal product. Different
types of cement are used for various jobs and conditions. For building structures in
most cases, Type Inormal portland cement, or Type IIIhigh-early-strength
portland cement are used.
While the basic raw materials of cement (limestone, clay, shale, etc.) are
relatively cheap, the making of cement, chiey the previously described baking
process, requires large amounts of energy (e.g., natural gas). Thus the cement is by
far the most expensive component of concrete. To save cement, other materials that
have hydraulic properties can be substituted for some part of the cement. Substitu-
tion of up to 35 % to 45 % by weight may be permitted. Fly ash, a by-product of
coal-red power plants, and ground blast furnace slag are two such commonly used
substitutes.
Aggregates, such as the ller material, make up the bulk of the volume in concrete.
Thus it is important that the aggregates be of good quality, strong and resistant to
the environmental forces (physical and chemical) that will affect the concrete
throughout its intended life. Aggregates should not contain chemicals or materials
that might lead to the destruction of the inner structure of the concrete.
As mentioned before, approximately 6575 % of the total volume of concrete is
aggregates. In a somewhat arbitrary way they are divided into two classes. The
particles that pass a #4 sieve, that is, less than 0.25 in. are called ne aggregates or
sand. Coarse aggregatesnatural gravel or crushed stoneare particles that are
larger than 0.25 in. Aggregates are mostly dug or dredged from a pit, river, lake, or
seabed. They are also produced by crushing rocks (limestone, dolomite, etc.) and
boulders.
Producers usually wish to ll most of the volume with the cheaper ingredients,
that is, the aggregates, so they rst carefully separate the different grain sizes (the
ne and the coarse), then mix them in desirable proportions. In the resulting particle
6 1 Reinforced Concrete Technology
distribution, the successively smaller particles ll the voids between the larger
parts. This is referred to as good gradation.
At a certain point however, the smallness of a ne aggregate particle becomes
deleterious to the quality. The cement paste must coat all the particles to bind them
together. The smaller the particle, the larger is its surface-to-volume ratio. Thus
particles less than 0.006 in. are undesirable. If necessary, aggregates are carefully
washed to rid them of adhered clay or mud particles.
On the other end of the scale, the maximum size of the coarse aggregate must be
controlled as well. The gravel in the concrete mix must pass between closely spaced
reinforcing bars, and the concrete must smoothly ll often-intricate forms. In
general, the maximum-size aggregate should be no larger than one-fth of the
narrowest dimension of the concrete form. Furthermore, in building structures,
where the minimum allowable clear spacing between reinforcing bars is 1 in., the
maximum size of coarse aggregate particles is usually limited to about 3/4 in. in the
concrete mix.
The unit weight of concrete made with gravel (or crushed stone) and sand
aggregates varies between 140 and 150 pounds per cubic foot (pcf). In calculations
an average weight of 145 pcf is used for unreinforced concrete, while for the weight
of reinforced concrete structures a value of 150 pcf is used. The difference between
these two values tends to account for the greater unit weight of the embedded steel
reinforcement.
The last 50 years has also seen a growing development in the use of lightweight
aggregates. In concrete structures, because the self-weight of the structure is a much
larger component of the total loads than in steel or wood-framed structures, it is
often desirable to use lighter aggregates than gravel or stone. Concretes made with
lightweight aggregates also have better insulating properties. Most of these aggre-
gates are articially produced. For structural purposes expanded shales and clays
are used almost exclusively. Their use enables the production of structural
(as opposed to insulating) concretes with only 110115 pcf unit weight. Light-
weight structural concrete is more expensive than normal weight concrete, but its
lighter weight often reduces the overall cost of the structure.
Water Water is an important and necessary part of making concrete. The water used
to make concrete has to be free of chemicals and unwanted elements. In general, if
the water is drinkable, it can be used to make concrete, although some waters that
are not t for drinking may also be suitable for concrete. Two important aspects
about the role of water in concrete need to be discussed: the hydration process and
the water/cementitious materials ratio.
The Hydration Process When water is mixed with cement, a chemical reaction
starts between them. This is called hydration, which creates the binding quality of
the paste. The two calcium silicates that make up about 75 % of portland cement
1.3 Concrete Ingredients 7
react with water to form two new compounds: calcium hydroxide and the more
important calcium silicate hydrate. The latter compound rst appears as a gel,
which later turns into a solid. The surface area of the calcium silicate hydrate is
enormous. Its crystals can be discerned only in a scanning electron microscope.
These crystals adhere to each other, as well as to the grains of sand and gravel,
cementing (gluing) all parts together.
The hydration process develops in three stages. These are setting, hardening, and
strength development. They all are related to the rate of reaction between the
cement and the water. This rate of reaction must be carefully determined and
regulated to allow sufcient time for the concrete to be transported, placed, and
nished. When the hydration advances to a certain stage of setting, the concrete
becomes difcult or impossible to handle. Thus the concrete must be placed and
consolidated in the forms, usually within 2 hours after batching (the mixing of
ingredients with the water). Temperature also has a major inuence on the rate of
the hydration, so various chemical admixtures may be added to either retard or
accelerate the process.
From the age of about 2 hours to about 68 hours, the hardening stage takes
place. After hardening, one may step on the concrete without leaving an imprint on
the surface. The concrete is far from being strong at this stage, however. Thereafter
begins the third stage, that is, the strength development that is quite rapid in the
early days and gradually becomes slower.
Hydration continues throughout the life of a concrete structure as long as free
moisture is available to react with unhydrated parts of cement particles.
Water/Cementitious Materials Ratio The water/cementitious (w/cm) materials
ratio is of paramount importance. It greatly inuences the quality of the paste,
hence the quality of the concrete. It is dened as the weight of the total water to the
weight of the cement (or cementitious products) in the mix. The total water must
also account for the water contained by moist aggregates. The free water adhering
to the aggregates can be signicant, so it must be carefully determined, and the
weight of the additional water into the mix must be adjusted accordingly.
For complete hydration only an approximate w/cm ratio of 0.25 (25 lb of water
for every 100 lb of cementitious material) is needed. This is a theoretical value only.
Evaporation of water from the mix cannot be prevented, thus reducing the amount
available for the hydration process. Furthermore, concrete made with such a small
w/cm ratio is too dry and unworkable. More water (higher w/cm ratio) is needed to
produce a concrete that is workable. Therefore, a minimum w/cm ratio between
0.35 and 0.40 is usually required.
Workability is not a scientically denable term. It refers to the ease of placing,
consolidating, and nishing fresh concrete. It is true that more water in the mix
tends to increase the workability, but excess water creates all sorts of problems.
Practically all desirable properties of concrete, such as strength and durability, are
adversely affected by high w/cm ratios.
To begin with, concrete with excess water has a tendency to segregate. When the
fresh concrete is too uid the heavier particles (coarse aggregate) settle on the
bottom of the form, that is, they segregate from the ideal distribution of particles.
8 1 Reinforced Concrete Technology
6,000
Non-air-entrained concrete
Compressive strength, psi
5,000
28-day
4,000
3,000
7-day
2,000
3-day
1,000
1-day
Figure 1.3 Changes of concrete compressive strength with w/cm ratio and age
Then the excess water migrates to the surface in a process called bleeding,
producing a weak top layer. Large amounts of bleed water also make it difcult
to properly nish the top surface of oors. In the later stages of hardening and
strength gain, the excess water, that is, the water that was not used during hydration,
will evaporate from the concrete through tiny capillaries. This results in voids that
weaken the concrete. Figure 1.3 illustrates the dramatic strength loss with increas-
ing w/cm ratio while all other parameters, such as total cement content, are kept
constant in a given mix.
A balance must be struck between having too little water and an unworkable
mix, and having too much water that results in loss of strength and durability. Thus
an optimum water content must be used. Optimum water content is the minimum
amount necessary in a mix to maintain good workability. As will be discussed in the
section on admixtures, there are certain chemical compounds that, when added to
the fresh concrete, temporarily increase its uidity. These are called water reducing
agents (plasticizers) or high-range water reducing agents (superplasticizers). The
former agents reduce the water requirement by 510 %, while the latter ones reduce
it by as much as 2030 % without the loss of workability.
Air All concrete, even after the most careful consolidation, contains some air. Two
types of air may be present in a concrete mix: unwanted or bad air, and wanted or
good air. Bad air is basically large bubbles of air entrapped inside the mix. These
bubbles create discontinuity in the concretes texture and weaken its strength. Every
effort is made to minimize this type of air in hardened concrete. At the time of
placement the aim is to consolidate the concrete to the maximum possible density
and bring these unwanted air bubbles to the surface. Unfortunately, even with the
best consolidation efforts, these large air bubbles (well visible to the naked eye) get
stuck in the concrete by adhering to aggregates, reinforcing bars, and most often to
1.3 Concrete Ingredients 9
the inner surface of the formwork, especially in columns and sides of deep beams.
There they become visible on the surface of the hardened concrete and are com-
monly referred to as bug-holes. In well-consolidated concrete, these may repre-
sent about 1 % or less of the total volume. Their presence is not a source of major
weakness.
The second type of air, good air, is deliberately introduced into the concrete.
This process is called air entrainment. Properly used air-entraining agents distribute
tiny (microscopic) air bubbles uniformly in the concrete. The size of the bubbles
ranges between 0.0004 and 0.004 in., in excess of three billion air bubbles per cubic
foot of concrete. The chemicals used to create them are added to the concrete using
special admixtures. Air-entrainment makes the concrete mix more workable (thus
requiring less water), slightly decreases the weight of the concrete, and, most
importantly, increases the durability of the concrete.
1.3.4 Admixtures
Admixtures are chemicals added to the concrete batch during mixing or just prior to
placement to enhance properties such as rate of setting, hydration, workability,
strength, and so on. Four main types of admixtures are discussed here.
Air-entraining Admixtures Air entraining agents are hydrophobic, that is, they
repel water. Thus a lm (e.g., soap lm) forms on the surface of the bubbles that
prevents them from collapsing or coalescing. The lm also keeps water out of the
bubbles. These bubbles are nely dispersed throughout the concrete during mixing.
They do increase the workability of the concrete, but their most important role is to
increase concrete durability. Durability in this context refers to concretes resis-
tance to the destructive process of freeze and thaw cycles that occur in certain
climates.
Hardened concrete contains ne capillaries that enable moisture to penetrate. As
free water in moist concrete freezes, it expands. The expansion is signicant. Ice
takes up about 9 % more volume than unfrozen water. This expansion exerts
hydraulic pressure on the yet unfrozen water, which in turn exerts pressure on the
surrounding paste structure. If this pressure is too great for the tensile strength of
the paste to withstand, the paste structure will rupture and collapse to provide the
excess room needed for the ice. This creates more volume the next time around for
the penetrating water, thus even more room is needed to accommodate the
expanding ice. This cyclic phenomenon continues, resulting in scaling and crum-
bling of the concrete.
Entrained air voids act as relief reservoirs in the paste structure. The expanding
water in the capillaries can enter the storage space provided by these well-dispersed
tiny bubbles by overcoming the air pressure existing within the bubbles. On
thawing, the water, driven out by the compressed air, returns to the capillaries.
To impart proper freeze/thaw resistance to building structures the accepted range
is to have about 57 % entrained air in the hardened concrete volume. Air content of
10 1 Reinforced Concrete Technology
the freshly mixed concrete can be measured right at the job site prior to placement.
This ensures that the hardened concrete will have an appropriate amount of
entrained air. The actual air content can also be established on core samples
taken from the already hardened concrete. The two different tests may give
different results, because some air inevitably will be lost during placement and
consolidation.
Accelerating Admixtures (Accelerators) This admixture hastens the setting of
concrete by speeding up the hydration process, which in turn makes the concrete
gain strength faster, especially at an early age. Similar results may be obtained by
using Type III, or high-early-strength portland cement, by lowering the w/cm ratio,
or by curing the concrete at higher temperatures. Accelerators are traditionally used
in cold weather construction. Cement hydration is an exothermic process, that is, it
generates heat. The use of accelerators reduces the setting and hardening times. The
accelerated hydration produces a larger amount of heat, which helps to prevent the
concrete from freezing. In the past, calcium chloride was used as an accelerator, and
some products on the market still contain calcium chloride. This chemical, how-
ever, has many potentially dangerous side effects (e.g., chloride ions in the presence
of moisture enhance corrosion of the reinforcing), so its use is strongly discouraged.
Several, non-chloridebased and noncorrosive accelerators are available for use.
Superplasticizers This admixture reduces the water needed to create a owing
concrete as well as the water that otherwise would be needed for proper workability.
The reduced w/cm ratio results in a higher strength concrete with the same amount
of cement. Superplasticizers are indispensable when concrete is pumped between
the point of discharge from the delivery truck and the point of placement. Normal
structural concrete is said to be workable when the slump is about 3 in. (see
Section 1.5.1 for the slump test that is used to check consistency and workability).
Such concrete is too stiff to ow through a 5-in. or 6-in. diameter hose. Adding a
superplasticizer will temporarily increase a 3 in. slump to 8 or 9 in.; thus the
concrete behaves like a liquid for a short time.
Retarding Admixtures (Retarders) As the name implies, retarders have an effect
opposite to that of accelerators. They slow down concrete hydration and increase
the setting time. Retarders are used for hot weather construction because the
hydration process is much faster at elevated temperatures. Their use enables
the contractor to place and nish the concrete before advancing hydration makes
the concrete difcult to handle. Retarders are also used to make exposed aggregate
elements in the precast concrete industry. A layer of retarder paste is smeared on the
inside of the form prior to the placement of the concrete. In about 1224 hours
(depending on the curing technique used), the precast concrete element is removed
from the form and the retarder paste is washed away, exposing the surface of the
underlying aggregate structure. (Note: For additional information, refer to ACI
212.3R: Chemical Admixtures for Concrete, reported by ACI Committee 212.)
1.5 Testing Concrete 11
1.4 Curing
Freshly placed concrete is consolidated to bring large entrapped air bubbles and
excess water to its surface. This is usually done by high-speed vibrators. The
vibrating action reduces the friction between the particles and makes the concrete
behave like a thick uid. At the same time, entrapped air bubbles and excess water
are forced to rise to the surface. Vibrators are elongated cylinders with an unbal-
anced weight rotating inside at a high frequency. Vibrators are of different diam-
eters, from about 3/4 up to 6 in.; the most frequently used ones are 2 to 21/2 in. in
diameter. Frequencies may vary from 5,000 to 15,000 cycles per minute. Vibrators
should be rapidly lowered into the concrete and then slowly withdrawn for best
effect.
After the freshly placed concrete is nished, it is necessary to create the best
possible environment for the concrete to harden and gain strength. This process is
called curing. Hydration and strength gain will continue as long as unhydrated
cement particles and adequate moisture are present for the chemical reaction. Thus
the moisture in the concrete after the consolidation and nishing processes must
remain in the concrete. If the concrete dries out (i.e., the relative humidity inside
drops below 80 %), the hydration stops. Similarly, if the moisture in the concrete
freezes, the hydration will stop and the expansion of ice will destroy the paste
matrix, which is at its early stages of formation.
So the concrete should be kept moist and comfortably warm. Concrete is kept
moist by covering it to prevent evaporation from the surface, or sprinkling it several
times daily. Chemical curing compounds also are available. These are sprayed on
the concrete to form a lm that prevents moisture from escaping.
In the wintertime, freshly placed concrete is covered with insulation blankets. It
is also a usual practice to enclose the space below the fresh concrete and heat the
space with propane space heaters. This process not only prevents the freshly placed
concrete from freezing but enhances the speed of the hydration. (Note: Detailed
information may be found in the ACI Standards, ACI 305Hot Weather Concret-
ing and ACI 306Cold Weather Concreting. These contain state-of-the-art recom-
mendations regarding the topics). Figures B1.3 and B1.4 in the Appendix B show
two different methods of concrete placement.
Testing of concrete aims (1) to ensure that it has the required properties called for in
the design documents and specications, and (2) to determine the properties of
concrete in an existing structure.
Many tests can be performed to evaluate certain properties of fresh or hardened
concrete. The three most commonly used tests are as follows.
12 1 Reinforced Concrete Technology
The slump test measures the consistency and workability of concrete. This test is
performed on fresh concrete either as it is discharged from the truck (known as
testing at the point of delivery), or after it has been conveyed to the point of
placement. The distinction is sometimes important, for signicant slump loss may
occur during conveyance. The device used in this test is a 12 in. high truncated
metal cone, 4 in. wide at the top and 8 in. wide at the base (see Figure 1.4a). The
method of sampling the fresh concrete, and of lling and consolidating the concrete
inside the slump cone is standardized in the ASTM (American Society for Testing
and Materials) C143 standard.
a 4 in.
12 in.
8 in.
b
Ruler
Slump
Slump cone
Settled concrete
The cone is lled with concrete in three equal-volume layers, and each layer is
consolidated within the cone by 25 strokes of a 5/8-in. diameter rod with a rounded
end. After the third layer is lled, the excess concrete is struck off with the steel rod,
and the cone is carefully lifted off. The cone is then placed upside down next to the
concrete, and the steel rod is placed across its top. The distance measured from the
bottom edge of the rod to the original center of the slumped concrete mass is the
slump (Figures 1.4b and 1.5). The slump recommended for good workability and an
acceptable w/cm ratio depends on the type of construction. The common range of
slump in building structures is 34 in. (unless superplasticizers are used).
14 1 Reinforced Concrete Technology
The most important property of hardened concrete is its compressive strength, fc0 .
This value refers to the cylinder compressive strength of the concrete at the age of
28 days and forms the basis of the design of a structure. In the United States
compressive strength of concrete is measured on 6 in. diameter by 12 in. high
cylinders. (Note: The sampling of the fresh concrete is governed by ASTM C172,
Method of Sampling Freshly Mixed Concrete. The making and curing of the
cylinders are governed by ASTM C31, Practice for Making and Curing Concrete
Cylinder Test Specimens.)
The architect or design engineer species the number of cylinders cast for
testing. Typically one set of cylinders is made from about every 50100 yd3 of
concrete, but not less than one set from each days pour. Usually three cylinders
comprise one set. After the concrete is hardened, the cylinders are transported to a
testing laboratory where they are placed in a curing chamber. The temperature
inside the curing chamber is kept at 72 F (room temperature) with 100 % relative
humidity. These cylinders thus treated are called lab-cured cylinders. They indicate
how good the concrete mix was, not how good the concrete is in the structure, as the
contractor may not maintain ideal curing conditions on the site. To determine the
strength development of the concrete in the eld, extra cylinders may be cast and
kept in the eld to cure under the same conditions as those of the structure. These
are known as eld-cured cylinders. Comparing the strength of eld-cured cylinders
to that of lab-cured cylinders helps to determine how successful the contractors
efforts were in providing good curing.
The strength test is performed in accordance with ASTM C39, Test Method for
Compressive Strength of Cylindrical Concrete Specimens. Compression force is
applied to the prepared concrete cylinder by a hydraulic jack (Figures 1.6 and 1.7).
The load is increased progressively at a rate of 35 5 psi (pounds per square inch)
per second until the concrete cylinder fails. The load required to break the cylinder
is noted, then divided by the cross-sectional area of the cylinder. The result gives
the breaking stress, or cylinder strength. A strength test is the average strength of
two cylinders cast from the same sample.
The acceptance of the concrete (from the strength point of view) is regulated by
the ACI Code. The strength of the concrete is considered satisfactory when:
The arithmetic average of any three consecutive strength tests equals or exceeds
fc0 , (i.e., the specied design strength)
No individual strength test (the average of two cylinders) falls below fc0 by more
than 500 psi when fc0 is 5,000 psi or less; or by more than 0.10 fc0 when fc0 is more
than 5,000 psi
Often extra cylinders are cast and tested at an earlier age (7 days) to evaluate the
development of strength. Although different cements may gain strength at some-
what different rates depending on the relative proportions of the main chemical
compounds, the 28-day strength can be estimated by extrapolating early test data.
1.5 Testing Concrete 15
6 in.
The problem with the 28-day strength test is that if the results are unsatisfactory,
the remedy is usually difcult and expensive. In any major project, construction
progresses far in 28 days, often resulting in two or three additional oors. Thus,
removing the weak concrete and replacing it is rarely an option. The various
strengthening methods of the structure are generally very expensive. So it is of
paramount importance to have good quality control throughout the process from
16 1 Reinforced Concrete Technology
mixing (making sure that all the required ingredients are there in the right
proportions), to transporting, placing, nishing, and curing.
Core-cylinder tests are used to evaluate the strength of the concrete in an existing
structure. The sample is obtained by coring the hardened concrete. The size of the
sample is typically 2 in. in diameter and 4 in. high, and its compressive strength is
determined in a manner similar to that for a normal cylinder test. Larger-diameter
and longer cylinders may be cored; however, the ratio of height to diameter should
preferably equal 2.
Nondestructive tests exist, that is, tests that do not require the removal of a
sample. The most popular is the rebound hammer test. This test uses a calibrated
spring-loaded device that shoots a rod against the concrete surface. A dial gage
measures the rebound that is correlated to the concretes modulus of elasticity, from
which an estimate of the compressive strength can be made.
Concrete is very strong in compression. In the United States, cylinder tests are used
to study the behavior of concrete in compression. In other parts of the world,
compression testing is typically done on 20-cm cube samples. Results obtained
from the two different tests are different for samples made of the same concrete.
This is due to the shape and proportions of the samples.
The deformation of the sample under load during testing may also be measured
to establish the stress-strain diagram. Axial
compression
stress is dened as the
force divided by the cross-sectional area f PA , which has units of psi (pounds
per square inch) or ksi (kips per square inch). In SI (International System) units, the
stress may be measured in KPa (kilopascal) or MPa (megapascal) units. Strain is the
deformation of a unit length of the member and is dened as
, where is
the change in the length, and is the original length. Strain is a dimensionless
number, for example, inch/inch.
It must be emphasized that the cylinder test used to evaluate the strength of
concrete is only a representative sample and provides only an indicative and
correlative value of how the concrete may behave in the structure. The cylinder,
when tested, is free to expand laterally. The concrete in a structure may be conned
by its surrounding. A conned sample of concrete is much stronger. Note also that
the cylinder test determines the strength of the concrete under short-term loading.
1.6 Mechanical Behavior of Concrete 17
Research indicates that under long-term loading (e.g., within building structures),
the strength of the concrete is less than that exhibited by the cylinder testing.
Different concrete mixes exhibit not only different strengths, but very different
deformation characters. Figure 1.8 shows the stress-strain diagrams of typical
concrete mixes.
fc (psi)
2,000
1,000
0 ec
0.001 0.002 0.003 0.004
fc (psi)
6,000
5,000
4,000
3,000
2,000
1,000
0 Time
28 days 1year 5 years
although the hydration process between cement and water may continue in the
presence of available free water.
Modulus of Elasticity of Concrete In elastic materials (or materials that behave in
an elastic way up to a certain stress level), a denite linear relationship exists
between stress and strain. The coefcient in the relationship is called the modulus of
elasticity. The capital letter E is used to denote this modulus, and the relationship is
dened as stress modulus of elasticity strain.
The behavior of concrete, as described by the typical stress-strain curves in
Figure 1.8, is not this simple. The diagrams are not linear; thus, the E value (i.e., the
slope of the tangent to the curve at any point) is changing continuously. To simplify
the matter and establish a value that can be used in calculations, a substitute E value
is used. The E value used is the secant modulus, which is the slope of a line
connecting the point of zero stress and zero strain to the stress point of 0.45 fc0
and its corresponding strain (Figure 1.10). By denition, this value is the modulus
of elasticity of concrete.
This value is different for different strength concretes: Stronger concretes have
greater E. Furthermore, concretes made with different aggregates (normal-weight
concrete, lightweight structural concrete, etc.) also exhibit different moduli of
elasticity.
The value for the modulus of elasticity is needed when calculating instantaneous
(also called elastic) deformations of structures under load, such as the deection of
a beam. This is justied, for at stress levels that exist during normal use of
structures, the concrete responds in a quasi-elastic manner to short-term loads.
After studying the results of hundreds of tests and applying statistical analysis
(tting a mean curve to the values), researchers have determined that Equation (1.1)
provides a reasonable approximation for the modulus of elasticity of concrete, Ec
for wc between 90 and 160 pcf (ACI Code, Equation 19.2.2.1.a).
1.6 Mechanical Behavior of Concrete 19
fc
f c
0.45 f c
Esecant
p
Ec 33w1:5
c fc0 1:1
where wc the weight of the concrete in pounds per cubic foot and fc0 the
ultimate cylinder strength, or specied compressive strength of concrete in pounds
per square inch. The resulting unit for Ec is psi, and substitution into the equation
must be made using the units as dened.
Unreinforced normal-weight concrete is about 145 pcf. When wc 145 pcf is
substituted in Equation (1.1), the result, after rounding, is Equation (1.2) (ACI
Code, Equation 19.2.2.1.b):
p
Ec 57,000 fc0 1:2
As discussed above, there are two ways of determining the modulus of elasticity:
(1) by testing, and (2) by using the approximate equation provided by the ACI
Code. Because the concrete that will go into the structure has not been made,
placed, and cured at the time of design, the designer is invariably forced to use the
accepted approximate equation.
Example 1.1 Find the modulus of elasticity of a concrete mix with the compres-
sion strength, fc0 3,500 psi. Assume the mix is lightweight structural concrete
with a unit weight of 110 pcf.
Solution The ACI approximate equation for Ec is:
20 1 Reinforced Concrete Technology
p
Ec 33w1:5
c fc0
In this equation fc0 and fr are in psi units. (Substitution of fc0 must be made in psi,
otherwise the formula will produce erroneous results). is a modication factor to
adjust the equation when lightweight aggregates are used and varies depending on
the type of aggregates. is equal to 0.75 for the all-lightweight concrete;
0.750.85 for the ne blend-lightweight concrete; 0.85 for the sand-lightweight
concrete; 0.851.0 for coarse blend sand-lightweight concrete; and 1.0 for the
normal weight concrete. This formula is simple to use, but in most cases, it
overestimates the true tensile strength of a concrete element.
1.6 Mechanical Behavior of Concrete 21
P = 150 lb P =150 lb
Solution
(a) Test Results The beam is subjected to two loads: its weight, and the two
concentrated loads as shown in Figure 1.11. The beam self-weight is a
uniformly distributed load, with a magnitude of
(3)(6)
w = (150) = 18.75 lb/ft
(12)(12)
Conversion factor for in2 to ft2
The maximum moment for the beam occurs at the midspan. The equations for
the maximum moments are as follows:
w2
Mmax for the beam with uniform load
8
P
Mmax for the beam with concentrated loads
3
w2 P
Mtotal
8 3
18:7592 1509
Mtotal
8 3
Mtotal 190 450 640 ft-lb
22 1 Reinforced Concrete Technology
w = 18.75 lb/ft A
P = 150 lb
b =3 in.
Section AA fb
A
4'-6"
Mtotal is the internal moment at the mid-span of the beam. This moment creates a set
of compression stresses at the top, and tensile stresses at the bottom of the beam as
shown in Figure 1.12. The maximum bending stress occurs at the top (compressive)
and the bottom (tensile) of the cross section. The equation for the maximum
bending stress is:
Mc M
fb
I Sm
where c is the distance from the neutral axis (where stress is zero) to the top or
bottom of the beam, I is the moment of inertia of the section about its neutral axis,
and Sm is the elastic section modulus. For a rectangular shape, Sm is:
bh2
Sm
6
Therefore,
3 6 2
Sm 18 in:3
6
Conversion factor for feet to inches
M (640)(12)
fr = fb = = = 427 psi
Sm 18
Concrete, like most other materials, expands with rising temperature and contracts
with falling temperature. Suppose a concrete element with the length is restrained
at only one end (A) (see Figure 1.13a). Under an increase in temperature of T
(degrees of Fahrenheit), the element expands and has an increase in length equal to
. This increase in the length can be calculated using Equation (1.4).
T 1:4
A B
A B
Figure 1.13 Effects of temperature on concrete: (a) free to move, (b) restrained
24 1 Reinforced Concrete Technology
f
1:5
Ec
f
T
Ec
f Ec T 1:6
where Ec is the modulus of elasticity of the material in psi (or ksi), and f is the
resulting stress in psi (or ksi) developed in the restrained element due to a change in
temperature equal to T.
As shown by Equation (1.6), large stresses can build up if the length change is
restrained. The buckling of pavements often seen on hot days is the result of two
neighboring pavement slabs pressing each other (in the absence of a wide enough
expansion joint) while trying to expand. The buckling relieves the prevented
expansion. On the other hand, tensile stresses will build up when concrete tries to
shorten with dropping temperatures if the free contraction is somehow hindered.
For example, for an fc0 4,000 psi concrete:
f Ec T
p
f 57,000 4,000 6 106 1:0 21:6 psi
for each degree of temperature change (T 1 F), if the length change is fully
prevented. If the concrete in the above example has an ultimate tensile strength of
p
7:5 4,000 474 psi, the theoretical value of the temperature drop that will crack
this concrete is only 474/21.6 21.9 F, a rather small temperature change.
Admittedly it is very rare that concrete is fully restrained against movement due
to temperature change. But the unsightly cracking of concrete structures all around
us provides ample testimony to the results of restrained volumetric changes.
The value of for concrete is quite similar to that of steel (6.5 106 in./in./ F).
Thus, the reinforcing steel inside the concrete will expand or contract at about the
same rate as the surrounding concrete, without signicant stresses resulting from
expanding or contracting at a different rate. Aluminum, for example, has a coef-
cient of expansion roughly twice that of steel. Thus, the use of aluminum as
reinforcement for concrete is not a good idea; for when the temperature rises, the
aluminum rod expands at twice the rate of the surrounding concrete at the interface
between the two materials. The conicting expansion rates cause all kinds of
weird stresses at the interface, breaking down the necessary bond between the
two materials.
1.7 Volume Changes in Concrete 25
Shrinkage means that the concrete becomes smaller in volume. There are many
causes of shrinkage, but the most signicant contributor to this phenomenon is the
loss of water.
As previously discussed, more water is needed in a concrete mix than the cement
uses for hydration. Some of this excess water bleeds and evaporates during and
immediately following consolidation while the concrete is still plastic. The heavier
parts in the still-uid concrete tend to settle, causing what is known as setting
shrinkage or plastic shrinkage. Reinforcing bars or large aggregates near the
surface obstruct the uniform settlement of the concrete, thus enhancing the forma-
tion of thin hairline cracks on the surface. These hairline cracks look like cobwebs:
lots of relatively short, thin cracks in all directions. Their depth is usually limited to
small fractions of an inch.
After the concrete hardens, it still contains free water in the capillaries and water
adsorbed on the surface of particles. As this water slowly evaporates, the concrete
continues to shrink, not unlike a sponge shrinks as it dries. This causes what is
known as drying shrinkage. The rate of the drying shrinkage is tied to the speed of
the evaporation, which in turn depends on the porosity of the concrete and the
environment, that is, temperature and relative humidity. Concrete in highly humid
climates shrinks less than corresponding concrete does in arid climates.
More than 90 % of the drying shrinkage happens within the rst few weeks after
casting. Drying shrinkage, however, is partially reversible. Thus, if the concrete
gets soaked it swells, and when it dries out again it shrinks. If the drying shrinkage
could take place without any restraint whatsoever (a theoretical proposition rather
than what really occurs), no stress buildup would result. Because, however, free
shrinkage is usually restrained (i.e., something prevents the concrete from shorten-
ing in any direction), tensile stresses start to develop and build up. In moderate
climates the average dimensional change is about 300 millionths (300 106 in./
in.). Compared to the length change due to a decrease in temperature, the effect of
the average shrinkage value is similar to that of a 50 F temperature drop. If the
developing tensile stress is greater than the tensile strength of the concrete at any
point, the concrete will crack. The crack should be thought of as a relief from
tension caused by the prevention of free movement.
An example is a long wall that has been cast on top of its footing. The footing has
already cured and hardened. When the wall tries to shrink, the footing restrains its
bottom edge from moving. The top of the wall, however, is free to shrink length-
wise. Thus, a tug of war results between the top and the bottom of the wall, resulting
in one or more cracks with diminishing width from top to bottom (Figure 1.14a).
Another example is a oor that is cast over walls placed earlier, thus hardened.
As the oor shrinks and tries to change its long dimensions, it cannot because the
walls restrain it. The buildup of tensile stresses results in cracks, especially in the
corners, where the edges of the slab try to move in two different directions. The
relief comes as diagonal cracks in the corners (Figure 1.14b).
26 1 Reinforced Concrete Technology
a
Cracks
Crack
Figure 1.14 Shrinkage cracks: (a) Concrete wall; (b) Concrete oor
Plastic shrinkage cracks are characterized as random surface cracks, that is, they
do not penetrate the full thickness of the concrete element. Drying shrinkage cracks,
on the other hand, are usually full depth and quite wide (1/8 in. or more is not
unusual).
The mitigation of the effects of shrinkage requires good design and construction
practices. The following actions help to minimize cracking in slabs and walls due to
shrinkage:
Use the minimum amount of water in the concrete mix. The concrete should
have not only the smallest w/cm ratio, but also the smallest amount of water in
absolute terms. This also means using the smallest amount of cement necessary
to achieve the desired concrete strength, because more cement introduces more
excess water in the mix. Such a tactic is also good for keeping costs down.
Use good curing technique. Moist curing helps keep the excess water from
evaporating too soon (i.e., before the concrete has a chance to develop its tensile
strength).
Limit the size of the pour to about 6080 ft maximum length in any direction.
The construction is broken up into segments by the use of construction joints
(Figure 1.15a). If the second pour is 34 days after the rst pour, some
shrinkage has already taken place in the rst pour. On some projects the pouring
sequence may follow a checkerboard pattern. Other construction techniques
leave a gap between two neighboring pours (1224 in.) that is lled in when the
larger pours have undergone most of their shrinkage.
Provide reinforcing steel (shrinkage reinforcement). Because steel bars are
bonded to the concrete, they restrain and limit the change of length of the
concrete.
Use shrinkage compensating cement (Type K). This particular cement type
expands during the early stages of hydration, before any drying shrinkage
occurs due to moisture loss. Reinforcing is also provided in both directions in
a wall or a slab, and the expansion of the concrete at the early stage of hardening
1.7 Volume Changes in Concrete 27
a b Dowel, greased on
Steel bars the second pour side
First pour Second pour First pour Second pour
Expansion
joint material
Figure 1.15 (a) Construction joint, (b) expansion joint, (c) control joint
induces tension in the steel (i.e., the expanding concrete tries to elongate the
reinforcing bars). If we recall Newtons law on action and opposite and equal
reaction, it is easy to understand that the steel in turn will compress the concrete.
When shrinkage sets in and causes tension in the concrete, it rst will have to
overcome the precompression in the concrete. Thus, the forces will either
completely cancel each other out, or at least the resulting tensile stresses will
be greatly reduced.
Provide expansion/contraction or control joints (Figure 1.15b, c, respectively).
At an expansion joint the longitudinal reinforcing is interrupted. The joint is lled
with an elastomeric material that can be compressed when the concrete expands,
and permits the free movement of the two parts relative to each other when the
concrete shrinks. A key-way (or a dowel that is greased on one side of the joint to
prevent bonding) forces the two parts to stay together in the out-of-plane sense
while still allowing them to move freely longitudinally. This prevents one side
from moving higher or lower than the other and thus creating a trip hazard or a
step. The role of the control joint is different. A weakening groove, usually 1/8 in.
wide and about one-fourth to one-fth of the slab thickness, is either tooled into the
freshly nished concrete, or cut with a saw into the concrete as soon as it hardens
enough so as not to leave an imprint on the surface. This allows the shrinking
concrete to crack along that straight line where the section is weakened. Control
joints essentially locate shrinkage cracks along predetermined paths instead of
letting them naturally meander all over the slab or wall.
A structure deforms when it is subjected to loads. For example, beams and slabs
deect, columns become shorter, and so on. For every stress level, there is a
corresponding strain. Strain is nothing else than the deformation of a unit length
of the material.
28 1 Reinforced Concrete Technology
The causes of creep are complex. Interestingly, one contributor is the loss of
adsorbed water. In drying shrinkage the loss occurs due to the lower relative
humidity of the ambient atmosphere, and this loss leads to the shrinkage. In the
case of creep, the sustained compression on the concrete squeezes some of the
moisture out of the concrete. This in turn lets the solids consolidate even more.
The second major cause of creep is thought to be microfracturing in the hardened
paste around sharp edges of aggregates under the effect of compression.
Creep deformations can be very signicant. They are caused mainly by dead
loads or sustained loads, because the self weight and some permanently attached
superimposed dead loads are dominant in most concrete structures, whereas the
transitory (live) loads are less signicant. Creep could account for an additional
100300 % of the instantaneous deformations. Thus, a beams original deection of
1 in. may grow to anywhere between 2 and 4 in. If this additional deformation is not
accounted for in the detailing of attached items, such as partitions, it may cause
serious distress in them.
Reinforcing steel is used to overcome the weakness of concrete in tension. The role
of the reinforcing is to resist the tension in structures. Thus, a hybrid structural
composite called reinforced concrete is created, where each material does the work
it is well suited for. Concrete takes care of the compression, while the steel takes
care of the tension.
1.8 Reinforcing Steel 29
To better understand the material that will be discussed in the following chapters,
we review the behavior of steel under stress. This review will also help to famil-
iarize the reader with the terminology that will be used later.
Consider Figure 1.17, which is a typical stress-strain diagram for steel in tension.
There are four distinct zones in the stress-strain diagram for steel. First is the
elastic zone, where steel under stress will go back to its original length if it is
released. In this zone the stress in the material is linearly proportional to the strain.
(Robert Hooke formulated this relationship, so we refer to it as Hookes law. Hooke
worked with Christopher Wren on the construction of St. Pauls Cathedral in
London, England.) When steel is pulled beyond the elastic zone (elastic or propor-
tional limit) it yields. Yielding is an elongation of the steel with no appreciable
change in stress. The onset of yielding (elastic or proportional limit) is the begin-
ning of the plastic zone. When the steel is pulled beyond the proportional limit it
will not return to its original length, but remain permanently deformed.
The stress at which steel yields is called yield stress and it is noted as fy. The
corresponding strain is called yield strain or y. By the time yielding ends, the
corresponding strain is about eight to ten times the strain at the proportional limit.
After yielding, the steels stress/strain curve starts to climb again in a curvilinear
mode until it reaches a plateau called the ultimate strength ( fu). This curvilinear
zone of the stress/strain curve is called strain hardening.
fs
Strain-
Elastic hardening Failure
zone Plastic zone zone zone
fu
Failure
fy
Es = f e
es
ey
Plain bars do not have any deformations and rely on surface bonding only to
prevent relative slippage. These are no longer in use, although they may be
encountered in old structures.
Different grades of steel are used, typically made from either new steel, scrap
metal, or a mixture of both. Grades of steel represent their guaranteed minimum
yield stress in ksi units. For example, Grade 60 steel refers to reinforcing steel with
a guaranteed minimum yield stress of fy 60 ksi. ACI Section 20.2.1.3 provides
the various types of steel that are used for the production of deformed bars.
Table A1.1 in Appendix A lists the different types of steel used as reinforcing
bars along with their mechanical properties. Each type of steel in Table A1.1 has an
ASTM designation such as A615. Different types of steel in Table A1.1 are:
Carbon steel (A615)This is the most common type. (Type S steel).
Low-alloy steel (A706)This type of steel provides enhanced weldability.
(Type W steel).
1.8 Reinforcing Steel 31
Stainless steel (A955)This type of steel has application where high corrosion
resistance or controlled magnetic permeability is required. Their property
requirements are the same as those for carbon steel. (Type SS steel).
Rail steel (A996)This is made of recycled railroad track. (Type R steel).
Axle steel (A996)Similar to rail steel, but made from axle as scrap metal.
(Type A steel).
Low-carbon chromium steel (A1035)This is a high-strength material that is
permitted for use as transverse reinforcement in special earthquakeresistant
structural systems and spirals in columns. (Type CS steel).
Of all these reinforcing steels, most construction uses A615 Grade 60 (carbon)
steel. Grades 75 and 80 are sometimes used in columns. Grades 100 and 120 are
mainly used for transverse reinforcements of structures in seismic prone regions.
Grade 50 has not been around for quite a while. Grade 40 is almost never used, for it
has only two-thirds of the strength of Grade 60 steel, and its cost in place per lb is
the same.
Bar Sizes Steel bars are made in different sizes. Bar size, in general, represents the
diameter of the steel bar in inches. From #3 to #8 (#1 or #2 bars do not exist), each
number represents the diameter of a bar in fractions of 1/8 in. For example, #3 bar
means that the diameter of the bar is 3/8 in., and #8 is 8/8 in. 1 in. diameter.
The heavier (larger-diameter) #9, #10, and #11 bars do not precisely follow the
1/8 in. rule, but they are close.
In addition, there are #14 and #18 bars, which are very large, heavy bars. They
are used mostly in large columns in high-rise construction and are available on
special order.
Table A1.2 includes the diameters and areas of the available steel bars.
Identication of Steel Bars Steel bars used in concrete construction have special
identication marks rolled on them. These marks provide information such as where
the bars were produced, the bar size, type of steel, and their grade (see Figure 1.19).
Epoxy-Coated Bars The highly alkaline environment that the concrete provides for
the embedded reinforcing usually protects it from corrosion. Some structures,
however, such as bridge decks, parking structures, coastal structures, and so on
are often exposed to moisture containing chloride salts. The deicing salts (road
salts) that are used on roads and bridges, and carried into parking structures by the
automobiles, contain large amounts of soluble chlorides. When such solutions get
into contact with the reinforcing, the result is corrosion (oxidization or rusting) of
the steel.
The rust (ferrous oxide) grows to about eight to tenfold the volume of the
original steel. Thus, as the rust tries to create elbow room for itself, the internal
pressure starts to crack, split, and spall the concrete around it. This in turn provides
more access to the dangerous chloride-laden moisture.
One way to protect reinforcing in this kind of environment is the use of epoxy-
coated bars. Epoxy resin is an excellent adhesive and protects the steel from
chloride attacks. A note of caution is in order, however: Such bars must be handled
carefully to prevent nicks or cracks in the coating. Such places are especially
attractive to chloride ions and often become nodes of violent and rapid corrosion
in the reinforcement.
2. Welded Wire Reinforcements (WWR) In certain situations it is more econom-
ical to use welded wire reinforcements (WWR) in lieu of a series of small-diameter
bars. WWR are thin wires spaced at certain distances in two orthogonal directions
and fabricated in either large sheets, or in long rolls in the case of light-gage wires.
They are welded together at intersection points, usually by the electric resistance
welding method. The chief advantage of using WWR is the labor saving. Individual
reinforcing bars are placed one by one and are secured by tying them together at
every intersection. This ensures that they will remain at the desired location
throughout the concrete placement, consolidation, and nishing process.
WWR are available in commonly standardized wire sizes and spacing. Table 1.1
lists some of the commonly used styles of WWR. The standard designation of the
reinforcement represents the spacing and the wire sizes. In the modern designation
system the W-number represents the approximate cross-sectional area of the wire in
multiples of 0.01 in.2. Thus, the cross-sectional area of a W4.0 wire is about
0.04 in.2. As an example, 6 12 W4.0 W2.5 represents wires with cross-
sectional areas of 0.04 and 0.025 in.2 in a rectangular grid of 6 in. 12 in. as
shown in Figure 1.20.
A = 0.040 in2
12 in.
A = 0.021 in2
12 in.
AA
A
Problems
Self-Experiments
In the self-experiments of this chapter, you learn about the different aspects of
making concrete by using simple tools.
5. Tap water
6. A 3/8 or 1/2 in. diameter wood dowel, about 1215 in. long
Make three samples: (1) cement sample, (2) concrete with a w/cm ratio of 1.0,
and (3) concrete with a w/cm ratio of 0.5.
1. Cement Sample
Pour 10 oz of cement with 6 oz of water into a bowl and mix them thoroughly.
Note how much effort is used to mix the cement with water. Then place the mix
in can number 1. Consolidate the mix in the can by prodding it with the dowel
about 1215 times.
2. w/cm 1.0 Concrete Sample
Pour 3 oz of cement, 10 oz of sand, 10 oz of gravel, and 3 oz of water into a bowl,
and mix them thoroughly. Again note how much effort is needed to mix the
materials. Then place the concrete in can number 2. Consolidate the mix in the
can by prodding it with the dowel about 1215 times.
3. w/cm 0.5 Concrete
Pour 4 oz of cement, 10 oz of sand, 10 oz of gravel, and 2 oz of water into a bowl,
and mix them thoroughly. As in the rst two cases, pay attention to the amount of
effort needed to make the mix. Then place the concrete in can number 3. Con-
solidate the mix in the can by prodding it with the dowel about 1215 times.
Leave the three samples for approximately 6 hours at room temperature. Check
them every 6 hours for 3 days. Record any observations. Answer the following
questions:
Which mix was easiest to make (i.e., which one was most workable)?
Which mix resulted in the most bleeding?
Was any sign of hydration observed?
At the end of the 3 days cut the tin cans to completely expose the samples.
Answer the following questions:
What are the differences in the textures of the three samples?
Which sample has the most uniformity of material?
Chapter 2
Rectangular Beams and One-Way Slabs
2.1 Introduction
This chapter covers the analysis (checking the strength) and the design (sizing the
concrete and steel) of reinforced concrete beams and slabs that span primarily
one way.
The previous chapter emphasized that concrete is very weak in tension, but
strong in compression. As a result, reinforcements are used to supply tensile
strength in concrete members (most commonly in the form of round reinforcing
bars or rebars). Like any other building system, reinforced concrete structures have
advantages and disadvantages.
1. Can be cast into any shape This is the main advantage of reinforced concrete
compared to other building materials. Concrete members can be made into any
desired shape by using forms. Figure B2.1 in Appendix B shows the pleasing
exterior of a reinforced concrete building.
2. Has great resistance to re and water Concrete loses its structural integrity
much more slowly than wood or steel when subjected to high temperature. In
fact, concrete is often used as reproong material. Concrete also better resists
exposure to water, does not corrode like steel, and does not lose strength as wood
does. Certain chemicals in water, however, can harm concrete.
3. Is a low-maintenance material Concrete does not corrode, so it does not need to
be painted and regularly maintained when exposed in the environment.
4. Has very long service life Reinforced concrete structures that are well designed
and built last a very long time.
1. Has very low tensile strength Concrete has a very low tensile strength in
comparison to its compressive strength. Consequently, reinforcing steel bars
are needed to counteract the development of tensions in concrete structures.
2. Requires shoring and forms This is a major disadvantage of concrete because it
raises the cost of concrete structures, especially in countries such as the United
States where labor costs are high. Shoring and formwork often constitute more
than half the total cost of the structure.
3. Has variations in properties The mechanical and physical properties of concrete
are sensitive and require careful proportioning, mixing, curing, and so
on. Eliminating large variation in these properties demands carefully monitored
procedures.
4. Results in heavy structural members Reinforced concrete structures are heavier
than similar steel or wood structures. This results in larger building dead loads,
which in turn result in larger foundations. Concrete structures are also more
sensitive to differential settlements. Thus, concrete structures require relatively
good soil conditions.
Before attending to the main topic of this chapter, which is the analysis and design
of bending members, a discussion on the concept of design is appropriate.
Ask ten people about the meaning of the word design and you probably will
get ten different answers. Design also has very different meanings to architects and
to engineers. And to top it all off, design is often viewed as synonymous with sizing
of members. So we hope that readers will forgive the rather loose usage of the term
design.
Structural design of reinforced concrete structures is an iterative process. It
begins with the layout of the structure or, in other words, with the selection of the
structural system. Any practitioner will admit that this initial step is by far the
hardest part of the process. It requires the designer to come up with a synthesized
whole for the building, laying out all the component elements (columns, girders,
beams (or joists), and slabs). Furthermore, the designer must also estimate the sizes
of the elements within the space in order to go to the next step, that is, to analysis.
The owchart of Figure 2.1 presents a somewhat simplied picture of the
process. Oddly enough, it begins with a step in synthesis, or the conception of the
structure. This step is nonmathematical, for the aim of the study at this point is to
look at what the building structure should do. What spaces are required? What is the
minimum column spacing required to t the architectural program?
But before we reach the part designated as Analysis or Design, we must
complete another exercise: identifying the loads that the structure may be subjected
to in its life span.
2.4 On the Nature of the Design Process 39
Structure is conceived
Input:
Functional requirements (space layout, column spacing, etc.)
Aesthetic requirements
Loads
Self-weight
Superimposed dead loads [floor finishes (or roofing and
Analysis insulation as appropriate), partitions, ceilings, suspended
mechanical and electrical equipment]
Live loads (building code requirements)
Economy
Serviceability (short- and long-term deflections)
Durability requirements
Fire rating requirements
Loads generally fall into two major categories: gravity loads and lateral loads.
Gravity loads are further divided into two major groups: dead loads and live
loads. One can only guess how this nomenclature came into usage. Perhaps
people originally identied loads that were stationary as dead, and loads that
moved as live. Today, we make a somewhat different distinction between these
two loads. Dead loads are those that remain permanently attached to the structure,
while other loads that are transitory in nature are referred to as live loads. Thus,
furniture and stored items as well as loads from peoples activities are in the latter
category. For example, most of the weight in a librarys stack area is from the
stored books with only a very small part of the oor loads coming from the
visitors; nevertheless, the stacks and the books are considered live loads. In
addition, environmental effects such as moisture or temperature changes may
create stresses in the structure, so they also may be loosely dened as loads that
the structure must safely withstand.
Before any meaningful analysis can be performed to calculate and appropriately
size any component element within a structure, designers must establish the
loads that such an element can safely support, or at least must reasonably
approximate them.
In a concrete structure, the self-weight is a very signicant part of the dead loads.
Because self-weight depends on the size of the particular member, a reasonable
estimate must be made on the size. After the designer estimates the size, he or she
can calculate the loads from the self-weight, assuming that reinforced concrete
weighs about 150 lb/ft3. At this point we do not want to tax the students attention
with detailed discussion on the selection of an appropriately sized beam or slab, and
40 2 Rectangular Beams and One-Way Slabs
all of the reasons thereof. This subject will be discussed later in this chapter. In any
case, if during the design process the designer determines that an initial estimate of
the members size, and thus the self-weight, was signicantly in error, he or she has
to re-analyze the member, taking into account the newly adjusted size; thus, the
iterative nature of the design and sizing.
Superimposed dead loads (SDL) are somewhat ambiguous. Often these items
and their precise location in space are not completely known at this stage of the
design (see Figure 2.1). Partition layouts have not been decided yet, or may change
in the future. Ductwork, piping, and light xtures may go anywhere. So the designer
is forced to make a blanket estimate on these. Most practitioners estimate that the
combination of these items will exert about 1520 lb/ft2 of oor area. (The only
areas that need more careful attention are those where some special ooring, such as
stone or terrazzo, is planned. These items exert about 1213 lb/ft2/in. thickness.
Thus, a 2 in. terrazzo ooring weighs about 25 psf.)
Live loads (LL) are prescribed by building codes for the particular usage of a
space. These loads are listed as uniformly distributed minimum loads and represent
the current professional wisdom. Because live loads are not uniformly distributed
except in very isolated cases, they have very little, if anything, to do with the real
loads that may occur on structures. Actual surveys show that total loads, uniformly
averaged out over the whole oor area, amount to only about 1520 % of the codes
mandated minimums in spaces like hotels, residential buildings, and ofces. These
minimums, however, represent a statistical probability of the loads that the structure
may experience in a projected lifetime of 50 or 100 years. Furthermore, these code-
prescribed live loads also try to account for the dynamic nature of many loads by
treating them as equivalent static loads.
This discussion of loads should sufce to show that any calculation made during
the load analysis phase will contain unavoidable inaccuracies and uncertainties.
These errors are inevitable no matter how carefully the designer tries to evaluate the
currently envisioned, but essentially future loads.
Example 2.1 In this simple oor plan, beams 12 in. wide and 20 in. deep are
spanning 30 ft. The beams are located 90 000 center to center. A 5-in. thick slab
spans from beam to beam. (See Figure 2.2.) The oor structure will be used in a
general ofce building, thus (per Code) the minimum uniformly distributed live
load is 50 lb/ft2. Calculate the dead and live loads that one interior beam has to
carry. Assume 20 psf for the superimposed dead load for the partitions, mechanical
and electrical systems, and so on.
Solution The beams are 9 ft apart, so each beam is assumed to be responsible for
the loads that occur 4.5 ft from either side of the beams centerline. Thus, each
linear foot of beam will support loads from 9 ft2 of oor in addition to the weight of
the stem.
2.5 Live Load Reduction Factors 41
30'-0" 5 in.
A A
15 in.
In addition, the beam will support live loads from 9 ft2 of oor area on each
linear foot of beam. Thus:
TOTAL LIVE LOADS: wL 9 50 psf 450 lb/ft
Summary: See Figure 2.3.
wL 450 lb/ft
wD 930 lb/ft
30'-0"
We complete this discussion of loads by dealing with the concept of live load
reduction factors. These are derived from statistical analyses of the probability of
having the maximum amount of live loads everywhere on a oor of a building.
42 2 Rectangular Beams and One-Way Slabs
A B C D E F
B-1
G-1 C-1
2
(AI)B-1 (AI)G-1
(AI)C-1
Studies indicate that the larger the oor area that contributes loads to a particular
member, the less likely it is that every square foot of that area will bear the
maximum amount of live loads.
Different codes deal with this concept somewhat differently. Some codes relate
the live load reduction to the tributary area (AT), or the area directly loading the
particular element under investigation. Other codes relate the live load reduction to
the so-called inuence area (AI), the area in which a part, however small, of any
load may contribute to the loading of a particular element under investigation. In
other words, the inuence area for a structural member is the part of the building
structure that may fail if that member is removed.
As an example consider Figure 2.4, which shows the oor framing plan for
a reinforced concrete building. To determine the inuence area for beam B-1,
assume that this beam is removed. This will cause the slabs supported by B-1 to
fail. As a result, the inuence area for B-1 is (AI)B-1, the area between column
lines 1, 2, A, and B. Following this logic, if we remove girder G-1, the beams it
supports will fail, and consequently the slabs supported by the beams. Thus, the
area between column lines 1, 2, B, and D (AI)G-1 will collapse. A similar study will
show that the inuence area for column C-1 is the area between column lines
1, 3, D, and F.
One variation of the live load reduction formula is given in Equation (2.1):
15
Lred L0 0:25 p 2:1
AI
where
Lred the reduced design live load per square foot of area supported by the member
L0 the unreduced design live load per square foot of area supported by the
member
AI the inuence area of the member in square feet
2.6 Continuity in Reinforced Concrete Construction 43
Equation (2.1) is applicable whenever AI > 400 ft2. The usage of live load
reduction is limited in that the reduction cannot exceed 50 % (Lred 0.5 L0) for
members supporting one oor and cannot exceed 60 % (Lred 0.4 L0) for members
supporting two or more oors. Live load reductions do not apply for live loads in
excess of 100 psf, except for members supporting two or more oors, in which case
the live load can only be reduced up to 20 %.
Example 2.2 For the interior beam of Example 2.1, determine the reduced
live loads.
Solution The inuence area, AI, for the beam is:
AI 2 9 30 540 ft2
Because this area is larger than 400 ft2, a reduced live load may be used in the
design of the beam. The reduced design live load is:
15
L 50 0:25 p 50 0:895 44:8 psf
540
Many readers may have encountered only statically determined structural elements.
These are simply supported beams (with or without cantilevers at their ends),
cantilevers xed at one end and free to move at the other, simple posts, and so
on. These elements P are all characterized
P by
Pneeding only the equations representing
static equilibrium H 0, V 0, M 0 to solve for the reactions.
A review of what reactions means may be needed here. A building element
does not exist in a stand-alone vacuum. It is connected to other elements. At a point
of connection the free relative displacement between the element under study and
the rest of the structure is denied. This denial of free movement results in the
transmission of a force (or moment) at the connection between the supporting and
the supported elements. Look at Figure 2.5a for example. Here a beam end is
supported on a wall. Elsewhere within the span the beam is free to deect, or move
vertically. But this ability to displace vertically is denied at the place of the support.
Figures 2.5b, c show the symbols of a hinge type of support and a roller. In the
hinge support, the two relative displacement components (vertical and horizontal)
are denied between the beam (the member under investigation) and the support
44 2 Rectangular Beams and One-Way Slabs
d e
a b c
Figure 2.5 The meaning of the different support conditions: (a) wall supporting a beam (roller),
(b) a hinge support, (c) a roller support, (d) wall supporting a beam (xed), (e) a xed support
below it. Thus, vertical and horizontal forces could be transmitted at the point
between the beam and the support. (The forces coming from the support to the
supported member are called reaction forces.) At a roller support (Figure 2.5c) only
relative vertical displacement is denied; the beam could still freely roll horizontally
without resistance. Correspondingly only a vertical force could be transmitted
between the beam and the support. Figure 2.5d shows a beam end built into a
large mass. The beam end cannot move horizontally or vertically, and it cannot
rotate with respect to the mass. This condition is called xity. The usual symbol of
xity is shown in Figure 2.5e. In this condition, horizontal force, vertical force, and
a moment may be transmitted between the member and the support at that location.
All of these support conditions are quite familiar to students who have had a rst
course in structures. These support conditions represent what may be called abso-
lute conditions: The displacement (vertical, horizontal, or rotational) is either freely
available, or completely denied. As will be pointed out later, there is an innite
number of conditions in between, especially as related to rotations. Consider, for
example, a exible joist supported by a wall or beam at its ends (Figure 2.6). The
mere supporting certainly precludes vertical displacement of the joist, thus a force
transfer occurs. An action force is transmitted from the joist to the wall or beam, and
an equal but opposite reaction force is transmitted from the supporting element to
the joist. As the joist deects under load, its supported ends can rotate freely; thus,
the moments at the ends are zero.
Reinforced concrete construction is monolithic, which means that members are
intimately built together with neighboring members. Slabs are continuous over
supporting beams and girders; beams and girders are continuous over supporting
interior columns, and so on.
Figure 2.7 illustrates the point. The slab in the beam and slab structure is
continuous in both horizontal directions over the beams. The beams are continuous
over other beams or columns.
A simple problem is presented here to clarify the concept. Admittedly, this
problem does not occur in reinforced concrete structures, but it serves to illustrate
the concept. A continuous structural member is represented by an imaginary center
2.6 Continuity in Reinforced Concrete Construction 45
a b c
w2
w1
Span 1 Span 2
Angle of rotation
Moment diagram
line (see Figure 2.8). On this two-span beam, Span 2 is larger than Span 1. If the
loads are about the same, Span 2 will deect more. Consequently this deection
will try to force Span 1 to curve upward slightly near the center support to follow
Span 2. (The tangent to the deformation curve will rotate toward Span 2.) Study of
the deformation curve shows that the beam bends into an upward curvature, that is,
tension develops at the top of the beam, between the two points of inection (where
the moment in the beam is zero), whereas elsewhere the beam bends downward,
resulting in tensions at the bottom. The moment diagram is shown below the
deformation line of the beam. The moments are referred to as positive when tension
is on the bottom, and negative when tension is on the top.
46 2 Rectangular Beams and One-Way Slabs
The deformation line in Figure 2.8 shows that the longer span (Span 2) will force
the beam to rotate toward itself at the center support. The resistance against this
rotation comes from the bending stiffness of the member in Span 1. Stiffness is the
ability of a member to resist deformation. There are several different types of
stiffness, such as exural, shear, axial, and torsional. Each type refers to a specic
ability to resist a certain type of deformation. The greater the stiffness, the more is
the effort required to bring about the specic deformation.
The exural stiffness of a member is linearly related to the moment of inertia, I,
which is a cross-sectional property, and to the modulus of elasticity, E, the ease of
extendibility or compressibility of the material; and is inversely related to the
EI
length, , of the member. Thus, if K represents the exural stiffness, K k ,
where k is a numerical constant that depends on the support conditions of the other
end of the member.
In the simple beam shown in Figure 2.8, if the exural stiffness of Span 1 is
innitely large, it will resist any attempt by Span 2 to rotate the section over the
center support toward itself. Hence the condition for Span 2 will approach that of
full xity at its left end. On the other hand, if the stiffness of Span 1 is very small, it
will offer very little resistance against the efforts of Span 2 to rotate freely at the
center support. Thus, as far as Span 2 is concerned, such a condition might be a
simple support, regardless of the continuity.
The free-body diagrams that resulted from the continuity are shown in Figure 2.9.
Double subscripts identify the locations of shears and moments. Thus, if the rst
span is from a to b then Vab represents the shear in that span at end a, and so on.
The two-span continuous beam is dissected to show the propagation of loads and
moments. Each cut shows every force and every moment as they act on the part
under consideration. For example, Mba is shown as a clockwise arrow on Span
1, whereas it is shown as a counterclockwise arrow on the small part over the
b support. These are two manifestations of the same moment, a concept well known
from Newtonian physics (action and reaction). Similarly, Vba is shown at the same
cut as an upward force on Span 1 that comes from the support to the beam, as well
as a downward force that comes from the beam to the support.
a b c
Vab Vba Vbc w2 Vcb
w1
Mab Mba Mbc Mcb
A B C D
13 14 15 16
9 10 11 12
5 6 7 8
1 2 3 4
P13-9 P14-10
V13-9 V14-10
M13-9 M14-10
M 9-13 M 10-14
V9-13 V10-14
P9-13 P10-14
V9-13
M10-14 V10-14
M 9-13
M9-5 M10-6
V9-5 V10-6
P9-5 P10-6
Figure 2.11 The propagation of forces and moments between beams and columns
The columns on the two upper oors, due to the forced rotations of their ends,
will bend into a double curve (S curve). Depending on the amount of xity available
at the footing level, the lower columns will bend either into a double curve when the
xity at the base is signicant, or into a single curve when the resistance against
rotation at the base approaches that of a hinge. P
P Figure 2.11P shows free-body diagrams for part of the frame. Again H 0,
V 0, and M 0 apply for each individual part. Thus, the axial force in
beam 1314 must equal the shear at the top of column 913 for Node 13 to be in
equilibrium. The axial force in the column equals the shear at the left end of beam
1314. And the moment at the end of column 913 must maintain equilibrium with
the moment at the left end of beam 1314. Mathematically:
X
For H 0 V 13-9 P13-14 0
X
For V 0 P13-9 V 13-14 0
X
For M 0 M13-14 M13-9 0
The reader may want to study and write out the equilibrium equations for other free-
body parts.
2.8 On the Fickleness of Live Loads 49
As stated earlier, loads permanently attached to the structure are referred to as dead
loads, and transitory loads are referred to as live loads. The nature of live loads is
that sometimes they are there and sometimes they are not, so it is entirely possible
that the live loads are fully present in one bay, while completely missing in other
bays. Figures 2.12ad show the effects of loading one span at a time on a four-bay
continuous beam. In each case the deformation and the moment diagram are shown
schematically under different live loading conditions. Deformations are shown as
dashed lines.
A study of the deformation lines and the moment diagrams of these four different
cases leads to the following conclusions:
1. The largest positive moments due to live loads in a given span occur when live
loads are on that span and on every second span on either side. This is known as a
checker-board pattern loading. See Figure 2.13a, b.
2. The largest negative moments due to live loads near a support occur when live
loads are on neighboring spans and on every other span on either side. See
Figure 2.13ce.
Thus, on a continuous beam the number of live loading patterns that result in
maximum moment effects equals the number of supports. For example, in a four-
span beam with ve supports, ve different live loading patterns need to be
considered to nd the possible absolute maximums in each of the positive and
negative moment zones.
These are only the moments that are due to the effects of the live loads. The
cases, shown in Figure 2.13ae must be combined with the moments resulting from
the dead loads, that is, the loads that are permanently present on the structure,
whose effects are not variable. The combinations of the dead load moments and the
live load moments will result in a maximum possible moment at every location
along the beam. The live and dead loads, when plotted into a graph such as the one
shown in Figure 2.14, produce a diagram that represents all these combinations.
This is called the diagram of maximum moments or the moment envelope.
Two important points must be noted here. Figure 2.14 shows that in some
portions of each span, only positive moments occur, and in others, only negative
moments, regardless of the distribution of the live loads. There are portions of each
span, however, where either positive or negative moments may occur. This fact is
signicant in that it affects how a continuous beam must be reinforced.
The second point is that so far we have assumed that the continuous beam is
similar to a mathematical line supported on knife-edge supports. The result of such
a simplied assumption is that the reactions appear as concentrated forces and the
moment diagram has a sharp peak (cusp) at those points. This result, however, is not
in conformance with the physical reality. Supports (columns) have a width over
which the reactions are distributed. This modies the moment diagram within
the width of the support to something similar to the sketch shown in Figure 2.15.
50 2 Rectangular Beams and One-Way Slabs
a
A B C D E
b
A B C D E
c
A B C D E
d
A B C D E
Figure 2.12 (a) The effects of live loads on span A-B. (b) The effects of live loads on span B-C.
(c) The effects of live loads on span C-D. (d) The effects of live loads on span D-E
The exact shape of the moment diagram at this location is quite immaterial, for both
theoretical studies and numerous test results clearly show that the critical negative
moments in the beam occur at the faces of the supports. (Refer to ACI Code,
Section 7.4.2.1 and Section 9.4.2.1)
2.8 On the Fickleness of Live Loads 51
Figure 2.13 (a) Live loads in the rst and third bays. Largest positive moments in rst and
third spans. (b) Live loads in the second and fourth bays. Largest positive moments in second and
fourth spans. (c) Live loads in the rst, second, and fourth bays. Largest negative moments
at second support. (d) Live loads in the second and third bays. Largest negative moment at
third support. (e) Live loads in the rst, third, and fourth bays. Largest negative moment at
fourth support
52 2 Rectangular Beams and One-Way Slabs
Live loads
Dead loads
Figure 2.14 Maximum moments due to dead loads and different combinations of live loads
Width of
support
Design moment
required by the
ACI Code
Some likely
distribution
Moment calculated
with knife-edge
support
The complexities involved in the design of a very simple continuous beam may
seem quite bewildering. In practice, however, a vastly simplied procedure is
available in most cases.
Any moment along a span may be expressed as follows:
Mu coefficient wu 2n 2:2
where
wu is the intensity of the total factored load (see Section 2.10), or the load per unit
length. This variable should be evaluated and applied separately for each span
if the live loads are different in each one
n is the net (clear) span for positive moment or shear, or the average of adjacent
net (clear) spans for negative moment
2.10 The Concept of Strength Design 53
When certain conditions are satised, the ACI Code permits the use of approx-
imate moments and shears in the design of continuous beams and one-way slabs in
lieu of the detailed analysis for maximum moments outlined in the previous section.
Approximate moments and shears usually provide reasonable and sufciently
conservative values for the design of these horizontal exural elements.
ACI Code Section 6.5.1 requires the following conditions for the use of these
coefcients:
There are two or more spans The beam or slab is continuous; that is, the
approximation does not apply to a single span only.
Spans are approximately equal, with the longer of two adjacent spans not
greater than the shorter by more than 20 % The larger span tends to pull the
shorter neighboring span upward if there are signicant differences between
adjacent spans.
Loads are uniformly distributed.
Unit live load does not exceed three times the unit dead load This is usually the
case with reinforced concrete structures.
Members are prismatic This means that the cross section is constant along the
length of the span.
The ACI Code design moments and shears are applicable when these precondi-
tions are satised. Table A2.1 and the accompanying gure list the coefcients for
the moments and shears according to the end conditions and number of spans. In the
authors experience, the ACI coefcients are somewhat more conservative than
values obtained from detailed computerized analysis; thus, their use will result in
additional safety for the structure.
In actual practice the use of simplied methods to nd the design moments and
shears is in decline. Many proprietary computer programs are available that not
only help evaluate all the most critical loading combinations, but also aid in the
design of the required reinforcing. These programs require the sizes of the members
as input, for the analysis of an indeterminate structure. (The result, or the output,
depends on the relative stiffnesses of the members.) Thus, the application of these
coefcients is still very useful for obtaining quick results that can be used in
preliminary sizing of the members, which in turn enables the development of
input data for a more detailed computerized analysis.
The rst design theory of reinforced concrete, developed near the end of the
nineteenth century, simply borrowed its approach from the prevailing theory of
elasticity. The method assumed that reinforced concrete elements at usual actual
loads will have stress levels that might be considered to fall within the elastic zone.
Figure 1.8 indicates that concrete in compression may follow an approximately
linear stress/strain relationship as long as the stress level does not exceed 50 % of its
54 2 Rectangular Beams and One-Way Slabs
ultimate strength level. Steel reinforcing behaves elastically below its yield point.
So the concept of working stress design (WSD) was not an unreasonable method-
ology, and the underlying calculation technique is still used when estimating
deformations (deections) in structural elements. (See Section 3.3 for a more
detailed discussion.)
The WSD method, however, has many conceptual drawbacks. First and fore-
most, it does not account for differences between dead and live loads. Rather, it
simply lumps them together and assigns a collective margin of safety, regardless
of the origin of the load. Dead loads can be estimated much more accurately than
can live (transitory) loads; thus, logic dictates that the part of the load that comes
from dead loads could use a much smaller safety factor against failure. On the other
hand, the magnitude and the distribution of the live loads are much more uncertain.
Another, and equally important, drawback of the WSD method is that it inac-
curately assumes that concrete behaves in a linear fashion with increasing stress
levels. Merely knowing a stress level does not ensure a correct prediction of an
undesirable level of stress (i.e., failure), because steel has a linear stress response to
strain whereas concrete has a nonlinear one.
The third, and perhaps the most signicant, drawback of the WSD method is that
it is unimportant to know the stress level in a structure at a given loading. What is
important is to know how much overload the structure can take before it fails.
Strength is needed to have a safe design, or adequate strength, so that the
structure does not fail whether the actually occurring loads were underestimated
or excess load is placed on the structure. Thus, load factors (i.e., values used to
magnify the actual loads [called working or service loads]), or moments or shears
therefrom, are used so as to create a demand on the strength. The concept of demand
states, for example, that the structure (or, more precisely, a given element under
investigation) must have an ultimate strength (i.e., before it fails) not less than those
given by Equation (2.3a) (ACI Code, Section 5.3.1).
U 1:4D
or U 1:2D 1:6L 0:5Lr or S or R
or U 1:2D 1:6Lr or S or R 1:0L or 0:5W
or U 1:2D 1:0W 1:0L 0:5Lr or S or R 2:3a
or U 1:2D 1:0E 1:0L 0:2S
or U 0:9D 1:0W
or U 0:9D 1:0E
where
U required (ultimate) strength
D effect from dead loads
L effect from live loads
W effect from wind loads
E effect from seismic (earthquake) loads
2.11 Design (Ultimate) Strength 55
where D includes the effects from all the dead loads and L is due to all the live loads.
For members that support both oor and roof loads (neglecting the effects of wind
or earthquake loads), the governing load combination from Equation (2.3a) are
(these typically include columns and walls):
The effects of uid, F, lateral earth pressure, H, and forces due to restraint of
volume change and differential settlement, T, can also be incorporated in the load
combination with their corresponding load factors. Refer to ACI Section 5.3 for
details.
The ultimate strength of a section within a structure (as discussed in detail later for
separate and combined cases of bending moment, shear, torsion, and axial load) is
calculated from the sizes (dimensions) of the section, the materials (steel and
concrete) employed, and the amount of reinforcing used. This calculation gives
56 2 Rectangular Beams and One-Way Slabs
us the supply, or the resisting strength furnished by the section. In exural design,
for example, this calculated quantity is designated as Mn, which is called nominal
moment strength or nominal resisting moment. Nominal strength is the calculated
strength, provided that everything goes according to plan; that is, the concrete is at
least as strong as assumed in the design, the dimensions of the beam, slab, or any
designed element is exactly as shown on the plans, the required reinforcing is
placed exactly where it was assumed in the calculations, and so on. But experience
shows that there is no such thing as perfectly executed plans, even in the best
circumstances. ACI 117-90, Standard Tolerances for Concrete Construction and
Materials lists tolerances that are reasonable to expect when good workmanship is
provided. Furthermore, the calculation processes employ simplied mathematical
models that should be considered as only reasonable approximations of reality. The
design methodology also tries to reect the relative importance of different struc-
tural components. The failure of columns, for example, may result in collapse of an
entire building, but the failure of a beam typically causes only limited local damage.
In light of all these possible detrimental effects to the assumed strength, a
strength reduction factor (-factor), sometimes referred to as an under-strength
factor, is introduced to the above dened nominal strength. This factor accounts for
the fact that the sections strength may be less than assumed in the analysis.
Thus, we arrive at the concept of useable strength (or supply), which is the
product of the nominal strength and the strength reduction factor.
Different factors are used for different types of effects. Equation (2.4) gives
some factors.
Flexure 0:90
Shear and torsion 0:75 2:4
Axial compression columns 0:65
Hence the ultimate strength design (USD) method can be stated as the following
inequality:
Demand Supply
or required ultimate strength useable design strength
or effects of loads resisting capacity of member
And so for a beam subjected to gravity (dead and live) loads, for example,
Equations (2.5)(2.8) represent this concept.
and
Mu Mn 2:6
2.11 Design (Ultimate) Strength 57
MR Mn 2:7
Mu MR 2:8
On the left side of Equation (2.8) is the demand. The demand depends only on
the span, the type of support (e.g., simply supported, cantilevered, etc.), and the
loads. All this information comes from the static analysis.
On the right side of Equation (2.8) stands the supplied strength of the section
(design resisting moment, MR), which depends on the size and shape of the cross
0
section, the quality of the materials employed (fc and fy), and the amount of
reinforcing furnished. Thus, the left side of the inequality is unique, but the right
side is undened. An innite number of different sizes, shapes, and reinforcing
combinations could satisfy a given problem. The only rule is that the supplied
useable strength be larger than (or at least equal to) the required strength.
Example 2.3 Assume that the beam in Example 2.1 is simply supported. Calculate
the required ultimate exural strength (factored moment from the loads). Use the
permitted reduced live load.
Solution
Thus:
The same result could be obtained by using factored loads (the loads multiplied by
their respective load factors).
and
Notice that when nding factored loads from service or working loads, the
nature of the loads does not change; only their magnitudes are multiplied by the
corresponding load factors. If a service load is distributed, its factored value is also
distributed; if the service load is concentrated, its corresponding factored load is
also concentrated. The following example claries this point.
58 2 Rectangular Beams and One-Way Slabs
PL = 10 kip
wL = 2.0 kip/ft
wD = 1.0 kip/ft
20'-0" 20'-0"
Pu 16 kip
wu1 4.4 kip/ft
wu2 1.2 kip/ft
20'-0" 20'-0"
Example 2.4 Determine factored loads for the beam shown in Figure 2.16.
Solution For the left half of the beam:
Pu 1:2PD 1:6PL
Pu 1:2 0 1:6 10 16 kip
To this point we have discussed the calculations for the left side of the design
Equation (2.8) (demand) in some detail. In this section we develop the right side of
the design equation. To establish the supply, or the ultimate exural strength, of a
2.12 Assumptions for the Flexural Design of Reinforced Concrete Beams 59
d
h
As
reinforced concrete section, we must discuss the stages of stress that a reinforced
concrete section experiences before reaching failure. This discussion of these
different stages of stress under increasing bending moments will also illuminate
the assumptions made in developing expressions for calculating the ultimate
strength of the section. To keep the discussion simple, we will examine a beam
with a rectangular cross-section like the one shown in Figure 2.18.
The symbols in Figure 2.18 will be used throughout this book. They are the
standard ACI symbols used with reinforced concrete. Thus:
b width of the section
h the overall depth of a section
d the effective depth of a section, or the depth from the centroid of the tension
reinforcement to the compression face
As the sum of the cross-sectional areas of the reinforcing bars
Notice that the reinforcement is not placed at the very bottom of the beam. The
rst and foremost reason for this placement is to provide corrosion protection to the
reinforcement. The inner environment of concrete is highly alkaline (high pH
value) and helps to protect the reinforcement. The concrete cover also provides
re protection to the reinforcement. Furthermore, the concrete surrounds the
reinforcing steel, which enables intimate bonding and allows the concrete and the
steel, two individual materials, to work together. The required minimum concrete
cover is given in Section 20.6.1.3.1 of the ACI Code. For unexposed beams it is
1.5 in. to the stirrups. (The stirrups, usually made out of #3 or #4 bars, will be
discussed in Chapter 4.)
Figure 2.19 shows a simply supported beam that has a simple rectangular cross
section made of plain concrete (homogeneous material). This type of beam is
almost never used in an actual building, but it will give us insight into the behavior
of concrete beams.
The uniformly distributed load (Figure 2.19a) represents the self weight plus
some superimposed load. The slightly exaggerated deected shape is shown in
Figure 2.19b, and the moment diagram in Figure 2.19c. Attention will be directed to
the section where the bending moment is the greatest. This location is where the
stresses and the strains are also the largest.
60 2 Rectangular Beams and One-Way Slabs
M max
b
Section Strains Stresses
Figure 2.20 shows the cross section of the beam and the distribution of strains
and stresses if the beam is unreinforced. Figure 2.21 illustrates the distribution of
the strains and stresses in a 3-D form. As long as the bending moments are
small, that is, the resulting tensile stresses at the bottom are less than the ultimate
tensile strength of the concrete, the section will behave as if it were made of a
homogeneous, quasi-elastic material. The bottom is in tension, and the top is in
compression.
Direct your attention to the strain diagram rst. Strain represents changes in
length. The strain distribution is linear from bottom to top.
The farther up or down a point is from the imaginary center, the greater the strain
in the beam. The largest tensile strains are at the bottom, whereas the largest
compressive strains are at the top. There is a line across the section where the strain
is zero. This is called the neutral axis. The straight-line distribution of strains is
known as the BernoulliNavier hypothesis. This distribution is called a hypoth-
esis because it results not from mathematical derivation, but from careful mea-
surements made on countless tests of many different materials, including concrete.
The distribution of stresses is also linear when the material follows Hookes law, as
2.12 Assumptions for the Flexural Design of Reinforced Concrete Beams 61
z 2h/3
steel does below the so-called proportional limit. Stresses are forces acting on a unit
area. Thus, it is possible to determine the resultant for these forces. The resultant,
which is a tensile (T) or compressive (C) force, is equal to the volume of the stress
block, For example, if the largest compressive stress is fcmax, then the sum of all the
compressive forces is given by Equation (2.9).
h i
C 1=2 fcmax =2 b
h
2:9
These resultants will be located at the centroid of the wedge-shaped stress blocks,
as shown in Figure 2.22. Equilibrium requires that these resultants be equal in
62 2 Rectangular Beams and One-Way Slabs
magnitude, and together they form an internal couple. The internal couple is
equivalent to the bending moment at the section.
Example 2.5 For the beam of Figure 2.22, assume b 12 in., h 24 in. and
Mmax 38.4 kip-ft. Determine the bending stresses and the equivalent tensile and
compression forces acting on the section.
