Professional Documents
Culture Documents
Wa S cammn 3)
Hansurjniin wn ntciecosva inant
serietteeseeneee ay
SOUULEAEE ALAA _-
4
Digitized by the Internet Archive
in 2021 with funding from
Kahle/Austin Foundation
https://archive.org/details/structuralanalysOO0Ochaj
Structural Analysis
PRENTICE-HALL INTERNATIONAL SERIES
IN CrvIL ENGINEERING AND ENGINEERING MECHANICS
William J. Hall, Editor .
Au and Christiano, Structural Analysis
Barson and Rolfe, Fracture and Fatigue Control in Structures, 2/e
Bathe, Finite Element Procedures in Engineering Analysis
Berg, Elements of Structural Dynamics
Biggs, Introduction to Structural Engineering
Chajes, Structural Analysis, 2/e
Cooper and Chen, Designing Steel Structures
Cording, et al., The Art and Science of Geotechnical Engineering
Gallagher, Finite Element Analysis
Hendrickson and Au, Project Management for Construction
Higdon, et al., Engineering Mechanics, 2nd Vector Edition
Holtz and Kovacs, Introduction to Geotechnical Engineering
Humar, Dynamics of Structures
Johnston, Lin, and Galambos, Basic Steel Design, 3/e
Kelkar and Sewell, Fundamentals of the Analysis and Design of Shell Structures
MacGregor, Reinforced Concrete: Mechanics and Design
Mehta, Concrete: Structure, Properties and Materials
Melosh, Structural Engineering Analysis by Finite Elements
Meredith, et al., Design and Planning of Engineering Systems
Mindess and Young, Concrete
Nawy, Prestressed Concrete
Nawy, Reinforced Concrete: A Fundamental Approach, 2/e
Popov, Engineering Mechanics of Solids
Popov, Introduction to the Mechanics of Solids
Popov, Mechanics of Materials, 2/e
Schneider and Dickey, Reinforced Masonry Design, 2/e
Wang and Salmon, Jntroductory Structural Analysis
Weaver and Johnson, Finite Elements for Structural Analysis ‘
Weaver and Johnson, Structural Dynamics by Finite Elements
Wolf, Dynamic Soil-Structure Interaction
Wray, Measuring Engineering Properties of Soils
Yang, Finite Element Structural Analysis
second edition
Structural Analysis
ALEXANDER CHAJES
Department of Civil Engineering
University of Massachusetts
= ,
Chajes, Alexander
Structural analysis / Alexander Chajes. — 2nd ed.
cm. — (Prentice-Hall international series in civil
engineering and engineering mechanics)
ISBN 0-13-855073-5
1. Structural analysis (Engineering) In Ditles I senies:
TA645.C37 1990
624.1°71—dc20 seed
The author and publisher of this book have used their best efforts in preparing this
book. These efforts include the development, research, and testing of the theories
and programs to determine their effectiveness. The author and publisher make no
warranty of any kind, expressed or implied, with regard to these programs or the
documentation contained in this book. The author and publisher shall not be liable
in any event for incidental or consequential damages in connection with, or arising
out of, the furnishing, performance, or use of these programs.
ISBN: 0-13-855073-5
Preface xi
Introduction 1
1.1 Structural Engineering 2
1.2 Structural Form 2
1.3. Materials 10
1.4 Loads 13
1.5 History of Structural Engineering 16
Calculation of Reactions 21
2.1 Introduction 22
2.2 Equations of Equilibrium 22
2.3 Types of Supports and Restraints 23
2.4. Stability and Determinacy 24
2.5 Calculation of Reactions: Simple Structures 28
2.6 Calculation of Reactions: Compound Structures 30
Problems 34
Contents
vi
Plane Trusses 41
Introduction 42
Simplifying Assumptions 42
Basic Concepts 44
Method of Joints 45
Method of Sections 51
Methods of Joints and Sections Combined 54
Stability and Determinacy of Trusses 58
Computer Solution 60
Problems 68
Space Trusses 77
4.1 Introduction 78
4.2 Calculation of Reactions 78
4.3 Calculation of Member Forces 79
Problems 86
Index 467
fy a ia
~ my
. _
- !
at ae ~
jm
‘a oP do we
: in ag
Ai \
‘a Aust 4 e oi
mie »* z
i “pes
vane af
peryy <) “4 7
io 24 “Pre I ; a
rh 73
a toiceh » ar Hoi Out tages "pcre
ty iy epeinwcer ¥;
ee
e ee
The aim of this book is to present the fundamentals of structural analysis and to
serve as a textbook for one or more courses in the subject. The material covered
by the book can be subdivided into three parts. The first part, which includes
Chapters 1 through 9, deals with the analysis of simple, determinate structures.
This section covers the analysis of trusses, beams, frames, arches, and cables, as
well as methods for calculating deflections and the use of influence lines for
moving loads. The second part, which includes Chapters 10 through 14, presents
the classical methods of analyzing indeterminate structures. Two force methods,
the method of consistent deformation and the method of least work, and two
deformation methods, the slope-deflection method and moment distribution, are
included. The last part of the book, Chapters 15 and 16, deals with matrix
analysis of structures. Both the flexibility method and the stiffness method are
presented.
The first two sections of the book, dealing with the analysis of determinate
and indeterminate structures by classical methods, are presented without recourse
to matrix algebra. It is felt that the inclusion of matrix algebra in this part of the
book would only detract from the principal objective, which is to introduce the
student to the fundamentals of structural analysis. Once the student has mastered
the basic principles presented in the first part of the book, he or she should have
no difficulty with the concepts of matrix analysis covered in the latter part of the
xi
xii Preface
Alexander Chajes
Introduction
,
Mackinac Straits Bridge, St. Ignace, Mich.
(Courtesy of American Bridge Division, U.S. Steel Corporation.)
1.1 STRUCTURAL ENGINEERING
The purpose of this book is to introduce the student to the principles of structural
analysis. However, structural analysis is part of the larger subject of structural
engineering, and it is with the latter that we will begin.
Structural engineers may find themselves involved in the design of a large
variety of projects, including buildings, bridges, dams, ships, airplanes, sports
stadiums, and nuclear power plants. Although each of these projects has a
different purpose, they all have some common characteristics. To begin with, they
are all built to satisfy human needs. Buildings are erected to provide shelter,
bridges to facilitate the movement of people across rivers, dams to contain large
quantities of water, and airplanes to provide rapid transportation. Furthermore,
in each of these examples the structure, while carrying out its intended function,
is subjected to various loads. The structure must always support the load due to
its own weight. In addition, it must resist loads such as the weight of snow on the
roof of a building or the dome of a sports stadium, the weight of people and
vehicles on a bridge, the pressure of water on the sides of a dam, and the force of
the wind on the wings of an airplane. It is the responsibility of the structural
engineer to see to it that these loads can be resisted by the structure both safely
and economically.
The structural design of a project can usually be broken’down into the
following four steps:
1. Selection of the type of structural form to be used and the material out of
which the structure is to be made.
2. Determination of the external loads that can be expected to act on the
structure.
3. Calculation of the stresses and deformations that are produced in the
individual members of the structure by the external loads.
4. Determination of the size of individual members so that existing stresses
and deformations do not exceed allowable values for the given material.
It has been pointed out that a structure, while carryinggout its intended function,
is subjected to various loads that it must resist without either collapsing or
deforming excessively. To put it very simply, resisting a load means transferring
2
Sec. 1.2 Structural Form 3
the load from the point at which it is applied to some type of rigid foundation.
For example, the frame of a building transmits the loads due to wind and snow,
which are applied to the sides and the roof of the building, to the foundation in
the ground. Similarly, a bridge transmits the weights of vehicles applied along its
span to the abutments at the ends of the span. The manner in which a structure
carries Out this transmission of loads depends largely on its geometrical shape or
form. Although there exist an endless number of different shapes and forms that
a given structure can assume, it can be shown that these are to a great extent
simply variations of a handful of basic forms.
The most important characteristic of a structural form is the type of stress it
develops in resisting applied loads, that is, uniform axial stress, bending stresses,
or a combination of the two. Accordingly, structural forms can be divided into
the following categories.
A. Cables
Probably the simplest way of transmitting a load P laterally to two supports A and
B is by means of a cable as shown in Fig. 1.la. In this system the load is
maintained in equilibrium solely by means of uniform tension stresses in the
cable. Since the applied load is balanced by the vertical components of the cable
forces, the cable must have a sag in order to resist the load. The larger the sag,
the smaller will be the tension that the load induces in the cable. For a cable
system to function properly, the supports must be able to resist the pull of the
cables.
B
A Cable Y
(b)
Figure 1.1
4 Introduction Chap. 1
oi Mert
#4] Roadway
in cet,
Cable
anchorage
Figure 1.2
If the position of the load changes or if several loads act simultaneously, the
geometry of the cable changes accordingly, as shown in Fig. 1.1b. This ability of
the cable to adjust its shape to different loading conditions is an extremely
important characteristic of the cable. It allows the cable always to assume that
shape for which only tension stresses are needed to resist the applied load. The
shape a cable assumes when subjected to a given loading is called the funicular
curve for that loading. For example, the funicular curve for a load that is
uniformly distributed in the horizontal direction is a parabola. The load that is
applied to the cables of a suspension bridge is very nearly uniform in the
horizontal direction, and the cables accordingly assume shapes that are ap-
proximately parabolic. If the load is uniformly distributed along the length of the
cable, the cable will take on the form of a catenary. A cable of uniform cross
section subjected only to its own weight will assume the form of a catenary.
Probably the most familiar example of a cable structure is the suspension
bridge. As shown in Fig. 1.2, the basic structural system consists of a horizontal
deck or roadway, two vertical towers, and a pair of cables draped over the
towers. The roadway is suspended from the cables by a series of vertical rods
called hangers. The cables in turn are held up by the towers and anchored at their
extremities by large concrete foundations. The entire weight of the deck, plus
that of any vehicles thereon, is thus transferred to the towers and the anchorages
at the ends of the cables by tension stresses in the cables.