Solution The section modulus is:
S b h2 =6 1,152 in:3
Then
460:8 kip-in:
C z T z 28:8 16 38:4 kip-ft
12
s
fs
Figure 2.24 Strain and stress distribution of a reinforced concrete beam prior to cracking
n.a.
s fs
cracks. At the location of the reinforcing steel, the concrete and the steel have
identical strains. The steel is bonded to the concrete, thus they must deform
together. But the two different materials respond differently to deformation because
they have a different modulus of elasticity, so the stresses will be different. In this
particular case the stress in the steel will be much larger than that in the concrete.
p
For example, assume a concrete with fc0 3,000 psi. Then Ec 57,000 3,000
3,122,000 psi 3,122 ksi. The modulus of elasticity of the reinforcing steel is
Es 29,000 ksi. According to Hookes law the stress equals the product of the
modulus of elasticity and the strain. So it follows that the stress in the steel will be
about nine times higher (the ratio of the two moduli of elasticity values) than the
stress in the concrete in the immediate vicinity. This ratio is usually designated as
n Es/Ec and is called the modular ratio.
The concrete cracks under increasing applied forces, and it is the reinforcement
that carries the tension across the crack. The crack travels up to a height, then stops
somewhere below the neutral axis as seen in Figure 2.25. The shaded area represents
the uncracked part of the section. Where the strains are still small near the neutral
axis, the concrete is still able to transfer some tensile stresses (albeit very small), even
in the cracked section; however, the amount of tensile force represented by the still
un-cracked tensile stress volume is so small that it is simply neglected.
Assuming, therefore, that the concrete does not carry any tension after cracking,
the bending moment in the section is transferred across from one side of the crack to
64 2 Rectangular Beams and One-Way Slabs
n.a.
s fs
T
the other via the tension in the steel and the compression in the concrete, as seen
in Figure 2.26. This assumption simplies the development of an appropriate
formula for the internal couple. The tensile component of this couple is at the
centroid of the reinforcing steel, while the compressive component is at the centroid
of the wedge-shaped compression block. Comparing Figures 2.22 and 2.26 indi-
cates that the T force now is concentrated at the centroid of the reinforcing.
In Figure 2.26 the compression stress block is represented as a triangular wedge
shape. This representation is more or less accurate as long as the compressive
stresses in the concrete remain quite low. Figure 1.8 shows the generic shapes of
the stress-strain curve of concrete in compression, and the assumption of linear
distribution of stresses may be justied up to approximately 0.5fc0 .
As the applied loads increase, there is a corresponding increase in bending
moments throughout the beam. Thus, many more sections away from the location
of the maximum moment will develop tensile stresses that exceed the concretes
ultimate tensile strength, resulting in the development of more cracks. While
theoretically the spacing between cracks is very small, it does not happen that
way, because the formation of a crack relieves tensile strains in the concrete in its
immediate neighborhood. Initially the cracks are very ne hairline cracks, and a
magnifying glass may be needed to locate them. These hairline cracks do not
indicate that there is anything wrong with the beam: They occur naturally in
reinforced concrete beams subjected to exure under normal working load condi-
tions. In fact, the reinforcement does not even do much work until after the concrete
has cracked.
As the bending moment at the section increases, the magnitude of T and C, the
tension and compression components of the internal couple, must also increase. In
the reinforcement this is simply reected as an increase in stresses. Correspond-
ingly, the steel also will experience greater strains and elongation. As long as the
strains in the reinforcing are less than the yield strain, the relationship between
stresses and strains remains linear.
In the concrete, however, the increased compression strains result in a nonlinear
response of the stresses while maintaining the required increase in the volume of the
stress block. The concrete stress block becomes more and more bounded by a
curvilinear surface. Ultimately, the contour will resemble the one shown in
Figure 2.27. This diagram is the same as the ones shown in Figure 1.8, except the
2.13 Different Failure Modes 65
Strains
0.003
0.002
0.001
Stresses
1,000
2,000
3,000
4,000
Figure 2.27 Typical curvilinear stress distribution in the concrete at ultimate strength
axes are reversed. At the origin, the strains and stresses are zero, just like on the
beam section at its neutral axis. At the top there is a strain value of 0.003, which is a
value selected by the ACI Code (somewhat arbitrarily) as the ultimate useful strain.
Somewhere between these two limits (in the neighborhood of 0.002) the peak stress
(the maximum compressive strength or simply compressive strength) occurs.
In calculations this value is designated as fc0 ; it is the specied compression strength
of the concrete, as already mentioned in Section 1.6.1.
On the tension side (i.e., at the reinforcement), Figure 1.17 shows the stress-strain
curve of the reinforcing steel, or the response of the steel to increasing strain values.
This curve clearly shows that the steel has signicant residual strength even after it
has yielded, but this residual strength (the strength gained in the strain hardening
zone) is neglected. Thus, we assume that the stresses will linearly increase with
increasing strains up to yield, after which ever-increasing strains produce no
corresponding increase in stresses. Scientically, this curve is known as a bilinear
stress-strain diagram, and the response of the steel as elasto-plastic behavior. Fig-
ure 2.28 shows the assumed stress-strain diagram for 40 and 60 ksi steel, respectively.
As a rst case assume that a beam has a relatively small amount of reinforcing steel.
Such a beam is shown in Figure 2.29. With increasing demand on the internal
couple the stresses in the steel will reach yield before the demand on the concrete
66 2 Rectangular Beams and One-Way Slabs
Stress
60 ksi
40 ksi
Strain
0.00138
0.00207
Compression zone
Yielding
b
0.003
cb
fy/E
A quick look at Figure 2.27, or at its 3-D representation in Figure 2.31, should
convince anyone that it would be impractical to calculate the value of C by guring
out the volume of the stress block. The calculation would require integral calculus,
68 2 Rectangular Beams and One-Way Slabs
T
Figure 2.31 True stress distribution in the concrete at ultimate strength
even if there was an easy way to express the shape of the curve mathematically.
A reasonable approximation can be obtained by substituting a stress block whose
volume is about the same as the true stress volume enclosed in Figure 2.31, and
whose centroid is fairly close to that of the true stress volume. This is known as the
equivalent stress block, and is shown in Figure 2.32.
The relationship between the true stress block and the equivalent stress block has
been established by studying many concrete stress-strain curves. The simple rect-
angular block has been adopted for its simplicity and ease of calculation. If a
uniform stress value of 0.85fc0 is adopted, then only the relationship between the
depth of the equivalent stress block a and the distance of the neutral axis from the
top c is needed. This relationship is given in Equation (2.11).
a 1 c 2:11
To account for the somewhat different shapes of the stress-strain curves of different
strengths of (refer to Figure 1.8) concrete, 1 is given by the ACI Code (Section
22.2.2.4.3) as follows:
2.14 The Equivalent Stress Block 69
b 0.85 fc
1 0.85 for concrete strength fc0 up to and including 4,000 psi. For strengths above
4,000 psi, 1 shall be reduced at a rate of 0.05 for each 1,000 psi
of strength in excess of 4,000 psi, but 1 shall not be taken less than 0.65
Equation (2.12) gives the expression to calculate 1 for fc0 > 4,000 psi.
0
fc 4,000
1 0:85 0:05 0:65 2:12
1,000
The equivalent stress block makes it extremely easy to manipulate the expression to
calculate the ultimate (design) resisting moment of a given section. The moment
arm of the internal couple, z, can be calculated using Equation (2.13).
a
zd 2:13
2
The numerical value of the internal couple can be expressed in two different
ways, using the designation of Mn for the nominal resisting moment and MR for the
design resisting moment. These moments can be calculated using Equations (2.14)
and (2.15), respectively.
Mn Tz or Mn Cz 2:14
MR Mn Tz Cz 2:15
70 2 Rectangular Beams and One-Way Slabs
where
As fy 0:85fc0 ab 2:16
Solving this equation for a gives Equation (2.17) for calculating the depth of the
equivalent stress block.
As fy
a 2:17
0:85fc0 b
As
2:18
bd
0.003
cb
d N.A.
dt
dt cb
ty
Section 2.13 discussed the two possible different failure modes of reinforced
concrete beams in bending. The theoretical dividing point between them, the
balanced failure, was also discussed. In this case the steel in the outermost
layer (if there is more than one layer) reaches its yield strain exactly when the
maximum compressive strain in the concrete reaches the 0.003 value. The strain
distribution at balanced failure resembles the one shown in Figure 2.33. In order to
cover the more general (although not so frequent) case of multilayer reinforcing in
the beam, a distinction is made between d, the working depth, and dt, the depth to
the outermost layer of reinforcing on the tension side. When there is only one layer
of reinforcement, d dt.
From the similarity of the two triangles above and below the neutral axis,
cb, the depth of the neutral axis at balanced failure can be expressed as a function
of dt and fy.
cb 0:003
2:19
d t cb ty
Solving for cb
0:003dt
cb 2:20
0:003 ty
because
fy fy
ty 2:21
Es 29,000,000
72 2 Rectangular Beams and One-Way Slabs
87,000
cb dt 2:22
87,000 fy
Asb fy
a b 1 c b 2:23
0:85fc0 b
87,000dt b bdfy
1
87,000 fy 0:85fc0 b
the steel ratio for balanced failure, b, can be calculated using Equation (2.24).
0:85fc0 87,000 dt
b 1 2:24
fy 87,000 fy d
If dt d, which means there is only one layer of reinforcing steel (by far the most
frequent case), then Equation (2.24) becomes Equation (2.25).
0:85fc0 87,000
b 1 2:25
fy 87,000 fy
Note that the value of b depends only on the selected materials ( fc0 and fy) and is
dt
independent of the size of the section. (The ratio becomes necessary only when
d
there is more than one layer of reinforcement.)
In an effort to generalize the approach for members subject to both bending and
axial compressive forces, the ACI Code strives to treat these combination cases
together. The different failure modes were discussed in Section 2.13. These modes
are distinguished by whether the primary failure is due to yielding of the steel or to
2.17 Elaboration on the Net Tensile Strain in Steel (t) 73
A1 B1t
C D
0.90
B
A
0.65 Flexural members
Compression
controlled Transition Tension controlled
t
ty 0.004 0.005
crushing of the concrete. The former is called tension-controlled failure, and the
latter is compression-controlled failure. It was also previously noted that tension-
controlled failure results in highly desirable ductility, whereas compression-
controlled failure is abrupt and nonductile in nature. Unfortunately, as will be
discussed later in Chapter 5, the desire to have only ductile tension-controlled
failure modes cannot always be satised. But in exural members, at least we can
control the failure behavior by using no more steel than an amount that ensures the
desirable ductility. In the past this was accomplished by limiting the reinforcement
ratio, , to 75b in exural members. Since 2002 the ACI Code has adopted a new
approach that is a better integration of dealing with members subject to axial
stresses whether from exure, or axial compression, or both. If ductile failure
mode cannot always be assured, then the use of a larger safety factor against a
nonductile type of failure is warranted. This larger safety factor is obtained by
regulating the ratio between the useful ultimate moment or design resisting moment
(MR Mn) and the nominal ultimate moment (Mn). This requires only an adjust-
ment in the (strength reduction) factor.
The ACI Code (Section 21.2.2) denes three different types of section behavior:
tension-controlled, compression-controlled, and a transition zone, which is the
zone between the tension- and the compression-controlled failure zones. Figure 2.34
shows a graphical representation of these zones, and denes and separates the three
regions. Theoretically the division between compression-controlled failure and
tension-controlled failure is where t ty. In other words, the section is
compression-controlled if the strain in the steel is less than the yield strain; and is
tension-controlled if the strain in the steel is greater than the yield strain when the
compression strain in the concrete reaches the limit of 0.003. For design purposes,
however, the ACI Code requires a safely assured tension-controlled section; thus, it
74 2 Rectangular Beams and One-Way Slabs
a b c
c 0.003 c 0.003 c 0.003
c
c
c
d dt
t ty t 0.005 ty t 0.005
Figure 2.35 Strain distribution and net tensile strain (t) at behavior limits: (a) compression-
controlled sections; (b) tension-controlled sections; (c) transition-controlled sections
denes a section as tension-controlled only when the steel strain at ultimate strength
is greater than 0.005. Between the two limits, yield strain (ty) and 0.005, the Code
denes a transition zone with lowered values.
Note that the ACI Code allows t for exural members to be as small as 0.004 at
ultimate strength. A somewhat diminished factor, however, is required in
conjunction.
It may be helpful here to repeat what was discussed in Section 2.13 in a
somewhat different format. Figure 2.35 denes graphically the behavior of
reinforced concrete sections.
1. A compression-controlled section is a reinforced concrete section in which the
strain in the concrete reaches 0.003 at ultimate strength, but the strain in the steel
(t) is less than the yield strain (ty). (See Figure 2.35a.) In other words, at the
ultimate strength of the member, the concrete compressive strain reaches 0.003
before the steel in tension yields. This condition results in a brittle or sudden
failure of beams and should be avoided. In reinforced concrete columns, how-
ever, a design based on compression-controlled failure behavior cannot be
avoided. As shown in Figure 2.34, 0.65 is mandated for this case, which is
considerably less than the 0.90 that is used for tension-controlled sections.
The reasons for this additional factor of safety are: (1) compression-controlled
sections have less ductility; (2) these sections are more sensitive to variations
in concrete strength; and (3) the compression-controlled sections generally occur
in members that support larger loaded areas than do members with tension-
controlled sections.
2. A tension-controlled section is a reinforced concrete section in which the tensile
strain in steel (t) is more than 0.005 when the compression strain in concrete
reaches 0.003 (see Figure 2.35b). In other words, when a section is tension-
controlled at ultimate strength, steel yields in tension well before the strain in the
concrete reaches 0.003. Flexural members with tension-controlled sections have
2.18 The Location of the Neutral Axis and Limit Positions 75
ductile behavior. As a result, these sections may give warning prior to failure by
excessive deection or excessive cracking, or both. Not all tension-controlled
sections will give both types of warning, but most tension-controlled sections
should give at least one type of warning. Both types of warnings, excessive
deection and cracking, are functions of the strain, particularly the strain on the
tension side. Because tensile strains are larger than compressive strains in
tension-controlled sections at failure, the ACI Code allows a larger factor
(0.90) for these types of members.
3. A transition-controlled section is a reinforced concrete section in which the net
tensile strain in the steel (t) is between yield strain (ty) and 0.005 when the
compression strain in the concrete reaches 0.003. (See Figure 2.35c.) Some
sections, such as those with a limited axial load and large bending moment,
may have net tensile strain in the extreme steel (t) between these limits. These
sections are in a transition region between compression- and tension-controlled
sections. In Figure 2.34, the line AC represents the Code-dened relationship
between and t in the transition-controlled zone. The value of in the
transition zone can be calculated using Equation (2.26).
A 1 B1 t 2:26
0:00325 0:9 ty
A1
0:005 ty
0:25
B1
0:005 ty
Table A2.2a in Appendix A lists the values for the coefcients A1 and B1 for
commonly used reinforcing steels.
Consider the strain diagram shown in Figure 2.36. The location of the neutral axis at
ultimate strength (c) depends upon the net tensile strain of the steel. Observe the
solid and the dotted lines. Because the strain at the compression face is constant
(0.003), c becomes smaller as the steel strain increases. Using similar triangles of
the strains above and below the neutral axis, an expression can be derived to
calculate the depth of the neutral axis, c.
c 0:003
dt c t 2:27
ct 0:003dt c
76 2 Rectangular Beams and One-Way Slabs
ec 0.003
dt
et
Figure 2.36 Variation of the location of the neutral axis (c) with the tensile strain in steel (t)
0:003
c dt 2:28
0:003 t
The ratio of c/dt, given in Equation (2.29), is often used to check if a section is
tension-controlled.
c 0:003
2:29
dt 0:003 t
Two values of t are of special interest. The rst one is t 0.004. This is the absolute
minimum steel strain permitted by the ACI Code for members in exure. (Refer to
Figure 2.34 and ACI Code, Section 7.3.3.1 for one way slabs, and Section 9.3.3.1 for
beams). Substituting this t value into Equation (2.29) gives us Equation (2.30).
c 3
0:429 or c0:429dt 2:30
dt 7
Equation (2.30) gives the lowest permissible value of the neutral axis depth. In
other words, this denes the largest permissible concrete area in compression
(c 0.429dt).
The second value of interest is t 0.005. Solving Equation (2.29) for this case,
we obtain Equation (2.31) for the lowest location of the neutral axis depth for
tension-controlled sections.
c 0:003 3
0:375 or c0:375dt 2:31
d t 0:003 0:005 8
2.20 Limitations on the Steel Percentage () for Flexural Members 77
A2 B2 ( dct ) C D
0.90
A B Flexural members
0.65
Compression
controlled Transition Tension controlled
dt
dt 7 8 c
cb 3 3
dt
Figure 2.37 Variation of versus
c
Equation (2.29) shows that the ratio of either c/dt, or its inverse, dt/c, are in direct
relationship with the steel tensile strain t. Then it is possible to modify Figure 2.34
to show the ACI Codeprescribed strength reduction factors (the factors)
variation in terms of the dt/c ratio. (For convenience of graphing, the relationship
is shown in terms of dt/c.) Figure 2.37 expresses the changing values with respect
to the ratio dt/c. Note that the ratio dt/cb is the ratio of d/c at the balanced failure
point.
Table A2.2b in Appendix A of this text lists the values for the coefcients A2 and B2
that describe the variations in values through the transition zone. The limiting ratios
between the depth of the member and the location of the neutral axis (dt/c) and its
inverse at the balanced failure point (i.e., dt/cb or cb/dt) are also included.
With the help of Equations (2.30) and (2.31), the corresponding largest values (i.e.,
the steel percentages that satisfy those limiting conditions) can be determined. For
t 0.004 (lowest permitted steel strain value at ultimate strength of exural mem-
bers), the maximum depth of the neutral axis is calculated using Equation (2.32).
78 2 Rectangular Beams and One-Way Slabs
3
cmax d t 2:32
7
The corresponding depth of the equivalent stress block (refer to Equations (2.11)
and (2.17)) is given by Equation (2.33).
As, max fy 3
amax dt 2:33
0:85fc0 b 7 1
max bdfy 3
1 d t 2:34
0:85fc0 b 7
3 f 0 dt
max 0:851 c
7 fy d
or
fc0 dt
max 0:3641 2:35
fy d
fc0
max 0:3641 2:36
fy
In a similar way, we can determine the value of that will ensure an t 0.005,
the upper limit of needed to ensure a tension-controlled (ductile) failure in beams
at their ultimate strength. Designate this value of as tc. After changing the right
side of Equation (2.34) accordingly (see Equations (2.30) and (2.31)), then the
value of tc can be calculated using Equation (2.38) (or Equation (2.39) for the
special case of a section with only one layer of reinforcement).
tc bdfy 3
0
1 d t 2:37
0:85fc b 8
2.21 Minimum Steel Ratio (min) for Reinforced Concrete Beams 79
fc0 d t
tc 0:3191 2:38
fy d
fc0
tc 0:3191 2:39
fy
Table A2.3 in Appendix A lists the values of max and tc for various grades of steel
( fy) and concrete strength (fc0 ) combinations. The value of the strength reduction
factor () is shown in the right column of the table. This value varies when
tc < < max, or the beams failure mode is in the transition zone (see
Section 2.17). Table A2.3 indicates that not much is gained in terms of useable
moment capacity with the required reductions in the values and when the
reinforcing percentage is increased from tc to max, especially when higher
strength steels are used.
This minimum amount of steel (As,min) provides enough reinforcement to ensure that
the moment strength of the reinforced concrete section is more than that of
an unreinforced concrete section, which can be calculated from its modulus of rupture.
200
In the past, the ACI Code required only an As, min bd. For concrete strength
fy
greater than about 4,440 psi, however, this is not sufcient to ensure the desired aim;
80 2 Rectangular Beams and One-Way Slabs
p
3 fc0
bd recties this condition. Because As,min min bd, Equation (2.40) may be
fy
expressed mathematically in terms of min (minimum steel ratio) as shown in
Equation (2.41).
( p )
3 fc0 200
min max , 2:41
fy fy
Table A2.4 in Appendix A provides values of min for different grades of steel
and compressive strengths of concrete.
This method closely follows the already discussed and established formulae.
Figure 2.38 shows the stress and strain distributions for a reinforced concrete
rectangular beam at ultimate strength. For the most general case, a beam section
with multilayer reinforcing is shown.
The resisting moment can be calculated from the internal couple and using
Equations (2.42)(2.44).
2.22 Analysis of Rectangular Reinforced Concrete Sections 81
a b
0.85fc
b
c
a C
c
d
dt
z
fy
T
t ty
T As fy C 0:85fc0 ba z d a=2
2:42
Mn Tz As fy d a=2
MR Mn 2:44
Step 3. Calculate the location of the neutral axis from Equation (2.11):
a
c
1
82 2 Rectangular Beams and One-Way Slabs
B2
A2 c refer to Table A2:2b for A2 and B2
dt
a
Step 5. Calculate MR Mn As fy d refer to Equations 2:42 and
2
2:44
Figure 2.39 summarizes the analysis steps.
1.
As No Beam is illegal ( min)
min?
bd per current ACI Code.
max?
Yes
2. Asfy
a
0.85fc b
No
3. a
c
1
4. Warning! max
c 3 Only the part of the reinforcing that
? is equal to max may be taken into
dt 8
account in the nominal strength calculations.
No
Yes
0.90 B2
A2 c
dt
5.
MR Asfy (d a)
2
Example 2.6 Use Method I to determine the design resisting moment, MR, of
the reinforced concrete beam section shown below. fc0 4 ksi, and fy 40ksi.
The reinforcement is 3 #9 bars, As 3.00 in2.
b = 10 in.
d = 22 in.
Step 3. From the depth of the equivalent stress block, determine the location of the
neutral axis, c:
a 3:53
c 4:15 in:
1 0:85
Example 2.7 Repeat Example 2.6 for fy 60 ksi, and fc0 3 ksi.
Solution
Step 1.
As
0:0136
bd
Table A2:4 ! min 0:0033 < 0:0136 ok
Table A2:3 ! max 0:0155 > 0:0136 ok
Step 2.
As fy 3 60
a 0
7:06 in:
0:85fc b 0:85 3 10
Step 3. a 7:06
c 8:30 in:
1 0:85
Step 4.
c 8:30
0:377 > 0:375
dt 22
B2
A2 c
dt
0:25
0:233 0:90
0:377
Step 5.
7:06
0:90 3 60 22
2
MR 249 ft-kip
12
2.22 Analysis of Rectangular Reinforced Concrete Sections 85
Example 2.8 Calculate MR for the reinforced concrete beam section shown below.
fy 60 ksi, fc0 4 ksi, As 7.62 in2.
12 in.
6 # 10
Solution
Step 1.
As 7:62
0:0203
bd 12 31:25
From Table A2:4 ! min 0:0033 < 0:0203 ok
From Table A2:3 ! max 0:0207 > 0:0203 ok
Step 2.
As fy 7:62 60
a 0
11:21 in:
0:85fc b 0:85 4 12
Step 3.
a 11:21
c 13:19 in:
1 0:85
Step 4.
c 13:19
0:406 > 0:375
dt 32:5
B2 0:25
A2 c 0:233 0:85
0:406
dt
Step 5.
11:21
0:85 7:62 60 31:25
2
MR 831 kip-ft
12
This method results in the development of design aid tables, which are more user-
friendly. The tables will also be useful when the aim is to design beam sections to
satisfy a given Mu demand instead of analyzing.
The expressions for the components of the internal couple are
T As fy C 0:85fc0 ba z d a=2
As fy
a
0:85fc0 b
Substituting from Equation (2.18), As bd. Equation (2.45) can be used to calcu-
late a.
bdfy dfy
a 2:45
0:85fc b 0:85fc0
0
dfy
1 c
0:85fc0
2:46
fy d
c
0:85fc0 1
or
c fy d
2:47
dt 0:85fc0 1 dt
If 3/8 c/dt 3/7 the strength reduction factor, , must be adjusted accordingly.
B2
A2 c
dt
Substituting for c/dt from Equation (2.47), Equation (2.48) can be used to
calculate the adjusted strength reduction factor.
0:85fc0 1 dt
A2 B 2 2:48
fy d
Equation (2.48) provides the values of in the transition zone. In order to simplify
the equation, introduce a new steel ratio, t:
As
t
bdt
(Note that dt d and t when the beam has only a single layer of steel.)
dt
Substituting t Equation (2.48) can be rewritten as Equation (2.48a).
d
0:85fc0 1
A2 B2 2:48a
t fy
MR Mn Cz
a
MR 0:85fc0 ba d
2
fy
R fy 1 2:50
1:7fc0
MR bd2 R 2:51
It is clear from Equation (2.50) that R depends on the materials used (i.e., fc0 , fy
and the steel ratio () in the beam), but it is independent of the dimensions of the
section. Thus tables for R can be developed in terms of for the various combina-
tions of materials. Values of R can be found from Tables A2.5 through A2.7. min
for beams are indicated on each table. Reinforcement ratio () values less than min
may not be used in beams, but may be used in slabs and footings.
These tables were developed with R in psi. Using R in psi and beam dimensions
b and d in inches results in lb-in. units for MR. Because kip-ft are usually used in
moment calculations, appropriate conversions must be made between lb-in. and
kip-ft for the correct use of the tables.
Rpsi
MR ft-kip b in:d in:2
12,000
The tables must be used with care, especially when large values result in
the section being in the transition zone. The value of depends on fc0 , fy, , and
dt
, thus if
d
tc ! 0:90
and if
0:85fc0 1 d t
max > tc ! A2 B2
fy d
The values of tc and max for common grades of steel and concrete strength are
listed in Table A2.3.
An important note here is that Tables A2.5 to A2.7 have been developed based
on (i.e., beams with a single layer of reinforcement). If the beam has multiple
layers of reinforcement (t 6 ), the R value must be modied by adjusting it to an
R0 value based on t. This can be easily done by using Equation (2.51a).
0
R0 R 2:51a
The values of 0 , which are listed in Tables A2.5 to A2.7, correspond to the
values of t.
2.22 Analysis of Rectangular Reinforced Concrete Sections 89
The use of Method II for analysis of reinforced concrete beam sections involves
the following steps:
Step 1. Determine whether min; if not, then the beam does not satisfy the
minimum requirements of the ACI Code and its use for load carrying is not
permitted.
Determine whether max; if not, the beam has too much reinforcing
and does not satisfy the latest ACI Codes limitations. A practical solution
for this is to disregard the excessive amount of reinforcement, assume that
the section is in the transition zone, and continue the calculations with the
maximum permissible amount of reinforcing.
Step 2. Use , fc0 and fy to obtain R and values from the appropriate Tables A2.5
to A2.7. If the beam has a single layer of steel or 0.90, nd MR from
Step 3. Otherwise move to Step 4.
bd2 R
Step 3. Calculate MR b, d in:; R psi; MR ft-kip
12,000
As
Step 4. For beams with multiple layers of reinforcement, calculate t and
bdt
obtain the corresponding strength reduction factor (0 ) from Tables A2.5 to
A2.7.
0
Step 5. Calculate the modied value of the coefcient of resistance R0 R .
bd2 R0
Step 6. Calculate MR .
12,000
The owchart for Method II is shown in Figure 2.40.
Method II has fewer steps to follow, so it is easier to use. Method I, however, is
more general as it does not require the use of design tables (which may not be
readily available) and it is adaptable to any grade of steel or compressive strength of
concrete, not just the ones listed in the tables.
90 2 Rectangular Beams and One-Way Slabs
1.
min,beam is illegal per ACI Code.
As No max,only the part of the reinforcing
min bd max? that is equal to max, may be taken into
account in the nominal strength calculations.
Yes
2. Use , f , f to find R and from
y c
Tables A2.5 through A2.7.
Beam has No
single layer of
bars (dt d )?
Yes
3.
bd 2R Yes
MR 0.90?
12,000
No
4. As
Use t to find from
bdt
Tables A2.5 through A2.7.
5. R R
6.
bd 2R
MR
12,000
Step 2. Using 0.0136, fy 40 ksi, and fc0 4 ksi, obtain the resistance coef-
cient, R, from Table A2.5b:
bd 2 R 10 222 450
MR
12,000 12,000
MR 182 ft-k
bd2 R
MR
12,000
10 222 615
MR 248 ft-kip
12,000
0:0203
fc0 4 ksi ! Table A2:6b ! R 825 psi
fy 60 ksi 0:82
Step 3. Because the beam has two layers of reinforcement and is not equal to
0.90, determine t and 0 and adjust the resistance coefcient, R:
92 2 Rectangular Beams and One-Way Slabs
Step 4.
As 7:62
t 0:0195
bdt 12 32:5
From Table A2:6b ! 0 0:85
0 0:85
R0 R 825 855 psi
0:82
This result is about the same as that from using Method I. The difference
is insignicant and is due to rounding errors in the calculations.
The selection of a beams depth is almost always a controversial issue. On the one
hand, the building designer wants to minimize the depth of the structure in order to
maximize the headroom without unduly increasing the height of the building. On
the other hand, structural elements that are too shallow lead to increased short- and
long-term deections. These, in turn, may be detrimental to attached nonstructural
building elements. Excessive deections of concrete structures may result in
cracked walls and partitions, non-functioning doors, and so on.
To guide in the design of well-functioning structures, the ACI Code (Sections
7.3.1.1 and 9.3.1.1) recommend a set of span/depth ratios, with the comment that
the designer does not have to calculate deections (an involved and somewhat
uncertain process) if the utilized depth is at least equal to the values provided in ACI
Table 7.3.1.1 for one-way slabs, and Table 9.3.1.1 for beams. These values are
summarized graphically in Figures 2.41 and 2.42.
Note from Figures 2.41 and 2.42 that the recommended minimum depth for
simply supported beams is span/16, whereas for one-way slabs this value is span/20.
These types of support conditions are quite rare in monolithic reinforced concrete
construction, because in most cases either continuity or some other type of restraint
is available at the supports. If the member is continuous at both ends, hmin span/21
for beams and hmin span/28 for one-way slabs. Finally, if the beam is continuous
at only one end, the minimum depth is span/18.5, and for one-way slabs is span/24.
A cautionary note is in order here. Span 2 (2) in Figures 2.41 and 2.42 is shown
as Both ends continuous. This assumption is valid only if the cantilever at the left
end of 2 is long enough to develop a signicant end moment. Experience shows
2.23 Selection of Appropriate Dimensions for Reinforced Concrete Beams. . . 93
Simply-supported
hmin /
1 8 hmin /
2 21 hmin /
3 21 hmin /
4 18.5
1 2 3 4
Cantilever Both ends continuous Both ends continuous One end continuous
hmin
hmin /20
Simply-supported
hmin /
1 10 hmin /
2 28 hmin /
3 28 hmin /
4 24
1 2 3 4
Cantilever Both ends continuous Both ends continuous One end continuous
Figure 2.42 Minimum depth requirements for reinforced concrete one-way slabs
that when the cantilever length is at least 2/3, the span 2 may safely be assumed as
both ends continuous from the point of view of satisfactory deection control.
The values shown in Figures 2.41 and 2.42 are applicable only to normal-weight
concrete (wc 145 lb/ft3) and Grade 60 reinforcement. For other conditions, the
ACI Code Section 7.3.1 recommends the following modications:
(a) For lightweight concrete in the range of 90115 pcf, the values in Figures 2.41
and 2.42 need to be multiplied by (1.65 0.005wc) where wc unit weight of
concrete in lb/ft3. This factor should not be less than 1.09. For a typical
lightweight structural concrete, wc 115 pcf. Then the multiplier is
1.65 0.005 115 1.075 < 1.09. Use a multiplier equal to 1.09.
(b) For fy other than 60,000 psi, the values obtained from Figures 2.41 and 2.42
shall be multiplied by:
94 2 Rectangular Beams and One-Way Slabs
fy
0:4 2:52
100,000
If the selected beam depth is less than the recommended hmin, the beam deec-
tion has to be calculated and checked against the ACI Code requirements. There-
fore, if a beam does not satisfy the minimum depth requirements, it may still be
acceptable if computation of deection proves it to be satisfactory.
Note that in most building structure applications (save for footings and founda-
tions) the usual concrete mix limits the size of the aggregate to in. Thus, a 1 in.
minimum spacing satises the third of the spacing requirements.
a b
s min
s min
1 in.
Figure 2.43 Minimum spacing between reinforcing bars: (a) single layer; (b) multiple layer
2.24 Crack Control in Reinforced Concrete Beams and One-Way Slabs 95
bmin
Minimum Space for Multiple Layers of Bars Where reinforcement is placed in two
or more layers (see Figure 2.43b), bars in the upper layers shall be placed directly
above bars in the lower layer with clear distance between layers not less than 1 in. In
addition, the requirements of single-layer bars must also be satised.
Minimum Width (bmin) of Reinforced Concrete Beams We use the minimum
required space between bars in a single layer to calculate the minimum beam
width needed to provide enough room for a specic number and size of bars. To
compute bmin, consider Figure 2.44. Usually #3 or #4 bars are used for stirrups.
Also, the minimum cover for bars in beams is 1.5 in. Therefore, we can calculate
bmin by adding the minimum required spaces and the bar diameters.
As an example, suppose that the beam in Figure 2.44 is reinforced with 4 #8 bars.
Assuming #4 stirrups, the minimum width for this beam is:
1
bmin = 2 1.5 in. + 2 in. + 4 1 in. + 3 1 in. = 11 in.
2
Cover Stirrups Main bars smin
Note that smin 1 in. was used; this assumes that 4/3 of the maximum aggregate
size is less than or equal to 1 in. Table A2.8, based on the above example, shows
bmin for different numbers and sizes of bars in a single layer.
It was previously mentioned that a reinforced concrete member will always crack
when subjected to bending. In fact, the reinforcing really starts working only after
the development of cracks. Nevertheless, designers try to minimize the size of the
cracks. Limitation of crack width is desirable for three main reasons: (1) appear-
ance; (2) limitation of corrosion of the reinforcement; and (3) water-tightness.
96 2 Rectangular Beams and One-Way Slabs
Laboratory experiments have shown that several parameters inuence the width
and spacing of exural cracks. The rst is the concrete cover over the reinforcing.
The smaller the cover, the smaller the crack width will be. The cover cannot
be reduced beyond a certain limit, however, because a minimum cover is needed
for re and corrosion protection. Thus, the Code requires a minimum cover
of 1.5 in. over the stirrups for interior beams, 2 in. for exposed exterior beams
(see Figure B2.2 in Appendix B) and in. for joists and slabs. The 1.5 in. cover
over the stirrups results in a cover of 1 7=8 in. to 2 in. over the main reinforcement.
The second important parameter is the maximum stress in the reinforcement
(directly related to the strain, or the elongation of the steel) at service load levels.
This value may be assumed to be roughly 0.66 fy. The higher the stress level is in the
steel, the wider the cracks are expected to be. Thus, using more reinforcing than
required to satisfy the ultimate strength capacity can reduce the width of cracks by
reducing the stresses (and strains) at working load levels. This is not an economical
choice, however. The same is true if steel with fy 40,000 psi is used instead of
steel with fy 60,000 psi. The section would need 50 % more steel, but the much
lower levels of stress at service load levels would help limit the crack width.
Another important parameter is the maximum spacing of the reinforcing bars.
For minimizing the width of cracks, placing more and smaller bars closer together is
preferable to placing a few large bars farther apart. The ACI Code (Sections 24.3.2,
and 24.3.3) limits the maximum spacing of the tensile reinforcement in beams and
one-way slabs. The empirical formula for maximum spacing, given in Equa-
tion (2.53), is based on the tensile stress in the steel and the concrete cover.
40,000 40,000
s 15 2:5cc 12 2:53
fs fs
a b
w1 w2
Concrete surface
ws cc1
c c2
Rebar
ws
Rebar
40,000 40,000
s 15 2:5cc 12
fs fs
! !
40,000 40,000
s 15 2 2:51:5 0:5 12 2
=3 60,000 =3 60,000
s 10 in: 12 in: ! s 10 in:
The ultimate strength of a beam depends on ve parameters. These are the materials
( fc0 and fy), the dimensions of the section (b and d ), and the amount of reinforcement
(As). The last three parameters may be expressed in the form of the steel ratio
As/bd.
Whichever way these parameters are expressed, they are always ve in number.
There is only one equation (or, more precisely, one inequality), however, that
expresses the problem:
Mu MR
The left side of this inequality depends only on the applied loads. The right side
of the inequality, on the other hand, depends on all ve of the variables listed above.
Thus, this problem has an innite number of solutions. But if four out of the ve
parameters are preselected or assumed, the inequality can be readily solved.
As an example, contemplate the following considerations. In a oor of a given
structure, it would be quite impractical to vary the quality of the concrete. Conse-
quently, every beam and slab of the oors of the structure is usually cast with the
same quality concrete (same fc0 ) throughout. (In columns, the use of a different
quality concrete may be warranted; but even then all columns in a given oor level
would have the same concrete mix.) So preselecting the concrete quality for the
slabs and beams throughout a building is standard practice.
The same is true with the reinforcement. Labor is the dominant factor in the price
of the in-place reinforcing steel. And the basic cost per ton of reinforcing steel
with fy 40 ksi and fy 60 ksi is very near the same, so there is no economic
incentive to use the former. In fact, 60 ksi steel provides 50 % more strength than
40 ksi steel, thus making it cheaper to use.
Of the three remaining variables, b (the width of the section), d (the working
depth of the section), and As (the amount of reinforcement), two must still be
preselected in order to solve for the remaining unknown quantity. Generally
speaking, practitioners select a concrete section (b and h) and then solve for a
minimum required amount of reinforcement to satisfy the demanded factored
moment requirements. Often all beams have the same depth and width to enable
the contractor to reuse the forms. In other cases keeping the depth of all beams
uniform satises the minimum headroom requirement throughout the structure.
In general, two types of problems arise: (1) The beams sizes (b and h) are set
using the considerations stated above and the designer needs only to determine the
required area of steel (As); this is by far the most common problem. (2) The beams
sizes (b and h) and area of steel (As) are all unknown and determined by the designer
during the process; this problem is more academic than practical.
b, h known, As unknown
The owchart in Figure 2.46 shows the steps for the design process.
2.25 Design of Beams 99
2. d h y; y 2.5 in.
assumed y
3.
12,000Mu
R
bd 2
Stop!
Yes Need to
R Rmax? increase
beam size.
No
Yes
min? min
No
5.
As bd, then select bar size
and numbers from Table A2.9.
6.
Calculate y and find d h y.
No
d dassumed
Yes
End
Figure 2.46 Flowchart for the design of reinforced concrete rectangular beams (b, h known,
As unknown)
100 2 Rectangular Beams and One-Way Slabs
bd2 R
MR
12,000
Set Mu MR:
bd 2 R
Mu MR
12,000
12,000Mu
R
bd 2
Step 4. Use R, fy, and fc0 to determine from Tables A2.5 to A2.7. If R is greater
than the maximum R value (Rmax) to be found in the tables, it means that the
selected sizes are too small and must be increased.
If the value obtained is less than min, it means that the beam sizes b and
h are larger than needed to carry the loads with minimum reinforcement.
This may happen when other considerations dictate the beam sizes. In this
case use min from Table A2.4, because the beam must always have the
required minimum reinforcement.
Step 5. Determine how much steel is needed and select bars using Table A2.9. It is
also helpful to use Table A2.8 here, because it lists how many of a certain
size of bar may be tted into the selected b in a single layer.
Step 6. Once the bar sizes are known, the exact effective depth (d) can be
calculated. If this depth is greater than what was assumed at the beginning
of process, the design will be conservative as it will have more moment
capacity than what was demanded. If the effective depth is less than the
assumed value (e.g., the section needs multiple layers of reinforcements),
then the process needs to be repeated with a new value of d. Insignicant
differences in the assumed and recalculated values in d (less than 3/8 in. in
slabs and 1/2 in. in beams) may be neglected and the reinforcing need not
be redesigned.
Note that having multiple layers of reinforcing bars may inuence the
value of the strength reduction factor, .
2.25 Design of Beams 101
Example 2.12 Figure 2.47a shows the partial framing plan of a beam-girder
reinforced concrete oor system. The slab is 6 in. thick, and is subjected to a
superimposed dead load of 30 psf. The oor live load is 100 psf. Beam B-2 has
a width of 12 in. (b 12 in.), and a total depth of 30 in. (including the slab
thickness). Determine the steel required at Section 1.1. Use the ACI Code coef-
cients to calculate moments. Assume that the beam end is integral with the column.
Use fc0 4 ksi, fy 60 ksi, and assume that the unit weight of concrete is 150 pcf.
Stirrups are #3 bars.
A B C
30'-0" 30'-0"
15'-0"
15'-0" 1
15'-0"
15'-0"
6 in.
#3 stirrup 24 in.
12 in.
Section 11
Solution
Step 1. Before calculating the moments at the selected location, we must determine
the oor loads:
102 2 Rectangular Beams and One-Way Slabs
6
Weight of slab 150 75 psf
12
Superimposed dead load 30 psf
Total dead load 105 psf
Live load 100 psf
The tributary width for beam B-2 is 150 000 ; therefore, the uniform dead
and live loads are:
Beam weight
12 24
150
105 15 12 12 = 1.88 kip/ft
wD = +
1,000 1,000
100 15 Note : Reduction of live load
wL 1:5 kip=ft
1,000 is neglected here:
wu 1:2wD 1:6wL 1:2 1:88 1:6 1:5 4:65 kip=ft
wu 2n 4:65292
Mu 391 ft-kip
10 10
Section to be designed
wu 2n wu 2n wu 2n
16 14 10
Step 2. Assuming the distance (y) from the edge of the beam in tension to the center
of tensile steel is 2.5 in.:
d h y 30 in: 2:5 in: 27:5 in:
Step 3. The required resistance coefcient, R, is:
12,000Mu 12,000 391
R
bd 2 1227:52
R 517 psi
2.25 Design of Beams 103
Step 4.
R 517 psi
fc0 4 ksi ! Table A2:6b ! 0:0106
fy 60 ksi
Table A2:8 ! bmin 10:5 in: < 12 in: < bmax 24 in: ok
6 in.
3 #10
24 in.
12 in.
b, h, As unknown
There is still only one design equation, but the problem now is formulated
differently. It is somewhat more contorted than the previous one, for if the
designer does not like the results obtained with the assumed cross section and the
104 2 Rectangular Beams and One-Way Slabs
corresponding reinforcement, he or she can just change the width or the depth
(or both) and recalculate the reinforcement until satised with the design.
A rst assumption may be an arbitrary selection of the steel ratio . When ratios
close to the max value are chosen, the amount of steel required creates a rather
congested layout, especially in the positive moment regions (steel is placed in the
bottom of the beam). On the other hand, an unnecessarily large concrete section
may result if the sections moment requirement can be satised with min. Most
practical designs have steel ratios somewhere between max and min.
Generally speaking, if is assumed to be about 0.6max or less the beam
proportions will likely be such that excessive deection will not be a problem.
Therefore, Table 2.1 is provided as an aid for the designer. In this table, des was
calculated as 0.6max as a starting point.
Then the corresponding R value may be obtained from Tables A2.5 to A2.7.
The value bd2 can be determined using Mu:
Mu
Mu Rbd 2 ! bd2
R
Two unknowns remain, however: b and d. There are no ACI Code requirements on
the geometrical proportioning of beams. But it is more economical to design beams
as deep and narrow rather than wide and shallow sections. This means that the
effective depth, d, should be larger than the width, b. Generally speaking, the most
economical beam sections for spans up to 25 ft usually have a d/b ratio between 1.5
and 2.5. For longer spans, a d/b ratio of 34 may be more suitable. Economy for a
specic beam (or set of beams) is not the same as economy for the overall building.
In fact, sometimes it is more economical to design wide and shallow beam sections
due to the savings in the oor-to-oor height, even though this design will require
more reinforcing steel.
Figure 2.48 summarizes the steps of the design process:
Step 1. Find the factored loads and moments.
Step 2. Use fy and fc0 to select a des value from Table 2.1. Then nd the
corresponding R value from the appropriate design table (Tables A2.5
to A2.7).
Step 3. The formula for MR is:
bd2 R
MR
12,000
2.25 Design of Beams 105
1. Calculate maximum Mu
9. As bd 5. h d y
Select the size and
number of bars using
Tables A2.8 and A2.9. 6. No
h hmin
h hmin?
Yes
7.
No Beam weight
included ?
Yes
End
Figure 2.48 Flowchart for the design of reinforced concrete rectangular beams (b, h, As unknown)
and the design of the beam requires that MR Mu. For the most economical
case, Mu MR; therefore
bd2 R
Mu
12,000
12,000Mu
bd 2
R
certain proportion between d and b, for example, d/b 2; then the problem
again becomes straightforward.
Step 4. Use the values of b and d from above to nd the required area of reinforce-
ment (As):
As bd
and select the size and number of bars using Tables A2.8 and A2.9.
Step 5. Now nd the beams total depth (h) using the effective depth (d) from Step
3 and size of bars:
hdy
12,000Mu
R
bd2
As bd
and select the numbers and sizes of bars from Tables A2.8 and A2.9.
Example 2.13 Determine the required area of steel for a reinforced concrete
rectangular beam subject to a total factored moment, Mu 400 ft-kip, that already
includes the estimated weight of the beam. fc0 4,000 psi and fy 60,000 psi and
use des 0.0124 from Table 2.1.
Solution
Step 1
Mu 400 ft-kip
2.25 Design of Beams 107
There are an innite number of solutions, that is, an innite number of concrete
cross sections that will satisfy the design problem, even with the provision that
0.0124 (1.24 %). The table below lists a few solutions. Take your pick!
bd 2 8,054 in:3
d
b
2
d 2 d3
d 8,054
2 2
p
d 3 2 8,054 25:3 in:
d 25:3
b 12:65 in: ! Select b 13 in:
2 2
h 25:3 2:5 27:8 in: ! Select h 28 in:
As bd 0:0124 12:65 25:3 3:97 in:2
108 2 Rectangular Beams and One-Way Slabs
Example 2.14 Use the oor framing plan and loadings of Example 2.12
(Figure 2.47a) to design the reinforced concrete rectangular beam along grid line
2. Assuming that the beam width b 12 in., determine the beam depth, h, and
required steel for the location of the maximum bending moment. Use ACI Code
coefcients for calculation of moments. Assume that the beam end is integral with
the column, fy 60 ksi, fc0 4 ksi, and the unit weight of the concrete is 150 pcf.
The stirrups are #3 bars.
Solution
Step 1. Find the maximum ultimate moment, Mu.
From Example 2.12:
105 15
wD 1:58 kip=ft without the weight of the beams stem
1,000
100 15
wL 1:5 kip=ft without the use of live load reduction
1,000
wu 1:2wD 1:6 wL 1:2 1:58 1:6 1:5 4:3 kip =ft
1 1
n 30 ft 29 ft
2 2
wu 2n 4:3292
Mu 362 ft-kip
10 10
Step 2. From Table 2:1 ! fc0 4 ksi, fy 60 ksi ! des 0.0124 From Table A2.6b
! R 596 psi
wu 2 wu 2 wu 2 wu 2 wu 2 wu 2
n n n n n n
16 14 10 11 16 11
n n
Figure 2.49 (a) Moments using the ACI coefcients (Example 2.14)
2.25 Design of Beams 109
Step 4. Calculate the required area of steel, and select the number and size of the
reinforcing bars:
Step 5. Use the selected bar sizes and the effective depth (d ) to calculate the total
beam depth (h):
1 3 1:128
y1 2:44 in:
2 8 2
h d y 24:7 2:44 27:14 in:
h 28 in:
Step 6. Check to see if the beam depth is more than the recommended minimum for
deection control. The case for the beam with one end continuous results in
the largest required depth (see Figure 2.41):
30 12
hmin 19:5 in: < 28 in: ok
18:5 18:5
Step 7. Calculate the correct beam weight. The total beam depth is 28 in. The
concrete slab, however, is 6 in. thick; therefore, the beam depth (the stem)
below the slab is 28 in. 6 in. 22 in.
12 22
150
12 12
Stem weight 0:28 kip=ft
1,000
110 2 Rectangular Beams and One-Way Slabs
Step 8.
12,000Mu
R
bd2
12,000 390
R
1224:72
R 639 psi
The selected reinforcement is the same as it was for the previous design
cycle. Figure 2.49b shows the sketch of the beam.
6 in.
4 #9
28 in.
12 in.
2.26 Slabs
1'0"
other strips behave the same way, that is, they need the same amount of reinforcing.
Figure 2.50 also illustrates that if only one imaginary strip is loaded, the adjacent
slab strips will have to help. This is because it is impossible for a monolithic structure
to get the deformation diagram shown on the right of the gure.
Figures 2.51 and 2.52 show the framing plan of different reinforced concrete
oor/roof systems. In Figure 2.51a, slab S-1 is supported by the surrounding beams
Columns
a
(S-1) Slabs
(S-1) (S-1) (S-1)
(typical)
Girders
(typical)
Figure 2.51 (a) Slabs in beam girder oor system; (b) at plate slab
2.27 Behavior of Reinforced Concrete Slabs Under Loads 113
A
1'-0" wide strip
Wall (typical)
(S-3)
A
Plan Section A-A
and girders. In Figure 2.51b slab S-2 is part of a at plate oor system, in which
slabs are directly supported by columns. In Figure 2.52 slab S-3 is supported by two
parallel walls, which can be made of concrete or masonry.
Depending on the geometry and location of the supports, most slabs are divided into
two groups: one-way slabs, and two-way slabs.
One-way slabs bend mainly in one direction. If the supporting elements of the
slab are only two parallel members such as beams or walls, the slab is forced to
bend in a perpendicular direction. Figure 2.52 shows the plan view of a slab
supported by two parallel walls. Because every 1 ft wide strip can be considered
to be the same as all the others, only a single 1 ft wide strip of slab needs to be
considered in analysis and design.
The slabs geometry is an important factor that affects its behavior under loads.
Figure 2.53a shows a slab supported by edge beams B-1 and B-2. Determining the
distribution of loads from the slab to the supporting beams can be simplied by
assuming that the load is transferred to the nearest beam. Such an assumption is
represented by drawing 45-degree lines from each slab corner. The enclosed areas
show the tributary loads to be carried by each beam. Beam B-1 will carry large
trapezoidal loads compared to the triangular loads that will be carried by beam B-2.
As the ratio of longer span () to shorter span (s) increases, B-1 carries more loads
than does B-2, that is, more loads are transferred in the shorter span
of the slab.
In fact, if the ratio is greater than or equal to 2.0 2:0 , the load carried
s s
by B-2 is quite small, and it can be neglected altogether. Therefore, if 2:0,
s
the slab behaves as a one-way slab for all practical purposes, even though the slab is
supported on all four edges.
114 2 Rectangular Beams and One-Way Slabs
45 B-1
B-2 B-2 s
45
B-1
1'-0"
B-1
B-2 ws B-2
1'-0" w s
B-1
b ws w
Figure 2.53 (a) Slab (edge supported); (b) slab load distribution
5ws 4s
s 2:54
384EI
2.28 Reinforcement in One-Way Slabs 115
5w 4
2:55
384EI
The two deections must be equal. Thus, an expression may be developed that
relates the loads and spans, as shown in Equation (2.56).
s
5ws 4s 5w 4
384EI 384EI
2:56
ws 4s w 4
4
ws 4
4
w s s
In general, two types of reinforcement are used in one-way slabs: main reinforce-
ment, and shrinkage and temperature reinforcement.
The main reinforcement resists the bending moments. It is designed to act in the
direction of the one-way slabs bending, which is along the shorter span length.
Figure 2.54 shows the main reinforcement in a one-way slab supported by two
parallel walls. The slab is assumed to be simply supported by the walls. In other
words, no moment is transferred from the slab to the walls. Because the bottom
portion of slab is in tension, the main reinforcement is placed in the bottom.
116 2 Rectangular Beams and One-Way Slabs
B
A
1'-0"
A
B
Wall
S&T reinforcement
S&T reinforcement
Main reinforcement
Main reinforcement
1'-0"
Section B-B Section A-A
As discussed in Chapter 1, fresh concrete loses water and shrinks soon after
placement. In addition, variations in temperature cause the concrete to expand
and contract. These volume changes, when restrained, may result in cracking of
concrete, especially in the early stages of strength development. Reinforcing bars
are used to resist developing tensions in order to minimize cracks in concrete
caused by shrinkage and temperature changes. The main longitudinal reinforcement
in beams plays that role as well. Because the cross-sectional dimensions of beams
are relatively small and beams may freely change their cross-sectional dimensions
2.28 Reinforcement in One-Way Slabs 117
a
S&T reinforcements Main reinforcements
Reinforcement distribution
b
(M)
Figure 2.55 One-way slab reinforcement (continuous construction). (a) Reinforcement distribu-
tion. (b) Moment diagram for continuous slab (refer also to Figure 2.14)
without restraint, shrinkage and temperature reinforcement are not needed perpen-
dicular to the main bars.
This is not the case in reinforced concrete slabs. Slabs typically have large
dimensions in two directions, thus they need shrinkage and temperature reinforce-
ment, which is placed in the direction perpendicular to the main reinforcement.
Figures 2.54 and 2.55 show such reinforcement for simple-span one-way slabs and
continuous one-way slabs, respectively. In addition, temperature and shrinkage
reinforcement helps distribute concentrated loads to a wide zone transversely to
the one-way direction. (This is necessary in bridges, for example, to distribute large
wheel loads onto a much wider strip than the one directly affected by the concen-
trated load.)
As discussed above, two types of reinforcement are used in one-way slabs. The ACI
Code sets the following minimum reinforcement criteria for both the main and the
shrinkage and temperature reinforcements.
118 2 Rectangular Beams and One-Way Slabs
In other words, if the calculated main reinforcement is less than that required for
shrinkage and temperature reinforcement, the designer must use at least the latter
amount.
Minimum Shrinkage and Temperature Reinforcement The ACI Code
(Section 7.6.1.1) requires shrinkage and temperature reinforcement based on the
grade of steel, as given in Equations (2.58)(2.60).
0:0018 60
For fy > 60 ksi ! AsS&T bh 0:0014bh 2:60
fy
Therefore, the maximum main reinforcement spacing with fy 60 ksi steel for
one-way slabs is given by Equation (2.61a).
Similarly, when using fy 40 ksi steel as main reinforcement, Equation (2.53) will
simplify to Equation (2.61b).
A 1 ft (12 in.) wide strip of slab is typically used for the analysis and design of
one-way slabs. Thus, it is advantageous to dene the amount of steel in a 1 ft wide
strip as a function of the bar size and the spacing.
Table A2.10 lists spacing and bar sizes for slabs. The table provides the areas of
reinforcement averaged out to 1 ft width for different sizes and spacing of bars.
(One can interpolate for in. spacing increments, if so desired.)
For example, with #5@8 in. o.c. (#5 bar at 8 in. on-center spacing), the table,
under #5 bars spaced at 8 in., provides the area of steel per foot of section 0.47 in.2.
In other words, 0.47 in.2/ft is equivalent to one #5 bar every 8 in.
Another example: If 0.50 in.2 of reinforcement is required for a 1 ft wide strip of
a slab, the table offers several options, including #4@4 in. (As 0.60 in.2),
#5@7 in. (As 0.53 in.2), #6@ 10 in. (As 0.53 in.2), and so on.
In general, one-way slabs and reinforced concrete beams are analyzed very simi-
larly. There are a few differences, however. These are listed below:
1. For the analysis of one-way slabs, b is always 12 in.
2. Slabs require a different amount of concrete cover over the reinforcement.
3. Slabs require shrinkage and temperature reinforcement.
4. The Code-specied minimum amounts of reinforcing steel for slabs and beams
are different.
5. Minimum required depth/span ratios for adequate control of deection are
different.
6. Bar spacing requirements are different.
Figure 2.57 summarizes the steps for the analysis of reinforced concrete one-way
slabs. They are as follows:
As
Step 1. Calculate the steel ratio, . As is the area of steel in a 1 ft wide strip
bd
of slab from Table A2.10. Compare with max from Table A2.3. The
maximum permitted steel ratio is the same for beams and slabs.
Step 2. Compare As with As,min, which is the minimum required area of steel for
the control of shrinkage and temperature-induced volumetric changes.
If As As,min, the proportioning of steel and concrete is not acceptable
according to the current ACI Code and the slabs use is illegal. If As As,min,
however, then one of the following methods can be used to check the
adequacy of the slab:
2.30 Analysis of Reinforced Concrete One-Way Slabs 121
Analysis of Reinforced
Concrete One-Way Slabs
1.
Slab is illegal As You may not legally use all
No
per current max
the reinforcing in finding MR.
ACI Code. bd
Yes
2.
No
3. Asfy As As,min ? 3.
a Use , fy and fc to find R
0.85 fc b from Tables A2.5 through A2.7.
Yes
(1) EITHER (2) OR
c 3
?
dt 8
Yes 5.
No
4.
bd 2 R
MR Mu ? MR
0.90 12,000
B2
A2 c
dt
4.
MR Asfy (d a (
2
No Yes
8.
Check slab thickness for deflection control.
END
Figure 2.57 Flowchart for the analysis of reinforced concrete one-way slabs
Method I
Step 3. Calculate the depth of the compression zone:
As fy
a
0:85fc0 b
c 3
If , the section is tension-controlled and 0.90. Otherwise, the
dt 8
section will be in the transition zone. Calculate the strength reduction
factor, :
B2
A2 c
dt
(If As is in in.2, fy ksi, d and a in., then MR will have kip-in. unit.
Divide the result by 12 to obtain MR in the customary units of kip-ft).
Step 5. Compare MR with the maximum factored moment from the applied loads. If
MR < Mu the slab is not adequate to carry the assumed loads. Proceed to
calculate a new permissible live load that the slab may legally support. If
MR Mu, the section can take the assumed loads, but the reinforcing still
needs to be checked for conformance with other Code requirements.
Step 6. Check spacing requirements. The maximum allowable spacing of main
reinforcement is min{3h, 12 in.}, or min{3h, 18 in.} for fy 60 ksi and
fy 40 ksi steel, respectively.
Step 7. Check the amount and spacing of shrinkage and temperature reinforcement,
(As)S &T (Refer to Equations (2.58)(2.60).)
Step 8. Check the thickness of the slab against the minimum thickness of one-way
slabs for desirable deformation control (see Figure 2.42).
If the slab thickness is less than the above limits, calculate the deection
and check it against the Codes serviceability requirements.
2.30 Analysis of Reinforced Concrete One-Way Slabs 123
Method II
Steps 1 and 2 are the same as in Method I.
Step 3 Use fy, fc0 , and the calculated steel ratio () to obtain the resistance coefcient,
R, from Tables A2.5 to A2.7.
Step 4 Use the R value to calculate the sections resisting moment.
bd2 R
MR
12,000
6 in
Use a concrete cover of in., fc0 3:0 ksi, and fy 40.0 ksi.
Solution
Step 1. Check the reinforcement ratio in the slab:
Diameter of #5 bars
5
3 8
y = + = 1.06 in.
4 2
Cover
d = h - y = 6 in. - 1.06 in.=4.94 in.
As 0:53
0:00894
bd 12 4:94
(a) Method I
Step 3. Calculate the depth of the compression zone:
As fy 0:53 40
a
0:85 fc0 b 0:85 3 12
a 0:69 in:
a 0:69
c 0:81 in:
1 0:85
dt d 4:94 in:
c 0:81
0:164 < 0:375 0:90
dt 4:94
Step 4.
a
MR Mn As fy d
2
0:69
MR 0:90:5340 4:94
2
87:7 in:-kip
MR 7:3 ft-kip
12 in:=ft
Mu 1:2MD 1:6ML
Mu 1:2 3:0 1:6 2:0 6:8 ft-kip < 7:3 ft-kip ok
2.30 Analysis of Reinforced Concrete One-Way Slabs 125
From Table A2:10 ! #3@12 in: ! As 0:11 in:2 =ft < 0:14 in:2 =ft N:G:
Therefore, the shrinkage and temperature reinforcement in the slab
does not satisfy the current ACI Codes minimum requirement.
(b) Method II
0.00894
Step 3. fc0 3 ksi ! Table A2:5a ! R 299 psi (by interpolation)
fy 40 ksi
Step 4.
bd 2 R
MR
12,000
124:942 299
MR
12,000
MR 7:3 ft-kip
Example 2.16 Figure 2.59 shows the partial oor framing plan and section of a
reinforced concrete oor system. The weight of the ceiling and oor nishing is
5 psf, the mechanical and electrical systems are 5 psf, and the partitions are 15 psf.
The oor live load is 150 psf. The concrete is normal weight, fc0 4 ksi, and
fy 60 ksi. Check the adequacy of slab S-1 in the exterior bay at (a) midspan,
and (b) over the interior supporting beam. Assume the slab is cast integrally with
the supporting beams and use ACI code coefcients to calculate moments. Use
in. cover for the slab.
126 2 Rectangular Beams and One-Way Slabs
A B C
25'-0" 25'-0"
A
1
12'-0" (S-1)
12'-0" A
12'-0"
12'-0"
6 in.
#4 @ 10 in.
#3 @ 10 in. 24 in.
14 in. 14 in.
12'-0 in.
Section A-A
Solution
(a) Check the Slab at the Midspan
Step 1. The main reinforcement at the midspan (positive moment) is #4@
10 in.
4
3 8
y 1:0 in:
4 2
d h y 6 in: 1 in: 5 in:
As 0:24
0:0040
bd 125
fc0 4 ksi ! Table A2:3 ! max 0:0207 > 0:004 ok
fy 60 ksi
Step 2.
As, min AsS&T 0:0018bh fy 60 ksi
As, min 0:0018126 0:13 in:2 =ft
As 0:24 in:2 =ft > 0:13 in:2 =ft ok
Step 4.
bd2 R
MR
12,000
1252 208
MR
12,000
MR 5:2 ft-kip
wu 2n
Mu
14
0:3610:832
Mu
14
Mu 3:0 ft-kip < MR 5:2 ft-kip ok
12 12
hmin 6 in: 6 in: ok
24 24
Slab is ok at mid-span:
Step 2.
As, min AsS &T 0:0018bh fy 60 ksi
Step 3.
0:005
fc0 4 ksi ! Table A2:6b ! R 258 psi
fy 60 ksi
Step 4.
bd2 R
MR
12,000
1252 258
MR
12,000
MR 6:5 ft-kip
130 2 Rectangular Beams and One-Way Slabs
wu 2n
Mu
10
0:3610:832
Mu
10
Mu 4:2 ft-kip < MR 6:5 ft-kip ok
y 1:12 in:
2.31 Design of Reinforced Concrete One-Way Slabs 131
4. 12,000Mu
(Mu ft-kip) R
bd 2
b 12 in.
No
Increase slab thickness. R Rmax?
Yes No
As bd 7.
dactual dassumed?
No 6. A As,min ?
As As,min s
As,min As(S&T)
Yes
Yes
Figure 2.60 Flowchart for the design of reinforced concrete one-way slabs
132 2 Rectangular Beams and One-Way Slabs
d h 1:12 in:
12,000Mu
Rpsi
bd 2
As bd
If As < As,min, the slab requires only a small amount of reinforcing steel,
As. Use at least As,min, however. Select the bar size and spacing from
Table A2.10.
Step 7. Check for actual depth (dactual) based on the bar selected. If
dactual < dassumed, go back to Step 4 and revise. Repeat if the difference is
too large (larger than 1/8 in. for slabs h 6 in. and 1/4 in. for h > 6 in.).
Step 8. Check bar spacing. The spacing of bars selected in Step 6 has to be checked
against the ACI Code requirements for maximum allowable spacing.
Step 9. Design the shrinkage and temperature reinforcements according to the ACI
Code requirements.
Example 2.17 Design the one-way slab (S-1) of Example 2.16. Determine the
reinforcement at (a) the midspan and (b) the supports.
Solution
(a) Slab Design at the Midspan
Step 1. Because S-1 is one end continuous, the minimum slab thickness (hmin) is:
12 12
hmin 6 in:
24 24
2.31 Design of Reinforced Concrete One-Way Slabs 133
On a 1 ft wide strip
100 1
wD 0:10 kip=ft
1,000
150 1
wL 0:15 kip=ft
1,000
wu 1:2wD 1:6wL 1:2 0:10 1:6 0:15
wu 0:36 kip=ft
14 in:
n 12 ft 10:83 ft
12
The maximum factored moment at the midspan of S-1 (see Figure 2.61) is:
wu 2n
Mu
14
0:3610:832
Mu
14
Mu 3:0 ft-kip
wu n2 wu n2 wu n2 wu n2
24 14 10 11
Figure 2.61 Design factored moments for slab S-1 of Example 2.17 using ACI Code coefcients
from Table A2.1
134 2 Rectangular Beams and One-Way Slabs
12,000Mu
R
bd2
12,000 3:0
R 126 psi
124:882
R 126 psi
fc0 4 ksi ! Table A2:6b ! 0:0024
fy 60 ksi
As bd 0:0024124:88
As 0:14 in:2 =ft
Step 6. The minimum amount of reinforcement for slabs cannot be less than the
required shrinkage and temperature reinforcement steel:
Note that according to Section 2.28, the smallest size bar for main
reinforcement is #4.
Step 7. Check for the actual effective depth.
4
3 8
dactual 6 5:0 in: > d assumed 4:88 in: ok
4 2
Therefore:
Use #4@ 12 in. for the main reinforcement at the midspan.
Step 9. Calculate the required shrinkage and temperature reinforcement.
Therefore,
Use #3@10 in. for the shrinkage and temperature reinforcement.
(b) Slab Design at the Supports
Step 1. From Step 1 of part a:
hmin 6 in:
Step 2. The factored uniformly distributed load on the slab (wu) from Step 2 of part
a is:
wu 0:36 kip=ft
n 10:83 ft
From Figure 2.61, the moments at the exterior and interior supports are:
wu 2n 0:3610:832
M
u 1:76 ft-kip exterior support
24 24
wu 2n 0:3610:832
M
u 4:22 ft-kip interior support
10 10
Step 3.
Assume d h 1:12 in: 6 1:12 4:88 in:
136 2 Rectangular Beams and One-Way Slabs
Step 4.
12,000Mu 12,000 1:76
R 74 psi exterior support
bd 2 124:882
12,000 Mu 12,000 4:22
R 177 psi interior support
bd 2 124:882
Step 5. 8
> R 74 psi
<
For exterior support fc0 4 ksi ! Table A2:6b ! ext: 0:0014
>
:
fy 60 ksi
8
> R 177 psi
<
For interior support fc0 4 ksi ! Table A2:6b ! int: 0:0034
>
:
fy 60 ksi
Therefore:
Therefore, use
#4 @ 12 in. #4 @ 12 in.
#3 @ 10 in. #4 @ 12 in.
12'0"
Problems
In the following problems, unless noted otherwise, use normal weight concrete with
a unit weight of 150 pcf, 1.5 in. for beam clear concrete cover, and 0.75 in. for slab
clear concrete cover.
2.1 Consider a section with a width (b) of 14 in. and reinforced with 4 #9 bars in
a single layer. fc0 4,000 psi, and fy 60,000 psi. Determine the moment
capacity of the section, MR, using Method I or II, for the following cases:
(a) d 28 in.
(b) d 32 in.
(c) d 36 in.
(d) d 40 in.
Show the changes in MR with respect to the sections effective depth.
Calculate the percentages of increase in MR versus d.
138 2 Rectangular Beams and One-Way Slabs
12 in. 8#9
32 in.
Problems 139
2.7 The rectangular reinforced concrete beam shown below is subjected to a dead
load moment of 180 ft-kip and live load moment of 90 ft-kip. Determine
whether the beam is adequate for moment capacity. fc0 4,000 psi, and
fy 60,000 psi. The stirrups are #3 bars.
12 in.
30 in.
4 #9
2.8 The beam below supports 500 lb/ft service dead loads and 600 lb/ft service live
loads in addition to its self-weight. Calculate the maximum simply-supported
span ( ?) for the beam. Use Method II in the calculations. Use fc0 5,000 psi
and fy 60,000 psi.
15.5 in.
3 #8
2.5 in.
12 in.
Pu Pu
2.10 The beam shown below is part of a beam-girder oor system. It is subjected to
a superimposed dead load of 4.0 kip/ft (excluding the beam weight) and a
live load of 2.0 kip/ft. Check the adequacy of this beam. Use fc0 4,000 psi,
fy 60,000 psi, and #3 stirrups. Assume knife edge type supports at the
centers of the walls.
wD 4.0kip/ft
wL 2.0kip/ft A B
A B
5'-3" 18 in. 18 in. 5'-3"
21'-0"
12 in. 12 in.
4 #9
24 in. 24 in.
3 #10
A-A B-B
Note: Check both sections A-A and B-B. Neglect the reinforcement in the
bottom of the beam at section A-A.
2.11 Determine the moment capacity, MR, of the reinforced concrete section
shown below if subjected to a negative moment. The stirrups are #3 bars.
Use fc0 4,000 psi and fy 60,000 psi.
Problems 141
4 #10
28 in.
13 in.
2.12 The gure below shows the cross section of a oor system consisting of a
reinforced concrete beam supporting precast concrete planks. The beam span
is 200 000 with 160 000 spacing. Calculate the maximum service live load per
square foot of oor area. Use fc0 4,000 psi and fy 60,000 psi. The unit
weight of lightweight (LW) concrete used is 108 pcf. Assume the beam is
simply-supported.
24 in.
3#9
12 in.
2.13 The 16 in. 27 in. rectangular reinforced concrete beam shown below is
reinforced with 4 #10 bars in the positive moment region and 3 #11 bars in
the negative moment region. Determine the maximum factored uniformly
distributed load, wu, for this beam. Stirrups are #4, fc0 5,000 psi, and
fy 60,000 psi.
142 2 Rectangular Beams and One-Way Slabs
wu
28'-0" 12'-0"
2.14 The beam of Problem 2.11 is part of a beam-girder oor system shown below
(beam B-1). The oor slab is 6 in. thick concrete, and the weight of the
mechanical/electrical systems is 5 psf. Assume 15 psf for partition loads,
and miscellaneous dead loads of 5 psf. What is the maximum allowable live
load for this oor? Consider only the negative moment capacity of the section.
(Note: Use the ACI moment coefcients. Live load is not to be reduced.)
(B-1)
40'-0"
40'-0"
2.15 Calculate the required areas of reinforcement for the following beams. Use
fc0 4,000 psi and fy 60,000 psi.
(a) b 10 in., d 20 in., Mu 200 ft-kip
(b) b 12 in., d 24 in., Mu 300 ft-kip
(c) b 18 in., d 36 in., Mu 500 ft-kip
Problems 143
P P
w
wu 4.0 kip/ft
#4 @ 9 in. #4 @ 12 in.
6 in. 6 in.
10'-0"
12 in. 12 in.
2.24 The gures below show the framing plan and section of a reinforced concrete
oor system. The weight of the ceiling and oor nishing is 5 psf, that of the
mechanical and electrical systems is 5 psf, and the weight of the partitions is
20 psf. The oor live load is 80 psf. The 6 in.-thick slab exterior bay (S-1) is
reinforced with #6@9 in. as the main reinforcement at the midspan and #4@
12 in. for the shrinkage and temperature reinforcement. Check the adequacy
of the slab. Use the ACI moment coefcients. Use fc0 4,000 psi and
fy 60,000 psi.
Problems 145
25'-0" 25'-0"
10'-0" (S-1)
A
10'-0" (S-2)
10'-0"
Framing Plan
6 in.
30 in.
12 in. 12 in.
10'-0"
Section A-A
146 2 Rectangular Beams and One-Way Slabs
2.25 Design a 6 in.-thick one-way slab for a factored moment, Mu 10 ft-kip. Use
fc0 4,000 psi and fy 60,000 psi.
2.26 Find the reinforcements for the midspan and supports for an interior 6 in.-thick
slab (S-2) of the oor of Problem 2.24. Sketch the slab and show the
reinforcements including the shrinkage and temperature reinforcement steel.
Self-Experiments
P
Slot
Place the beam on two supports and add a load at the center as shown in Figure SE
2.1. Answer the following questions:
1. What happened to the slots at the top and bottom of the beam?
2. Did the slots stay straight after adding the load?
3. Any other observations?
Experiment 2
You must start and perform Experiments 2 and 3 at the same time. In this
experiment, you nd the modulus of rupture for a plain concrete beam and learn
about concrete curing and gaining strength with time.
For this experiment you will build four beams using concrete with w/cm
ratio 0.5. Size the beams as you wish, but do not make them excessively small
or large (for practical reasons). After forming the beams (you can use cardboard or
wood for your forms, depending on the beam size), spray water on two of the beams
Self-Experiments 147
while keeping the other two dry. Keep your concrete beams indoors, as the concrete
may freeze and stop the hydration process. After 2 days, test two of your test
beams (one kept dry and one kept wet) by placing loads on them, as shown in
Figure SE 2.2.
P P
3 3 3
Increase the loads until the beams fail. Record the loads at which the two
specimens fail.
After seven days, repeat the tests with the remaining two beams and record the
loads at which they fail.
Experiment 3
In this experiment, you will learn about the importance of reinforcing steel in
concrete beams and compare the results with those of Experiment 2.
When you pour the four plain concrete beams for Experiment 2, build two
reinforced concrete beams with the same dimensions as those of the plain concrete
beams. You can use steel wires for the reinforcement (depending on your beam
size). Place these wires on only one side of the beam (singly-reinforced beam).
After 2 days, place one of the beams on two supports and apply loads as shown in
Figure SE 2.3a. Increase the load, and record your observations.
P P
3 3 3
Repeat this test for the remaining reinforced concrete beam after seven days.
(Perform these tests at the same time as Experiment 2.) DO NOT TRY TO FAIL
THE REINFORCED CONCRETE BEAMS! Turn the beams upside down
(Figure SE 2.3b) and repeat the tests. Add loads until the beams fail. Record
your observations.
148 2 Rectangular Beams and One-Way Slabs
P P
3 3 3
a b
Figure SE 2.4 Slabs under loads: (a) two parallel supports; (b) supports along all edges
Self-Experiments 149
Test 2
Repeat Test 1 using the one-way slab model. Record your observations.
Experiment 5
This experiment deals with the reinforcement in slabs.
Cast two slab models with a thickness of approximately 1 in. and a width of at
least 12 in. Make one from plain concrete and the other from concrete reinforced
with a grid of thin wires (provide about in. cover).
One week after making the samples, compare the two slabs in terms of crack
formation. Which one has more surface cracks?
Chapter 3
Special Topics in Flexure
3.1 T-beams
3.1.1 Introduction
In cast-in-place reinforced concrete systems, the concrete for beams and slabs is
poured at the same time. As a result, a monolithic system is obtained, that is, beams
and slabs working together to carry the loads.
There are several different types of reinforced concrete oor systems, as we will
discuss in detail later in Chapter 6. Here we will use a beam-girder oor system to
study T-beams. Figure 3.1 shows the oor framing plan and the section of a typical
beam-girder oor system. The oor beams (B-1) support the one-way slab (S-1).
The slab transfers the load to the beams (B-1); then the girders (G-1) carry the
loads from the beams. The girders are supported by columns (C-1). Because the
one-way slab is continuously supported by the beams, the load on the beams is a
uniformly distributed load. The girders, however, support the beams at their ends,
so the loads on the girders are concentrated. Thus, the ow of the gravity loads is
from the slab to the beams, from the beams to the girders, from the girders to the
columns, from the columns to the footings, and from the footings to the ground.
In cast-in-place concrete construction, concrete is poured in the forms after the
form-work is built and the rebars are placed, creating a monolithic system of slabs,
beams, and girders. There is no physical separation between beams and slabs as in
steel construction. So when a beam bends, part of the slab attached to the beam
works with the beam and helps the beam carry the load. At the midspan the top part
of the beam is in compression. As a result the slab, which is attached to the top of
the beam, is subjected to compression stress. But at the support, the top portion of
beam, including the neighboring slab, is in tension. Therefore, the slab does not
help carry the beam load because the concrete does not take any tensile stresses.
Figure 3.2 shows cross sections and moments for a typical beam (B-1). At the
midspan the moment is positive, so steel reinforcement is needed at the bottom of
(G-1)
(C-1) (C-1)
A A
(B-1)
(B-1)
(B-1)
(B-1)
(S-1) (S-1) (S-1)
(G-1)
(C-1) (C-1)
(B-1) (G-1)
Figure 3.1 Beam-girder oor system. (a) Typical oor framing plan. (b) Section A-A
the beam A s . In this case the concrete slab and part of the beam web are in
compression. The shape of the compression zone looks like a T-shape, so it is called
a T-beam. Over the supports, however, there are negative moments. This requires
steel reinforcement at the top of the beam A s . In certain special cases, the ACI
3.1 T-beams 153
B A-
s A
0" A+s
B A
n
+
+
Moment diagram
Flange
A-
s
N.A. N.A.
A-A B-B
Figure 3.2 Beam behavior at midspan (T-beam) and over the support (rectangular beam)
Code requires part of the positive reinforcements A s to be extended over the
supports. In these cases reinforcing is used in the compression zone, resulting in a
doubly-reinforced beam (see Section 3.2).
The attached slab zone of a T-beam is referred to as the ange of the beam. The
portion below the ange is called the web. How much of the slab width acts as part
of the beam is a rather complex matter. It depends on many parameters that dene
how much of a slabs width is dragged into compression by the beam. The
phenomenon that dissipates the compression in the slab that lies farther away
from the beams web is known as shear lag.
The ACI Code simplies the matter by dening an effective ange width (beff),
in which the stresses due to bending are assumed to be uniform. Figure 3.3, which
shows the oor framing plan and a section through the midspan of a reinforced
concrete oor system, also shows the effective width for an edge beam and for an
154 3 Special Topics in Flexure
A A
n
b
beff beff
hf
d d
As As
bw sw bw
Figure 3.3 Effective ange widths for T- and L-beams. (a) Plan. (b) Section A-A
interior beam. The edge beam is called an L-beam because the compression zone
has an L shape. The interior beam is a T-beam. The beams clear span is n, and their
clear distance between adjacent webs is designated by sw. The slab or ange
thickness is designated by hf.
The effective ange widths of T- and L-beams are based on Sections 6.3.2.1 of
the ACI Code and are given in Equations (3.1) and (3.2).
a. beff for T-beams:
n
beff min bw , bw 16hf , bw sw 3:1
4
3.1 T-beams 155
The minimum amount of steel for a T-beam is the same as that for a rectangular
beam having working dimensions of bW (width of web) and d. Equation (3.3) gives
the minimum amount of steel required.
( p )
3 fc0 200
As, min max bw d, bw d 3:3
fy fy
The behavior of the T-beam (or L-beam) depends on the shape of the compression
zone. The depth of the equivalent stress block (a) may be above or below the bottom
of the ange, depending on the proportioning of the beam and the slab and the
amount of reinforcement used. Figure 3.4a, b show these two cases, respectively.
When the neutral axis is within the anges depth, the T-beam (or L-beam) acts
like a wide rectangular beam with a rectangular compression zone of size beff a.
In the rare cases when a small beff is coupled with relatively large positive moments,
the beff hf zone is not adequate to develop the compression part of the internal
couple. Then a part of the web becomes in compression to aid the compression
zone. The analysis and design of such beams are somewhat different from those of
rectangular beams (Figure 3.5).