B. Arches
If the cable in Fig. 1.1a is inverted and replaced by two rigid bars, we have the
simple arch shown in Fig. 1.3. In this system the load P is transmitted to the
supports at A and C by means of uniform compression stresses in each of the
struts. As was the case with a cable, it is necessary in the design of an arch to
ensure that the foundations are strong enough to supp®rt the system adequately.
In contrast to the cable, which pulls on the supports, the arch exerts outward
thrusts on its supports.
For any single loading condition, it is possible to design an arch that can
resist that loading by developing only axial compression. However, most arches
Sec. 1.2 Structural Form 5
Figure 1.3
are subjected to several different loading conditions during their life, and since an
arch, unlike a cable, is a rigid structure that cannot change its shape when
subjected to different loadings, it will develop bending stresses in addition to axial
compression. In designing an arch it is customary to choose its shape so that the
predominant loading will produce a uniform compression stress. The bending that
then results from other loads will be minimal. For example, the main loading of a
bridge is its own weight, and since this load roughly corresponds to a uniformly
distributed load, most arch bridges are approximately parabolic in shape.
C. Trusses
The lower ends of the two slanting members of the arch in Fig. 1.3 are prevented
from moving outward by the supports at A and C. Alternatively, the lower ends
of the structure can be kept from spreading apart by connecting them with a third
bar as indicated in Fig. 1.4. If this is done, we have, instead of an arch, a truss.
The slanting members are still in compression, but there is now an additional
member, the horizontal bar, that is in tension.
In general, a truss is a structure composed of axially stressed bars, some of
which are in tension and some in compression, and in which the bars are arranged
to form one or more triangles.
Figure 1.4
6 Introduction Chap. 1
Figure 1.5
D. Beams
The beam shown in Fig. 1.7 transfers the load P laterally to the supports at A and
B by bending. A bent member takes on a curved shape, causing the fibers along
the concave edge to be shortened and in compression and the fibers along the
convex edge to be lengthened and in tension. An advantage of the beam over the
arch and the cable is that it requires only vertical support at its ends. Unlike the
arch and the cable, it does not require any horizontal restraint. The abutments on
which the ends of a beam rest are therefore easier to design than those required
ANNSNSSA
Figure 1.6
Sec. 1.2 Structural Form 7
>
——— Compression
> Tension
Figure 1.7
by the arch and the cable. In this respect the truss is similar to the beam. It also
does not require horizontal restraints at its ends.
Another important advantage of the beam is its compactness. A beam
usually occupies less vertical space than any of the structural forms considered so
far. This attribute of the beam makes it especially attractive for supporting floors
and ceilings in multistory buildings where vertical space is limited.
A disadvantage of the beam is that it sometimes uses material less
economically than other structural systems. In an axially loaded member such as a
cable or the bar of a truss, the entire cross-sectional area of the member is
subjected to a uniform stress. Thus every element of the section can be stressed
up to the allowable strength of the material. By comparison, the magnitude of the
axial stress varies from a maximum at the upper and lower edges of a beam to
zero at the center of the cross section. As a consequence, fibers near the center of
the beam are stressed relatively slightly when the outer fibers are working near
their maximum capacity. The fact that a truss could be obtained by removing
much of the material of a beam, as shown in Fig. 1.8, demonstrates that a beam
uses material less efficiently than a truss.
ped
ee || Beam
NNN
Truss
Figure 1.8
8 Introduction Chap. 1
Figure 1.9
Figure 1.10
Sec. 1.2 Structural Form 9
Y a
A] x
A ~y
Curved surface
Flat plate
Figure 1.11
(INS
Cylindrical
vault
Dome
Figure 1.12
10 Introduction Chap. 1
cylindrical barrel shell and the dome in Fig. 1.12 can be formed, respectively, by
placing several arches side by side and by rotating an arch about a vertical axis
through its apex.
1.3 MATERIALS
To a very large degree, the capacity of a structure to resist external loads without
collapsing or deforming excessively depends on the material out of which it is
made. The properties of a material that most interest the structural engineer are
its strength and its deformation characteristics.
The strength of a material is a measure of its ability to resist stress without
fracturing. Most important is the capacity of the material to resist tension and
compression stress. Steel is probably the best generally available structural
material because of its ability to resist equally well high tension and compression
stresses without fracturing. By comparison, stone and concrete are fairly strong in
compression but relatively weak in tension. Whereas structural steel can safely
withstand tension and compression stresses of approximately 30 ksi (207 MN/m’),
concrete can resist 3 to 7ksi (21 to 48 MN/m’) in compression but only about
0.5 ksi (3.5 MN/m’) in tension.
Since a structure must always support its own weight, the strength-to-weight
ratio of a material is of considerable importance. Wood, for example, has a
strength along the grain only one-tenth that of structural steel. However, the unit
weight of wood is also only 10% of the unit weight of steel. Per pound of
material, wood is thus able to support as much load as steel. Because a high
strength-to-weight ratio is of paramount importance in airplanes, aluminum alloys
are used in their construction instead of steel. These alloys, although they are
more expensive than steel, have about the same strength as structural steel but
weigh only one-third as much.
For a material to be suitable as a structural component, it must in addition
to possessing sufficient strength have acceptable deformation characteristics.
When subjected to stresses whose magnitudes are comparable to those expected
to occur during the everyday life of the structure, the material should not deform
excessively. In other words, the material must be relatively stiff. For this reason
rubber, which is very flexible, is not a satisfactory structural material. Excessive
deformations are objectionable because they tend to impair the proper function-
ing of the structure. For example, large deformations in the framework of a
building may prevent doors and windows from opening and closing properly.
Similarly, plaster ceilings and walls would crack if the members supporting these
surfaces were to deform excessively.
In addition to requiring that a structural material be stiff, it is necessary that
deformations resulting from moderate, everyday loads disappear after these loads
cease to exist. If this is not the case and deformations are cumulative, repeated
applications of even small loads will eventually lead to excessive deformations
Sec. 1.3 Materials 11
regardless of whether the material is stiff or not. The property that enables a
material to recover its original size and shape after a load is removed and
prevents deformations from being cumulative is elasticity. All commonly used
structural materials, including steel, concrete, and wood, behave elastically for a
significant portion of their load-carrying capacity. By comparison, clay is a plastic
material; deformations in clay resulting from the application of a load do not
disappear when the load is removed.
Although all structural materials, if stressed sufficiently, will eventually
fracture, the manner in which the fracture occurs differs from one material to
another. Some materials, even though they are stiff at low stresses, experience
very large deformations before they actually break, while others fracture
suddenly without any indication that the material is approaching its limit of
resistance to stress. The materials that remain stiff up to the instant of fracture
are known as brittle materials. Concrete, stone, cast iron, and glass are examples
of this type of material. By comparison, steel, aluminum, and most metals are
ductile. These materials, although they are stiff at low stresses, become very
flexible and deform considerably before they actually break. In general, a ductile
material is preferable to a brittle one because the ductile material gives ample
warning of an impending failure, whereas the brittle one does not. Either the
excessive deformations in the ductile material are noted visually and the load that
is acting is physically removed, or, if this is not possible, the structure itself
redistributes the stresses from the members that are being overstressed to others
that still have additional capacity to resist load. Neither of these alternatives is
possible in a brittle material.
In addition to brittleness, there are many other undesirable material
characteristics. In fact, almost every structural material has some shortcomings.
Steel loses a significant part of its strength at high temperatures and is very
susceptible to corrosion. Concrete is brittle and can resist only small amounts of
tension, and unreinforced plastics possess neither adequate strength nor adequate
stiffness. Fortunately, it is possible by judiciously combining several materials to
overcome most of these as well as other material deficiencies. A layer of paint
will prevent steel from corroding, and a cover of a few inches of concrete will
enable steel to maintain its strength during a fire. A somewhat more sophisticated
procedure for enhancing the properties of a material is the combination of
concrete and steel to form reinforced concrete. The load-carrying capacity of a
beam composed solely of concrete is severely limited by the inability of the
concrete to resist the tension stresses that occur on the convex side of the beam.
As soon as these tension stresses exceed the minimal tensile strength that
concrete possesses, the member cracks and collapses. To remedy the situation, it
is customary to embed steel bars in the convex side of the beam, where the
tension stress occurs. These steel bars then resist the tension stress that exists on
the convex side of the beam, while the concrete itself resists the compression
stress on the concave side. Thus it is possible, by reinforcing a concrete beam
with a sufficient amount of steel on its convex side, to overcome the weakness of
12 Introduction Chap. 1
concrete in tension and produce a beam that can resist tension and compression
stresses equally well.
The practice of reinforcing one material with another, used to such great
advantage in the production of reinforced concrete, has been applied in other
instances as well. In recent years a new group of materials, commonly referred to
as plastics, has been used extensively in a variety of situations. The attributes of
plastics that have led to their widespread use are the ease with which they can be
formed into even the most intricate shapes and their high resistance to most forms
of corrosion. However, by themselves they are neither strong nor stiff enough to
make a satisfactory structural material. They are weak in both tension and
compression and deform excessively when subjected to loads. To overcome these
shortcomings, it is customary to reinforce the plastics with glass fibers and form a
new group of materials called fiber-reinforced plastics. By themselves neither glass
nor plastics are satisfactory structural materials; the former is too brittle and the
latter too weak and too flexible. Together, however, they form a group of
composite materials that are relatively strong and stiff and that behave in a ductile
manner.
Most material properties that interest the structural engineer can be
obtained from a stress-strain diagram of the material. For example, let us consider
the stress-strain diagram of A36 structural carbon steel shown in Fig. 1.13. The
curve indicates that the ultimate strength of the material is approximately 60 ksi.
In other words, 60 ksi is the maximum stress that the material can resist without
rupturing. Although the material does not fail until the stress reaches 60 ksi, it is
undesirable for the stress under ordinary conditions to exceed 36 ksi, the yield
Ultimate strength
60
Yield strength
ow oO
Stress
(ksi)
20
Elastic range
0.20 0.30
Strain (in/in) y
Figure 1.13
Sec. 1.4 Loads 13
strength of the material. As long as the stress remains below the yield strength,
the material behaves in an elastic manner, that is, there are no permanent
deformations upon removal of the load. Furthermore, by limiting the stress to the
region below the yield strength, where the material is very stiff, it is possible to
prevent excessive deformations.
The slope of the stress-strain diagram-is a measure of the stiffness of the
material. Within the elastic range, which is the range of stress in which the
material is expected to be working, it takes a stress of 30 ksi to stretch steel
one-thousandth of an inch per inch. Steel is thus an extremely stiff material for
the range of stresses in which it is expected to perform. Steel is also a very ductile
material, as indicated by the fact that it can undergo a total elongation of
approximately 25% before it ruptures.
1.4 LOADS
After a decision has been reached regarding the overall size and shape of a
structure and the material out of which the structure is to be made, it is necessary
to determine the loads that the structure must resist. All loads that can
reasonably be expected to act on the structure during its entire life must be
considered. This includes the construction phase as well as the time during which
the finished structure is in service. From among these loads, we choose those that
produce the largest stresses in the structure and then design the individual parts
of the structure so that the stresses can be safely resisted.
Loads are usually separated into specific categories. For example, there are
static and dynamic loads. A static load is one that is applied so slowly that the
structure remains at rest during the application of the load. By comparison, a
dynamic load is one that is applied rapidly enough to cause the structure to
accelerate and as a consequence gives rise to inertia forces. Thus, when dealing
with static loads, we have only the load itself to consider, whereas a structure
subjected to dynamic loads must resist the combined effect of the applied load
and the inertia forces.
Another important distinction made among loads is whether they are dead
loads or live loads. Dead loads are those whose magnitude remains constant and
whose position does not change with time. They consist of the weight of the
structure itself plus the weight of any objects permanently attached to the
structure. For example, the dead loads in a building include the weight of the
beams and columns that make up the frame, the weight of the walls, floors, and
roof, and the weight of permanent fixtures such as plumbing, heating, and
air-conditioning equipment.
The weights of a few common building materials are given in Table 1.1.
In comparison with dead loads, live loads are those whose magnitude and
position do not remain constant. Thus, when we consider live loads, we must
determine not only the maximum magnitude of the load but also the location of
14 Introduction Chap. 1
Brick 120 19
Concrete 150 24
Sand 100 16
Steel 490 78
Timber 40 6.4
the load that results in the largest stresses. Some examples of live loads are the
weight of occupants in a building, the weights of vehicles on a bridge, and the
forces exerted by wind on various structures. In each of these instances, the
position of the force, as well as its magnitude, is subject to change.
Some commonly encountered live loads are briefly discussed in the
following paragraphs.
Live loads in buildings are primarily due to the weight of people and movable
objects. It is customary to represent these forces by uniformly distributed loads.
For example, Table 1.2 lists some loads recommended by the American National
Standards Institute (ANSI) Code to be used in the design of buildings. The
intensity of the load is seen to depend on the use to which the area in question
will be put.
The live loads that act on a highway bridge are mainly due to the weights of the
vehicles that cross the bridge. Since bridges are subjected to various combinations
of many different vehicles during their lifetime, design codes have developed
Sec. 1.4 Loads 15
H20-44 loading
Figure 1.14
standard equivalent loadings whose use leads to a safe and economical structure.
For example, the specification of the American Association of State Highway and
Transportation Officials (AASHTO) recommends that highway bridges be
designed to support standard trucks like the one shown in Fig. 1.14. The first
number in the truck designation stands for the weight of the vehicle in tons, and
the second number specifies the year that the loading was adopted.
Snow Loads
The weight of snow is an important live load in the design of roofs. Two factors
that govern the intensity of the snow load to be used are the geographical location
of the building and the slope of the roof. Table 1.3 gives snow loads for several
areas in the United States. As is to be expected, the snow load is heavier in the
northern part of the country than in the south. In addition, the steeper the roof,
the smaller the amount of snow that can collect.
Slope of roof
New England
and 45 30 10
Great Lakes region
Central states 30 20 10
Southern states 20 15 10
16 Introduction Chap. 1
Wind Loads
All exposed surfaces of structures are subject to wind loads. The larger the
exposed area, the larger will be the total force to which the structure is subjected.
Wind loads are especially important for structures such as high-rise buildings,
radio towers, and suspension bridges.
The pressure exerted by the wind on a surface depends in general on the
velocity of the wind and on the nature of the surface. A relation giving the
pressure is
p = 0.00256CV*
where p = pressure in psf
C = a factor that depends on the nature of the surface
V = wind velocity in mph
or
p = 0.0000473CV7
where p is measured in kN/m? and V in km/hr. :
The value of C for buildings with vertical sides has been found to be 1.3.
This is a combination of 0.8 due to pressure on the windward side and 0.5 due to
suction on the leeward side. Assuming a wind velocity of 100 mph, the wind load
on a vertical building would be 33.3 psf.
The live loads to which structures can be subjected are almost too numerous to
mention. They include forces produced by earthquakes, water pressure on coastal
structures, soil pressure on underground structures, and forces that arise when
the temperature changes and the structure is not free to expand or contract.
If many structures of the type being designed have already been built, the
chances are high that the necessary loads will be given in readily available design
codes. However, the loads to which new types of structures are subjected may not
be recorded anywhere and will consequently have to be determined specially for
the structure being built.
they employed beams and columns as their main structural elements. Unfortun-
ately, stone is relatively weak in tension and is therefore a poor material for
making beams. To avoid the formation of cracks on the tension side of the
member, it is necessary to make the beam relatively short. The resulting
structures (Fig. 1.15) consisted of numerous columns with relatively little useful
space between them. :
In comparison with stone beams, stone arches represent a_ structural
engineering achievement of the first order. Although the arch was discovered
prior to the Roman era, it was not used extensively until the time of the Roman
Empire. The arch is primarily stressed in compression, and stone is an excellent
material for resisting compression. The stone arch thus represents an ideal
marriage of form and material. Compared to stone beams, whose lengths may not
have exceeded 25 feet, stone arches spanned distances of 100 feet and more.
In addition to using the arch to build bridges and aqueducts, the Romans
employed it to construct vaulted roofs and domes. The Pantheon, one of the truly
magnificent structures built in Roman times, was topped by a dome spanning 142
feet. Structures spanning distances of this magnitude would not be built again
until modern times.
With the decline of the Roman Empire in the fifth century came a decrease
in structural engineering activity. The period lasting from about a.p.500 to
A.D. 1500 is known as the Middle Ages or more appropriately the Dark Ages.
During this time, Western civilization experienced a general decline, and it is thus
not surprising that no significant progress was made in the area of structural
engineering. The structures that were built continued to employ the stone arch as
the major structural form. The most noteworthy among these are the Gothic
cathedrals build during the twelfth and thirteenth centuries. These towering
structures, soaring toward the heavens, symbolized the spiritual tone of the day.
A most interesting aspect of the cathedrals is the manner in which the roof
is supported. Unlike the arches in Roman bridges, which rested on massive
abutments, the arched roofs of the cathedrals were supported on tall, thin walls
containing glass windows over a sizable portion of their surface. Since walls such
as these are incapable of resting the outward thrust that the legs of an arch exert
on their supports, they were braced by inclined members called flying buttresses
Figure 1.15
Introduction Chap. 1
18
Flying
buttress
Figure 1.16
(Fig. 1.16). The flying buttress was an ingenious device without which many of
the splendid cathedrals could not have been constructed.
The fifteenth and sixteenth centuries, known as the Renaissance, were a
period of intense activity for Western civilization. It was during these years that
many of the foundations of modern society were established. During the Middle
Ages people had been more preoccupied with the spiritual than the secular. All
this now began to change. The human was no longer considered to be a sinful
being whose nature had to be repressed. Instead people felt encouraged to
express their individuality and to participate in the molding of their destiny.
As far as structural engineering is concerned, the renaissance had its
greatest impact in the realm of science. It was at this time that Galileo
(1564-1642) laid the groundwork for the science of structural analysis. In the first
book ever written on the subject, Galileo discussed the concepts of force and
moment and analyzed the strength of structural members. Although not all of his
work was accurate, he stated correctly that the strength of an axially loaded bar
depends on its area and that the strength of a beam is proportional to its depth
squared. Another of Galileo’s contributions was the influence he exerted on
others. Among those following in his footsteps were Robert Hooke (1635-1703),
who formulated the law of linear behavior of materials bearing his name; Sir
Isaac Newton (1642-1727), who formulated the laws of motion; and Leonhard
Euler (1707-1783), who was the first to analyze correctly the buckling of
columns.
In addition to the basic work in structural meghanics that was initiated
during the Renaissance, an Italian architect, atti is believed to have
introduced the use of the truss. The trusses that he and his contemporaries built
were modest timber structures whose purpose was more to take the place of
simple wooden beams than to replace stone arches. In fact, it was another 200
Sec. 1.5 History of Structural Engineering 19
years before the use of trusses became widespread. However, the seed of a new
and major structural form had been planted.
If the work of men like Galileo and his followers provided the scientific base
of the modern era of structural engineering, then the introduction of iron as a
building material during the Industrial Revolution can be said to have provided
the physical material for modern structural engineering. Humans had been using
iron for limited purposes ever since the time of the pyramids. However, it was not
until the Industrial Revolution occurred, during the latter half of the eighteenth
century, that iron was produced in sufficient quantity to make its use as a building
material feasible.
The first major structure built of iron was a bridge over the Severn River
near the English town of Coalbrookdale. Although the cast iron out of which this
bridge was constructed is a crude material compared to the steel we use today, a
new era in structural engineering had been inaugurated.
The first few iron bridges, like the one at Coalbrookdale, employed the old
familiar form of the arch. However, engineers soon recognized the tremendous
potential of the new material and started to use it in a variety of novel ways. An
especially important use of iron was made in conjunction with the construction of
suspension bridges. The early part of the nineteenth century witnessed the
erection of several suspension bridges both in England and in America, using iron
chains to support the roadway. Among these were Thomas Telford’s bridge over
the Menai Straits in Wales, Brunel’s Clifton Bridge in Bristol, England, and
Finley’s bridge over the Merrimack River in Newburyport, Massachusetts.
Telford’s bridge had a span of 579 feet, the longest in the world at the time and
probably at least twice as long as any stone arch or timber truss built during the
eighteenth century.