Thus, two slightly different sets of procedures are used for the analysis of
T- (or L-) beams based on the shape of the compression zone. The owchart in
156 3 Special Topics in Flexure
a b
beff beff
hf a
d
d
h
bw bw
Figure 3.4 Different types of T-beams. (a) T-beam with compression zone in the ange. (b)
T-beam with compression zone in the web
a b
beff
hf y
c a
d z
dt
bw
Figure 3.5 T-beam with neutral axis below the ange. (a) Assumed area of compression zone at
ultimate moment. (b) y is the location of the centroid of the compression zone at ultimate moment
Figure 3.6 summarizes the different steps of analysis of T- (or L-) beams. They are
as follows:
Step 1. Calculate the effective ange width (beff).
Step 2. Check the minimum area of steel As,min or the minimum steel ratio from
Table A2.4. Note if the areas of reinforcing satisfy the current ACI Codes
requirements.
Step 3. Assume that the steel yields in tension before the concrete crushes in
compression (i.e., fs fy). Then calculate the total tensile force, T:
T As fy
1. Find beff :
beff = min bw + n , bw + 16hf , bw + s w
{ { (T-beam)
4
s
beff = min bw + n , bw + 6hf , bw + w
{ { (L-beam)
12 2
5. As 5. Calculate a:
= TCf
beffd a= + hf
Beam has more No 0.85fc bw
c 3
6. Use f , f and than maximum ?
y c dt 7
reinforcement. 6.
to find R from Tables a
Yes c=
A2.5 through A2.7
B2 1
No c 3
= A2 + ?
c /dt dt 8
Yes
6a.
= 0.9 Locate y:
Beam has
No Ay
single layer y = and
of bars? A
7.M = Tz
R z= dy
= 0.9? Yes
No Yes 7. M = b d 2R /12,000
R eff
8. No Beam is not
MR Mu?
Adequate!
Calculate t , and Yes
find ' from Tables
A2.5 through A2.7. Beam is ok.
7. R
MR = beffd 2
12,000
Cf 0:85fc0 beff hf
The internal couple requires that T C, that is, the compression force
and the tensile force must be equal.
If T < Cf (or a < hf), the full depth of the ange thickness is not needed to
develop the compression part of the internal couple. In that case the depth
of the equivalent stress block is less than the thickness of the ange, and
case a below is applicable; otherwise, use case b.
Case a: The compression zone is within the ange (a hf); the beam behaves like a
rectangular beam.
Step 5. Determine the steel ratio, :
As
beff d
Step 6. Use fy, fc0 , and to obtain the resistance coefcient, R, from Tables A2.5
through A2.7.
The resistance coefcient obtained, R, is only applicable for beams
with a single layer of reinforcement (dt d). If the beam has multiple
layers of reinforcement, R may need to be revised. If the value of the
strength reduction factor, , in the last step is 0.90, no change in the
As
value of R is necessary. If < 0.90, however, then compute t
beff d t
and obtain the corresponding value of from Tables A2.5 through A2.7.
Then calculate R0 R0 R0 =.
Step 7. Calculate MR:
MR Mn beff d 2 R=12,000
or
MR Mn beff d2 R0 =12,000
T Cf 0:85fc0 bw a hf
T Cf 0:85fc0 bw hf 0:85fc0 bw a
T Cf 0:85fc0 bw hf 3:5
a
0:85fc0 bw
T Cf
a hf
0:85fc0 bw
Step 6. Locate the neutral axis (c) and check to ensure the section satises the ACI
Codes requirements for being in the tension-controlled or transition zones.
The neutral axis is located at:
a
c
1
c 3
If > the section does not satisfy the ductile failure requirements,
dt 7
as t < 0.004 when c 0.003.
c 3
If , determine the strength reduction factor, using the relation-
dt 7
ships below:
c 3
if ! 0:90
dt 8
c 3 B2
if > ! A2
dt 8 c=dt
Ay
y 3:6
A
y is the distance from the top of the beam to the centroid of the
compression zone.
The moment arm (z), which is the distance between the tensile and
compression forces, is:
zdy
MR Mn Tz 3:7
160 3 Special Topics in Flexure
Step 8. After computing MR, check to ensure the beam is adequate. Also check the
depth of the beam for deection (Figure 2.41).
Example 3.1 Figure 3.7 shows the partial oor framing plan and sections of a
reinforced concrete oor system. The slab is 4 in. thick, and the weight of mechan-
ical/electrical systems, ceiling, and oor nishing is 24 psf. The oor live load is
200 psf. The beam ends are integral with their support, fc0 3 ksi, fy 60 ksi, and a
24 in. x 24 in.
Column (typical)
A A B C
32'-0" 32'-0"
1
(B-1)
10'-0" (S-1)
(S-1)
(B-2)
10'-0" (S-1) (S-1)
A
2
(B-2)
10'-0" (S-1)
(S-1)
4 in.
18 in. 18 in.
10'-0"
Figure 3.7 Floor framing plan and section for Example 3.1. (a) Partial oor framing plan.
(b) Section A-A
3.1 T-beams 161
unit weight of concrete of 150 pcf. Stirrups are #4 bars. Use ACI coefcients for
calculation of bending moments.
a. Check the adequacy of the edge beam (B-1) at midspan.
b. Check the adequacy of the interior beam (B-2) at midspan.
Solution Use the owchart of Figure 3.6.
(a) Edge Beam (B-1) B-1 is an L-beam for positive moment (midspan):
Step 1. Calculate the effective ange width:
n sw
beff min bw , bw 6hf , bw
12 2
where
n 32 2 12 360 in:
bw 18 in:
hf 4 in:
sw 10 12 18 102 in:
360 102
beff min 18 , 18 64, 18
12 2
beff minf48 in:, 42 in:, 69 ing 42 in:
Step 4. Determine the total compression force, Cf, assuming that the compression
zone is within the ange:
Cf 0:85fc0 beff hf
Cf 0:853424
Cf 428k
T 468k > Cf 428 kip
162 3 Special Topics in Flexure
Because T > Cf , the assumption in step 4 was not correct, and the
compression zone has to be larger in order for Cf to be equal to T. Thus,
the compression zone extends below the ange.
Step 5. Determine the depth of the compression zone, a:
T Cf
a hf
0:85fc0 bw
468 428
a 4
0:85 3 18
a 4:87 in:
a 4:87
c
1 0:85
c 5:73 in:
c 5:73 3
0:210 < 0:429 ok
dt 27:3 7
3
0:210 < 0:375 0:90
8
0.85f c
beff = 42 in.
26 in.
5 #11
T fy
bw = 18 in.
Ay
y
A
4 0:87
42 4 18 0:87 4
2 2
y
42 4 18 0:87
y 2:21 in:
10,568 kip-in
MR Tz 0:90 468 25:09 881 ft-kip
12
Step 8. To ensure that the beam can carry the loads, calculate the maximum
factored moment after determining the loads.
4
Weight of slab 150 50 psf
12
Superimposed dead loads 24 psf
Total dead load 74 psf
18 26
74 5:75 150
12 12
wD 0:913 kip=ft
1,000
200 5:75
wL 1:15 kip=ft
1,000
wu 1:2wD 1:6wL 1:2 0:913 1:6 1:15 2:94 kip=ft
n 32 2 30 ft
wu 2n 2:94302
M u 189 ft-kip
14 14
Mu 189 ft-kip < MR 881 ft-kip ok
Check the beam depth for deection. See Figure 2.41. (B-1 is a one-end
continuous beam):
32 12
hmin
18:5 18:5
hmin 21 in: < h 30 in: ok
164 3 Special Topics in Flexure
B 1 is ok:
(b) Interior Beam (B-2) B-2 is a T-beam for positive moment at midspan.
Step 1. Determine the effective ange width:
n
beff min bw , bw 16hf , bw sw
4
where
As 9:36
0:0044
beff d 82 26:1
Step 6.
0:0044
R 225 psi
fc0 3 ksi ! Table A2:6a !
0:90
fy 60 ksi
Step 7. Calculate the design resisting moment, MR:
beff d2 R
MR
12,000
8226:12 225
MR
12,000
MR 1,047 ft-kip
18 26
74 10 150
12 12
wD 1:23 kip=ft
1,000
200 10
wL 2:0 kip=ft
1,000
wu 1:2wD 1:6wL
wu 1:2 1:23 1:6 2:0 4:68 kip=ft
wu 2n 4:68302
M u 301 ft-kip
14 14
Mu 301 ft-kip < MR 1,047 ft-kip ok
In theory, the design of T-beams involves nding the ange thickness, the width
and depth of the web, and the amount of reinforcement required. In practice,
however, the ange thickness is determined when designing the slab. The size of
166 3 Special Topics in Flexure
the web is selected to resist not only the moments at the supports (no T-beam
action), but to provide adequate shear capacity, and to simplify formwork layout for
ease of construction.
Hence, when designing a T-beam, the geometric dimensions of the beam
typically are known. The only unknown is the amount of steel required to resist
the loads. The T-beam design procedure, like beam analysis, depends on the
required depth of the equivalent stress block. In most cases, the compression zone
is within the ange area; so the design follows that of a simple rectangular beam,
with a width equal to the effective width of the ange.
In some rare cases, however, the compression zone available within the depth of
the ange, may not be adequate to develop the necessary factored moment. The
difference then must be compensated by having an additional compression zone
below the bottom of the ange (within the web).
The steps for the design of T- and L-beams follow. These are summarized in the
owchart of Figure 3.10.
Step 1. Calculate the maximum factored moment that the beam must carry (Mu).
Step 2. Determine the effective ange width (beff) based on the ACI requirements.
Step 3. Assume a single layer of reinforcement y 2:5 in: and the effective depth,
d h y. In addition, assume 0.90.
Step 4. Calculate MRf using Equation (3.8). MRf is the moment capacity when the
compression zone is only within the ange.
MRf Mnf 0:85fc0 beff hf d hf =2 3:8
Step 5. Case a: If Mu MRf ! the compression zone is entirely within the ange.
Case b: If Mu > MRf ! the ange area is not adequate to develop the
required factored moment.
Case a: Compression zone is within the ange (a hf).
Step 6. Calculate the resistance coefcient, R:
12,000Mu
R
beff d2
Step 7. Use fy, fc0 , and R to obtain and from Tables A2.5 through A2.7.
Step 8. Calculate the required area of steel, As:
As beff d
d hy
a
beff
hf
a
As
bw
b beff c
Cf
Cw
h
df = d zf
+ hw
dw
A sf Asw
Asf fy Asw fy
bw
Figure 3.9 T-beam design where the compression zone extends below the bottom of the ange.
(a) Compression zone extends below the ange. (b) Compression in ange only. (c) Compression
in web only
168 3 Special Topics in Flexure
12MRf
Asf 3:9
fy zf
1.
Find the maximum Mu
2.
Find beff:
beff = min {bw + n /4, bw + 16hf , bw + sw} (T-beam)
beff = min {bw + n /12, bw + 6hf , bw + sw /2} (L-beam)
3.
Assume y = 2.5 in. and
= 0.90, d = h - y.
4.
Find MRf = Mnf (design resisting moment
for compression zone covering the entire flange):
h
(
MRf = (0.85 f c ) beffhf d - f
2 )
Case a Case b
5.
Compression zone is No Yes Compression zone extends
Mu > MRf?
within flange (a hf ). below flange (a > hf )
6.
6. 12,000Mu Assume df = h - 2.5 in.
R = h .
beffd 2 and zf = df - f
2
7. 12MRf
7.
Use fy, f c , and R to find Asf =
fy zf
and from Tables
A2.5 through A2.7.
8.
hw = h - hf and
assume dw = hw - 2.5 in.
8. As = beffd
12,000(Mu - MRf )
Rw =
As,min = minbw d bw d w2
C A B
C A B
9.
No Use fy , fc and Rw to find
As >As,min ? As = As,min w fromTables A2.5
through A2.7.
Asw = w bw dw
Yes
10.
As = Asf + Asw
11.
Select the size and
Select the size and
number of bars using
number of bars using Tables A2.8 and A2.9.
Tables A2.8 and A2.9.
9. 12.
d = h-y
d = h-y
No
No d dassumed?
d dassumed?
Yes Yes
10. 13.
Check the beam depth Check the beam
for deflection. depth for deflection.
Step 8. In order to calculate the area of required steel for the part of the compres-
sion zone that is below the ange (Asw), consider only the depth of the stem
that is below the ange (hw h hf). Assume that the effective depth of the
stem (dw) is dw hw 2.5 in. See Figure 3.9c.
Then use Equation (3.10) to calculate the resistance coefcient for the
required area of steel in the web (Rw).
12,000 Mu MRf
Rw 3:10
bw d2w
Step 9. Use fy, fc0 , and Rw to obtain w (steel ratio for the web) from Tables A2.5
through A2.7, and calculate the required area of steel in the web (Asw)
using Equation (3.11).
Asw w bw dw 3:11
Step 11. Select the size and number of bars using Tables A2.8 and A2.9.
170 3 Special Topics in Flexure
Step 12. Based on the size and number of the selected bars, compute the actual
beam effective depth d h y. If this value is larger than what was
assumed in step 3, the design is a little conservative. Otherwise, revise the
design as needed by using the new value of effective depth.
Step 13. Finally, check the beam depth (h) for deection requirements.
Example 3.2 Because beams B-1 and B-2 of Example 3.1 were overdesigned,
redesign these L- and T-beams respectively for the maximum positive moments at
midspan.
Solution Use the owchart of Figure 3.10.
(a) Edge Beams (B-1)
Step 1. From Example 3.1a, step 8:
Mu 189 ft-kip
Step 2. From Example 3.1a, step 1, the effective ange width (beff) is:
beff 42 in:
Step 4. Equation (3.8) gives the design resisting moment if the entire ange is in
compression (MRf):
hf
MRf Mnf 0:85fc0 beff hf
d
2
4
MRf 0:900:85 3424 27:5
2
9,832 in-kip
MRf 819 ft-kip
12
Step 5. Because Mu 189 ft-kip < MRf 819 ft-kip, the compression zone will be
within the ange (a < hf).
Step 6. Calculate the resistance coefcient, R:
12,000Mu
R
beff d2
12,000 189
R 71 psi
42 27:52
Step 7.
fc0 3 ksi
fy 60 ksi ! Table A2:6a ! 0:0014
R 71 psi
3.1 T-beams 171
4 0:875
y 1:5 2:44 in:
8 2
d h y 30 in: 2:44 in: 27:56 in: dassumed 27:5 in: ok
hmin one-end continuous beam
18:5
32 12
hmin 21 in: < h 30 in: ok
18:5
4 in.
30 in.
3 #7
18 in.
Mu 301 ft-kip
172 3 Special Topics in Flexure
Step 2. From Example 3.1b, step 1, the effective ange width (beff) is:
beff 82 in:
Step 4. Equation (3.8) gives the design resisting moment for the beam with the
entire ange in compression (MRf):
hf
MRf Mnf 0:85fc0 beff hf d
2
MRf 0:900:85 3824 27:5 4 2
19,195 in- kip
MRf 1,600 ft-kip
12
Step 5. Because Mu 301 ft-kip < MRf 1,600 ft-kip, the compression zone will
be within the ange.
Step 6. Calculate the resistance coefcient, R:
12,000Mu
R
beff d2
12,000 301
R 58 psi
82 27:52
Step 7.
fc0 3 ksi
fy 60 ksi ! Table A2:6a ! 0:0011
R 58 psi
Step 8.
As beff d 0:0011 82 27:5
As 2:48 in:2
From Table A2:4 ! min 0:0033
As, min 0:0033 18 27:5 1:63 in:2
From Table A2:9 ! 3 #9 barsAs 3 in:2
From Table A2:8 ! bmin 10 in: < 18 in: ok
From Table A2:8 ! bmax 24 in: > 18 in: ok
4 1:128
y 1:5 2:56 in:
8 2
d h y 30 in: 2:56 in: 27:44 in: dassumed 27:5 in: ok
3.1 T-beams 173
32 12
hmin 21 in: < h 30 in: ok
18:5 18:5
4 in.
30 in.
3 #9
18 in.
Example 3.3 Design the T-beam shown in Figure 3.13. Assume that the effective
ange width is 54 in. The T-beam is subjected to a total factored positive moment,
Mu 950 ft-kip. Use fc0 3 ksi, and fy 60 ksi. Assume #4 stirrups.
beff = 54 in.
3 in.
30 in.
20 in.
Step 5. Because Mu 950 ft-kip > MRf 806 ft-kip, the compression zone will
extend into the web area (a > hf). Use Case b.
Step 6. First calculate the amount of steel needed to work with the entire ange in
compression (Asf), and then the reinforcing needed to work with the part of
the web that is in compression (Asw). The total required area of steel (As)
will then be:
As Asf Asw
12MRf
Asf
fy zf
12 806
Asf 6:89 in:2
0:9 60 26
Then
The resistance coefcient for the area of steel required for the part of
the compression in the web (Rw) is:
3.1 T-beams 175
12,000 Mu MRf
Rw
bw d2w
12,000950 806
Rw 144 psi
2024:52
Step 9.
fc0 3 ksi
fy 60 ksi ! Table A2:6a ! 0:0028
Rw 144 psi
Asw w bw dw 0:00282024:5
Asw 1:37 in:2
Step 11.
From Table A2:9 ! Try 6 #11 barsAs 9:36 in:2
From Table A2:8 ! bmin 19:5 in: < 20 in: ok
From Table A2:8 ! bmax 54 in: > 20 in: ok
3 in.
30 in.
6 #11
20 in.
3.2.1 Introduction
To this point we have shown the use of steel reinforcement only for the tension part
of a reinforced concrete beam (tension steel). When reinforcement is also used in
the compression zone of a reinforced concrete section (compression steel), the
beam is referred to as a doubly-reinforced beam. Even though such a section in
general is not economical, the use of compression steel has several advantages and
applications, including the following:
1. It allows the use of a cross section smaller than that of a singly-reinforced beam.
This is especially useful if the beam size is limited for architectural or aesthetic
purposes.
2. It helps in reducing long-term deections.
3. It can support stirrups or shear reinforcement by tying them to compression bars.
4. It adds signicantly to the ductility of beams. Compression reinforcement
enables the beam to withstand large levels of movement and deformation
under extreme loading conditions that might occur during earthquakes.
5. It is frequently used where beams span more than two supports due to practical
considerations. The ACI Code requires a percentage of the tensile steel at
midspan to continue into the supports, and by a small extension this steel can
easily be used as compression reinforcement at the face of the supporting
column.
0.85fc'
b
a C1 C2
A's
As
= +
T1 T2
fy
The following notations will be used in this section, and are shown in
Figure 3.16.
As As
As
= A s1
+ As2
b c = 0.003
d s
As c
d dt
As
c 0.85fc
b
a C1 = 0.85f c ba
d a
z1 = d
2
As1
T1 = As1fy
fy
d
d
C2 = A sf s
As
d
z2 = d d
As 2
T2 = As2fy
C2 T 2
3:13
As0 fs0 As2 fy
The following steps present the analysis of a doubly-reinforced beam. Figure 3.17
summarizes these steps in a owchart.
Step 1. Assume that the compression steel has yielded s0 y before the concrete
in compression has reached its ultimate strain. Therefore, fs0 fy and
As2 fy As0 fy
from which
As2 As0
Because
As As1 As2
From Figure 3.16c, which is the part of the beam represented by the
concrete-tensile steel couple, Equation (3.15) can be written.
C1 T 1 3:15
0:85fc0 ab As1 fy
Step 2. Calculate the depth of the compression zone (a) using Equation (3.16).
As1 fy
a 3:16
0:85fc0 b
Step 3. Determine the strain levels for the tensile steel (t) from the similarity of
triangles (see Figure 3.16b).
180 3 Special Topics in Flexure
1.
Assume that all steel in tension and compression yields.
(fs = f s = fy)
As 2 = As
As1 = As As
2. As1fy
a=
0.85 f c b
c= a
1
3.
t = 0.003(dt c)
c
4. 0.003(c d )
s =
Yes c
3a.
= A1 + B1 t t < 0.005?
No
= 0.90
Yes No
Case 1 s y? Case 2
5. a 5.
(
Mn1 = As1f y d 2 ) (0.85f c b 1)c 2 + (87As Asfy)c 87d As = 0
Solve for c.
6.
Mn 2 = Asfy (d d ) 6. 0.003(d c)
t = t
c
7. 6a.
Mn = Mn1 + Mn2 7. (c d ) Yes
f s = (87) t < 0.005?
c
8.
M R = Mn a = 1c No = A1 + B1 t
8. = 0.9
Mn1 = (0.85f c ba) d a
( )
2
Mn2 = As f s (d d )
9.
Yes
Section is ok. MR Mu? Mn = Mn1 + Mn2
No
Check beam MR = Mn
Section is N.G.
depth for
deflection.
t 0:003
dt c c
3:17a
0:003dt c
t
c
s0 0:003
0
cd c
3:17b
0:003 c d 0
s0
c
s0 y 3:18
This indicates that the compression steel yielded. In other words, fs0 fy or the
assumption made earlier in step 1 is correct. Hence, proceed directly to calculating
the resisting moment of the section.
Step 5. Calculate the nominal resisting moment from the concretetensile steel
couple according to Equation (3.19).
a
Mn1 As 1 fy z1 As 1 fy d 3:19
2
Step 6. Calculate the nominal resisting moment from the compression steeltensile
steel couple according to Equation (3.20).
Step 7. Calculate the nominal resisting moment for the doubly-reinforced beam
according to Equation (3.21).
Step 8. Calculate the design resisting moment (MR) using the strength reduction
factor () and Equation (3.22).
182 3 Special Topics in Flexure
MR Mn 3:22
s0 < y 3:23
Step 5. Because s0 < y , the compression steel did not yield when the strain at the
extreme compression edge on the concrete section reached 0.003. From
similar triangles (see Figure 3.16b) the strain in the compression steel can
be calculated using Equation (3.24).
0:003c d0
s0 3:24
c
The stress in the compression steel ( fs0 ) can then be calculated using
Equation (3.25).
0:003c d0
fs0 s0 Es Es 3:25
c
Thus, the assumption made in step 1 is not correct. The force provided by
the compression steel is less than was assumed. Hence a smaller amount of
tensile steel will work in the compression steeltensile steel couple, and a
new location has to be determined for the neutral axis.
Equilibrium requires that the total compression on the section be equal to
the total tension, as expressed by Equations (3.26) and (3.27).
C1 C2 T 1 T 2 3:26
0:85fc0 ab As0 fs0 As1 fy As2 fy
3:27
0:85fc0 ab As0 fs0 As1 As2 fy
Substituting a 1c and fs0 from Equation (3.25) into the above equation:
0:003c d0
0:85fc0 1 cb As0 Es As fy 3:29
c
3.2 Doubly-Reinforced Beams 183
Substituting Es 29,000 ksi, the location of the neutral axis (c) can be
determined from the quadratic Equation (3.31).
0:85fc0 b1 c2 87A0s As fy c 87d 0 As0 0 3:31
0:003dt c
t 3:32
c
a 1 c
Calculate the component forces of the internal couples (see Figure 3.16c
and d) and determine whether equilibrium is satised, as expressed in
Equation (3.35).
C1 0:85fc0 ba
C2 As0 fs0
T 1 T 2 As fy 3:34
184 3 Special Topics in Flexure
T 1 T 2 C1 C2 3:35
If Equation (3.35) is not satised, then most likely an error was made in
the computation of c.
Step 8. Calculate the nominal resisting moment of the doubly-reinforced section by
adding the concretetensile steel and compression steeltensile steel cou-
ples as shown in Equations (3.36)(3.38) (see Figure 3.16c, d).
a a
Mn1 C1 z1 C1 d 0:85 fc0 ba d 3:36
2 2
0 0 0 0
Mn2 C2 z2 C2 d d As fs d d 3:37
Mn Mn1 Mn2 3:38
MR Mn 3:39
Step 9. Once MR is calculated, determine whether the beam has enough capacity by
comparing MR with the maximum factored moment, Mu (i.e., the demand):
MR Mu Beam is adequate:
MR < Mu Beam is not adequate:
Check if the beam depth is large enough so that deection does not need
to be computed.
Example 3.4 Calculate the design resisting moment, MR, of the doubly-reinforced
beam shown in Figure 3.18. fc0 4 ksi, fy 40 ksi, Es 29,000 ksi. The stirrups
are #4 bars. The beam is subjected to a positive bending moment.
Solution Use the owchart of Figure 3.17.
14 in.
2 #7
#4 Stirrups
30 in.
6 #9
Step 1. Assume that the tension and compression steel yield. The validity of this
assumption will be checked later in the analysis.
fs fs0 fy 40 ksi
From Table A2:9 ! 6#9 ! As 6:0 in:2
! 2#7 ! As0 1:2 in:2
0
As 1 As As 6 1:2 4:8 in:2
Step 2. Calculate the depth of the equivalent stress block (a) in the concrete for the
section:
As1 fy 4:8 40
a 4:03 in:
0:85 fc0 b 0:85 4 14
a 4:03 in:
c 4:75 in:
1 0:85
Step 3. Calculate the strain in the tension and compression steel and check for the
validity of the assumption made in step 1. Make a sketch of the strain
distribution as shown in Figure 3.19 and calculate the strains from similar
triangles.
4 1:128
dt 30 in: 1:5 27:44 in:
8 2
0:003dt c
t
c
0:00327:44 4:75
t
4:75
t 0:0143
fy 40
t 0:0143 > y 0:00138
Es 29,000
14 in.
c = 0.003
s = 0.0015
c = 4.75 in. d = 2.44 in. 2 #7
dt d
30 in.
6 #9
t = 0.0143
Step 3a.
t 0:0143 > 0:005 0:90
4 0:875
d0 1:5 2:44 in:
8 2
0:003c d0
s0
c
0:0034:75 2:44
s0
4:75
s0 0:00146 > y 0:00138
Therefore, the compression steel yields, and the assumption in step 1 was
correct! Because the compression steel yields, follow the process under
case 1:
Step 5. Calculate the effective depth (d).
4 1
d 30 1:5 1:128 26:37 in:
8 2
Note that the clear vertical space between the bars is 1.0 in.
Calculate the nominal resisting moment from the concretetensile steel
couple, Mn1.
a
Mn1 As1 fy d
2
4:03
4:840 26:37
2
Mn1
12
Mn1 390 ft-kip
Step 6. Calculate the nominal resisting moment for the compression steeltensile
steel couple, Mn2.
Mn2 As0 fy d d 0
1:24026:37 2:44
Mn2
12
Mn2 95:7 ft-kip
Step 7. Calculate the total nominal resisting moment, Mn, which is the sum of the
concretetensile steel (Mn1) and compression steeltensile steel (Mn2)
couples:
3.2 Doubly-Reinforced Beams 187
Mn Mn1 Mn2
Mn 390 95:7 485:7 ft-kip
Example 3.5 Determine the design resisting moment, MR, for the doubly-
reinforced beam with 6 #8 bars used for steel in tension as shown in Figure 3.20.
Use fc0 4 ksi and fy 40 ksi, Es 29,000 ksi. The stirrups are #4 bars. The beam is
subject to a positive bending moment.
14 in.
2 #7
#4 Stirrups
30 in.
6 #8
As1 fy 3:54 40
a 0
0:85fc b 0:85 4 14
a 2:97 in:
a 2:97
c 3:5 in:
1 0:85
188 3 Special Topics in Flexure
0:003d t c
t
c
0:00327:5 in: 3:5 in:
t
3:5 in:
40
t 0:0206 > y 0:00138
29,000
Step 3a.
t 0:0206 > 0:005 0:90
4 0:875
d 0 1:5 2:44 in:
8 2
0:003c d 0
s0
c
0:003 3:5 2:44
s0
3:5
s0 0:00091 < y 0:00138
The compression steel does not yield when the strain in the concrete
reaches 0.003. Therefore, the assumption in step 1 was not correct. Hence,
follow the procedure outlined in case 2.
Step 5. Determine the location of the neutral axis, c, using Equation (3.31).
0:85fc0 b1 c2 87A0s As fy c 87d 0 A0s 0
0:85 4 14 0:85c2 87 1:2 4:74 40c 87 2:44 1:2 0
40:46c2 85:2c 254:7 0
Ax2 Bx C 0
(Quadratic equations have two roots. The one in the example has
c1 3.77 in. and c2 1.67 in. c cannot be a negative value, so the second
one obviously does not apply.)
Step 6. Determine the correct value of the net tensile strain at the extreme layer of
the reinforcement using Equation (3.32).
0:003d t c
t
c
0:00327:5 3:77
t
3:77
Step 6a.
t 0:0189 > 0:005 0:90
Step 7. Calculate the stress in the compression steel ( fs0 ) using Equation (3.33).
c d 0
fs0 87
c
3:77 2:44
fs0 87
3:77
fs0 30:69 ksi < fy 40 ksi
Step 8. Calculate Mn1 and Mn2, the nominal resisting moments for the concrete
tensile steel couple and the compression steeltensile steel couple,
respectively.
4 1
d 30 1:5 1 26:5 in:
8 2
a
Mn1 0:85fc0 ba d
2
3:2
0:85 4 14 3:2 26:5
2
Mn1
12
Mn1 316 ft-kip
Mn2 A0s fs0 d d0
1:2 30:6926:5 2:44
Mn2
12
Mn2 74:0 ft-kip
190 3 Special Topics in Flexure
Mn Mn1 Mn2
Mn 316 74:0 390 ft-kip
MR Mn 0:90 390
MR 351 ft-kip
If a singly-reinforced section cannot develop the required factored moment and the
beam size cannot be increased, a doubly-reinforced section may be appropriate. In
the design of doubly-reinforced beams, the section sizes are known, so only the
reinforcement needs to be determined.
The design of a doubly-reinforced section follows the same concept as that of the
analysis: Calculate the amount of steel necessary for the concretetensile steel and
compression steeltensile steel couples and add the results. The step-by-step design
procedure is outlined below and summarized in a owchart in Figure 3.21.
Step 1. Calculate the maximum factored moment, Mu, from the loads acting at the
section under consideration. Because the beam sizes (b and h) are known,
estimate the effective depth (d) as
1.
Calculate maximum Mu
2.
Use f c and fy to find tc (Table A2.3) and corresponding R from the resistance
coefficient tables (Tables A2.5 through A2.7).
3. Rbd 2
M n1=
12,000
No Design as a
Mn1 < Mu?
singly-reinforced beam.
Yes
4. As1 = tc bd
Mn2 = Mu Mn1
b
Mn2
C2 = d
(d d )
= 0.90 As
d dt
5. As
As1fy
a=
0.85f c b
y
a
c =
1
6.
0.003(c d )
s =
c
No Yes
f s = s Es s y ? f s = fy
7. C2
As =
f s
B A
B
A
f s As
As2 =
fy
8. A = A + A
s s1 s2
9.
d dassumed
No
and/or
d d assumed
Yes
End
As1 tc bd 3:41
Mn2 Mu Mn1
Mn2 C2 z2 C2 d d0 3:42
Mn2
C2 A0s fs0 3:43
d d 0
Step 5. To calculate A0s we must rst determine the value of fs0 , as the compression
strain in the reinforcing may be less than the yield strain. In order to do this,
determine the location of the neutral axis and check the strain level in the
compression steel (see Figure 3.16c):
3.2 Doubly-Reinforced Beams 193
C1 T1
0
0:85fc ba As1 fy
As 1 fy
a
0:85fc0 b
a
c
1
0:003c d0
s0
c
The result enables the selection of the size and number of bars for the
compression steel (A0s ) and the tension steel (As) using Table A2.9.
Step 9. After the selection of the tensile and compression steel, calculate the actual
values of d and d 0 , and compare to the assumed values in step 1. If
d dassumed or d 0 dassumed
0
, the assumptions are conservative. However,
if these relationships are violated by more than in., a recalculation of As
and A0s , is necessary using the adjusted d and d 0 values by repeating the
process from step 2.
Example 3.6 Figure 3.22 shows the oor framing plan and sections of a reinforced
concrete building. The slab is 6 in. thick and there is a superimposed dead load of
25 psf. The oor live load is 125 psf. Assume that the beams are integral with the
columns, fc0 4 ksi, fy 60 ksi, and the unit weight of the concrete is 150 pcf. The
194 3 Special Topics in Flexure
A B C
Columns 40'-0" 40'-0"
14 in. 14in.
(typical) A
1
20'-0"
-
A
20'-0"
-
2
20'-0"
B
-
20'-0"
6 in. 6 in.
14 in.
14 in.
Section A-A
Section B-B
Figure 3.22 Floor framing plan and sections for Example 3.6
stirrups are #4 bars. Design the reinforcements for the edge beam and the rst
interior beam along column lines 1 and 2 (as shown in the sections A-A and B-B of
Figure 3.22) where the maximum negative moments occur. Consider doubly-
reinforced beams if necessary. Use ACI coefcients for the calculation of moments.
Solution Use the owchart of Figure 3.21.
(a) Edge Beam Along Line 1
Step 1. Find the factored loads on the beam:
3.2 Doubly-Reinforced Beams 195
6
Weight of slab 150 75 psf
12
Superimposed dead loads 25 psf
Total dead load 100 psf
Live load 125 psf
The uniformly distributed dead and live loads on the beam are (tributary
width 10.58 ft):
14 24
wD 100 10:58 150 =1,000 1:41 kip=ft
12 12
125 10:58
wL 1:32 kip=ft
1,000
(Note that live load reduction does not apply for the beams, because the
unit live load is in excess of 100 psf.)
wu 2n 3:838:82
M u 572 ft-kip
10 10
Also,
Step 2. Use fc0 and fy to obtain the maximum tension-controlled steel ratio (tc) from
Table A2.3:
Step 3. Calculate the design resisting moment based on the limit of reinforcement
for tension steel (tc):
Rbd 2 8181427:52
Mn1
12,000 12,000
Mn1 722 ft-kip > 572 ft-kip
Mn1 > Mu Design as a singly-reinforced beam:
12,000Mu
R
bd 2
12,000 572
R 648 psi
1427:52
Step 4. From Table A2.6b, obtain the steel ratio () for this R value:
y
6 in.
30 in.
d
#4 Stirrups
14 in.
14 27
wD 100 20 150 =1,000 2:40 kip=ft
12 12
125 20
wL 2:5 kip=ft
1,000
wu 1:2wD 1:6wL 1:2 2:40 1:6 2:5 6:88 kip=ft
wu 2n 6:8838:82
Mu 1,036 ft-kip
10 10
dassumed h y 33 in: 2:5 30:5 in:
d0assumed 2:5 in:
Step 2.
fc0 4 ksi ! Table A2:3 ! tc 0:0180
fy 60 ksi
0:018 ! Table A2:6b ! R 818 psi
Rbd 2 8181430:52
Mn1
12,000 12,000
Mn1 888 ft-kip < Mu 1,036 ft-kip
Step 4.
As1 tc bd 0:0181430:5
As1 7:69 in:2
Mn2 Mu Mn1
Mn2 1,036 888 148 ft-kip
Mn2 148 12
C2 70:5 kip
d d0 0:930:5 2:5
As1 fy 7:69 60
a
0:85fc0 b 0:85 4 14
a 9:69 in:
a 9:69
c 11:4 in:
1 0:85
0:003c d 0
s0
c
0:003 11:4 2:5
s0 0:0023
11:4
fy 60
y 0:00207 < 0:0023
Es 29,000
fs0 fy 60 ksi
Step 7. Calculate the required compression steel (As0 ) and the additional tension
steel for the compression steel-tensile steel couple (As2).
C2 70:5
A0s 1:18 in:2
fs0 60
fs0 A0s 60 1:18
As2 1:18 in:2
fy 60
Step 8. Calculate the total required tensile steel (As) and select the bars.
3.2 Doubly-Reinforced Beams 199
As As1 As2
As 7:69 1:18 8:87 in:2
Step 9. Calculate the actual values of d and d 0 , and compare to the assumed values.
4 1:128
d h y 33 1:5
8 2
30:44 in: d assumed 30:5 in:
4 0:875 0
d0 1:5 2:44 in: < dassumed 2:5 in: ok
8 2
6 in.
9 #9
#4 Stirrups 27 in.
2 #7
14 in.
where db is the diameter of the main bars, dt is the diameter of the transverse
reinforcement (stirrups), and bmin is the smaller dimension of the beam section.
200 3 Special Topics in Flexure
b A
s A-A
A
Figure 3.25 Lateral support for compression steel in doubly-reinforced beams. (a) Possible
buckling of compression reinforcement without adequate stirrups. (b) Stirrups for doubly-
reinforced beams
3.3.1 Introduction
5 w4
384 EI
where wc is the weight of the concrete in pounds per cubic ft and fc0 is the 28-day
cylinder strength of the concrete in psi. (Normal-weight concrete is about 145 pcf.)
The Code formula for the modulus of elasticity is accurate only within a range of
about 15 %.
Calculating the moment of inertia is even more problematic. Concrete exural
members, as discussed earlier, develop cracks while subject to normal service load
conditions. Between the cracked sections and the points where the moments are less
than the cracking moment (Mcr), there is the full concrete section augmented by the
reinforcing. At the cracked sections, however, only a much smaller moment of
inertia is available. Correspondingly, the center region of a beam has considerably
less rigidity as shown in Figure 3.26.
Mcr
The third major uncertainty is due to the creep behavior of concrete in compres-
sion. The rst two uncertainties inuence the ambiguity of calculating the so-called
instantaneous deections, but creep inuences long-term deformation (i.e., a grad-
ually increasing deformation under sustained loads). Fortunately, the rate of
increase of deformation dissipates with time, and it virtually stops after about
5 years.
202 3 Special Topics in Flexure
The ACI Code simplies the complex problem posed by uncracked and cracked
sections in different regions of beams by assuming that the effective moment of
inertia (Ie) lies somewhere between the gross sections moment of inertia (Ig) and
the cracked sections moment of inertia (Icr). Equation (3.48) [ACI Equation
24.2.3.5a] presents the ACI Code (Section 24.2.3.5) formula to calculate Ie.
" #
Mcr 3 Mcr 3
Ie Ig 1 I cr 3:48
Ma Ma
where
Ig is the moment of inertia of the gross concrete section about its centroidal axis,
neglecting reinforcement
Icr is the moment of inertia of the cracked concrete section
Mcr is the cracking moment
Ma is the actual (unfactored) maximum moment in the member
The ACI Code (Section 24.2.3.6) recommends using an average of values
obtained from Equation (3.48) for the critical positive and negative moment
sections in calculating Ie for continuous beams. This averaged value should be
used in the appropriate deection formulae for continuous beams.
a. Equation (3.49) gives the gross moment of inertia for a rectangular section.
bh3
Ig 3:49
12
fr I g
Mcr 3:50
yt
where fr is the modulus of rupture given by Equation (3.51) (for values of refer
to Section 1.6)
p
fr 7:5 fc0 3:51
and yt is the distance from the sections centroidal axis (neglecting reinforce-
ment) to the extreme ber in tension for a rectangular section, as shown in
Figure 3.27:
h
yt
2
3.3 Deection of Reinforced Concrete Beams 203
d
h
yt
bw
d
h
yt
bw
b. For a typical T-beam section like the one shown in Figure 3.28, calculating yt
and the gross moment of inertia requires considerable computational effort. To
ease the difculty, Table 3.1 is provided, which gives coefcients (Cyt) as a
function of the t/h and bw/b ratios. Then we can calculate the distance from the
sections centroidal axis to the bottom using Equation (3.52).
yt Cyt h 3:52
The gross moment of inertia of T-beams about the centroidal axis can be
determined with the help of Table 3.2, which gives coefcients (CIg) for differ-
ent ratios of t/h and bw/b. Then we can use Equation (3.53) to calculate the gross
moment of inertia.
I g CIg bh3 3:53
Example 3.7 Calculate yt, Ig, fr, and Mcr for a T-beam made of normal-weight
concrete with the following data: b 60 in., bw 12 in., t 4 in., h 24 in., and
fc0 4,000 psi:
Solution
bw 12 t 4
0:2, 0:167
b 60 h 24
The cracking moment in the positive moment regions (i.e., tension at the bottom) is:
fr I g 474 25,522
Mcr 757,984 in:-lb 63:2 ft-kip
yt 15:96
3.3 Deection of Reinforced Concrete Beams 205
The cracking moment in the negative moment regions (i.e., tension at the top) is:
fr I g 474 25,522
Mcr 1,504,655 in:-lb 125:4 ft-kip
h yt 24 15:96
fc kd b kdb
C Ec c and T As fs As Es s
2 2
Substitute the area of steel with a special kind of material (As)tr, which can take
tension and has an elastic response similar to that of concrete. The tension force
then can be calculated as:
T As Es s As tr Ec s
The transformed steel area (As)tr is shown in Figure 3.30 and can be calculated
using Equation (3.54).
Es
As tr As nAs 3:54
Ec
a b c
b
ec fc = E c e c
C = fc(kd )b/2
kd
d kd
es fs = E s e s T = Asfs
Figure 3.29 The cracked rectangular section: (a) strains, (b) stresses, and (c) the internal couple
206 3 Special Topics in Flexure
kd
d - kd
(As )tr = nAs
where n is the modular ratio, which is the ratio of the steels and the concretes
modulus of elasticity. The value of n Es/Ec may be rounded as shown in Table 3.3.
The centroidal axis (measured as kd from the top) is located where the rst moments
of the areas above and below that axis balance each other.
kd
bkd As tr d kd nAs d kd
2
By substituting As bd, the value for k can be calculated using Equation (3.55).
q
k 2n n2 n 3:55
The location of the neutral axis depends on only two parameters: the value of n that
depends on the concretes quality (because the modulus of elasticity of steel, Es, is
relatively constant and equal to 29,000 ksi), and the steel ratio, , employed in the
section.
Then the moment of inertia about the centroidal axis can be expressed as:
bkd3
I cr nbdd kd2
3
k3
I cr bd 3
n1 k2 3:56
3
3.3 Deection of Reinforced Concrete Beams 207
If the expression within the bracket is designated by Cr, the cracked moment of
inertia can be calculated easily from Equation (3.57) using values obtained from
Table 3.4.
I cr Cr bd3 3:57
bh3 12 203
Ig 8,000 in:4
12 12
As 2:37
0:0113
bd 12 17:5
Then:
n 8 0:0113 0:0904
Hence, the location of the neutral axis from the top is:
T-section Figure 3.31 shows a typical T-shaped concrete beam reinforced for
positive moment. The expression for k and the cracked section moment of inertia are
quite complicated for T-beams; however, there are easy solutions with certain
simplifying assumptions. During service load conditions, especially with large
amounts of reinforcing, the neutral axis may fall below the bottom of the ange
(in other words kd > t). Figure 3.32 shows the neutral axis and stresses for the general
case. If kd t, or n is less than the value shown in Table 3.5, the neutral axis is
within the ange and Equation (3.57) and Table 3.4 can be used to calculate Icr.
In the introduction we stated that the calculation of deections contains many
uncertainties, so the errors introduced with simplifying assumptions are minimal
and do not seriously inuence the validity of the results. The main simplication for
calculating deection for T-beams is that when the neutral axis at service load
conditions falls below the bottom of the ange, the portion of the compressive zone
that is within the web is neglected. As shown in Figure 3.32, this is usually a small
3.3 Deection of Reinforced Concrete Beams 209
d
h
As
bw
b
fc
t
kd
d kd
nAs
fs
bw
area combined with small stresses, and so the error is small. Thus, the neutral axis is
located where the rst moments of the transformed areas from above and from
below are equal:
t
bt kd nAs d kd
2
Introducing
As bd
210 3 Special Topics in Flexure
1 t 2
n
k 2 d 3:58
t
n
d
Table 3.6 provides k values for different n and t/d values. When kd > t, or n is
greater than the value shown in Table 3.5, the neutral axis is below the ange and
Table 3.6 should be used to nd parameters required to calculate Icr. The heavy
horizontal line in each column of Table 3.6 represents the limiting values of
Table 3.5.
I cr CrT bd 3 3:61
Example 3.9 The cross section used in Example 3.7 has 6 #11 bars in two rows,
As 6 1.56 9.36 in.2. Calculate the cracked section moment of inertia of the
T-beam.
Solution
Calculate the parameters required in Equation (3.60) for CrT. Because there are two
rows of reinforcing, use d h 4 in.; thus, d 24 4 20 in.
0:3143
CrT 0:2 0:0624 1 0:3142 1 0:2
3 3
0:2
0:22 0:314 0:2 0:3142 0:0393
3
Then
3.3.4 Applications
b 12 in:, h p
20 in:, d 17:5 in:, As 2:37 in:2 , I g 8,000 in:4 , I cr 3,376 in:4
fr 7:51:0 4,000 474 psi
0:7 0:4302
Ma 123:75 ft-kip
8
fr I g 474 8,000
Mcr 379,200 in:-lb 31:6 ft-kip
yt 20
2
The effective moment of inertia from Equation (3.48) is:
" #
Mcr 3 Mcr 3
Ie Ig 1 I cr
Ma Ma
" #
31:6 3 31:6 3
8,000 1 3,376 3,453 in:4
123:75 123:75
This value is only about 2.3 % higher than the cracked moment of inertia (Icr).
It hardly seems worth the trouble to go through the calculations.
LL = 0.4 kip/ft
DL = 0.7 kip/ft
30'-0"
About 1 in. of this deection is due to dead loads. The rest is due to live loads.
Example 3.11 Given the T-beam used in Examples 3.7 and 3.9, calculate the
instantaneous deections due to the dead and live loads shown in Figure 3.34.
Solution From dead loads:
2:0 402
MDL 400:0 ft-kip
8
0:8 402
MLL 160 ft-kip
8
I cr 18,864 in:4
LL = 0.8 kip/ft
DL = 2.0 kip/ft
40'-0"
Note again that the result is only slightly different from the value of Icr.
fc0 4,000 psi, from Example 3.10, E 3.64 106 psi, and therefore the instanta-
neous deections are:
2,000
5 40 124
DL 12 1:68 in:
384 3:64 106 18,890
800
LL 1:68 0:67 in:
2,000
The calculation of the live load deection (LL not exactly according to the ACI
requirements, as this value has to be computed by subtracting the dead load
deection from the total dead and live load deection. This total deection is
computed using the effective amount of inertia (Ie) based on the applied dead and
live load amounts. The difference in the results, however, is negligible.
The values of the cracked section moment of inertia and the effective moment of
inertia (as dened by the ACI Code) usually differ only slightly, as observed in
Examples 3.10 and 3.11. To make it easy to understand the reason, we now rewrite
Equation (3.48) as Equation (3.62).
" #
Mcr 3 Mcr 3 Mcr 3
Ie Ig 1 I cr I cr I g I cr 3:62
Ma Ma Ma
In other words, Ie is equal to Icr plus a fraction of the difference between Ig and Icr.
Because the cracking moment (Mcr) is usually much smaller than the actual
moment (Ma), their ratio, raised to the third power, is a small number. In building
structures the actual moment is about 6575 % of the ultimate moment; and in most
members the ratio of the cracking moment to the actual moment is less than 0.3
where the required reinforcing is at least two or three times the minimum As,min.
Hence, the multiplier to the (Ig Icr) is only 0.027 or less. Thus
I e I cr 0:03 I g I cr I cr
3:63
1 500
where 0 is the
ratio of compressive
reinforcing (if any) in positive moment regions
0
A
of the beam i:e:, 0
s
; and the time-dependent factor for sustained loads is
bd
equal to one of the following:
Tests have shown that the presence of compression reinforcing steel decreases
the additional long-term deformation. If no reinforcing exists on the compression
side, the deection due to sustained loads may grow to three times the instantaneous
deection in 5 years or more. Fortunately the rate of growth dissipates and becomes
very slow after about 3 years. The growth in deections virtually disappears after
about 5 years.
These additional deformations apply only to the part of the instantaneous
deections that the structure must sustain on a continuous basis. Thus they apply
to the dead loads and the part of the live loads that is continuously present. For
example, in a residential structure or an ofce structure, probably less than 15 % of
the design live loads are present continuously. In a library stack area or a storage
facility, on the other hand, 7580 % of the design live loads are present all the time;
thus, the live loads in these facilities contribute a great deal to the long-term
deformations as well.
Table 3.7 summarizes deections that are permissible according to the ACI
Code Section 24.2.2. Note that the main concern is damage to nonstructural
elements that are supported by, or are attached to the concrete structure. These
elements most frequently are walls, or, in some rare occasions, ceilings. So aside
from the fact that some shallow elements with really long spans may also exhibit
undesirable vibrations (very rare in concrete structures), the issue is not the mag-
nitude of the deection, but what it may cause. For example, when a beam or a slab
deects, a partition wall may unintentionally become a support to the beam or slab.
If the partition wall cannot take that load without cracking or buckling, then that
partition wall will fail, while nothing terrible happens to the beam or slab whose
action caused the failure. So the designers job is to evaluate the consequences
arising from the inevitable deections and take steps to avoid potential harm to
216 3 Special Topics in Flexure
Table 3.7 Maximum permissible computed deections [ACI Code Table 24.2.2]
Deection
Type of member Deection to be considered limitation
Flat roofs not supporting or attached Immediate deection due to live, /180
to nonstructural elements likely to be snow or rain loads
damaged by large deections
Floors not supporting or attached to Immediate deection due to live load /360
nonstructural elements likely to be
damaged by large deections
Roof or oor construction supporting That part of the total deection /480
or attached to nonstructural elements occurring after attachment of
likely to be damaged by large nonstructural elements (sum of the
deections time-dependent deection due to all
Roof or oor construction supporting sustained loads and the immediate /240
or attached to nonstructural elements deection due to any additional live
not likely to be damaged by large load)
deections
neighboring elements. In the example cited above, the easy solution is to connect
the partition wall at its top in such a way that it permits the deection of the
structure above it and, at the same time, provides lateral support to the wall.
The integrity of reinforced concrete requires that there be no slippage between the
reinforcement and the surrounding concrete. The whole theory of design is based on
that assumption.
Figure 3.35 shows a small piece of a beam with applied moments. As the
moment changes along the length of a beam, so does the tension in the reinforcing
steel.
When the reinforcing is isolated, as shown in Figure 3.35b, the role of the bond
stresses becomes quite clear. They transfer the difference in the tensile force, T,
from the steel to the concrete surface surrounding the bar, and vice versa.
The magnitude of the bond stresses varies along the length of the beam with the
rate of change in the moments. Where the moments change rapidly, the bond
stresses are high; and moments change rapidly where shears are high. Hence,
where shears are high, the bond stresses also are high.
The use of deformed bars results in three distinct effects that resist relative
slippage between the surface of the reinforcement and the concrete. The rst is
chemical adhesion between the two materials. The second is friction on the surface
of the bar. (Reinforcing bars are not smooth; in fact, they have a rather rough
3.4 Reinforcement Development and Splices 217
b
M M + M
Bond stresses
T T + T T T + T
Figure 3.35 (a) A x long portion of a beam and (b) isolated reinforcing as a free body
a b
Chemical adhesion and friction Reactions from concrete to ribs
T T + T T T + T
Figure 3.36 (a) Adhesion and friction forces on the bar surface and (b) reactions on the
deformation ribs
surface.) The third comes from the concrete bearing on the ridges of the deforma-
tions. These effects are shown schematically in Figure 3.36.
Research has shown that the following sequence occurs at the bar/concrete
interface. Initially, chemical adhesion bonds the two together. After the adhesion
breaks down, friction and the reactions on the ribs become engaged. These reactions
are at an angle to the axis of the bar, as seen in Figure 3.36b. The angle depends on
the slope of the ribs surface and the rib conguration. For simplicity ribs are shown
perpendicular to the bar, although they very often have different orientations.
Forces of the same magnitude but with opposite sense act on the surrounding
concrete. The component of these forces that acts parallel with the axis of the bar
counteracts T. The component that is perpendicular to the bar axis, however,
develops outward pressures from the bar to the concrete. Figure 3.37 shows these
two components of the reaction. The perpendicular component, in turn, results in
circumferential tensions in the concrete, similar to those in a pipe under pressure.
The circumferential tensions affect a cylindrical portion of the concrete that surrounds
the bar, as shown in Figure 3.37b. If the bar is too close to the outside of the concrete,
the cylinder is too thin, and cracks may appear on the side or bottom of the beam,
indicating a splitting failure. If bars are too close to each other, the two cylinders
overlap, and a split may develop in a horizontal plane between the bars. Thus, the
closer the bar is to the surface (small concrete cover), or the closer parallel bars are to
each other, the greater is the likelihood of splitting failure due to bond stresses.
Bond stresses change along even a small length of the beam. Research has shown
that bond stresses spike next to exural cracks (there is no bond across the crack
width), and also where a reinforcing bar terminates. These highly localized peak
218 3 Special Topics in Flexure
b Cylindrical zones of
circumferential tensions
a
surrounding the bars
Outward (radial)
Reactions from component of
ribs to concrete reaction on
concrete
Possible fractures
Figure 3.37 (a) Reactions on the surrounding concrete from the ribs and (b) outward-pointing
radial pressures on the concrete
bond stresses do not signicantly endanger the safety of the structure provided that
an adequate length of bar extends beyond where the bar will be fully stressed to
yield at ultimate strength. This extra bar length is called embedment length and is
dened as the length necessary for the bar to develop its full capacity. Another,
more common name for this length is development length.
The ACI Code provides two ways to determine the required development length for
deformed bars and deformed wires.
The rst method (Section 25.4.2.2 of the ACI Code) is a simplied one, whereas
the second method is more involved. In both of these methods, however, the
formulae include all the important variables that inuence the bond strength. The
latter method, which is based on Equation (3.64) (ACI Code, Equation 25.4.2.3a of
Section 25.4.2.3), is the general approach used to calculate the development length.
Table A3.1 provides a description and values of the different factors.
2 3
6 3 fy 7
d 6
440 p t e s 7 5db 12 in: 3:64
fc0 c b K tr
db
where
cb K tr
2:5
db
Table A3.2 summarizes the simplied equations allowed by the ACI for the
calculation of the required development length for reinforcing bars in tension in
3.4 Reinforcement Development and Splices 219
lieu of using Equation (3.64). The required development length cannot be less than
12 in. (d 12 in.).
Table A3.3 shows the tensile bar development lengths (d) for fy 60 ksi and
fc0 3 ksi or 4 ksi. The reinforcing bars are assumed to be uncoated (e 1.0) and
not top bars (t 1.0), and the concrete is normal weight ( 1.0).
Example 3.12 Calculate the required development length, d, for a #7
epoxy-coated bottom bar. Assume normal-weight concrete, fc0 4,000 psi,
fy 60,000 psi, #3 stirrups, 1.5 in. concrete cover over the stirrups, and a 5 in.
center-to-center spacing of bars.
Solution Both methods will be used here:
(a) Using Equation (3.64):
Step 1. Obtain the factors values from Table A3.1.
cb K tr 2:31 0
2:64 > 2:5 Use 2:5
db 0:875
Step 3. Use values from steps 1 and 2 to calculate the required development
length.
When there is not enough space or length to transfer stresses using the entire
development or embedment length (e.g., when a beam terminates into a column),
the ACI Code permits the use of hooks or mechanical anchorage devices, or a
combination of these. The bending radii and the extensions for hooks are standard-
ized by the ACI Code and are shown in Figure 3.38. Equation (3.65) (ACI Code,
Sections 25.4.3.1 and 25.4.3.2) gives the required development length, dh, for bars
in tension when the end is terminated in a hook:
0 1
Bfy e c r C
dh @ q Adb maxf8d b , 6 in:g 3:65
0
50 f c
a
dh dh
D D
12db
b
6db for #3 through #5 bars
12db for #6 through #8 bars
135
6db for #3 through #8 bars
D D
Figure 3.38 ACI Code standard hooks. (a) For primary reinforcement. (b) For stirrups and ties
3.4 Reinforcement Development and Splices 221
of the outside of the bend, and r 1:0, where db is the diameter of the hooked bar
(ACI Code Section 25.4.3.3).
Example 3.13 Calculate the development length (dh) for the bar in Example 3.12
if it is terminated in a standard hook.
Solution From Table A3.4:
e 1:2
c 1:0
r 1:0
1:0
60,000 1:2 1:0 1:0
dh p 0:875 19:9 in: > maxf80:875, 6 in:g 7 in:
501:0 4,000
Example 3.14 Figure 3.39 shows a beam/column connection. The beam clear span
is 30 ft, wu 2.2 kip/ft, Mu 194 kip-ft, and Vu 33 kip at the face of the column.
Calculate the cutoff points for the top bars. fc0 4,000 psi, fy 60,000 psi. Assume
uncoated bars with normal-weight concrete.
Solution Check to determine if there is any excess reinforcement:
As 3:16
0:0127
bd 16 15:5
From Table A2:6b ! R 609 psi
MR bd 2 R=12,000 16 15:52 609=12,000 195 kip-ft
18 in.
> d
ext1 ext2
> dh 4 #8 bars
> d h = 18 in.
d = 15.5 in.
b = 16 in.
104 kip-ft
194 kip-ft
3.03 ft
8.03 ft
needed; in other words, where M 0. Cutting a section at the distance x from the
face of column, the equation for moment, Mx, is:
Mx Mu V u x wu x2 =2 0
2:2x2
194 33x 0 ! x 8:03 ft
2
Calculate the theoretical point where two of the four bars may be terminated. With
only 2 #8 bars, 0.0064, R 326 psi, and MR 104 kip-ft. Then
2:2x2
194 33x 104 ! x 3:03 ft
2
Calculate d using the general Equation (3.64). From Table A3.1 the necessary
factors are:
where n is the clear (net) span of the beam. The n/16 requirement covers the
uncertainty of the true location of the point of inection.
Thus, the two outer bars may be cut off at 8.03 + 1.88 9.91 ft 90 -1100 from the
face of the column.
These bars have to be long enough to develop their strength between their cut off
point and the theoretical cutoff point for the rst two bars. This length is
9.91 3.03 6.88 ft, which is considerably greater than the required development
length of 3.08 ft.
If the bars were extended straight into the column, they would not have the
length needed to develop their strength because d 3.08 ft, but the width of the
column is only 18 in. These bars must be bent into the column with a 90 hook, as
shown in Figure 3.39. Considering the factors of Table A3.4, check for the
adequacy of the available length to develop dh.
e 1:0uncoated bars
c 0:7side cover 2:5in: with concrete cover 2 in: beyond the 90-degree hook
1:0 normal-weight concrete
60,000 1:0 0:7 1:0 00
dh p 13:3 in: > maxf81, 6 in:g 8
501:0 4,000
where the constant 0.0003 has the unit of in.2/lb. The calculated length (dc) may be
multiplied by the reduction factors given in Table A3.5.
Table A3.6 gives the compression bar development length (dc) for fy 60 ksi,
and fc0 35 ksi, or more for r 1:0.
224 3 Special Topics in Flexure
Splices are often needed in construction, either because the required length of a bar
cannot be supplied, or because of practical construction considerations, such as
splicing column reinforcing just above the most recently cast oor.
Splices for #11 or smaller bars may be made either by simple overlap, by butt or
lap welding them together, or by using a proprietary splicing device.
Tension Splices
Lap splices The ACI Code (Section 25.5.2.1) gives the minimum length of lap for
tensile reinforcing as
Class A splice . . . . . . . . . max{1.0 d ; 12 in.}
Class B splice . . . . . . . . . max{1.3 d ; 12 in.}
Generally speaking, Class B splice is required for most cases. Class A splice is
permitted only when both of the following conditions are satised:
a. The area of reinforcement provided is at least twice that required by analysis
everywhere along the length of the splice;
b. Only one-half or less of the total reinforcing is spliced within the required lap
length.
Welded splices
Splices may be butt welded or lap welded, as shown in Figure 3.40. The ACI Code
(Section 25.5.7.1) requires that they be able to develop 125 % of the yield strength
of the reinforcing. Welding must conform to the Structural Welding Code
Reinforcing Steel (ANSI/AWS D1.4).
Butt-welded splices are preferred over welded lap splices. In the former the
tensile force travels in a straight path. In the latter there is an eccentricity equal to
the bar diameter. The resulting moment develops forces on the concrete perpen-
dicular to the spliced bar. These forces may result in local cracking along the bars in
the lap zone.
Welded splices are expensive, as they are very labor intensive. Butt-welded bars
usually require extensive preparation of the ends, and lap-welded splices take more
time to weld.
Problems 225
0
When fc < 3,000 psi, the calculated lap length must be increased by one-third.
Welded splices Welded compression splices are permitted. The rules for the
welding are the same as those for welding tension splices.
Problems
6 in.
30 in.
5 #9
16 in.
3.5. The gure below shows a cross section of the interior bay of a oor system.
The beam has a clear span of 240 -000 . The superimposed dead load is 20 psf.
What is the maximum allowable service live load on the oor in psf based on
the moment capacity of the beam at the midspan? For simplicity assume
the beam is simply-supported. Assume #4 stirrups and 1.5 in. cover. Use
fc0 4,000 psi, fy 60,000 psi. Neglect the top bars. Do not consider live
load reduction.
4 in.
18 in.
2 #11 2 #11
12 in.
10'-0"
3.6. Consider the oor system of Problem 3.5 with fc0 4,000 psi, fy 60,000 psi,
a superimposed dead load of 25 psf, and a live load of 60 psf. What is the
maximum allowable clear span for the beam? For simplicity assume the beam
is simply-supported. Do not consider live load reduction.
3.7. What are the required areas of reinforcement for the following T-beams? Use
fc0 4,000 psi, and fy 60,000 psi.
(a) beff 66 in., bw 12 in., hf 4 in., h 20 in., Mu 200 ft-kip
(b) beff 48 in., bw 12 in., hf 4 in., h 18 in., Mu 150 ft-kip
(c) beff 32 in., bw 10 in., hf 3 in., h 16 in., Mu 100 ft-kip
3.8. Select the reinforcement for the beam of Problem 3.5 if the superimposed
dead load is 40 psf and live load is 40 psf. Assume the beam is singly
reinforced with bottom bars (positive moment). Consider live load reduction.
Problems 227
3.9. What is the moment capacity, MR, of the doubly-reinforced beam shown
below? Use fy 60,000 psi, #4 stirrups, 1.5 in. cover, and d0 2.5 in.
(a) b 12 in., h 24 in., As 3 #9 and, A0s 2 #6; fc0 4,000 psi
(b) b 12 in., h 30 in., As 4 #9 and, A0s 2 #7; fc0 3,000 psi
(c) b 16 in., h 34 in., As 5 #10 and, A0s 2 #7; fc0 3,000 psi
Note: Reinforcements are in single layers.
As
As
3.10. Calculate the moment capacity, MR, of the rectangular beam shown below.
How much will this capacity increase if 3 #9 bars are added as compression
reinforcement? Assume d0 2.5 in., fc0 4,000 psi, fy 60,000 psi, and #4
stirrups and a cover of 1.5 in.
32 in.
4 #11
16 in.
14 in.
#3 @ 12 in. c/c
Stirrups
24 in.
4 #9
1.5 in.
Clear cover
3.13. Calculate the cracked and effective moments of inertia for the beam of
Problem 3.4 in the positive moment region (tension in the bottom). The
actual service load moment is Ma 500 kip-ft.
3.14. Calculate instantaneous deections due to the dead and live loads for the
T-beam of Problem 3.5. The oor live load is 100 psf.
3.15. The following rectangular reinforced concrete beam has a width bw 12 in.,
and a total depth h 24 in. It is reinforced with 2 #9 bars and #3@10 in.
c/c stirrups. Use fc0 3 ksi, fy 60 ksi, and clear cover 1.5 in. Assume
normal-weight concrete.
Answer the following questions:
(a) Use Table A3.3 to see whether sufcient development length is avail-
able for the 2 #9 bars.
(b) Check to see whether sufcient development length is available if 2 #7
bars were used in lieu of the 2 #9 bars.
(c) Use the simplied formula given in Table A3.2 to see whether sufcient
development length is available for the 2 #9 bars if fc0 5 ksi was used
in the beam.
(d) Calculate the required development length for the 2 #9 bars when
Equation (3.64) is used.
2 #9
24in.
5'-6"
#3@10 in.
Beam Section
3.16. A 6-ft-wide wall footing supports a 12 in. thick concrete wall. fc0 3 ksi and
fy 60 ksi. The maximum moment in the footing occurs at the face of the
wall. Concrete is normal weight. Answer the following questions:
Problems 229
3 in.
3.17. The reinforced concrete beam shown in the elevation below has a width of
15 in. and a total depth of 24 in. It is subjected to a factored moment of
260 kip-ft at the face of the column. Use fc0 4 ksi, fy 60 ksi, #3 stirrups,
and 1.5 in. clear cover. Answer the following questions:
(a) Determine the required development length (d) for the 4 #8 epoxy-
coated top bars using the simplied formulae shown in Table A3.2.
Calculate and use the permitted excess reinforcement factor (refer to
Table A3.1).
(b) Recalculate the required development length by using Equation (3.64).
Refer to Table A3.1 for denitions of the factors.
4 #8
24 in.
230 3 Special Topics in Flexure
3.18. Calculate the location (measured from the face of the support) where two of
the 4 #8 bars may be terminated. Use fc0 4 ksi, and fy 60 ksi.
4 #8
#3 Stirrups
18 in.
12'-0"
Self-Experiments
In the following self-experiments, you will learn the behavior of T-beams and
doubly-reinforced beams. Include in the nal report all the test details (sizes, time
of day you cast the concrete, amounts of water/cement/aggregate, problems
encountered, etc., with images showing steps of the tests).
Experiment 1
Cut several pieces of Styrofoam in the form of rectangular beams and a slab. Place
the beams and slab on two supports as shown in Figure SE 3.1.
Apply a load (a few pounds) on top of the slab and observe how much the beams
bend. Record the maximum deection of the beams.
Next, glue the rectangular beams to the slab and repeat the test. How much do
the beams deect? Compare the maximum deection between the two cases.
Explain the difference, if any, in the results obtained. What is the importance of
gluing the pieces together?
Self-Experiments 231
Slab
Loading
Beam
Supports
Experiment 2
In this experiment, make the T-beams of Experiment 1 from concrete. First, build
the forms for the beams and slab. Then, place wires at the bottom of the beams and
slab (with about in. cover), and place concrete in the form. Make stirrups to hold
the beam wires together as shown in Figure SE 3.2. Record all the different stages
of casting the beams and slab and placing the wires. Record your observations and
any problems encountered.
Stirrups
Main bars
Experiment 3
This experiment demonstrates the behavior of doubly-reinforced beams. Get a
piece of Styrofoam and cut it as a rectangular beam. Make two holes at the bottom
of the beam using a heated wire. Then pass wires through these holes. Place the
beams on two supports and apply a load at the center of beam. Record how much
and the manner in which the beam bends under the load.
Now, remove the wires and apply some glue to them. Again pass the wires
through the holes and wait for the glue to harden. Load the beam as before and
record how much and the manner in which the beam deects under the load. Is there
a change in the amount of deection? Why?
Make two holes at the top of the beam, apply some glue to the two wires, pass the
wires through the holes and wait for the glue to harden. Load the beam as before and
record how much and the manner in which the beam deects under the load. Do you
notice any differences? Does the addition of top wires help the beam in resisting the
load?
Experiment 4
In this experiment we construct a doubly-reinforced concrete beam. Using wood
and cardboard, make forms for the beam. Place two rows of wires at the top and
bottom and tie them together with smaller-sized wire representing stirrups
according to Figure SE 3.4.
Self-Experiments 233
Record all the different stages of making forms, placing bars, and casting
concrete. Record your observations and any problems encountered.
4.1 Introduction
Figure 4.2 shows a reinforced concrete beam and its shear and moment diagrams
due to some applied load. A small length of the beam, dx, bounded by sections 1 and
2, is selected. The shear and the moment are different at the two respective sections
(i.e., V1 > V2 and M1 < M2).
The change in the moment equals the area under the shear diagram, and the rate
of change in the moment equals the magnitude of the shear. This is expressed
mathematically as
dM M2 M1
V or V
dx dx
V
M
1 2
dx
V1 V2
M1 M2
The internal couples are substituted for the moments (see Figure 4.3);
(i.e., M1 T1z C1z and M2 T2z C2z), therefore T1 < T2 because M1 < M2.
When a small part of the beam that is below the horizontal section is isolated
(see lower part of Figure 4.3), equilibrium requires that the horizontal force acting
on that section balance the applied loads. The area of that horizontal section is bdx,
and if the stress (i.e., the force per unit area) is designated by v, we can derive
the following relationship:
M2 M1 dM
T2 T1 4:1
z z z
From equilibrium:
T 2 T 1 vb dx 4:2
4.2 Shear in Beams 237
C1 C2
V1
Horizontal z
section
V2
T1 T2
dx
v b dx
T1 T2
Figure 4.3 The internal couples on a short (dx) length of reinforced concrete beam
Hence:
dM
vb dx 4:3
z
dM
V vbz 4:4
dx
V
v 4:5
bz
v v
Figure 4.4 shows an isolated part of the beam in elevation. Inside this portion of
the beam a small 1 in. 1 in. 1 in. cube is selected. The shear stresses are
indicated on the elevation of this cube. Previously we showed what causes the
horizontal shears. The horizontal shears form a couple (shown in Figure 4.4 as a
counterclockwise couple). Because a couple can be kept in equilibrium only by
another couple, a clockwise couple must be acting on this cube. The clockwise
couple is furnished by equal-magnitude shears on the vertical sides of the cube.
238 4 Shear in Reinforced Concrete Beams
The existence of shears on both the horizontal and vertical sections of a beam is
known as the duality of shears. This means that shears of equal magnitude are
always present on both the horizontal and the vertical surfaces of a small cube
inside the beam.
Shears do not cause a problem for concrete. As a matter of fact, concrete is quite
strong in shear. Figure 4.5 shows the cube with the components of the shear stresses
parallel to the diagonals.
Components of v
v
In Figure 4.6 we substitute the shears with their components. Then we cut the
cube into two triangular wedges along the diagonal planes. Figure 4.7 shows that
tensile stresses are generated along one of the diagonal faces to maintain equilib-
rium. Figure 4.8 shows a section of the cube along the opposite diagonal. In this
case, compression stresses will exist to maintain the equilibrium. So the conclusion
here is that horizontal and vertical shears cause tension and compression in the
diagonal directions.
v
2
v
2 v
2
v
2
By looking at one of the wedges in Figure 4.7 we can derive the magnitude of the
p
diagonal tension. Because the area of the diagonal plane is 2 (the cube is
4.2 Shear in Beams 239
v
2
Tensile stresses
v
2
v
2
v
2
p
1 in. 1 in.) the magnitude of the diagonal tension is t 2, where t is the diagonal
tensile stress. The equilibrium equation for forces along the diagonal is:
p
v
t 2 2 p 4:6
2
Thus
tv 4:7
In other words, the magnitude of the diagonal tensile stresses equals that of the
shear stresses.
We can perform a similar calculation for the wedges shown in Figure 4.8.
The result would show that the magnitude of the compressive stresses on the
diagonal face equals that of the shear stresses.
The above conclusion is valid only on a unit cube that is not subject to axial
stresses, as these do not occur at the neutral axis. Flexural compressive stresses
occur above the neutral axis while below the neutral axis we have exural tensile
stresses, for the beam shown in Figure 4.2.
v
2
Compressive stresses
v
2
v
2
v
2
Figure 4.8 Equilibrium on the perpendicular triangular wedge, resulting in compressive stresses
240 4 Shear in Reinforced Concrete Beams
v f
f v
Figure 4.9 shows a cube that is above the neutral axis. This cube, therefore, is
subject to axial compression as well as to shears. Imagine now a series of sections
cut through this cube. These sections are rotated by an angle, , from the horizontal
section. A detailed mathematical analysis can show that among all the possible
planes there exists a pair of planes, perpendicular to each other, where the resulting
normal stresses are the largest compressions or tensions, respectively. These planes
are called the principal planes. The stresses that act on the planes are the principal
stresses. Figures 4.10 and 4.11 show the orientation of the principal compression
and tensile stresses, respectively, on a section above the neutral axis.
The angle can be calculated as:
2v
tan 2 4:8
f
Figure 4.10 Orientation of the principal compressions above the neutral axis
4.2 Shear in Beams 241
Figure 4.11 Orientation of the principal tensions above the neutral axis
The axial stress below the neutral axis is tension (see Figure 4.12). Thus, the
orientations of the principal tensions will be similar to those shown in Figure 4.13.
v f
f v
Because the magnitudes of the exural stress and the shearing stress vary along
the beam as well as in relation to their distance from the neutral axis, the orientation
and the magnitude of the principal stresses also vary accordingly. Of the two
principal stresses, the tensile stress is the main concern here, as concrete is weak
in tension. The diagonal tensions, therefore, may tear the beam apart. A potential
crack starts out vertically at the bottom surface (because there is no shear at the
outer edge), then changes orientation gradually as shear is introduced, causing a
change in the principal stress direction. It crosses the neutral axis at 45 , because no
axial forces exist at that location (pure shear), and then attens out as it invades the
zone of larger compressive stresses. Figure 4.14 shows two such cracks that follow
the principal tensile stresses.
Shear (or, more precisely, diagonal tension) is a very complex problem. Thus a
simplied approach is used in the analysis and design of beams for shear. Although
simplied, the approach has been shown to provide safe and satisfactory design.
242 4 Shear in Reinforced Concrete Beams
Figure 4.13 Orientation of the principal tensions below the neutral axis
R
Figure 4.14 Potential cracks perpendicular to the principal tensions
The basic concept of shear reinforcement is the same as that of exural reinforcing.
If cracks begin to open due to lack of tensile strength in the concrete, reinforcement
is needed to transfer the tensile forces across the crack.
Vertical stirrups are used almost exclusively in modern concrete construction
for shear reinforcement in beams (see Figure 4.15). The terminology shear
reinforcement comes from the fact that shear is used as a measure of the diagonal
tension. (To confuse the issue even further, shear reinforcement is often also
referred to as web reinforcement.) Vertical stirrups typically are U-shaped #3
or #4 bars. They surround the tensile reinforcing on the bottom and are anchored
into the compression zone by a hook at each end. (See the beam section
in Figure 4.15).
4.3 The Design of Shear Reinforcement 243
s s
Vc
Vertical
stirrup
d
Vs
Vn
Vd
bw
Approximate d
The relationship between the design resisting shear, VR, and the nominal
resisting shear is:
VR Vn 4:10a
where is the strength reduction factor. As discussed in Chapter 2, the ACI Code
uses this factor to account for possible understrength of the materials and construc-
tion inaccuracies. The factor for shear (ACI Code, Section 21.2.1) is:
0:75 4:10b
which is smaller than the value for bending ( 0.90). The main reason for the
difference is that reinforced concrete beams are less ductile in shear than in
bending.
The design principle is to supply a greater strength than the required strength.
Expressed mathematically (Section 9.5.1.1 of the ACI Code):
VR Vu 4:11
The left side of the equation, VR, is the design shear strength of the section under
investigation. The right side, Vu, is the demand, or the shear acting on the section.
To describe the Vn, or the nominal strength of the section against shear failure,
Figure 4.15 shows a simple model that has been adapted in lieu of the very complex
interaction of the concrete and the various reinforcements. Other theoretical models
try to provide a practical solution to the problem. The empirical model adapted by
the ACI Code and discussed here is an easy-to-follow representation of the different
components of the available strength.
1. The rst component of the model is the shear strength of the concrete section, Vc.
The compression zone provides resistance due to friction and aggregate
244 4 Shear in Reinforced Concrete Beams
interlock. To make calculation easy, the ACI Code relates the value of Vc to an
average shear over the whole working section of the concrete beam (ACI Code,
Equation 22.5.5.1):
p
Vc 2 fc0 bw d 4:12
In this expression fc0 must be entered in psi, and bw and d are in inches. is the
light-weight concrete factor, which is equal to 0.75 for concrete made of all-
light weight aggregates, and is equal to 0.85 for sand-light weight concrete.
It is equal to 1.0 for normal weight concrete. The resulting Vc is in pounds.
Tables A4.1a, A4.1b, and A4.1c include Vc for different sizes of beam (bw and h)
and compressive strengths ( fc0 ) of normal weight concrete.
2. The second component of the model is the sum of the tensions developed by the
vertical component of the diagonal tensions in the stirrup legs (Vs). All stirrup
legs that cross a potential crack (Figure 4.15 shows three stirrups with two legs
each) will provide this strength. Thus
Vs n Av fyt 4:13
where n is the number of stirrups crossing the potential 45 crack. Because the
stirrups are placed at a spacing of s, and ns is approximately equal to d, we can
calculate the strength, Vs, (ACI Equation 22.5.10.5.3) as:
Av fyt d
Vs 4:14
s
where
Av sum of the cross sectional areas of the stirrup legs in square inches
s spacing of the stirrups in inches
bw width of the web of concrete beams in inches
d distance from the extreme compression ber to the centroid of the tensile
reinforcement in inches
fyt the specied yield strength of the transverse reinforcing steel (i.e., stirrups)
in ksi or psi (consistent units must be used)
Tables A4.2a and A4.2b show Vs for #3 and #4 stirrups with different s and h,
and fyt 60,000 psi.
3. The third component of the model that provides strength against shear is the
so-called dowel action. This results from the vertical shear resistance of
the horizontal reinforcing after the sides of the crack are vertically separated.
The contribution from this source is neglected by the ACI Code.
The nal design equation (based on ACI Code Sections 9.5.1.1, 9.5.3.1, and
22.5.1.1) is:
4.3 The Design of Shear Reinforcement 245
VR Vn Vc Vs Vu 4:15
From what was mentioned above, if Vu Vc (Vc is the shear that concrete can
carry), theoretically we do not need any stirrups. The ACI Code
(Section 9.6.3.1), however, requires a minimum amount of stirrups where
Vu > Vc/2. There are exceptions to this requirement as indicated in Table
9.6.3.1 of the ACI Code. This minimum amount of stirrups (ACI Code
Section 9.6.3.3) is:
( )
p bw s 50bw s
Av, min max 0:75 fc0 , 4:16
fyt fyt
The use of minimum amount of stirrups required by the code prevents
sudden shear failures when inclined cracking occurs. This rule has a few
exceptions, such as for slabs, footings, and concrete joist construction; in these
cases there is a possibility of load sharing between the weak and strong areas, so
no shear reinforcements are needed when Vu Vc (more on these topics in
Chapters 6 and 7). In any case, the ACI Code requires no shear reinforcements
where Vu Vc /2 in reinforced concrete beams.
In general, therefore, a reinforced concrete beam has three possibilities
(or zones) when designing for stirrups. Figure 4.16 shows these zones.
Note that design of beams for shear involves nding the spacing of stirrups
because almost all construction uses the same size stirrups for the entire beam.
The spacing is changed based on the level of shear force the stirrups have
to resist.
Critical section wu
2 in. or 3 in.
Zone Zone Zone
3 2 1
Vu Vc
Vc / 2
(Shear diagram)
No stirrups are needed where Vu Vc/2. For a symmetrically loaded beam this
condition usually occurs in a region close to the center of the beam, as shown in
Figure 4.16. Although the ACI Code does not require any stirrups in this zone, a few
stirrups are used to hold the main reinforcements in place.
This is a zone where theoretically no stirrups would be needed. The ACI Code,
however, requires a minimum area of stirrups. Because our objective here is to
determine the stirrup spacings, we rewrite Equation (4.16) to obtain the maximum
allowable spacing (s1) in terms of the selected stirrup size, the width of the beams
web, and the materials used in the beam.
( )
Av fyt Av fyt
s1 min p , 4:17
0:75 fc0 bw 50bw
The units used for this empirical equation are as follows: Av in:2 , fyt psi,
fc0 psi, bw in:, and s1 (in.).