In addition to its use in suspension bridges, iron was employed extensively
during the nineteenth century for the construction of truss bridges. At the
beginning of the eighteenth century, most highway bridges were made out of
lumber. Many of the covered bridges that are still standing today in parts of the
United States date from that period. However, these bridges were neither strong
nor stiff enough to serve the needs of the railroads that were starting to develop
at the time. As a consequence, by the latter part of the eighteenth century
numerous iron truss bridges were being built to replace the old timber structures.
The iron truss was a popular structure. It was reasonably strong, easy to build,
and not too expensive. Unfortunately, many of these structures were poorly
designed, and the iron being used at the time was not a very reliable material. As
a result, numerous failures occurred. In fact, between 1870 and 1880 as many as
25 bridges failed in the United States during a single year.
The problems plaguing these bridges were eventually eliminated when iron
was replaced by steel and structural analysis as we know it today evolved. In
1856, Henry Bessemer, an Englishman, developed the first successful process for
producing large quantities of steel economically. Basically, steel is a mixture of
iron and a small, carefully controlled amount of carbon. By comparison, the
20 Introduction Chap. 1
so-called iron used before the introduction of steel contained excessive amounts
of carbon, which made it brittle, and various impurities, which made its strength
unreliable. ,
In addition to the introduction of steel as a structural material, the latter
part of the nineteenth century saw tremendous progress in the science of
structural analysis. Men such as Mohr and Muller-Breslau in Germany, Maxwell
in England, Castigliano in Italy, and Greene in America developed routine
procedures for analyzing both determinate and indeterminate structures. To-
gether with the general availability of steel, these advances in structural analysis
paved the way for structural engineering in the twentieth century. Long-span
suspension bridges and high-rise buildings, two of the most significant achieve-
ments of twentieth-century structural engineers, would not have been possible
without the availability of a reliable high-strength material such as steel and
reliable methods of structural analysis.
Calculation of Reactions
Nt S
: A bridge in Oregon.
(Courtesy of Oregon Department of Transportation.)
2.1 INTRODUCTION
The first step in the analysis of a structure usually consists of drawing a free-body
diagram. This is a simplified picture of the structure, isolated from its supports,
on which are shown all the external forces that act on the structure. In general,
these forces include the applied loads and the reactions that the supports exert on
the structure. For example, Fig. 2.1b depicts the free-body diagram of the beam
in Fig. 2.1a. Shown acting on the free-body diagram are the two known loads and
the three unknown reactions that the supports apply to the beam. Since the beam
is in equilibrium under the action of these forces, we can use the conditions of
equilibrium to solve for the unknown reactions in terms of the known loads. A
review of the conditions of equilibrium as well as the manner in which they are
used to determine unknown reactions is the main subject of this chapter.
10k 5k
(b)
Figure 2.1
In this book we will primarily be concerned with structures at rest. For such
a
structure, the sum of the forces in any direction must vanish,
and the moment
about any axis of all the forces acting on the structure must be zero.
In statics it
was demonstrated that these conditions of equilibrium are satisfied
if the sum of
the forces along any three mutually perpendicular axes are zero
and if the
moments of the forces about these three axes are also zero. If we
denote the axes
by x, Y, and z, the above requirements can be expressed analytical
ly by the
equations
X+A=0 SR=0, See =0
2M.=0, SM,=0, > M,=0 CM
22
Sec. 2.3 Types of Supports and Restraints 23
Shy, eRe
for a two-dimensional structure.
TABLE 2.1
|
Types of supports
||
fA
|
Hinged — —_>
support 7
support {
A structure for which all the unknown reactions can be determined using the
equations of equilibrium is referred to as a determinate structure. For example,
the structure in Fig. 2.2a possesses three unknown reactions, and, since it is a
plane structure, three equations of equilibrium are available to solve for these
reactions. Thus the structure in Fig. 2.2a is said to be determinate. If the roller at
C is replaced by a hinge, as indicated in Fig. 2.2b, the number of unknown
reactions is increased from three to four. However, there are still only three
independent equations of equilibrium that can be used to evaluate the reactions.
A structure of this type, which possesses more unknown reactions than equations
of equilibrium, is referred to as an indeterminate structure. It is still possible to
determine the reactions in such a structure. However, equations other than those
dealing with equilibrium must be employed in carrying out the solution. We will
consider this type of structure in the latter part of the book.
A third possibility is that the pin at A is replaced by a roller as shown in Fig.
2.2c. In this case there is nothing to prevent the structure from moving toward the
right when subjected to the applied loads. A structure such as this, in which there
are an insufficient number of reactions to prevent motion from taking place, is
called an unstable structure.
In general, a plane structure is both stable and determinate if it is supported
by three reaction components that are neither parallel nor concurrent. Two
reaction components may be able to prevent motion of a structure for a specific
loading but they are usually unable to do this for all loadings. For example, the
structure in Fig. 2.3a is stable as long as it is only subjected to a vertical load.
However, the structure is unable to resist a horizontal load and it is therefore said
to be unstable. In other words, a structure is classified as being stable only if it
can resist any and all possible loads without moving. Although at least three
reaction components are necessary to provide stability, this number is not always
Sec. 2.4 Stability and Determinancy 25
Determinate Indeterminate
structure structure
15 k
10k— >» I,
Unstable
structure
Figure 2.2
sufficient. The structure in Fig. 2.3b has three reactions. Nevertheless, because
the reactions are all vertical, the structure cannot resist a horizontal load and is
therefore unstable. Similarly, the structure in Fig. 2.3c has three reactions, yet it
is unstable. The lines of action of the three reactions all pass through a common
point, making it impossible for them to resist an applied moment about this point.
The structure in Fig. 2.4a is determinate because it has three unknown
reactions and three equations of equilibrium are available to determine the
reactions. By comparison, the structure in Fig. 2.4b is indeterminate. It has four
unknown reactions, but only three equations of equilibrium are applicable. It may
appear from these two examples that a plane structure possessing more than three
reactions is always indeterminate. However, this is not the case. Sometimes,
compound structures, formed by connecting in a nonrigid manner two or more
simple structures, can have more than three reactions and still be determinate.
For example, the structure in Fig. 2.5 has four reactions but it is nevertheless
Calculation of Reactions Chap.2
26
(c)
Figure 2.3
(a)
(b)
Figure 2.4
Figure 2.5
Sec. 2.4 Stability and Determinancy 27
Re
Rs | A | C Rq
aa A eS oo
Rs
Re R3
10k
ae ae
Hinge
at A
—— B
R, t
R,
(a)
Re 4 10k
7
Rs A Cc :
—————O SS =e
A 4
Re M, {
R3
R
10ae
k é
No hinge
at A
—— B
(b)
Figure 2.6
28 Calculation of Reactions Chap. 2
Once the free-body diagram of a structure has been drawn, the calculation of the
reactions is usually the next step. Since most of the analysis that follows the
calculation of the reactions depends on the reactions, the importance of this
initial step cannot be overemphasized.
Example 2.1
As an illustration of the procedure followed in determining reactions, let us consider
the structure in Fig. 2.7a. The structure is acted on by two known loads and
supported by a hinge at A and a roller at B. In Fig. 2.7b, the supports have been
replaced by a horizontal and a vertical reaction at A and a vertical reaction at B. In
addition, the 60 k load has been resolved into its vertical and horizontal components.
For a plane structure such as the one being considered, three independent equations
of equilibrium are available: )} F. = 0, ) F, = 0, and Y M, = 0. Since the number
of reactions, three in this case, does not exceed the number of available equations of
equilibrium, the structure is determinate and the reactions can be evaluated using
the equations of equilibrium.
Although the calculation of the reactions can be carried out in a haphazard
manner, it is possible by applying the equations of equilibrium judiciously to keep
the numerical work to a minimum. For example, the equation of horizontal
equilibrium and a moment equation about point A each contain only one unknown
reaction. Using these equations, it is therefore possible to immediately solve for R,
and R;. Thus
DF =0
Res 30i— A ie 16
el ] 16k—>
> M,=0
—R;(20) + 48(10) — 36(12) + 20(6) = 0
R;
= 8.4
Ro = 84kf
Once R; is known, the equation of vertical equilibrium can be used to calculate Re
dB =0
Rz = 48 — 8.4 = 39.6
R, = 39.6k}
Although only three independent equations can be written to determine unknown
reactions for a plane structure, we have a choice regarding which equations to use.
For instance, it is usually possible to substitute momagt equations for horizontal and
vertical equilibrium equations. Thus we could have solved for R, by writing a
moment equation about point B instead of using the equation of vertical
equilibrium.
Because of the importance of obtaining the correct values for the reactions, we
should always take the time required to check these calculations. The check is
Sec. 2.5 Calculation of Reactions: Simple Structures
20 k
6!
anilks
(a)
48 k
fen (SO k B
Rs
20 k
R, A
ir
(b)
48 k
eee 26K
8.4k
20 k
16k A
ie k
(c)
Figure 2.7
Calculation of Reactions Chap. 2
30
Thus
performed using an equation not employed in the calculation of the reactions.
we will check the values obtained for the reactions by writing a moment equation
about point B.
> My = 0
39.6(20)— 16(12)— 20(6)— 48(10) = 0
0=0
O.K.
Whereas some structures, like the one in Fig. 2.7, consist of a single rigid
member, others are made up of a number of members connected to one another
in a nonrigid manner. For example, the system depicted in Fig. 2.8a consists of
two bars, AC and BD, connected to each other by a pin at B. At first sight this
structure, having four unknown reactions, would appear to be indeterminate.
However, the pin connection at B provides a condition of connection that when
combined with the three basic equations of equilibrium for a plane structure
makes it possible to determine all four reactions.
Example 2.2
To illustrate the procedure used to calculate reactions for a compound structure, let
us determine the reactions for the structure in Fig. 2.8a. A very direct solution to the
problem would be simply to write equations of vertical, horizontal, and moment
equilibrium for the entire structure in addition to summing moments about the pin at
B of the forces either to the left or to the right of B. The disadvantage of this
approach is that it entails the solution of simultaneous equations, which is
undesirable if the calculations are being carried out without the aid of a computer.
We will therefore approach the solution in a somewhat different manner. To begin,
let us separate the structure into its basic components as indicated in Fig. 2.8b. We
can now write equations of equilibrium for either of the two free bodies in Fig. 2.8b
or for the entire structure in Fig. 2.8a. It should be noted that the free bodies of the
individual bars have acting on them, in addition to the applied loads and reactions at
C and D, the two components of the force that they apply to one another at B. In
accordance with the principle of action and reaction, the forces at B acting on bar
BD are equal and opposite to those acting on bar AC.
Starting with the free body of bar ABC, we can sum moments about B and
solve for R;. Thus a
SM; =0
R3(6)
+ 50(4) II S
R, Io
R; = 33.3kKN—>
Sec. 2.6 Calculation of Reactions: Compound Structures 31
10 kKN/m 4m
(a)
Fe
10 kKN/m
R, Fe
=~ ee || (FY
D B | Fe
Ro Fs
| |
a
| ia
~=~—'!C la
R; {
Ra
(b)
A
— 50 KN
10 kN/m
83.3 kN poe! 5 B
40 kN
C la 33.3 KN
iikN
(c)
Figure 2.8
Calculation of Reactions Chap. 2
32
direction of R; is
The negative sign in front of the answer indicates that the assumed
incorrect.
turn to the
At this point we can proceed in one of two ways. We can either
and solve for R, by summing moments about D and
free body of the entire structure
R, and R, by writing equations of horizontal and vertical equilibrium, or we can
, we
continue using the free bodies of the individual bars. Using the former approach
turn to the free body of the entire structure.
> Mp = 0
>, Bi
R, + 40 — 10(8) = 0
Ike SE 40
Ro = 40 kN7f
We will now check the values obtained for the four reactions by summing moments
about point A for the entire structure.
iM, =0
(—33.3)(10) — 10(8)(4) + 40(8) + 83.3(4) = 0
0=0
O.K.
A free-body diagram of the entire structure, including all the reactions, is shown in
Fig. 2.8c.
Example 2.3
Determine the reactions for the structure in Fig. 2.9.
The structure in Fig. 2.9a is known as a cantilever structure. The presence of
two hinges in the center span makes the structure determinate. This type of structure
appealed to nineteenth-century bridge builders, who had not yet mastered the
science of analyzing indeterminate structures.
It is best to begin the analysis of the structure by considering a free body of
the section between the hinges, as shown in Fig. 2.9b. Since hinges cannot transmit
moments, only vertical forces are applied to member @D by the adjacent segments
of the structure. Because of symmetry, each of these forces is equal to 160k.
We are now ready to consider the free body of span ABC, shown in Fig. 2.9c.
This member has acting on it, in addition to the applied distributed load and the
reactions at A and B, a 160k force at C. In accordance with the principle of action
Problems 33
2 k/ft
& of ane iy X
A
160 K
ak: 2 k/ft
Cc D A
- C
160’
160 k 160 k Rin Re
x 120’ ae 60’ a
(b) a
2 k/ft
A F
B E
10 k 510 k 510 k 10 k
(d)
Figure 2.9
and reaction, if ABC pushes up on CD with a force of 160k, then CD must push
down on ABC with a force of 160k. Taking moments about point A, we obtain
A free body of the entire structure, showing all reactions, appears in Fig. 2.9d.
34 Calculation of Reactions Chap.2
PROBLEMS
ie 14’
20'
PAP 20 kN/m
10 KN
q B
a 20 kN Y
ep
2.3. 4
30 kN Nv
16kN/m aly
A B
+2 m ++— 8m —+-— 3m —++— 3m a
2.4. 2 Skit
A B
- 10 - 14 — 8—
2.5. 30 kKN/m
10 kN/m
Problems 35
Cable
wz |)
6m
A 60 kN |
|.—6m —-2m
3.10. 10kN/m
>
2 3
$—
bS-+— ow
|}+_—3
36 Calculation of Reactions Chap. 2
2.11. 30 kN/m
4
3m
te
x 10m +
16° =re)
f “i
2.13.
2.14. 20 kKN/m
ze eae 8 kN/m
A Hinge &
7 77 B 7 C
Problems 37
2.17.
2.18.
1 k/ft
2.19.
Calculation of Reactions Chap.2
38
2.20.
2.21.
~ Hinges re
10 kKN/m
2.24.
L—29 1.5 k/ft
Problems 39
2.25. 10 kKN/m
PAs ams 2 member > og
whats + ae
_
~
‘em ‘= . ==
1 ee @
7 - “| -
= Te - AS) Tes,
; a » ~ oo
) Vqge—wleliede SL 7
7 ae " ae he -
LB. i on
iy | a. bee
Plane Trusses
1. The members are connected to each other at their ends by frictionless pins;
that is, only a force and no moment can be transferred from one member to
another.
2. External loads are applied to the truss only at its joints.
3. The centroidal axes of the members meeting at a joint all intersect at a
common point, that is, the point where the members are assumed to be
pinned to one another.
Fa
(b)
Figure 3.1
42
Sec. 3.2 Simplifying Assumptions 43
Figure 3.2
44 Plane Trusses Chap. 3
Bridge deck
Floor beams
Figure 3.3
The object of analyzing a truss is to determine the forces produced in its members
by a set of external loads. Since these forces are internal when one considers the
truss as a whole, it is necessary in the course of the analysis to take free bodies of
portions of the truss. Such a free body, of a portion of a truss, will be acted on by
the bar forces corresponding to the members that have been cut in the process of
creating the free body. If we then write equations of equilibrium for the free
body, we can use these equations to evaluate the bar forces.
There are two well-known methods of passing Sections through a truss in
order to produce useful free bodies of portions of the truss. In one, known as the
method of joints, we consider free bodies of individual joints; in the other, known
as the method of sections, we pass a section that bisects the truss and consider the
Sec. 3.4 Method of Joints 45
Figure 3.4
free body on either side of the section. It is usually advantageous to employ the
method of sections when we wish to determine the forces in only a few isolated
members and to use the method of joints when it is desired to calculate all the bar
forces in a truss.
Although we are interested in determining the bar forces themselves, it ‘is
necessary to deal with their horizontal and vertical components during the
analysis. To see how the components of a bar force are obtained, let us consider
the force in member AB of the truss depicted in Fig. 3.4. The horizontal
component of the force is given by
a
Fie = F 4, C0OS a= Fol“)
c
and the vertical component by
; b
Fapy = Fag Sina = Faa(-)
In the method of joints, a section is passed completely around a joint, cutting all
the bars meeting at the joint and isolating the joint from the rest of the truss, as
46 Plane Trusses Chap. p 3
PNG
= |
he Foc
{ B
Compression ;
member Tension
member
Fac
vee Tension t
A
Zi 6 <
member |Cc
eer? | —P- Fop
Fac A C Fac
R
P
(b)
Figure 3.5
shown in Fig. 3.5a. The free body of the joint produced in this manner will have
acting on it a set of concurrent forces consisting of bar forces and externally
applied loads or reactions. For example, the free body of joint A, shown in Fig.
3.5a, has acting on it the left-hand reaction of the truss and bar forces Fy, and
F,c. If we consider free bodies of the bars as well as the joints (Fig. 3.5b), it
becomes evident that members in compression such as bar AB push on the joints
at their extremities and that tension members such as bar AC pull on the joints at
their extremities. @
Since joint A is in equilibrium, the equations )) F. = 0 and »} F, = 0 can be
used to determine the two unknown forces acting on the joint. Not all joints in a
truss involve only two member forces. However, for most trusses we can proceed
Sec. 3.4 Method of Joints 47
from one joint to another in such a manner that there are never more than two
unknown bar forces acting on the joint being investigated. To see how the
method of joints is carried out, let us consider Example 3.1.
Example 3.1
It is required to calculate the bar forces for the truss in Fig. 3.6a using the method of
joints. Z
The first step, in this as in any other analysis, is to determine the reactions.
Since an error in the reactions renders the remainder of the analysis useless, we
cannot overemphasize the importance of this step. To minimize the chances of error,
the calculation of the reactions should always be checked. We will obtain the
reactions by writing two moment equations and then check the results using an
B C
Ra
|, Eas rel
12k 24k if
3 @ 20’ = 60’ _
(a)
0.6 Fag
Fag
16 16
(b)
Ferg
12
(c)
‘ Figure 3.6
48 Plane Trusses Chap.3
B —> Fac
—> Fac
oi [pe 213 jon0.8 Fee
LP aS Fee Kat
12 16 iZ 0.6 Fee
ee,
Eacaeed
Snowe
G
(e)
F cp = 33.3 20
SSI 26.7 |
(f)
(g)
B 26.7(C) S
26.7(C) 33.3(€)
aF = 0
16 + 0.6F4, = 0
Phos == —26.7
DF =0
Fir + 0.8F,, = 0
Fur = 21.3k(T)
50 Plane Trusses Chap. 3
This time the answer is positive, indicating that member AF pulls on the joint as
assumed and that the member is therefore in tension.
At this stage in the analysis, joint B has three unknown forces acting on it,
whereas joint F has only two unknown forces acting on it. We will therefore consider
F next. As shown in Fig. 3.6c, this joint has acting on it a 12k load, the two
unknown bar forces F;,; and F;,, and the previously determined force Fy, = 21.3k.
The unknown bar forces are readily determined using equations of vertical and
horizontal equilibrium. Thus,
Fre — 12 = 0
En = 12
Fre = 12k(T)
ME= 0
lies = IS} Sw
Fe 23
Fre = 21.3k (T)
Since of the four bar forces acting on joint B only two are still unknown, we
can now proceed with the analysis of joint B. The free body of this joint is shown in
Fig. 3.6d. It should be noted that, whereas the unknown forces are assumed to be
tension, the known forces in bars AB and FB are shown acting in their correct
directions. As before, we replace all forces by their horizontal and vertical
components and write equations of horizontal and vertical equilibrium.
DF=0
For = 6.67k(T)
dF,=0
21.3 + 0.8(6.67) + Fee = 0
> £.=,0
0.8Fep + 26.7 = 0
Fep = —33.3
Fep = 33.3k(C)
De 10
Fer + 0.6(—33.3) = 0 ’
For = 20.0
Fer = 20.0k(T)
Sec. 3.5 Method of Sections 51
Finally, the value of F,, is obtained using the free body of joint D, shown in
Fig. 3.6f.