The ACI Code (Section 9.7.6.2.2) places a further restriction on the maximum
allowable spacing in this zone (smax):
d
smax min s1 , , 24 in: 4:18
2
This requirement ensures that each 45 potential crack is intercepted by at least one
stirrup (Figure 4.15). Therefore, Equation (4.18) determines the stirrup spacing in
Zone 2 of the beam.
This is the only part of the beam for which we need to design the stirrups (i.e., this
zone may require closer stirrup spacing than the allowable maximum found in
Equation (4.18)). To determine the spacing of the stirrups in this zone, we need to
calculate how much shear the stirrups must carry (Vs). This is accomplished by
rearranging Equation (4.15):
4.3 The Design of Shear Reinforcement 247
Vu Vc Vs
Vu 4:19
Vs Vc
Vu
The rst term on the right side of this equation is the total factored shear on
the beam at the section under consideration magnied by the strength reduction
factor in the denominator and the second term (Vc) is the shear to be carried by the
concrete. The remainder is to be resisted by the stirrups (Vs). Rearranging
Equation (4.14) to nd the stirrup spacing:
Av fyt d
s 4:20
Vs
Usually Av, fyt, and d are the same for the entire beam. Therefore, the stirrup spacing
(s) changes with the shear to be resisted by the stirrups (Vs). Clearly the stirrup
spacing is smaller near the supports as Vs is larger. The calculated stirrup spacing
increases continuously as we move toward the midspan and the shear diminishes.
Although theoretically this is true, in reality only a few (two or three) different
spacings are used. While we could save a few stirrups by continuously varying the
stirrup spacings, this complicates construction, as locating and placing the stirrups
become difcult. As Figure 4.16 shows, the rst stirrup is usually placed 2 in. or
3 in. from the face of the support.
Because stirrups cannot resist shear unless they cross an inclined crack, the
ACI Code (Section 9.7.6.2.2) limits the maximum stirrup spacing. The maximum
allowable stirrup spacing is:
p d
if 0
Vs 4 fc bw d or Vs 2Vc ! smax min s1 , , 24 in:
2
4:21
p d
if 0
Vs > 4 fc bw d or Vs > 2Vc ! smax min s1 , , 12 in:
4
The rst part of Equation (4.21) limits the stirrup spacing such that each
potential 45 crack will be intercepted by at least one stirrup (Figure 4.15).
Where the shears are so large that the stirrups are required to carry Vs > 2Vc, the
maximum allowable spacing is limited to that shown in the second part of
Equation (4.21). This is necessary to provide better control of the width of the
potential inclined cracks.
Example 4.1 Determine the total resisting shear, VR, for the beam shown in
Figure 4.17. The shear reinforcements provided are #3 stirrups @ 8 in. on center.
Assume fc0 4,000 psi and fyt 60,000 psi. Concrete is normal weight.
248 4 Shear in Reinforced Concrete Beams
d 18.5 in.
h 21 in.
bw 12 in.
Solution
1.0 (normal weight concrete)
From Equation (4.12)
p
Vc 2 1:0 4,000 12 18:5 28,081 lb 28:1 kip
Vc 28:1 kip
Vs 30:5 kip
Thus
Solution
dest h 2:5 24 2:5 21:5 in:
Vu 60
Vs Vc 35:3 44:7 kip
0:75
2 0:11 60 21:5
s 6:35 in:
44:7
Rounding down to the nearest in., we use stirrups at 6 in. on centers at this
section.
Solution Using Tables
From Table A4.1a (interpolating for bw 15 in.)
Vc 35:3 kip
Then
60
Vs 35:3 44:7 kip
0:75
Entering into Table A4.2a with h 24 in., #3 stirrups with two legs at 6 in. spacing
will provide Vs 47:3 kip, which is slightly more than we need.
(b) The design yield strength of the shear reinforcing bars is limited to 60,000 psi
(ACI Code, Section 20.2.2.4). This requirement limits the crack width. The
limit is 80,000 psi when welded wire reinforcement is used as shear
reinforcing.
p
(c) The value of Vs is limited to 8 fc0 bw d (ACI Code, Section 22.5.1.2). This
p
provision effectively limits the maximum value of Vn to 10 fc0 bw d.
Stating it differently, Vs may not exceed 4Vc (Vs 4Vc). Thus, if Vu is too
large to satisfy this requirement, the concrete section must be enlarged by
making the beam wider
or deeper. Note that Vs,max for a beam usually is at its
critical section Vs .
(d) The critical section for stirrup design (within zone 3) may be taken at distance
d from the face of the support in beams and joists, when the loads are applied
onto the top of the beam. In the portion between the face of the support and the
critical section the support reaction introduces vertical compressions into the
end zone of the member, which signicantly increases the shear strength in
that region. Sections located between the face of the support and the critical
section may be designed for Vu at the critical section Vu . This means that for
design purposes the shear force from the critical section to the face of the
support is taken as, Vu , as shown in Figure 4.18. Note that we use the bar here
to indicate the value at the critical section (i.e., Vu , Vs and s represent the total
shear at the critical section, shear to be resisted by the stirrups at the critical
section, and the required stirrup spacing at the critical section, respectively).
Critical section
(e) Limit the stirrups size to #3, #4, or #5, as these bar sizes are easier to bend.
(This is only a recommendation, not an ACI requirement.) Also, the bend radii
at the corners of the stirrups require a minimum beam width for each size of
stirrup, as shown in Table 4.1.
Table 4.1 Recommended Stirrup size Minimum beam width (bw) (in.)
minimum beam width to
#3 10
accommodate different
stirrup sizes #4 12
#5 14
4.5 Stirrup Design Procedure 251
bw 48 in.
Six legs
Av 6Ast
Sometimes, to avoid very small (less than 3 in.) required spacing, the designer
may employ four, six, or more legs for stirrups. This increases Av in Equation (4.20),
and consequently the calculated spacing, s. The use of multiple legs is also
recommended in wide beams, as shown in Figure 4.19.
The steps for designing stirrups are summarized in Figure 4.20 and given below:
Step 1. Determine the distribution of shear along the beam to calculate the stirrup
spacing. (You can do this by drawing the shear force diagram.) Use either
the beam clear span or the center-to-center span. If you use the center-to-
center span, include half the support width when locating the critical
section.
Step 2. Determine the shear at the critical section Vu . As mentioned in Section 4.4,
the critical section is at a distance d from the face of the support. Vu is the
largest shear acting on the beam that needs to be considered.
Step 3. Calculate the shear capacity of concrete (Vc):
252 4 Shear in Reinforced Concrete Beams
1. Draw V diagram.
u
3.
Vc 2fc bw d
4.
Vu No Section size is
Vs Vc 4Vc? inadequate:
revise.
0.75
Yes
5.
Vc No
Vu ? No stirrups are needed.
2
Yes
6. Determine the location of Vc 2
on the Vu diagram.
7. Av fyt Av fyt
s1 min ,
0.75 fc bw 50bw
8.
Place stirrups at No ^
Vs V s@smax?
smax up to the location of Vc 2
Yes
p
Vc 2 fc0 bw d
Av fyt d
^s
Step 8. Calculate V ^s , only minimum stirrups (at smax spacing)
. If Vs V
smax
are required. Place them up to the point of Vc/2. If Vs > V ^s , go to step 9.
Step 9. Calculate the stirrup spacing:
Av fyt d
s
Vs
wu 5 kip/ft
24 in.
30'-0" 12 in.
Beam section
Vu 62.9 kip
75 kip Vc
/2 10.6 kip
(Vu )
2.42 ft
12.9 ft
Solution
Step 1. Draw the Vu diagram:
wu 530
R 75 kip
2 2
The critical section is at the distance d from the column face. Thus, from
the column centerline this distance is:
15
x 21:5 29 in: 2:42 ft
2
Because the factored shear decreases from the support to the midspan
(Figure 4.21) at a rate of 5kip/ft
(the slope of the shear is the load), the
shear at the critical section Vu is:
Step 4.
0:75
Vu 62:9
Vs Vc 28:3 55:6 kip
0:75
4Vc 428:3 113:2 kip > 55:6 kip ok
Vc 0:7528:3
10:6 kip < Vu 62:9 kip
2 2
Stirrups are needed!
Step 6. Determine the location of Vc/2 on the Vu diagram.
Write the equation for the shear force diagram:
Vu 75 5x
Vc
Vu
2
75 5x 10:6
x 12:9 ft
Step 8. Determine whether the beam needs more than minimum stirrups. With #3
stirrups at smax 10.5 in.
Av fyt d 0:226021:5
^s
V 27:0 kip < Vs 55:6 kip
smax 10:5
Av fyt d 0:226021:5
s 5:1 in:
Vs 55:6
Av fyt d 0:226021:5
Vs 35:5 kip
s 8
Vu Vc Vs 0:7528:3 35:5 47:9 kip
The s 8 in. starts (or s 5 in. ends) where Vu 47.9 kip. Determine the location of
this point on the shear diagram:
75 5x 47:9
x 5:42 ft
5.42(12) 7.5 2
N = = 11.1
5
Av fyt d 0:226021:5
Vs 27 kip
s 10:5
Vu Vc Vs 0:7528:3 27 41:5 kip
75 5x 41:5
x 6:7 ft
This is the end point of s 8 in. The number of 8 in. spacings (N ) is:
6:712 7:5 2 12 5
N 1:4 2 @ 8 in:
8
Finally, we will determine the portion of the beam in which a stirrup spacing of
10.5 in. can be used. Because 10.5 in. is the maximum spacing and no stirrup is
needed after that, the end point is x 12.9 ft (found in step 6 for Vu Vc /2). The
number of 10.5 in. spacings is:
12:912 7:5 2 12 5 2 8
N 6:6 7 @ 10:5 in:
10:5
Therefore, the stirrup spacing from each end of the beam is 1 @ 2 in. + 12 @ 5 in.
+ 2 @ 8 in. + 7 @ 10.5 in. Figure 4.22 shows the resulting stirrup layout.
15'-0"
Example 4.4 Design stirrups for the oor beam shown in Figure 4.23. The
loads include the beams self-weight. The columns are 15 in. 15 in. Use
fc0 4,000 psi, and fyt 60,000 psi. Assume that the minimum cover is 1.5 in.
Concrete is normal weight.
a Pu 75 kip Pu 75 kip
wu 1.0 kip/ft
A
A
5'-0" 5'-0" 5'-0"
15'-0"
Elevation
27 in.
2 #10
15 in.
Section A-A
Figure 4.23 Elevation and section of beam in Example 4.4. (a) Elevation. (b) Section A-A
Solution
Step 1. Draw the Vu diagram (see Figure 4.24):
n 15 15 12 13:75 ft
75 2 1 15
R 82:5 kip
2
75 2 1 13:75
R 81:88 kip
2
Step 2. Calculate Vu :
d h 1:5 3 8 1:27=2
27 2:5 24:5 in:
4.5 Stirrup Design Procedure 259
75 kip 75 kip
1.0 kip/ft
Critical section
82.5 Vu 79.8 kip
77.5 kip
Vc / 2 17.4 kip
2.5
(Vu )
x 2.67 ft
x
The location of the critical section from the column center is:
The shear at the critical section using the clear span is:
As expected, the shear at the critical section, Vu , is the same when using
the center-to-center span or clear span.
Step 3. Calculate the concrete resisting shear, Vc:
1.0 (normal weight concrete)
p p
Vc 2 fc0 bw d 21:0 4,0001524:5=1,000 46:5 kip
Step 4. Calculate the shear to be resisted by the stirrups (determine whether the
beam size is adequate):
Vu 79:8
Vs Vc 46:5 59:9 kip
0:75
4Vc 446:5 186 kip > 59:9 kip ok
Vc 0:7546:5
17:4 kip < Vu 79:8 kip
2 2
Stirrups are required:
Step 6. Locate Vc/2 (17.4 kip) on the Vu diagram. This point lies on the vertical
part of the shear force diagram (Figure 4.24). Determine the stirrup spacing
for the rst 5 ft of beam from each end; the 5 ft at the center do not require
any stirrups for shear.
Step 7. Calculate the maximum stirrup spacing:
Use #3 stirrups with two legs (bw 15 in., see Figure 4.19),
Av 2 0.11 0.22 in.2
( )
Av fyt Av fyt
s1 min p ,
0:75 fc0 bw 50bw
0:2260,000 0:2260,000
s1 min p ,
0:75 4,00015 5015
s1 minf18:6; 17:6g 17:6 in:
Av fy d
^s
V
smax
0:22 60 24:5
12
27:0 kip < Vs 59:9 kip
Av fyt d 0:226024:5
s 5:4 in: s 5 in:
Vs 59:9
Find the stirrup spacing at the end of this zone, where Vu 77.5 kip:
Vu 77:5
Vs Vc 46:5 56:8 kip
0:75
Av fyt d 0:226024:5
s 5:7 in:
Vs 56:8
Because Vs 59:9 kip < 2Vc 246:5 93 kip the maximum stirrup
spacing is the same value found in step 7.
512 7:5 2
N 10:1 Use 11@ 5 in:
5
2 in.
11@5 in.
7'-6"
where
Vu d
1:0
Mu
The presence of signicant axial compression (in addition to exure and shear)
increases the shear strength of a section. This is because the compressive loads can
prevent cracks from developing. The ACI Code provides the following equation
(ACI Code, Section 22.5.6.1) to account for the contribution of axial compression:
Nu p
Vc 2 1 fc0 bw d 4:23
2,000Ag
where Nu is the axial compression calculated from factored loads and Ag is the gross
cross-sectional area of the concrete section. In the formula Nu/Ag must be expressed
in psi. Nu is positive for compression. A more detailed calculation of Vc considering
Nu is given in the ACI Code Table 22.5.6.1.
The presence of signicant axial tension (in addition to exure and shear) decreases
the shear strength of the section. The ACI Code mandates the use of the following
4.6 Additional Formulas to Calculate the Shear Strength of a Beam Section 263
equation (ACI Code, Equation 22.5.7.1) to account for the presence of axial
tension:
Nu p
Vc 2 1 fc0 bw d 4:24
500Ag
where Nu is negative for tension. Nu/Ag must be expressed in psi. Vc cannot be less
than zero.
Example 4.5 Calculate the nominal shear capacity, Vc, of the section shown below
for the following cases (normal weight concrete with fc0 3,000 psi):
12 in.
d 22 in.
25 in.
3 #9
Solution
1.0 (normal weight concrete)
(a) The shear capacity of the concrete section according to Equation (4.12) is:
p p
Vc 2 fc0 bw d 21:0 3,0001222=1,000 Vc 28:9 kip
As 3:0
w 0:0114
bw d 1222
Vu d 2022
0:37 < 1:0 ok
Mu 10012
p
Vc 1:91:0 3,000 2,5000:01140:37 1222=1,000
p
3:51:0 3,0001222=1,000
Vc 30:3 kip 50:6 kip
Vc 30:3 kip
(c) Using Equation (4.23):
Nu p
Vc 2 1 fc0 bw d
2,000Ag
1001,000 p
Vc 2 1 1:0 3,0001222=1,000
2,0001225
Vc 33:7 kip
The ACI Code has special provisions for brackets and corbels. Figure 4.26 shows a
typical corbel. These are special elements on the side of a column or at the end of a
wall. In Figure 4.26, Vu is the factored vertical load from some building element,
which may be a precast or prestressed building girder, or a crane girder. Nuc is the
factored tension force on the corbel acting simultaneously with Vu. This horizontal
tension force results from any restraint against free relative horizontal movement
between the bracket and the supported element. Most often Nuc comes from
frictional restraint that occurs in the presence of volumetric changes in
the supported girder. The use of special bearing pads helps to minimize the
magnitude of Nuc.
4.7 Corbels and Brackets 265
Vu
av
Nuc
T
0.5 d
Shear
plane h
d
Imaginary
compression
strut
The ACI Codes design methodology (Section 16.5) is based on the satisfying
of the equilibrium of four forces (assumed to be concurrent). The method is
applicable when the following conditions are satised: (1) av /d 1.0, (2) Nuc Vu,
and (3) the depth of the bracket at the front is not less than d/2. (ACI Code Sections
16.5.1.116.5.2.2).
From Figure 4.26 it is clear that the shear plane at the level of the primary tension
reinforcement is subject to a moment:
Mu Vu av N uc h d 4:25
It is also subject to the tensile force, Nuc, and the shear force, Vu.
Figure 4.27 shows the typical reinforcement of a corbel. The required amount of
primary reinforcement, Asc, is determined from two parts. The rst part, Af, resists
the moment in Equation (4.25). Its design follows the procedure of the exural
design of rectangular sections. The second part, An, resists the tensile force, Nuc.
Hence:
Asc Af An 4:26
where
N uc
An , 0:90 4:27
fy
266 4 Shear in Reinforced Concrete Beams
Anchor bar
2d/3
d
Avf
Shear-friction
Shear-friction Avf reinforcement
reinforcement
perpendicular
at an angle
to the crack
The design of Ah, total area of closed stirrups parallel to the primary tension
reinforcement, is based on the shear-friction concept. If a crack forms at a shear
plane, reinforcing is needed to prevent a relative displacement (slippage) between
the surfaces. This type of reinforcement is shown in Figure 4.28. The reinforcement
ties together the two halves and ensures that the friction resistance parallel to the
crack is maintained.
When the shear-friction reinforcing is perpendicular to the shear plane, as is the
case for the corbel shown in Figure 4.27, the shear strength can be calculated as
(ACI Code Equation 22.9.4.2):
4.7 Corbels and Brackets 267
Vn Avf fy , 0:75 4:28
where is the coefcient of friction dened by the ACI Code for different types of
concrete and different pouring sequence scenarios. Avf is the area of reinforcement
crossing the shear plane to resist shear. (Refer to Section 22.9.4.2 of the ACI Code.)
For a corbel cast monolithically with the column (always the case), 1.4 for
normal-weight concrete.
The ACI Code imposes the following limitations to ensure that the corbel will
act in concurrence with the proposed design model:
1. The depth of the corbel at the outside edge of the bearing area shall be not less
than d/2. (ACI Code Section 16.5.2.2.)
2. The corbel must be deep enough so that Vn for normal-weight concrete may not
exceed the smaller of 0.2 fc0 bwd, 480 0:08fc0 bw d, and 1600 bwd (ACI Code
Section 16.5.2.4).
3. The corbels minimum primary reinforcement, Asc,min, must be the greater of the
following (ACI Code Section 16.5.5.1):
Af An , 2Avf =3 An , and 0:04 fc0 =fy bw d
for corbels cast monolithically with normal-weight concrete. The minimum total
area of closed stirrups parallel to the tension reinforcement, Ah,min, is equal to (ACI
Code Equation 16.5.5.2):
A few words must also be said about the anchor bar. The tensile reinforcement
(Asc) must develop its strength between the outer edge of the corbel and the face of
the column. (For a discussion of development of tensile reinforcing, see Chapter 3.)
This length is not adequate in most cases, hence some device is needed to add
mechanical anchorage. One such device is a largediameter bar (#9 or larger) to
which the reinforcing bars representing Asc are welded (see Figure 4.27). Another
way to provide mechanical anchorage is to weld the bars to an edge angle, as shown
in Figure 4.29.
268 4 Shear in Reinforced Concrete Beams
Edge angle
Example 4.6 Design the required reinforcement for the corbel shown in Figure 4.30.
Assume the following data: Vu 62 kip, Nuc 6 kip, av 8 in., b 16 in.,
fc0 5,000 psi, and fy 60,000 psi.
Solution
Step 1. Assume 1.5 in. cover and #6 bars for the primary reinforcement.
0:75
d 18 1:5 16:12 in:
2
From Equation (4.25):
0:0023
Af required bd 0:0023 16 16:12 0:59 in:2
4.7 Corbels and Brackets 269
8 in.
62 kip
6 kip
10 in.
18 in.
N uc 6
An required 0:11 in:2
fy 0:9 60
Hence
Vu 62
Avf 0:98 in:2
required 1:05fy 1:05 60
Using #4 stirrups ! two legs provide 2 0.2 0.4 in.2 Thus, the
required number of horizontal stirrups is
Because (2/3)d (2/3) 16.12 10.7 in., place stirrups at 3.5 in.
center-to-center to have the three stirrups within the ACI Code-required
distance 2/3d. Then use additional stirrups at the same spacing to the
bottom of the corbel.
Step 3. Check for Asc,min:
#9 Anchor bar.
Weld 2 #6 bars to
2 #6 the anchor bar.
4 Spaces
@ 3.5 in. c/c
Asc, min 2=3 Avf An 2=3 0:98 0:11 0:76 in:2 < 2 # 6 bars
0:88 in:2 ok
Asc, min 0:04 fc0 =fy bw d 0:045,000=60,0001616:12
0:86 in:2 < 0:88 in:2 o:k:
Ah, min 0:5Asc An
Ah, min 0:50:7 0:11
0:3 in:2 < 4#4 stirrups 4 0:4 1:6 in:2 o:k:
Problems
In the following problems, assume concrete is normal-weight unless noted otherwise.
4.1. A rectangular reinforced concrete beam has been designed for moment only,
without any stirrups for shear. It is, however, subjected to a shear at the critical
section, Vu 10 kip. The beam width, b, is 12 in., and the effective depth, d, is
26 in. Use fc0 4,000 psi. Determine whether this beam is adequate.
4.2. A beam is subjected to a uniformly distributed load and has a maximum shear of
60 kip at the face of its supports. The beam clear span is 30 ft0 in., b 12 in.,
d 24 in., fc0 4,000 psi, and fyt 60,000 psi. What is the shear at the critical
section? What is the required spacing for #3 stirrups at the critical section?
4.3. Design stirrups for the beam shown below. The dead load includes the beams
self-weight. Use fc0 4,000 psi, fyt 60,000 psi, and 1 in. cover.
16 in.
32 in.
3 #11
A-A
272 4 Shear in Reinforced Concrete Beams
4.4. Rework Problem 4.3 for a beam subjected only to the concentrated load on the
left in addition to the distributed loads.
4.5. The shear force at the critical section, Vu , of a reinforced concrete beam is
60 kip. If the beam has bw 14 in., fc0 3,000 psi, and fyt 60,000 psi,
what is the required effective depth, d, so that the minimum spacing of #3
stirrups is 9 in.?
4.6. Design stirrups for the beam shown below. The dead load is 0.70 kip/ft (beam
weight not included), and the live load is 1.5 kip/ft. Use fc0 4,000 psi,
fyt 60,000 psi, and 1 in. cover. The unit weight of the concrete is 150 lb/ft3.
A
12 in.
2 #8
A
24 in.
5'-0" 12". 20'-0" 12". 5'-0"
3 #8
A-A
4.7. A 6 in. thick one-way reinforced concrete slab has #6 @ 8 in. main reinforce-
ment. The cover is in. and fc0 3,000 psi. The unit weight of the concrete
is 150 pcf. Answer the following questions:
(a) What is the maximum shear (Vu) the slab can carry?
(b) What is the maximum live load the slab can support based on shear
requirements? Assume that the slab is simply-supported and has a clear
span of 10 ft0 in.
4.8. Design stirrups for the interior beam (B-1) shown below. The mechanical/
electrical systems weigh 5 psf, the partitions are 20 psf, and the ceiling,
carpeting, and so on, weigh 5 psf. The oor live load is 80 psf. Consider live
load reduction, if applicable. Use fc0 4,000 psi, fyt 60,000 psi, and 1 in.
for cover. Use ACI Code coefcients to determine the beam shear force. The
unit weight of the concrete is 150 pcf.
Problems 273
30'-0" 30'-0"
10'-0" A
(typical)
(B-1)
Columns
12 in. 12 in.
(typical)
Plan
6 in.
18 in.
2 #11
(typical)
12 in.
A-A
4.9 Design stirrups for the beam shown below. The applied loads do not include
the beams self-weight. Use fc0 4,000 psi, fyt 60,000 psi, and 1 in. for
cover. The concrete is all-lightweight with a unit weight of 110 pcf.
A
27 in.
10'-0" 10'-0"
3 #8
A-A
274 4 Shear in Reinforced Concrete Beams
4.10 Rework Problem 4.3, considering the effects of moment on the shear strength
of concrete. Use the moment and the shear at the critical section for purposes
of simplication.
4.11 Rework Problem 4.3 for a beam subjected to an axial compressive live load of
150 kip. Compare the results with Problem 4.3.
4.12 Rework Problem 4.3 for a beam subjected to an axial tensile live load of
50 kip. Compare the results with Problem 4.3.
4.13 Design the required reinforcement for the corbel shown below. Use
fc0 4,000 psi, fy fyt 60,000 psi, b 18 in., and 1 in. for cover.
Vu = 50 kip
9 in.
16 in.
Self-Experiments
In these self-experiments you learn about shear in beams. You will use both
Styrofoam and reinforced concrete models. Remember to include in your report
all the details of your tests (sizes, time of day you poured, amount of water/cement/
aggregate, problems that you encountered, etc.) together with images showing the
steps of the tests.
Self-Experiments 275
Experiment 1
In this experiment we use Styrofoam models to learn about the vertical and
horizontal components of shear in beams.
Test 1: Horizontal Shear
Stack several layers of Styrofoam, one on top of the other, and place them on two
supports, as shown in Figure SE 4.1a. Apply a load, P, at the center of the beam.
Measure how much the beam deects at the center under the load. Now glue the
layers together and repeat the test. Compare the measured deection for the two
cases, and discuss your observation.
P
Plastic wire
cures, apply a load at the center of the beam. Increase the load until you notice the
concrete cracks. Record your observations and any problems encountered.
A
Main bars
A-A
Figure SE 4.2 Reinforced concrete beam with main and shear reinforcements
Chapter 5
Columns
5.1 Introduction
Figure 5.2 shows the various classications of reinforced concrete columns. Columns
can be classied by the type of their reinforcement (main and lateral), by their shape,
by the type of loads that they will resist, by the type of structural system of which they
are part, and by their length. We now will study each class of columns.
Three main types of columns fall in this category: tied columns, spiral columns, and
composite columns.
b
b
a b
h
3
h b
h
3
h b
Because columns are subjected mainly to axial loads, they are reinforced with
longitudinal bars along their length. Because these bars are very slender, they need
to be laterally supported to keep them in place during concrete placement, and they
need lateral support when subjected to loads. Small-diameter (#3 or #4) bars,
referred to as ties, are used to fulll these requirements. Columns that use ties for
lateral reinforcement are called tied columns. Figure 5.3a shows a square tied
column. The ties are wired to the longitudinal bars to make a cage, which then is
placed into the form and properly positioned before casting the concrete. The cage
of bars and ties keep the longitudinal bars straight and the ties provide resistance
against buckling. Ties generally follow the perimeter of the columns cross section
(rectangular in rectangular columns and circular in circular columns).
Tied columns are the most common because their construction costs are lower
than those for spiral and composite columns. In fact, over 95 % of all columns in
concrete buildings located outside earthquake-prone regions are tied columns. The
area inside the ties is called the core, and the area outside them is the shell of the
column (see Figure 5.3a).
Spirals are used in spirally reinforced columns to provide lateral support to the
main reinforcements. Spirals are helical-shape wires, which are placed around
the main reinforcements as shown in Figure 5.3b. Because most spiral columns
5.2 Types of Columns 279
Spiral columns
Composite columns
Square columns
Circular columns
Based on shape
Rectangular columns
Braced columns
Based on structural system
Unbraced columns
Short columns
Based on length
Slender (long) columns
are circular in shape, a spirally reinforced core sometimes may be placed inside a
square cross section. Spiral reinforcing is a more expensive construction (about
twice as much) than using ties. Spiral columns provide larger capacity than do tied
columns, but their main advantage is their ductility and toughness when large
overloads, such as loads occur in earthquakes, are expected. Similar to tied
columns, the area conned by the spirals is the core, and the area outside them is
the shell (see Figure 5.3b).
280 5 Columns
a Core Tie
Shell
A A
A-A
Tied column
b
Spiral
Core
Shell
B B
B-B
Spiral column
c
I-shape
Pipe
Composite column
Square and rectangular shapes are the most common, as they are the simplest to
form and construct. Circular columns may be formed by using cardboard or
plastic tubes, or by using hinged steel forms, which can be removed easily. Other
column shapes besides circular and rectangular are also used. Figure 5.4 shows a
few of them. Figure B5.1 in Appendix B shows the tapered reinforced concrete
columns used for a train station.
282 5 Columns
Columns primarily carry loads in compression. But they can also be subjected to
moments, depending on the buildings geometry and loading. Therefore, columns
are grouped into two classes: axially loaded columns, and eccentrically loaded
columns. (Sometimes these are referred to as columns with small eccentricity, and
columns with large eccentricity, respectively).
A concentric axial load and a moment can be combined into an eccentric load.
The term eccentricity refers to the distance between the point of load application
and the center of the section. To better understand the consequence of an eccen-
tricity on the behavior of columns, consider Figure 5.5a, which shows a column
subjected to a compression force, P, acting at the center of the section (point 1).
Because the force acts at the center of the section, the internal compression stresses
are distributed uniformly on the section.
If we move P to a new location (point 2) at a distance e from the column center,
as shown in Figure 5.5b, the load generates bending stresses in addition to axial
compressive stresses. The bending stress is the result of the moment caused by the
off-center load (M Pe). The action of P at the eccentricity, e, is equivalent to the
load P acting at the center and an additional moment, M Pe.
If the moment acting on the column is negligible compared to its axial load, we
consider the column to be an axially loaded column, or a column with small
eccentricity. If the applied moment is large, the column is an eccentrically loaded
column, or a column with large eccentricity. In former ACI Codes, tied and spiral
columns were considered to be axially loaded columns when the eccentricity was
less than 0.1 h and 0.05 h, respectively (h the cross-sectional dimension in the
direction of the eccentricity).
Now that we know that the effects of eccentric loads are the same as adding
moments on columns and vice versa, let us review the sources of moments or
eccentricities. Figure 5.6a shows a reinforced concrete building frame under gravity
loads (Refer to Figures 2.10 and 2.11). The column on line B is subjected to
moments from the adjoining beams. If the beam spans and loads are equal, the
applied moments have the same magnitude but opposite directions, thus canceling
each other. As a result, the column on line B is subjected only to an axial load. Even
though this is theoretically correct, in reality there is always some moment on the
column because the loads on the neighboring beams are never the same and the
column is not perfectly straight.
The columns on lines A and C, on the other hand, are subjected to moments from
the beams on one side in addition to axial loads. Therefore, these columns are
subjected to large moments or have large eccentricities. Also, the column between
the two bays (column B) will be subjected to moment in addition to axial loads if
the live load is larger on one span than on its adjacent span.
Another example of a column with large eccentricity is shown in Figure 5.6b, in
which the column is part of a precast concrete structure. The beams and columns are
cast off the construction site, and then transported to the site for assembly.
5.2 Types of Columns 283
a P P
1 1
A
View A
b P P
eP e
M Pe
2 2
A
View A
Figure 5.5 (a) Axially loaded column (column with small or no eccentricity); (b) eccentrically
loaded column
In precast construction, beams are often placed on brackets and connected together
through steel plates embedded in both the beam and the bracket. As a result, there is
always an eccentricity between where the beam is supported and the column
centerline. This eccentricity generates a moment on the column, which needs to
be considered in the analysis and design of the column.
Lateral loads, such as high winds and earthquakes, can generate large moments
on the columns of monolithic concrete structures. Such columns usually have large
eccentricity. Figure 5.6c shows how the columns of a two-story building undergo
284 5 Columns
a
A B C
b c
e
P
Figure 5.6 (a) and (b) Gravity loading, (c) lateral loading
large bending moments when subjected to lateral loads. Columns subjected to loads
with large eccentricity will be studied in greater detail later in this chapter.
Column and beam assemblies can be divided into two categories, depending on the
building structural framing systems used: braced frames and unbraced frames. The
columns within such systems are called braced columns and unbraced columns,
respectively.
In a braced frame, lateral loads are resisted by shear walls, elevator or stairwell
shafts, diagonal braces, or a combination thereof. The large stiffness of these
elements prevents the columns of such a frame from undergoing large lateral
motion or sidesway, and from experiencing signicant moments due to lateral
loads. In an unbraced frame, on the other hand, the columns (unbraced columns)
are subject to large bending moments due to the lateral loads and have to withstand
large lateral motions. These columns generally have large eccentricities. Figure 5.7
shows braced and unbraced columns in two different structural framing systems.
5.3 Behavior of Short Columns with Small Eccentricity Under Load 285
Unbraced columns
Columns may be divided into two groups based on their length, or more accurately
their slenderness ratio. Slenderness ratio (k/r) is the ratio of the columns effective
length (k) to the least radius of gyration (r) of the section. Short columns are
columns whose slenderness ratio is low enough that their failure occurs from
excessive stress levels rather than by buckling. Slender columns, on the other
hand, may buckle when subjected to large axial loads.
Most reinforced concrete columns in normal building structures are short col-
umns. In fact, the results of a study conducted by the ACI (Notes on ACI 318-71,
Building Code with Design Applications, p. 10-2) indicate that 90 % of braced
columns and about 40 % of unbraced columns could be considered to be short
columns. As a result, the emphasis in this book will be on short columns.
Figure 5.8 shows the failure mechanisms of an axially loaded square tied column
and round spiral column. When a short tied column is subjected to increasing axial
loads, the column fails suddenly. First the longitudinal reinforcing reaches yield,
286 5 Columns
a P
P
b P P
P P
Figure 5.8 Short columns under axial loads: (a) tied columns; (b) spiral columns
and then the concrete fails when the ultimate strain is reached. The failure is usually
accompanied by plastic buckling of the longitudinal bars. Figure 5.9 shows a typical
load-deformation relationship for tied and spiral columns. The tied column reaches
the maximum capacity at point A, and fails soon thereafter at point B.
A spiral column, on the other hand, does not fail suddenly because the closely
spaced spirals keep the core conned while the column shell spalls (Figure 5.8b).
This connement does increase the columns deformability signicantly. The outer
5.3 Behavior of Short Columns with Small Eccentricity Under Load 287
P
A Spiral column C
D
B
Tied column
shell is not conned, thus it falls away readily. The inner core, however, is still able
to carry loads, even after the concrete has been crushed by large compressive
stresses. The column behaves like a bag of our: As long as the paper sack does
not burst, the our column will support loads. For this reason, in this type of
column, the ACI Code requires a minimum spiral reinforcement that will prevent
the column from bursting until well after the concrete has reached its assumed
ultimate strain of 0.003. Thus, a typical spiral column will have a second maximum
point in its load-deformation diagram (point C in Figure 5.9). The yielding of the
spirals makes the column failure ductile, which makes a spiral column ideal for
unexpected large overloads such as seismic loads.
288 5 Columns
The ACI Code has several requirements for the design of columns:
1. Limits on the amount of longitudinal reinforcements. Column steel ratio, g, is
dened as the ratio of the area of the longitudinal reinforcement (Ast) to the gross
area of the column (Ag):
Ast
g 5:1
Ag
The ACI 318-14 no longer uses the column steel ratio notation (g); however, the
authors have kept it for purposes of clarity.
The ACI Code (Section 10.6.1.1) limits the area of the longitudinal reinforce-
ment, Ast, in columns between 0.01 Ag to 0.08 Ag. This means that the steel ratio,
g, can average as shown in Equation (5.2):
The minimum steel ratio of 0.01 provides resistance to bending, which may exist
whether or not calculations show the column is subjected to bending moments.
In addition, a minimum amount of steel reduces creep and shrinkage of the
concrete under sustained compression loads. It is common practice to use a
minimum bar size of #5 for the longitudinal reinforcement.
Although the maximum steel ratio is 0.08, in practice it is very difcult to use
such a high amount of steel in the column, especially where the bars are spliced
above a oor level. Such congestion may be avoided by using #14 or #18 bars.
(A #14 bar has a cross sectional area of 2.25 in.2, a #18 bar has one of 4.0 in2.)
These bars are not used in beams, but are very useful in columns.
Bundled bars may be used if the column is subjected to a large load and a large
number of bars is needed. Bundles consist of three or four bars (a maximum of
four bars according to the ACI Code, Section 25.6.1.1) tied together in direct
contact, and are usually placed at the corners of the column, as seen in
Figure 5.10. Each bundle of bars is treated as if it were a single round bar of
area equal to the sum of the areas of the bundled bars. The main drawback of
bundled bars is that they cannot be lap-spliced.
Figure 5.11 Minimum number of bars: (a) tied columns, (b) spiral columns
2. Limit on the number of bars. According to the ACI Code (Section 10.7.3.1), the
minimum number of main longitudinal bars is four for rectangular or circular tied
columns, three for triangular tied columns, and six for spiral columns (see
Figure 5.11).
3. Limit on the clear cover. According to the ACI Code (Section 20.6.1.3.1), the
minimum clear cover for columns is 1.5 in., measured from the edge of the
column to the transverse reinforcement. This cover is for interior columns that
are not exposed to weather or in contact with the ground. The clear cover is 2 in.
for formed surfaces exposed to weather or in contact with the ground. If the
concrete is cast directly against the earth without forming (as in drilled piles), the
cover must be increased to 3 in.
4. Limit on tie spacing. In general, there are four main reasons for having ties in a
column:
A. They hold the longitudinal reinforcement in place during construction.
B. They provide a conned core and, as a result, increase the columns strength
and ductility.
C. They act as shear reinforcement.
D. They provide lateral support for the longitudinal bars and prevent them from
elastic buckling. Columns need sufciently large tie sizes that are well
connected to the longitudinal bars at sufciently close vertical spacings (s).
To satisfy the above requirements, the ACI Code (Sections 10.7.6.1.2 and
25.7.2.2) requires that at least #3 ties be used for #10 or smaller longitudinal
bars, and at least #4 ties be used for #11, #14, and #18 and bundled bars.
290 5 Columns
In addition, the ACI Code (Sections 10.7.6.1.2 and 25.7.2.1) requires that the
vertical center-to-center spacing of ties (s) be limited to:
where db is the diameter of the longitudinal bars, dt is the diameter of the ties,
and bmin is the minimum dimension of the column. The clear spacing of ties has
to be at least equal to 4/3 of the maximum aggregate size (dagg). According to the
ACI Code (Sections 10.7.6.1.2 and 25.7.2.3), a bar is adequately supported
laterally if it is located at a corner of a tie, with an enclosed angle not exceeding
135 , or if it is located between laterally supported bars with a clear spacing of
6 in. or less. Figure 5.12 shows typical tie arrangements that satisfy this
a b c
d e f
g h
6 in.
6 in.
10 bars 12 bars
requirement. Figure B5.2 in Appendix B shows the main reinforcements and ties
for arrangement e before concrete placement.
5. Limit on longitudinal bar spacing. Sections 10.7.2.1 and 25.2.3 of the ACI Code
requires that the clear distance between longitudinal bars be at least equal to the
greatest of 1.5 times the bar diameter (db), 1.5 in. or 4/3 of maximum aggregate
size (dagg), to allow concrete to ow between the reinforcements. Reinforced
concrete columns in multistory buildings are generally cast one level at a time.
Therefore, the longitudinal reinforcements in columns typically are spliced
above every oor. An exception is columns in seismic zones where splicing is
usually near midheight between oors. There are different methods of splicing
bars in columns. Figure 5.13 shows one common method of splicing reinforce-
ments in a multistory building. Table A5.1 lists the maximum number of bars
that can be placed in a square or circular column based on minimum bar spacing
requirements and the splicing method shown in Figure 5.13. The spiral and tie
sizes are assumed to be #4 with 1.5 in. cover. Figure B5.3 in Appendix B shows
the reinforcements at the end of a concrete column during construction.
6. Limit on spacing and amount of spiral reinforcement. Spirals are often made of
smooth bars rather than deformed bars; and the spacing and amount of spirals need
to be such that they conne the column core. For these reasons, the ACI Code
(Sections 10.7.6.1.2 and 25.7.3.1) requires that clear spacing between spirals
(sclear) be at least equal to 4=3 dagg and between 1 and 3 in. (see Figure 5.14a). In
cast-in-place construction spirals must be at least 3=8 in. in diameter.
Spiral steel ratio is dened as:
If the diameter of the spiral steel is dsp and the area of the spiral steel is Asp, the
volume of column core in height s (see Figure 5.14b) is:
h2c
s
4
hs Asp
hs Asp 4Asp hs
s 2
h2 hc s
c s
4
Because hs hcdsp, and dsp is negligible compared to hc, we can assume that hs
and hc are approximately equal (hs hc). Substituting hc for hs into the above
292 5 Columns
A A
Section A-A
equation allows us to calculate the spiral steel ratio for columns, s, using the
simplied Equation (5.4).
4Asp
s 5:4
hc s
The ACI Code requires a minimum spiral steel ratio to ensure ductility and
toughness. According to Equation (25.7.3.3) in Section 25.7.3.3 of the ACI
Code, the minimum spiral steel ratio (s,min) is:
Ag fc0
s, min 0:45 1
Ach fyt 5:5
fyt 100 ksi
5.4 General ACI Code Requirements for Columns 293
b
hc h
hs
hc
Section A-A
h2
Ag
4
Ach is the area of the core (measured from outside-to-outside edge of spiral
reinforcement). See Figure 5.14:
h2c
Ach
4
and fyt is the specied yield strength of the transverse (spiral) reinforcement. The
required spiral pitch, s, can then be calculated from Equation (5.4) based on an
assumed spiral size, which must be at least 3/8 in. in diameter.
294 5 Columns
Most reinforced concrete columns are categorized as short columns. This means
that they will fail in compression under large loads rather than undergo elastic
buckling. In this section we will study the load carrying capacity and the design of
short columns with small eccentricity.
5.6 Analysis of Short Columns with Small Eccentricity 295
a Po b Po
Ag
Ast
0.85 fc
fy fy
Po
We rst nd the axial load strength of a column. Figure 5.15a shows a typical
column subject to a concentrated load at its center. From equilibrium of forces in
the vertical direction (along the column axis), the axial load capacity of a column
(Po) is equal to the sum of the volume of stresses in the concrete and the steel (see
Figure 5.15b). Based on the results of tests carried out at the University of Illinois
and Lehigh University from 1927 to 1933, the ACI Code uses 0.85fc0 for the
ultimate concrete compression stress. At ultimate load the stress in the steel is
equal to the yield stress ( fy). Therefore, according to ACI Equation 22.4.2.2:
The stress in the concrete (0.85fc0 ) is applied on the net column area (AgAst), which
is the area of the concrete. In reality, however, the loads acting on columns always
have an eccentricity (e.g., due to vertical misalignment of the form). The ACI Code
accounts for accidental eccentricity by requiring that the theoretical capacity be
reduced by 20 % for tied columns and 15 % for spirally reinforced columns. Then
the nominal load capacity of columns is:
h i
Pn 0:8Po 0:8 0:85fc0 Ag Ast fy Ast tied columns 5:7
h i
Pn 0:85Po 0:85 0:85fc0 Ag Ast fy Ast spiral columns 5:8
To nd the design resisting load, PR, we must reduce Pn by the strength reduction
factor, :
296 5 Columns
PR Pn 5:9
According to the ACI Code (Sections 10.5.1.2 and 21.2.2), columns with small
eccentricity have the following strength reduction factors:
The values of for pure compression are less than those used for beams in
bending (0.90). The main reasons that the strength reduction factors are consider-
ably lower are the following:
1. A column failure is a much more severe event than the local failure of a beam,
because a column supports larger areas of a building.
2. The quality of concrete used in columns is less reliable than that used in beams
and slabs. The difculty of consolidating the concrete in narrow column forms
and between the longitudinal and lateral reinforcements often leads to honey-
combs that are difcult to repair (even when visible).
3. The strength of the concrete has a much greater role in the ultimate strength of a
column than it does in beams and slabs, where the reinforcing has the most
inuence on the ultimate strength.
After introducing the factor, we calculate the strength of an axially loaded
column (ACI Code Sections 10.5.2.1, 22.4.2.1 and 22.4.2.2) as follows:
h i
PR Pn 0:8 0:85 fc0 Ag Ast fy Ast tied columns 5:11
h i
PR Pn 0:85 0:85 fc0 Ag Ast fy Ast spiral columns 5:12
The steps of the analysis of short columns with small eccentricity are shown in
Figure 5.16 and are as follows:
Step 1. Check the steel ratio. When we analyze a column, we know its dimensions
and the size and number of its reinforcements. Therefore, we have the gross
area of concrete (Ag) and the total area of steel (Ast), from which we can
determine the column steel ratio (g):
Ast
g
Ag
0:01 g 0:08
If the steel ratio does not fall within these limits, the column does not
conform to the current ACI Code requirements.
5.6 Analysis of Short Columns with Small Eccentricity 297
Analysis of Columns
(short with small eccentricity)
2.
Check the maximum number of bars (Table A5.1).
3. 3.
Calculate PR Pn where 0.65: Calculate PR Pn where 0.75:
PR 0.8 [0.85fc (Ag Ast) fy Ast ] PR 0.85 [0.85fc (Ag Ast) fy Ast ]
4. 4.
No No
PR Pu? Column is N.G. PR Pu?
Yes Yes
5. 5.
Check tie spacing: Check the spiral steel:
smax min {16db , 48dt , bmin} 4Asp
s
hcs
Ag fc
s,min 0.45 1
Ach fyt
s s,min?
6. 6.
Check the tie arrangement Check the clear vertical
using Figure 5.12. space of the spirals:
1 in. sclear 3 in.
Step 2. Determine whether the spacing between the longitudinal bars meets the
ACI Code requirements by obtaining the maximum number of bars that can
be placed in the column according to Table A5.1. The remaining steps
differ depending on whether the column is tied or spiral:
(a) Tied Columns
298 5 Columns
0:65
h i
PR Pn 0:8 0:85fc0 Ag Ast fy Ast
Step 4. Calculate the factored loads, Pu, and determine whether the column can
resist the applied loads (PR Pu).
As indicated in Chapter 2, Section 2.10, depending on the value of the total
oor load, PL, compared to the total roof live load, PLr, one of the load
combinations of Equation (2.3d) is used to compute Pu:
0:75
h i
PR Pn 0:85 0:85fc0 Ag Ast fy Ast
Step 4. Calculate the factored loads, Pu, and determine whether the column can
resist the applied loads (PR Pu).
Step 5. Check the spiral steel. Calculate the spiral steel ratio (s) and compare it
with the minimum amount required by the ACI Code:
4Asp
s
hc s
0
Ag f
s, min 0:45 1 c
Ach fyt
s s, min
Step 6. Check the clear space (sclear) between each turn of the spirals:
Example 5.1 Determine the maximum factored axial load that a short tied column
with the cross section shown below can resist. There is no moment on the column.
Determine whether the ties are appropriate. The compressive strength of the
concrete is 4,000 psi, and the reinforcement is A615 grade 60 steel.
5.6 Analysis of Short Columns with Small Eccentricity 299
PR ?
14 in.
A A
8 #8 14 in.
Solution
Step 1. Determine and check the steel ratio, g:
Step 2. Check the spacing of the longitudinal bars by obtaining the maximum
number of #8 bars that can be placed into the column from Table A5.1:
Therefore, the maximum design (factored) load for this column is 533 kip.
Step 4. We skip this step because we need only the load capacity.
Step 5. Check the adequacy of the ties. The maximum spacing of the ties (smax) is:
300 5 Columns
Cover Tie #8
# # #
3
14 21:5 2 31:0
Clear space 8
2
Clear space 3:6 in: < 6 in:
Because the clear space between the bars is less than 6 in., no additional
ties are necessary on the non-corner longitudinal reinforcing. Therefore, the
tie arrangement meets the ACI Code requirements.
Example 5.2 The circular spiral column shown below is subjected to a dead load
of 200 kip and a roof live load of 225 kip. The eccentricity of the loads is small.
The compressive strength of the concrete is 4,000 psi, and the reinforcement is
A615 grade 60 steel. Check the adequacy of the column including the spirals.
PD 200 kip
PLr 225 kip
6 #9
A A
16 in.
Section A-A
5.6 Analysis of Short Columns with Small Eccentricity 301
Solution
Step 1. Determine and check the steel ratio:
4Asp 40:11
s 0:0169
hc s 132:0
PR Pn Pu
The useful capacity of the column (PR) must be at least equal to the factored load
on the column (Pu). Thus,
5.7 Design of Short Columns with Small Eccentricity 303
h i
Pu 0:8 0:85fc0 Ag Ast fy Ast
Because the area of concrete (Ag) is known we can solve the above equation for the
area of steel (Ast):
h i
Ast 0:8 fy 0:85 fc0 Pu 0:8 0:85 fc0 Ag
0:65
Pu 0:8 0:85 fc0 Ag 5:13
Ast
0:8 fy 0:85fc0
Similarly, the required area of steel, Ast, for spiral columns is:
0:75
Pu 0:85 0:85fc0 Ag 5:14
Ast
0:85 fy 0:85fc0
Note that if the numerator in Equations (5.13) or (5.14) results in a negative value,
the column requires only 1 % longitudinal reinforcement (min). The design steps
are shown in Figure 5.17 and are as follows:
Step 1. Determine the factored axial load on the column, Pu (as indicated in the
previous section):
Which steps you perform next depends on whether the column is to be tied
or spirally reinforced:
Tied Columns
Step 2. Calculate the required area of steel using Equation (5.13). Use 0.65.
Step 3. Use Tables A2.9 and A5.1 to select bars. The minimum number of bars
for tied square columns is four. Determine whether 0.01 g 0.08. If
g < 0.01, use g 0.01. Also, if g > 0.08, or you cannot nd any arrange-
ments of bars to t inside the column, the column dimensions are not
enough and its cross-sectional area (Ag) must be increased.
Step 4. Design the ties. Use #3 ties for #10 and smaller longitudinal bars. Other-
wise, use #4 ties. The tie spacing, smax, is:
Round down smax to the nearest 0.5 in. Check the tie arrangement using
Figure 5.12.
304 5 Columns
Design of Columns
(short with small eccentricity)
Ag known, Ast unknown
1.
Calculate Pu
0.65 0.75
2. 2.
Calculate Ast : Calculate Ast :
P u 0.8 (0.85fcAg) P u 0.85 (0.85fcAg)
Ast Ast
0.8 (fy 0.85fc ) 0.85 (fy 0.85fc )
3. 3.
Use Tables A2.9 Use Tables A2.9
and A5.1 to select bars. and A5.1 to select bars.
Figure 5.17 Flowchart for the design of a short column with small eccentricity (Ag known,
Ast unknown)
Spiral Columns
Step 2. Calculate the required area of steel using Equation (5.14). Use 0.75.
Step 3. Use Tables A2.9 and A5.1 to select bars. The minimum number of bars for
spiral columns is six. Determine whether 0.01 g 0.08. Similar to
the tied columns, if g < 0.01, use g 0.01; and if g > 0.08 or bars do
not t inside the column, increase the column cross-sectional sizes.
the spiral steel ratio (s) to s,min (use
Step 4. Design the spiral steel by equating
a minimum spiral diameter of 3 8 in.):
5.7 Design of Short Columns with Small Eccentricity 305
Ag fc0
s, min 0:45 1
Ach fyt
4Asp
s
hc s
s s, min
Solve for s (spiral pitch) and round down to the nearest 1/4 in. Check
the clear pitch where sclear sdsp, which must be between 1 and 3 in.
If sclear is less than 1 in., increase the spiral size; if sclear is more than 3 in.,
use 3 in.
Example 5.3 Figure 5.18a shows the typical partial oor plan and sections of a
three-story reinforced concrete ofce building. The mechanical and electrical
systems for the oor and the roong and insulation weigh 5 psf. The weight of
the partitions is 15 psf. The oor live load is 50 psf and the roof snow load is 30 psf.
Assume fc0 4,000 psi and fy 60,000 psi Design the square 16 in. 16 in. tied
interior columns between the ground and second levels. Moments acting on the
columns are not signicant, and you should not consider live load reduction in load
calculations. Assume that the unit weight of the concrete is 150 pcf. Neglect the
self-weight of the columns.
Solution
Step 1. Determine the loads acting on the columns:
Floor Loads
(B-1)
(S-1)
A
10'-0"
(B-1)
B B
(S-1)
(B-2)
(B-2)
A 10'-0"
(B-1)
(S-1)
A A
10'-0"
(B-1)
Columns
16 in. 16 in. (typical)
30'-0"
Typical partial framing plan
6 in.
24 in.
16 in.
Section A-A
Roof
EL 39'-0"
3rd level
EL 27'-0"
2nd level
EL 15'-0"
Ground
EL 0.00
Section B-B
Roof Loads
6
Weight of concrete slab 150 12
75 psf
Roofing and insulation 5 psf
Roof dead load 80 psf
The tributary area of the columns at roof level is the same as that of a
oor (900 ft2). Therefore
Step 3. Using Table A2.9, select 8 #8 bars (As 6.32 in2). Based on Table A5.1, the
maximum number of #8 bars for a 16 in. 16 in. column is 12. Therefore,
8 #8 bars are ok. The provided column steel ratio, g, is:
Ast 6:32
g 0:0247
Ag 16 16
0:01 < g 0:0247 < 0:08 ok
16 in.
8 #8
16 in.
#3 @ 16 in.
Because we have to determine both the size of the column and the required area of
steel, and only one equation denes the column load capacity, we must assume
one unknown. According to the ACI Code, g can vary between 0.01 and 0.08.
If g 0.01, the column size may be excessively large. On the other hand, g 0.08
is not practical as the reinforcement will be very congested. Exceeding g 0.04 is
not recommended, so for this process we use g 0.03.
5.7 Design of Short Columns with Small Eccentricity 309
PR Pn Pu
For design, we consider PR Pu. The column steel ratio, g, is dened as:
Ast
g ! Ast g Ag
Ag
and simplifying:
h i
Pu 0:8Ag 0:85 fc0 1 g fy g
P
Ag h u i 0:65 5:15
0
0:85 0:85fc 1 g fy g
Similarly, for spiral columns, the required column area, Ag, is:
P
Ag h u i 0:75 5:16
0:85 0:85fc0 1 g fy g
Now that we have determined the column area, we can calculate the column
dimensions, h and b, as follows:
p
Ag h2 !h Ag square column
Ag h b rectangular column
r
h 2
Ag
Ag !h 2 round column
4
310 5 Columns
The steps for the design of short columns with small eccentricity are shown in
Figure 5.19 and are as follows:
Step 1. Determine the factored axial load on column, Pu:
P
Ag h u i 0:65
0:8 0:85fc 1 g fy g
0
Design of Columns
(short with small eccentricity)
Ag , Ast unknown
1.
Calculate Pu .
Assume g 0.03
0.65 0.75
2. 2.
Calculate Ag : Calculate Ag :
Pu Pu
Ag Ag
0.8 [0.85fc (1 g) fy g] 0.85 [0.85fc (1 g ) fy g]
3. 3.
Determine the column size, b Determine the column size, h.
and h. Round to the nearest inch. Round h to the nearest inch.
4.
4. Calculate Ast :
Calculate Ast :
Pu 0.85 (0.85fc Ag)
Pu 0.8 (0.85fc Ag) Ast
Ast 0.85 (fy 0.85fc )
0.8 (fy 0.85fc )
5.
Use Tables A2.9
5.
Use Tables A2.9 and A5.1 to select bars.
and A5.1 to select bars.
6.
Design spirals:
6. 4Asp
Design ties:
s hcs
smax min{16db , 48dt , bmin}
Check the tie arrangement Ag fc
using Figure 5.12. s,min 0.45 1
Ach fyt
Set s s,min and solve for
pitch(s). Round down to the
nearest 1 4 in.
1 in. sclear 3 in.
Figure 5.19 Flowchart for the design of a short column with small eccentricity
Step 5. Use Tables A2.9 and A5.1 to select the size and number of longitudinal
bars. Remember that the minimum number of bars for square tied columns
is four.
Step 6. Design the ties. Use #3 ties for #10 and smaller longitudinal bars.
Otherwise, use #4 ties. The tie spacing, smax, is:
312 5 Columns
Round down smax to the nearest 1/2 in. Check the tie arrangement using
Figure 5.12.
Spiral Columns
Step 2. Calculate the required gross area of the column, Ag:
P
Ag h u i 0:75
0:85 0:85 fc0 1 g f y g
Step 5. Use Tables A2.9 and A5.1 to select the size and number of longitudinal
bars. Remember that the minimum number of bars for spiral columns is six.
Step 6. Design the spiral steel by equating the spiral steel ratio, s, to s,min. Use a
minimum spiral diameter of 3/8 in.:
0
Ag f
s, min 0:45 1 c
Ach fyt
4Asp
s
hc s
s s, min
Solve for s (spiral pitch) and round down to the nearest 1/4 in. Check the
clear pitch, sclear sdsp, which must be between 1 and 3 in. If sclear is less
than 1 in., increase the spiral size; if sclear is more than 3 in., use 3 in.
Example 5.4 Design a short square tied column to carry an axial dead load of
300 kip and a oor live load of 175 kip and a roof live load of 80 kip. Assume
that the applied moments on the column are negligible. Use fc0 4;000 psi and
fy 60,000 psi.
5.7 Design of Short Columns with Small Eccentricity 313
Solution
Step 1. Compute the factored load, Pu.
1.83PLr 1.83(80) 146.4kip < PL 175kip, therefore from Equation
2.3d:
Pu 1:2PD 1:6PL 0:5PLr
Pu 1:2300 1:6175 0:580
Pu 680 kip
Assume g 0.03.
Step 2. The required area of the column, Ag, is:
P
Ag h u i
0:8 0:85 fc0 1 g fy g
680
Ag
0:800:650:85 41 0:03 600:03
Ag 257 in:2
Step 5. Select the size and number of bars. For a square column with bars uniformly
distributed along the edges, we keep the number of bars as multiples of
four. Using Table A2.9, 8 #9 bars (As 8 in2) are selected.
Step 6. Because the longitudinal bars are #9, select #3 bars for the ties. The
maximum spacing of the ties (smax) is:
314 5 Columns
16 in
16 in
#3 @ 16 in.
8 #9
Example 5.5 Solve Example 5.4 for a circular spiral column. fyt 60,000 psi.
Solution
Step 1. The factored load was determined in Example 5.4: Pu 680 kip. Assume
g 0.03.
Step 2. The required gross area of column, Ag, is:
P
Ag h u i
0:85 0:85fc0 1 g fy g
680
Ag
0:850:750:8541 0:03 600:03
Ag 209 in:2
5.7 Design of Short Columns with Small Eccentricity 315
h 16 in:
h2 3:14162
Ag 201 in:2
4 4
Step 4. The area of steel required, Ast, is:
Pu 0:85 0:85fc0 Ag
Ast
0:85 fy 0:85 fc0
680 0:85 0:750:85 4 201
Ast
0:85 0:7560 0:85 4
Ast 6:77 in:2
Step 5. Using Table A2.9, select 7 #9 bars. The provided area of steel is 7.00 in2.
4Asp 40:11
s
hc s 13s
In the above equation, the pitch of spiral, s, is the unknown. The
minimum required spiral steel ratio, s,min, is:
0
Ag f
s, min 0:45 1 c
Ach fyt
h2c 3:14132
Ach 132:7 in:2
4 4
316 5 Columns
s s , min
40:11
0:0154
13smax
40:11
smax 2:20 in:
130:0154
Use s 2:0 in:
In addition, the spiral clear pitch, sclear, should be between 1 and 3 in.:
sclear s d sp 2 3 8 1:625 in:
1 in: < 1:625 in: < 3 in: ok
Therefore, the spiral to be used for this column is 3/8 in. diameter at 2 in.
The following gure shows the nal design of the column.
7 #9
16 in.
There are two types of columns, based on the applied loads: axially loaded and
eccentrically loaded. In monolithic concrete construction, most columns are eccen-
trically loaded, which means that the applied load is not acting at the center of the
column. In other words, the column is subjected to a moment in addition to the
axial load. In the following, we explore the behavior of such columns in more
detail.
5.8 Behavior of Short Columns Under Eccentric Loads 317
a Pu b Pu Pu c Pu
e e
Mu Pue
Pu
Figure 5.20a shows a column subject to a load, Pu, at a distance e from the center
of the column. By adding two equal and opposite forces of magnitude Pu at the
center of the column, as shown in Figure 5.20b, we nullify the net effect because
these forces cancel each other out.
The two equal and opposite forces at a distance e form a couple or moment with
magnitude Mu Pue, and a concurrent axial load of Pu applied concentrically as
shown in Figure 5.20c. Hence, we conclude that a column subjected to a load Pu at a
distance e from its center is equivalent to a concentric load, Pu, and a moment,
Mu Pu e. Similarly, a concentric load, Pu, and a moment, Mu, may be represented by
an eccentric load, Pu, at an eccentricity e equal to Mu /Pu from the centroid.
The concentric load, Pu, creates a uniform compression stress while the applied
moment, Mu, adds bending stresses, as shown in Figure 5.5. Suppose a column has a
nominal axial load strength of Pn. If the load is applied at an eccentricity, e, the
column axial load capacity, Pn, will be reduced because it is subjected to a moment
in addition to the load. The moment, as shown on Figure 5.20c, adds compressive
stresses to the already compressed column. Thus, as the eccentricity of the load
increases, the applied moment increases, and the axial load capacity of the column
decreases.
Figure 5.21 shows the deformations and strains of a typical column subjected to
an axial load, Pu, and a bending moment, Mu. When the load is concentric, the
deformations across the section are uniform (i.e., the section shortens uniformly).
Because strain () is dened as the ratio of the change in length to the original
length, the distribution of strain across such a section is uniform, as shown in
Figure 5.21a. Assume that the same column section now is subjected to only a
moment, Mu, which causes one side of the section to be in tension and the other side
in compression. The result is a linear distribution of deformation (0 ) or strain (),
as shown in Figure 5.21b.
318 5 Columns
a Pu b c Pu
Mu Mu
1
2
Strain
1
Figure 5.21 Deformations and strains from axial load and moment
When a column is subjected to the combined action of an axial load, Pu, and a
moment, Mu, the compressive deformations and strains due to Pu and those due to
Mu add up, while the compressive deformations and strains due to Pu and the tensile
deformations and strains due to Mu reduce each other. Depending on how large Mu
is in comparison to Pu, a part of the section may be in tension (large Mu), or the
entire section may be in compression (small Mu). In Figure 5.21c, 1 is the shape of
the strain distribution for a large Mu/Pu ratio, and 2 is the strain distribution for a
small Mu /Pu ratio. Therefore, a column with a given amount of reinforcing may fail
due either to excessive compression, where the effects of the load and the moment
are added up, or to excessive tension, where tension from a large moment over-
comes the compression from the axial force.
The ultimate useful strain in the concrete is assumed to be 0.003. Any reinforced
concrete column with a given amount of reinforcing has a combination of Pu and Mu
that causes the compressive strain in the concrete to reach 0.003 while tensile strain
in the steel at the opposite side of the section reaches the yield strain. This state is
called a balanced failure condition, which is somewhat similar to that dened for
reinforced concrete beams (see Chapter 2).
Example 5.6 Determine the nominal axial load strength, Pn, and the nominal
moment, Mn, for the short tied column shown in Figure 5.22a for the following
cases: (1) axial load (i.e., e 0.0); (2) e 5 in.; (3) balanced condition; (4) no load
but moment (i.e., e 1); and (5) axial tensile load. Assume fc0 4;000 psi,
5.8 Behavior of Short Columns Under Eccentric Loads 319
16 in.
y
3 in.
4 #10
x x 24 in.
3 in.
fy 60,000 psi, and bending about the xx axis. Do not consider reduction in Pn due
to accidental eccentricity.
Solution
1. Assuming the load is concentric, the nominal axial load capacity of the column is
the sum of the compressive strengths of the concrete and the steel:
Pn Po 0:85 fc0 Ag Ast fy Ast
Ag 1624 384 in:2
4#10 ! Table A2:9 ! 5:08 in:2
Pn 0:854384 5:08 605:08
Pn 1,593 kip
s0 > y !fs0 fy
t < y !fs < fy
320 5 Columns
5 in.
Pn y
x
x
y
a y b c
0.003 0.85fc
Location of the 16 in.
axial load
3 in. s
C2
As Pn C1 a 0.85c
Pn c
5 in.
d
18 in. x x
14 in.
As
3 in. T
t
y
Column section Strain distribution Stress distribution and forces
Figure 5.23 Assumed strain and stress distribution for e 5 in. for Example 5.6
Figure 5.23 shows the assumed distribution of strain and stress at failure for
this section. The strains in the tension and the compression steel t and s0
depend on the location of the neutral axis (c). From similarity of the triangles of
Figure 5.23b, determine the relationship between t and c:
t dc
0:003 c
dc
t 0:003
c
5.8 Behavior of Short Columns Under Eccentric Loads 321
Use the volumes under the stresses shown in Figure 5.23c to calculate the
compression and tensile forces acting on the section:
T fs As
Substituting fs:
dc
T 87 As
c
21 c 21 c
T 87 2:54 221 kip
c c
Pn C1 C2 T 0
or
Pn C1 C2 T
21 c
Pn 46:24c 143:8 221 kip
c
1 0:85c
Pn 46:24c 21 143:8 18
14 2
Pn 1:40c 69:36c 184:89 kip
2
322 5 Columns
c 18:93 in:
Pn 995 kip
Having determined c and Pn, we now check the correctness of our assump-
tions. First, calculate the strain in the compression steel s0 . From similarity of
the triangles of Figure 5.23b:
s0 c3
0:003 c
c3
s0 0:003
c
18:93 3
s0 0:003
18:93
s0 0:00252 > y 0:00207
Hence, the second assumption (i.e., the tensile steel has not yielded) was also
correct.
5.8 Behavior of Short Columns Under Eccentric Loads 323
0:003 cb
y d cb
0:003d cb y cb
0:003d
cb
0:003 y
0:00321
cb
0:003 0:00207
cb 12:43 in:
Also from similarity of the triangles, calculate the strain (and stress) in the
compression steel s0 and fs0 :
a Location of the
b c
applied load (to be found) 0.003 0.85fc
Pn
Pn Pb
3 in. X s
C2
As C1 ab 0.85cb
cb eb
21 in.
18 in.
d
As
3 in. T
fy
16 in. y 0.00207
s0 cb 3
0:003 cb
0:003cb 3
s0
cb
0:003 12:43 3
s0
12:43
0
s 0:00228
Because s0 0:00228 > y 0:00207, fs0 fy 60 ksi: The sum of the forces
acting on the section is Pn or Pb (see Figure 5.24c):
From the sum of moments about the tensile steel, determine the balanced
eccentricity, or eb:
0:85cb
Pn eb 9 C1 d C2 18
2
0:85 12:43
566eb 9 574:6 21 143:818
2
eb 9 20:53
eb 11:53 in:
Thus, for case 3, Pn 566 kip and Mn 544 ft-kip. Note that as the moment
increases, the axial load decreases.
4. In case 4 the column is subjected only to moment. This is obviously only a
theoretical case, as columns always have an axial load. Because the eccentricity
is the ratio of moment to applied load, this condition represents a very large
(innite) eccentricity. Essentially, columns subjected to pure moments behave
like doubly- reinforced beams. Assuming steel in tension yields (fs fy) before
5.8 Behavior of Short Columns Under Eccentric Loads 325
0.003 0.85fc
a b c
3 in. s C2
a 0.85c C1
As c
21 in. Mn
18 in.
As
T
3 in.
16 in. t > y
Figure 5.25 Strain, stress, and force distribution of column subjected only to moment
s0 c3
0:003 c
0:003 c 3
s0
c
0 0 0
Because s < y ! fs Es s
0:003c 3
fs0 29, 000
c
c 3
fs0 87
c
C1 0:85fc0 ab
C1 0:8540:85c16 46:24c kip
C2 As0 fs0 0:85fc0
c 3
C2 2:54 fs0 0:854 2:54 87 3:4
c
c 3
C2 221 8:64 kip
c
T As fy 2:5460 152:4 kip
Equilibrium of the forces acting on the section (Figure 5.25c) requires that:
C1 C 2 T
221c 3
46:24c 8:64 152:4
c
Simplifying the above equation, we get:
c 3 873:2 3
fs0 87
c 3:2
fs0 5:44 ksi
0:85c
M n C1 d C 2 d 3
2
0:85 3:2
Mn 148 21 5:221 3
2
Mn 2907 93 3000 in:-kip=12 250 ft-kip
Thus, for case 5, Pn 305 kip and Mn 0. The negative sign means that the
column is in tension.
The following table shows the calculated values of Pn, Mn, and corresponding e.
Figure 5.26 shows the plot of these values to better visualize the results
obtained. The horizontal axis in the graph is Mn Pne and the vertical axis is Pn.
The graph shows the combinations of moment, Mn, and load, Pn, at which the
column may fail. This graph is called a column interaction diagram. The
interaction diagram in Figure 5.26 is a unique property of a specic column
with given dimensions, materials, and amount of reinforcing. When the load is
applied with no eccentricity or moment, the column has a nominal axial load
capacity of Pn 1,593 kip. As the eccentricity, e, increases (or the moment on
the column increases), the axial load capacity of the column decreases until it
reaches the balanced failure condition, which is when the failure of concrete in
compression and the yielding of steel in tension occur simultaneously. Values of
Pn and Mn above the balanced condition cause the concrete to crush in compres-
sion (c 0.003) before the steel yields in tension. Therefore, the failure of the
column section in this region is compression controlled. If the eccentricity
increases from the balanced condition (eb), the failure of the section occurs at
decreasing values of Pn and Mn. This may seem odd; however, when the
eccentricity increases from a balanced condition (e > eb), the steel in tension
yields before concrete in compression crushes (i.e., the element, in fact, may act
as a exural or bending member). In this region the failure of the section is
328 5 Columns
Pn(kip)
2,000
(0, 1,593)
1,500
(415, 995)
1,000
. Compression-controlled
5 in (544, 566)
e
500 .
M
eb 11.53 in Tension-controlled
P e
Balanced condition
(250, 0)
Mn Pne (ft-kip)
100 200 300 400 500 600
(0, 305)
500
tension controlled. Because the column fails in tension in this region, an increas-
ing compression force, Pn, keeps the section from failing and results in an
increase in its moment capacity, Mn, as well.
The interaction diagram also shows that, if we draw a line connecting a point
on the diagram to the origin (Pn Mn 0), any point on this line represents a P
and M combination that has the same eccentricity (e), because the ratios of P and
M are constant.
Here we repeat Example 5.6 to gain a better understanding of the interaction
diagram and its relationship to the distribution of strain across the section. This
time, however, we examine Pn and Mn values for different levels of strain in the
tensile steel (t).
Example 5.7 Determine Mn and Pn from the interaction diagram of Example 5.6
for the following different levels of strain in the tensile steel: (1) t 0.0;
(2) t 0.25y; (3) t 0.50y; (4) t 0.75y; (5) t y; (6) t 0.0035;
(7) t 0.0040; and (8) t 0.0050.
Solution For fy 60,000 psi, the yield strain, y, is:
fy 60,000
y 0:00207
Es 29,000,000
if t < y ! fs Es t
5:18
if t y ! fs fy
Now, we consider the distribution of strain and stress on the column section, as
shown in Figure 5.27.
5.8 Behavior of Short Columns Under Eccentric Loads 329
21 in.
Pn Pn
e x e
x
d
9 in.
As
T
3 in. fs
t
Column section Strain distribution Stress distribution
and forces
Figure 5.27 Strain, stress, and force distribution of column of Example 5.7
c dc
0:003 t
5:19
0:003d 0:00321 0:063
c
0:003 t 0:003 t 0:003 t
c c3
0:003 s0
5:20
0:003 c 3
s0
c
if s0 < y ! fs0 Es s0
if s0 y ! fs0 f y
Pn C1 C2 T
5:21
Pn 46:24c 2:54 fs0 8:64 2:54 f s kip
For each strain case we can use Equations (5.18), (5.19), and (5.20) to determine
the location of the neutral axis (c) and the stress in the tensile and the compression
steels (fs and fs0 ). Having c, fs, and fs0 , we can calculate Pn, e, and Mn from
Equations (5.21) to (5.23). The table below shows the results for each case:
0
Case t s c (in.) Mn (ft-kip) Pn (kip) e (in.)
1 0.00 0.0026 21.0 357 1,115 3.84
2 0.25(0.00207) 0.0025 17.91 439 934 5.64
3 0.5(0.00207) 0.0024 15.62 488 790 7.41
4 0.75(0.00207) 0.0023 13.84 520 670 9.32
5 1.0(0.00207) 0.0023 12.43 544 566 11.53
6 0.0035 0.0021 9.69 517 440 14.1
7 0.0040 0.0020 9.0 502 402 14.96
8 0.0050 0.00186 7.875 473 340 16.7
Figure 5.28 shows the interaction diagram generated from the results of Exam-
ples 5.6 and 5.7. This diagram is the same as the one shown in Figure 5.26. The
levels of the tensile steel strain (t) and stress (fs) along the curve are also shown.
This interaction diagram shows the maximum nominal capacity of the column. Any
combination of Pu and Mu that lies inside the curve (e.g., point A) is safe for the
column; however, any combination of Pu and Mu that lies outside the interaction
diagram (e.g., point B) will cause the column to fail.
In Examples 5.6 and 5.7, the rectangular column had 4 #10 bars. But if we increase
the area of reinforcements (e.g., to 8 or 12 #10 bars), the shape of the interaction
diagram would remain approximately the same but would have larger values of Pn
and Mn, as shown in Figure 5.29.
5.9 ACI Column Interaction Diagrams 331
Pn(kip)
2,000
B
(0, 1,593)
1,500
(357, 1,115)
(1) t 0, fs 0
1,000
(439, 934)
(2) t 0.25 y , fs 0.25fy
(3) t 0.5 y , fs 0.5fy (488, 790)
A (520, 670)
(4) t 0.75 y , fs 0.75fy
(5) t y , fs fy (544, 566)[balanced condition]
500
(6) t 0.0035 (517, 440)
(7) t 0.004 (502, 402)
(8) t 0.005 (473, 340)
(250, 0)
Mn Pne (ft-kip)
100 200 300 400 500 600
Pn
12 #
10
8 #1
0
4#
10
Mn
Figure 5.29 Column interaction diagrams for different areas of steel reinforcements
332 5 Columns
Column interaction diagrams exist for rectangular and round columns with
different reinforcement arrangements, concrete compression strength, and steel
tensile strength. These curves are similar to the one shown in Figure 5.29. To
make these curves more versatile, however, the Pn values are substituted by
Pn
Kn 0 , which is a nondimensional parameter as long as the values are sub-
fc Ag
Mn
stituted with consistent units (e.g., kip and inches). Also, Rn 0 is used for the
fc Ag h
horizontal axis instead of Mn, which is a nondimensional value. Each set of curves
is made for a specic arrangement of reinforcement, compressive strength of
concrete ( fc0 ), yield strength of steel (fy), steel ratio (g), and parameter , which
represents the spread of reinforcements in the column:
h0
5:24
h
where h0 and h are the distance between the center to center of the extreme steel in
the column, and the total depth of the column, respectively, as shown in Figure 5.30.
The dimensions h0 and h are measured perpendicular to the bending axis of the
column.
Figure 5.31 shows the interaction diagram for a rectangular column with steel
reinforcement uniformly distributed around the column. It is from the ACI Design
Handbook, SP-17(11). In Figure 5.31, fc0 4 ksi; fy 60 ksi, and 0.6.
The interaction diagrams are for g 0.010.08. Figure 5.32 is a similar interaction
diagram for a circular column with fc0 4 ksi, fy 60 ksi, and 0.8. The
diagrams show the levels of stress in the tension steel (fs) as a fraction of the steel
yield strength (fy). In addition, they indicate tensile strains of t 0.0035 and
0.0050. Another value that is given is Kmax, which is the maximum useable nominal
axial load capacity for a tied column:
Bending axis
Bending axis
h h
h h
2.0
INTERACTION DIAGRAM R4-60.6 h
fc 4 ksi h
1.8 g fy 60 ksi
0.0 0.6
8
1.6 Kmax
0.0
7
0.0 e P
6 n
1.4
0.0
5
0.0
1.2 4
fs /fy 0
0.0
fc Ag
3
Pn
1.0 0.0
2
Kn
0.25
0.0
1
0.8
0.50
0.6
0.75
0.4
0.0
0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35
Pne
Rn
fc Ag h
h i
K max 0:80 0:85fc0 Ag Ast Ast fy 5:25
In essence Kmax represents a cut-off level for Kn (and consequently for Pn). The
value of Kmax in the diagrams is dened for tied columns, as just mentioned. For a
0:85
spiral column, the value from the diagrams has to be multiplied by 1:0625
0:80
(i.e., the ratio of the limiting coefcients that account for accidental moments).
These computer-generated interaction diagrams assume the reinforcement to be a
thin rectangular tube for rectangular cross sections that have longitudinal reinforce-
ments distributed along all four faces, and a thin circular tube for patterns of
longitudinal steel bars arranged in a circle.
334 5 Columns
2.0
INTERACTION DIAGRAM C4-60.8 h
h
fc 4 ksi
1.8 g fy 60 ksi
0.0
8 0.8
0.0
1.2 4
0.0
3 0.25
fc Ag
Pn
1.0 0.0
2
Kn
0.0
1 0.50
0.8
0.75
0.6
1.0
0.4
et = 0.0035
0.2
et = 0.0050
0.0
0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35
Pne
Rn
fc Agh
The latest ACI column interaction diagrams [ACI Design Handbook, SP-17(11)]
take no consideration of the strength reduction factor, , which is a signicant
change from previous versions. This change is intended mainly to make the dia-
grams as universal as possible.
As in the case of beams, the design resisting moment of columns (MR) and their
design axial load strength (PR) are:
MR Mn
PR Pn
5.10 Design Axial Load Strength (Pn), and Moment Capacity (Mn) 335
0.9
Spiral columns
0.75
0.65
Tied columns
Compression Tension
controlled Transition controlled
et
ety 0.005
Figure 5.33 Variation of strength reduction factor () with the net tensile strain in steel (t)
where is the column strength reduction factor, which depends on the level of
strain in the tension steel. Figure 5.33 shows the column strength reduction factor as
a function of net tensile steel strain (t) for tied and spiral columns. Most building
columns are compression controlled, that is, the concrete reaches the ultimate
useable compressive strain of 0.003 before the strain in the tensile steel reaches
the yield strain (t). The factor is constant for compression-controlled sections
(0.75 for spiral columns and 0.65 for tied columns). If the moment on the column is
relatively large compared to the axial load, the column section may be in the
transition zone. Then the factor varies between 0.75 for spiral columns, or 0.65
for tied columns, and 0.90 as t varies between ty and 0.005. If t is more than
0.005, the section is tension controlled, and is constant and equal to 0.9. Such a
section acts like a exure member (beam) rather than a compression member
(column).
If the column section is compression controlled, we can easily calculate the
design resisting axial load, PR, and the design resisting moment, MR; however, when
the column is in the transition zone, we must calculate from t as given below
from ACI 318-14, Table 21.2.2
t ty
0:75 0:15 spiral column 5:26
0:005 ty
336 5 Columns
0.4
fc Ag
Pn
g 0.01
0.2
Kn
0.0
2
0.0
3
0.0
4
0.
05
0.
06
0.
07
0.