DF =0
|Bsr = 26.7
Fog = 26.7k(T)
Having determined all the bar forces, we should now check our results. In
obtaining the bar forces, we did not make use of the equation of vertical equilibrium
at joint D and the equations of vertical and horizontal equilibrium at joint E. These
are not independent equations. However, they can be used to check our calcula-
tions. The reason we did not make use of these three equations is that the reactions
that could have been determined using joint equations were instead determined
using a free body of the entire truss.
For joint D, shown in Fig. 3.6f,
pagal
20 — 20 = 0
and for joint E, shown in Fig. 3.6g,
Bie
4+ 20 —-24=0
DF =0
DOT = 8s = SES) = Osi O.K.
When the analysis of a truss has been completed, it is useful to record all the
bar forces on a diagram of the truss, as shown in Fig. 3.6h.
Example 3.2
Using the method of sections, it is required to determine the forces in bars BC, HG,
and BG for the truss in Fig. 3.7a.
52 Plane Trusses Chap. 3
1 kN 40 kN 40 kN 40 kN
}+__—4@8m=32m
(a)
B-
——> Fac
4
3 Nafes
—
A ae ee
{ | Eye
60 40
+ 8m- a 8m -
(b)
Figure 3.7
The first step is to calculate the reactions at A and E. Next a section is passed
through the truss cutting members BC, BG, and HG, whose forces we wish to
determine. To calculate these forces, we will consider the free body to the left of the
section, shown in Fig. 3.7b. As before, the unknown bar forces are initially assumed
to be tension forces. If we write a moment equation about G, a point through which
the lines of action of two of the unknown forces go, we can immediately solve for
Fyc. Thus
Me 0
6Fc + 60(16) — 40(8) ll 0
Fac —106.7
Fac = 106.7 kN (C)
In a similar manner, we can solve directly for F,, by summing moments about
point B.
S) M, = 0
60(8) — 6Fyg = 0
Fig = 80
Fug = 80 kN (T)
Sec. 3.5 Method of Sections 53
De
60 — 40 — 0.673, = 0
Feg = 33.3
Fg = 33.3 KN (T)
To check our results, we write an equation of horizontal equilibrium.
ee
Fae + Fag + 0.8Frc =0
— 106.7 + 80 + 0.8(33.3) = 0
—0.1 =0
Example 3.3
Using the method of sections, determine the forces in bars BC, BG, and HG for the
truss in Fig. 3.8a.
« 4 @ 20' = 80' |
(a)
(b)
Figure 3.8
54 Plane Trusses Chap. 3
As in the previous example, the reactions are calculated first. A section is then
passed through the truss as indicated, resulting in the free body shown in Fig. 3.8b.
To determine the force in member HG, we sum moments about point B. Thus
> Mm, =0
30(20) — Fc(10) = 0
Fug = 60
Fic = 60k (T)
In a similar manner, we can calculate Fz: by taking moments about point G. This
requires taking moments of force Fz- about G. Although it would be possible for us
to obtain this moment by determining the perpendicular distance from G to Fc, it is
easier to resolve Fg- into vertical and horizontal components at either B or C and
then, knowing the distances from B and C to G, calculate the moments of these
components about G. In fact, only the horizontal component of Fz- has a nonzero
moment about G, if we resolve Fz: into components at point C. Thus
SMa =0
30(40) — 20(20) + 0.89F,<(20) = 0
eee — —44.9
34-= 0
20(20) + 0.45F5,(40) = 0
love = —22.2
Fog = 22.2k(C)
To check our results, we write an equation of horizontal equilibrium.
Y= 0
Fig + 0.89Fg¢ + 0.89Fg- = 0
60 + 0.89(—22.2) + 0.89(—44.9) = 0
025"
Even when we wish to determine all the bar forces in a truss, it may be desirable
to use both the method of joints and the method of Sections. In the following
example, we will employ the method of joints to determine some of the
bar forces
and the method of sections to calculate others. .
Sec. 3.6 Method of Joints and Sections Combined 55
(b) f- 6 +}x—6 ~
a el Qa
67.1 | Fep B
For 59.7
(d) (e)
Figure 3.9
56 Plane Trusses Chap. p 3
(g)
20.0
20 4.7 t 33.0
67.1 oe | ~\ Wik
oN kK
Lia aN82.0 (he
F
For ,
(h) (j)
C
67.1(C) 67.1(C)
D
82.0(C)
F
63.4(T) 41.0(T)
(k )
Example 3.4
Determine all the bar forces for the truss in Fig. 3.9a. %
Let us begin by calculating F,, and F,, using the free body of joint A in Fig.
3.9b. Since both forces have horizontal as well as vertical components, it is not
Sec. 3.6 Method of Joints and Sections Combined 57
DA =0
0.447F,, + 0.894F,- — 30 = 0
> Fe =0
0.894F,, + 0.447F,- + 25 = 0
SM; = 0
0.894F,-(9) + 25(12) + 30(6) + 30(6) — 20(6) = 0
Fre = —67.1 kN
DR =0
0.894(59.7) — 20 — 0.447(67.1) — 0.707F5- = 0
Fgr = 4.7KN
To determine F;;, we make use of the free body in Fig. 3.9f.
> My =0
0.894F,--(6) + 0.447Fe-(6) — 55.0(6) = 0
Fea 0
DE =0
0.894(41.0) + 0.447Fep = 0
Fen cS —82.0 kN
a =0
20 + 0.447(67.1) — 0.894(82.0) + 0.707Fp- = 0
For = 33.0
The calculations are checked using an equation of vertical equilibrium at joint F
Plane Trusses Chap. 3
58
ae =0
4.7(0.707) + 33.0(0.707) + 20.0 — 63.4(0.447) — 41.0(0.447) = 0
0=0
A summary of the results is given in Fig. 3.9k.
In Section 2.4 a determinate structure was defined as one for which the number of
equations of equilibrium are equal to the number of unknown reactions. By
comparison, it was stated that an indeterminate structure is one that possesses
more unknown reactions than equations of equilibrium, and an unstable structure
is one that has an insufficient number of reactions to prevent motion. In
extending these criteria to trusses, we will see that stability and determinacy of
trusses depend not only on the number of reactions but also on the number and
arrangement of individual members. A truss is externally stable and determinate
if it possesses the proper number of reactions and internally stable and
determinate if it has the correct number of bars arranged in a proper manner.
In applying the method of joints, we have seen that two independent
equations of equilibrium can be written at-each joint of a truss and that these two
equations can be used to solve for two unknown forces at that joint. In other
words, if a truss possesses j joints, we can write 27 independent equations and
solve for 27 unknown forces. Although we did not solve for the reactions using
joint equations in the preceding illustrative problems, it is certainly possible and
sometimes even preferable to do so. If this procedure is employed, it follows that
a truss is determinate if the number of member forces and reactions is equal to
twice the number of joints. That is,
j=oamtr Gat)
P— P——
Y “N ~ ~
R, Ro R, R,
j=4,r=3,m=5 j=4,r=3,m=4
2) =r 2) >mt+r
(a) (b)
j=4,r=3,m=6
2] <r
(c)
Figure 3.10
and determinate, while those in Figs. 3.10b and 3.10c are indeterminate and
unstable, respectively.
The truss in Fig. 3.10a is externally stable and determinate because it has
three nonparallel, nonconcurrent reactions that suffice to prevent motion of the
structure as a whole and that can be evaluated using the three equations of
equilibrium available for a plane structure. The structure is also internally stable
and determinate. As long as the bars are arranged to form one or more triangles,
a rigid, stable system of members results. Furthermore, the bar forces can be
evaluated using equations of equilibrium. For example, employing the method of
joints, we could begin the analysis at either joint B or D, where there are only
two unknowns, and then proceed to other joints to determine the remaining bar
forces.
Like the truss in Fig. 3.10a, those in Figs. 3.10b and 3.10c are externally
stable and determinate. The truss in Fig. 3.10b is, however, internally unstable.
60 Plane Trusses Chap. 3
Because the bars are hinged to one another, the rectangle that they form can
collapse. By comparison, the truss in Fig. 3.10c is internally indeterminate. That
is, the bar forces cannot be evaluated using only equations of equilibrium. The
presence of three unknown bar forces at each joint makes it impossible to use the
method of joints. The method of sections is likewise inapplicable. Whereas the
truss in Fig. 3.10b needs an interior diagonal to make it stable, the truss in Fig.
3.10c has one too many diagonals for determinacy.
Unlike the trusses in Figs. 3.10b and 3.10c, the one in Fig. 3.10d does satisfy
Eq. (3.1). Nevertheless, it is neither stable nor determinate. It is externally
unstable because the two vertical reactions are insufficient to prevent motion of
the structure as a whole, and it is internally indeterminate because it has one
more diagonal than necessary.
The foregoing examples illustrate the fact that Eq. (3.1) is a necessary but
not sufficient condition for stability and determinacy. The trusses in Figs. 3.10b
and 3.10c, which do not satisfy Eq. (3.1), are indeed unstable and indeterminate,
respectively. However, the truss in Fig. 3.10d, although it satisfies Eq. (3.1), is
neither stable nor determinate. The conclusion to be drawn from these examples
is that a truss will be stable and determinate if, in addition to satisfying Eq. (3.1),
it is stable and determinate both internally and externally. This means that the
bars must be arranged in the form of triangles or some other stable configuration,
and their number must be such that the bar forces can be determined using either
the method of joints or the method of sections. In addition, there must be a
sufficient number of reactions to prevent the structure from moving as a whole
but not more than can be evaluated using the available equations of equilibrium.
For most trusses, stability and determinacy can be ascertained by using the
foregoing principles. However, there are some trusses for which instability or
indeterminacy becomes evident only after we attempt to analyze them.
not only solve the equations but also set them up. The latter procedure obviously
necessitates the writing of a more involved program than does the former
procedure. However, once the program has been written, it obviates the need for
a considerable amount of long-hand calculation, which is after all the primary
purpose of using a computer.
To illustrate the first procedure, let us consider the truss in Fig. 3.11a. If we
assume the reactions to be in the directions indicated and assume the bars to be
stressed in tension, then the joints will have acting on them the forces shown in
Fig. 3.11b. Applying equations of vertical and horizontal equilibrium to each of
tn 40 kN
Ved 20
Ray Rey
(b)
Figure 3.11
Plane Trusses Chap. 3
62
Using the coefficients of the unknown bar forces and reactions and the values of
the known applied loads as input, we may now employ any standard computer
program to solve this set of simultaneous equations for the bar forces and the
reactions. If this is done, we obtain the following results:
Fig = —20.83
Fic = +36.67
Fac = —45.83
Rie =a 12.50
Rax = —20.0
Rey = +27.50
Negative signs for the reactions indicate that they act in a direction opposite to
the one assumed in the figure. For the bar forces, positive signs denote tension
and negative signs denote compression.
Let us now consider the second procedure, in which we let the computer set
up the simultaneous equations as well as solve them. In other words, if we were
analyzing the truss in Fig. 3.11, the computer would form the set of Eqs. (3.2) in
addition to solving them. Before we consider the details of this procedure, let us
note exactly what Eqs. (3.2) consist of.
For the computer to be able to form the joint equilibrium equations, it must
be given two sets of information. First the computer must be told which forces act
in the x- and y- directions at each joint. In other words, we must tell the
computer which nonzero terms appear in each equation. Second, the computer
Sec. 3.8 Computer Solution 63
must be able to calculate the coefficients of the unknowns on the left-hand sides
of the equations.
There are two ways of dealing with the question of which forces act at any
given joint. We can either proceed from joint to joint and simply note which
members frame into the joint and which reactions and applied loads act on it, or
we can consider the members, reactions, and applied loads one at a time and note
which joints lie at their extremities. It turns out that the latter procedure is better
suited to our task, and we will consequently use it. Thus we will tell the computer
which forces act on each joint by considering each member, reaction, and applied
load, one at a time, and note which joints lie at the extremities of each.
Having determined which forces appear in each joint equation, the
computer must then calculate the coefficients of each unknown on the left-hand
sides of the equations. In every instance the component of a force acting in either
the x- or the y-direction at a joint is equal to the force multiplied by the cosine of
the angle that the force makes with the x- or y-direction. In other words, the
coefficients of the unknown forces in the equations are equal to the cosines of the
angles between these forces and the x- and y-directions. These quantities, known
as direction cosines, can be determined for the members by the computer if it is
given the coordinates of the joints at the ends of the members.
We are now ready to consider the program that will set up and solve a set of
joint equilibrium equations of the form
where the F’s represent the unknown bar forces and reactions, the A’s represent
the coefficients of these unknowns, and the P’s are the known loads.
The program that follows determines the bar forces and reactions for a
plane truss. A sample problem, whose purpose is to explain how the input data
are to be entered into the program, follows the program.
00010¢
oe adi aiAiAiAiAIAIAIAIAIAIAIAiaiAiAiAIAiMIMIAIAIAIAIAIAIAIAIAIAiRiaIAIAiAIAIAiRiRididiaiaii
000400 # FORTRAN - 77 PROGRAM FOR ANALYZING TRUSSES #
00050¢ = Ht
Pe ae a AiAiAiAidiAiaiAiAiaiAiAiMiaiMiAiAiAiAAIAIAIAIAIAIAiAininiAiMiaARiAIAiAiaiaiaiaiaiaiAil
00080 PROGRAM TRUSS (INPUT, OUTPUT)
00090 DIMENSION A(50, 50), P(50), F(50), C(50)
00100 PRINT *, ENTER THE NUMBER OF SIMULTANEOUS EQUATIONS;’
00110 READ », N
00120 PRINT +, ENTER THE NUMBER OF MEMBERS;’
00130 READ *, NM
00140 PRINT +, ENTER THE VALUES OF APPLIED LOADS;’
00150 pO5I=1,N
00160 ae PRINTS P(e. te)
Plane Trusses Chap.3
64
00700 102k
00710 15 CONTINUE
00720 IF (PIVOT .NE. 0.) GO TO 20
00730 PRINT 1000
00740 G0 TO 100
00750 20 IF (IR .EQ. I) GO TO 30
00760 DO 25 J =I, N 4
00770 AA = A(I, J)
00780 ACL, 3)’ UAE i)
00790 25 Vany Apcey
00800 S31)
Sec. 3.8 Computer Solution 65
Example 3.5
Using the preceding program, analyze the truss in Fig. 3.12a.
We begin with the following steps.
Be 20) KN
aiee
Hz
(b)
Figure 3.12
we now have two vectors associated with each joint. For example, the x-vector
at joint B has been assigned the number 3 and the y-vector the number 4. The
numbering of the joint vectors is independent of the member numbering.
6. Choose a convenient origin and use it to determine the coordinates of the
joints. Assign the x-coordinate of each joint to the x-vector and the
y-coordinate to the y-vector. For example, if we assume the origin of the truss
to be A, the coordinates at B would be C(3) = 10 and C(4) = 15.
We are now ready to consider the entering of data into the program. Each
piece of data that is required is entered by the user in response to a question put to
him or her by the computer.
Sec. 3.8 Computer Solution 67
- The first question asks for the number of simultaneous equations (see page
62). This will always be equal to twice the number of joints. Thus the user
would respond by entering the number 6 for the sample problem being
considered.
. The second question asks for the number of members. To this question, the
user would respond by entering the number 3.
. The third question asks for the values of the applied loads, acting in the x- and
y-directions at each joint. Since there are no applied loads acting in the 1 and
2 directions at joint A, the response, when the computer asks for the values of
P(1) and P(2), would be 0 and 0. By comparison at joint B, there are applied
loads of 20k and 40k acting in the 3 and 4 directions. Hence the user would
enter the numbers 20 and —40 when the computer asks for P(3) and P(4). The
applied loads are considered to be positive when they act in the positive x- and
y-directions.
. The fourth question asks for the x- and y-coordinate at each joint. Remem-
bering that we have chosen A as our origin and that we are using the two
numbers of the vectors at each joint to designate the x- and y-coordinates of
that joint, the responses for C(1) and C(2) would be 0 and 0, and the
responses for C(3) and C(4) would be 4 and 3.
. The fifth question asks for the numbers of the four joint vectors at the
extremities of each member. We will designate these by J, J, K, and L, where
I and J are the numbers corresponding to the x and y joint vectors at the tail of
the member vector, and K and L are the numbers corresponding to the x and
y joint vectors at the head of the member vector. Thus the user would enter
the numbers 1, 2, 3, and 4 for member 1, the numbers 3, 4, 5, and 6 for
member 2, and the numbers 1, 2, 5, and 6 for member 3.
. The sixth and final question asks for the number of the joint vectors
corresponding to the reactions. There are three reactions located at A and C,
as shown in Fig. 3.12b. The user will thus respond with the number 1 when
asked for the joint vector number for F,, which is the horizontal reaction at A,
and with the numbers 2 and 6 when asked for the joint vector numbers for F;
and F.
The following is the input and output for the sample problem.
INPUT
Biome
70
RG 6):
can)
ENTER THE X AND Y COORDINATES AT EACH JOINT
Ye al jre
PROBLEMS
3.1 to 3.10. Use the method of joints to determine all the bar forces.
80 kN
Problems 69
3.2.
3.4.
3:55 B ic D
20’
A SAL
H G F
7, Y,
|20 k 16 k
«4 @20'- ae
3.6.
Plane Trusses Chap.3
70
E a 40)KN
—
fore) =}
80 kN
oo
Problems 71
3.10.
3.11 to 3.21. Use the method of sections to determine the bar forces for the indicated
members.
16 k 16k 16k
“A 4
80 kN
72 Plane Trusses Chap.3
B c D
a15!
A E
“ll {f
LAs
20 k 20 k
6m
6m
60 kN 100 kN
8m 8m
10k 10k 5 k
un slag ea
el 10°
D B 4
i
F g
Problems 73
=|
b-&
12°
12°
=
G
4 @24’= 96"
20k
rz
A B a7
12’
C
|50%
|-— ‘29°—+—
29° —+|
74 Plane Trusses Chap.3
}+—15° —ol+—
15° +
30kN 30kN
20 cca
7
8m
40 kN 4
8m
40 kN
8m
eee €S
10m ~
le
4@10m=40m =
Problems 75
3.25.96
K 15'- an
Plane Trusses Chap. 3
76
3.26. 12 kN
Sato 20 k 20k
Space Trusses
1. A ball and socket joint that can resist translation in three perpendicular
directions and can therefore apply forces to the structure in the xX-, y-, and
z-directions.
2. A support that is able to resist movement in only two directions and can
therefore apply forces to the structure only in these two directions.
3. A support that can prevent motion only, in one directio
n and can
consequently apply a reaction only in that direction.
78
Sec. 4.3 Calculation of Member Forces 79
We will assume that the members of a space truss are connected to one another
by ball and socket joints and that external loads are applied to the truss only at its
joints. The bars of the truss will consequently be subjected to axial forces only.
We will determine the forces in individual members of the truss by using the
method of joints. A joint in a three-dimensional truss can have acting on it a set
of concurrent forces having components in the x-, y-, and z-directions. Such a
joint will be in equilibrium if the sum of the forces in each of these three
directions vanishes, that is, if
Figure 4.1
80 Space Trusses Chap. 4
where 9, is the angle between the member AB and its projection AC along the
x-axis. Similarly,
AF AE
Fapy = FaB yp? Fan.z = Faeap (4.4)
The lengths of the projections of the member along the three coordinate axes are
determined from the coordinates of the end points of the member. Having
calculated these projections, we can then obtain the length of the member itself
from
Example 4.1
As an illustration of the principles involved in analyzing a space truss, let us
determine the reactions and the bar forces for the truss in Fig. 4.2.
As shown in Fig. 4.2a, the truss consists of six bars arranged in the form of a
three-sided pyramid. The elevation and plan views of the truss are depicted in Figs.