08
0.0
0.65 0.70 0.75 0.80 0.85 0.90
Figure 5.34 Kn vs. diagram for the interaction diagram of Figure 5.31
t ty
0:65 0:25 tied column 5:27
0:005 ty
These values have signicance in only a small area of the interaction diagrams (i.e.,
f
the zone between s 1:0 and t 0.005).
fy
Graphs that relate and Kn have been developed for each interaction diagram to
simplify the computation of . Figures 5.34 and 5.35 are examples of such graphs
and are to be used in conjunction with the interaction diagrams of Figures 5.31 and
5.32, respectively. Appendix A contains additional interaction diagrams and their
corresponding Kn versus graphs.
0.6
0.4
fc Ag
Pn
g = 0
Kn
.01
0.2 0.02
0.03
0.04
0.05
0.06
0.07
0.08
0
0.75 0.76 0.77 0.78 0.79 0.8 0.81 0.82 0.83 0.84 0.85 0.86 0.87 0.88 0.89 0.9
Figure 5.35 Kn vs. diagram for the interaction diagram of Figure 5.32
given factored load may safely have? Regardless of the approach, one can always
take advantage of the interaction diagrams.
The value of is constant if the section is compression controlled. For columns
in the transition zone, however, must be adjusted accordingly. Therefore, the
procedures for the analysis of columns in compressioncontrolled and noncom-
pression-controlled zones are somewhat different.
The following are steps for the analysis of compression-controlled members which
are summarized in Figure 5.36:
Step 1. Calculate and check the column steel ratio, g:
Ast
0:01 g 0:08
Ag
The strength reduction factor, , is equal to 0.65 for tied, and 0.75 for
spiral columns.
h0
Step 2. Calculate (see Figure 5.30) and the nondimensional factors Kn or Rn:
h
Pu
Kn
fc0 Ag
Mu
Rn
fc0 Ag h
338 5 Columns
Analysis of Columns
(short with large eccentricity)
(compression - controlled section)
1.
Calculate and check the steel ratio:
0.01 g 0.08
= 0.65 (tied column)
= 0.75 (spiral column)
2. h Pu Mu
Calculate = , Kn = (or Rn = )
h f c Ag f c Ag h
Select the interaction diagram based on fc , fy , and , and find Rn (or Kn)
for the computed g (Appendix A).
Yes
3.
Calculate Pn = KnfcAg, PR = Pn
Mn = Rnf c Agh , MR = Mn
4.
5. PR Pu
The section is ok. Yes No
or The section is N.G.
Check the ties/spirals.
MR Mu?
Figure 5.36 Flowchart for the analysis of columns with compression-controlled section
If we know both Pu and Mu, then we may take these calculated Kn and Rn
values as the demand on the section. We now enter into the appropriate
interaction diagram (based on fc0 , fy, and ) and locate the point dened by
the calculated Kn and Rn. If this point falls within the curve dened by g
(calculated in step 1), the Pu and Mu combination is safe for this column.
Another approach is to calculate only Kn (or Rn) and, from the appropri-
ate interaction diagram, obtain the corresponding Rn (or Kn) using the
calculated g.
5.11 Analysis of Short Columns with Large Eccentricity Using Interaction Diagrams 339
Pn K n fc0 Ag , PR Pn
Mn Rn fc0 Ag h, MR Mn
0.65 for the tied columns and 0.75 for the spiral columns.
Step 4. Check the column capacity. For the column to be adequate, its axial load
capacity, PR, has to be greater than the applied load, Pu or the column
moment capacity, MR, has to be larger than the applied moment, Mu:
PR Pu or MR Mu
Step 5. Check the ties or spirals. If the section is ok, we can check the ties or spirals
as we did for the columns with small eccentricity (step 5 of Figure 5.16).
Analysis of Columns
(short with large eccentricity)
(noncompression-controlled section)
2a.
Estimate the value from the
appropriate K n versus diagram.
Pu Mu
Calculate K n or R n
fc Ag fc Ag h
2b.
Recalculate the value
using the K n versus diagrams.
Is the new
No value the
same as the
previous one?
Yes
3.
Calculate P n K n fc Ag , PR Pn or
Mn R nfc Ag h , MR Mn
4.
5. PR Pu
The section is ok. Yes No
or The section is N.G.
Check the ties/spirals. MR M u?
Figure 5.37 Flowchart for the analysis of columns with noncompression-controlled section
5.11 Analysis of Short Columns with Large Eccentricity Using Interaction Diagrams 341
Pu
Kn
fc0 Ag
Mu
Rn 0
fc Ag h
Use fc0 , fy, and to select the appropriate interaction diagram. Having
Kn or Rn and, g we can either locate the corresponding point on the
diagram, when both Kn and Rn are known, or obtain the corresponding
Rn (or Kn), when only one of them is known.
Step 2b. Recalculate the value. At this point, recalculate the value using the Kn
value obtained in step 2a. We can determine the corrected factor by
using Kn versus diagrams such as those in Figures 5.34 and 5.35. If the
new factor is close to the previous estimate, proceed to step 3. Otherwise,
move back to step 2a and use the new factor to revise Kn or Rn, then
repeat the process.
Step 3. Calculate the columns resisting load and moment:
Pn K n fc0 Ag PR Pn
Mn Rn fc0 Ag h MR Mn
Step 4. Check the adequacy of the column. The following relationships must be
satised for the column to be adequate:
PR Pu or MR Mu
Step 5. Check the ties or spirals. If the column section is adequate, we can check
the ties or spirals (step 5 of Figure 5.16).
Example 5.8 Determine the maximum axial load that can be applied on the short
tied column section shown below (this is the column used in Example 5.1). The
applied dead and roof live load moments are MD ML 35 ft-kip. Use fc0 4,000
psi and fy 60,000 psi.
14 in.
8 #8 14 in.
#3 @ 14 in.
Solution
Step 1. Check the column steel ratio, g:
Ast 6:32
g 0:032
Ag 196
0:01 < g 0:032 < 0:08 ok
h 14 in: ! Table A5:1 ! Maximum of 12 #8 bars ok
fc0 4 ksi, fy 60 ksi, and 0.66; therefore, we use the interaction dia-
grams of Figures A5.1a and A5.2a and interpolate:
Pn K n fc0 Ag
Pn 0:704196 549 kip
PR Pn 0:65549 357 kip
If we compare PR 357 kip with the result of Example 5.1 for the same
column but with small eccentricity, PR 533 kip, the effect of added
moment on the reduction of the column axial load capacity is evident.
Step 5. The procedure of checking the ties is the same as that shown in
Example 5.1.
5.11 Analysis of Short Columns with Large Eccentricity Using Interaction Diagrams 343
Example 5.9 Check the adequacy of the short spiral column shown below (the
same as the one used in Example 5.2) if it is subjected to MD 36 ft-kip and
MLr 43 ft-kip. Use PD 200 kip, PLr 225 kip, fc0 4000 psi, and fy 60,000 psi.
6 #9
Solution
Step 1. Check the column steel ratio, g:
Ast 6:0
g 0:03
Ag 201:1
0:01 < g 0:03 < 0:08 ok
K n 0:66
Thus
and
In Example 5.2, the axial load capacity of this column, PR, was 652 kip,
which was satisfactory. Comparison of the axial load capacities again
shows the signicant decrease due to the applied moment.
Alternatively, the following question could be asked: How large an
eccentricity may the given loads safely have? Then we proceed as
follows.
For Kn = 0.99 and g = 0.03, the corresponding Rn from the interaction diagram is
Rn 0:08
Thus
and
MR 64:4 12
e 1:29 in:
Pu 600
Example 5.10 Determine whether the tied column shown below is adequate. The
applied factored axial load and bending moment are Pu 70 kip, and Mu 60 ft-kip,
respectively. Use fc0 4 ksi; and fy 60 ksi. Also, check the adequacy of the ties.
The typical clear cover is 1.5 in.
5.11 Analysis of Short Columns with Large Eccentricity Using Interaction Diagrams 345
12 in.
8 #6 12 in.
#4 @ 12 in.
Solution
Step 1. Check the steel ratio, g:
Figure A5.1a is the interaction diagram for fc0 4 ksi; fy 60 ksi, and 0.60 with
uniformly distributed reinforcements for a rectangular column. The point for g 0.024
f
and Kn 0.19 falls in the transition zone (i.e., the region between s 1:0 and
fy
t 0.005). Therefore, continue with the analysis for columns with non-compression-
controlled sections using the owchart of Figure 5.37.
Step 2a. Use Figure A5.1b with g 0.024 and Kn 0.19 to obtain the strength
reduction factor, :
0:735
346 5 Columns
Pu 70
Kn 0:165
fc0 Ag 0:7354:012 12
Pu 70
Kn 0:160
fc0 Ag 0:764:012 12
Pu 70
Kn 0:158
fc0 Ag 0:774:012 12
From Figure A5.1a, we can conclude that the section is in the transition
zone and obtain Rn 0.152. Use Figure A5.1b to obtain 0.77. Because
the new value is about the same as the one obtained in the previous
iteration, we proceed with step 3.
Step 3.
Mn Rn fc0 Ag h
0:152412 1212
Mn
12
Mn 87:6 ft-kip
MR Mn 0:7787:6 67:5 ft-kip
Step 4.
MR 67:5 ft-kip > Mu 60 ft-kip ok
The design of columns with large eccentricity, similar to the analysis, depends on
the column behavior under the load and moment. If the column is compression
controlled, the factor is constant, and the design process is straightforward. If the
column is not compression controlled, however, the factor varies with the tensile
strain in the steel (t). Thus, using the value as if the column were compression
controlled would lead to a conservative design. An iterative approach is necessary
to calculate the correct value.
The following are steps for the design of compression-controlled columns, which
are also summarized in the owchart of Figure 5.38:
Step 1. Determine the factored loads and moments acting on the column. The loads
on the column come from the beams and slabs connected to it. The
moments acting on the column are due to either gravity loads or lateral
loads such as wind or earthquake loads. If gravity loads are considered (PLr
is the roof live load, and MLr is the roof live load moment):
Step 2. Estimate the column size. In most cases, the column size is preselected
based on architectural considerations or ease of construction. If it is
necessary to estimate the column size, however, we must make reasonable
assumptions because both the area of the column (Ag) and the area of steel
(Ast) are unknown. Different simplifying assumptions may be used to
348 5 Columns
Design of Columns
(short with large eccentricity)
(compression-controlled section)
1.
Determine Pu and Mu
4.
Yes
5.
Calculate the area of steel,
Ast = g Ag , and select the bars.
Yes
6.
Design the ties/spirals.
obtain a preliminary size. The most common assumptions are that the
columns capacity is the same as that of an axially loaded column and
that the area of steel, Ast, is neglected. The latter is made in an attempt to
account for the effects of the moments. The preliminary design equations
are given below:
5.12 Design of Short Columns with Large Eccentricity 349
h i
PR 0:80 0:85 fc0 Ag Ast fy Ast tied column
h i
PR 0:85 0:85 fc0 Ag Ast fy Ast spiral column
Pu
Ag tied column
0:80:85fc0
Pu
Ag spiral column
0:850:85fc0
After calculating the gross area of the column, Ag, we can select the
column size. Round the column size to the nearest inch.
Step 3 and 4 Calculate , Kn, and Rn. In order to utilize interaction diagrams it is
h0
necessary to calculate : We assume that the centerline of the longi-
h
tudinal reinforcing is 2.5 in. from the face of the column.
Thus
h0 h 22:5 h 5 in:
Using fc0 , fy, and , we select the appropriate interaction diagram, and then
compute Kn and Rn:
Pu Mu
Kn Rn
fc0 Ag fc0 Ag h
Entering the interaction diagram with the known values of Kn and Rn, we
determine whether the column is compression controlled. If the column is
not compression controlled or if Rn is out of the range of the interaction
diagram, we proceed to the owchart for the design of noncompression-
controlled sections (Figure 5.39). Otherwise, use Kn and Rn to obtain g and
move to step 5.
Step 5. Calculate the required area of steel, Ast:
Ast g Ag
350 5 Columns
Select bars using Tables A2.9 and A5.1. If the selected bars can t into
the column based on Table A5.1, we proceed to step 6. Otherwise, increase
the column size, which will increase the space between the bars, and
repeat the process starting back in step 3.
Step 6. Design the ties or spirals. This step is the same as that for columns with
small eccentricity (as shown in Figure 5.17).
MR Rn fc0 Ag h
To resize the column, make MR Mu and solve for Agh as shown below:
Mu Rn fc0 Ag h
Mu
Ag h
fc0 Rn
Design of Columns
(short with large eccentricity)
(noncompression-controlled section)
4a.
Assume a value between
0.75 and 0.90 (start with 0.80)
4b.
Is the point on
the interaction diagram No
within the range of
curves?
4c.
Select R n in the
transition zone for g 0.02.
Yes
4e. h
Calculate , and: 4d.
h Recalculate the
Pu column size (h):
Kn
fc Ag Mu
Ag h
Mu fc R n
Rn
fc Ag h
5.
Is the
No
same as the
previous value?
Yes
6.
Design the ties/ Calculate the area of steel,
spirals. Ast g Ag , and select the bars.
Step 4e. Determine , Kn, and Rn. Once you know the new size of the column,
h0
determine (As in step 3, assume h0 h 2(2.5) h 5 in.). Also,
h
calculate Kn and Rn as
Pu Mu
Kn Rn
fc0 Ag fc0 Ag h
352 5 Columns
Step 4f. Obtain g from the appropriate interaction diagram, and update the
value. To do so, use , fc0 , and fy, and the assumed value to select
the appropriate interaction diagram. Then use the values of Rn and Kn
(determined in the previous step) to obtain g from the interaction
diagram. Also, determine from the Kn versus diagram.
Step 5. Determine whether convergence has been achieved. If the new factor is
not approximately the same as the one obtained in the previous cycle,
repeat the process with this new value starting in step 4e. Repeat the
procedure cycle until the value converges. After achieving convergence,
compute the required area of steel by using the steel ratio, g, determined
in the last cycle:
Ast g Ag
Select the reinforcing bars using Table A2.9 and check the layout using
Table A5.1.
Step 6. Design the ties or spirals. This step is similar to that of the design of
compression-controlled columns.
Example 5.11 Design a short tied square column to carry PD 300 kip,
PL 175 kip, PLr 80 kip, MD 150 ft-kip, ML 90 ft-kip, and MLr 32 ft-kip.
Assume fc0 4,000 psi; fy 60,000 psi, and that the main reinforcements are
distributed uniformly around the column edges.
Solution
Step 1. Calculate the factored load, Pu, and the factored moment, Mu:
With reference to Equation 2.3d, since
1.83 PLr 1.83(80) 146.4 k < PL 175 k
and
therefore,
Pu
Ag
0:80:85 fc0
680
Ag 385 in2
0:80:650:854:0
p p
h Ag 385 19:6 in:
Try a 20 in: 20 in: column
Step 3.
h0 h 22:5 20 22:5 15 in:
h0 15
0:75
h 20
Pu 680
Kn 0
0:65
fc Ag 0:65420 20
Mu 34012
Rn 0 0:20
fc Ag h 0:65420 2020
0:034 0:04
g 0:037
2
Step 4. We now check our assumption for column behavior. Because the column is
in the compression-controlled region of the interaction diagram, our
assumption for the factor is correct. Go to step 5.
Step 5. The required area of steel, Ast, is:
Using Tables A2.9 and A5.1 ! Use 12 #10 bars (Ast 15.24 in2)
Step 6. Design of the ties: Using #3 ties, the maximum spacing of the ties is:
Based on Figure 5.12, this column requires three sets of ties, as shown in
the gure below.
20 in.
12 #10
20 in.
#3 @ 18 in.
Example 5.12 Solve Example 5.11 for a round spiral column. fyt 60,000 psi.
Solution
Step 1. From Example 5.11, Pu 680 kip, and Mu 340 ft-kip.
Step 2. Estimating the column size:
Pu
Ag
0:850:85 fc0
680
Ag 314 in:2
r 0:85
0:85 0:75 4:0
r
Ag 314
h2 2 20:0 in:
3:14
202
Assume h 20 in: ! Ag 314 in:2
4
Steps 3 and 4
h0 h 22:5 20 22:5 15in:
h0 15
0:75
h 20
Pu 680
Kn 0:72
fc0 Ag 0:754314
Mu 34012
Rn 0:22
fc0 Ag h 0:75431420
Step 5.
Ast g Ag 0:062314 19:5 in:2
Step 5.
Ast g Ag 0:034380 12:9 in:2
Use 13 #9 bars Ast 13:0 in:2
hc 22 21:5 19 in:
192
Ach 284 in2
4
0
Ag f
s, min 0:45 1 c
Ach fyt
380 4:0
s, min 0:45 1 0:0101
284 60
356 5 Columns
4 Asp 40:11
s
hc s 19s
s, min s
40:11
0:0101
19s
s 2:29 in: ! s 21=4 in:
sclear 21=4 in: 3=8 in: 17=8 < 3 in: ok
s 2 1=4 in:
13 #9
22 in.
Example 5.13 Design a circular spiral column to resist Pu 300 kip, and
Mu 400 ft-kip. Use fc0 4 ksi and fy 60 ksi. Use a maximum steel ratio of
0.02. The design of the spirals is not required.
Solution
Step 1. Pu 300 kip, and Mu 400 ft-kip.
Step 2. Estimate the column size (h):
Pu 300
Ag 138:4 in:2
0:850:85fc0 0:850:750:85 4
r r
Ag 138:4
h2 2 13:3 in:
3:14
Try h 14 in:
142
Ag 154 in:2
4
5.12 Design of Short Columns with Large Eccentricity 357
Step 3.
h0 h 22:5 14 22:5 9 in:
h0 9
0:64
h 14
Pu 300
Kn 0 0:65
fc Ag 0:754154
Mu 40012
Rn 0
0:74 > 0:30 Out of range:
fc Ag h 0:75415414
Because Rn is out of the range of the interaction diagram, use the owchart
for non-compression-controlled columns (Figure 5.39).
Step 4a. Assume 0.80.
Step 4b. As noted in step 3, the value of Rn is not within the diagrams range;
therefore, proceed to step 4c.
Step 4c. In order to resize the column, we must have a reasonable value of Rn.
Using the interaction diagram of Figure A5.10a ( fc0 4 ksi, fy 60 ksi,
and 0.70) for a steel ratio, g 0:02, we obtain the maximum value of
Rn. This value is about 0.14. We use this value to select a reasonable size
for the column.
Step 4d.
Mu
Ag h 0
fc Rn
40012
Ag h 10, 714 in3
0:8040:14
2
h
h 10, 714
4
h3
10, 714 ! h3 13, 641 in3
4
h 23:9 in: Try h 24 in
Step 4e.
h0 24 22:5 19 in:
h0 19
0:79 0:80
h 24
242
Ag 452 in:2
4
Pu 300
Kn 0 0:207
fc Ag 0:804452
Mu 40012
Rn 0 0:138
fc Ag h 0:80445224
358 5 Columns
Step 4f. Since fc0 4 ksi, fy 60 ksi, and 0.80, and the column is a round
section; therefore use the interaction diagram of Figure A5.11a. Using Kn
and Rn, we obtain a steel ratio, g 0.019. In addition, using the Kn versus
diagram of Figure A5.11b and Kn 0.207, we obtain a value of
0.825.
Step 5. Because the new factor (0.825) is higher than that of the previous cycle
(0.80), repeat step 4e with the new .
Step 4e.
Pu 300
Kn 0 0:20
fc Ag 0:8254452
Mu 40012
Rn 0 0:134
fc Ag h 0:825445224
Step 4f. From Figure A5.11a, g 0:017. From Figure A5.11b, 0:84.
Step 5. Because the new value of is close to that of the previous step (0.84
vs. 0.825), accept this value and calculate the area of steel (Ast):
Using Tables A2.9 and A5.1, select 7 #10 bars (As 8.89 in.2). The
following shows a sketch of the nal design.
7 #10
24 in.
Columns are divided into two classes based on their slenderness: short columns and
slender columns. Short columns crush under large axial force, whereas columns
with great slenderness may buckle before they fail in crushing.
Figure 5.40 shows a slender column subjected to an increasing axial compres-
sion force, P. As P increases the column may buckle (i.e., suddenly show large
lateral movement, see Figure 5.40). The stress at which the column starts buckling
is called the Euler buckling stress (fE):
5.13 Slender Columns 359
2 E
f E 2 5:28
ku
r
where
E modulus of elasticity of the concrete
k effective length factor for the column
u the unsupported length of the column, which is the clear distance between the
oor slabs, beams, or other members that provide lateral support for the column
s
Ig
r radius of gyration of column section, equal to in which Ig is the
Ag
moment of inertia of gross concrete section neglecting reinforcement and
Ag is the gross area of the column. The ACI Code (Section 6.2.5.1) suggests
a rounded value of r of 0.3h for rectangular columns, and 0.25h for circular
columns, where h is the dimension of the column in the direction perpen-
dicular to the axis of bending.
ku
The term is called the slenderness ratio. From the expression of the Euler
r
buckling stress we can see that as the slenderness ratio increases, the stress at which
the column buckles decreases. In other words, the column buckles at a lower stress
when it is more slender. The numerator of the slenderness ratio (ku) is called the
effective length, which depends on the unsupported column length (u), the type of
end supports (i.e., pinned, partially xed, or fully xed), and whether or not the
360 5 Columns
b Beam
Point of inflection
Column k u
u
Figure 5.41 Relationship between relative member sizes and end conditions of columns
column is allowed to move laterally (i.e., unbraced or braced). The effective length
is the length of that portion of the column that lies between two points of inection
of the buckled shape.
Column ends are connected to a foundation (at the base of the building), or to
slabs, beams and girders, or to both. Thus, there is no such thing as a true pin or a
fully xed support for columns. The amount of the columns xity depends on the
relative stiffness of the building elements at the ends of the column. For example, if
the size of the beams and girders at the ends of column is small compared to the
5.13 Slender Columns 361
column size, the end supports can be assumed to be pinned because the beams and
girders at the ends will provide little restraint to the free rotation of the column ends
when buckling. This condition is depicted in Figure 5.41a. But if the size of beams
and girders is large compared to the column, the end supports may be treated as
xed because the beams and girders will prevent the column ends from rotating, as
shown in Figure 5.41b. When the columns ends are pinned, the entire column
buckles; as a result, the columns effective length, ku, is almost the same as the
column length, u, in other words k 1.0. When the column ends are xed, its
buckled shape has two inection points; thus, ku is smaller than u. In the
theoretical case of perfect xity, ku 0.5u (i.e., k 0.5).
Another major factor inuencing the effective length of a column is the type of
structural system. If a column is part of the lateral load carrying system, it is subject
to sidesway, which means it can have signicant lateral motion and is called an
unbraced column. When other elements such as shear walls, however, are used as
the lateral load carrying system, the column will not have signicant lateral motion
and is called a braced column. The ends of an unbraced column have signicant
relative horizontal motion or sidesway. This relative movement is small when the
column is braced. A braced column is referred to as one without sidesway, and an
unbraced column as one with sidesway.
Sidesway affects the columns effective lengths. Figure 5.42 shows two columns
with xed-end supports. The column in Figure 5.42a is without sidesway and
a b
Point of inflection
Point of inflection
u k u u
k u
2
Point of inflection
k u
Point of inflection
Figure 5.42 Effective lengths for (a) column without sidesway, and (b) column with sidesway
362 5 Columns
5.13.2 P Effects
There is no such thing as a perfectly straight and vertical column. Applied moments
at the ends also bend the columns into a curvilinear shape. When a slender column
bends into a curve while subjected to an axial load, P, added moments, M, are
generated on the column. These moments have a magnitude of P multiplied by the
lateral deformation, . This is called the P effect.
Figures 5.43 and 5.44 show P effects on columns. The P effects on
columns with sidesway are more severe than those on columns without
sidesway.
The ACI Code requires a magnication of moments on slender columns due to
P effects. ACI Code (Section 6.2.5) allows P effectsto be ignored for
ku
columns without sidesway if the columns slenderness ratio satises the
r
following relationship (ACI Equation (6.2.5b)):
ku M1
34 12 5:29
r M2
M1
where M1 is the smaller end moment. is positive if M1 and M2 are acting in the
M2
same direction (column is bent into a double curve; see Figure 5.43b), and is negative
if M1 and M2 act in opposite directions (column is bent into a single curve;
see Figure 5.43a). The right side of Equation (5.29) is limited to a maximum value
of 40.
The ACI Code (Section 6.2.5) permits the effects of slenderness to be ignored for
ku
columns with sidesway when is less than 22. Figure 5.44 shows the P effects
r
on columns with sidesway.
Because of the complexity of the design of slender columns (as well as for
visual reasons), designers prefer to work with column dimensions that are not
slender. In general, we can ignore the slenderness effect of braced frames (columns
u
without sidesway) if is equal to 10 or less on lower oors, and 12 or less on upper
h
5.13 Slender Columns 363
a
P
M1
M1
M2
M2
P Moment diagram
Column bent into a single curve by the end moments
b
P
M1
M1
P
M2
M2
P
Moment diagram
Column bent into a double curve by the end moments
Figure 5.43 P effect on columns without sidesway superimposed over the end moments
u
oors. For unbraced columns must be smaller than 6 to have negligible slenderness
h
effects.
Designing slender reinforced concrete columns is a complex procedure.
Computer software is available to help the structural designer analyze and design
slender columns. The detailed discussion of this subject is beyond the scope of
this text.
364 5 Columns
a Sidesway
P
P
M1
M1
M2
M2
P Moment diagram
Column with sidesway bent into a single curve by the end moments
b Sidesway
P
P M1
M1
M2
M2
P Moment diagram
Column with sidesway bent into a double curve by the end moments
Figure 5.44 P effect on columns with sidesway superimposed over the end moments
Problems
3
d e
8 in. diameter @ 2 in.
6 #7
8 #8
16 in. 14 in.
5.2. Rework Problem 5.1 for fc0 4,000 psi and fc0 5,000 psi. What is the
percentage of change in the axial load capacity for each case? Do not check
the ties/spirals.
5.3. Repeat Problem 5.1 for fy 40,000 psi and fy 75,000 psi. What is the
percentage of change in the axial load capacity for each case? Do not check
the ties/spirals.
5.4. The square reinforced concrete tied column shown below is subjected to a dead
load of 200 kip and a roof live load of 220 kip. Determine whether the column
is adequate. The clear cover is 1.5 in. The loads eccentricity is negligible. Use
fc0 4,000 psi and fy 60,000 psi. Do not check the ties.
16 in.
#3 @ 14 in.
16 in.
8 #8
366 5 Columns
5.5. Rework Problem 5.4 for the following circular spiral column. Do not check
spirals.
8 #8
16 in.
Compare the PR determined in this problem with that of the square column of
Problem 5.4.
5.6. Determine the required reinforcements for a 12 in. 12 in. tied reinforced
concrete column subjected to a dead load of 20 kip and a roof live load of
30 kip. Assume that the loads have small eccentricity. Use fc0 5,000 psi
fy 60,000 psi, and 2 in. clear cover.
5.7. Redesign the column of Problem 5.6 for a dead load of 125 kip, a oor live
load of 175 kip, and a roof live load of 80 kip.
5.8. Design a square tied reinforced concrete column subjected to a dead load of
250 kip, a oor live load of 240 kip, and a roof live load of 150 kip. The
moments due to the loads are negligible. Use fc0 4,000 psi; fy 60,000 psi,
and 1.5 in. clear cover.
5.9. Redesign the column of Problem 5.8 as a circular spiral reinforced concrete
section. Assume fyt 60,000 psi.
5.10. A 12 in. 12 in. column reinforced with 4 #9 bars is subjected to axial roof
dead and live loads with small eccentricity. If the ratio PD/PLr 1.5, deter-
mine the maximum compressive axial service loads that the column can
carry. Use fc0 4,000 psi and fy 60,000 psi. Use #3 ties, and 1.5 in. clear
cover.
5.11. The following gures show the typical framing plan, elevation, and beam
section of a three-story reinforced concrete building. The oor live load is
50 psf and the roof live load is 15 psf. Assume 5 psf for mechanical/electrical
systems and 20 psf for partitions. Determine the required reinforcements for a
typical interior tied column between the ground and second levels. Use
fc0 4,000 psi, fy 60,000 psi, and 1.5 in. for cover. Do not reduce live
loads. Assume small eccentricity for the loads. The unit weight of the
concrete is 150 pcf. Neglect the self-weight of the column.
5.12. Rework Problem 5.4. Assume the column is subjected to a dead load moment,
MD 20 ft-kip, and a roof live load moment, MLr 30 ft-kip.
5.13. Determine the maximum factored moment, MR, that a 24 in. 24 in. column
with 12 #10 bars distributed uniformly around the column can carry when
Problems 367
subjected to a factored axial load, Pu 750 kip. Use fc0 4,000 psi,
fy 60,000 psi, #3 ties, and 1 1=2 in. for cover.
30'-0" 30'-0"
A A
10'-0" A
(typical)
16 in. 16 in.
column (typical)
6 in.
26 in.
16 in.
Section A-A
Roof level
12'-0"
3rd level
12'-0"
2nd level
15'-0"
Ground level
Elevation
368 5 Columns
5.14. Rework Problem 5.10. Assume that the column is subjected to a factored
moment, Mu 50 ft-kip.
5.15. Rework Problem 5.11. Assume that the column is subjected to a factored
moment, Mu 110 ft-kip. Place the longitudinal reinforcing uniformly
around the four faces.
5.16. Determine PR values for the columns shown below subjected to the factored
moments indicated. Use fc0 4,000 psi, fy 60,000 psi, and 2.0 in. cover.
Do not check ties or spirals.
a 20 in.
b
y
3 in. Diameter spiral
8
x x 20 in.
8 #8
8 #8
#4 Ties
y 20 in.
Mux = 200 ft-kip Mu = 200 ft-kip
c d 16 in.
18 in. y
y
x x x 20 in.
x 18 in.
6 #10 #4 Ties
y 8 #11 #4 Ties
y
Mux = 140 ft-kip
Muy = 180 ft-kip
5.17. A square tied column is subjected to a factored load, Pu 250 kip, and a
factored moment, Mu 50 ft-kip. Design this column. Use fc0 4,000 psi,
fy 60,000 psi, and 1.5 in. cover. Do not design the ties. Place the longitu-
dinal reinforcing uniformly in the four faces.
5.18. A circular spiral column is subjected to a factored load, Pu 400 kip, and a
factored moment, Mu 150 ft-kip. Design this column. Use fc0 4,000 psi,
fy 60,000 psi, and 1.5 in. for cover. Do not design the spirals.
Self-Experiments 369
Self-Experiments
Moving load
A A
A-A
6.1 Introduction
The appropriate selection of a oor system heavily inuences the overall cost of a
building. A designer has to take many factors into account when making such a
selection; and, unfortunately, the structural scheme cheapest to construct may not
be the best bargain in terms of overall construction cost of the building.
The rst element that forms the oor surface is the slab. Beams or columns, in
mathematical abstraction can be described by a single line. These are called linear
elements, because one of their three dimensions, the length, is much greater than the
other two (i.e., the dimensions of the cross-section). The load-path of linear
elements is easy to describe: They carry their loads along their length to the
supports.
A single line, however, cannot describe slabs or plates. As discussed in Chapter 2,
slabs have two dimensions that are signicantly larger than the third one, the
thickness. They are usually described mathematically as thin plates. The exact
bending theory of thin plates, based on the theory of elasticity, requires the solution of
a partial differential equation of the fourth order. This is completely impossible for
any practicing engineer or architect. Furthermore, so-called exact solutions fail to
deal with everyday realities, such as reinforced concrete that does not follow strict
elastic behavior, or load distributions that are not nice and uniform, and so on.
Fortunately, we can understand how a slab behaves by carefully considering how
it deforms under loads. Slabs bend in two directions, so a single line cannot describe
the bent shape. A way of describing a bent surface is therefore necessary for
understanding slab behavior. The load-path (i.e., the way a slab transfers any
load to the supports) depends on the way the slab bends between those supports,
which in turn depends on the way the designer chooses to support the slab.
The types of supports for slabs are divided into three groups.
1. Point supports. These consist of columns, posts, suspension points, and so
on. Slabs supported by supports of this type are referred to as at slabs or at
plates.
2. Line supports. Examples of line supports are beams, girders, and walls. Slabs
supported by supports of this type are referred to as one- or two-way slabs.
3. Continuous media (Slabs on grade supported by soil).
Admittedly the classications of the rst two types of support are somewhat
arbitrary. Often a designer may employ linear support elements (beams and walls)
in conjunction with point support elements, which makes the referencing more
difcult.
The overall cost of a monolithic concrete oor system depends on several
factors. First and foremost among these is the cost of shoring and forming. A strong
shoring system must be constructed. It should safely support the weight of the
freshly poured concrete and the associated construction loads (i.e., the people and
the equipment necessary for placing and nishing the wet concrete). The forms that
will serve as the mold must also be built.
The costs in a reinforced concrete oor system usually break down as follows:
These gures are based on current material and labor costs in the United States and
may not necessarily be the same elsewhere in the world. They clearly show,
however, that the cost of labor in the United States usually is greater than the raw
cost of materials. Thus, the selection of the right oor system for a building is not an
easy task.
Figure 6.1 shows a typical at plate oor system. Flat plate is a slab of uniform
thickness resting on column supports. It is the most economical system to form, as it
requires only a wood deck on adequate shoring. It also provides for the least
structural depth and thus for minimal oor-to-oor height, which is a very impor-
tant cost consideration in the overall economy of a multistory building. Figure B6.1
in Appendix B shows a high-rise building with at plate oor system under
construction.
Table 6.1, which is Table 8.3.1.1 of the ACI Code, provides guidelines for the
selection of minimum thickness of slabs without interior beams or at plates as a
function of the clear span. The guidelines enable the designer to select a slab thick
enough to prevent excessive short- and long-term deections in the system. The
6.2 Flat Slabs and Plates 373
Table 6.1 Minimum slab thicknesses recommended by the ACI Code Table 8.3.1.1 (without
interior beams)
Without drop panels With drop panels
Interior Interior
Yield Exterior panels panels Exterior panels panels
strength, Without With edge Without With edge
fy (psi) edge beams beams edge beams beams
40,000 n n n n n n
33 36 36 36 40 40
60,000 n n n n n n
30 33 33 33 36 36
75,000 n n n n n n
28 31 31 31 34 34
Figure 6.3 Flat slab with drop panels and column capitals
Because loads must travel toward the columns, the available zone through which
shear forces must travel becomes smaller and smaller; thus, the shear stress
increases, reaching a maximum at or near the interface of the column and the
slab, as shown in Figure 6.5. The large shears also indicate a sharp change in the
moments that occur around the columns. Shears cause diagonal tensions in concrete
structures that are subject to exure, and because concrete is quite weak in resisting
tension, failure can result. The failure surface may be envisioned as a truncated
pyramid similar to the one shown in Figure 6.6. This phenomenon is known as
punching shear: The column punches through the slab, or, more precisely, the
slab fails and falls down around the column.
Diagonal tension
This type of failure (i.e., punching) can be quite catastrophic. Many spectacular
failures in the history of construction have happened due to punching shear.
The intensity of the shear stress and the resulting diagonal tension depends on the
cross-sectional area through which the shear forces must travel toward the column.
This cross-sectional area, or shear surface area, depends on two parameters:
the thickness of the slab around the column, and the cross-sectional size of the
column. So the selection of these dimensions plays a very important role in
the preliminary design of the system. The column size is also inuenced by the
loads and moments that the column must resist at the oor level under consider-
ation. In this discussion, however, the column size is considered only from the point
of view of the punching shear in the slab.
Figure 6.7 shows a plan view of a typical interior column. The ACI Code
approach is based on a simple analytical model. It assumes that the critical shear
surface lies at a d/2 distance from the face of the column, where d is the effective
depth of the slab. The shear surface area then is the length of the critical periphery
(or perimeter of the critical section) multiplied by d.
d /2
c1
d /2
Critical shear
periphery
c2
Column section
The ACI Code gives the maximum factored shear to be transferred by stresses on
the concrete from the slab to the column as the smallest of Equations (6.1), (6.2),
and (6.3) (ACI Code Table 22.6.5.2).
4 p0
V c 2 fc bo d 6:1
where is the ratio of the long side to the short side of the columns cross section,
and bo is the shear periphery (perimeter of the critical section for shear). is the
lightweight concrete factor ( 0.75 for all-lightweight, 0.85 for sand-light-
weight, and 1.0 for normal weight concrete). As in Chapter 4, 0.75 for
shear. Equation (6.1) is the governing formula when the columns cross section is
an elongated rectangle, with the ratio of longer side to shorter side greater than 2.
s d p
Vc 2 fc0 bo d 6:2
bo
6.3 Shears in Flat Slabs and Plates 377
where s is 40 for interior columns, 30 for edge columns, and 20 for corner
columns.
p
Vc 4 fc0 bo d 6:3
Figures 6.8 and 6.9 show the shear periphery for a corner column and for an edge
column, respectively.
Experience has shown that vertical chases, ducts, pipes, and so on are somehow
attracted to columns. Although structures can tolerate openings near columns, open-
ings reduce the available shear periphery. Figure 6.10 shows some examples of openings
near a column and the resulting reduction in the effectiveness of the shear transfer.
Critical shear d /2
c1
periphery
d /2
c2
d
Figure 6.8 Denition of the shear periphery at a corner column, bo c1 c2 2
2
Critical shear d /2
c1
periphery
d /2
c2
d
Figure 6.9 Denition of the shear periphery at an edge column, bo c1 2c2 4
2
Ineffective
Assume free
edge
Critical shear
periphery
Ineffective
Ineffective
Example 6.1 For a oor structure with typical 24 ft by 26 ft bay sizes, the
superimposed dead loads are 20 psf and the superimposed live loads are 100 psf.
As a preliminary design, select an appropriate at slab thickness
(without drop panels) and a column size as governed by punching shear. Assume
fc0 4,000 psi and fy 60,000 psi. The concrete is normal weight with a unit
weight of 150 pcf.
Solution
The slab thickness will be governed by the longer span. Because we do not yet
know the column size, assume 18 in. 18 in. square. Even if we under- or
overestimate the column size, the error will have little effect on the slab thickness
selection. Thus, the larger net (clear) span is:
26 12 18 294 in:
Pu qu A 334 24 ft 26 ft 208,416 lb
This value is larger than the actual shear that must be transmitted through the
critical shear periphery because the loads within the periphery do not contribute to
it. We will use this approximate value, however, for the factored shear as well as to
estimate the size of the required column.
Thus:
Vu 208,416 lb
Hence:
bo 129 in:
Vu 2
Vu 2
B
Vu1
vuM
B Mu
Vu 2
Vu 2
Vu 1
Critical sections
Figure 6.11 Model of moment transfer from slab to column via shears
380 6 Floor Systems
model. These shears then must be combined with those from the gravity loads that
were discussed above.
Example 6.2 Assume that it is necessary to transfer a factored moment of
Mu 120 kip-ft between a at slab and an interior column. The moment acts
clockwise, as shown on Figure 6.11.
The column is 20 in. 20 in. and d 8.5 in. for the slab. Calculate the factored
shear stress due to the moment.
Solution
Per the ACI Code, the shears will transfer 40 % of the moment. Thus, the shears will
be responsible for an Mu 0.4 120 48 kip-ft.
Designating the maximum shear as vuM, then:
28:5 2 28:5
48 12,000 2 242:25vuM 4 60:56vuM
2 3 2
576,000 9,205vuM ! vuM 62:6 psi
If the shear stresses exceed what the ACI Code allows, we can reinforce the
column/slab interface. The reinforcing may be a shear head manufactured from
crossing steel shapes, or sets of closed stirrups. The last two decades have witnessed
the development of proprietary premanufactured shear reinforcement.
Flexure in at slabs and plates is a very complex problem. The simple representa-
tion of the deformation shown in Figure 6.4 does not truly describe the deections,
which have a rather intricate topography. But the magnitude of bending moments in
any direction is related to the slope of the deection curve, so the largest moments
occur where the curvature of the deection surface is greatest. Because the surface
curves in all directions, bending moments will occur in all directions at any location
on the slab. Moments in any direction can be represented by their component
moments in a preselected coordinate system, so the ACI Code uses a design method
based on a simple model that is easy to visualize and proven to be safe. (Refer to
Figure 6.12 in the following discussion of this model).
6.4 Flexure in Flat Slabs and Plates 381
Column strip
Zone Zone
B Zone B
C
Middle strip
Zone Zone Zone
1
C A C
Column strip
Zone Zone
B Zone B
C
Figure 6.12 Denition of column strips and middle strips in at plates and slabs
Each slab bay is divided into strips in both directions, as shown in Figure 6.12.
By studying the deection pattern shown in Figure 6.4, we easily understand how
exures occur in the structure. In the zone where two middle strips cross (Zone A),
the slab bends downward in both directions; thus, there will be tensions in the
bottom in both directions (positive moment regions). Where two column strips
cross (Zone B), the slab bends upward in both directions, generating tensions at
the top in both directions (negative moment regions). Where a middle strip crosses
a column strip (Zone C), the slab bends downward in the direction of the column
strip, but bends upward in the direction of the middle strip; thus, there will be
positive moments in the column strips direction and negative moments in the
middle strips direction.
The ACI Code suggests two methods for the exural analysis of at slabs and
plates. The rst (and simpler) is called the direct design method; the second is
called the equivalent frame method. These methods are not exclusively for at slabs
or at plates. They may also be used when beams exist on the column lines, which
are commonly known as two-way slabs on beams. In this chapter we discuss only
the direct design method. (Discussion of the equivalent frame method is beyond the
scope of this book. The interested reader is referred to Section 8.11.2 of the ACI
Code).
The direct design method may be used only when the plan geometry conforms to
the following set of limitations (ACI Code, Section 8.10.2)
382 6 Floor Systems
qu 2 2n
Mo
8
where
qu the factored load on a unit area (psf);
n the clear (net) span length in the direction for which moments are
being determined;
2 the length of span (center to center) transverse to n;
Step 2. Divide Mo into positive and negative moments.
The value of Mo is divided between total factored positive and negative
moments in the span under consideration. Figure 6.13 shows a schematic
moment diagram for a slab and the value of Mo. The values assigned are not
the result of theoretical studies, but rather observations from testing. They
appear to be safe and reasonable values given the highly indeterminate
nature of the problem. (A few percentage points of difference one way or
the other does not change the overall ultimate strength of the system).
n
Net span
Span
Table 6.2 Percent distribution of moments into positive moments and negative moments in an
end bay (ACI Code Table 8.10.4.2)
(1) (2) (3) (4) (5)
Slabs without
beams between
interior supports
Exterior Slab with beams Without With Exterior
edge between all edge edge edge fully
unrestrained supports beam beam restrained
Interior negative 0.75 0.70 0.70 0.70 0.65
factored moment
Positive factored 0.63 0.57 0.52 0.50 0.35
moment
Exterior negative 0 0.16 0.26 0.30 0.65
factored moment
Step 3. Divide positive and negative moments between column strips and middle
strips.
Now that we have determined the values of the positive and negative
moments across the full width (2) of the panel, we divide these moments to
the appropriate column strips and middle strips.
The calculations are slightly more involved when beams are incorpo-
rated into the oor system.
The ACI Code denes a coefcient, f, which is the ratio of exural
stiffness of a beam section to the exural stiffness of a width of slab
bounded laterally by the centerlines of adjacent panels on each side of the
beam. The coefcient f is calculated using Equation (6.4) (ACI Equation
8.10.2.7b).
Ecb I b
f 6:4
Ecs I s
where Ecb and Ecs are the modulus of elasticity of the concrete in the beam
and slab respectively (these two values are usually the same in cast-in-place
concrete construction); and Ib and Is are the moment of inertia of the gross
concrete section of the beam and slab, respectively. For at plates and
slabs, f 0.0.
384 6 Floor Systems
In Table 6.5, t is the ratio of the torsional stiffness of the beam to the
exural stiffness of a width of slab equal to the span length of the beam.
The term beam here refers to a T-section attached to a certain amount of
the slab that helps to increase the beams torsional stiffness. See Figure 6.14
for an illustration of the T-section.
The cross-sectional constant of the combined stem and attached slabs
may be evaluated from rectangular parts as given by Equation (6.5) (ACI
Code Equation 8.10.5.2b).
6.4 Flexure in Flat Slabs and Plates 385
hf
hf
hb
hb
bw bw
X
x x3 y
C 1 0:63 6:5
y 3
where x and y are the shorter and longer overall dimensions, respectively,
of the rectangular part of the cross section. The value for t can then be
calculated using Equation (6.6) (ACI Equation 8.10.5.2a).
Ecb C
t 6:6
2Ecs I s
In typical cast-in-place concrete construction, Ecb Ecs.
Step 4. Determine reinforcement.
From the moments calculated in step 3, determine the reinforcement
required in the column strips and middle strips using the exural design
methods discussed in Chapter 2.
Example 6.3 Design the reinforcement for a typical interior bay of the at plate
oor system of Example 6.1. Assume that the columns are 20 in. 20 in. and that
the slabs thickness is 10 in. Use fc0 4,000 psi and fy 60,000 psi.
The solution shows detailed calculations for the longer span (26 ft)
direction only.
Solution
Step 1. Calculate Mo.
Using the coefcient from Table 6.4, the factored positive moment
assigned to the column strip is:
12,000 96:4
R 102 psi ! from Table A2:6b ! 0:0019
144 8:872
then As 0:0019 144 8:87 2:43 in:2
From Table A2:9, select 13 #4 bars As 2:60 in:2 :
12,000 124:6
R 132 psi ! from Table A2:6b ! 0:0025
144 8:872
then As 0:0025 144 8:87 3:19 in:2
From Table A2:9, select 16 #4 bars As 3:20 in:2 :
12,000 83:1
R 88 psi ! from Table A2:6b ! 0:0017
144 8:872
then As 0:0017 144 8:87 2:17 in:2
From Table A2.9, select 11 #4 bars (As 2.20 in2). Figure 6.15 shows the
selected reinforcing in the different strip zones. Final results include the
required reinforcing in the 24-ft span direction as well. The resulting
moments in the short span strips are somewhat less (due to the shorter
span), but the reinforcing required is almost identical to that of the long
span strip. This reinforcement will be placed in a second layer, with the
working depth d estimated as only 8.12 in. The student is encouraged to
verify these reinforcing requirements. The placement order (i.e., which
layer of reinforcement must be laid rst and which layer onto the second
layer) must be clearly noted by the designer on the structural plans.
Example 6.4 Calculate the column and middle strip moments for an end bay of the
oor system in Example 6.3 with the addition of a 12 in. wide by 20 in. deep edge
beam. Figure 6.16 shows the slab divided into column and middle strips. As in
Example 6.3 only the 26 ft span direction calculations are shown.
Solution
26 ft.
13 #4 Top
13 #6 Top
12 ft.
11 #4 Bottom
16 #4 Bottom
24 ft. 12 ft.
13 #4 Top
13 #6 Top
12 ft.
Thus:
The total factored negative moment at the rst interior support is:
Step 3. Distribute the moments to the column strips and the middle strips.
(a) Negative moments at the rst interior support:
Because f1 0 (no beams in the span direction), from Table 6.3:
6.4 Flexure in Flat Slabs and Plates 389
26'-0"
12'-0"
12 in. 20 in.
edge beam
24'-0" 12'-0"
12'-0"
Figure 6.16 Column strips and middle strips in the exterior bay
10 in.
20 in.
12 in. 10 in.
20 12 10 10 10 5
yt 8:53 in, A 20 12 10 10 340 in:2
20 12 10 10
203 12 103 10
Ib 340 8:532 10,595 in:4
3 3
24 12 103
Is 24,000 in:4
12 3
12 12 20 10 103 10
C 1 0:63 1 0:63 8,399 in:4
20 3 10 3
8,399
t 0:175
2 24,000
This value is very small due to the relatively small torsional stiff-
ness of the edge beam.
Flat plates are usually the most economical choice when spans are about 26 ft or
less. Beyond this length the slab thickness required to control deections becomes
too large, thus making the slab too heavy. Moments and shears are highest in the
areas around the columns. Hence, it makes eminent sense to increase the depth of
the plate in these critical areas, as shown in Figures 6.2 and 6.3. Although the use
of drop panels may be attractive from the standpoint of structural behavior, the
6.5 Flat Slabs and the Use of Drop Panels 391
associated forming costs are considerable. On the other hand, drop panels allow
thinner slabs to be used in most of the areas. This results in weight and concrete
savings that offset some of the excess forming costs. Figure B6.2 in Appendix B
shows the forming of a drop panel for a at slab oor system.
According to the ACI Code (Section 8.2.4), drop panels must extend at least
one-sixth of the span length in each direction from the column center line, and their
thickness must be at least 25 % of the slab thickness beyond them.
Example 6.5 Calculate the appropriate size of a at slab system for the oor in
Example 6.3. The columns are 20 in. 20 in. Figure 6.18 shows the plan layout
indicating the outlines of the drop panels.
Solution
The minimum plan dimension of the drop panel is
1 1
26 2 80800 by 24 2 80000
6 6
From Table 6.1 the recommended minimum slab thickness is:
n 26 12 20
hmin 8:11 in: select h 8:5 in:
36 36
8'-8"
8'-0"
24'-0"
26'-0"
Thus, the total thickness within the drop panel will be 8.5 + 2.5 11 in.
The total volume of concrete within one typical bay is:
8:5 2:5
24 26 8:67 8 456:5 ft3
12 12
Figure 6.19 shows a typical wafe slab or two-way joist oor structure. Wafe slab
oor structures are thick at plate structures, with the concrete removed from zones
where it is not required by strength considerations. Wafe slab provides economical
structures for spans up to 60 ft, in square bays, loaded with light and moderate
loads. The voids are formed by steel (or berglass) domes that are reusable, thus
very economical. These forms are available in standard sizes, as shown in
Figure 6.20, although wider or odd-shaped domes are also used to satisfy some
design objectives. The sides of the domes are tapered (usually 1 in 12) to permit
easy removal after the concrete has cured. The two-way joists, when carefully
nished after the removal of the forms, provide a pleasing appearance as well.
Figure B6.3 in Appendix B shows an exposed wafe slab oor system.
Depth of void
19 in. or 30 in.
24 in. or 36 in.
Standard depth:
6 in., 8 in., 10 in., 12 in.
for 19 in.-wide voids
8 in., 10 in., 12 in., 14 in., 16 in., 20 in.
for 30 in.-wide voids
The lips on the domes, when laid out side by side, form 5 in.-wide joists for the
19 in.-wide voids and 6 in.-wide joists for the 30 in.-wide voids. But the designer
does not have to work with 24 in. or 36 in. planning modules. Because the domes
are always laid out on a at plywood deck, the spacing between the domes can be
easily adjusted to make the joists wider than standard at the base. This accommo-
dates virtually any column spacing while maintaining a uniform appearance.
Leaving out the domes around the columns forms a shear head that provides
increased shear strength as well as concrete in the bottom for the high negative
moments. The slab over the domes is typically 34.5 in. thick, unless large
concentrated loads, increased re rating requirements, or embedded electrical
boxes and conduits warrant the use of a thicker slab. The 3 in. minimum is quite
adequate for roofs. The slab is reinforced with a light welded wire reinforcement to
prevent shrinkage and temperature cracks. Figure B6.4 in Appendix B shows the
domes and reinforcements during the construction of a wafe slab oor system.
For overall depth selection, the span/depth ratios given in Table 6.1 under the
heading Without Drop Panels will result in a very serviceable structure. The solid
concrete area around the column ideally should approach the size required for drop
panels; in other words, it should extend about one-sixth of the span length measured
from the column centerline.
Example 6.6 Select an appropriate wafe slab oor structure for 36 ft 36 ft bays.
Columns are 20 in. 20 in. fy 60,000 psi
Solution
Select a 4.5 in.-thick slab, anticipating electrical conduits or junction boxes in the
slab Also, use the same depth structure in the end bays without edge beams.
From Table 6.1:
n 36 12 20
h 13:73 in:
30 30
Select 10 in.-deep pans and a 4.5 in. slab for a total structural depth of 14.5 in.
Figure 6.21 shows the resulting layout.
394 6 Floor Systems
Column strip
Middle strip
Column strip
Column strip Middle strip Column strip
Figure 6.21 Plan view of a wafe slab showing the dened column and middle strips
Each 30 in. 30 in. 10 in. dome displaces 4.92 ft3 of concrete (CRSI Design
Handbook, p. 11-1). A typical bay contains 128 domes. Thus, the total volume of
concrete in the bay is:
14:5
Volume 36 36 128 4:92 936:2 ft3
12
936:2
tavg 12 8:67 in:
36 36
The analysis of the system is very similar to that of at plates. When nding the
reinforcement take into account that, with the exception of the lled areas around
the columns, the slab is no longer solid, but rather a set of joists. Thus:
(a) In positive moment areas, divide the strip moment by the number of joists and
design the joists for that fraction. Figure 6.22 shows that joists in these areas
6.7 One-Way Joists 395
b 36 in.
d1 d2
are like T-beams, and the slab on top provides a wide compression ange. The
joists will have two layers of reinforcement, one for each joist direction; thus,
the working depth will be slightly less in one direction.
(b) In negative moment regions (where middle strips bend upwards over column
strips), the bottom width of the stem is in compression and the slab on the top
is in tension. Thus, distributed reinforcing is used over the width of the ange,
as shown in Figure 6.23.
(c) Shear and bending must be determined in the joists around the solid section
surrounding the column. On rare occasions, shear reinforcing is necessary in
the joists along a short distance.
d = 13.5 in.
b = 6 in.
Figure 6.24 shows a typical one-way joist system. One-way joists spanning between
beams are essentially closely spaced beam elements. The clear space between them
must not exceed 30 in. in order to qualify for the joist designation used by the ACI
Code. The forms are made of various materials, such as steel, berglass, ber board,
and corrugated cardboard, and are made with or without the edge lip, as shown in
Figure 6.25. Forms without the edge lip, however, tend to bulge sideways during
construction under the lateral pressure of the freshly placed concrete, and the
396 6 Floor Systems
20 in. or 30 in.
Lip
Standard depth:
8 in., 10 in., 12 in., 14 in., 16 in., 20 in.
resulting joist widths are uneven. Forms with square or tapered ends are also
available, as shown in Figure 6.26. The tapered ends provide increased shear
capacity as well as increased moment capacity at the negative moment regions.
Figure B6.5 in Appendix B shows a one-way joist oor system under construction.
One-way joist systems are often used when the bays are elongated (i.e., the
column spacing in one direction exceeds the spacing in the other direction by about
40 % or more). At such span ratios the advantage of two-way behavior is greatly
reduced, and it is more economical to use one-way systems in which beams span
between columns, and joists span between the beams. It is also more economical to
orient the beams in the shorter spans and the joists in the longer span. For ease of
forming, the selected depth of the beams is often equal to that of the joists. In order
to provide for the necessary shear and moment capacity, the beams are made
considerably wider than the faces of the columns into which they frame. These
wide and shallow beams are not as efcient as deeper beams, but the savings
achieved by the reduced forming cost more than make up for that. Beams deeper
than the joists, such as those shown in Figure 6.27, occupy additional ceiling space
and require additional forming cost. Figure B6.6 in Appendix B shows an exposed
one-way joist oor system.
The slab over the voids is typically 3 in. thick, unless large concentrated loads or
increased re rating requirements warrant the use of a thicker slab. The slab is
6.7 One-Way Joists 397
2 in.
3'-0"
Wide beam: joists and beam are Beam is deeper than joists, more
of equal depth, simple forming complicated forming
Table 6.6 Recommended minimum span/depth ratios [ACI Code, Tables 7.3.1.1 and 9.3.1.1].
(minimum thickness of non-prestressed beams or one-way slabs unless deections are computed.)
Simply One end Both ends
Member supported continuous continuous Cantilever
Solid one-way Span/20 Span/24 Span/28 Span/10
slabs
Beams or joists Span/16 Span/18.5 Span/21 Span/8
bear on the walls when deecting. Furthermore, the deection of the supporting
beams should not exceed span/600 to ensure crack-free masonry walls.
If the depth must be minimized beyond the values listed in Table 6.6, the
designer may use Grade 40 ( fy 40,000 psi) reinforcement. This will result in
about 50 % more required reinforcement, but will reduce the strain in the
reinforcing steel in service load condition. Reduced strain in the reinforcement
provides reduced deection. Because using a different grade of reinforcement for a
few selected members on a project is not recommended, it is permissible to use
Grade 60 steel equal in cross-sectional area to the calculated amount of Grade
40 steel that would be necessary. Values shown in Table 6.6 should be used directly
for members with normal-weight concrete and Grade 60 reinforcement. For other
conditions the values should be modied as follows:
(a) For structural lightweight concrete that has a unit weight in the range of
90115 pcf, the values shall be multiplied by max(1.650.005wc, 1.09),
where wc is the unit weight in pcf.
(b) For fy other than 60,000 psi, the values shall be multiplied by (0.4 +
fy/100,000).
Figure 6.28 shows a typical beam and one-way slab system. A beam and one-way
slab system is an economical choice when the bays are elongated and the
superimposed loads are large. The system is especially economical when the
structure is subject to large line loads such as heavy partitions. Large openings
through the slab can be easily accommodated virtually anywhere in the oor. Beam
and one-way slab systems have a larger structural depth than do the other oor
systems, and their forming cost is usually higher. These disadvantages are
Figure 6.29 shows a typical two-way slab system. When the aspect ratio (the ratio of
the longer span to the shorter span) of a slab that is supported on all four sides is less
than about 1.50, the slab exhibits a signicant two-way behavior. As discussed in
detail in Chapter 2, this means that the slab will carry the loads in both directions. In
plan, the load distribution from the slab to the beams may be approximated, as shown
on Figure 6.30. The shorter beam supports much less load than the longer span does.
So if the aspect ratio is signicantly larger than 1.0, the use of stronger beams in the
long direction than in the short direction is recommended. As the aspect ratio
approaches 1.0, the load division between the beams is more evenly distributed.
Area loading
long beam
Area loading
short beam
The ACI Code provides recommendations (Section 8.3.1.2) for the minimum
thickness of slabs supported on all four sides. These ACI Code formulae are
somewhat cumbersome. In the experience of the authors, h n/35 to n/40 is a
reasonable value for preliminary design (n is the longer clear span of the slab from
face of beam to face of beam). The preliminary selection of beams is governed by
considerations similar to those for one-way slabs and beams.
Figure 6.31 shows a typical two-way joist, or wafe slab, with slab-band beams
system. This oor system is an interesting variation of the two-way slabs on beams.
Wide beams form a two-way grid of beams (often referred to as slab bands)
between the columns. The depth of the beams is equal to the depth of the two-
way joist system. This arrangement provides a somewhat easier layout of reinforce-
ment in the negative moment regions around the columns. This system may also
have a seismic performance better than that of ordinary wafe slabs.
Problems 401
Problems
(a) Determine an appropriate slab thickness for an interior panel. Round the
thickness to the nearest 1/2 in.
24 in. 24 in.
Column
(typical)
25'-0"
25'-0"
25'-0" 25'-0"
Plan of Problem 6-6
(b) Check the shear around a typical interior column. Assume #6 bars in both
direction and use the average d when calculating the shear strength.
Assume no moment transfer between the column and the slab.
(c) Calculate the required outer layer reinforcing for (1) positive moment in a
column strip, (2) positive moment in a middle strip, (3) negative moment
in a column strip, and (4) negative moment in a middle strip.
Self-Experiments
Experiment 1
Make small-scale reinforced concrete models of a one-way joist, a wafe slab, and
a at plate oor system. Place wires for their reinforcements. Record the procedure
and your observations. Which system required the least effort in building the forms
and placing the reinforcement?
Experiment 2
Identify three different concrete oor systems from local buildings. Record the
range of the spans and the bay shapes (square, rectangular, etc.). In addition, record
their occupancy types. Write a report that summarizes your ndings and includes
photos.
Chapter 7
Foundations and Earth Supporting Walls
7.1 Introduction
Any building structure requires a foundation system in order to transfer the loads to
the supporting soil. The strength of concrete typically is 400800 kip per square feet
(ksf). Soils typically however, can safely withstand only pressures of 310 ksf. As a
result the foundation system has to spread the load over a large surface area to
reduce the pressure when it transfers loads from columns and walls to the
supporting soil.
Foundations were constructed of stone and masonry before concrete was used as
a building material. Application of concrete has improved the foundation system
signicantly. Today, virtually all foundations are made of plain or reinforced
concrete.
Because the design of foundations requires an understanding of the soil-structure
interaction, we must study the different types of soil and their behavior under
loading. Therefore, the following sections present an overview of the different
types of soil, their classications, the exploration methods, and the laboratory
tests for nding the allowable bearing capacity of soil.
Subsequently, this chapter deals with the different types of foundation systems,
with a focus on the design and analysis of wall and column footings. The last part of
this chapter discusses the different types of earth supporting walls, with an empha-
sis on basement walls and cantilever retaining walls.
In general, all subsurface materials fall into one of two groups: rock or soil. But in
reality soils are made up mostly from rock eroded by air and water and settled over
many millennia. Soils are divided into two main categories: coarse-grained soils
and ne-grained soils. Coarse-grained soils consist mainly of gravel and sand. The
particle sizes are large enough to be seen with the naked eye. Coarse-grained soils
are also called noncohesive soils, as their grains do not stick to each other when
oven-dried. Fine-grained soils are classied as clay or silt. A magnifying glass is
needed in order to see their particles. Fine-grained soils are also called cohesive
soils because their particles stick to each other. Cohesive soils expand when
subjected to moisture and shrink when dried.
In addition to these major categories of soil, soils are classied as organic or
inorganic soils. Organic soil consists of decayed vegetable or animal remains. The
top soil used to grow plants and vegetations is an organic soil. Inorganic soil, in
contrast, is almost completely free of organic materials. Organic soil is not suitable
for supporting building structures or even to be used as backll against basement or
retaining walls. If it is encountered in a construction site, it must be replaced with
appropriate compacted engineered ll.
A soil classication called the Unied Soil Classication System (USCS) has been
devised to specify the soil mix and its condition. It is based on the work of Professor
Arthur Casagrande at Harvard University. Each designation in this classication
consists of two letters. The rst letter represents the type of soil: G (gravel), S
(sand), M (silt), C (clay), O (organic), and P (peat). The second letter shows the soil
condition, for example, W (well-graded) or P (poorly graded). In this classication,
soils are divided into 15 types, as shown in Table 7.1. A well-graded soil consists of
both large and small grains, with the small particles lling the voids between the
large ones (sand and gravel). Well graded and compacted sand and gravel are very
good substrata. The poorly graded soil refers to a soil that does not have the right
proportioning of sand and gravel and, as a result, has large voids between adjacent
grains.
2030 ft below the foundation level, with one deep boring to search for hidden
weak deposits. For high-rise buildings the spacing of borings is closer, around
4050 ft, and the depth often descends to the underlying bedrock.
Boring is performed by an auger drill. A hollow pipe, called the casing, is advanced
to prevent the soil from collapsing into the borehole. As the bore hole is advanced, the
soil is tested in situ at certain locations. This testing is usually performed wherever
the driller experiences a different stratum, or at 5-ft intervals within the same stratum.
The test used most often is the Standard Penetration Test (SPT).
The SPT uses a device called a split-barrel sampler. Figure 7.1 shows a schematic
drawing of a split-barrel sampler. It is a hollow cylinder 2 in. in diameter, made up of
two tting half cylinders, which are held together by two threaded end-pieces. It is
placed at the bottom of the bore hole and driven through the soil by a 140-lb hammer
that has a free fall of 30 in. The number of blows needed to move the sampler three
times 6 in. into the soil is recorded. The numbers of the blows from the second and
third 6-in. advancements are added up. This sum gives the so-called N value. The
blows from the rst 6-in. advancement do not reveal the true characteristics of the soil
in situ, because the auger tends to leave disturbed soil at the bottom of the hole.
In addition, when the sampler is withdrawn at the end of the test, the device is
taken apart and the soil sample contained in the cylinder (part B on Figure 7.1) is
placed into a sealed and labeled jar. The sample is then taken to the laboratory for
further testing.
The blow count, N, is related to the soil condition. Table 7.2 shows a general
classication relating soil condition and the blow count. After careful laboratory
analysis of the samples, the geotechnical engineer prepares a boring log of each
boring performed at the site. Figure 7.2 shows a sample boring log.
406 7 Foundations and Earth Supporting Walls
F
E
C D
G A
Tube Ball Vent
(2 at 3 8 in.
B diameter)
A 1.0 to 2.0 in. (25 to 50 mm)
B 18.0 to 30.0 in. (0.457 to 0.762 m)
C 1.375 0.005 in. (34.93 0.13 mm)
D 1.50 0.05 0.00 in. (38.1 1.3 0.0 mm)
E 0.10 0.02 in. (2.54 0.25 mm)
F 2.00 0.05 0.00 in. (50.8 1.3 0.0 mm)
G 16.0 to 23.0
The 1 1 2 in. (38 mm) inside diameter split barrel may be used with a 16-gage wall thickness split
liner.The penetrating end of the drive shoe may be slightly rounded. Metal or plastic retainers
may be used to retain soil samples.
Figure 7.1 Schematic drawing of a split-barrel sampler for the Standard Penetration Test
(copyright ASTM International. Reprinted with permission)
Soil, like any other material, has a certain load bearing capacity. If the pressure
from the footing exceeds this limit, the soil will fail. This would cause the footing
to sink into the soil, which may have disastrous consequences on a supported
building.
Figure 7.3 shows a simple theoretical failure mechanism for soils under pressure.
A wedge is formed directly under the footing (Zone I). This wedge is pushed down
by the footing, which in turn pushes Zone II outward. Zone II rotates about a pivot
at the top and pushes Zone III sideways and up. The bottom parts of Zones II and III
7.5 Soil Failure Under Footings 407
SAMPLE DISTANCE
PLASTIC WATER LIQUID
LIMIT % CONTENT % LIMIT %
DEPTH (FT)
RECOVERY
10 20 30 40 50
LBS/F T 3
STANDARD
PENETRATION BLOWS/(FT)
SURFACE ELEVATION 10 20 30 40 50
PA
15.0
13
5 SS
PA
20.0 5/6"
6 SS Seams of clay in Sample 6.
21.0
6A SS Silty clay, trace fine gravel and sand - gray - very 7/6"
25.0 14
7 SS
PA
14
8 SS
30.0 30.0
End of Boring * Calibrated Penetrometer
Borehole backfilled with chips and cuttings; patch
pavement
Automatic-Diedrich Hammer used for Standard
Penetration Tests
The stratification lines represent the approximate boundary lines between soil types: In situ, the transition may be gradual.
WL BORING STARTED
8.0 ft WS 3/1/06
WL BORING COMPLETED ENTERED BY SHEET NO. OF
5.0 ft AB 3/1/06 1 1
WL RIG/FOREMAN APPD BY JOB NO.
form a shear plane along which the wedges move. This plane provides shear
resistance against the movement. The weight of Zone III also provides resistance
against the rotation of Zone II. Thus, placing loads on Zone III (surcharge) or
moving the footing deeper into the ground will inhibit the movement of the wedges.
As a result, the soil will have more bearing capacity.
408 7 Foundations and Earth Supporting Walls
Load
Footing
I
III III
II II
The pressure at the bottom of the footing propagates through the soil mass. The
pressure is most intense directly under the footing, and decreases at increasing
horizontal and vertical distances from the footing. This is known as the pressure
bulb effect as shown in Figure 7.4a for a circular footing. Figure 7.4b shows how the
pressure bulb extends in all directions like a balloon. The pressure bulb for a square
footing has a shape somewhat similar to the one shown in Figure 7.4. In the case of
a continuous strip footing, the pressure distribution extends along the footing with
proportions that are cylindrical rather than spherical.
Soils compress into a smaller volume when subjected to pressure. This leads to
settlement, or a downward movement of the footings. The amount of settling
depends on several factors such as the pressure level under the footing, the size
of the footing, and the properties of the soil. The volume of soil affected by the
footing is basically the limit of the pressure bulb shown in Figure 7.4. Therefore,
settlement also depends on the shape and size of the footing. Settlement can never
be completely eliminated unless the footing is directly supported by bedrock. So we
design footings to limit the detrimental effects of settlement on the structure.
Soils are composed of three major constituents: solid particles, a system of voids
between these particles, and air (gas) or water that lls the voids. When the soil is
loose, it has a large void content.
Settlement or consolidation in soils is associated with the squeezing of the
moisture or air out of the voids. This permits the solid particles to move closer to
each other, resulting in a denser structure.
In granular soils the movement of water (or smaller particles) is easy. Thus,
settlement of these soils will take place quickly as the structure is built on top of
them and loads are added onto the footings. About 90 % of the expected settling will
have taken place by the time the building is completed.
In cohesive soils (clays), the movement of moisture is slow, thus the consolida-
tion (settlement) of the soil is also slow. Settlement may take place long after the
completion and occupancy of the building.
7.6 Pressure Distribution Under Footing and Soil Settlement 409
b W
q
0 0.5W 0.75W W
0.9q
0.8q
0.7q
0.6q 0.5W
0.5q
0.4q
0.3q
1.0W
0.2q
0.15q
1.5W
0.1q
2.0W
0.05q
2.5W
Figure 7.4 (a) Isometric view of pressure bulb under column footing; (b) graph of pressure bulb
under column footing
410 7 Foundations and Earth Supporting Walls
Because differential settlement can cause severe distortion and structural problems,
various methods are used to reduce its effect. One common method is to design the
footings so that each applies approximately the same pressure on the soil under the
most usual loads. These loads consist of dead loads and an average percentage of
the live load depending on the occupancy type.
There is a limit to the pressure that soils can safely support. This limit is called
allowable soil bearing pressure or simply soil bearing capacity, and is based on
two criteria: (1) the soil does not fail, and (2) the settlement is not excessive.
Building codes recommend soil bearing capacities for specic conditions. These
recommended values are generally approximate. Table 7.3 shows approximate soil
bearing capacities for each soil type. The proper method of establishing the
allowable soil bearing pressure is a soil investigation program conducted by a
qualied geotechnical engineer. Usually the structural engineer selects and designs
the foundation system based on a soil report from a geotechnical engineer.
For smaller projects, however, the structural engineer may use the presumptive
bearing capacities recommended by the local building codes such as those shown in
Table 7.3 when foundations are supported by soil whose properties are known.
Foundations fall into two main categories: shallow foundations and deep founda-
tions. Each of these foundation systems consists of different subsystems, as shown
in Figure 7.5. Many factors need to be considered in the selection of a foundation
system. These include soil strength, soil type, the location of the water table,
variation of the soil with depth, and so on. In general, foundations are constructed
of plain or reinforced concrete. The typical strength of concrete used in footings is
fc0 3,000 psi. In rare cases, concrete with higher strength may be used to reduce
the footing depth and weight.
Shallow foundations are usually located no more than 6 ft below the lowest nished
oor. These are the most economical and most common type of foundation. A
shallow foundation system generally is used when (1) the soil close to the surface of
the ground has sufcient strength, and (2) underlying weaker strata do not result in
undue settlement.
Figure 7.5 divides shallow foundations into ve major types: wall footings,
isolated column spread footings, combined footings, strap footings, and mat or
raft foundations.
Foundation Systems
Isolated
Wall Combined Strap Mat
spread
footing footing footing foundation
footing
Depending on the condition and the type of the supporting soil, shallow foundations
are cast either into a neat excavation in the soil or into wood forms. It is most economical
to cast the concrete into earth forms. But this is only possible with a cohesive soil,
such as clay, which remains stable during the concrete placement. If the soil is granular,
the concrete is cast using wood side forms. The forms are removed after the concrete
gains strength, and the area around the footing is backlled and compacted.
Wall Footings Wall footings support walls made of wood, masonry, or concrete.
They are made of plain or reinforced concrete and are continuous under the entire
length of the wall. Similar to slabs, structural analysis and design are performed on
a 1-ft-long strip of the footing (assuming the wall is evenly loaded). The supported
wall is usually placed at the center of the footing to avoid any eccentricity and
rotation of the footing. Figure 7.6a shows a typical wall footing.
a b
c
d
Figure 7.6 Different types of shallow foundations: (a) wall footing, (b) isolated spread footing,
(c) combined footing (rectangular), (d) combined footing (trapezoidal), (e) strap footing, and (f) mat
foundation
7.8 Types of Foundations 413
Isolated spread footings can be made of plain concrete if they are subjected only
to gravity loads and are not located in earthquake-prone areas. Plain concrete
footings are usually used only with light loads.
Combined Footings One footing may be used to support two columns when the
columns are close to each other and the isolated spread footings for one column
would overlap the other. Such a footing is called a combined footing. Figure 7.6c
shows a rectangular combined footing. If the column loads are signicantly differ-
ent, we may use a trapezoidal footing such as the one shown in Figure 7.6d. The
larger width of the footing will be closer to the column supporting the heavier load.
When designing a combined footing, it is important to size the footing so that it
exerts an approximately uniform pressure on the soil. To achieve this, the footing is
proportioned so that its centroid is at or near the resultant of the column loads.
Strap Footings A strap or cantilever footing is a special type of combined footing
that uses a strap or beam to connect the two footings together. The application of
this footing is similar to that of the combined footing. Strap footings may also be
useful when underground utility lines prevent the use of rectangular combined
footings. Figure 7.6e shows a typical strap footing. The strap acts as a cantilever
beam that partially resists the moment from the eccentrically loaded exterior footing.
This ensures that the soil pressure is uniform underneath the entire strap footing.
Mat Foundation A mat or raft foundation consists of a large and thick continuous
reinforced concrete slab that supports the entire building. This system is used when
the soil bearing capacity is low or column loads are heavy, resulting in more than
50 % of the building plan area being required for individual footings. An advantage
of a mat foundation is that it drastically reduces differential settlement between
columns. Mat foundations are usually made of heavily reinforced concrete slabs at
least 24 in. thick. Figure 7.6f shows a typical mat foundation.
The use of shallow foundations may not be economical or even possible if the soil
bearing capacity close to the surface is too low. Deep foundations are used in these
situations to transfer the loads to a strong layer, which may be located at a
signicant depth below the ground surface. The load is transferred through skin
friction and end bearing as shown in Figure 7.7.
There are two main types of deep foundations: piles and piers (also called
caissons), as shown in Figure 7.5.
Pile Foundations Piles usually have small cross-section sizes, ranging from 6 to
24 in., and capacities of up to 500 kip. They are made of treated timber, steel, or
concrete in different shapes. Piles typically are driven into the ground using pile
driving hammers. This process causes noise and vibration, which may disturb
sensitive adjacent structures, such as hospitals.
7.8 Types of Foundations 415
Skin friction
Hard soil/bedrock
End bearing
When a pile is driven into the soil, it displaces the soil that is in direct contact
with it. The soil around the pile becomes signicantly compacted and lateral
pressure on the pile increases. This results in friction forces between the soil and
the pile.
Timber piles have been used since ancient times. The piles used today are about
2535 ft long although it is possible to splice them for longer length. Timber piles
act mainly as friction piles because their end bearing is not signicant. These piles
have load capacities in the range of 3050 kip.
Steel and precast concrete piles are normally used to carry large loads. For long,
slender piles the end bearing on soil is insignicant compared to the resistance from
the skin friction. But if the piles are driven to the underlying bedrock, they act as
end-bearing piles because the end-bearing resistance contributes a large percentage
of the total resistance.
Piles are commonly used in groups with a pile cap connecting the tops of the
piles and providing a surface area for placing building columns. Figure 7.8a shows a
pile group supporting a column. The piles in a group should be separated far enough
that the load carrying action of each pile does not affect that of an adjacent one.
Typically, at least three piles are used in a group, but two piles are often acceptable
in certain conditions. Figure 7.8b shows typical layouts of piles in groups with their
associated pile caps.
It is common to use battered piles if the piles are to be subjected to large lateral
loads due to wind or earthquake loads. These piles resist the effects of lateral
loads through axial tension and compression forces. Figure 7.8c shows a pile
group with battered piles.
Auger-cast piles may be used to alleviate the noise and harmful vibrations
associated with pile driving. A long hollow stem screw is drilled into the ground.
The auger is withdrawn at the desired depth while, simultaneously, concrete is
pumped into the bottom through the hollow stem. The withdrawing auger removes
the soil and replaces it with a concrete column. The pile can be reinforced by
416 7 Foundations and Earth Supporting Walls
Column
Pile cap
Piles
b Pile (typical)
Pile cap
(typical)
Column (typical)
3 Piles 4 Piles
5 Piles
6 Piles
7 Piles
c
P
H
Figure 7.8 Pile foundation: (a) pile group, (b) typical pile layouts, and (c) battered piles
7.8 Types of Foundations 417
Column
Reinforcements
Shaft
Bell
lowering a wide ange steel section into the fresh concrete immediately after
withdrawal of the auger.
Pier Foundations Piers are typically made by using a large-diameter auger to drill a
round hole in the ground and placing concrete into the hole. The drilling process for
piers is much less noisy than pile driving. The shaft diameter is usually at least
36 in. This provides enough room to lower a person to inspect and test the soil
before placing the concrete. A protective steel cylinder (called a casing) is used to
prevent the collapse of the sides during drilling.
A drilled pier is sometimes called a caisson, which is French for box. A special
device is often used at the bottom of the pier to enlarge the base, creating a belled
caisson instead of a straight-shaft caisson. The main purpose of a bell is to increase
the bearing area of the caisson. Figure 7.9 shows a typical belled caisson with its
components. Bells can be made only in cohesive soils such as clay. It is common
practice when designing a caisson to ignore the skin friction between the shaft and
soil and use only the end bearing capacity.
Caissons support many tall buildings in Chicago. Each column of the John
Hancock Center is supported by a 140-ft-long caisson, which transfers the load to
the bedrock. The Willis Tower (old Sears Tower) sits on 114 caissons that are 6 ft in
diameter and over 100 ft long to reach the bedrock.
418 7 Foundations and Earth Supporting Walls
Several issues must be considered when selecting a footing type. These issues,
however, mainly affect the placement of shallow foundations.
Adjacent Property Lines Buildings often have columns or walls close to or right on
a property line. Typically building codes and legal considerations do not allow any
part of a footing to extend beyond the property line. A good way to avoid this
problem is by setting the buildings supporting elements away from the property
line and letting the supported structure cantilever to the legal limit. But if the design
demands supports at or near a property line, the structural designer may need to use
an elongated rectangular footing, a combined footing, or a strap footing.
Depth for Frost Penetration The moisture in the soil underneath a footing may
freeze during the cold season if the bottom of an exterior footing is located too close
to the ground surface. Water expands when it freezes, and the magnitude of the
expansion is about 10 %. The expansion takes place toward the least resistance,
which is usually upward. This phenomenon is called frost heaving. This can push
the footing upward, which in turn can distort and crack the footing and damage the
supported building structure.
This problem is simple to prevent. Exterior footings have to be placed below the
frost line, as shown in Figure 7.10a. The frost line is the distance measured from the
nished exterior grade to the bottom of the expected maximum depth of frost
penetration.
Historic data are available on the depth of the frost line in different locations. For
example, 42 in. is a safe depth in most parts of the Midwest. In the northern part of
the Great Lakes and in many northern states, exterior footings must be placed at
least 6072 in. below the exterior grade. The geotechnical engineer usually pro-
vides the necessary information relating to the local frost line. The map in
Figure 7.10b shows the variation of the frost line depth for different areas of the
United States, based on the values recommended in city building codes. (This map
should be used only for general information because it is not necessarily accurate
for specic localities.)
In heated buildings, this requirement is mandatory only for exterior footings, as
frost does not travel horizontally to affect the interior footings. An exception to this
is when the foundations are constructed in the winter, or are left unprotected from
freezing during cold spells. In those cases even the interior footings should be
placed below the frost line and protected by backlling or insulating blankets.
Different Elevations of Adjacent Footings If the elevations of two adjacent
footings are different, as shown in Figure 7.11, the pressure on the soil from the
upper footing may increase the pressure under the lower footing. Therefore, a limit
called the proximity line is placed on the slope of the line joining the footings when
placing adjacent footings at different elevations.
This slope should preferably be limited to 1:2 if the soil is mainly granular (sand,
gravel). If the footings are on good clays, however, the slope may be increased to
7.8 Types of Foundations 419
Exterior grade
Frost line
7'
6'
5'
4'
3'
2'
1'
0'
Figure 7.10 Frost line (minimum required depth to prevent frost penetration): (a) frost line
denition, and (b) maximum anticipated depth of freezing from the city building codes
Footing
Footing
Proximity line
Slope
close to 1:1. The designer should consult the geotechnical engineer regarding the
safe elevation difference between neighboring footings.
Presence of Expansive Soil Expansive soil is a type of clay that undergoes signicant
volumetric changes with moisture variations. For example, a vast area of the southern
United States is covered by a clay deposit known as the Yazoo clay, which is an
expansive soil. Foundations placed directly on expansive soil may experience large
upward pressures that could cause serious distortion and structural damage through-
out the building. The moisture variation affects only the top few feet of an expansive
soil; thus, footings are usually placed at an elevation below which the periodic
moisture variation is insignicant. Short drilled piers are also commonly used.
Presence of Organic Layers Construction on soils that have signicant organic
content or underlying layers of organic soils (e.g., peat, marl, etc.) can cause serious
problems. Organic matter is highly compressible and in a state of long-term
decomposition.
Two different strategies for dealing with organic soil are available to the designer.
One is completely removing the soil to the full extent of the organic layers and
replacing it with a so-called engineered backll. The second is to use a deep
foundation, usually piles. It is also advisable to design the lowest level, which is
normally just a slab on grade, as a structural oor. Otherwise, the slab may settle
unevenly, cracking and distorting attached nonstructural elements, such as partitions.
Footings apply pressure on their supporting soil. This pressure has to be limited to a
certain allowable level (soil bearing capacity). There is also a reaction from the soil
acting on the footing when it presses the soil.
The true theoretical distribution of the reaction pressure from the soil on the
footing depends on the type of supporting soil. Figure 7.12 shows typical pressure
a b c
Figure 7.12 Soil pressure distributions: (a) sandy soil, (b) clayey soil, and (c) design assumption
7.10 Design of Wall Footings 421
distributions for different types of soil. The pressures are larger under the center of
footing and smaller along the edges if the soil is sandy. This is because the sand
along the edges does not have good lateral support and can easily move laterally
(see Figure 7.12a). The shape of the theoretical distribution of pressures in clays is
shown in Figure 7.12b. In practice, an average uniform soil pressure distribution,
like the one shown in Figure 7.12c, is assumed. This is much simpler and the results
have been proven to provide adequate and safe designs.
Because wall footings are long continuous members, designers use a 1-ft-long strip
of wall (b0 12 in. and its footing to represent the whole length for design purposes.
Figure 7.13 illustrates this concept. Wall footings are made of plain concrete or
reinforced concrete. Plain concrete footings are commonly used to support light
loads such as residential construction.
b
"
1'-0
To design a plain concrete wall footing we need only to determine the depth and
width of the footing such that (1) the soil pressure beneath the footing is less
than the allowable value (bearing capacity), and (2) the bending and shear strength
of the concrete footing is adequate. Generally, the footing width is calculated to
satisfy the rst requirement, while the depth is computed to satisfy the second
requirement.
Plain Concrete Wall Footing Design We must perform the following steps in order
to design a plain concrete wall footing. They are summarized in the owchart of
Figure 7.17.
422 7 Foundations and Earth Supporting Walls
qs qa
wT wD wL wLr
wT
qs qa
b
wT
b
qa 7:1
In this equation wD, wL, wLr and wT are the unfactored dead, oor live,
roof live, and total loads, respectively. The footing width (b) is usually
rounded up to the nearest even inch.
Step 2. Estimate the footing thickness (h).
The rule of thumb for the thickness (h) of the plain concrete wall
footing is:
7.10 Design of Wall Footings 423
bt
h 7:2
2
where t is the wall thickness (see Figure 7.14). The thickness (h) is usually
rounded up to the nearest inch.
According to the ACI Code (Section 14.3.2.1) the wall footing has to
have a minimum thickness of 8 in. In practice, we use this value and the
thickness of the supported wall, whichever is larger. The footing has to be
at least as wide as the wall thickness. In addition, the geotechnical report
often states a minimum acceptable footing width, usually at least 16 in.
Figure 7.15 summarizes these requirements.
h max {t , 8 in.}
b t
Check the footing dimensions against these minimum values. If they are
smaller than the minimum, use the minimum width and thickness.
Step 3. Calculate and check the moment.
The bending moment is the critical factor in determining the required
thickness (h) for plain concrete wall footings. The critical sections for
moment (where moments are the largest) are based on the type of wall
being supported. According to the ACI Code (Section 14.4.3.2.1) the
critical section for moment for a masonry wall (more exible than a
concrete wall) is at a distance t/4 to the inside of the wall, as shown in
Figure 7.16a. The critical section for a poured concrete wall is at the face of
the wall, as shown in Figure 7.16b.
The weight of soil above the footing and the weight of the footing do not
cause any bending or shear in the footing. In this respect, a footing is like a
424 7 Foundations and Earth Supporting Walls
mattress lying at on a bed. It does not bend; but when you stand on it, you
notice how it deforms and bends. Similarly, the only loads that cause
bending and shear in the footing are the dead (wD) and the oor and roof
live (wL and wLr) loads. As mentioned in Chapter Two, Section 2.10,
wu 1.2wD + 1.6wL + 0.5wLr (if wL 1.83wLr) and wu 1.2wD + 1.6wLr +
1.0wL (if wL < 1.83wLr). Figure 7.16c shows the bending of the footing
subjected to the applied loads. The ultimate tensile stress is calculated using
factored loads. Therefore, the factored pressure acting from soil on the
footing, qu, is:
wu
qu 7:3
b
a t
b
Masonry wall Concrete wall
D D
c d
Mu
qu
D b 1'-0"
qu
b
Figure 7.16 Moment in wall footings: (a) critical section for masonry wall footing, (b) critical
section for concrete wall footing, (c) bending of wall footing, and (d) moment at the critical section
7.10 Design of Wall Footings 425
Figure 7.16d shows the moment at the critical section, which can be
found from the equilibrium of the sum of moments:
D D2
Mu qu D qu 7:4
2 2
The footing is constructed of plain concrete, so there is no clear denition
of the effective depth (d). But ACI Code (Section 14.5.1.7) requires a
reduction of the overall footing thickness by 2 in. to allow for unevenness
of excavation and possible contamination of the concrete adjacent to the soil:
d h 2 in: 7:5
The moment, Mu, acts on a section (12 in. d). Therefore, the elastic
section modulus, Sm, is:
b0 d2 12d 2
Sm 7:6
6 6
2. b
t
Estimate footing thickness, h , and compare
2
the footing dimensions with the minimum sizes.
No
Mu MR ?
Yes
END
Figure 7.17 Flowchart for the design of plain concrete wall footings
426 7 Foundations and Earth Supporting Walls
MR Mn 7:8
MR Mu 7:9
The strength reduction factor for exure, compression, shear, and bearing
of structural plain concrete (ACI Code, Section 21.2.1) is:
0:60
Example 7.1 A 12 in. load-bearing CMU (concrete masonry unit) wall supports an
outdoor canopy. The wall will support a dead load of 10 kip/ft (including the weight
of the wall), and a roof live load of 5 kip/ft. Design a plain concrete footing for
this wall. The compressive strength of the concrete is 3,000 psi. The net bearing
capacity of the soil is 3,000 psf. The frost line is 4 ft from the grade. Concrete is
normal weight.
12 in.
CMU wall
4'-0"
b
7.10 Design of Wall Footings 427
Solution
Step 1. Determine the footing width.
wT wD wLr 10 5 15 k=ft
wT 15:0
Approximate footing width b 5 ft
qa 3:0
b t 60 12
h 24 in:
2 2
wu 1:2wD 1:6wLr
wu 1:210 1:65 20 k=ft
Determine the factored pressure from soil acting on the footing (qu):
wu
qu
b
20
qu 4:0 ksf
5
The critical section for moment, shown in Figure 7.18, is at t/4 from the
face of the CMU wall. Therefore, the distance, D, from the footing edge to
this location is:
bt t
D
2 4
5 12 12 12
D 27 in:
2 4
2
27
D2 12
Mu qu 4:0
2 2
Mu 10:13 ft-kip
d h 2 in: 24 in: 2 in: 22 in:
428 7 Foundations and Earth Supporting Walls
12 in.
D 27 in. wu
hest 24 in.
qu 4.0 ksf
b 5'-0"
Figure 7.18 Plain concrete wall footing of Example 7.1 (checking moment)
b0 d 2
Sm elastic section modulus
6
12222
Sm 968 in:3
6
Reinforced concrete wall footings usually support larger loads than do their plain
concrete counterparts. They are typically reinforced in the short direction.
7.10 Design of Wall Footings 429
12 in.
2'-0"
2'-0"
5'-0"
wT wD wL wLr
wT
Footing width b
qa
a b
d Vu
qu
'-0 "
C b 1
C
Figure 7.20 Shear in wall footings: (a) location of critical section, and (b) shear at the critical
section
According to the ACI Code (Sections 13.2.7.2), the critical section for
shear is located at a distance d from the face of the wall. Figure 7.20 shows
the critical section and the shear at the critical section. To calculate Vu we
cut the footing at this location and write the equilibrium of forces
(Figure 7.20b):
Vu qu C 7:12
where
bt
C d 7:13
2
7.10 Design of Wall Footings 431
Design of Reinforced
Concrete Wall Footings
No
Vu Vc ?
Yes
Figure 7.21 Flowchart for the design of reinforced concrete wall footings
The nominal shear strength of concrete, Vc, (ACI Equation 22.5.5.1) is:
p
Vc 2 fc0 b0 d
7:14
b0 12 in:
0:75
432 7 Foundations and Earth Supporting Walls
Vc Vu 7:15
qu D2
Mu
2
Using Tables A2.5 through A2.7, we obtain the steel ratio and calculate
the required area of steel:
As b0 d 7:17
Check the bars development length by using Table A3.3 and the
applicable modication factors of Table A3.1. The bar length from
the critical section for moment to the edge of the footing has to be more
than d. Otherwise, the bars have to be hooked at their ends. Therefore,
Note that we can use Equation (3.64) and the corresponding modication
factors of Table A3.1 instead of Table A3.3 to obtain the bar development
length. This method usually results in a smaller required development
length.
Normally, 1 #4 or #5 longitudinal bar is used per foot of width as
distributor bars.
Example 7.2 Design the wall footing of Example 7.1 using reinforced concrete.
Assume fy 60,000 psi.
7.10 Design of Wall Footings 433
Solution
wT wD wLr 10 5 15 k=ft
wT
b
qa
15
b 5 ft
3
Calculate the effective depth of the footing. Assume #6 bars with 3 in.
minimum clear cover. Therefore,
The critical section for shear is located at the distance d from the face of
wall, as shown in Figure 7.22a; from Equation (7.13):
bt
C d
2
512 12
C 8:62
2
C 15:38 in: 1:28 ft
a 12 in.
3
wu
d 12 3 8 8.62 in.
C 15.38 in.
1'-0"
qu 4.0 ksf
b 5'-0"
D 27 in. wu
t/4 3 in.
1'-0"
qu 4.0 ksf
b 5'-0"
Figure 7.22 Reinforced concrete wall footing of Example 10.2: (a) check shear, and (b) check
moment
Vu qu C
Vu 4:01:28 5:12 kip
21:0 3,000128:62
Vc 0:75
1,000
Vc 8:50 kip > 5:12 kip
As 0:0018b0 h 0:00181212
As 0:26 in:2 =ft < 0:27 in:2 =ft
d 28 in:
2 3
6 3 fy 7
d 6
440 p t e s 7 5db 12 in:
fc0 c b K tr
db
where
cb K tr
2:5
db
0:625
cb concrete cover to bar center 3 3:313 in:
2
cb K tr 3:313 0
5:3 > 2:5 Use 2:5:
db 0:625
Square spread footings are the most common type of column footing. Regardless of
the material used for the column construction, reinforced concrete (or, rarely, plain
concrete) spread footings are used to support columns. Figure B7.1 of Appendix B
shows a large square spread footing under construction.
7.11 Reinforced Concrete Square Spread Footing Design 437
1'-0"
5 #4
4'-0"
(Distributor
#5 @ 13 in. bars)
1'-0"
3 in. clear cover
5'-0"
The steps for the design of reinforced concrete square spread footings are
summarized in the owchart of Figure 7.28 and are as follows:
Step 1. Calculate the required area and select the size of the footing.
As shown in Figure 7.24, the footing is sized such that the pressure on the
soil (qs) is less than the soil bearing capacity (qa). PD, PL, PLr, and PT
are the applied service dead, oor live, roof live and total loads,
respectively.
PT PD PL PLr
PT
qs qa
A
PD , PL , PLr
qs qa
b b
or
PT p
Arequired ! brequired Arequired 7:20
qa
where qa is the allowable soil pressure. The value for b is usually rounded
up to the nearest even inch.
Step 2. Estimate the footing thickness.
In a square or rectangular footing, a reasonable estimate of the required
thickness is about one-half of the larger overhanging (O.H.) length of the
footing. Therefore, for a square footing, the estimated required thickness is:
bt
h 0:5O:H: 0:5 7:21
2
PD is the total dead load. PL and PLr are the applied service oor and roof
live loads, respectively.
The factored pressure on the footing from the soil, qu, is:
Pu
qu 7:22
b2
Step 3a. Check the two-way (punching) shear.
Typically, the two-way (punching) shear (refer to Chapter 6) is the
controlling factor in determining the required thickness.
The critical sections for the two-way shear action are located at the
distance d/2 from the faces of the concrete column. For a steel column this
distance is measured from the midpoint between the face of column and
the edge of the base plate, as shown in Figure 7.25.
The footing bends in two perpendicular directions, so it requires rein-
forcement in the form of a grid. The average effective depth, d, may be
taken from the top of footing (which is in compression) to the location
between the two layers of bars as follows:
The concrete cover for footings is 3 in. Assuming #8 bars, the distance
d can be calculated as:
7.11 Reinforced Concrete Square Spread Footing Design 439
a
t
Concrete column
d/2
C d
b
t
Steel column
d/2
C d
Figure 7.25 Critical sections for square spread footings: (a) concrete column, and (b) steel column
a
B
Vu2
B
b
qu
b
Vu1
b
qu
C
Mu
b
qu
D
Figure 7.26 Shear forces and bending moments at the critical sections: (a) two-way shear, (b)
one-way shear, (c) bending moment
The total factored shear acting on the critical shear surface is:
Vu2 qu b2 B2 7:24
where
Vu1 qu bC 7:27
bt
C d 7:28
2
qu bD2
Mu qu bDD=2 7:31
2
The required resistance coefcient, R, is:
12,000Mu
R
bd2
Using the Tables A2.5 through A2.7, we obtain for R. The required area
of steel (As) then is:
As bd
Use Table A2.9 to select the number and size of bars. The minimum area
of steel is:
D 3 in: cover d
Note that we can use (3.64) and the corresponding modication factors
of Table A3.1 instead of using Table A3.3 to obtain the bar development
length. This method usually results in a smaller required development
length.
Step 5. Determine the required dowel bars.
The column at its base transfers the load to the footing on an area equal to
the columns cross-sectional area (Ag). This generates a bearing pressure
that the footing must resist.
The bearing capacity of the concrete at the column footprint, Nbearing, is
given by the ACI Code (Section 22.8.3.2):
N 1 0:85fc0 A1 7:32
r
A2
N 2 min 0:85fc0 Ag 0
, 20:85fc A1 7:33
A1
7.11 Reinforced Concrete Square Spread Footing Design 443
a
b A2 b2
A1 t2 t b
Column
Dowels
sc
Footing reinforcement
dc
Figure 7.27 Dowel reinforcements: (a) bearing areas A1 and A2, and (b) dowel bars between the
column and footing
where
0.65 (ACI Code, Section 21.2.1)
A1 column bearing area, which for a column directly bearing on
the footing is equal to Ag of the column
A2 area of the part of the footing that is geometrically similar to,
and concentric with the column bearing area, A1 (see
Figure 7.27a)
N1 bearing capacity of the column
N2 bearing capacity of the footing
444 7 Foundations and Earth Supporting Walls
3.
Calculate and check the shear:
Pu
qu
b2
No
Vc2 Vu 2 ?
Yes
3b. Check one-way shear:
Vu1 qubC
Vc1 2 l fc bd
= 1.0, 0.75
No
Vc1 Vu1?
Yes
4. Determine the required reinforcement:
5. Determine the qubD 2 12,000Mu
Mu ;R
required dowel 2 bd 2
bars. Use Tables A2.5 to A2.7 to find , As = bd.
Select the size and number of bars.
Check bars' development length using Tables A3.1 and A3.3 or Equation 3.64.
Figure 7.28 Flowchart for the design of reinforced concrete square spread footings
7.11 Reinforced Concrete Square Spread Footing Design 445
Pd Pu N bearing 7:35
( )
Pd
Asd max , 0:005 Ag 7:36
fy
The ACI Code, Section 16.3.4.1 requires a minimum amount of dowel rein-
forcement equal to 0.005Ag (Ag is the gross area of column) to transfer the loads
from the column to the footing. This is in the form of a minimum of four bars
placed at the corners of the column. A minimum development length for the
dowels in compression equal to dc is required. This minimum length has to be
provided from the column bearing area extending into the column and the footing,
as shown in Figure 7.27b. The dowels are commonly hooked and tied to the
footing main reinforcements for ease of construction. The dowels have to be lap
spliced in compression to the column reinforcement based on requirements given
in Chapter 3. The length of the dowel in the column is the larger of the compres-
sion lap splice for the dowel bars (sc), as given in Section 3.4.5 of Chapter 3, and
the development length for the compression reinforcements in the column (dc).
Use Table A3.6 to obtain the development length for compression bars, dc, and
Table A3.5 to obtain the applicable modication factors.
Example 7.3 Design a square reinforced concrete footing for the 16 in. square
interior concrete column shown below. The dead load is 200 kip, the oor live load
is 80 kip, and the roof live load is 20 kip. The allowable net soil pressure (bearing
capacity) is 3,500 psf. Use fc0 3,000 psi for the footing, fc0 4,000 psi for the
column, and fy 60,000 psi. Assume the column has 4 #7 bars. Concrete is normal
weight.
446 7 Foundations and Earth Supporting Walls
16 in.
Solution
Step 1. Determine the required size of the footing.
1 9 12 4 16
hest 24 in:
2 2
Pu
qu
b2
378
qu
9:332
qu 4:34 ksf
7.11 Reinforced Concrete Square Spread Footing Design 447
3'-0"
Vu2 24 in.
3'-0"
9'-4"
qu 4.34 ksf
9'-4"
The critical section for the two-way shear is at a distance d/2 from
the face of the concrete column. Therefore, one side of the critical
section, B, is:
B t d 16 20 36 in: 30000
The shear at the critical sections, Vu2, is shown in Figure 7.29, and is
calculated as follows:
Vu2 qu b2 B2
Vu2 4:34 9:332 32
Vu2 339 kip
4 p0 s d p p
Vc2 min 2 fc bo d, 2 fc0 bo d, 4 fc0 bo d
bo
16=16 1:0
s 40 column in the center of footing
bo 4B 4 36 144 in:
1:0
p 40 20
Vc2 min 2 4 1:0 3,000 14420=1,000, 2 1:0
144
p p
3,000 14420=1,000, 41:0 3,000 14420=1,000
Vc2 minf946 kip, 1,192 kip, 631 kipg 631 kip
Vc2 0:75 631 473 kip > 339 kip ok
bt
C d
2
9:33 12 16
C 20
2
C 28 in: 20 400 2:33 ft
If we cut the footing at the critical section, as shown in Figure 7.30, the
shear at this location, Vu1, is:
Vu1
b 9'-4"
qu 4.34 ksf (9.33 ft)
C 2'-4"
(2.33 ft)
Vu1 qu bC
Vu1 4:349:332:33
Vu1 94:5 kip
Therefore, the footing has enough capacity against the one-way shear.
Step 4. Determine the required reinforcement.
To calculate the required reinforcement, we rst calculate the bending
moment at the critical section for moment (the face of the column):
b t 9:33 12 16
D
2 2
D 48 in: 4:00 ft
Mu qu bD2 =2
Mu 4:349:334:02 =2
Mu 324 ft-kip
12,000Mu
R
bd 2
12,000324
R 87:0 psi
9:33 12202
Mu
qu 4.34 ksf
b
9'-4"
D
4'-0"
From Table A2.6a fc0 3,000 psi, fy 60,000 psi ! 0:0017
(conservatively) and the required area of steel, As, is:
As bd
As 0:00179:33 1220
As 3:8 in:2
As, min 0:0018 bh
As, min 0:00189:33 1224
As, min 4:84 in:2 > 3:8 in:2
As 4:84 in:2
d 48 in: #7 bars
As, required
From Table A3.1, d can be reduced by , therefore :
As, provided
3:8
d 48 33:8 in:
5:40
The bar length measured from the critical section for moment is:
r
A2
N 2 min 0:85fc0 Ag, 20:85fc A1 ; fc0 footing 3:0 ksi
0
A1
( " r#
112 112
N 2 min 0:65 0:853:016 16 ,
16 16
20:650:853:016 16 g
N 2 minf2,970, 848:6g 848:6 kip
N bearing minfN 1 , N 2 g 565:8 kip
Pu 378 kip < 565:8 kip
Use minimum area for dowels:
Asd 0:005 Ag 0:005 16 16 1:28 in:2
(Table A2.9 ! use 4 #6 (As 1.76 in.2)
It has to be noted that for practical purposes the dowel bar size is usually
selected to match the column main reinforcements, which for a
16 in. 16 in. column it is expected to be larger than #6 bars. We have,
however, selected #6 bars here for consistency and clarity in the solution.
The required development length in the footing from Table A3.6 (com-
pression bars) for fc0 3,000 psi is:
dc 17 in: #6 bars
This is the length of dowel to be extended in the footing. From Table A3.6
the development length for #7 main reinforcements in the column
0
(fc 4,000 psi) is dc 17 in:
The required lap splice length, sc, for #6 dowels in the column is
(Chapter 3 Section 3.4.5):
sc 0:0005 fy d b 12 in:
0:000560,0000:75 22:5 in: 23 in:
We use the larger of 17 and 23 in., which is 23 in. for the dowel extension in
the column.
Figure 7.32 shows the nal design of this footing.
452 7 Foundations and Earth Supporting Walls
16 in.
4 #6 Dowels
23 in.
Example 7.4 Design the square reinforced concrete footing shown in Figure 7.33
for the interior column of Example 5.3. The bearing capacity of the soil is 8,000 psf,
fc0 3;000 psi for the footing and fy 60,000 psi. Concrete is normal weight.
16 in.
Pu 621:9
qu
b2 8:02
qu 9:72 ksf
The critical sections, as shown in Figure 7.34, are at distanced d/2 from the
face of the column.
2'-8"
20 in
Vu2
2'-8"
8'-0"
qu 9.72 ksf
8'-0"
The nominal shear capacity of the concrete for the two-way action is:
4 p s d p p
Vc2 min 2 fc0 bo d, 2 fc0 bo d, 4 fc0 bo d
bo
16
1, s 40 column at the center of footing, 1:0
16
4 p
Vc2 min 2 1:0 3,000 128 16=1,000,
1
40 16 p
2 1:0 3,000 128 16=1,000,
128
p
41:0 3,000 128 16=1,000
qu 9:72 ksf
d 27 4 23 in:
B t d 16 23
39 in: 30 300 3:25 ft
bo 439 156 in:
7.11 Reinforced Concrete Square Spread Footing Design 455
3'-3"
27 in.
Vu2
3'-3"
8'-0"
qu 9.72 ksf
8'-0"
The new footing is deep enough, and we continue with checking the
one-way shear action.
Step 3b. One-way shear.
The critical section for the one-way shear, as shown in Figure 7.36, is at a
distance d(23 in.) from the face of the column.
8:0 12 16
C 23 17 in: 1:42 ft
2
456 7 Foundations and Earth Supporting Walls
16 in.
C 17 in. d 23 in.
27 in.
8'-0"
Vu1 qu bC 9:728:01:42
Vu1 110 kip
Vu1
b 8'-0"
qu 9.72 ksf
C 1.42 ft
16 in.
D 40 in.
8'-0"
Mu
b 8'-0"
qu 9.72 ksf
D 3.33 ft
From Table A2.6a fc0 3,000 psi, fy 60,000 psi ! 0:0020
As bd 0:00208:0 1223
As 4:42 in:2
As, min 0:0018 bh
As, min 0:00188:0 1227 4:67 in:2 > 4:42 in:2
d 48 in: #7 bars
As, required
From Table A3.1, d can be reduced by A to get the required
s, provided
length (req). Therefore,
4:42
req 48 44 in:
4:80
The bar length measured from the critical section for moment is:
Use Equation (3.64) to calculate the more accurate required bar develop-
ment length:
0:875
cb 3 3:44
2
K tr 0 conservatively
cb K tr 3:44
3:93 > 2:5 Use 2:5
db 0:875
2 3
6 3 fy t e s 7 As, required
6
d 4 p 7
db 12 in:
40 fc0 cb K tr 5 As, provided
db
3 60,000 1:01:01:0 4:67
p db 31:97 db
40 1:0 3,000 2:5 4:80
31:970:875 28 in: > 12 in:
D 3 37 in: > 28 in: The bar development length is adequate:
7.11 Reinforced Concrete Square Spread Footing Design 459
N 1 0:85fc0 A1
N 1 0:65 0:854:016 16 565:8 kip
r
A2
N 2 min 0:85fc0 Ag , 2 0:85fc0 A1
A1
( " r# )
96 96
N 2 min 0:65 0:853:016 16 , 20:650:853:016 16
16 16
dc 17 in:
As , required 1:28
dowel dc 17 13 in:
As , provided 1:76
Therefore, use 13 in. minimum dowel length in the footing. The compres-
sion lap splice in the column for fy 60,000 is (see Chapter 3):
sc 0:0005fy db 12 in:
0:000560,0000:75 22:5 in:
16 in.
23 in. min.
13 in. min. 4 #6 dowels
8'-0" 8'-0"
PT PD PL PLr
PT
Arequired
qa
If the footing area is b bs (b is the longer dimension, and bs the
shorter, as shown in Figure 7.41) and the side (bs) is limited to a certain
A
known value, then b . Round up b to the nearest even inch and
bs
calculate the footing contact area (A):
A b bs
Step 2. Estimate the footing thickness.
7.12 Rectangular Reinforced Concrete Footing 461
Pu
qu
A
d
t 2
B bs
which is the same equation (Equation (7.25)) as for square footings. (Refer
to the section on square footings for the denitions of the parameters.)
In order to satisfy the ACI Codes requirements, the shear capacity of
the concrete has to be greater than the applied shear force:
Vc2 Vu2
If this condition is not satised, we need to increase the footing depth, h,
and repeat the process.
Step 3b. Check one-way shear.
The one-way shear requirements for rectangular footings are also
similar to those for square footings. But, the rectangular shape of the
footing places the critical section for the one-way shear at the distance
d from the face of the column in the long direction, as shown in
Figure 7.42a. Therefore, the distance, C, from the edge of the footing to
the critical section is:
b t
C d
2
Figure 7.42b shows the applied loads at the critical section for shear,
Vu1, which can be calculated as:
Vu1 qu bs C
a
Critical section for one-way shear
b
t
Vu1
bs t
Column
qu
d
bs
C
C
b
Figure 7.42 One-way shear in rectangular reinforced concrete footings: (a) plan view, and (b) cut
at the critical section
7.12 Rectangular Reinforced Concrete Footing 463
a
Critical section for moment (long direction)
Ds
bs
Critical section for moment
(short direction)
Column
D
b
b c
Mus
Mu
qu
qu
b
bs
D
Ds
Figure 7.43 (a) Plan view of critical sections for moment; (b) moment at the critical section in the
long direction; (c) moment at the critical section in the short direction
and the moment at the critical section, Mu, shown in Figure 7.43b, is:
D
Mu qu bs D
2
D2
Mu qu bs
2
Use Tables A2.5 through A2.7 to obtain , and calculate the required
area of steel in the long direction:
As bs d
bs t
Ds
2
and the moment at the critical section, Mus, shown in Figure 7.43c is:
Ds
Mus qu b Ds
2
D2s
Mus qu b
2
The coefcient of resistance, R, can be calculated as follows:
12,000Mus 12,000Mus
R
bd2 b d 2
Using Tables A2.5 through A2.7, we obtain . Then, the required area of
steel in the short direction is:
As b d
2
of reinforcement in this band
1
bs
Column
bs /2 bs /2
b
2
s 7:37
1
where
No
Vc2 Vu2 ?
Yes
3b. Check one-way shear:
Vu1 qubsC
Vc1 2 l f c bsd
l 1.0, 0.75
No
Vc1 Vu1 ?
Yes
4. Determine the required
Long Direction Short Direction
reinforcement.
qubsD 2 qub D 2s
Mu Mus
2 2
12,000Mu 12,000Mus
R 5. Determine the required R
bsd 2 b d2
dowel bars.
From Tables A2.5 to A2.7 From Tables A2.5 to A2.7
find ; As b sd find ; As bd
Select the size and number of bars. Select the size and number of bars.
Check the bar development length. Check the bar development length.
2
Distribute s of
1
reinforcement over bs centered
on the footing.
Figure 7.45 Flowchart for the design of reinforced concrete rectangular spread footing
7.12 Rectangular Reinforced Concrete Footing 467
16 in.
4'-0"
A 50
b 8:33 ft
bs 6
A 8:33 6 50 ft2
Pu 1:2PD 1:6PLr
Pu 1:2100 1:675 240 kip
Pu
qu
A
240
qu
50
qu 4:8 ksf
d h 4 in: 18 4 14 in:
468 7 Foundations and Earth Supporting Walls
2.5 ft
Vu2
2.5 ft
18 in.
The length of one side of the critical section for two-way shear, B, is:
B t d 16 14 30 in: 2:5 ft
Figure 7.47 shows the forces acting at the critical two-way shear sections,
Vu2, which can be calculated as follows:
Vu2 qu A B2
h i
Vu2 4:8 50 2:52
Vu2 210 kip
Therefore, the footing thickness is adequate for the two-way shear action.
Step 3b. Check one-way shear.
The critical section for one-way shear, as shown in Figure 7.48, is at a
distance d from the face of the column in the long direction:
7.12 Rectangular Reinforced Concrete Footing 469
Vu1
qu 4.8 ksf
bs 6'-0"
C 2.33 ft
b t
C d
2
8:33 12 16
C 14 28 in: 2:33 ft
2
Vu1 qu bs C
Vu1 4:86:02:33
Vu1 67:1 kip
p
Vc1 2 fc0 bs d
p 6 1214
Vc1 21:0 3,000
1,000
Vc1 110:4 kip
Vc1 0:75110:4 82:8 kip > 67:1 kip ok
Long Direction
The location of the critical section for moment, shown in Figure 7.49, is:
b t 8:33 12 16
D 42 in: 3:50 ft
2 2
2
D
Mu qu bs
2
3:502
Mu 4:86 176:4 ft-kip
2
12,000176:4
R 150 psi
6 12142
470 7 Foundations and Earth Supporting Walls
Mu
qu 4.8 ksf
bs 6'-0"
D 3.50 ft
From Table A2.6a fc0 3,000 psi, fy 60,000 psi ! 0:0029
rounded up
As bs d
As 0:00296 1214
As 2:92 in:2
As, min 0:0018bs h
As, min 0:00186 1218
As, min 2:33 in:2 < 2:92 in:2
Short Direction
The location of the critical section for moment, shown in Figure 7.50, is:
bs t 6 12 16
Ds 28 in: 2:33 ft
2 2
Mus
qu 4.8 ksf
b 8.33 ft
Ds 2.33 ft
D2s
Mus qu b
2
2:332
Mus 4:88:33 109 ft-kip
2
12,000Mus
R
b d 2
12,000109
R 67 psi
8:33 12142
As b d 0:00138:33 1214
As 1:82 in:2
As, min 0:0018b h
As, min 0:00188:33 1218
As, min 3:24 in:2 > 1:82 in:2
b 8:33
1:39
bs 6
s As 0:848 6:7
Use 7 #6 bars in the 60 000 center bandwidth and one #6 bars on each
side. This results in 9 #6 bars (As 3.96 in.2).
Check the bars development length:
Bar spacing in the long direction
612 23 0:75
10:9 in:
6
472 7 Foundations and Earth Supporting Walls
From Table A3.2, because cover >0.75 in. and clear space >2(0.75 in.),
use condition A. From Table A3.3:
d 33 in: #6 bar
As , required
From Table A3.1, d can be reduced by :
As, provided
For the long direction:
2:92
d 33 31 in:
3:08
1:82
d 33 15:2 in:
3:96
0:75
cb 3 3:38 in:
2
K tr 0
cb K tr 3:38 in:
4:5 > 2:5 use 2:5
db 0:75
2 3
6 3 fy t e s 7 A
d 6 p
440 f 0 cb K tr 5 7 db s , required 12 in:
c As, provided
db
3 60,000 1:0 1:0 0:8 1:82
p db 12:08 db
40 1:0 3,000 2:5 3:96
12:08 0:75 9 in: < 25 in: ok
The bars are long enough:
7.12 Rectangular Reinforced Concrete Footing 473
N 1 0:85 fc0 A1
N 1 0:650:85 3 16 r 424:3 kip
16
A
N 2 min 0:85 fc0 Ag , 20:85 fc0 A1
2
( " A 1
r# )
72 72
N 2 min 0:65 0:85 3:0 16 16 , 2424:3
16 16
N 2 minf1,909, 848:6g 848:6 kip
N bearing minfN 1 , N 2 g min f424:3, 848:6g 424:3 kip
Pu 240 kip < 424:3 kip
Use the minimum area for dowels:
Asd 0:005Ag 0:00516 16 1:28 in:2
d 17 in:
As, required 1:28
dowel d 17 12:4 in:
As, provided 1:76
Use 13 in: minimum:
Figure 7.51 shows the nal design of the footing.
16 in.
4 #6 dowels
13 in. minimum #6 (typical)
7 #6
18 in.
3 in. clear cover
1 #6 #6 @ 12 in. c/c 1 #6
Basement walls and retaining walls are two common concrete (plain or reinforced)
structural systems. Sometimes, however, they are made of concrete masonry units.
These structural elements have to resist lateral soil pressure. Therefore, it is
important to understand the action of soil on them. This section briey explains
lateral soil pressure, then discusses the different aspects of the design and analysis
of basement and retaining walls.
Soil that is retained on one side of a wall is conned on the higher grade and
prevented from moving freely. Figure 7.52 shows a vertical section of a retaining
wall. A wedge-shaped part of the soil in this vertical cut is pulled downward by
gravity and tries to slide down along a plane of rupture. The wall, however, prevents
these downward and outward movements, resulting in a lateral pressure on the wall.
Frictional resistance occurs along the plane of rupture as the grains of the soil try
to slide by one another. If the soil has clay content, cohesion increases this sliding
resistance. The plane also supports part of the weight (W ) of the wedge. The
combination of the weight support and the sliding resistance results in the force R.
The earth pressure (E) on the back of the wall is the resultant of the soil friction
on the wall and the lateral earth pressure. The reaction to this force, E0 , acts on the
soil wedge, which is in equilibrium with the weight of the wedge (W ) and the
R force. The inclination of force E is due to the frictional resistance the wall offers
Plane of rupture
Soil wedge W
Retaining wall
E
h
E R
to the sliding wedge. If the wall surface is practically frictionless (e.g., smooth
waterproong on a basement wall), then the earth pressure on the wall is horizontal.
The magnitude and the distribution of the earth pressure on the back of the wall
depend on many factors. The most important of these are the type of the retained
soil (granular or cohesive), its moisture content, and the slope of the backll
(if any). In addition, there may be loads on the upper surface, called surcharge
loads. These are caused by stored materials, trafc, or permanent installations such
as neighboring building foundations. All these factors increase the gravity loads on
the sliding wedge, which in turn increase the pressure on the wall.
A detailed discussion of the theory of lateral earth pressure is beyond the scope
of this book, as it belongs to the eld of soil mechanics. Our intention is to
familiarize you sufciently with the results of the theory.
There are three different types of earth pressure, distinguished from each other
by their pressure coefcients. The rst type is earth pressure at rest. This is the
theoretical pressure on an essentially immovable object. The second type is the
so-called active earth pressure. This occurs when the wall moves ever so slightly.
This very slight movement activates the sliding resistance along the plane of
rupture, which in turn reduces the pressure on the back of the wall. The active
earth pressure, which occurs in the direction of wall movement, is signicantly less
than the at-rest earth pressure. The third type is the passive earth pressure. This
happens when the wall is moving against the soil.
This section of the text is concerned mainly with the active earth pressure. We use
Rankines theory here, which is the easiest and simplest theoretical solution for
calculating active earth pressure. The theory assumes that the rupture plane is a straight
line and the backll material is cohesionless. It also assumes that the frictional resis-
tance at the back of the wall is nonexistent (i.e., the wall is smooth). Thus, our primary
concern is with the horizontal component of the pressure that is exerted on the wall.
The theory assumes that the distribution of the pressure on the back of the wall is
triangular, as shown in Figure 7.53. In the absence of surcharge loads, the pressure
is zero at the top and increases linearly with depth.
pa Ka soilh
Level Backll, No Surcharge The pressure at any depth ( y) can be expressed as:
where
pa K a soil h 7:40
The value of Ka depends on the angle of internal friction within the soil (). This
relationship is:
K a tan 2
45 7:41
2
pmax Ka soilh
Figure 7.54 Horizontal component of the pressure distribution, sloping granular drained backll
pmax K a soil h
The Effect of Surcharge Any additional load surcharge atop the surface increases
the gravity force on the sliding wedge. This in turn increases the lateral pressure on
the back of the wall, as illustrated in Figure 7.55. The lateral pressure at any depth is
pa Kasoily, where the product soily is the weight of the soil above level y, so the
increased lateral pressure from a distributed surcharge load (wsc) will be
The second part of the equation represents the increased lateral pressure from the
surcharge, which is independent of the depth. The surcharge may also be concep-
tualized as having an additional height (hsc) of soil atop the nish surface. If we
express the surcharge with the unit weight of the soil as
wsc
pa Ka soil h
psc Kawsc
hsc
pa Ka soil h
So if the unit weight of a ctitious liquid is a Kasoil, and we substitute this value
into Equations (7.39) and (7.40) that express the lateral pressure, we obtain what is
referred to as equivalent uid pressure. Geotechnical engineers usually make their
recommendations regarding lateral pressures on walls in terms of the equivalent
uid density (a).
The unit weight of compacted granular backll is between 105 and 115 pcf.
From the average values of Ka the calculated equivalent uid density for granular
backlls is between 30 and 40 pcf, with a mean value of about 35 pcf.
Note that these values hold for soils that are well drained. Clay soils or saturated
soils may produce much higher pressure values. The designer should always
consult with a geotechnical engineer to verify the most likely equivalent uid
density prior to designing retaining structures.
Basement walls are earth retaining walls that are supported laterally by the rst
oor construction at their top and by the basement slab on grade at their bottom. In
addition, they are vertically supported on wall footings.
Figure 7.57 shows a schematic section through a basement wall. The wall will be
stable only after the rst oor construction is complete, so no backll (or only a
very limited height of backll) should be placed against the wall until after the rst
First floor
Backfill
Basement wall
Undisturbed earth
Drain tile
Footing
Crushed stone
oor is in place. The backll should be a compacted granular ll that will drain well
into the footing drain. The footing drain (drain tile) is made of perforated tiles or
plastic drain pipes. This drain tile is then connected into either the storm drains or a
sump pit out of which the water is pumped. This prevents water from accumulating
behind the basement wall, and thereby prevents the increase of the lateral pressure.
Basement walls are usually made of concrete (either reinforced or unreinforced).
In residential construction they are sometimes built using concrete masonry units
(CMU) and hence are called CMU walls. These also may be reinforced or
unreinforced.
Design of Basement Walls The structural behavior of a basement wall is similar to
that of a simply-supported one-way slab spanning vertically between the slab on
grade at the base and the rst oor at the top. Similar to slabs, only a 1-ft-long strip
of the wall, as shown in Figure 7.58, is considered in the design of these walls.
The minimum thickness commonly used for unreinforced concrete basement
walls is 10 in. It is difcult to properly consolidate the concrete within the forms for
thinner walls. In addition, basement walls have to be thick enough to provide width
for placement of members such as stud walls, brick veneer, and so on. In
unreinforced concrete walls, it is advisable to use vertical control joints at a
maximum spacing of 20 ft. The control joints prevent the random cracking of the
wall due to volumetric changes.
Unreinforced Concrete Basement Wall Design The steps to design unreinforced
(plain) concrete basement walls are as follows. They are summarized in
Figure 7.61:
1'-0"
a
R1 (V ) (M )
R1
0.577h
Mmax
R2
pmax
R2
h
E pmax Vmax R2
2
E 2E
R1 R2 Mmax (@0.577h from grade) 0.128 Eh
3 3
b wsc
p1 R1 (V ) (M )
R1
x
Mmax
R2
p2
R2
p1 Kawsc p2 p1 Ka soilh Vmax R2 Mmax (@ x from grade)
h x [h 2 (2p
E (p1 p2) 1 p 2) 3hp1x (p2 p1)x 2]
2 6h
h
R1 (2p1 p2) R2 E R1 x is found by solving:
6
(p2 p1)
x2 2p1x 2R1 0
h
Figure 7.59 Shear force and bending moment in basement wall with full height backll: (a)
without surcharge, and (b) with surcharge
482 7 Foundations and Earth Supporting Walls
a
R1 (V ) (M )
R1
h1
h x
Mmax
h2
E
R2
pmax
R2
p max h 2
E Vmax R2
2
Eh2
R1 Eh2 2 h2
3h Mmax h1 x at x h2
3h 3 3h
R2 E R1
b R1 (V ) (M )
R1
h1 wsc
p1
h x
Mmax
h2
E
R2
p2
R2
p1 Kawsc Vmax R2
h2 Mmax (@ x from grade)
E (p1 p2)
2
h 22 (2p1 p2) h 22 x2
R1 (2p1 p2)(h1 x) [3h2p1 x (p2 p1)]
6h 6h 6h 2
R2 E R1 x is found from solving:
(p 2 p1)
x2 2p1x 2R1 0
h2
Figure 7.60 Shear force and bending moment in basement wall with partial backll: (a) without
surcharge, and (b) with surcharge
1.
Calculate the maximum moment (Mu, max):
Use Figures 7.59 and 7. 60
Figure 7.61 Flowchart for the design of plain concrete basement walls
p
Mn 5 fc0 Sm
where is the light weight concrete factor. Typically, basement walls are
made of normal weight concrete ( 1:0). Sm is the elastic section modulus
of 1-ft-long wall, or:
bt2
Sm , b 12 in:
6
MR Mn Mu , max
Solving for t:
s
12,000Mu , max
t p 7:47
6 fc0
10'-0"
7'-0"
coefcient of active soil pressure, Ka 0.33. Consider two cases: (a) without a
surcharge; (b) with a surcharge of 100 psf acting on the backll. Use fc0 4;000 psi.
Solution
(a) Without Surcharge
Step 1 Calculate the maximum moment.
This is Case (a) on Figure 7.60, a basement wall with partial backll
and without surcharge. The lateral soil pressure and resulting maxi-
mum moments are:
R1
3'-0"
h1
10'-0"
7'-0"
h2
E
R2
pu 0.45 ksf
pu h2 0:45 7
Eu 1:58 kip=ft of wall
2 2
Eu h2 1:58 7
Ru1 0:37 kip=ft
3h 3 10
Ru2 Eu Ru1 1:58 0:37 1:21 kip=ft
r
h2
and the location of the maximum moment, x h2 :
3h
r
7:0
x 7:0 3:38 ft
3 10
Eu h 2 2
Mu , max h1 x
3h 3
1:58 7 2
Mu , max 3 3:38
3 10 3
Mu , max 1:94 ft-kip=ft
s
12,000Mu, max
t p
6 fc0
s
12,000 1:94
t p
61:0 4,000
t 7:8 in:
Use t 10 in: minimum wall thickness
pa a h 40 7 280 lb=ft2
Obtain the equations for the reactions and the maximum moment
from Figure 7.60:
pu2 pu1 2
x 2pu1 x 2Ru1 0
h2
0:503 0:053 2
x 20:053x 20:50 0
7:0
0:0643x2 0:106x 1:0 0
q
0:106 0:1062 40:0643 1:0
x
2 0:0643
x 3:20 ft
h22 x2
Mu , max 2 pu1 pu2 h1 x 3h2 pu1 xpu2 pu1
6h 6h2
7:02
Mu , max 2 0:053 0:503 3 3:20
610
3:202
37:0 0:053 3:20 0:503 0:053
67:0
Mu , max 2:46 ft-kip=ft
providing for adequate support for the rst oor construction, often results in much
thicker walls than would be required by strict structural considerations only.
Hence, the thickness of reinforced concrete basement walls is usually
preselected by the designer, and the wall is strengthened by providing the needed
amount of reinforcement. An absolute minimum thickness in a reinforced concrete
wall is 8 in. As with plain concrete basement walls, only exure needs to be
considered; the shear stresses in normal basement walls are never excessive.
The steps in the design are as follows and are summarized in the owchart of
Figure 7.64.
Design of Reinforced
Concrete Basement Wall
2.
Select the wall thickness t :
Minimum wall thickness 8 in.
Assume y 1.13 in. (#6 bars with 3/4 in. cover)
3.
Determine the required reinforcement.
(a) Design the vertical reinforcements:
12,000Mu, max
d t y Rreq
bd 2
Use Table A2.5 through A2.7 to find . (Check for min from Table A2.4)
As bd
Use Table A2.10 to find the size and spacing of the bars.
(b) Design the horizontal shrinkage and temperature
reinforcements:
Ash 0.002bt (#5 and smaller)
Ash 0.0025bt (otherwise)
Figure 7.64 Flowchart for the design of reinforced concrete basement walls
12,000Mu
Rreq
bd 2
Use the appropriate fc0 and fy in Tables A2.5 through A2.7 to obtain the
steel ratio, . Then the required area of vertical reinforcement, As is:
As bd
According to the ACI Code (Section 9.6.1.2), the minimum area of the
vertical exural reinforcements is:
( p )
3 fc0 200
As , min min bd max , bd 7:48
fy fy
The bar spacing for the vertical and horizontal reinforcements (ACI Code, Sections
11.7.2.1 and 11.7.3.1) is limited to:
For basement walls with a thickness of 10 in. or less, shrinkage and temperature
horizontal reinforcement typically is placed on only one face. For thicker walls, we
distribute the required horizontal reinforcing evenly between the inside and the
outside faces.
490 7 Foundations and Earth Supporting Walls
Example 7.7 Design the reinforced concrete basement wall shown in Figure 7.65.
The unit weight of backll is soil 115 pcf, and the coefcient of active soil
pressure is Ka 0.33. The surcharge on the backll is 150 psf. Use fc0 4;000 psi
and fy 60,000 psi.
Solution
Step 1. Calculate the maximum moment.
This basement wall is subjected to full backll with surcharge. This is
Case (b) in Figure 7.59.
The equivalent uid density and pressure are:
12'-6"
x 7.05 ft
12'-6"
Eu 5.75 kip/ft
Location of Mu, max
pu2 pu1 2
x 2pu1 x 2Ru1 0
h
0:84 0:08 2
x 20:08x 22:08 0
12:5
0:061x2 0:16x 4:16 0
q
0:16 0:162 4 0:061 4:16
x
2 0:061
x 7:05 ft
492 7 Foundations and Earth Supporting Walls
12,000Mu 12,0009:15
Rreq 194 psi
bd 2 126:872
req 0:0038
#5 @ 12 in.
vertical
12'-6"
#4 @ 12 in.
horizontal
t 8 in.
The behavior of retaining walls is very different from that of basement walls.
Basement walls are vertical simply-supported slabs bending between two supports
(i.e., the basement oor and the rst oor). Unlike basement walls, retaining walls
are not supported at the top. They must have substantial weight to prevent toppling
over from the earth pressure.
Figure 7.68 shows the acting forces on a retaining wall. In addition to the weight
(W ), the earth pressure (E) is applied to the back side of the wall. If the back of the
wall is smooth (i.e., frictionless), the E force is horizontal. The two forces, W and E,
are combined into the resultant (R).
The E force wants to overturn the wall, or to pivot it around the toe point. In
Figure 7.68 E exerts a counterclockwise moment on the toe. The W force (i.e., the
weight of the wall) wants to prevent the overturning by applying a clockwise
moment about the toe. As long as the resisting moment, Mr , of W is greater than the
overturning moment, Mot , from E, the wall will be stable. Another way to express
the same concept is that the wall is stable as long as the resultant force, R, intercepts
the base of the wall inside the bottom width, as shown in Figure 7.68. On the other
hand, the wall will tip over if the R force intercepts the base line outside the bottom
494 7 Foundations and Earth Supporting Walls
W
R W
E
R E
The force polygon
Toe
width. Figure 7.69 shows a typical gravity retaining wall, and the applied forces
(assuming a smooth wall), with their corresponding application locations.
F
c
h
Mot E 7:52
3
Mr Wc 7:53
where FSot is the factor of safety against overturning. Section 1807.2.3 of the
International Building Code (IBC 2015) requires a minimum factor of safety
against overturning of
A second failure mode besides overturning exists. The earth pressure (E) tries to
push the wall (from right to left in Figure 7.69) to make it slide along its base.
Resistance against sliding comes from two sources. The rst is frictional resistance
(F) at the bottom of the wall. The magnitude of this force is equal to the weight of
the wall (W ) multiplied by the coefcient of friction () between the two materials
(i.e., the wall and the soil). The larger the weight, the larger is the frictional
resistance. The second force that resists sliding is the passive earth pressure in the
front of the wall. The bottom of the retaining wall, as with any other footing, is
usually placed below the frost line. The ll at the front provides passive resistance,
which can be considerable when the ll is there. Sometimes, however, this ll in the
front is removed for one reason or another. Thus, its continuous presence is not a
given, and most designers disregard it.
To ensure safety against sliding:
F W FSs E 7:55
where
the coefcient of friction between the bottom of the wall and the soil
FSs the factor of safety against sliding. (The recommended minimum safety
factor against sliding is 1.5, per IBC 2015, Section 1807.2.3)
All retaining walls in essence are gravity walls, although only one type is
designated as such. They differ only in the way we provide the mass needed to
safely retain the soil at the upper elevation. These differences in design, however,
increase the diversity of structural behavior within the wall structures themselves.
Figure 7.70 shows various types of retaining walls, and discussed as follows.
1. Gravity walls These walls are constructed of plain concrete, stone, or brick
masonry. Their extensive use of material and labor costs limit their economy
to relatively low heights of about 8 ft above the low grade (Figure 7.70a).
2. Cantilever retaining walls These are by far the most common type of retaining
wall. They are constructed with reinforced concrete or reinforced masonry, and
are economical to use for heights up to about 20 ft. Figure 7.70b shows a typical
cantilever retaining wall. The backll above the heel provides much of the
weight needed for the stability of the wall. There are, however, variations of
these walls such as cantilever walls without heel or toe as shown in Figures 7.70c,
d. They are used when property lines or other limitations prevent the footing from
extending beyond one side of the wall. These walls are not as efcient as the
typical cantilever retaining walls.
3. Counterfort walls The wall stem will be subjected to very large bending
moments if a cantilever retaining wall is higher than 20 ft. In such cases it
may be economical to construct the wall with counterforts (walls perpendicular
to the stem, spaced about 1215 ft apart) that attach the stem to the heel. The
reinforced counterforts act like tension members supporting the stem. They also
greatly increase the bending strength of the stem. Figure 7.70e shows a
counterfort wall.
496 7 Foundations and Earth Supporting Walls
a b
Toe
Stem
Heel
c d
e f
Figure 7.70 Different types of retaining walls (a) gravity wall (b) cantilever retaining wall (c)
cantilever wall without heel (d) cantilever wall without toe (e) counterfort wall (f) buttress wall
4. Buttress walls These walls are similar to counterfort walls, except that the
buttresses, which attach the stem to the toe, are located in front of the wall
(see Figure 7.70f). The buttersses are in compression
Vertical Soil Pressure Under the Base of a Retaining Wall So far we have
discussed only wall and column footings that are concentrically loaded (i.e., the
load acts at the centroid of the footing, and the distribution of the pressures on the
soil is uniform). In general, concentric loading cannot be achieved under retaining
walls. The resultant force (R), as shown in Figure 7.68, does not intercept the
footing at its center, but rather is eccentric to it.
Figure 7.71 shows three different possibilities of pressure distribution under a
footing. In Case I the load is concentrically applied to the footing. In Case II the
load is applied at a small eccentricity (e), which is less than b=6. In Case III the load
is applied at an eccentricity larger than b=6. Because a 1-ft-long strip is considered
in a wall footing:
7.13 Earth Supporting Walls 497
1 b2 3
A b 1 b ft2 and Sm ft
6
The pressures for the individual cases then can be found as:
P P 6e 2P
fmax f 1
b max 3c
max f
b b
P 6e 7:56
fmin 1
b b
Case I Case II Case III
b b b
P P
P c e
c b/2 c e
e 0
3c
f max f min
f max f max
b
e 0; c
2 b b b b
P e ; c e ; c
f f max 6 3 6 3
A
f max P Pe From V 0, P 3c f
A Sm 2 max
P Pe f max 2P
f min
A Sm 3c
If 1 6e
b < 0 in Case II, the expression for fmin becomes negative, which
indicates that theoretically there is tension between the footing and the soil. But
tension cannot develop between the bottom of the footing and the soil, as a gap
would appear. This is an impossible and inadmissible situation, so Case III applies.
With a straight-line pressure distribution, the pressure volume under the footing
must be in equilibrium with the P force. Hence, the resultant of the reaction pressure
must be colinear with the P force. Then the neutral axis (i.e., where the pressures
become zero) must be located at a distance 3c from the toe (see Figure 7.71).
Shear keys and weep holes We can easily increase inadequate sliding resistance
in a retaining wall by using a shear key at the bottom of the footing, as shown in
Figure 7.72. The passive earth pressure in front of the key provides a sure and
economical resistance.
As with basement walls, it is important to prevent water saturation of the backll
behind retaining walls. Saturated backll increases the earth pressure dramatically
498 7 Foundations and Earth Supporting Walls
Weep hole
Shear key
and may endanger the stability of the wall. Providing weep holes in the wall at
regular spacing is the easiest and safest way to drain the backll. A typical retaining
wall weep hole is shown in Figure 7.72. Figure 7.73 shows a owchart for the
stability analysis and design of cantilever retaining walls.
7.
Check the bar development length.
Figure 7.73 Flowchart for the stability analysis and design of retaining walls
7.13 Earth Supporting Walls 499
Example 7.8 Analyze the stability of the gravity retaining wall shown in
Figure 7.74. The backll is sandy gravel, soil 120 pcf; the coefcient of the
lateral active earth pressure is Ka 0.32; and the coefcient of friction at the base
is 0.52. The wall is constructed of concrete, which weighs 150 pcf. A sur-
charge load of 150 psf exists on the upper elevation. Calculate the toe pressure on
the soil in addition to the stability analysis. Disregard the passive pressure in front
of the toe in the analysis.
4'-0" 2'-0"
wsc 150 psf
10'-0"
2 1 E1
12'-0"
W2 W1
4'-0" E2
3
Solution
Step 1. Calculate the lateral soil pressures and the overturning moment.
The overturning moment about the toe is the sum of the moments caused by
E1 and E2:
Mr 33,800
FSot 2:33 ! 2:33 > 1:5 ok
Mot 14,520
E
W
Rh E
0
R Rv W
intercepts the base of the wall. The vertical component of the resultant is
Rv W, and the horizontal component is Rh E; (W, E, Rh, and Rv are the
forces acting on the wall). The moment of a resultant about any point must
equal the sum of the moments of the composing forces about the same
point.
The moments of the composing forces are already known. The moment
of W about the toe is the resisting moment (Mr). The moment of E is the
overturning moment (Mot).
Hence, from the equilibrium of forces acting on the wall:
X
Mo 0
Wc Mr Mot 0
Wc Mr Mot 7:57
Mr Mot
c
W
33,800 14,520
c 2:0 ft
9,600
b
Because b 6.0 ft, this is Case III of Figure 7.71 c . Find the
3
maximum pressure, fmax:
2 9,600
fmax 3,200 psf
3 2:0
Figure 7.76 Soil pressure distribution below the wall in Example 7.8
502 7 Foundations and Earth Supporting Walls
Example 7.9 Analyze the stability of the reinforced concrete cantilever retaining
wall shown in Figure 7.77. Calculate the reinforcement required in the wall and the
footing. Disregard the passive resistance of the soil in front of the toe. Assume
Ka 0.32, soil 115 pcf, and 0.50. Use fc0 3,000 psi and fy 60,000 psi.
Concrete is normal weight.
Wwall
hw 12'-0" Wsoil
10'-6"
Wftg
1'-0"
pmax
1'-0" 1'-0" 3'-6"
5'-6"
Solution
Step 1. Calculate the lateral soil pressure and the overturning moment:
Step 2. Calculate the weight and the resisting moment. Include the weight of the
backll atop the heel of the wall and treat that as an integral part of the
retaining wall.
7.13 Earth Supporting Walls 503
Mr 19,309
FSot 2:72 > 1:5 ok
Mot 7,102
W 0:50 6,449
FSs 1:59 > 1:5 ok
E 2,029
Step 4. Calculate the soil pressure under the footing. Determine the location at
which the resultant force intersects the bottom of the footing and calculate
the resulting soil pressures.
So the eccentricity is
b 5:50
e c 1:89 0:86 ft
2 2
Because
b 5:5
e 0:86 < 0:92 ft
6 6
Therefore, Case II (see Figure 7.71) is applicable. The soil pressures under
the footing are:
W 6e 6,449 6 0:86
fmax 1 1 2;273 psf
b b 5:50 5:50
W 6e 6,449 6 0:86
fmin 1 1 72 psf
b b 5:50 5:50
Step 5. Design the required reinforcement in the stem. Calculate the factored
bending moment in the stem, as illustrated in Figure 7.78. The maximum
pressure at the base of the stem is:
pu 559.4 psf
pu hstem hstem 559:4 9:5 9:5
Mu
2 3 2 3
8,414 ft-lb=ft 8:41 ft-kip=ft
The minimum concrete cover required for #6 or larger bars (ACI Code,
Section 20.6.1.3.1) is 2 in., as the back of the wall is exposed to the soil.
Thus (assuming #6 bars):
12,000Mu 12,0008:41
R 91 psi
bd 2 129:632
Using fc0 3,000 psi concrete and fy 60,000 psi steel from Table A2.6a:
7.13 Earth Supporting Walls 505
or
#6 @12 in: c=c As 0:44 in:2 =ft
From Table A2.10 select #5 @ 12 in. c/c (As 0.31 in.2/ft) for the footing.
Use #4 @ 16 in. on each face of the stem, as walls thicker than 10 in. require
two layers of reinforcement (total As 2 0.15 0.30 in.2/ft).
Step 6. Design the reinforcement required in the heel and toe.
The heel acts like a cantilever from the back of the stem to the end of the
heel, as shown in Figure 7.79. The loads acting on it are the weight of the
soil from above, its self-weight, and the upward reaction pressures at its
bottom (found in step 4):
p1 72 psf
2,273 72
p2 72 3:5
5:5
p2 1,473 psf
The moment at the intersection of the heel and the stem is:
506 7 Foundations and Earth Supporting Walls
Location of maximum
moment in heel Self-weight 150 psf
72 psf
1,473 psf
3.5 ft
3:5 3:5
M 1, 093 150 3:5 72 3:5
2 2
3:5 3:5
1, 473 72 4;312 ft-lb=ft
2 3
The reinforcement will be placed at the top of the heel. Thus, 2 in. cover
(as in the case of stem) is required, therefore, d 9.63 in. For a 1-ft length
of the heel:
12,000Mu 12,0006:9
R 74 psi
bd2 129:632
Per the ACI Code, Sections 13.3.2.1 and 7.6.1.1 the minimum reinforce-
ment is equal to the required shrinkage and temperature reinforcements.
1'-0" 1'-0"
The toe is only 10 000 long, so the reinforcement required will not be
signicant. For the sake of thoroughness, however, we will also determine
the reinforcement in the toe.
p1 2,273 psf
2,273 72
p2 72 4:5 1,873 psf
5:5
Neglecting, conservatively, the soil on the toe, the moment at the intersec-
tion of the toe and the stem, as shown in Figure 7.80, is:
1:0 1:0 2
M 1,873 1501:0 2;273 1;873 1:0 995 ft-lb=ft
2 2 3
Mu 1:6995 1, 592 ft-lb 1:6 ft-kip=ft
0:75
d 12 3footing cover 8:63 in:assuming #6 bars
2
12,0001:6
R 22 psi ! Table A2:6a ! req < 0:0010
128:632
use minimum steel
As 0:001812 12 0:26 in:2 =ft
Table A2.10 ! Use 6 @ 12 in. c/c to dowel bars from the footing into the
stem (As 0.44 in.2/ft). These bars will also serve for reinforcement in
the toe.
Step 7. Check the bar development length.
(a) Stem Reinforcement For the reinforcements in the stem, we must lap
splice the dowels in the footing to the main reinforcement. From
Table A3.3, for #6 bars, d 33 in., which can be reduced by
As, required
:
As, provided
508 7 Foundations and Earth Supporting Walls
Therefore
0:38
required 33 29 in:
0:44
The footing is too small to provide the above length, so the bars are
hooked into the toe. The development length (per Equation (3.65)) is:
!
fy e c r
dh p d b
50 fc0
60,0001:00:71:0
dh p db 15:3db
501:0 3,000
dh 15:30:75 11:5 in:
As, required
From Table A3.4, we can reduce dh by A . Therefore, the
s, provided
required length is:
As, required
required = dh
As, provided
0.38
required = 11.5 = 9.9 in. > min{8(0.75), 6 in.} \ok
0.44
required = 9.9 in. < provided = 24 - (3 + 2) = 19 in. \ok
0:26
d 28 27 in:
0:27
Hooks are required in the toe area, as the toe is not long enough.
7.13 Earth Supporting Walls 509
#4 @ 18 in. vertical
#4 @ 16 in. horizontal
#6 @ 12 in. vertical
#4 @ 16 in. horizontal
#6 @ 12 in. dowels
3'-2" #5 @ 14 in.
#5 @ 12 in.
!
fy e c r 60,0001:00:71:0
dh p db p 0:625 9:6 in:
50 fc 0 501:0 3,000
0:26
required 9:6 9:2 in: > minf80:625, 6 in:g
0:27
provided 24 in: 3 in: 21 in: > 9:2 in: ok
(c) Toe Reinforcement The dowels from the stem are used as reinforce-
ment for the toe. From Equation (3.65), the required length of bars
with hook for #6 bars is:
!
fy e c r
dh p db
50 fc0
60,0001:00:71:0
dh p 0:75 11:5 in:
501:0 3,000
510 7 Foundations and Earth Supporting Walls
0:26
required 11:5 6:8 in: minf80:75, 6 in:g
0:44
provided 12 in: 3 in: 9 in: > 6:8 in: ok
Figure 7.81 shows the retaining wall and the details of the
reinforcements.
Problems
24'-0" 24'-0"
10'-0"
10'-0"
10'-0"
10'-0"
4'-0"
Building section
Problems 511
7.5. A 16 in. 16 in. reinforced concrete column supports 150 kip dead load and
75 kip roof live load. The allowable soil bearing pressure is 4,000 psf. Design a
square footing to support the column. Use fc0 3,000 psi for the column and
the footing and fy 60,000 psi. Assume the column to have 4 #6 bars as the
main reinforcement.
7.6. Design a square reinforced concrete spread footing for the interior columns of
Problem 5.11. The soil bearing capacity is 6,000 psf. Use fc0 3,000 psi for the
footing, fc0 4,000 psi for the column, and fy 60,000 psi. Column has 8 #9
main reinforcement. Neglect the self-weight of column.
7.7. Design a square reinforced concrete spread footing to support a 24 in. 24 in.
column carrying a 600 kip dead load and a 400 kip roof live load. The soil bearing
capacity is 10,000 psf. Use fc0 3,000 psi for the footing, fc0 4,000 psi for
the column, and fy 60,000 psi. The columns main reinforcements are 8 #11.
7.8. Redesign the footing of Example 7.3, if one of the horizontal dimensions of
the footing is limited to 70 000 due to the proximity of an adjacent property line.
7.9. Redesign the footing of Problem 7.7 if one of the horizontal dimensions of the
footing is limited to 80 000 .
7.10. Determine the thickness of the unreinforced basement wall shown below for
the following cases. The unit weight of the backll material is 120 pcf, and the
coefcient of active pressure Ka 0.33. Show the soil lateral pressure and
draw the shear force and bending moment diagrams for the applied loads. Use
fc0 4;000 psi:
(a) h2 80 000 , no surcharge (wsc 0) (c) h2 70 000 , wsc 200 psf
(b) h2 60 000 , no surcharge (d) h2 100 000 , wsc 200 psf
wsc
h 10'-0"
h2
512 7 Foundations and Earth Supporting Walls
7.11. Rework Problem 7.10 for h 120 000 . For case d, use h2 120 000 .
7.12. Design the plain concrete basement wall shown. The equivalent uid active
density of the backll material is 36 pcf. The unit weight of the soil is 120 pcf.
Consider two cases: (a) without surcharge, and (b) a surcharge of 150 psf
acting on the backll. Use fc0 3,000 psi.
7.13. Redesign the basement wall of Problem 7.12 using reinforced concrete. Use
fy 60,000 psi.
7.14. Check the adequacy of the 10-in.-thick reinforced concrete basement wall
shown below. Use fc0 4,000 psi and fy 60,000 psi. The clear cover is in.
The unit weight of the backll is 100 pcf, and the coefcient of active soil
pressure Ka 0.30.
12'-0"
6'-0"
t 10 in.
#5 @ 12 in. horizontal
14'-0"
#6 @ 12 in. vertical
7.15. What is the maximum allowable surcharge that can be placed on the outside
grade of the basement wall of Problem 7.14?
7.16. Check the stability of the concrete gravity retaining wall shown below. Also,
determine the soil pressure distribution on the base. The unit weight of the
backll is 120 pcf, the coefcient of active soil pressure is 0.30, and the
coefcient of friction at the base is 0.50. The unit weight of the concrete is
150 pcf. The applied surcharge on the backll is 100 psf. Disregard the
passive pressure action on the wall.
3'0" 6'0"
12'0"
3'0"
7.17. Check the stability of the concrete gravity retaining wall shown below. Also,
determine the soil pressure distribution on the base. The unit weight of the
backll is 115 pcf, the coefcient of active soil pressure is 0.33, and the
coefcient of friction at the base is 0.45. The unit weight of the concrete is
150 pcf. The applied surcharge on the backll is 130 psf. Disregard the
passive pressure action on the wall.
14'0"
3'6"
2'0"
7.18. Check the stability of the reinforced concrete cantilever retaining wall shown
below. Disregard the passive resistance of the soil in front of the toe. Assume
Ka 0.3, soil 120 pcf, and 0.52. The unit weight of the concrete is
150 pcf.
16'6"
1'6"
2'6" 1'6" 6'6"
10'6"
7.19. Design the cantilever retaining wall of Problem 7.18. Use fc0 3,000 psi and
fy 60,000 psi. Use the ACI Code-recommended minimum covers.
Self-Experiments
Experiment 2
To gain a better understanding of the behavior of different types of soil under a
foundation, we repeat Experiment 1 in the following order:
1. Fill a dish with dry sand. Compact the sand by gently pounding it with the
bottom of a bottle. Smooth the top, and place a wood block (representing a
square footing) on the sand. Load the block with an increasing load. Note what
happens to the sand around the loaded block.
2. Repeat step 1 using wet sand.
3. Repeat step 1 using wet clay. The clay must be wet enough to be moldable.
4. Repeat step 3 after letting the clay dry for a few days.
Styrofoam column
Rubber footing
Soil
Experiment 3
Form and cast a reinforced concrete square spread column footing. We will
use wires to represent the two required sets of reinforcement, as illustrated in
Figure SE 7.2. Document all your problems and observations in casting the footing.
516 7 Foundations and Earth Supporting Walls
Experiment 4
Using the concrete footing of Experiment 3, repeat Experiment 1 by adding a
square reinforced concrete column at the center of the footing. Remember that you
need dowel bars to tie the column to the footing. Document all problems and your
observations in the construction of the column and footing model.
What is the approximate capacity of the footing if it is placed on a soil with a
bearing capacity of 2,000 psf?
Chapter 8
Formwork for Monolithic Concrete
Construction
8.1 Introduction
This Chapter discusses the issues that need to be studied and understood by an
aspiring architectural or construction engineer, regardless whether works as a
designer, or is engaged in construction.
We encourage the readers to consult the National Design Specication for Wood
Construction, which provides comprehensive instructions for the use and the
allowable values for metal connectors (nails, bolts, etc.) in wood construction.
Besides offering detailed guidance to the analysis and design of formwork
elements, this Chapter discusses the design of wood shores. We have also included
many useful Tables and step-by-step numerical examples for the easier compre-
hension and safe design of different wood and plywood formwork component
elements. The scope and available space unfortunately does not permit us to include
here the many, mostly patented, metal shoring and forming systems offered by a
large number of manufacturers. A search of the Internet will help those readers,
who may wish to expand their knowledge in the subject beyond wood materials
used in formwork and shoring.
Formwork and shoring that supports it are the largest cost-component in mono-
lithic concrete construction, often reaching 50 %-or more-of the total. Even more
important is to know that the formwork materials costs amount to about 1/6th of
that and labor cost associated with the formwork in the U.S.-and other countries
with relatively high wages-is about 5/6th.
Formwork should be thought of as a mold into which rst the reinforcement is
assembled, then the freshly mixed concrete is poured, consolidated and cured. The
shoring should be thought of as the temporary structure necessary to support the
formwork, the weight of the workers during the construction process, the weight of
the fresh concrete and the weight of the equipment used during the concreting
operation.
The most important reference material for the design and construction of formwork
is the American Concrete Institutes Guide to Formwork for Concrete, ACI 347R-
14. In addition, a much more comprehensive and authoritative treatment of the
subject may be found in the ACI SP-4 7th Ed., a book titled Formwork for Concrete,
written by Mary K. Hurd.
Shoring and formwork are essentially temporary structures that must be built to
provide a mold into which the uid concrete is poured, consolidated and cured.
Thus they have to be carefully designed to provide for the essential attributes of all
structural designs, which are quality, safety and economy.
Quality in formwork means:
(a) the quality of the facing material, which is an important constituent in achiev-
ing the desired nish of the formed components;
(b) the formwork must be watertight. Butt joints, or corner joints between the
plywood sheets must be sealed to prevent leakage from the fresh concrete.
Leakage leads to unsightly cement ns at such locations, which when broken
off clearly show a discontinuity of appearance. It may also lead to
honeycombing in the concrete, which is rather difcult and expensive to
repair.
The sealing may be done by using a thin adhesive tape, or an appropriate
caulk. Corner joints are often protected from leakage by sealed wooden
(or plastic) chamfers.
(c) the accuracy, i.e. size, thickness, and geometrical conformance of the nished
concrete construct to the design documents. The reader is referred to ACI
Standard Specication for Tolerances for Concrete Construction, ACI
117, which lists the acceptable deviations from the dimensions provided for
in the Contract Documents.
Forms that are not designed and/or constructed to produce elements satis-
fying the tolerance requirements, or the nish requirements set forth in the
project specications, may result in expensive renishing (or in some cases
demolishing and re-construction)!
Safety refers to the requirements:
(a) that the temporary structure must be designed and built to safely withstand all
the loads (gravity and lateral) that it is subjected to during the construction
process;
(b) that it provides for the safety of the construction personnel.
Economy in the design and construction of the formwork looms huge among the
attributes, when one looks at the share of its cost in the overall concrete construction
cost. In the construction process there are three major components contributing to
the total cost of monolithic reinforced concrete construction in architectural struc-
tures (refer to Figure 8.1):
8.3 Loads on Formwork 519
For the purposes of design, Dead Loads are dened as the weight of the formwork,
shoring and scaffolding, the weight of the reinforcement and the freshly placed
concrete. In the authors experience an allowance of 10 psf to 15 psf is generally
sufcient to account for the weight of the formwork, scaffolds and shores. To
account for the weight of concrete and reinforcement, use the customary 150 pcf
for the design.
Live loads are more difcult to predict. Those represent the weight of workers,
concrete buggies or other concrete conveying equipment, pumping hoses,
generators, compressors, consolidating and nishing equipment, etc. The ACI
347 recommends the use of a minimum 50 psf uniformly distributed live load to
account for these, and minimum 75 psf, when motorized carts are used for the
conveyance of concrete.
520 8 Formwork for Monolithic Concrete Construction
Example 8.1 Calculate the uniformly distributed load a formwork (and its
supporting post-shores) must be designed for a 10 in. thick normal weight concrete
slab. The concrete will be delivered via a concrete pump.
Solution
Fresh concrete behaves like a uid when placed, thus exerts hydrostatic pressure
against the sides of the forms. The value of the pressure depends upon many
variables. If the concrete is poured to full height of the form within a time that is
less than the initial set, then the formwork will experience full hydrostatic pressure
in the form of
p wc h 8:1
where:
wc unit weight of concrete, lb/ft3
h depth of uid or plastic concrete from top of placement to point of consider-
ation in the form, ft
However, when the rate of pour is slower than the time required for the initial set
of the concrete mix, the pressure at the bottom of the formwork will diminish from
the full hydrostatic pressure. The parameters inuencing the magnitude are the
temperature of the concrete inside the form (higher temperature accelerates the
setting time) and the chemistry of the concrete mix (pozzolans used as cement
replacement, retarding admixtures, etc.). In combining all these effects into one
empirical formula, ACI 347 recommends
pmax wc h 8:3
where
R rate of placement, ft/hr
T temperature of concrete during placement, F
Cw unit weight coefcient (see Table 8.1)
Cc chemistry coefcient (see Table 8.2)
8.3 Loads on Formwork 521
For walls higher than 14 ft, or for placement rates between 7 and 15 ft per hour
the following empirical formula is recommended:
Example 8.2 Assume a 24 in. 24 in. column form 14 ft high. The project
specications require normal weight concrete with Type I cement and allow the
use of maximum 25 % y ash cement replacement. Calculate the maximum lateral
pressure at the base of the forms.
Solution The total volume of concrete will be 2 2 14 56 cubic feet !
2:07 cubic yards. This is a small amount, hence it is more than likely that it will
be poured faster than the initial setabout 2 hours after mixingregardless of the
temperature of the concrete. Thus Equation (8.3) will apply:
Example 8.3 Calculate the maximum pressure on a wall-form for an 18 in. thick,
60 ft long and 12 ft high wall. Normal weight concrete with Type I cement is
specied with a 40 % GBFS (Ground Blast Furnace Slag) cement replacement
without retarders. The average temperature during the pour is expected to be
about 80 F and the rate of pour is estimated to be about 4 ft/hr.
Solution
Cw 1:0
Cc 1:2 (From Table 8.2 (fourth row): blend with less than 70 % slag and no
retarder)
Using Equation (8.2): pmax 1:0 1:2 150 9,000 4=80 720 psf
522 8 Formwork for Monolithic Concrete Construction
Since this is greater than 600 Cw 600 psf (refer to Equation (8.2)), use
720 psf for the design.
Example 8.4 Same data as in Example 8.3, but the rate of pour is 5 ft/hr and the
temperature is expected to be only 50 F.
Then
A comparison of the results in Examples 8.2 and 8.3 shows, rather dramatically,
the importance of the R and T parameters. While the rst, i.e. the rate of pour can be
more accurately planned and enforced on the eld, the temperature on the date of
the pour is more difcult to assess with much accuracy. Hence, the designer should
always be cautious. Many authorities in the eld recommend that unless the
economy of the formwork design absolutely forbids it, the most conservative
design, i.e. Equation (8.3) be used.
The temporary structure of the shoring, bracing and forming assembly also must be
designed for lateral loads, such as wind and/or seismic loads, just like any other
structure. For the evaluation of such loads, the reader is referred to the SEI/ASCE
37-10 Standard Design Loads on Structures during Construction.
ACI 347R-14 Paragraph 4.2.3.1 states the following:
Formwork exposed to the elements should be designed for wind pressures
determined in accordance with ASCE/SEI 7 with adjustments as provided in
ASCE/SEI 37 for shorter recurrence interval. Alternatively, formwork may be
designed for the local building code-required lateral wind pressure, but not
less than 15 psf. Consideration should be given to possible wind uplift on
formwork.
In addition, concrete formwork is subject to other horizontal loads from the
starting and stopping of motorized equipment, from the concrete dumping opera-
tions, and other not-easily-calculable effects. ACI 347 recommends a minimum
horizontal force of 100 lb per linear ft (lb/ft) of oor edge in any direction be used to
account for these effects. This force is an alternative requirement to the minimum
wind load cited above, for it is unlikely that concreting operations would occur
during a wind storm.
In concrete construction usually the vertical elements are poured rst, thus they
are often available to provide lateral support to the shoring and forming elements.
The designer of the formwork must, however, carefully analyze the available
strength of the already-in-place elements when considering their utilization for
providing the necessary stability of the temporary structure.
8.4 Materials for Formwork 523
Form panels should be thought of as the lining of the mold, i.e. the surface that will
be in intimate contact with the concrete. While many different materials may be
used as form panels, the overwhelming majority of them are APA (American
Plywood Association) Exterior type Plywood. These products are manufactured
using moisture resistant adhesives.
The industry also manufactures a trademarked special product known as
Plyform, manufactured for the specic purpose of forming for concrete.
Plyform-and indeed almost all plywood-panels are manufactured in odd num-
ber of layers (also called plies). The layers are laid up with the grain perpendicular
in adjacent layers, and bonded together under high pressure using adhesives. The
adhesives are selected on the basis of intended use, i.e., interior (not exposed to
weather), or exterior use. Plyform panels are always made with exterior quality
(waterproof) adhesives.
The alternating direction of the grain in the adjacent layers helps to minimize
the shrinking and warping of the panels. The typical plywood panels are
manufactured in 4 ft 8 ft size, although they may be available in larger sizes
on special order. The grain orientation in the outer plies is always in the long
direction.
Plyform panels are always manufactured with exterior bond classications and
in three basic grades: Plyform Class I, Plyform Class II and Structural I Plyform.
The Class refers to the strength of the Plyform, which in turn depends upon the
Group wood species that form the outer ply.
Overlaid Plyform panels are also manufactured for concrete form use. Two
types MDO (Medium Density Overlay) and HDO (High Density Overlay) are
available. During the fabrication process thermo-setting phenolic resins are
bonded-usually on one side only-to the surface of the plywood using high heat
and pressure. The overlay produces a smooth, hard, semi-opaque surface and
increases the durability, hence the re-use, of the forming panels manifold. It can
help to create concrete to appear nearly like a polished surface.
Since many different species of wood are used in manufacturing plywood and
due to its inner construction grains in adjacent layers are perpendicular to each other
and thus have different strength and stiffness characteristics, the cross-sectional
properties cannot be simply calculated like we do when using homogenous mate-
rials. Table 8.4 shows the calculated cross section properties of a 1 ft (12 in.) wide
section of Plyform. Thus the designer may need to select only the appropriate
effective section properties and the allowable stresses for the face ply in order to
perform any required calculations.
Table 8.3 shows the allowable design stress values and the Modulus of
Elasticity of the various Plyform materials. Note that the listed values are adjusted
values, used specically in concrete forming. In wood products the Codes require
524 8 Formwork for Monolithic Concrete Construction
the application of several adjustment factors, like wet service factor, load duration
factor, etc. So, the values taken from the Plywood Design Specications allow-
able stress are adjusted to wet use (outdoor construction must assume that the
formwork is in wet condition); however the load duration is a relatively short
one, for as the concrete stiffens and begins to carry its own weight, the pressures
on the plywood diminish. Thus, a special Cs 1.65 concrete setting factor was
applied together with the wet design stresses. Table 8.3 shows the allowable
design stresses already adjusted to account for the wet service and load duration
factors.
Table 8.3 Allowable stresses and moduli of elasticity used with Plyform
Plyform Plyform Structural
Class I Class II I Plyform
Modulus of elasticityE 1,650,000 psi 1,430,000 psi 1,650,000 psi
(adjusted, use for bending
and deection calculation)
Modulus of elasticityEs 1,500,000 psi 1,300,000 psi 1,500,000 psi
(unadjusted, use for shear
deection calculation)
Bending stressFb 1,930 psi 1,330 psi 1,930 psi
Rolling shear stressFs 72 psi 72 psi 102 psi
The term Rolling Shear Constant (see Table 8.4) maybe unfamiliar to the reader.
The bers in an inner layer of the plywood, laid perpendicular to the face layer, tend
to roll over each other, (not unlike a layer made out of toothpicks), when subjected
to horizontal shear due to bending.
Horizontal shear in members in bending is calculated as:
VQ V
v 8:5
Ib Ib
Q
The term in parenthesis in the denominator is referred as the rolling shear constant.
Due to the non-homogenous and non-isotropic nature of the plywood material,
calculations of the cross-sectional properties are far from being simple. Conve-
niently, however, all the design-applicable information is tabulated and made
readily available. (Refer to Table 8.4).
8.4 Materials for Formwork 525
Table 8.4 Section properties of plyform (courtesy of APAThe Engineered Wood Association)
Section properties for Plyform Class I and Class II, and Structural I Plyforma
Properties for stress applied Properties for stress applied
parallel with face grain perpendicular to face grain
Effective Rolling Effective Rolling
section shear section shear
Approx. Nominal Moment modulus constant Moment modulus constant
Performance weight thickness of inertia KS Ib/Q of inertia KS Ib/Q
category (psf) t (in.) I (in.4/ft) (in.3/ft) (in.2/ft) I (in.4/ft) (in.3/ft) (in.2/ft)
Class I
15/32 1.4 0.469 0.066 0.244 4.743 0.018 0.107 2.419
1/2 1.5 0.500 0.077 0.268 5.153 0.024 0.130 2.739
19/32 1.7 0.594 0.115 0.335 5.438 0.029 0.146 2.834
5/8 1.8 0.625 0.130 0.358 5.717 0.038 0.175 3.094
11/16 2.0 0.688 0.164 0.409 6.175 0.044 0.183 3.524
23/32 2.1 0.719 0.180 0.430 7.009 0.072 0.247 3.798
3/4 2.2 0.750 0.199 0.455 7.187 0.092 0.306 4.063
7/8 2.6 0.875 0.296 0.584 8.555 0.151 0.422 6.028
1 3.0 1.000 0.427 0.737 9.374 0.270 0.634 7.014
11/8 3.3 1.125 0.554 0.849 10.430 0.398 0.799 8.419
Class II
15/32 1.4 0.469 0.063 0.243 4.499 0.015 0.138 2.434
1/2 1.5 0.500 0.075 0.267 4.891 0.020 0.167 2.727
19/32 1.7 0.594 0.115 0.334 5.326 0.025 0.188 2.812
5/8 1.8 0.625 0.130 0.357 5.593 0.032 0.225 3.074
11/16 2.0 0.688 0.164 0.409 6.020 0.036 0.236 3.496
23/32 2.1 0.719 0.180 0.430 6.504 0.060 0.317 3.781
3/4 2.2 0.750 0.198 0.454 6.631 0.075 0.392 4.049
7/8 2.6 0.875 0.300 0.591 7.990 0.123 0.542 5.997
1 3.0 1.000 0.421 0.754 8.614 0.220 0.812 6.987
11/8 3.3 1.125 0.566 0.869 9.571 0.323 1.023 8.388
Structural I
15/32 1.4 0.469 0.067 0.246 4.503 0.021 0.147 2.405
1/2 1.5 0.500 0.078 0.271 4.908 0.029 0.178 2.725
19/32 1.7 0.594 0.116 0.338 5.018 0.034 0.199 2.811
5/8 1.8 0.625 0.131 0.361 5.258 0.045 0.238 3.073
11/16 2.0 0.688 0.167 0.418 5.621 0.051 0.249 3.493
23/32 2.1 0.719 0.183 0.439 6.109 0.085 0.338 3.780
3/4 2.2 0.750 0.202 0.464 6.189 0.108 0.418 4.047
7/8 2.6 0.875 0.317 0.626 7.539 0.179 0.579 5.991
1 3.0 1.000 0.479 0.827 7.978 0.321 0.870 6.981
11/8 3.3 1.125 0.623 0.955 8.841 0.474 1.098 8.377
a
The section properties presented here are specically for Plyform, with its special layup restric-
tions. For other grades, section properties are listed in the APAs Plywood Design Specication,
Form Y510
526 8 Formwork for Monolithic Concrete Construction
8.4.2 Lumber
Practically any formwork construction and consequently its design involve lumber.
Although the material selection may use any species listed in the National Design
Specication for Wood Construction and its Supplement: Design Values for Wood
Construction, published by the American Forest & Paper Association (AFPA) and the
American Wood Council (AWC), only a few species are typically used in the con-
struction industry. These are: Douglas Fir-Larch North; Douglas Fir South; Hem-Fir;
Spruce-Pine-Fir and Southern Pine. Within all these species each piece of cut-to-size
lumber is either visually or machine graded and classied as to their stress-grade.
Table 8.5 shows the sizes and section-properties of lumber typically used in
formwork construction. The listed values are for (S4SSurfaced 4 Sides) dressing,
which means that after rough sawing from the timber, the boards are run through a
planing machine to obtain smooth surface on all four sides and uniform cross-section.
Table 8.6 shows the Reference Design Stress values for 2 in.4 in. thick
lumber of selected species and grades. These values are subject to a number of
Adjustment Factors. We list only those factors that are typically used in formwork
design.
Table 8.5 Properties of selected lumber sizes, typically used in formwork construction
Area of
Nominal Actual Approx. cross- Moment Section Moment Section
size size weight section of inertia modulus of inertia modulus
(in.) (in.) (lb/ft) (in.2) Ix (in.4) Sx (in.3) Iy (in.4) Sy (in.3)
24 1.5 3.5 1.3 5.25 5.36 3.06 0.98 1.31
26 1.5 5.5 2.0 8.25 20.80 7.56 1.55 2.06
28 1.5 7.25 2.6 10.87 47.63 13.14 2.04 2.72
2 10 1.5 9.25 3.4 13.87 98.93 21.39 2.60 3.47
2 12 1.5 11.25 4.1 16.87 177.97 31.64 3.16 4.21
34 2.5 3.5 2.1 8.75 8.93 5.10 4.56 3.65
36 2.5 5.5 3.4 13.75 34.66 12.60 7.16 5.73
38 2.5 7.25 4.4 18.12 79.39 21.90 9.44 7.55
3 10 2.5 9.25 5.6 23.12 164.89 35.65 12.04 9.63
3 12 2.5 11.25 6.8 28.12 296.63 52.73 14.65 11.72
44 3.5 3.5 3.0 12.25 12.50 7.15 12.50 7.15
46 3.5 5.5 4.7 19.25 48.53 17.65 19.65 11.23
48 3.5 7.25 6.2 25.38 111.15 30.66 25.90 14.80
4 10 3.5 9.25 7.9 32.38 230.84 49.91 33.05 18.88
4 12 3.5 11.25 9.6 39.38 415.28 73.83 40.20 22.97
66 5.5 5.5 7.4 30.25 76.26 27.73 76.26 27.73
68 5.5 7.5 10.0 41.25 193.36 51.56 103.98 37.81
88 7.5 7.5 13.7 56.25 263.67 70.31 263.67 70.31
Notes: (1). Weights shown assume dry condition, approx. 35 lb/ft3; (2). Ix and Sx are about the
strong axis of the section; (3). Iy and Sy are about the weak axis of the section; and (4). 6 and
8 sizes are mostly used as shore-posts.
8.4 Materials for Formwork 527
Table 8.6 Reference design stresses for selected species of visually graded lumber
2 in. to 4 in. thick
Shear Compression Compression Modulus
parallel perpendicular parallel to of elasticity
Species and Size Bending to grain to grain grain (psi)
grade classication Fb (psi) Fv (psi) Fc (psi) Fc (psi) E (psi) Emin (psi)
Douglas Fir-Larch (North)
No. 2 2 in. and wider 850 180 625 1,400 1,600,000 580,000
Construction 2 in.4 in. wide 950 180 625 1,800 1,500,000 550,000
Douglas Fir (South)
No. 2 2 in. and wider 850 180 520 1,350 1,200,000 440,000
Construction 2 in. and wider 975 180 520 1,650 1,200,000 440,000
Hem-Fir
No. 2 2 in. and wider 850 150 405 1,300 1,300,000 470,000
Construction 2 in.4 in. wide 975 150 405 1,550 1,300,000 470,000
Spruce-Pine-Fir
No. 2 2 in. and wider 775 135 335 1,000 1,100,000 400,000
Construction 2 in.4 in. wide 875 135 335 1,200 1,000,000 370,000
Southern Pine
No. 2 2 in.4 in. wide 1,500 175 565 1,650 1,600,000 580,000
5 in.6 in. wide 1,250 175 565 1,600 1,600,000 580,000
8 in. wide 1,200 175 565 1,550 1,600,000 580,000
10 in. wide 1,050 175 565 1,500 1,600,000 580,000
12 in. wide 975 175 565 1,450 1,600,000 580,000
Construction 4 in. wide 1,100 175 565 1,800 1,500,000 550,000
Note: Emin is used with the calculation of the Column Stability Factor, CP
CDLoad Duration Factor. The stress level that wood may safely sustain for
short periods of time is higher than those from loads that are permanent. Thus for
concrete formwork a CD 1.25 (load duration 7 days or less) applies.
CMWet Service Factor. The strength and stiffness of wood is adversely
affected, when the moisture content in the material is greater than about 19 %.
The reduction of the allowable stress is different from the type of stress, for
example perpendicular to grain compression is more affected than shear
stress, etc.
CtTemperature Factor. Sustained temperatures above 100 F adversely affect
some properties of wood. In very hot climates, the stresses and the Modulus of
Elasticity (used in deection calculations) should be adjusted accordingly.
CFSize Factor. The NDS Code permits the use of this adjustment factor (see
Table 8.8) for joists, beams and studs. As its name implies, the factor depends
upon the size of the member. The magnitude of the factor is based on probability
studies of reliability and is applicable to all species of timber used in formwork
construction, with the exception of Southern Pine, where the size factors are
already included in the tabulated reference design stresses.
CrRepetitive Members Factor. The NDS Code permits the increase of the
allowable bending stresses (only) for 2 in.4 in. wide joists, beams, studs by
15 %, provided that there are at least three such members, spaced not more than
528 8 Formwork for Monolithic Concrete Construction
24 in. apart and are joined by oor, roof or other load distributing elements.
Minimum two span, appropriately designed plywood qualies for that role. Thus
Cr 1.15. However, ACI 347 does not recommend the use of this factor for
dimensional lumber, when CD and/or CF factors are also utilized.
CpColumn Stability Factor. This factor applies to elements (primarily
shoring posts) and applies to allowable compression parallel-to-grain stress values.
CfuFlat use factor. This factor is applicable to the Fb reference design bending
stress, when the load is applied to the wide face of the member. (See Table 8.8)
CbBearing Area Factor. Allowable perpendicular to grain compression stresses
are permitted to be multiplied by a factor of
b 0:375
Cb 8:6
b
where b is the bearing length measured in inches parallel with the grain, provided
the bearing length is less than 6 in. and it is not less than 3 in. from the end of the
member. Cb 1.0 for bearing at the end of a member and/or b > 6 in.
It can be disregarded conservatively.
To make it easier to comprehend, we summarize the use and applicability of
these adjustment factors in Table 8.9.
Thus the general formula to compute allowable stresses (F0 ) is the reference
design stress (F) multiplied by all the applicable adjustment factors:
F0 F CD CM Ct CF Cr Cp Cfu Cb 8:7
Note that only a selected few adjustment factors are typically applicable,
depending on condition and type of stress involved. See Table 8.9. (Refer to the
NDS Code for additional information).
Table 8.7 Reference design stresses for selected species of visually graded lumber
Posts and Timbers 5 in. 5 in. or larger
Shear Compression Compression Modulus of elasticity
Species parallel perpendicular parallel to
and Bending to grain to grain grain
grade Fb (psi) Fv (psi) Fc (psi) Fc (psi) E (psi) Emin (psi)
Douglas Fir-Larch (North)
No. 2 725 170 625 700 1,300,000 470,000
Douglas Fir (South)
No. 2 675 165 520 650 1,000,000 370,000
Hem-Fir
No. 2 575 140 405 575 1,100,000 400,000
Spruce-Pine-Fir
No. 2 500 125 425 500 1,000,000 370,000
Southern Pine
No. 2 850 165 375 525 1,200,000 440,000
8.4 Materials for Formwork 529
Table 8.8 Size and at use adjustment factors for grades #1 and #2
(for construction grade lumber the size factor is 1.00 for all widthsa)
Bending stress adjustment factor
Size factor Flat use factor
Width of 2 in.3 in. 4 in. 2 in.3 in. 4 in. Compression parallel
lumber thick thick thick thick to grain adjustment factor
2 in. and 3 in. 1.50 1.50 1.00 1.15
4 in. 1.50 1.50 1.10 1.00 1.15
6 in. 1.30 1.30 1.15 1.05 1.10
8 in. 1.20 1.30 1.15 1.05 1.05
10 in. 1.10 1.20 1.20 1.10 1.00
12 in. 1.00 1.10 1.20 1.10 1.00
a
Size and Flat use factors are not applicable for Southern Pine
Table 8.9 Tabulation of typical adjustment factors to basic material property values for dimen-
sion lumber used in formwork design
Type of stress (or property) Adjustment factors Comment
Bending stress, Fb Load duration factor, CD 1.25 Typically used
Size factor, CF see Table 8.8 Typically used
Flat use factor, Cfu see Table 8.8 Used as applicable
Repetitive member factor, Cr 1.15 Not recommended
by ACI 347
Shear stress, Fv Load duration factor, CD 1.25 Typically used
Wet service factor, CM 0.97 Used as applicable
Compression Bearing area factor, Cb see Used as applicable
perpendicular to grain, Fc Equation (8.6)
Wet service factor, CM 0.67 Used as applicable
Compression Load duration factor, CD 1.25* Typically used
parallel with grain, Fc Wet service factor, CM 0.8 Used as applicable
Size factor, CF see Table 8.8 Typically used
Column stability factor, CP see Always used
Equation (8.9)
Modulus of elasticity, E Wet service factor, CM 0.9 Used as applicable
*CD 1.0 for shoring posts
Formwork accessories are hardware items that are typically used in the construction
of the formwork. The reader is encouraged to search the Internet, where numerous
companies list and exhibit their proprietary products.
TIES are used in holding the opposite sides of the form secure against the lateral
pressure from the fresh concrete. (See Section 8.3.2.) There are many different,
usually patented devices. Some typical form ties are shown in Figure 8.2. They pass
through the concrete and fastened onto the formwork on each side. Manufacturers
530 8 Formwork for Monolithic Concrete Construction
cone shaped
plastic spreader.
wedge anchor
optional
plate.
water stop.
break back
point.
wing nut.
wing nut.
cone shaped
plastic spreader.
provide information on the safe working loads, which may range from about
2,000 lb for light duty to in excess of 60,000 lb for some super-heavy duty ties.
The published working load values carry a typically used safety factor of 2.
After the concrete hardened, they are either fully, or partially withdrawn, usually
1224 hours after placement of the concrete. Snap ties (as the name implies) have a
weakened point about 1.5 in.2 in. inside the concrete surface, called break-back
8.4 Materials for Formwork 531
points. The idea is that after breaking the tie, the outer part is withdrawn, but the rest
remains inside. A similar idea is represented by the so-called she-bolt type of tie,
where the outer parts are unscrewed from the inner tie-rod. The crimping in the ties
prevents the inner tie from rotating.
It is very important to note that the holes left in the concrete, whether just partial
on the surfaces, or all the way through, as in the case fully withdrawn types of ties,
be thoroughly lled with grout or pressure-grouted. ACI 347 recommends that no
corrodible metal be left in the concrete closer than 1.5 in. from the surface,
especially on concrete surfaces left exposed to view.
SPREADERS are used to keep the opposing sides of the formwork apart at a
xed dimension during concrete placement. Sometime they are just wooden blocks,
or special adjustable metal products, not left in the concrete, but removed either
during, or immediately after the placement of the fresh concrete. Some tie systems
(see Figure 8.2) incorporate plastic cone or hemisphere shaped spreaders secured to
the tie. After the removal of these plastic appurtenances and the withdrawal of the
outside portions of the ties, the remaining indentations provide good base for the
grouting of the holes in a visually acceptable manner.
SPACERS are usually small plastic attachments that are snapped onto the
reinforcement and used to maintain the specied concrete cover distance from the
vertical form.
CLAMPS are used to secure column forms. (See Figures 8.3 and 8.4; also
Example 8.6 for more detailed explanation).
HANGERS are primarily used to hang slab formwork from steel or pre-cast
concrete beams. Their use eliminate the need for shoring under the slab forms.
In addition to these most often used accessories, many different purpose and
design inserts, anchors, etc. are offered by manufacturers. All aim to help with
building the temporary structure speedily and economically and to enable the
plywood sheeting
clamp
batten
Figure 8.3 Three dimensional view of the formwork components for a concrete column
532 8 Formwork for Monolithic Concrete Construction
chamfer
if req'd.
contractor in forming the desired mold, regardless how complicated and/or unusual
is the designed concrete element. Discussion about these accessories is beyond the
scope of this chapter.
Fresh concrete will adhere to the surfaces of the formwork, unless a special coating,
or release agent is applied to the contact area. Coatings are differentiated from
release agents, for they also provide other sometimes desirable benets.
Coatings (sealers) are used to:
modify the texture of the concrete surface;
enhance the durability of the nished concrete surface;
prevent the fresh concrete adhering to the form material; and
seal the concrete surface from moisture intrusion.
Release agents (often commonly referred to as form-oils, as a reference to the
past, when petroleum based products were used exclusively) serve only as a bond-
breaker between the fresh concrete and the form surfaces, thus facilitating the easy
removal and preservation of the formwork material. Release agents may be applied
to the form material during its manufacture, or applied in the eld. Care must be
taken during eld application to prevent the release agent from coating reinforcing
steel.
There are two different types of release agents available on the market:
(a) barrier release agents and (b) reactive release agents.
8.5 Design of Formwork Elements 533
(a) Barrier release agents, as the name clearly implies, develop a lm (barrier)
on the contact surface, thus preventing bond from the fresh concrete to the
forms.
(b) Reactive release agents interact with the free lime in the cement matrix and
through a chemical reaction process a bond-preventing-lm is created. This
lm is considered bio-degradable, hence their use is better for the
environment.
When designing formwork for vertical structural members, such as walls and
columns, four elements: sheathing, studs, wales and ties need to be carefully
analyzed and selected. (Refer to Figure 8.5 for the visual representation of these
designations used in the industry.) The sheathing (Plyform1, or plywood) retains
the lateral pressure from the fresh concrete, the studs support the sheathing and the
wales support the studs. Finally, the ties are attached to the outside of the wales to
hold together the two faces. Figure B8.1 in Appendix B shows the process for a
large wall forming.
In the design of horizontal structural members elements (slabs and beams), the
elements that need to be carefully analyzed and selected are: sheathing, joists,
stringers (beams) and shores. The sheathing supports the weight of the concrete
(plus the construction loads), the joists support the sheathing, the stringers support
the joists and the shores support the stringers.
With the exception of the ties and the shores, all other elements are exural
elements and conventional beam formulas are used in their analysis. Since form-
work design involves many assumptions about the loads, the quality of the mate-
rials used and the workmanship in the actual construction, a simplied design
approach is justied. Thus, the following assumptions are commonly accepted
and used:
(a) Assume that all loads are uniformly distributed. The loads (pressures) on the
sheathing are truly distributed (albeit not always uniformly). So are the loads
on the studs and the joists that support the sheathing.
(b) Wales and stringers are actually loaded by point loads (typically closely
spaced) and can be approximated by an equivalent distributed load.
(c) Flexural elements that are continuous over three or more spans can be safely
designed by the formulas that apply for the 3-span condition.
Three governing conditions need to be considered in the design of exural
elements. These are: bending stresses, shear stresses and deection.
534 8 Formwork for Monolithic Concrete Construction
a 100 lb/ft
alternate
lateral load
plywood
wales
4'-0"
studs
form ties
15 psf wind load min.
2x4 strut
lateral
brace.
8'-0"
anchor
2x4 brace
toe plate
6'-0"
b
wale spacing
ing
ac
sp
tie
ing
pac
ds
stu
Figure 8.5 Typical formwork components for a concrete wall: (a) Section, (b) Three-dimen-
sional view
8.5 Design of Formwork Elements 535
The formulas in Table 8.10 are for calculating the maximum bending moment and
shear-force that in turn are used in the analysis of the stresses in the member under
investigation. In additon, this table includes the equations to compute the maxi-
mum deections. The numerical coefcients are adjusted to the fact that the load is
input in lb/ft (pounds per lineal foot) units. The formulas for Mmax are those for the
absolute maximum that will occur; i.e. at mid-span for the simple span; at an
intermediate support for the two or three span condition. The formulas for Vmax
include the fact, that the critical shear does not occur at the point support, but a d
distance away from it.
The formulas shown in Table 8.10 can be solved for the allowable maximum
span for a pre-selected type and thick Plyform sheathing, or a pre-selected size and
species of lumber. The resulting formulas are shown in Table 8.11. Among the
Table 8.10 Design formulas for bending members
Simple span Two spans Three or more spans
w lb/ft w lb/ft w lb/ft
formulas for deection check, we list three different cases, the rst two are
applicable when the deection limit is specied as a fraction of the span, i.e.,
L/240 or L/360. The third case provides for when the deection limit is specied
as a denite value, for example max 1/16 in.
P AFc0 8:8
where
A cross-section area of the solid post;
Fc0 allowable compression stress, which is an adjusted reference design stress
parallel with grain in compression (Fc) for the selected species and grade;
The following adjustment factors apply:
CD 1.0 (Load Duration Factor)-Shoring posts may be left in place for an extended
period of time, hence ACI 347 does not recommend an increased value for the
design of posts.
CM (Wet Service Factor) CM 0.8 when used in Equation (8.10), CM 0.9 for
Equation (8.12). This is a conservative approach, again due to the likelihood that
shoring posts may be used for an extended period of time.
Ct 1.0 (Temperature Factor)-It is unlikely that the shoring will serve in temper-
atures in excess of 150 F.
CF (Size factor)-Applicable species listed in Table 8.7, except Southern Pine. Refer
to Table 8.8 for applicable values.
CP (Column Stability Factor):
v u2
FcE u 32 FcE
1 u FcE
* u 6 1 F*
Fc* 7
Fc u
CP t4 5 c 8:9
2c c
2c
where
Fc* Fc CD CM Ct CF 8:10
0
0:822Emin
FcE 8:11
e =d2
c 0:8 for sawn lumber
e the effective length of compression member
d least dimension of rectangular compression member
0
Emin Emin CD CM Ct CF 8:12
e =d slenderness ratio of compression member
8.6 Wall Formwork Design 537
Note: while in wood construction the maximum slenderness ratio for solid
columns is limited to 50, during construction this limit may be raised to 75.
Equation (8.9) may be simplied by introducing:
FcE
a 8:13
Fc*
Then
s
1a 1a 2 a
CP 8:14
2c 2c c
Example 8.5 Find the maximum allowable load on a 10.0 ft high 6 6 shore post
using Douglas Fir-Larch (North) No. 2 grade material.
Solution
e =d 10 12=5:5 21:8 < 75 o:k:
805
a 1:31
616
s
s
Example 8.6 Design the formwork for the wall in Example 8.3. The walls face
will be exposed to view, hence a maximum allowable deection for the sheathing
and the studs of span/360 is desired. Use in. Structural I Plyform for the
538 8 Formwork for Monolithic Concrete Construction
sheathing and Douglas Fir-Larch - North #2 for the studs and wales. The maxi-
mum internal formwork pressure of 720 psf was found in Example 8.3. Assume
dry service conditions for the design of all formwork elements, i.e., CM 1.0.
Design the necessary lateral bracing for the ACI recommended 15 psf minimum
wind load on the formwork, or the alternate lateral load of 100 lb/ft. (See
Figure 8.5).
Solution
1. Sheathing design
The sheathing is pre-selected in the problem statement, hence the rst element of
the design requires the nding of the stud-spacing, i.e., the maximum span that the
sheathing can safely span. Since stud spacing rarely exceeds 24 in., the sheathing
will span over three or more spans.
Three criteria must be satised: bending strength, shear strength and deection.
Step 1. Using the sheathing to span horizontally between the studs, the
section properties of the in. Structural I Plyform from Table 8.4 are:
I 0.202 in.4/ft
KS 0.464 in.3/ft
Ib/Q 6.189 in.2/ft
Step 2. The Material Properties from Table 8.3:
Fb 1,930 psi
Fs 102 psi
E 1,650,000 psi
Step 3. From Table 8.11 for 3-span condition, the maximum allowable span for the
sheathing:
(a) Based on bending:
r r
120Fb0 S 120 1,930 0:464
Lmax 12:2in:
w 720
Ib
20Fv0
Q 20 102 6:189
Lmax 2d 2 0:75 19:0 in:
w 720
2. Studs design
In most wall forming the studs are 2 4 (rarely 2 6) lumber, because the
spacing of the wales is limited by the form-tie layout. So the design of the studs,
after pre-selecting the species and the size of the lumber, is simplied into
nding the maximum allowable span of the studs between the wales,
i.e., nding the wales spacing.
Step 1. Select 2 4 and nd its section properties from Table 8.5:
A 5.25 in.2
I 5.36 in.4
S 3.06 in.3
Step 2. Find the material properties applicable to Douglas Fir-Larch (North)
No. 2 grade lumber.
(a) From Table 8.6, reference design bending stress: Fb 850 psi.
Applicable adjustment factors from Tables 8.8 and 8.9
CF 1.50 and CD 1.25
Thus
Fb0 850 1:25 1:50 1,594 psi
(b) From Table 8.6, reference design shear stress: Fv 180 psi.
Applicable adjustment factors:
CD 1.25
Thus
Fv0 1:25 180 225 psi
(c) Modulus of elasticity: E 1,600,000 psi
No adjustment factor applies.
Step 3. Find the loads on the stud.
Since the studs are 12 in. c/c, the maximum load on the studs is 720 lb/ft.
Step 4. As stated above, in nding the maximum wales-spacing, based on the
strength of the studs, 3-span condition will apply. Using the formulas
listed in Table 8.11, the maximum allowable span for the studs is:
r r
120Fb0 S 1,594 3:06
Based on bending: Lmax 120 28:5in:
w 720
Based on shear:
13:33Fv0 A 13:33 225 5:25
Lmax 2d 2 3:5 28:9 in:
w 720
Based on deection:
r r
3 EI 3 1,600,000 5:36
Lmax 1:69 1:69 38:6 in:
w 720
The maximum wale spacing is governed by bending. For practical
purposes select 24 in. spacing.
540 8 Formwork for Monolithic Concrete Construction
Step 5. Check the bearing stress (compression perpendicular to the grain) at the
interface of the stud and the wales.
The reaction force on the stud equals (the pressure) (stud
spacing) (wales spacing). Thus:
R 720 1 2 1,440 lb
Wales are assumed to be made of two 2 4s.
The bearing area A 2 1:5 1:5 4:50 in:2
1,440
The bearing pressure fbrg 320 psi
4:5
The reference design stress from Table 8.6 is 625 psi. The applicable
adjustment factor from Equation (8.6) is
Based on shear:
Based on deection:
r r
3 EI 3 1,600,000 10:72
Lmax 1:69 1:69 38:6 in:
w 1,440
Step 2. Checking the bearing stress at the interface of the tie anchorage (wedge
or other device) is also required. As it was mentioned before, ties and
their anchorages are proprietary items and thus the bearing area at the
surface of the wales and the selected anchorage becomes known only
after the selection of the tie system.
542 8 Formwork for Monolithic Concrete Construction
If, for example, the anchor wedge is 3.50 in. high and 1.5 in. wide,
then the bearing area equals 2 1:5 in: 1:5 in: 4:50 in:2 and
2,880
fbearing 640 psi
4:50
The reference design stress from Table 8.6 is 625 psi. The applicable
adjustment factor from Equation (8.6) is b 1:5 in:
1:5 0:375
Cb 1:25
1:5
Thus:
0
Fc 625 1:25 781 psi > 640 psi o:k:
a b
100 lb / ft
4'-0"
4'-0"
H H
str
str
ut
ut
for
for
15 psf
ce
ce
V V
8'-0"
8'-0"
6'-0" 6'-0"
V V
Figure 8.6 Minimum lateral loads and brace forces: (a) Lateral wind pressure (15 psf), (b)
Horizontal force (100 lb/ft)
8.6 Wall Formwork Design 543
The effect of the wind load to the base of the 12 ft high formwork per ft of
wall is:
W 15 12 180 lb=ft
12
Its overturning moment to the base is M 180 1,080 ft-lb=ft
2
This moment needs to be kept in equilibrium by the horizontal component of
the strut force (H) by providing a balancing moment with its moment arm 8 ft.
Hence:
And the force in the strut (from similar triangles-see Figure 8.6) is:
The effect of the alternative 100 lb/ft horizontal force to the base of the 12 in.
high formwork of wall is:
Since this is greater than the force from the 15 psf wind-load, we will use this
as the basis of designing the bracing strut.
When we follow the ow of forces, we nd the following: the 100 lb/ft force at
the top creates bending in the studs, which in turn create bending in the double
wale at the level where the strut is connected. The available bending capacity of
those wales then is used to determine the strut spacing.
1. Bending in the studs:
The cantilever length of the studs is 4 ft as shown on Figure 8.6(b). They are
spaced at 1 ft centers, hence the bending moment in each stud will be
544 8 Formwork for Monolithic Concrete Construction
400 12
f 1,569 psi
3:06
100 lb=ft 12 ft
H 150 lb=ft
8 ft
Based on shear:
The bending strength governs. Select the strut spacing as 60 -000 (72 in. <
79 in. ok).
3. Design of the lateral bracing struts:
Since the struts are spaced at 6 ft, the horizontal component of the force in a
strut will be:
H 150 6 900 lb
8.6 Wall Formwork Design 545
The strut-force is
C 250 6 1,500 lb
Try 2 4 Douglas Fir-Larch (North) #2 Grade. (1.5 in. 3.5 in. dressed
sizes or true dimensions.)
Step 1. Check
for maximum allowable slenderness.
e 10 12
80 > 75 N.G.
d1 1:5
Introduce a secondary brace (as shown on Figure 8.6) to the midpoint
of the strut, thereby reducing its slenderness by half in the weak direc-
tion. Additionally it will help to reduce the deection of the strut due to
its self-weight. Thus:
e 5 12 e 10 12
40 and 34:3
d1 1:5 d2 3:5
Step 2. From Table 8.3 the allowable stresses and the modulus of elasticity for the
specied Plyform Structural I sheathing material are:
Allowable bending stress Fb 1,930 psi
Allowable rolling shear stress Fs 102 psi
Modulus of elasticity E 1,650,000 psi
Step 3. From Table 8.4 the section properties of the specied sheathing in its weak
direction are:
8.7 Column Formwork Design 547
Step 5. It appears that the rolling-shear-strength will govern the design. However,
the listed formulas used in the calculations are based on the theoretical
knife-edge (or at least narrow width) supports concept. The at 2 4
battens are not such, but will provide support to the plywood in a wider
zone than the center-to-center span dimensions would imply. Thus, the use
of 6.5 in. span is quite conservative.
After a bit of trial and error type of deliberations, the batten layout
shown on Figure 8.7 was adopted.
1 3/4"
6"
6"
6"
1 3/4"
1 3/4"
Figure 8.7 Batten layout for the concrete column formwork of Example 8.7
548 8 Formwork for Monolithic Concrete Construction
The reference design stresses for Douglas-Fir (North) No. 2 Grade mate-
rial from Table 8.6 are:
6:5 in:
w1 1,575 853:1 lb=ft
12 in:=ft
8.7 Column Formwork Design 549
Then using the Three or more spans case from Table 8.11:
r r
120Fb0 S 120 1,753 1:31
Based on bending: Lmax 18:0 in:
w 853:1
Based on shear:
Based on deection:
r s
4 EImax 4 1,600,000 0:98 1
Lmax 6:46 6:46 16
21:1 in:
w 853:1
Bending strength governs. Select this spacing of 18 in.. Thus the second
clamp from the bottom is at (6 in. + 18 in.), i.e. 20 000 . Thus the distance
from the top to the second clamp from the bottom is:
z2 11 2 9 ft
and
and
6:5 in:
w2 1,350 731:3 lb=ft
12 in:=ft
Using this load to calculate the maximum spacing between the second
and third clamps, we obtain the following:
r r
120Fb0 S 120 1,753 1:31
Based on bending: Lmax 19:4 in:
w 731:3
Based on shear:
Based on deection:
v
u 1
r u
t1,600,000 0:98 16
4
4 EImax
Lmax 6:46 6:46 22:0 in:
w 731:3
550 8 Formwork for Monolithic Concrete Construction
z3 11 3:5 7:5 ft
and
and
6:5 in:
w3 1,125 609:4 lb=ft
12 in:=ft
Using this load to calculate the maximum spacing between the third and
fourth clamps, we obtain the following limitations:
r r
120Fb0 S 120 1,753 1:31
Based on bending: Lmax 21:3in:
w 609:4
Based on shear:
Based on deection:
v
u
r u 1
u1,600,000 0:98
4
4 EI max t 16
Lmax 6:46 6:46 23:0 in:
w 609:4
Example 8.8 Design the formwork elements: decking, joists, beams and shores
required for an 8 in. thick at slab oor. The bay sizes are 240 000 240 000 , the
oor to oor height is 12 ft. Assume normal weight concrete. Use in. thick Class
II Plyform decking (face grain perpendicular to the joists) and No. 2 Grade Hem-Fir
sawn lumber for the joists, beams and shore-posts. Assume dry service condition for
8.8 Floor Slab Formwork Design 551
6"
4x21"= 7'-0"
11'-0"
clamps
2x18"= 3'-0"
toe
6"
the design of the joists and the beams (stringers). Limit the maximum allowable
deection to L/360 in the decking, joists and beams. (Refer to Figure 8.9)
Solution
Step 1. Find the loads
Step 2. Find the section properties and the allowable stresses for the selected
decking material. This will enable the selection of an appropriate joist
spacing.
The section properties of in. Class II Plyform from Table 8.4
Moment of inertia I 0.198 in.4/ft
Section modulus S 0.454 in.3/ft
Rolling shear constant Ib/Q 6.631 in.2/ft
552 8 Formwork for Monolithic Concrete Construction
be
sp a
ac m
ing
ing
eck
o dd
wo
ply
st
joi ing
p ac
s
x-bracing
sh st
ore po
o re ing
pe spac post sh pac el
rpe in s rall s
to ndic g pa beam
be
am ular i t h
s w
19:2
w 160 256 lb=ft
12
The maximum allowable spans for the joists from Table 8.11 (assuming
three span conditions) are:
(a) Based on bending:
r r
120Fb0 S 120 1,594 3:06
Lmax 47:8 in:
w 256
While the maximum allowable length obtained from the governing crite-
rionbendingis a tiny bit less than the planned span of 48 in., the
difference amounts to less than %, the use of the 2 4 joists at 19.2 in.
spacing and 48 in. span can be accepted.
Step 5. Beam (Stringer) design.
As shown in Figure 8.9, the beams support the joists. We may approach this
problem either by assuming a size and then calculate the maximum allow-
able span, which is the same as nding the spacing of the shores under the
beams, or arbitrarily assume a shore spacing and then nd an acceptable
beam section. We will select the rst route, as it is the easier approach. In
the given bay size (24 ft) the beam spacing (which is based on the joists
spans) of 4 ft can provide an orderly layout of the shore-posts in one
direction.
Hence, we will select a trial beam size of 4 6, No. 2 Hem-Fir. This is
still small enough to likely be available in length sufcient for 3-span
condition.
The reaction forces from the joists do not produce a uniformly distrib-
uted loading pattern on the beams. However, since the joists are closely
spaced, it is customary to assume an equivalent distributed load derived
from the total loads.
Thus, the design loads on the beams will be the (psf oor loads) (beam
spacing).
The maximum allowable spans for the beams from Table 8.11 (assuming
three span conditions) are:
r
120 1,381 17:65
(a) Based on bending Lmax 67:6 in:
640
13:33 187 19:25
(b) Based on shear: Lmax 2 5:5 86:0 in:
640
r
3 1,300,000 48:53
(c) Based on deection: Lmax 1:69 78:1 in:
640
Studying the results we realize that 67.6 in. is a somewhat awkward
value, when we try to t it somehow into the 24 ft bay spacing. One simple
solution is to space the shores under the beams at 4.8 ft (57.6 in.), which
will produce ve rows of shores in a 24 ft bay.
We select this as the best compromise.
Step 6. Check the supporting shore-posts using 4 4 Hem-Fir No. 2 material.
From the previous calculations, each shore post has a tributary area of
Thus the load is: 19.2 ft2 160 psf 3,072 lb.
Since the oor-to-oor height was given as 12 ft, with reference to
Figure 8.10, the unsupported length of the shore post is:
e 126:25
36:1 < 75 o:k:
d 3:5
556 8 Formwork for Monolithic Concrete Construction
concrete:....... 8.00"
plywood:....... 0.75"
joists:............ 3.50"
beams:.......... 5.50"
total:............. 17.75"
shore post
Figure 8.10 Section through the formwork for the concrete oor slab in Example 8.8
The reference design values for compression parallel with the grain for
the selected No. 2 Hem-Fir material from Table 8.6:
The 44 shore post is safe for the load and height.
Beam forms can be developed in many different ways, depending on how the
formwork designer decides to integrate it within the slab forms. Figure 8.11
shows only one possible formation, however the issues that are associated with
the design of the elements involved therein are quite similar to any other type of
formwork layout for beams. Figure B8.3 in Appendix B shows the forming of
reinforced concrete beams and slabs.
The important items that will require our attention are the design of: (1) the
bottom form; (2) the side forms; (3) the runner joists; and (4) the kicker. [Refer to
Figure 8.11 for the identication of these elements.] The design (or checking) of the
other elements should be familiar by now, if the reader already studied the oor-
slab form design in Example 8.8.
4'-0" 4'-0"
ledger
side sheathing
bottom sheathing blocking
chamfer strip kicker
if required
runner joists
beams
x-braces at
selected
locations.
shore posts
2'-6"
Example 8.9 Design the formwork for a 16 in. wide, 24 in. deep beam as shown in
Figure 8.11. Slab is 6 in. thick. Use Class I Plyform for the sheathing and Grade
No. 2 Hem-Fir for the dimensional lumber. Assume dry service condition. Limit the
maximum deection of bottom sheathing to L/360, and the kickers to 1/32 in.
Solution
Step 1. Bottom sheathing design.
The bottom form needs to be designed to support the weight of the fresh
concrete above it. (24 in.)
24 in:
w 150 pcf 300 psf
12 in:=ft
Note: No construction Live Load is used for the design of the beam
bottom, for there is highly unlikely within the connes of the small space of
the beam forms. One may wish to add here the self-weight of the ply-
woodapproximately 3.3 psfbut it can be safely neglected in the design
of this element.
The bottom forms are usually cut in such a way that the plywood bends
in its strong direction. As it is shown in Figure 8.11, this bottom form spans
between the runner joists, which in turn are supported by beams. Assuming
a minimum 3-span condition, we will try to work with the thickest Plyform
decking available, 11/8 in. thick.
The section properties, i.e. the moment of inertia, the effective
section modulus and the rolling shear constant respectively from Table 8.4:
I 0:554 in:4 =ft KS 0:849 in:3 =ft Ib=Q 10:43 in:2 =ft
The material properties, i.e. the allowable bending stress, the allowable
rolling-shear stress and the modulus of elasticity from Table 8.3:
From Table 8.11 the maximum allowable spans for the 11/8 in. thick
bottom is the least of
r r
120Fb0 S 1201,9300:849
(a) Based on bending: Lmax 25:6 in:
w 300
(b) Based on shear:
20Fv0 Ib
Q 20 72 10:43
Lmax 2d 2 1:125 52:3 in:
w 300
8.9 Beam Formwork Design 559
Based on these results we select 24 in. span for the bottom, which
will be used as the spacing of the runner joists.
Step 2. Design the side forms.
The most conservative assumption is that the side-forms span vertically
from the base to the sheathing under the 6 in. slab (24 in. 6 in. 18 in.) in
our case, and the bending is in the weak direction of the plywood. The
lateral pressure from the fresh concrete will not be uniform, but vary with
the depth as shown on Figure 8.12.
75 psf
R1=112.5 lb / ft
18" (1.50')
R2=168.8 lb / ft
300 psf
Try in. plyform. The section properties of the assumed in. Class I
Plyform from Table 8.4, bending perpendicular to grain:
I 0:092 in:4 =ft KS 0:306 in:3 =ft Ib=Q 4:063 in:2 =ft
(a) Check for bending from the calculated Mmax 52.7 lb-ft
M 52:7 12
fb 2,067 psi > 1,930 psi N:G:
S 0:306
I 0:151 in:4 =ft KS 0:422 in:3 =ft Ib=Q 6:028 in:2 =ft
M 52:7 12
Re-check for bending: fb 1,499 psi < 1,930
S 0:422
psi o:k:
(b) Check for deection using the simplied assumption that the loads are
uniform:
From Table 8.10 for simple span
300 75
5 w L4 5 2 1:5 124
max
384 12 EI 384 12 1,650,000 0:151
0:0857 in:
1
150 12:5 psf
12
for each inch above the bottom, the lateral pressure at the location of
one inch above the bottom is 300 12.5 287.5 psf. Assuming the
critical section to be 1 in. (plywood thickness 7/8 in.), then
300 287:5 1
Vcrit 168:8 144:3 lb=ft
2 12
V 144:3
fs 24 psi < Fs 72 psi o:k:
Ib 6:028
Q
1,680
=1,120 plf
1.5
1'-6"
840 lb
840 lb
2'-6"
Figure 8.13 Load distribution on a runner joist supporting the beam formwork
Fb0 850 1:25 1:50 1,594 psi and Fv0 150 1:25 187 psi
Fb0 S 1,594 3:06 4,877 lb -in: 406:5 lb-ft < 735 lb-ft N:G:
Ma 1,594 6:12 9,755 lb-in: 813 lb-ft > 735 lb-ft o:k:
3 V 3 840
fvmax 120 psi < Fv0 187 psi o:k:
2 A 2 2 5:25
5 Mmax L2
max
48 EI
5 735 12 302
max 0:059 in:
48 1,300,000 2 5:36
L=360 30=360 0:083 in: > 0:059 in: o:k:
Problems
Assume minimum three span condition with the face grain perpendicular to
the supports. The maximum allowable deection in the sheathing is L/360.
8.4. A 9 ft high 20 in. 20 in. column form is planned. The form will be built from
in. Class I Plyform material, face grain vertical, without battens. Calculate
the required clamp spacing for the rst three clamps above the base. The
maximum deection between the clamps may not exceed 1/16 in.
8.5. Calculate the allowable safe load on a Spruce-Pine-Fir No. 2, 4 4 shore
post, if the laterally unsupported height is: (a) 80 000 and (b) 120 000 .
8.6. A 10 in. thick oor slab using normal weight concrete is formed with 23/32
in. thick Class I Plyform decking on 2 6 Douglas Fir Larch (DFL) (North)
No. 2 joists. The face grain will be perpendicular to the supports. The
maximum allowable deection of the decking may not exceed L/360. Calcu-
late the allowable joist spacing.
8.7. A 10 in. thick oor slab using normal weight concrete is formed with
plywood over 2 6 Douglas Fir-Larch (North) No. 2 joists at 16 in. center-
to-center. Assume 2-span condition. Calculate the maximum allowable spac-
ing of the beams made up from Double 2 10 DFL No. 2 material. The
maximum allowable deection of the joists may not exceed L/360.
8.8. Design the bottom form for a 24 in. wide by 30 in. deep beam using 1 in. thick
Class I Plyform. Assuming three span condition, calculate the allowable
spacing of the runner joists. The maximum allowable deection of beam
bottom form between the runner joists is 1/16 in.
8.9. Design the formwork components for a 10 ft high, 12 in. thick wall.
Assume Cw 1.0, Cc 1.0, temperature 75 F and 4 ft/hour rate of pour.
Materials will be 23/32 in. Class I Plyform for the sheathing and Construc-
tion Grade Hem-Fir 2 4 studs, double 2 4 s for wales. Limit the
deection of all individual components to L/360. Calculate the maximum
tie force. Make a sketch of your design identifying all your selections.
8.10. The wall form described in Problem 8.9 is planned to be braced by 4 4
Construction Grade Hem-Fir struts, the top end attached at 9 ft above the base
and slopes 45 . The spacing between the braces will be 8 ft. Check the
adequacy of the brace for a minimum 15 psf wind-load, or alternately for
ACI 347 recommended 100 lb/ft horizontal distributed load at the top of the
forms.
Self-Experiments
Experiment 1
Following the details shown in Figure 8.5, construct the formwork for a 24 in.
high 3 in. thick concrete wall. Calculate the required thickness of the plywood
sheathing and the size of the studs, wales, ties and struts. Remove the forms 2 days
after the pour. Discuss the various stages of this task and the challenges you were
faced with.
Self-Experiments 565
Experiment 2
Following the details shown in Figures 8.3 and 8.4, construct the formwork for a
8 in. 8 in. 24 in. high concrete column. Construct the forms without battens.
Calculate the required thickness of the sheathing, size and spacing of the clamps.
Remove the forms 2 days after casting the concrete. Discuss the various stages of
the form construction and concrete placement along with any challenges that you
were faced with.
Chapter 9
Overview of Prestressed Concrete
9.1 Introduction
Concrete has a considerable compressive strength, but its tensile strength is quite
limited. Thus, designers use reinforcement in conjunction with concrete to make
useful elements in buildings.
Early on, researchers realized that tensile stresses could be eliminated in con-
crete structures by adding sufcient compressive stresses to balance them out. Then
the element would have a stress distribution throughout that consisted of
compression only.
Figure 9.1 shows a simply-supported beam with an applied distributed load
w 800 lb/ft. The cross section of the beam is b 12 in. and h 18 in. If this
beam is made of a homogeneous elastic material, the stresses can be calculated as
follows:
0:8202
Mmax 40 kip-ft
8
12182
Sm 648 in:3
6
40 12,000
fmax 741 psi
648
This value represents the compression at the top edge of the section, and a similar
magnitude of tension at the bottom edge. Based on the discussions on the tensile
W 800 lb/ft
n.a. h 18 in.
20'-0"
12 in.
P P
20'-0"
The force P would produce uniform compressive stresses over the cross section.
The magnitude of the force P needed to eliminate any tension on this beam could be
determined as follows:
P
f 741 psi
A
Then
P 741 12 18 160,000 lb
n.a.
741 741 0
Figure 9.3 Stresses in the beam: (a) from force P, (b) from bending, (c) superimposed
n.a.
Now consider what will happen if the initial prestressing force is moved down-
ward at an eccentricity e from the centroid. As discussed in Chapter 5, an eccentric
force has the same effect as a concentric force plus a moment. See Figure 9.5. The
force P still causes a uniform compression on the section; however, the M P e
moment will cause tension at the top edge and compression at the bottom edge. So
how large an eccentricity is needed so that the tension from the moment and the
compression from the force P at the top edge cancel out each other? To answer this
question we must rst algebraically make the stresses due to the force equal to the
stresses due to the moment. Mathematically this is expressed as follows:
P M P Pe
0 or 0
A Sm bd bd 2
6
P
e
P M P e
Returning to the numerical problem: Assume e 3 in. and determine the amount of
eccentric prestress force, P, is needed to have zero tension after the application of
uniform loads of 800 lb/ft to the beam. The equation expressing this condition is:
P P e Mmax 12,000
0
A Sm Sm
P Pe
where is the compressive stress caused by the force P, is the compressive
A Sm
Mmax 12,000
stresses caused by the eccentricity of the force P, and is the tensile
Sm
stress caused by the maximum moment. P is assumed to be in lb, and Mmax is in
ft-kip. (Note that a positive value designates compressive stress and a negative
value designates tensile stress.)
Substituting e 3 in., A 12 18 216 in2, Sm 648 in3, and Mmax 40 kip-ft
and solving for P:
P 80,000 lb
This value is only half that needed to achieve the same result when the P force was
concentrically applied. So, providing a well-selected eccentricity to the prestressing
force can drastically reduce the magnitude of the force and still have the same effect
as a concentrically applied force.
Although the tensile strength of concrete is small compared to its compressive
strength, the ACI Code (Section 24.5.3.2) allows the section initially (at the time the
p
prestressing force is applied) to have tensile stress equal to ft 3 f ci0 : Here fci0 is
the specied compressive strength of the concrete at the time the prestress is
applied. This value is usually smaller than the nal design strength of the concrete,
as prestressing is usually accomplished before the concrete is completely cured.
p
Similarly, the ACI Code permits a tensile stress of ft 7:5 fc0 under full service
load condition in most applications.
For the nal introductory example we examine the use of the allowable tensile
stress at the bottom of the section at midspan. Assume that fc0 5,000 psi: Then the
p
beam can have ft 7:5 5,000 530:3 psi of tension in the nal service load
condition. Using e 3 in., we obtain the following equation for the service load
condition at midspan after the substitutions:
P P 3 40 12,000
530:3 psi
216 648 648
(Note: The negative sign in front of 530.3 indicates tension.) Solving for P:
P 22,728 lb
This is the amount of prestress force that the beam needs at service load condition.
9.2 Advantages of Prestressed Concrete Structures 571
Table 9.1 Recommended maximum span-depth ratios for prestressed oor structures with mod-
erate live loads
Single span Continuous spans
(oor) (oor)
Prestressed hollow core slabs 36 N.A.
Prestressed double tees 32 N.A.
Posttensioned one-way solid slabs 44 48
Posttensioned solid slab cantilevers 18 N.A.
Posttensioned at plates (supported on N.A. 45
columns)
Posttensioned wafe slabs (supported on N.A. 35
columns)
572 9 Overview of Prestressed Concrete
9.3.1 Pretensioning
elements in long (up to 600 ft) casting beds, which permits the simultaneous
fabrication of many elements with a single tensioning of the strands. A large saw
is used to cut each element to its individual required length. Accelerated curing
techniques permit the release of the strands in only 1618 hours after the placement
of the concrete; thus, a 24 hours manufacturing cycle can be maintained. The
production is highly mechanized and provides great quality control. Many standard-
ized proles are available in catalog form for spans and loading capacities. These
can be readily called out and specied by the designer. Figures 9.9 to 9.13 show
some typical proles that are popular in building construction.
Hollow core decks are a popular precast and prestressed building element. They
are typically available in standard depths of 6, 8, 10, and 12 in., and are
manufactured mostly in 40 000 width, although some manufacturers may supply
them in 20 000 or 80 000 widths. The shape of the cores may also differ from the
circles shown on Figure 9.9, as different patented manufacturing processes are used
to form them. Not all building designs can use the standard widths, so narrower
ller panels are made by slicing the panels lengthwise. Panels can also be cut at an
angle to accommodate supporting beams or girders that are not perpendicular to the
span of the panels.
Grout key
A grout key is formed at the sides of the individual decks. As the name indicates,
the formed keyway is grouted solid after the erection of the panels. The keyways
prevent individual panels from deecting differently after the grout cures. Thus, a
kind of lateral load transfer occurs if one of the oor panels is loaded much more
than its neighbors.
The top surface of hollow core decks is not smooth enough for oor structures,
which receive nish materials like tiles or carpets. Thus, in those types of applica-
tions, the decks usually receive a 2 in.-thick (nominal) concrete topping (see
Figure 9.10) that can be nished to the desired atness and smoothness. The
concrete topping bonds to the surface of the decks and becomes a composite part
of the whole. It also makes the oor thicker, and consequently considerably stiffer
than one made of untopped decks.
Concrete topping
4'-0"
8'-0"
10'-0"
12 to 36 in.
2'-0"
4'-0"
5'-0"
Double tees are quite light; the top slab is typically 2 in. thick only at the outer
edges. They are very economical for covering large spans (spans of 100 ft or more
are not uncommon). Such long elements, however, are difcult to transport in a
tight urban environment. Double tees are rather exible, so their camber or deec-
tion under load can be signicant, especially on longer spans. So the designer must
carefully plan the interface of double tees with other building elements. A double
tee on an 80 or 90 ft span element may have an initial camber of 34 in. (or more),
with perhaps a similar magnitude of deection under loads. That much movement
requires very careful consideration of details, such as when the design contemplates
a window-wall parallel with the span of a double tee.
Figures 9.12 and 9.13 show some other typical precast and prestressed elements
that are frequently used to support oor elements in building construction. The
shapes shown here are just a few of the many different shapes readily available to
the designer. Figure B9.1 in Appendix B shows prestressed I-Section beams.
9.3 Types of Prestressing 575
Bearing pad
Figure 9.13 Inverted T-beam and L-beam supporting hollow core decks
9.3.2 Posttensioning
Posttensioning has one vast advantage over pretensioning: It usually uses curved
tendons. This enables the designer to change the location of the prestressing force
from section to section along the length of the structure.
Figure 9.14 shows a single span beam with the strands draped in a parabolic
form. After tensioning, the eccentricity of the prestressing force to the neutral axis
is zero at the ends and maximum at the center. The moment resulting from the
prestressing causes compression at the bottom and tension at the top. The maximum
of these forces occurs at midspan, as illustrated in Figure 9.14. In fact, the forces
closely balance the effects of the gravity loads on the beam.
9.4.1 Concrete
The earliest experiments with prestressing failed mainly because they were
performed with ordinary steels (yield strength in the range of 3640 ksi). All, or
almost all, of the prestress was lost with the passage of time, due to a series of
contributory reasons. (Loss of prestress in normal applications may amount to
2535 ksi, or higher.)
The use of very high-strength steel wires helped to solve this problem, as these
wires, even after the considerable prestress losses, retained sufcient stress levels.
Although many proprietary prestressing (mostly post-tensioning) systems use
large-diameter, high-strength bars, most systems employ prestressing strands
manufactured from cold-drawn wires conforming to ASTM A421. Usually six
wires are wound tightly around a seventh (and usually slightly larger-diameter)
wire into a uniform pitch helix, as illustrated in Figure 1.21. The pitch is 1216
times the diameter of the wires. After manufacture, the strands are put through a
stress-relieving heat treatment to make them conform to the requirements of ASTM
A216, Standard Specications for Uncoated 7-Wire Stress-Relieved Strand for
Prestressed Concrete. They are also prestretched to increase their apparent mod-
ulus of elasticity. The strands may be manufactured in Grade 250 or in Grade
270, the numbers referring to the minimum ultimate strength of the strand in ksi.
Table 9.2 shows the properties of Grade 270 strands.
Figure 9.16 shows a typical stress-strain curve for prestressing strands. The
strands, unlike normal reinforcing steel, do not have a dened yield. They remain
elastic up to about 85 % of their ultimate strength. An arbitrary yield point is often
used for specication purposes. ASTM A216 requires a minimum value of 0.85fpu
at 1 % extension (or strain), where fpu is the minimum guaranteed ultimate strength.
By this denition, the yield for Grade 270K strand may be taken as
0.85 270 230 ksi.
The nal service level stresses in the prestressing strands will be signicantly lower
than they were at the time of the initial stressing. The contributory causes for this
prestress loss are numerous. The ve major ones are discussed below.
578 9 Overview of Prestressed Concrete
Stress
Prestressing strand
Strain
When prestress is applied to concrete (i.e., when the prestressing strands are
released in the pretensioning type of application), the concrete shortens due to the
compressive stresses transferred to it. As the concrete shortens, so do the strands
bonded inside the concrete. This shortening lessens the stress in the steel, and
correspondingly lessens the compression on the concrete. How much the concrete
member shortens depends on the concretes modulus of elasticity, which in turn
depends on the concretes strength at the time the prestress was applied. The higher
the strength of the concrete, the lesser the loss due to elastic shortening. In
posttensioning, however, very little loss is caused by the elastic shortening; as the
stressing and the shortening of the concrete take place simultaneously; so when the
force in the strand is measured, the change in the concretes length has already
taken place. (This is not exactly true, because as the tendons are pulled one by one,
each stressed and anchored tendon will lose some of its stress when its neighboring
tendon is stressed).
If the concrete shrinks due to loss of moisture after prestressing, the shrinkage will
shorten the member. Correspondingly, the stretched strands will also shorten by the
9.5 Loss of Prestressing 579
same amount. This shortening leads to a loss of stress in the strand. Nearly 80 % of
the shrinkage takes place in the rst year of life of the structure. The magnitude of
shrinkage depends on many variables, but it can be estimated to a reasonable degree
of accuracy based on experimental data.
Friction loss occurs only in posttensioning. Figures 9.14 and 9.15 show typical
paths of curved tendons. Sometimes posttensioned at plates use tendons that curve
in the horizontal direction to accommodate oor openings or ducts. As a result,
when a tendon is pulled from one end, it leans against the duct. Figure 9.17
illustrates the problem.
If the tendon is pulled from the right, P2 will be less than P1. The difference will
be the loss due to the frictional resistance at the contact surfaces. The loss depends
on the radius of curvature and the friction coefcient between the tendon and the
duct. Sharper curves and larger friction coefcients will result in larger loss of
prestress. In other words, the force in the tendon at locations away from the live end
P2
P1
Figure 9.17 Friction between the tendon and the duct in posttensioning
580 9 Overview of Prestressed Concrete
will be less than that measured at the stressing jack. In addition to the intentional
curving of the ducts, an unintentional curving also takes place during the concreting
operation. This is referred to as wobble. The so-called wobble (or wobble friction)
coefcient accounts for this curving, but is a rather vague value. In the past, the
ACI Code used to recommend a value between 0.0003 and 0.0020 for the wobble
coefcient of 7-wire pre-greased strands, and 0.05 to 0.15 for their curvature
friction coefcient. Values of the wobble and curvature friction coefcients to be
used for the particular types of prestressing reinforcement and ducts should be
obtained from the manufacturers of the tendons.
Equation (9.1) gives the formula for calculating frictional losses:
Kpx p px
Ppx Ppj e 9:1
where
Example 9.1 Figure 9.18 shows a curved tendon between two inection points in
an 8 in.-thick slab. The inection points are 15 ft apart. Between these two points
the tendon rises 2.75 in. Assume K 0.001 and p 0.10. Find the prestress loss
within these two points.
8 in.
15 ft
Solution We use the given data to calculate the radius of curvature, which is
122.8 ft. The value of px is then equal to 7 0.122 radians. Substituting these
values into Equation (9.1) we obtain the prestress loss due to friction between the
two points as a function of the prestressing force.
9.6 Ultimate Strength 581
This means that the loss of force along the curve is 2.69 %. In multiple curves the
losses will combine, and may become very signicant depending on the number of
spans.
Several techniques can be used to mitigate the frictional losses. One of them is to
stress the tendons from both ends. After the tendon has been stressed from the live
end, that end is anchored and the tendon is restressed by pulling from the former
dead end. Other techniques involve stressing very long tendons one section at a
time, coupling the next section to the already stressed segment.
Calculating the total losses of prestress is a very complex problem. Even with the
best available research information, we can obtain only approximate values. The
values of the parameters that inuence the loss from any of the major sources are
only approximate ones. The values listed in Table 9.3 quite closely approximate the
true average prestressing losses (at least in building construction).
For typical 270K strands that are initially stressed to the ACI Code-
recommended value of 0.7fpu for post-tensioning applications and 0.8fpu for
pretensioning applications, a good estimate of losses from volumetric changes for
average conditions (not including friction losses) is 1820 % for pretensioning, and
1516 % for posttensioning. This translates to the average loss values shown in
Table 9.3.
Prestressed elements are designed to limit stresses at service load conditions. The
ACI Code, however, also requires that prestressed elements satisfy ultimate
strength requirements as well.
The ultimate moment strength is calculated using equations similar to those used
with ordinary reinforcement. But these familiar equations substitute a calculated
(somewhat ctitious) yield value for fy in the calculations. This yield value, fps, can
be calculated using Equations in Section 20.3.2 of the ACI Code.
582 9 Overview of Prestressed Concrete
This introductory chapter on prestressing includes only the formulae that deal
with unbonded tendons. The reader should consult the ACI Code for further
information on the ultimate strength calculations for bonded tendons.
Thus, Equation (9.2) (ACI Code Section 20.3.2.4.1) can be used to calculate the
yield strength for members with unbonded tendons and with a span/depth ratio of
35 or less:
fc0
f ps fse 10,000 9:2
100p
and Equation (9.3) (ACI Code Section 20.3.2.4.1) is applicable for members with
unbonded tendons and with a span/depth ratio greater than 35:
fc0
f ps fse 10,000 9:3
300p
where
If the design uses unbonded tendons, the ACI Code (Sections 7.6.2.3 and 9.6.2.3)
require the addition of a minimum amount of bonded normal reinforcement. This
minimum amount is given below:
As 0:004Act
where Act is the area of the concrete section between the tension face and the
centroid of the gross concrete section.
The ACI Code also permits the use of nonprestressed reinforcing to help with the
required ultimate strength. The detailed discussion of this subject is beyond the
scope of this text.
In the simple numerical examples given in Section 9.1, the analyses of prestressed
sections were conducted by nding a sufciently large concentric, or eccentric,
prestressing force that eliminated, or greatly reduced, unwanted large tensile
9.7 The Concept of Load Balancing 583
stresses from the applied loads. This procedure is sometimes referred to as super-
position (i.e., adding axial compressive stresses to the ones caused by exure).
In posttensioning there is a much more easily visualized method of analysis. The
load balancing method, introduced by T. Y. Lin in 1963, is the most widely used
and most powerful analytical tool for the design of prestressed structures. This
method provides prestressing by using a system of stressed tendons selected to
impose loads on the element in opposition to the acting gravity loads; hence, the
name load balancing method.
This concept is illustrated in Figure 9.19. If the path of the tendon is a parabola,
its effect, after stressed on the simple span beam, is equivalent to an upward acting
uniform load. The balancing loads, wbal, can be calculated using simple statics as
follows. The moment caused by the horizontal component of the prestressing force,
as shown in Figure 9.19b, is (P cos). This prestress moment must balance the
moment caused by acting loads. Mathematically this is expressed as:
wbal 2
P cos P
8
Thus
wbal 2
P 9:4
8
P cos P cos
b
P sin P sin
P wbal P
P cos P cos
Figure 9.19 (a) Balancing uniform load, (b) free body diagram of the tendon
The loads on the beam do not have to be uniformly distributed in order to apply
the load balancing concept. This is illustrated in Figures 9.20 and 9.21. If we use
similar notations for the loading conditions shown in Figures 9.20 and 9.21 and
change the tendons path from a parabola to those shown, the balancing forces for
F
the midspan point load and one-third span point load conditions are P bal
4
Fbal
and P
3 , respectively. We can easily extend the load balancing concept to
continuous spans as well as to at plates or at slabs. The detailed discussion of the
intricacies involved, however, are beyond the scope of this text.
Acting load
F
P Fbal P
Balancing load
Acting loads
F F
P P
Fbal Fbal
Balancing loads
Figure 9.21 Balancing two equal loads at the third points of the span
12 16
wsw 150 200 lb=ft
144
To balance all the dead loads and one-half the live loads, wbal is
9.7 The Concept of Load Balancing 585
w 400 lb/ft
h 16 in.
6 in.
30'-0"
b 12 in.
1
wbal 200 400 400 lb=ft
2
Only the self-weight of the beam acts on the beam at the time of the transfer of
the prestressing force. Hence, the upward loads from the prestressing tendon and
the self-weight of the beam yield a net result of 200 lb/ft upward. So we calculate
the initial stresses due to prestress at midspan as follows:
0:2 302
M 22:5 ft-kip
8
bh2 12 162
Sm 512 in3
6 6
M 22:5 12,000
f 527 psi
Sm 512
tension at the top and compression at the bottom. (The net difference of wbal and the
self-weight acts upward.)
From the 90 kip axial compression load at the time of transfer of the prestress
(transferred at the anchorages at the ends):
586 9 Overview of Prestressed Concrete
P 90 1,000
f 469 psi compression
A 192
Combining the axial compression (469 psi) and the exural stresses due to the
net 200 lb/ft upward load, the initial stresses at midspan are:
Because these resulting stresses at the time of the load transfer are less than the
ACI Code (Sections 24.5.3.2 and 24.5.4.1) permits:
p p
3 fci0 3 3,750 184 psi tension
and
0:4 302
MS:I: 45 ft-kip
8
M 45 12,000
f 1,055 psi
Sm 512
compression at the top and tension at the bottom. Combining these stresses due to
total live load with the initial stresses gives us:
Now check the stress after long-term losses have taken place.
Assume long-term prestressing losses of 15 %.
The nal remaining prestressing force is only:
8 P 8 76,500 0:5
wbal 340 lb=ft
2 302
rather than the full 400 lb/ft that was rst used.
Problems 587
0:26 302
Mnet 29:25 ft-kip
8
This moment causes compression at the top and tension at the bottom. These
stresses, combined with the compression stresses from the remaining
posttensioning force, will give the nal stresses:
These stresses are lower than those permitted by the ACI Code (Sections 24.5.2.1
and 24.5.4.1) for service load stage.
0:45 fc0 0:45 5,000 2,250 psi compression > 1,084 psi
p p
7:5 fc0 7:5 5,000 530 psi tension > 287 psi
Problems
9.1. A plain concrete beam has a width of 14 in. and a total depth of 24 in., and is
simply supported with a span of 24 ft. What is the maximum tensile stress
acting on the beam due to its weight?
9.2. The beam of Problem 9.1 is prestressed with a straight tendon at the centroid of
the section to produce a prestressing force of 200 kip.
1. What will be the maximum stresses on the beam at midspan (a) at the top,
and (b) at the bottom?
2. How much uniformly distributed load may be placed on the beam if no
tension is permitted in the beam?
9.3. Assume that the straight prestressing tendon of Problem 9.2 is placed 4 in.
from the bottom of the beam.
1. What are the maximum stresses on the beam at midspan (a) at the top, and
(b) at the bottom, when only the beams self-weight acts?
2. How much uniformly distributed load may be placed on the beam if 424 psi
maximum tension is permitted?
588 9 Overview of Prestressed Concrete
9.4. The rectangular prestressed beam (16 in. 32 in.) shown below is subject to a
total dead and live load of 2.2 kip/ft (including the beam weight). The
parabolic tendon will have a sag of 10 in., as shown, to provide an upward
balancing load of 1.6 kip/ft.
1. Calculate the required nal prestressing force (after losses).
2. Calculate the nal stresses in the beam at midspan and at the ends (a) at the
top of the section, and (b) at the bottom of the section.
w 2.2 kip/ft
10 in.
60'-0"
Self-Experiments
In this self-experiment, you will study the behavior of prestressed and
posttensioned beams. Record all the details of the tests and include photos showing
different stages of the experiments in your report.
Experiment 1
In this experiment we study the behavior of prestressed beams using a styrofoam
beam. Place the beam between two supports. Apply a predetermined load (a few
pounds) on the beam and record the magnitude of the beam deection.
Make a hole at the bottom of the beam and pass a few plastic strings through
it. Anchor the strings at one end. Pull the strings from the other side and anchor
them as shown in Figure SE 9.1. Place the beam on the same supports and apply the
same load. Determine how much the beam deects. Compare the results with those
of the previous test.
P
A-A
A
Experiment 2
In this experiment, we study the behavior of prestressed beams using concrete
models. Cast two concrete beams of the same size. Reinforce one beam with regular
wires, and thread the ends of the wire for the other. Before placing concrete for the
prestressed beam, pull the wires from the one side as shown in Figure SE 9.2.
Compare the behavior of the two beams by placing them on two supports and
gradually loading them. Which one deects more? Why? Discuss your
observations.
a b
Figure SE 9.2 (a) Reinforced concrete beam, (b) prestressed concrete beam
Experiment 3
Here we use a styrofoam beam similar to that of Experiment 1 to study the behavior
of posttensioned beams. Make a hole of the same size as that of Experiment 1 on the
side of the beam using a hot wire. The hole should have a curved shape, as shown in
Figure SE 9.3. Pass a few plastic strings through this hole, anchor them to the beam
from one end, and pull and anchor to the other end. Now place the beam on the two
supports and apply the same load as in Experiment 1. Record the beam deection at
the midspan. Compare the results with those of Experiment 1. Which case resulted
in less deection?
Experiment 4
This experiment involves the use of a posttensioned concrete beam. As in Exper-
iment 2, cast two beams. One will use regular wire, and the other will be
posttensioned. For the posttensioned beam, place a plastic tube inside the beam
590 9 Overview of Prestressed Concrete
and cast the concrete. After the concrete is set (72 hours), insert steel wires,
anchor them to the beam from one end, and pull and anchor to the other end
(Figure SE 9.4). Place the two beams on the two supports and gradually load
them. Compare their behavior. Which one deects more? Why? Any other
observations?
a b
Figure SE 9.4 (a) Reinforced concrete beam, (b) posttensioned concrete beam
Chapter 10
Metric System in Reinforced Concrete Design
and Construction
10.1 Introduction
Efforts to change the U.S. measurement units to the metric system have been under
way for quite a while. This chapter briey discusses this matter as it relates to
reinforced concrete structure design and construction. We present a few examples
using this system of units so that you will better understand how to make the
conversions.
Historically, the United States has used the British system of measurements. Most
other countries, however, use variations of the metric system. To conform with the
rest of the world, and to increase the international competitiveness, productivity,
and quality of U.S. industry, the U.S. Congress enacted the Metric Conversion Act
of 1975. A version of the metric system called Le Syste`me International dUnites
(International System of Units), or the SI system, was adopted. Furthermore, in
1988, the U.S. Congress passed the Omnibus Trade and Competitiveness Act, which
resulted in the formation of the Construction Metrication Council. This council is
part of the National Institute of Building Sciences (NIBS) located in Washington,
DC. The Council has published documents on Construction Metrication, which
provide the latest efforts on system conversions. You can obtain a copy of these
documents from the NIBS Web site (www.nibs.org).
The American Concrete Institute has published an equivalent metric version of
the ACI Code since 1983. The current metric version of the ACI code is ACI
318M-14 (M stands for metric). Adopting the metric system has two major
ramications: (1) using metric units for structural calculations, and (2) changing
the physical sizes of products based on the metric system of units. The rst task can
be accomplished with relative simplicity, as this chapter will show. But the man-
ufacture of construction material for concrete structures, in particular reinforcing
bars, has been one of the major obstacles in the adoption of the SI system of units.
To prevent the costly maintenance of two different inventories of steel rein-
forcement (in British and SI units), the producers of reinforcing bars adopted a soft
metric conversion in 1997. This conversion allows mills to produce reinforcing bars
in the customary British unit sizes, but to designate them with their equivalent
metric values instead of multiples of 100 mm2 as required in the hard metric
conversion. As a result, nearly all reinforcing bars currently produced are marked
with the soft metric equivalent sizes. Table 10.1 shows the equivalent soft metric
bar size designations for the customary British unit sizes.
A familiarity with the SI units is required to convert British units to their equivalent
SI units. Table 10.2 shows the main SI units along with the most common prexes
used in the design of building structures.
Two important quantities that we need to understand well are mass and force.
The SI unit of mass is the kilogram, kg, which is used as the unit of force in other
versions of the metric system. The SI unit used for force is the newton, which is
equal to 1 kg-m/s2. Force is equal to mass (m) multiplied by the gravitational
acceleration (g). Thus,
Fmg
F 1:0 kg9:80665 m=s2
F 9:80665 kg-m=s2 9:81 newtons N
The British units of pound-mass (lbm) and pound-force (lbf) are also dened.
These are related to each other as follows:
F mg
lbf lbm32:174 ft=s2
lbf 32:174 lbm-ft=s2 32:2 lbm-ft=s2
1 ft 0:3048 m
Therefore, substituting:
Note that weight is dened in units of mass. To use weight as load we need to
consider the gravitational acceleration of 9.81 m/s2. The following is an important
conversion:
lbf 4:448 N
1 157:1 N=m3
ft3 0:30483 m3
lbm 0:45359 kg
1 16:02 kg=m3
ft3 0:3048 m 3
Table 10.3 shows the complete set of conversion factors between the SI and
the British systems of units. The following examples solve problems posed by
examples in previous chapters using the equivalent SI units. Since we must use the
soft metric reinforcing bar sizes, Table 10.4 shows their designations along with
their properties.
Table 10.3 Conversion factors between the SI and the British system of units
Unit Multiply By To get:
Length inch (in.) 25.4 millimeter (mm)
foot (ft) 0.3048 meter (m)
millimeter (mm) 0.03937 inch (in.)
meter (m) 3.281 foot (ft)
Area square inch (in.2) 645.2 square millimeter (mm2)
square foot (ft2) 0.0929 square meter (m2)
square millimeter (mm2) 0.00155 square inch (in.2)
square meter (m2) 10.764 square foot (ft2)
Volume cubic inch (in.3) 16,387 cubic millimeter (mm3)
cubic foot (ft3) 0.028317 cubic meter (m3)
cubic millimeter (mm3) 0.000061024 cubic inch (in.3)
cubic meter (m3) 35.315 cubic foot (ft3)
Mass pound-mass (lbm) 0.454 kilogram (kg)
kilogram (kg) 2.205 pound
Density pound per cubic foot 16.02 kilogram per cubic meter
(lb/ft3) (kg/m3)
kilogram per cubic meter 0.06243 pound per cubic foot
(kg/m3) (lb/ft3)
Force pound-force (lbf) 4.448 newton (N)
kip 4,448 newton (N)
pound per foot (lb/ft) 14.594 newton per meter (N/m)
kip per foot (kip/ft) 14.594 kilonewton per meter
(kN/m)
newton (N) 0.2248 pound-force (lbf)
newton (N) 0.0002248 kip
newton per meter (N/m) 0.06852 pound per foot (lb/ft)
kilonewton per meter 0.06852 kip per foot (kip/ft)
(kN/m)
Moment of inch4 (in.4) 416,231 millimeter4 (mm4)
inertia millimeter4 (mm4) 0.000002403 inch4 (in.4)
(continued)
10.3 Conversion to SI Units 595
Example 10.1 (SI Version of Example 1.2) A 75 mm 150 mm, 2.70 m-long plain
concrete beam was simply supported at its ends and tested to determine the modulus
of rupture of the concrete. Two concentrated loads, P, were placed at the third
points. The beam failed at P 670 N. The specied compressive strength of the
concrete is fc0 28 MPa: The concrete weight (mass) is 2,400 kg/m3 (normal
weight). Determine the modulus of rupture of the concrete using (a) the results of
the test, and (b) the ACI Code approximate equation.
596 10 Metric System in Reinforced Concrete Design and Construction
w = 265 N/m
P = 670 N P = 670 N
Solution
(a) Test results
Determine the loads acting on the beam shown in Figure 10.1:
75150
w 2,400 27:0 kg=m 9:81 m=s2 265 N=m
1,0001,000
w2 P
Mtotal
8 3
2652:702 6702:70
Mtotal
8 3
Mtotal 242 603 845 N-m 845,000 N-mm
The maximum tensile stress at the bottom of the beam ( fr) is:
Mc M
fr
I Sm
bh2 751502
Sm 281,250 mm3
6 6
M 845 1,000
fr 3:00 N=mm2
Sm 281,250
3:00 106 N=m2 3:00 MPa
Example 10.2 (SI Version of Example 2.8) Calculate MR for the reinforced con-
crete section shown in Figure 10.2. Use fy 420 MPa, and fc0 28 MPa:
As 6 #32 4,914 mm2.
b 305 mm
d 795 mm dt 825 mm
6 #32
Solution
Step 1.
As 4,914
0:0203
bd 305 795
Step 2.
As fy 4,914420
a 284 mm
0:85fc0 b 0:85 28 305
Step 3.
a 284
c 334 mm
1 0:85
Step 4.
c 334
0:405 > 0:375
dt 825
598 10 Metric System in Reinforced Concrete Design and Construction
B2
A2 c
dt
0:25
0:233
0:405
0:850
Step 5.
mm 2 N/mm2(MPa) mm
a 284
MR = As fy d = (0.850)(4,914)(420) 795
2 2
MR = 1,145,556,594 N-mm/10 = 1,146 kN-m
6
Example 10.3 (SI Version of Example 4.2) Determine the spacing of #10 stirrups
at the critical section for a reinforced concrete beam with bw 380 mm,
h 610 mm, and Vu 270 kN: Use fc0 21 MPa, and fyt 420 MPa. Concrete is
normal weight.
Solution
dest h 65 610 65 545 mm
The shear force to be resisted by the stirrups at the critical section, Vs ; is:
Vu
Vs Vc
270
Vs 158:2
0:75
Vs 201:8 kN
Av fyt d
s
Vs
2 71420545
s 161 mm
201:8 1,000
Example 10.4 (SI Version of Example 5.4) Design a short square tied column to
carry an axial dead load of 1,300 kN, a oor live load of 800 kN and a roof live load
of 320 kN. Assume that the applied moments on the column are negligible. Use
fc0 28 MPa, fy 420 MPa, and a concrete clear cover of 40 mm.
Solution
Step 1. The factored load, Pu, is:
3,000 1,000
Ag
0:80:650:85281 0:03 4200:03
Ag 161,667 mm2
Therefore, the column is 400 400 mm, and the column gross area, Ag, is:
Pu 0.8(0.85 fc Ag )
Ast = Conversion from MPa
0.8( fy 0.85 fc ) to kPa
Step 5. Using Table 10.4, select 8 #29 bars (As 8 645 5,160 mm2).
Step 6. Using #10 for the ties, the maximum spacing, smax, (ACI Code, Sections
10.7.6.1.2 and 25.7.2.1) is:
400 mm
400 mm
#10 @ 400 mm
8 #29
Example 10.5 (SI Version of Example 7.1) A 300 mm load-bearing CMU wall
supports an outdoor canopy. The dead load is 150 kN/m (including the wall weight)
and the roof live load is 75 kN/m. Design the plain concrete footing shown in
Figure 10.4 to support this wall. The compressive strength of the concrete is
21 MPa, and the net bearing capacity of the soil is 150 kPa. The frost line is at
1.20 m from the outside grade. Concrete is normal weight.
10.3 Conversion to SI Units 601
300 mm
Block wall
1.20 m
Solution
Step 1. Determine the footing width (b).
wT
Approximate footing width (b)=
qa
kN/m
225
= = 1.50 m
150
kPa = kN/m2
\ b = 1.5 m
b t 1:5 0:30
h 0:60 m
2 2
h 0:60 m 600 mm
602 10 Metric System in Reinforced Concrete Design and Construction
wu 1:2wD 1:6wLr
wu 1:2150 1:675 300 kN=m
wu
qu
b
300
qu 200 kN=m2 kPa
1:5
The distance from the edge of the footing to the critical section for moment
(D) is:
bt t
D
2 4
1:5 1,000 300 300
D
2 4
D 675 mm
675 2
D2 1,000
M u qu 200
2 2
Mu 45:6 kN-m
d h 50 600 50 550 mm
bd 2
Sm
6
1,0005502
Sm 50:42 106 mm3
6
The nominal resisting moment, Mn, (ACI 318M-14 Equation 14.5.2.1a) is:
300 mm
600 mm
600 mm
1.50 m
Problems
In the following problems assume concrete is normal weight unless noted
otherwise.
10.1 (SI version of Problem 1.7) Draw the bending moment and shear force dia-
grams for a 300 mm 600 mm concrete beam made of lightweight concrete
with a unit weight (mass) of 1,800 kg/m3 and subjected to a uniformly
distributed load of 15 kN/m. Assume that the beam is simply-supported and
has a 3.0 m span.
10.2 (SI version of Problem 1.10) Determine the maximum span for a 200 mm
300 mm simply-supported plain concrete beam constructed of normal-
weight concrete with a unit weight (mass) of 2,400 kg/m3 and loaded by a
uniformly distributed load of 30 kN/m just before it fails. The specied
compressive strength of the concrete is 28 MPa. Use the ACI Code
recommended value for the modulus of rupture.
10.3 (SI version of Problem 2.7) The rectangular reinforced concrete beam shown
below is subjected to a dead load moment of 250 kN-m and a live load moment
of 125 kN-m. Determine whether this beam is adequate for the applied
moment using the Method I. Use fc0 28 MPa and fy 420 MPa. The
stirrups are #10 and the cover is 40 mm.
604 10 Metric System in Reinforced Concrete Design and Construction
300 mm
750 mm
4 #29
10.4 (SI version of Problem 2.11) Determine the moment capacity, MR, of the
reinforced concrete section shown below if it is subjected to a negative
moment. Use the Method I. The stirrups are #10 and the cover is 40 mm.
Use fc0 28 MPa and fy 420 MPa.
4 #32
710 mm
330 mm
10.5 (SI version Problem 4.2) A beam is subjected to a uniformly distributed load and
has a maximum shear of 270 kN at the face of its supports. The beam clear span
is 9.0 m, bw 300 mm, and d 600 mm. Use fc0 28 MPa, and fyt 420 MPa.
Determine the shear at the critical section. Determine the spacing of #10 stirrups
at the critical section.
10.6 (SI version of Problem 4.5) The shear force at the critical section, Vu ,
of a reinforced concrete beam is 265 kN. If the beam has bw 360 mm,
fc0 21 MPa, and fyt 420 MPa, what is the required effective depth, d,
such that the minimum spacing of #10 stirrups is 230 mm?
Problems 605
10.7 (SI version of Problem 5.4) The square reinforced concrete tied column
shown below is subjected to a dead load of 900 kN and a roof live load of
1,000 kN. Determine whether this column is adequate. The clear cover is
40 mm and the load eccentricity is negligible. Use fc0 28 MPa and
fy 420 MPa. Checking the ties is not required.
400 mm
#10 @ 350 mm
400 mm
8 #25
10.8 (SI version of Problem 5.8) Design a square tied reinforced concrete column
subjected to a dead load of 1,100 kN, a oor live load of 1,000 kN, and a
roof live load of 675 kN. The moments due to the loads are negligible. Use
fc0 28 MPa; fy 420 MPa, and 40 mm clear cover.
10.9 (SI version of Problem 7.1) Design a plain concrete wall footing to support a
300 mm thick concrete wall. The dead load, including the weight of wall, is
70 kN/m, and the roof live load is 90 kN/m. The bearing capacity of the soil
is 120 kPa, and fc0 21 MPa:
10.10 (SI version of Problem 7.3) Rework Problem 10.9 for a reinforced concrete
wall footing. Use fy 420 MPa.
Appendix A
Tables and Diagrams
Table A2.1 ACI approximate design moments and shears for beams and one-way slabs
Positive End spans
moment Discontinuous end unrestrained wu2n /11
Discontinuous end integral with support wu2n /14
Interior spans wu2n /16
Negative At exterior face of the rst interior support
moment Two spans wu2n /9
More than two spans wu2n /10
At other faces of interior supports wu2n /11
At the face of all supports for SLABS with spans not exceeding wu2n /12
10 ft; and BEAMS where ratio of sum of column stiffnesses to
beam stiffness exceeds 8 at each end of the span
At interior face of exterior support for members built integrally
with supports
Where support is a spandrel beam wu2n /24
Where support is a column wu2n /16
Shear In end members at the face of the rst interior support 1.15 wun/2
At face of all other supports wun/2
wu
Spandrel beam
Column
n n n
wu n wu n wu n
wu n
2 2 2
2
(Vu )
1.15wu n wu n wu n
2 2 2
wu n2
(End unrestrained)
11
wu n2 wu n2 wu n2
(End integral with
14 support) 16 16
(Mu )
2
wu n wu 2
n wu 2
n
(Spandrel beam) (Two spans) (Two spans)
24 9 9
wu 2
n wu n2 wu 2
n wu n2 wu n2 wu n2
wu n2 (Others) (Others)
(Column) 10 11 11 11 11 11
16
wu n2 wu n2 wu n2 wu n2 wu n2 wu 2
n
12 12 12 12 12 12
Table A2.3 max and tc for common grades of steel and compressive
strength of concrete (single layer of steel, i.e., d dt)
fy (psi) fc0 3,000 psi fc0 4,000 psi fc0 5,000 psi
max (t 0.004)
40,000 0.0232 0.0310 0.0364 0.83
60,000 0.0155 0.0207 0.0243 0.81
75,000 0.0124 0.0165 0.0194 0.80
tc (t 0.005)
40,000 0.0203 0.0270 0.0318 0.90
60,000 0.0135 0.0180 0.0212 0.90
75,000 0.0108 0.0144 0.0169 0.90
dt
Note: For multiple layers of reinforcements, multiply the table values by
d
Table A2.5a Resistance coefcient R (in psi) versus reinforcement ratio (); fc0 3,000 psi,
fy 40,000 psi (for beams min 0.005)
R R R R R R
0.001 36 0.0051 176 0.0092 307 0.0133 429 0.0174 541 0.0215 643 0.87
0.0011 39 0.0052 180 0.0093 310 0.0134 432 0.0175 544 0.0216 646 0.87
0.0012 43 0.0053 183 0.0094 313 0.0135 435 0.0176 546 0.0217 648 0.86
0.0013 46 0.0054 186 0.0095 317 0.0136 437 0.0177 549 0.0218 651 0.86
0.0014 50 0.0055 189 0.0096 320 0.0137 440 0.0178 551 0.0219 653 0.86
0.0015 53 0.0056 193 0.0097 323 0.0138 443 0.0179 554 0.022 655 0.86
0.0016 57 0.0057 196 0.0098 326 0.0139 446 0.018 557 0.0221 658 0.86
0.0017 60 0.0058 199 0.0099 329 0.014 449 0.0181 559 0.0222 660 0.85
0.0018 64 0.0059 203 0.01 332 0.0141 451 0.0182 562 0.0223 662 0.85
0.0019 67 0.006 206 0.0101 335 0.0142 454 0.0183 564 0.0224 665 0.85
0.002 71 0.0061 209 0.0102 338 0.0143 457 0.0184 567 0.0225 667 0.85
0.0021 74 0.0062 212 0.0103 341 0.0144 460 0.0185 569 0.0226 669 0.84
0.0022 78 0.0063 216 0.0104 344 0.0145 463 0.0186 572 0.0227 672 0.84
0.0023 81 0.0064 219 0.0105 347 0.0146 465 0.0187 574 0.0228 674 0.84
0.0024 85 0.0065 222 0.0106 350 0.0147 468 0.0188 577 0.0229 676 0.84
0.0025 88 0.0066 225 0.0107 353 0.0148 471 0.0189 580 0.023 679 0.84
0.0026 92 0.0067 229 0.0108 356 0.0149 474 0.019 582 0.0231 681 0.83
0.0027 95 0.0068 232 0.0109 359 0.015 476 0.0191 585 0.0232 683 0.83
0.0028 99 0.0069 235 0.011 362 0.0151 479 0.0192 587
0.0029 102 0.007 238 0.0111 365 0.0152 482 0.0193 590
0.003 105 0.0071 241 0.0112 368 0.0153 485 0.0194 592
0.0031 109 0.0072 245 0.0113 371 0.0154 487 0.0195 595
0.0032 112 0.0073 248 0.0114 374 0.0155 490 0.0196 597
0.0033 116 0.0074 251 0.0115 377 0.0156 493 0.0197 600
0.0034 119 0.0075 254 0.0116 380 0.0157 496 0.0198 602
0.0035 123 0.0076 257 0.0117 383 0.0158 498 0.0199 605
0.0036 126 0.0077 260 0.0118 385 0.0159 501 0.02 607
0.0037 129 0.0078 264 0.0119 388 0.016 504 0.0201 610
0.0038 133 0.0079 267 0.012 391 0.0161 506 0.0202 612
0.0039 136 0.008 270 0.0121 394 0.0162 509 0.0203 614 tc
0.004 139 0.0081 273 0.0122 397 0.0163 512 0.0204 617 0.90
0.0041 143 0.0082 276 0.0123 400 0.0164 514 0.0205 619 0.90
0.0042 146 0.0083 279 0.0124 403 0.0165 517 0.0206 622 0.89
0.0043 150 0.0084 282 0.0125 406 0.0166 520 0.0207 624 0.89
0.0044 153 0.0085 286 0.0126 409 0.0167 522 0.0208 627 0.89
0.0045 156 0.0086 289 0.0127 412 0.0168 525 0.0209 629 0.88
0.0046 160 0.0087 292 0.0128 415 0.0169 528 0.021 631 0.88
0.0047 163 0.0088 295 0.0129 417 0.017 530 0.0211 634 0.88
0.0048 166 0.0089 298 0.013 420 0.0171 533 0.0212 636 0.88
0.0049 170 0.009 301 0.0131 423 0.0172 536 0.0213 639 0.87
0.005 173 0.0091 304 0.0132 426 0.0173 538 0.0214 641 0.87
Appendix A: Tables and Diagrams 611
Table A2.5b Resistance coefcient R (in psi) versus reinforcement ratio (); fc0 4,000 psi,
fy 40,000 psi (for beams min 0.005)
R R R R R
0.001 36 0.0051 178 0.0092 313 0.0133 441 0.0174 562
0.0011 39 0.0052 181 0.0093 316 0.0134 444 0.0175 565
0.0012 43 0.0053 185 0.0094 320 0.0135 447 0.0176 568
0.0013 46 0.0054 188 0.0095 323 0.0136 450 0.0177 571
0.0014 50 0.0055 192 0.0096 326 0.0137 453 0.0178 574
0.0015 54 0.0056 195 0.0097 329 0.0138 456 0.0179 577
0.0016 57 0.0057 198 0.0098 332 0.0139 459 0.018 579
0.0017 61 0.0058 202 0.0099 336 0.014 462 0.0181 582
0.0018 64 0.0059 205 0.01 339 0.0141 465 0.0182 585
0.0019 68 0.006 208 0.0101 342 0.0142 468 0.0183 588
0.002 71 0.0061 212 0.0102 345 0.0143 471 0.0184 591
0.0021 75 0.0062 215 0.0103 348 0.0144 474 0.0185 594
0.0022 78 0.0063 218 0.0104 351 0.0145 477 0.0186 596
0.0023 82 0.0064 222 0.0105 355 0.0146 480 0.0187 599
0.0024 85 0.0065 225 0.0106 358 0.0147 483 0.0188 602
0.0025 89 0.0066 228 0.0107 361 0.0148 486 0.0189 605
0.0026 92 0.0067 232 0.0108 364 0.0149 489 0.019 608
0.0027 96 0.0068 235 0.0109 367 0.015 492 0.0191 610
0.0028 99 0.0069 238 0.011 370 0.0151 495 0.0192 613
0.0029 103 0.007 242 0.0111 374 0.0152 498 0.0193 616
0.003 106 0.0071 245 0.0112 377 0.0153 501 0.0194 619
0.0031 110 0.0072 248 0.0113 380 0.0154 504 0.0195 621
0.0032 113 0.0073 252 0.0114 383 0.0155 507 0.0196 624
0.0033 116 0.0074 255 0.0115 386 0.0156 510 0.0197 627
0.0034 120 0.0075 258 0.0116 389 0.0157 513 0.0198 630
0.0035 123 0.0076 261 0.0117 392 0.0158 516 0.0199 633
0.0036 127 0.0077 265 0.0118 395 0.0159 519 0.02 635
0.0037 130 0.0078 268 0.0119 398 0.016 522 0.0201 638
0.0038 134 0.0079 271 0.012 402 0.0161 525 0.0202 641
0.0039 137 0.008 274 0.0121 405 0.0162 528 0.0203 644
0.004 141 0.0081 278 0.0122 408 0.0163 531 0.0204 646
0.0041 144 0.0082 281 0.0123 411 0.0164 533 0.0205 649
0.0042 147 0.0083 284 0.0124 414 0.0165 536 0.0206 652
0.0043 151 0.0084 287 0.0125 417 0.0166 539 0.0207 654
0.0044 154 0.0085 291 0.0126 420 0.0167 542 0.0208 657
0.0045 158 0.0086 294 0.0127 423 0.0168 545 0.0209 660
0.0046 161 0.0087 297 0.0128 426 0.0169 548 0.021 663
0.0047 165 0.0088 300 0.0129 429 0.017 551 0.0211 665
0.0048 168 0.0089 304 0.013 432 0.0171 554 0.0212 668
0.0049 171 0.009 307 0.0131 435 0.0172 557 0.0213 671
0.005 175 0.0091 310 0.0132 438 0.0173 559 0.0214 673
(continued)
612 Appendix A: Tables and Diagrams
Table A2.5c Resistance coefcient R (in psi) versus reinforcement ratio (); fc0 5,000 psi,
fy 40,000 psi (for beams min 0.0053)
R R R R R
0.001 36 0.0051 179 0.0092 317 0.0133 449 0.0174 575
0.0011 39 0.0052 183 0.0093 320 0.0134 452 0.0175 578
0.0012 43 0.0053 186 0.0094 323 0.0135 455 0.0176 581
0.0013 47 0.0054 189 0.0095 327 0.0136 458 0.0177 584
0.0014 50 0.0055 193 0.0096 330 0.0137 461 0.0178 587
0.0015 54 0.0056 196 0.0097 333 0.0138 465 0.0179 590
0.0016 57 0.0057 200 0.0098 337 0.0139 468 0.018 593
0.0017 61 0.0058 203 0.0099 340 0.014 471 0.0181 596
0.0018 64 0.0059 207 0.01 343 0.0141 474 0.0182 599
0.0019 68 0.006 210 0.0101 346 0.0142 477 0.0183 602
0.002 71 0.0061 213 0.0102 350 0.0143 480 0.0184 605
0.0021 75 0.0062 217 0.0103 353 0.0144 483 0.0185 608
0.0022 78 0.0063 220 0.0104 356 0.0145 486 0.0186 611
0.0023 82 0.0064 223 0.0105 359 0.0146 489 0.0187 614
0.0024 85 0.0065 227 0.0106 363 0.0147 493 0.0188 617
0.0025 89 0.0066 230 0.0107 366 0.0148 496 0.0189 620
0.0026 92 0.0067 234 0.0108 369 0.0149 499 0.019 623
0.0027 96 0.0068 237 0.0109 372 0.015 502 0.0191 626
0.0028 99 0.0069 240 0.011 376 0.0151 505 0.0192 629
0.0029 103 0.007 244 0.0111 379 0.0152 508 0.0193 632
0.003 106 0.0071 247 0.0112 382 0.0153 511 0.0194 635
0.0031 110 0.0072 250 0.0113 385 0.0154 514 0.0195 638
0.0032 113 0.0073 254 0.0114 388 0.0155 517 0.0196 641
0.0033 117 0.0074 257 0.0115 392 0.0156 520 0.0197 643
0.0034 120 0.0075 260 0.0116 395 0.0157 523 0.0198 646
0.0035 124 0.0076 264 0.0117 398 0.0158 527 0.0199 649
0.0036 127 0.0077 267 0.0118 401 0.0159 530 0.02 652
0.0037 131 0.0078 270 0.0119 404 0.016 533 0.0201 655
0.0038 134 0.0079 274 0.012 408 0.0161 536 0.0202 658
0.0039 138 0.008 277 0.0121 411 0.0162 539 0.0203 661
0.004 141 0.0081 280 0.0122 414 0.0163 542 0.0204 664
0.0041 145 0.0082 284 0.0123 417 0.0164 545 0.0205 667
0.0042 148 0.0083 287 0.0124 420 0.0165 548 0.0206 670
0.0043 152 0.0084 290 0.0125 424 0.0166 551 0.0207 673
0.0044 155 0.0085 294 0.0126 427 0.0167 554 0.0208 676
0.0045 159 0.0086 297 0.0127 430 0.0168 557 0.0209 678
0.0046 162 0.0087 300 0.0128 433 0.0169 560 0.021 681
0.0047 165 0.0088 304 0.0129 436 0.017 563 0.0211 684
0.0048 169 0.0089 307 0.013 439 0.0171 566 0.0212 687
0.0049 172 0.009 310 0.0131 443 0.0172 569 0.0213 690
0.005 176 0.0091 314 0.0132 446 0.0173 572 0.0214 693
(continued)
614 Appendix A: Tables and Diagrams
Table A2.6a Resistance coefcient R (in psi) versus reinforcement ratio (); fc0 3,000 psi,
fy 60,000 psi (for beams min 0.0033)
R R R R
0.001 53 0.0051 259 0.0092 443 0.0133 606
0.0011 59 0.0052 264 0.0093 447 0.0134 610
0.0012 64 0.0053 268 0.0094 451 0.0135 613 tc
0.0013 69 0.0054 273 0.0095 456 0.0136 615 0.90
0.0014 74 0.0055 278 0.0096 460 0.0137 615 0.89
0.0015 80 0.0056 282 0.0097 464 0.0138 616 0.89
0.0016 85 0.0057 287 0.0098 468 0.0139 616 0.88
0.0017 90 0.0058 292 0.0099 472 0.014 616 0.88
0.0018 95 0.0059 296 0.01 476 0.0141 616 0.87
0.0019 100 0.006 301 0.0101 481 0.0142 616 0.87
0.002 105 0.0061 306 0.0102 485 0.0143 617 0.86
0.0021 111 0.0062 310 0.0103 489 0.0144 617 0.86
0.0022 116 0.0063 315 0.0104 493 0.0145 617 0.86
0.0023 121 0.0064 320 0.0105 497 0.0146 617 0.85
0.0024 126 0.0065 324 0.0106 501 0.0147 617 0.85
0.0025 131 0.0066 329 0.0107 505 0.0148 618 0.84
0.0026 136 0.0067 333 0.0108 509 0.0149 618 0.84
0.0027 141 0.0068 338 0.0109 513 0.015 618 0.83
0.0028 146 0.0069 342 0.011 517 0.0151 618 0.83
0.0029 151 0.007 347 0.0111 521 0.0152 618 0.83
0.003 156 0.0071 351 0.0112 525 0.0153 619 0.82
0.0031 161 0.0072 356 0.0113 529 0.0154 619 0.82
0.0032 166 0.0073 360 0.0114 533 0.0155 619 0.81
0.0033 171 0.0074 365 0.0115 537
0.0034 176 0.0075 369 0.0116 541
0.0035 181 0.0076 374 0.0117 545
0.0036 186 0.0077 378 0.0118 549
0.0037 191 0.0078 383 0.0119 553
0.0038 196 0.0079 387 0.012 557
0.0039 201 0.008 391 0.0121 560
0.004 206 0.0081 396 0.0122 564
0.0041 211 0.0082 400 0.0123 568
0.0042 216 0.0083 404 0.0124 572
0.0043 220 0.0084 409 0.0125 576
0.0044 225 0.0085 413 0.0126 580
0.0045 230 0.0086 417 0.0127 583
0.0046 235 0.0087 422 0.0128 587
0.0047 240 0.0088 426 0.0129 591
0.0048 245 0.0089 430 0.013 595
0.0049 249 0.009 435 0.0131 598
0.005 254 0.0091 439 0.0132 602
616 Appendix A: Tables and Diagrams
Table A2.6b Resistance coefcient R (in psi) versus reinforcement ratio (); fc0 4,000 psi,
fy 60,000 psi (for beams min 0.0033)
R R R R R
0.001 54 0.0051 263 0.0092 456 0.0133 634 0.0174 795
0.0011 59 0.0052 268 0.0093 461 0.0134 638 0.0175 799
0.0012 64 0.0053 273 0.0094 465 0.0135 642 0.0176 803
0.0013 69 0.0054 278 0.0095 470 0.0136 646 0.0177 807
0.0014 75 0.0055 283 0.0096 474 0.0137 650 0.0178 810
0.0015 80 0.0056 287 0.0097 479 0.0138 654 0.0179 814
0.0016 85 0.0057 292 0.0098 483 0.0139 659 0.018 818 tc
0.0017 90 0.0058 297 0.0099 488 0.014 663 0.0181 820 0.90
0.0018 96 0.0059 302 0.01 492 0.0141 667 0.0182 820 0.89
0.0019 101 0.006 307 0.0101 497 0.0142 671 0.0183 820 0.89
0.002 106 0.0061 312 0.0102 501 0.0143 675 0.0184 821 0.89
0.0021 111 0.0062 316 0.0103 506 0.0144 679 0.0185 821 0.88
0.0022 116 0.0063 321 0.0104 510 0.0145 683 0.0186 821 0.88
0.0023 122 0.0064 326 0.0105 514 0.0146 687 0.0187 821 0.88
0.0024 127 0.0065 331 0.0106 519 0.0147 691 0.0188 822 0.87
0.0025 132 0.0066 336 0.0107 523 0.0148 695 0.0189 822 0.87
0.0026 137 0.0067 340 0.0108 528 0.0149 699 0.019 822 0.87
0.0027 142 0.0068 345 0.0109 532 0.015 703 0.0191 822 0.86
0.0028 147 0.0069 350 0.011 536 0.0151 707 0.0192 822 0.86
0.0029 153 0.007 355 0.0111 541 0.0152 711 0.0193 823 0.86
0.003 158 0.0071 359 0.0112 545 0.0153 715 0.0194 823 0.85
0.0031 163 0.0072 364 0.0113 549 0.0154 719 0.0195 823 0.85
0.0032 168 0.0073 369 0.0114 554 0.0155 723 0.0196 823 0.85
0.0033 173 0.0074 374 0.0115 558 0.0156 726 0.0197 823 0.84
0.0034 178 0.0075 378 0.0116 562 0.0157 730 0.0198 824 0.84
0.0035 183 0.0076 383 0.0117 567 0.0158 734 0.0199 824 0.84
0.0036 188 0.0077 388 0.0118 571 0.0159 738 0.02 824 0.83
0.0037 193 0.0078 392 0.0119 575 0.016 742 0.0201 824 0.83
0.0038 198 0.0079 397 0.012 579 0.0161 746 0.0202 824 0.83
0.0039 203 0.008 402 0.0121 584 0.0162 750 0.0203 825 0.82
0.004 208 0.0081 406 0.0122 588 0.0163 754 0.0204 825 0.82
0.0041 213 0.0082 411 0.0123 592 0.0164 757 0.0205 825 0.82
0.0042 218 0.0083 415 0.0124 596 0.0165 761 0.0206 825 0.82
0.0043 223 0.0084 420 0.0125 601 0.0166 765 0.0207 825 0.81
0.0044 228 0.0085 425 0.0126 605 0.0167 769
0.0045 233 0.0086 429 0.0127 609 0.0168 773
0.0046 238 0.0087 434 0.0128 613 0.0169 777
0.0047 243 0.0088 438 0.0129 617 0.017 780
0.0048 248 0.0089 443 0.013 621 0.0171 784
0.0049 253 0.009 447 0.0131 626 0.0172 788
0.005 258 0.0091 452 0.0132 630 0.0173 792
Appendix A: Tables and Diagrams 617
Table A2.6c Resistance coefcient R (in psi) versus reinforcement ratio (); fc0 5,000 psi,
fy 60,000 psi (for beams min 0.0035)
R R R R R R
0.001 54 0.0051 265 0.0092 465 0.0133 651 0.0174 824 0.0215 976 0.89
0.0011 59 0.0052 270 0.0093 469 0.0134 655 0.0175 828 0.0216 976 0.89
0.0012 64 0.0053 275 0.0094 474 0.0135 660 0.0176 832 0.0217 977 0.89
0.0013 70 0.0054 280 0.0095 479 0.0136 664 0.0177 836 0.0218 977 0.88
0.0014 75 0.0055 285 0.0096 483 0.0137 668 0.0178 840 0.0219 977 0.88
0.0015 80 0.0056 290 0.0097 488 0.0138 673 0.0179 844 0.022 978 0.88
0.0016 85 0.0057 295 0.0098 493 0.0139 677 0.018 848 0.0221 978 0.87
0.0017 91 0.0058 300 0.0099 497 0.014 681 0.0181 853 0.0222 978 0.87
0.0018 96 0.0059 305 0.01 502 0.0141 686 0.0182 857 0.0223 978 0.87
0.0019 101 0.006 310 0.0101 507 0.0142 690 0.0183 861 0.0224 979 0.86
0.002 106 0.0061 315 0.0102 511 0.0143 694 0.0184 865 0.0225 979 0.86
0.0021 112 0.0062 320 0.0103 516 0.0144 699 0.0185 869 0.0226 979 0.86
0.0022 117 0.0063 325 0.0104 520 0.0145 703 0.0186 873 0.0227 980 0.86
0.0023 122 0.0064 330 0.0105 525 0.0146 707 0.0187 877 0.0228 980 0.85
0.0024 127 0.0065 335 0.0106 530 0.0147 711 0.0188 880 0.0229 980 0.85
0.0025 133 0.0066 340 0.0107 534 0.0148 716 0.0189 884 0.023 980 0.85
0.0026 138 0.0067 345 0.0108 539 0.0149 720 0.019 888 0.0231 981 0.85
0.0027 143 0.0068 350 0.0109 543 0.015 724 0.0191 892 0.0232 981 0.84
0.0028 148 0.0069 354 0.011 548 0.0151 728 0.0192 896 0.0233 981 0.84
0.0029 153 0.007 359 0.0111 552 0.0152 733 0.0193 900 0.0234 981 0.84
0.003 159 0.0071 364 0.0112 557 0.0153 737 0.0194 904 0.0235 982 0.83
0.0031 164 0.0072 369 0.0113 562 0.0154 741 0.0195 908 0.0236 982 0.83
0.0032 169 0.0073 374 0.0114 566 0.0155 745 0.0196 912 0.0237 982 0.83
0.0033 174 0.0074 379 0.0115 571 0.0156 750 0.0197 916 0.0238 982 0.83
0.0034 179 0.0075 384 0.0116 575 0.0157 754 0.0198 920 0.0239 983 0.82
0.0035 184 0.0076 388 0.0117 580 0.0158 758 0.0199 924 0.024 983 0.82
0.0036 189 0.0077 393 0.0118 584 0.0159 762 0.02 928 0.0241 983 0.82
0.0037 195 0.0078 398 0.0119 589 0.016 766 0.0201 931 0.0242 983 0.82
0.0038 200 0.0079 403 0.012 593 0.0161 771 0.0202 935 0.0243 984 0.81
0.0039 205 0.008 408 0.0121 598 0.0162 775 0.0203 939 0.0244 984 0.81
0.004 210 0.0081 412 0.0122 602 0.0163 779 0.0204 943
0.0041 215 0.0082 417 0.0123 607 0.0164 783 0.0205 947
0.0042 220 0.0083 422 0.0124 611 0.0165 787 0.0206 951
0.0043 225 0.0084 427 0.0125 615 0.0166 791 0.0207 954
0.0044 230 0.0085 431 0.0126 620 0.0167 795 0.0208 958
0.0045 235 0.0086 436 0.0127 624 0.0168 800 0.0209 962
0.0046 240 0.0087 441 0.0128 629 0.0169 804 0.021 966
0.0047 245 0.0088 446 0.0129 633 0.017 808 0.0211 970
0.0048 250 0.0089 450 0.013 638 0.0171 812 0.0212 973 tc
0.0049 255 0.009 455 0.0131 642 0.0172 816 0.0213 976 0.90
0.005 260 0.0091 460 0.0132 646 0.0173 820 0.0214 976 0.90
618 Appendix A: Tables and Diagrams
Table A2.7a Resistance coefcient R (in psi) versus reinforcement ratio (); fc0 3,000 psi,
fy 75,000 psi (for beams min 0.0027)
R R R
0.001 67 0.0051 318 0.0092 537
0.0011 73 0.0052 324 0.0093 542
0.0012 80 0.0053 330 0.0094 547
0.0013 86 0.0054 336 0.0095 552
0.0014 93 0.0055 341 0.0096 557
0.0015 99 0.0056 347 0.0097 561
0.0016 105 0.0057 352 0.0098 566
0.0017 112 0.0058 358 0.0099 571
0.0018 118 0.0059 364 0.01 576
0.0019 125 0.006 369 0.0101 580
0.002 131 0.0061 375 0.0102 585
0.0021 137 0.0062 380 0.0103 590
0.0022 144 0.0063 386 0.0104 595
0.0023 150 0.0064 391 0.0105 599
0.0024 156 0.0065 397 0.0106 604
0.0025 163 0.0066 402 0.0107 609
0.0026 169 0.0067 408 0.0108 613 tc
0.0027 175 0.0068 413 0.0109 615 0.90
0.0028 181 0.0069 418 0.011 614 0.89
0.0029 187 0.007 424 0.0111 613 0.88
0.003 194 0.0071 429 0.0112 613 0.87
0.0031 200 0.0072 435 0.0113 612 0.87
0.0032 206 0.0073 440 0.0114 611 0.86
0.0033 212 0.0074 445 0.0115 611 0.85
0.0034 218 0.0075 450 0.0116 610 0.85
0.0035 224 0.0076 456 0.0117 609 0.84
0.0036 230 0.0077 461 0.0118 609 0.83
0.0037 236 0.0078 466 0.0119 608 0.83
0.0038 242 0.0079 471 0.012 608 0.82
0.0039 248 0.008 476 0.0121 607 0.81
0.004 254 0.0081 482 0.0122 606 0.81
0.0041 260 0.0082 487 0.0123 606 0.80
0.0042 266 0.0083 492 0.0124 605 0.80
0.0043 272 0.0084 497
0.0044 278 0.0085 502
0.0045 284 0.0086 507
0.0046 289 0.0087 512
0.0047 295 0.0088 517
0.0048 301 0.0089 522
0.0049 307 0.009 527
0.005 313 0.0091 532
Appendix A: Tables and Diagrams 619
Table A2.7b Resistance coefcient R (in psi) versus reinforcement ratio (); fc0 4,000 psi,
fy 75,000 psi (for beams min 0.0027)
R R R R
0.001 67 0.0051 325 0.0092 558 0.0133 766
0.0011 73 0.0052 331 0.0093 563 0.0134 771
0.0012 80 0.0053 337 0.0094 569 0.0135 776
0.0013 86 0.0054 343 0.0095 574 0.0136 780
0.0014 93 0.0055 349 0.0096 579 0.0137 785
0.0015 100 0.0056 355 0.0097 585 0.0138 790
0.0016 106 0.0057 361 0.0098 590 0.0139 794
0.0017 113 0.0058 366 0.0099 595 0.014 799
0.0018 119 0.0059 372 0.01 601 0.0141 804
0.0019 126 0.006 378 0.0101 606 0.0142 808
0.002 132 0.0061 384 0.0102 611 0.0143 813
0.0021 138 0.0062 390 0.0103 616 0.0144 818 tc
0.0022 145 0.0063 396 0.0104 621 0.0145 820 0.90
0.0023 151 0.0064 402 0.0105 627 0.0146 819 0.89
0.0024 158 0.0065 407 0.0106 632 0.0147 818 0.89
0.0025 164 0.0066 413 0.0107 637 0.0148 818 0.88
0.0026 170 0.0067 419 0.0108 642 0.0149 817 0.87
0.0027 177 0.0068 425 0.0109 647 0.015 816 0.87
0.0028 183 0.0069 430 0.011 652 0.0151 816 0.86
0.0029 189 0.007 436 0.0111 658 0.0152 815 0.86
0.003 196 0.0071 442 0.0112 663 0.0153 815 0.85
0.0031 202 0.0072 447 0.0113 668 0.0154 814 0.85
0.0032 208 0.0073 453 0.0114 673 0.0155 813 0.84
0.0033 215 0.0074 459 0.0115 678 0.0156 813 0.84
0.0034 221 0.0075 464 0.0116 683 0.0157 812 0.83
0.0035 227 0.0076 470 0.0117 688 0.0158 811 0.83
0.0036 233 0.0077 476 0.0118 693 0.0159 811 0.82
0.0037 240 0.0078 481 0.0119 698 0.016 810 0.82
0.0038 246 0.0079 487 0.012 703 0.0161 809 0.82
0.0039 252 0.008 492 0.0121 708 0.0162 809 0.81
0.004 258 0.0081 498 0.0122 713 0.0163 808 0.81
0.0041 264 0.0082 503 0.0123 718 0.0164 808 0.80
0.0042 270 0.0083 509 0.0124 723 0.0165 807 0.80
0.0043 276 0.0084 514 0.0125 727
0.0044 283 0.0085 520 0.0126 732
0.0045 289 0.0086 525 0.0127 737
0.0046 295 0.0087 531 0.0128 742
0.0047 301 0.0088 536 0.0129 747
0.0048 307 0.0089 542 0.013 752
0.0049 313 0.009 547 0.0131 756
0.005 319 0.0091 553 0.0132 761
620 Appendix A: Tables and Diagrams
Table A2.7c Resistance coefcient R (in psi) versus reinforcement ratio (); fc0 5,000 psi,
fy 75,000 psi (for beams min 0.0028)
R R R R R
0.001 67 0.0051 329 0.0092 571 0.0133 792 0.0174 973 0.88
0.0011 74 0.0052 335 0.0093 576 0.0134 798 0.0175 973 0.88
0.0012 80 0.0053 341 0.0094 582 0.0135 803 0.0176 972 0.87
0.0013 87 0.0054 347 0.0095 587 0.0136 808 0.0177 971 0.87
0.0014 93 0.0055 353 0.0096 593 0.0137 813 0.0178 971 0.86
0.0015 100 0.0056 359 0.0097 599 0.0138 818 0.0179 970 0.86
0.0016 106 0.0057 365 0.0098 604 0.0139 823 0.018 970 0.85
0.0017 113 0.0058 371 0.0099 610 0.014 828 0.0181 969 0.85
0.0018 120 0.0059 378 0.01 615 0.0141 833 0.0182 969 0.85
0.0019 126 0.006 384 0.0101 621 0.0142 838 0.0183 968 0.84
0.002 133 0.0061 390 0.0102 627 0.0143 843 0.0184 967 0.84
0.0021 139 0.0062 396 0.0103 632 0.0144 848 0.0185 967 0.83
0.0022 146 0.0063 402 0.0104 638 0.0145 854 0.0186 966 0.83
0.0023 152 0.0064 408 0.0105 643 0.0146 859 0.0187 966 0.82
0.0024 159 0.0065 414 0.0106 649 0.0147 864 0.0188 965 0.82
0.0025 165 0.0066 420 0.0107 654 0.0148 869 0.0189 965 0.82
0.0026 171 0.0067 426 0.0108 660 0.0149 874 0.019 964 0.81
0.0027 178 0.0068 431 0.0109 665 0.015 878 0.0191 963 0.81
0.0028 184 0.0069 437 0.011 670 0.0151 883 0.0192 963 0.81
0.0029 191 0.007 443 0.0111 676 0.0152 888 0.0193 962 0.80
0.003 197 0.0071 449 0.0112 681 0.0153 893 0.0194 962 0.80
0.0031 204 0.0072 455 0.0113 687 0.0154 898
0.0032 210 0.0073 461 0.0114 692 0.0155 903
0.0033 216 0.0074 467 0.0115 697 0.0156 908
0.0034 223 0.0075 473 0.0116 703 0.0157 913
0.0035 229 0.0076 479 0.0117 708 0.0158 918
0.0036 235 0.0077 484 0.0118 714 0.0159 923
0.0037 242 0.0078 490 0.0119 719 0.016 928
0.0038 248 0.0079 496 0.012 724 0.0161 932
0.0039 254 0.008 502 0.0121 730 0.0162 937
0.004 260 0.0081 508 0.0122 735 0.0163 942
0.0041 267 0.0082 513 0.0123 740 0.0164 947
0.0042 273 0.0083 519 0.0124 745 0.0165 952
0.0043 279 0.0084 525 0.0125 751 0.0166 956
0.0044 285 0.0085 531 0.0126 756 0.0167 961
0.0045 292 0.0086 536 0.0127 761 0.0168 966
0.0046 298 0.0087 542 0.0128 766 0.0169 971 tc
0.0047 304 0.0088 548 0.0129 772 0.017 975 0.90
0.0048 310 0.0089 554 0.013 777 0.0171 975 0.90
0.0049 316 0.009 559 0.0131 782 0.0172 974 0.89
0.005 323 0.0091 565 0.0132 787 0.0173 974 0.89
Appendix A: Tables and Diagrams 621
Table A2.8 bmin and bmax for reinforced concrete beams (in.)
Table A3.2 Simplied expression of development length, d, for bars in tension based on ACI
code Section 25.4.2.2
#6 and smaller bars #7 and larger
Conditions and deformed wires bars
A. Clear spacing of bars or wires being developed or ! !
spliced not less than db, clear cover not less than f y t e f y t e
db, and stirrups or ties throughout d not less than p db p db
25 fc0 20 fc0
the ACI Code minimum; or clear spacing of bars
or wires being developed or lap spliced at least
2db and clear cover at least db
! !
B. Other cases 3f y t e 3f y t e
p db p db
50 fc0 40 fc0
Note: The development length used may not be less than 12 in.
Table A3.3 Development length for tension bars (d) with fy 60 ksi (e t 1.0) [in.]
ld (in.)
fc0 3 ksi fc0 4 ksi
Bar size Condition A Condition B Condition A Condition B
#3 17 25 15 22
#4 22 33 19 29
#5 28 41 24 36
#6 33 50 29 43
#7 48 72 42 63
#8 55 83 48 72
#9 62 93 54 81
#10 70 105 61 91
#11 78 116 67 101
Note: Conditions A and B are based on Table A3.2
Appendix A: Tables and Diagrams 625
Table A3.4 Factors for development of hooked bars in tension [ACI Section 25.4.3.2]
Symbol Name Condition Value
Epoxy-coated or zinc and epoxy dual-coated
1.2
reinforcement
e Coating factor
Uncoated or zinc-coated (galvanized)
1.0
reinforcement
#11 Bar and smaller hooks with side cover (normal
to plane of hook) 2.5 in., and for 90 hook with 0.7
c Cover factor cover on bar extension beyond hook 2 in.
Other 1.0
#11 Bar and smaller 90 hooks:
(1) enclosed along dh within ties or stirrups per-
pendicular to dh at spacing 3db, or
(2) Enclosed along the bar extension beyond hook
including the bend within ties or stirrups per-
0.8
r Conning factor pendicular to the hook extension at spacing
3db
(3) #11 and smaller bars 180 hooks
Enclosed along dh within ties or stirrups per-
pendicular to dh at spacing 3db
Other 1.0
Lightweight aggre- Lightweight concrete 0.75
gate concrete factor Normal weight concrete 1.0
Excess Reinforcement used by excess of that required by As, required
reinforcement analysis As, provided
Note: The development length used may not be less than the smaller of 8db or 6 in.
Table A3.5 Factors for development of bars in compression [ACI Section 25.4.9.3]
Symbol Name Condition Value
Reinforcement enclosed within one of the
following:
(1) A spiral
(2) A circular continuously wound tie with 0.75
r Conning factor db 1/4 in., and pitch 4 in.
(3) #4 Ties spaced 4 in. on center
(4) Hoops spaced 4 in. on center
Other 1.0
Lightweight aggregate Lightweight concrete 0.75
concrete factor Normal-weight concrete 1.0
Reinforcement used in excess of that required As, required
Excess reinforcement
by analysis As, provided
Note: The development length used may not be less than 8 in.
626 Appendix A: Tables and Diagrams
Table A4.1a Values of Vc in kips fc0 3,000 psi [d h2.5 in. (+/)] ( 1.0)
fc0 3,000 psi
bw (in.)
h (in.) 6 8 10 12 14 16 18 20 22 24 26 28 30
10 4.9 6.6 8.2 9.9 11.5 13.1 14.8 16.4 18.1 19.7 21.4 23.0 24.6
12 6.2 8.3 10.4 12.5 14.6 16.7 18.7 20.8 22.9 25.0 27.1 29.1 31.2
14 7.6 10.1 12.6 15.1 17.6 20.2 22.7 25.2 27.7 30.2 32.8 35.3 37.8
16 8.9 11.8 14.8 17.7 20.7 23.7 26.6 29.6 32.5 35.5 38.5 41.4 44.4
18 10.2 13.6 17.0 20.4 23.8 27.2 30.6 34.0 37.4 40.8 44.1 47.5 50.9
20 11.5 15.3 19.2 23.0 26.8 30.7 34.5 38.3 42.2 46.0 49.8 53.7 57.5
22 12.8 17.1 21.4 25.6 29.9 34.2 38.5 42.7 47.0 51.3 55.5 59.8 64.1
24 14.1 18.8 23.6 28.3 33.0 37.7 42.4 47.1 51.8 56.5 61.2 65.9 70.7
26 15.4 20.6 25.7 30.9 36.0 41.2 46.3 51.5 56.6 61.8 66.9 72.1 77.2
28 16.8 22.3 27.9 33.5 39.1 44.7 50.3 55.9 61.5 67.0 72.6 78.2 83.8
30 18.1 24.1 30.1 36.1 42.2 48.2 54.2 60.2 66.3 72.3 78.3 84.3 90.4
32 19.4 25.9 32.3 38.8 45.2 51.7 58.2 64.6 71.1 77.6 84.0 90.5 96.9
34 20.7 27.6 34.5 41.4 48.3 55.2 62.1 69.0 75.9 82.8 89.7 96.6 103.5
36 22.0 29.4 36.7 44.0 51.4 58.7 66.1 73.4 80.7 88.1 95.4 102.8 110.1
38 23.3 31.1 38.9 46.7 54.4 62.2 70.0 77.8 85.6 93.3 101.1 108.9 116.7
40 24.6 32.9 41.1 49.3 57.5 65.7 73.9 82.2 90.4 98.6 106.8 115.0 123.2
42 26.0 34.6 43.3 51.9 60.6 69.2 77.9 86.5 95.2 103.8 112.5 121.2 129.8
Table A4.1b Values of Vc in kips fc0 4,000 psi [d h2.5 in. (+/)] ( 1.0)
fc0 4,000 psi
bw (in.)
h (in.) 6 8 10 12 14 16 18 20 22 24 26 28 30
10 5.7 7.6 9.5 11.4 13.3 15.2 17.1 19.0 20.9 22.8 24.7 26.6 28.5
12 7.2 9.6 12.0 14.4 16.8 19.2 21.6 24.0 26.4 28.8 31.2 33.6 36.0
14 8.7 11.6 14.5 17.5 20.4 23.3 26.2 29.1 32.0 34.9 37.8 40.7 43.6
16 10.2 13.7 17.1 20.5 23.9 27.3 30.7 34.2 37.6 41.0 44.4 47.8 51.2
18 11.8 15.7 19.6 23.5 27.4 31.4 35.3 39.2 43.1 47.1 51.0 54.9 58.8
20 13.3 17.7 22.1 26.6 31.0 35.4 39.8 44.3 48.7 53.1 57.6 62.0 66.4
22 14.8 19.7 24.7 29.6 34.5 39.5 44.4 49.3 54.3 59.2 64.1 69.1 74.0
24 16.3 21.8 27.2 32.6 38.1 43.5 49.0 54.4 59.8 65.3 70.7 76.1 81.6
26 17.8 23.8 29.7 35.7 41.6 47.6 53.5 59.5 65.4 71.3 77.3 83.2 89.2
28 19.4 25.8 32.3 38.7 45.2 51.6 58.1 64.5 71.0 77.4 83.9 90.3 96.8
30 20.9 27.8 34.8 41.7 48.7 55.7 62.6 69.6 76.5 83.5 90.4 97.4 104.4
32 22.4 29.9 37.3 44.8 52.2 59.7 67.2 74.6 82.1 89.6 97.0 104.5 111.9
34 23.9 31.9 39.8 47.8 55.8 63.8 71.7 79.7 87.7 95.6 103.6 111.6 119.5
36 25.4 33.9 42.4 50.8 59.3 67.8 76.3 84.7 93.2 101.7 110.2 118.6 127.1
38 26.9 35.9 44.9 53.9 62.9 71.8 80.8 89.8 98.8 107.8 116.8 125.7 134.7
40 28.5 37.9 47.4 56.9 66.4 75.9 85.4 94.9 104.4 113.8 123.3 132.8 142.3
42 30.0 40.0 50.0 60.0 69.9 79.9 89.9 99.9 109.9 119.9 129.9 139.9 149.9
628 Appendix A: Tables and Diagrams
Table A4.1c Values of Vc in kips fc0 5,000 psi [d h2.5 in. (+/)] ( 1.0)
fc0 5,000 psi
h bw (in.)
(in.) 6 8 10 12 14 16 18 20 22 24 26 28 30
10 6.4 8.5 10.6 12.7 14.8 17.0 19.1 21.2 23.3 25.5 27.6 29.7 31.8
12 8.1 10.7 13.4 16.1 18.8 21.5 24.2 26.9 29.6 32.2 34.9 37.6 40.3
14 9.8 13.0 16.3 19.5 22.8 26.0 29.3 32.5 35.8 39.0 42.3 45.5 48.8
16 11.5 15.3 19.1 22.9 26.7 30.5 34.4 38.2 42.0 45.8 49.6 53.5 57.3
18 13.2 17.5 21.9 26.3 30.7 35.1 39.5 43.8 48.2 52.6 57.0 61.4 65.8
20 14.8 19.8 24.7 29.7 34.6 39.6 44.5 49.5 54.4 59.4 64.3 69.3 74.2
22 16.5 22.1 27.6 33.1 38.6 44.1 49.6 55.2 60.7 66.2 71.7 77.2 82.7
24 18.2 24.3 30.4 36.5 42.6 48.6 54.7 60.8 66.9 73.0 79.1 85.1 91.2
26 19.9 26.6 33.2 39.9 46.5 53.2 59.8 66.5 73.1 79.8 86.4 93.1 99.7
28 21.6 28.8 36.1 43.3 50.5 57.7 64.9 72.1 79.3 86.5 93.8 101.0 108.2
30 23.3 31.1 38.9 46.7 54.4 62.2 70.0 77.8 85.6 93.3 101.1 108.9 116.7
32 25.0 33.4 41.7 50.1 58.4 66.8 75.1 83.4 91.8 100.1 108.5 116.8 125.2
34 26.7 35.6 44.5 53.5 62.4 71.3 80.2 89.1 98.0 106.9 115.8 124.7 133.6
36 28.4 37.9 47.4 56.9 66.3 75.8 85.3 94.8 104.2 113.7 123.2 132.7 142.1
38 30.1 40.2 50.2 60.2 70.3 80.3 90.4 100.4 110.5 120.5 130.5 140.6 150.6
40 31.8 42.4 53.0 63.6 74.2 84.9 95.5 106.1 116.7 127.3 137.9 148.5 159.1
42 33.5 44.7 55.9 67.0 78.2 89.4 100.6 111.7 122.9 134.1 145.2 156.4 167.6
Table A4.2a Values of Vs in kips, with 2 legs of #3 stirrups (fyt 60,000 psi)
#3 Stirrups2 legs
Spacing s (in.)
h (in.) 2 3 4 5 6 8 10 12 14 16 18
10 49.5 33.0 24.8
12 62.7 41.8 31.4 25.1
14 75.9 50.6 38.0 30.4 25.3
16 89.1 59.4 44.6 35.6 29.7
18 102.3 68.2 51.2 40.9 34.1 25.6
20 115.5 77.0 57.8 46.2 38.5 28.9
22 128.7 85.8 64.4 51.5 42.9 32.2 25.7
24 141.9 94.6 71.0 56.8 47.3 35.5 28.4
26 155.1 103.4 77.6 62.0 51.7 38.8 31.0 25.9
28 168.3 112.2 84.2 67.3 56.1 42.1 33.7 28.1
30 181.5 121.0 90.8 72.6 60.5 45.4 36.3 30.3 25.9
32 194.7 129.8 97.4 77.9 64.9 48.7 38.9 32.5 27.8
34 207.9 138.6 104.0 83.2 69.3 52.0 41.6 34.7 29.7 26.0
36 221.1 147.4 110.6 88.4 73.7 55.3 44.2 36.9 31.6 27.6
38 234.3 156.2 117.2 93.7 78.1 58.6 46.9 39.1 33.5 29.3 26.0
40 247.5 165.0 123.8 99.0 82.5 61.9 49.5 41.3 35.4 30.9 27.5
Note: Multiply table values by 2 for #3 stirrups with 4 legs
Appendix A: Tables and Diagrams 629
Table A4.2b Values of Vs in kips, with 2 legs of #4 stirrups (fyt 60,000 psi)
#4 Stirrups2 legs
Spacing s (in.)
h (in.) 2 3 4 5 6 8 10 12 14 16 18
10 90.0 60.0 45.0
12 114.0 76.0 57.0 45.6
14 138.0 92.0 69.0 55.2 46.0
16 162.0 108.0 81.0 64.8 54.0
18 186.0 124.0 93.0 74.4 62.0 46.5
20 210.0 140.0 105.0 84.0 70.0 52.5
22 234.0 156.0 117.0 93.6 78.0 58.5 46.8
24 258.0 172.0 129.0 103.2 86.0 64.5 51.6
26 282.0 188.0 141.0 112.8 94.0 70.5 56.4 47.0
28 306.0 204.0 153.0 122.4 102.0 76.5 61.2 51.0
30 330.0 220.0 165.0 132.0 110.0 82.5 66.0 55.0 47.1
32 354.0 236.0 177.0 141.6 118.0 88.5 70.8 59.0 50.6
34 378.0 252.0 189.0 151.2 126.0 94.5 75.6 63.0 54.0 47.3
36 402.0 268.0 201.0 160.8 134.0 100.5 80.4 67.0 57.4 50.3
38 426.0 284.0 213.0 170.4 142.0 106.5 85.2 71.0 60.9 53.3 47.3
40 450.0 300.0 225.0 180.0 150.0 112.5 90.0 75.0 64.3 56.3 50.0
Note: Multiply table values by 2 for #4 stirrups with 4 legs
630 Appendix A: Tables and Diagrams
h h
a 2.0
INTERACTION DIAGRAM R4-60.6 h
fc 4 ksi h
1.8 g fy 60 ksi
0.0 0.6
8
1.6 Kmax
0.0
7
0.0 e P
6 n
1.4
0.0
5
0.0
1.2 4
fs /fy 0
0.0
fc Ag
3
Pn
1.0 0.0
2
Kn
0.25
0.0
1
0.8
0.50
0.6
0.75
0.4
0.0
0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35
Pne
Rn
fc Agh
b 0.4
fc Ag
Pn
g 0.01
0.2
Kn
0.0
2
0.0
3
0.0
4
0.
05
0.
06
0.
07
0.
08
0.0
0.65 0.70 0.75 0.80 0.85 0.90
Figure A5.1 (a) ACI column interaction diagram [SP-17(11)], Courtesy of American Concrete
Institute. (b) Kn versus relationship
632 Appendix A: Tables and Diagrams
a 2.0
INTERACTION DIAGRAM R4-60.7 h
fc 4 ksi h
1.8 fy 60 ksi
g
0.0 0.7
8
Kmax
1.6
0.0
7
e P
0.0 n
1.4 6
0.0
5
0.0 fs /fy 0
1.2 4
0.0
fc A g
Pn
3
1.0 0.0
2 0.25
Kn
0.0
1
0.8
0.50
0.6
0.75
0.4
1.0
t = 0.0035
0.2 t = 0.0050
0.0
0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40
Pne
Rn
fc Agh
b 0.4
fc Ag
Pn
g 0.01
0.2
0.02
Kn
0.03
0.0
4
0.0
5
0.
06
0.
07
0.
08
0.0
0.65 0.70 0.75 0.80 0.85 0.90
Figure A5.2 (a) ACI column interaction diagram [SP-17(11)], Courtesy of American Concrete
Institute. (b) Kn versus relationship
Appendix A: Tables and Diagrams 633
a 2.0
INTERACTION DIAGRAM R4-60.8 h
fc 4 ksi h
1.8 fy 60 ksi
g
0.0 0.8
8
1.6 Kmax
0.0
7
e P
0.0 n
1.4 6
0.0
5 fs /fy 0
0.0
1.2 4
0.0
3
fc Ag
Pn
0.0 0.25
1.0 2
0.0
Kn
1
0.8 0.50
0.6 0.75
0.4 1.0
t = 0.0035
0.2 t = 0.0050
0.0
0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45
Pne
Rn
fc Ag h
b 0.4
g 0.01
fc Ag
Pn
0.2 0.02
Kn
0.03
0.04
0.0
5
0.0
6
0.0
7
0.0
8
0.0
0.65 0.70 0.75 0.80 0.85 0.90
Figure A5.3 (a) ACI column interaction diagram [SP-17(11)], Courtesy of American Concrete
Institute. (b) Kn versus relationship
634 Appendix A: Tables and Diagrams
a 2.0
INTERACTION DIAGRAM R4-60.9 h
fc 4 ksi h
1.8 g
0.0 fy 60 ksi
8 0.9
Kmax
1.6 0.0
7
0.0 e P
6 n
1.4 0.0
5 fs /fy 0
0.0
4
1.2 0.0
3
fc Ag
Pn
0.0 0.25
2
1.0
0.0
Kn
1
0.50
0.8
0.75
0.6
1.0
0.4
t = 0.0035
t = 0.0050
0.2
0.0
0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50
Pne
Rn
fc Agh
b 0.6
0.4
fc Ag
Pn
Kn
g 0.01
0.02
0.2 0.03
0.04
0.05
0.06
0.07
0.08
0.0
0.65 0.70 0.75 0.80 0.85 0.90
Figure A5.4 (a) ACI column interaction diagram [SP-17(11)], Courtesy of American Concrete
Institute. (b) Kn versus relationship
Appendix A: Tables and Diagrams 635
a 2.0
INTERACTION DIAGRAM L4-60.6 h
fc 4 ksi h
g
1.8 0.0 fy 60 ksi
8 0.6
1.6 Kmax
0.0
7
0.0 e P
6 n
1.4 0.0
5
0.0
4
1.2 0.0
3 fs /fy 0
fc Ag
0.0
Pn
2
1.0
0.0
1
Kn
0.25
0.8
0.50
0.6
0.75
0.4
t = 0.0035 1.0
0.2
t = 0.0050
0.0
0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45
Pne
Rn
fc Ag h
b 0.4
fc A g
Pn
0.2 g 0.01
Kn
0.02
0.03
0.0
4
0.0
5
0.0
6
0.0
0.0 7
8
0.0
0.65 0.70 0.75 0.80 0.85 0.90
Figure A5.5 (a) ACI column interaction diagram [SP-17(11)], Courtesy of American Concrete
Institute. (b) Kn versus relationship
636 Appendix A: Tables and Diagrams
a 2.0
INTERACTION DIAGRAM L4-60.7 h
fc 4 ksi h
1.8 fy 60 ksi
g
0. 0.7
08
0.0
Pn
2
1.0
0.0
1 0.25
Kn
0.8
0.50
0.6
0.75
0.4
1.0
t = 0.0035
0.2 t = 0.0050
0.0
0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50 0.55
Pne
Rn
fc A g h
b 0.4
g 0.01
fc Ag
Pn
0.2 0.02
0.03
Kn
0.04
0.05
0.06
0.07
0.0
8
0.0
0.65 0.70 0.75 0.80 0.85 0.90
Figure A5.6 (a) ACI column interaction diagram [SP-17(11)], Courtesy of American Concrete
Institute. (b) Kn versus relationship
Appendix A: Tables and Diagrams 637
a 2.0
INTERACTION DIAGRAM L4-60.8 h
fc 4 ksi h
1.8 g fy 60 ksi
0.
08 0.8
0.
1.0 02
0.25
Kn
0.0
1
0.8
0.50
0.6
0.75
0.4
1.0
t = 0.0035
t = 0.0050
0.2
0.0
0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50 0.55 0.60
Pne
Rn
fc Ag h
b 0.4
g 0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
fc Ag
Pn
Kn
0.2
0.65 0.70 0.75 0.80 0.85 0.90
Figure A5.7 (a) ACI column interaction diagram [SP-17(11)], Courtesy of American Concrete
Institute. (b) Kn versus relationship
638 Appendix A: Tables and Diagrams
a 2.0
INTERACTION DIAGRAM L4-60.9 h
fc 4 ksi h
1.8 fy 60 ksi
g
0.0 0.9
8
1.6 0. Kmax
07
0.0 e P
6 n
1.4 0.0
5
0. fs /fy 0
04
1.2
0.0
3
f c Ag
Pn
0.
02
1.0 0.25
Kn
0.
01
0.8 0.50
0.6
0.75
1.0
0.4
t = 0.0035
t = 0.0050
0.2
0.0
0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40 0.45 0.50 0.55 0.60 0.65
Pne
Rn
fc Agh
b 0.4
g 0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
fc Ag
Pn
Kn
0.2
0.65 0.70 0.75 0.80 0.85 0.90
Figure A5.8 (a) ACI column interaction diagram [SP-17(11)], Courtesy of American Concrete
Institute. (b) Kn versus relationship
Appendix A: Tables and Diagrams 639
a 2.0
INTERACTION DIAGRAM C4-60.6 h
fc 4 ksi h
1.8 g fy 60 ksi
0.0
8 0.60
0.02
1.0
0.01 0.25
Kn
0.8
0.50
0.6
0.75
0.4
1.0
0.2 t = 0.0035
t = 0.0050
0.0
0.000 0.025 0.050 0.075 0.100 0.125 0.150 0.175 0.200 0.225 0.250
Pne
Rn
fc Ag h
b
0.4
g
0.0
1
fc Ag
Pn
0.2 0.0
2
Kn
0.0
3
0.
04
0.
05
0.
06
0.
07
0.0
8
0
0.75 0.76 0.77 0.78 0.79 0.8 0.81 0.82 0.83 0.84 0.85 0.86 0.87 0.88 0.89 0.9
Figure A5.9 (a) ACI column interaction diagram [SP-17(11)], Courtesy of American Concrete
Institute. (b) Kn versus relationship
640 Appendix A: Tables and Diagrams
a 2.0
INTERACTION DIAGRAM C4-60.7 h
fc 4 ksi h
1.8 g fy 60 ksi
0.0 0.7
8
0.25
1.0 0.0
2
Kn
0.0
1
0.8 0.50
0.6
0.75
0.4
1.0
0.2 t = 0.0035
t = 0.0050
0.0
0.00 0.05 0.10 0.15 0.20 0.25 0.30
P ne
Rn
fc Agh
b 0.4
g
0.0
1
f c Ag
Pn
0.0
0.2 2
0.0
Kn
3
0.0
4
0.0
5
0.
06
0.
0
0. 7
08
0
0.75 0.76 0.77 0.78 0.79 0.8 0.81 0.82 0.83 0.84 0.85 0.86 0.87 0.88 0.89 0.9
Figure A5.10 (a) ACI column interaction diagram [SP-17(11)], Courtesy of American Concrete
Institute. (b) Kn versus relationship
Appendix A: Tables and Diagrams 641
a 2.0
INTERACTION DIAGRAM C4-60.8 h
h
fc 4 ksi
1.8 g fy 60 ksi
0.0
8 0.8
0.0
1.2 4
0.0
3 0.25
fc Ag
Pn
1.0 0.0
2
Kn
0.0
1 0.50
0.8
0.75
0.6
1.0
0.4
t = 0.0035
0.2
t = 0.0050
0.0
0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35
Pne
Rn
fc Ag h
b
0.6
0.4
fc Ag
Pn
Kn
g 0.01
0.2 0.02
0.03
0.04
0.05
0.06
0.0
7
0.0
8
0
0.75 0.76 0.77 0.78 0.79 0.8 0.81 0.82 0.83 0.84 0.85 0.86 0.87 0.88 0.89 0.9
Figure A5.11 (a) ACI column interaction diagram [SP-17(11)], Courtesy of American Concrete
Institute. (b) Kn versus relationship
642 Appendix A: Tables and Diagrams
a 2.0
INTERACTION DIAGRAM C4-60.9 h
fc 4 ksi h
1.8 g
0.0 fy 60 ksi
8 0.9
0.0 Kmax
1.6 7
0.0
6
fs /fy 0 e Pn
1.4 0.0
5
0.0
4
1.2
0.0 0.25
3
fc A g
Pn
1.0 0.0
2
0.50
Kn
0.0
1
0.8
0.75
0.6
1.0
0.4
t = 0.0035
0.2
t = 0.0050
0.0
0.00 0.05 0.10 0.15 0.20 0.25 0.30 0.35 0.40
Pne
Rn
fc Agh
b 0.6
0.4
fc Ag
Pn
Kn
g 0.01
0.2 0.02
0.03
0.04
0.05
0.0
6
0.07
0.0
8
0
0.75 0.76 0.77 0.78 0.79 0.8 0.81 0.82 0.83 0.84 0.85 0.86 0.87 0.88 0.89 0.9
Figure A5.12 (a) ACI column interaction diagram [SP-17(11)], Courtesy of American Concrete
Institute. (b) Kn versus relationship
Appendix B
Concrete Structure and Construction Images
Figure B5.2 Main reinforcements and Ties for a column with 8 bars (see Figure 5.12e)
Euler buckling stress, 358, 359 beams and one-way slabs, 398399
Expansion joints, 24, 27 cost, 371, 372
Expansive soil, 420 drop panels, 373, 374, 390392
at slabs and plates, 372373
introduction to, 371372
F one-way joists, 395398
Failure superimposed dead loads, 378, 401
balanced, 67, 71, 72, 77, 318, 323, 327 two-way joists with slab band beams, 400
compression-controlled, 73, 74, 327 two-way slabs on beams, 381, 399400
modes, 6567, 7173, 495 wafe slabs, 392395
primary cause, 66 Footings
simultaneous, 67 pressure bulb under, 409
tension-controlled, 66, 73, 78 pressure distribution, 408410,
Fine aggregates, 5, 6 420421, 496
Fine-grained soils, 404, 405 rectangular, 418, 438, 460463, 465, 473
Fixity, 44, 46, 48, 360, 361, 384 soil failure under, 406407
Flange square spread, 436459, 514, 515
dened, 153 wall, 228, 411, 412, 421437, 479, 496,
effective width, 153156, 161, 164, 166, 497, 510, 605
170, 172, 173 Force distribution
neutral axis below, 156 column, 325, 329
Flat plate oor system Formwork
exure, 375, 380390 accessories, 529532
illustrated, 373 design
interior bay reinforcement design, 385 Adjustment Factors, 526, 528,
use, 373, 390392 529, 536
Flat slabs. See also Slabs Reference Design Stresses, 527
ACI Code minimum slab thickness, 373 design of, 533537
column strips, 381, 383385 beam forms, 557563
dened, 111 column forms, 546551
economy, 373 slab forms, 550557
exure, 380390 wall forms, 537545
exure analysis, 382385 lateral pressure on, 520522
gravity load shear transfer, 375 loads on, 519520, 522
middle strips, 381, 383385 materials for, 523533
moment transfer, 379 section properties, 525, 526
plan, 376 Foundations
recommended minimum thickness, 373 adjacent property lines, 418
shear periphery, 376378 combined footings, 411, 414, 418
shears, 375380 depth of frost penetration, 418
slab width, 385 elevations of adjacent footings, 418
with drop panels, 374, 390392 expansive soil, 420
with drop panels and column capitals, 374 isolated spread footing, 413
Flexural stiffness, 46, 383, 384 mat, 411, 414
Flexural strength, 57, 58 organic layers, 420
Flexure pier, 414, 417, 420
direct design method, 381 pile, 414417
at slabs, 380390 placement, 418420
reinforced concrete beams, 20, 5865, 79 shallow, 411414, 418
Floor beams, 41, 151, 258 strap footings, 411, 414
Floor load, 39, 55, 305, 307, 553, 554 types of, 411420
Floor systems wall footings, 412, 421436, 479, 496
beam-girder, 101, 112, 140, 142, 151, 152 Friction loss, 581
Index 669
G dened, 62, 69
Geometrical proportioning, of beams, 104 propagation of, 4648
Girders, 38, 42, 44, 101, 111113, 140, 142, on short length reinforced concrete
151, 264, 277, 360, 361, 372, 398, beam, 237
399, 573 Internal forces
Gravity loading, 284 on a section of, 235, 236
Gravity walls. See also Retaining walls propagation of, 4648
dened, 495 International Building Code (IBC), 2
illustrated, 496 International Conference of Building
lateral soil pressures, 474, 502 Ofcials (ICBO), 2
resisting moment calculation, 500 Isolated spread footing, 413, 414
safety factor, 500
soil pressure under footing, 500
stability analysis, 498 K
Grout key, 573 Kilograms (kg), 592594
H L
High-range water reducing agents, 8 Lab-cured cylinders, 14
Hinge support, 43, 44 Lateral
Hollow core decks earth pressure, 474479, 483
dened, 573 loads, 39, 283, 284, 347, 361, 399, 415,
illustrated, 573, 575 486, 522, 538, 544, 563, 573
shape, 573, 575 L-beams. See also T-beams
Hook actual effective depth, 103
ACI Code standard, 220 analysis illustration, 155165
bars developed by, 220 analysis steps, 166, 168
bars in compression and, 223 compression zone, 154
dened, 220 compression zone centroid, 155
90-degree, 223 dened, 154
illustrated, 220 deection check, 158, 160
tension bars terminated in, 220223 depth, 154
Hookes law, 17, 29, 60, 63, 200 design example, 166, 168
Hydration. See also Water design owchart, 168
dened, 3, 6 design resisting moment, 158
exothermic process, 10 design steps, 166
hardening, 7, 8, 10 effective ange width, 154
setting, 7, 9, 10 moment arm calculation, 159
stages, 7 neutral axis location, 155
strength development, 7 required area of steel, 155
resistance coefcient, 158
tensile force calculation, 156
I total area of steel, 155
Inuence area, 42, 43 total compression force, 154
In situ tests, 16 Linear elements, 111, 371
Instantaneous deections. See also Line supports, 49, 111
Deections Live loads (LL)
additional deformations, 213, 215 dened, 40, 49
calculations, 228 effects, 39, 49, 50
dened, 201 "ckleness, 4952
values, 213, 214 illustrated, 51
Internal couple negative moments due to, 49
after cracking, 64 one-way slabs, 111
components, 205 reduced, determining, 43
cracked rectangular section, 205 reduction factors, 4143
670 Index
Load(s) M
assumed, 122 Main reinforcement
axial, 55, 75, 263, 278, 282, 285, 286, 295, allowable spacing, 122
298, 303, 310, 316319, 324, 327, dened, 115
332, 334336, 339, 341344, 362, deection control thickness, 119
364367, 369, 568 maximum bar spacing, 118, 119
column, determining, 299, 308, 414 minimum, 98, 117119, 132, 166, 506
concentric, 317 one-way slabs, 96, 115119, 132, 144
dead, 28, 38, 39, 41, 49, 5255, 106, at supports, 129
110, 123, 139, 140, 142, 143, Mat Foundation, 411, 412, 414
165, 193, 196, 213215, 271, Maximum beam width (bmax), 96
272, 300, 312, 365, 366, 401, Maximum bending stress, 22
410, 426, 445, 465, 510, 511, Metric system. See also SI system
571, 599, 600, 603, 604 adoption ramications, 591
oor, 55, 101, 307 ASTM standard reinforcing bars, 595
gravity, 39, 151, 347, 380, 414, 475, 519, United States adoption history, 591592
520, 576, 583, 587 Middle strips. See also Column strips
lateral, 39, 283, 284, 347, 361, calculating, 386, 402
399, 415, 486, 522, 538, 544, negative reinforcing, 386
563, 573 positive reinforcing, 387
live (LL), 28, 3943, 4952, 5458, 80, Minimum reinforcements
101, 122, 123, 125, 127, 130, bar spacing, 119
139142, 144, 160, 193, 195, 196, concrete cover, 94, 118
213215, 272, 274, 282, 300, 305, main, 118, 132
312, 341, 365, 366, 378, 392, 401, one-way slabs, 117119
410, 422, 426, 437, 445, 465, 510, shrinkage and temperature, 118, 132,
511, 519 134, 506
one-way slabs, determining, 53, 113 Minimum steel ratio, 7980, 155, 156, 288
propagation of, 46 Modular ratio, 63, 206
service, 54, 57, 58, 96, 200, 201, Modulus of elasticity
207, 208, 212, 366, 398, 570, dened, 18, 20
581, 583, 587 example, 19, 63
slab behavior under, 111, 371 Modulus of rupture, 20, 21, 34, 62, 79, 146,
superimposed dead (SDL), 40, 101, 140, 148, 202, 204, 595, 603
193, 378, 392, 401, 510 Moment(s)
working, 54, 57, 64, 96, 530 with ACI Code coefcients, 101, 108,
Load balancing 125, 133
at center of span, 584 ACI Code design, 53
concept, 582, 584 bending, 34, 55, 5964, 75, 100, 108, 115,
equal loads at third points of span, 584 161, 184, 187, 195, 235, 236, 277,
method, 583 283, 284, 288, 317, 344, 380, 423,
uniform load, 583 440, 449, 457, 463, 481, 482, 495,
Load factors, 54, 55, 57, 481, 484 511, 535, 543, 576, 603
Load-path, 371 design resisting, 57, 69, 73, 82, 159, 163,
Longitudinal bars 165, 170, 172, 181, 184, 187, 190,
diameter of, 278, 290, 291 196, 334, 335
minimum number, 289 diagram of maximum, 49
spacing, 290, 291, 297, 299, 308 expression, 69, 222
spacing limits, 289, 291 at panel bay, 382
Long-term deections maximum bending, 108, 535
additional, 215 maximum, due to dead loads, 52
multiplier, 215 negative, 47, 4952, 108, 116, 129, 140,
Long-term loading, 17 141, 152, 155, 194, 195, 202, 205,
Index 671
O
One-way joists. See also Floor systems P
ACI Code designation, 395 Passive earth pressure, 475, 495, 497
beam sections, 397 Pedestals, 277, 278
square-/tapered-end pan layouts, 397 Pier foundations
standard pans, 396 caissons, 417
One-way shear dened, 417
rectangular footing, 462, 469 illustrated, 417
square spread footing, 440, 441, 448 Pile foundation
One-way slab. See also Slabs battered piles, 415, 416
analysis, 53, 120 dened, 414
analysis owchart, 121 end-bearing piles, 415
672 Index
Soil Splices
allowable bearing pressure, 410, 511 compression, 225, 445, 459
classication, 404 of reinforcement, 224233
coarse-grained, 403, 405 proprietary mechanical, 225
cohesive, 404, 408, 412, 417, 475 tension, 218219, 224
constituents, 408 welded, 224, 225
expansive, 420 Split-barrel sampler, 405, 406
failure under footings, 406407 Square column, 303, 310, 313, 350, 352, 366,
ne-grained, 404, 405 379, 461
lateral pressures, 511 Square spread footing. See also Footings
noncohesive, 404 bending moment, 440, 449
settlement, 408410 critical sections, 438, 439
Standard Penetration Test (SPT), 404405 design, 436459
test borings, 404405 design examples, 451, 452
types, 410, 411, 476 design owchart, 466
vertical pressure under retaining wall, design steps, 437
496, 497 nominal shear capacity, 454
wedge, 407, 474 one-way shear, 440, 441
Soil pressure required area of steel, 438, 439, 442, 450,
allowable bearing, 403, 410411 463, 464
lateral, 474, 484, 499, 502 required reinforcement, 442, 449, 457,
under footing, 420421 463, 469
vertical, under retaining walls, 496 shear, 438441
Soil pressure under footings shear capacity calculation, 455
distribution, 420421 size determination, 446, 453
failure, 406407 thickness, 438
Southern Building Code Conference two-way shear, 438, 440, 441, 447
International (SBCCI), 3 use, 436, 438, 442, 445, 450
Span/depth ratio, 92, 393, 397, 582 Standard Penetration Test (SPT), 404406
Spiral(s) Steel. See also Reinforcing steel
clear space between, 302 behavior under stress, 29
composition, 277, 278 compression, 176, 177, 179, 181186, 189,
diameter, 315, 355 190, 192, 193, 198200, 319, 320,
maximum pitch, 316 322, 323, 325, 326, 329, 330
pitch, 293, 305, 312, 315, 316, 356 modulus of elasticity, 177, 206
short columns with large eccentricity, 337, net tensile strain, 7275, 183, 189, 335
339, 341, 350, 352, 354356, 358 percentage, limitations, 70
steel design, 304, 312 prestressing, 580
steel ratio, 291, 292, 298, 301, 304, reinforcing, 24, 26, 2835, 38, 59, 6366,
312, 315 70, 72, 75, 78, 79, 97, 104, 120,
Spiral columns. See also Columns 132, 147, 215, 216, 244, 398, 532,
dened, 304, 312 577, 578
design steps, 288, 304, 356 spiral, 291, 292, 298, 301, 304, 312, 315
failure, 285 stress-strain diagram, 16, 17, 29, 65, 66
gross area calculation, 288, 293, 310, tensile, 60, 6264, 75, 96, 239, 329,
312, 349 424, 426
illustrated, 286, 289, 293 Steel bars
minimum number of bars, 289, 304, 312 deformed, 29
required area of steel, 303, 304, 308, 312 epoxy-coated, 219, 229
required column area, 309 identication, 31
size, 312 mechanical properties, 30
Spiral reinforcement, spacing/amount plain, 30
limits, 291 sizes, 31
Index 677
Steel ratio ultimate, 17, 33, 5459, 65, 68, 74, 75,
balanced, 7172 7780, 96, 98, 218, 296, 382,
calculation, 70, 72, 120, 123, 165, 432 581, 582
dened, 70 useable, 56, 57
design, 104, 356 Strength reduction factor
minimum, 7980, 155, 156, 288 dened, 56
short columns, 350 obtaining, 89, 345
short columns with large eccentricity, value, 79, 100, 158
336358 Stress
spiral, 291, 292, 298, 301, 304, 312, 315 after cracking, 63
T-beam, 155 axial compressive, 282, 583
Stirrups. See also Shear reinforcement bending, 62, 282, 317, 524, 527529, 533,
closed, 246, 266, 380 535, 539, 546, 558
design owchart, 252 bond, 216218
design procedure, 251261 column, distribution, 325
design zones, 246, 250261 compression, 16, 64, 66, 151, 282, 295, 317
oor beam, 258 compression steel, 177, 182, 183, 189, 323,
horizontal, 270 325, 326, 330
layout illustration, 261 cracked rectangular section, 205
minimum amount, 245, 260 distribution, 6061, 64, 319
number of legs of, 251 distribution, at failure, 320
shear resistance, 250, 259 Euler buckling, 358, 359
size, limiting, 250 horizontal shear, 237
spacing, 220, 244, 245, 247, 253, maximum bending, 22
255257, 260 prestressed beam, 581
spacing, calculating, 246, 247 principal, 241, 581
spacing, maximum, 246, 247 steel behavior under, 2935
Strain tensile, 20, 22, 24, 25, 27, 33, 60, 62, 63, 96,
after cracking, 63 151, 235, 239, 241, 329, 424, 426,
columns, distribution, 325, 328, 329 567, 570, 582, 587, 596
compression steel, 177, 182, 185, 186, 188, on unit cube, 240, 241
192, 198, 319, 320, 322, 323, 329 units of, 87
cracked rectangular section, 205 yield, 30, 70, 295, 322
distribution, at failure, 67, 71, 320 Stress-strain diagram
doubly-reinforced beams, 177 bilinear, 65, 66
net tensile, 7275, 177, 183, 189, 335 illustrated, 17, 29
tensile steel, 179, 188, 321, 322, 328, zone, 29, 30
330, 339 Structural design
ultimate useful, 17, 65, 318, 335 iterative nature, 39
Strain distribution process, 3841
at balanced failure, 67, 71 Structural stiffness, 46
doubly-reinforced beams, 177 Structural Welding Code-Reinforcing
on reinforced concrete section, 80 Steel, 224
Strain hardening, 29, 65 Superimposed dead loads (SDL), 28, 3941,
Strap footing, 411, 414, 418 59, 101, 140, 144, 193, 378, 392,
Strength 401, 510
compressive, 8, 14, 1621, 38, 65, 66, 80, Superplasticizers, 8, 10, 13
294, 298, 300, 319, 332, 426, 445, Superposition, 176, 177, 583
567, 570, 582, 595, 600 Support(s)
design, 5355 absolute conditions, 44
need for, 9, 54 conditions, 44
nominal, 56 continuous media, 111
required, 57, 243, 541 hinge, 43
678 Index