4.2b and 4.2c. Two loads, acting in the z- and x-directions, respectively, are applied
at point D, the apex of the structure. The base of the truss is restrained by three
vertical reaction components, one each at points A, B, and C, two reaction
components in the y-direction at points A and B, and a single reaction component in
the x-direction at point C. The reactions are assumed to act in the directions shown
in the figure.
We begin the analysis by calculating the reactions. There are six of them, and
we have six independent equations to evaluate them. The vertical reaction at C,
Rcz, can be determined by taking moments about an x-axis passing through points
A and B. Thus
S\ M, = 0
(100)(4) — 8Rez = 0
Rex=S0
Rez = SOKN 4
> M, = 0
100(4) — 200(12) — 50(4) — 8RR ll °
Rpz = —275
Rez = 275 kN
Sec. 4.3 Calculation of Member Forces 81
100 kN
200 kN
Cam Rex
ma 4m x
200 kN
Plan view
A pi oe ck B ete y
Rav bre,
4 mote 4mm}
(c)
t7 0.301F ao
@
ie |Peace
Ray
= 100 a
(d)
Figure 4.2
Space Trusses Chap. 4
82
Rex — 200
Rex = 200kN —>
To determine R,y, we sum moments about a z-axis through point B.
SM, =0
—200(4) + 200(8) — 8Ray = 9
Rev = 100
Ray = 100kNi}
Finally, to calculate Ray, we sum forces in the y-direction.
ce
Ray + 100 = 0
IRtaee 3 —100
Ray = 100kN Tf
Having determined the reactions, we are now ready to calculate the bar forces.
As indicated by Eqs. (4.3) and (4.4), these calculations require a knowledge of the
lengths of the members, their x-, y-, and z-projections, and the ratios of these
quantities. It is convenient to list these terms in tabular form as shown in Table 4.1.
We begin the calculations by considering the free body of joint A shown in Fig.
4.2d. It should be noted that all bar forces have initially been assumed to be tension
TABLE 4.1
Ratio of projection
Projection to length
Rez Roz
Elevation view
(b)
Plan view
(c)
Figure 4.3
Space Trusses Chap. 4
that has a
forces. Since the force in bar AD is the only force acting on the joint
ly allows us to evaluate
z-component, summing forces in the z-direction immediate
Fp. Thus
pang FileMh
0.904F,p + 325 = 0
Fup = —359.5 kN
Next we write a force equation in the y-direction and determine the magnitude of
Fc:
LEO
100 — 0.301(—359.5) — 0.895 Fic = 0
Fic = 232.6 kN
Finally, summing forces in the x-direction gives F4,.
SE=0
F,, + 0.301(—359.5) + 0.447(232.6) = 0
Fug = 4.2KN
In a similar manner, using other joints, the remaining bar forces can be evaluated.
The magnitudes of all bar forces are listed in the last column of Table 4.1.
Having familiarized ourselves with some of the basic principles involved in
the analysis of a three-dimensional truss, let us now consider how we can simplify
the analysis of such a structure with the aid of a computer. The most direct
approach is to write three equations of equilibrium at every joint and then, using
a computer, solve this set of simultaneous equations for both the reactions and
the member forces.
Example 4.2
Determine the member forces and the reactions for the truss shown in Fig. 4.3.
The x-, y-, and z-projections of all bar lengths are listed in Table 4.2. Using
TABLE 4.2
Ratio of projection
Projection to length
AB 0 20 0 20 0 1.0 0
BD 50 0 0 50 1.0 0 0
DC 0 20 0 20 0 1.0 0
CA 50 0 0 50 1.0 0 0
AE 20 10 20 30 0.667 0.333 0.667
BE 20 10 20 30 0.667 0.333 0.667
DE 30 10 20 37.42 0.802 0.267 0.534
CE 30 10 20 37.42 0.802 % 0.267 0.534
F1VL
fv
85
86 Space Trusses Chap. 4
TABLE 4.4
Raz =) AB —4
Rzz 32 BD 32
Rez 8 DC 4
Rpz 8 CA 12
Ray 20 AE 12
Rax —20 BE —48
Rpx 20 DE —15
CE =ail)
this information, the lengths and the ratios of the projections to the lengths are
calculated and noted in the table.
Assuming the reactions to be in the directions indicated in Fig. 4.3 and the
bars to be in tension, the simultaneous equations given in Table 4.3 are obtained.
Each row in the table represents a force equation at one of the joints. For example,
the first row contains the x-components of the forces acting on joint A. Similarly, the
second and third row contain the y- and z-components of the forces acting on joint
A. Each column in the table represents an unknown bar force, an unknown reaction,
or a known load.
The results obtained by solving the equations in Table 4.3 are given in
Table 4.4.
PROBLEMS
4.1 to 4.5. Determine the reactions and all the bar forces.
4.1. 20k
Problems
87
4.2. 120 kN
D teal kN
Rez Rez
4.3. 30k
88 Space Trusses Chap. 4
Ray
beae Ra am 10’ |
Beams are relatively long and slender members that are usually loaded normal to
their longitudinal axis. They transfer loads by developing a combination of
bending and shear stresses. Some examples of beams are (1) the wooden joists
that support the floors in single-family dwellings, (2) the horizontal steel or
concrete members that are used to support floors in large multistory buildings,
and (3) the girders that support the roadway in small highway bridges. Beams are
used to span distances ranging from several feet to approximately 200 feet. For
very large spans, they are uneconomical and other structural forms are employed.
Beams are generally classified according to the manner in which they are
supported. Some of the more common types are illustrated in Fig. 5.1. A simply
supported beam is one that is supported by a hinge at one end and a roller at the
other end (Fig. 5.1a). A cantilever beam is completely fixed at one end and free at
the other end (Fig. 5.1b). If both ends of a beam are fixed, it is referred to as a
fixed-fixed beam (Fig. 5.1c). Sometimes a beam, in addition to being supported at
its ends, is also supported at one or more intermediary points (Fig. 5.1d). Such a
beam, if it is continuous over the intermediary supports, is called a continuous
beam. Beams, like any other structure, can be determinate or indeterminate
depending on whether or not there are enough equations of equilibrium to
determine all the unknown reactions. For example, the beams in Figs. 5.1a and
5.1b are determinate. They each possess three unknown reactions, and three
equations of equilibrium are available for determining these reactions. By
comparison, the beams in Figs. 5.1c and 5.1d are indeterminate because they each
have more than three unknown reactions but there are still only three equations
available for determining the reactions.
gtZ UY, = LG
Simply supported Cantilever
(a)
(b)
4g—+_1
—_y YZ
Ss aor 2
Fixed-fixed
Continuous
(c)
(d)
Figure 5.1
&
90
Sec. 5.2 Internal Forces 91
The analysis of a beam involves calculating the stresses produced by the applied
loads. The student may recall from the study of mechanics of materials that
stresses are proportional to internal forces. To determine the stresses in a
member, we must therefore know the magnitude of the internal forces. These
forces are in turn obtained by passing sections through the member and
considering the free body on either side of the section. For example, let us
consider the beam in Fig. 5.2. If we are interested in the internal forces at a
distance x from the left-hand support, we pass a section a—a through that point
and consider the free bodies thus formed. Since either the free body to the left or
to the right can be used, let us consider the left-hand free body. The forces that
the remainder of the beam exerts on this free body, to keep it in equilibrium,
consist of a vertical force V and a couple M. The force V, which puts the free
body in vertical equilibrium, is called the shear, and the couple M, which is
necessary for moment equilibrium, is called the bending moment.
Unlike the axial force in the bar of a truss, which remains constant, the
internal forces V and M in a beam vary along the span of the member. The shear
and bending-moment diagrams, which are our main concern in this chapter, are
devices for picturing how V and M vary. They are graphs of the magnitude of the
shear and bending moment plotted along the span of the member. Once these
diagrams have been plotted for a given beam, we can tell at a glance what the
values of V and M are at any point along the beam and where the maximum
values of V and M occur.
It is essential for the student to have a clear understanding of the difference
between the internal couple that we refer to as the bending moment and the
external moment about a point of all the forces acting on a member. The internal
Figure 5.2
92 Shear and Moment Diagrams for Beams and Frames Chap. 5
couple at any section is the moment required to keep the free body on either side
of the section in equilibrium. It is also a measure of how much the beam is being
bent. By comparison, the external moment about a point of all the forces acting
both to the right and to the left of the point is an indication of whether or not the
member will rotate. If the member is in equilibrium, this moment is equal to zero
for all points.
To be able to plot the variation of the shear and the bending moment along a
member, it is necessary to adopt a sign convention for these quantities. The sign
convention that we will use is illustrated in Fig. 5.3. As indicated, a shear force is
considered to be positive if it produces a clockwise moment about a point in the
free body on which it acts. Conversely, a negative shear force produces a
counterclockwise moment about a point in the free body on which it acts. The
bending moment is positive if it causes compression in the upper fibers of the
beam and tension in the lower fibers. By comparison, a negative bending moment
causes tension in the upper fibers and compression in the lower fibers. In other
words, positive bending causes the upper surface of the beam to take on a
concave shape, whereas negative bending causes a convex curve at the top of the
beam.
For a shear and moment sign convention to be useful, it must give the same
results regardless of whether we consider the free body to the left or right of the
section. That the sign convention we have adopted satisfies this criterion is
evident from Fig. 5.2. For this member, both the shear and the bending moment
are positive at section a—a, and this result can be obtained from either the
left-hand or the right-hand free body.
Vv Vv
ca Positi
sitive shear
ey a) (co iti
Positive moment
Vv Vv
[eae 0 |peered
Negative shear pee) (a,
q Negative moment
Figure 5.3
Sec. 5.4 Shear and Bending-Moment Diagrams
93
10k Bik
foe)
(e)
Figure 5.4
94 Shear and Moment Diagrams for Beams and Frames Chap. 5
(f)
a (g)
pt anf
Shear diagram
(h)
28
Moment diagram
(i)
along the beam can be obtained by passing a section through that point and by
considering either of the two free bodies thus formed. If this process is repeated
at regular intervals along the span of the member, curves of shear and moment
versus distance can be plotted. For example, let us consider the beam in Fig.
5.4a. If sections are passed through the beam at intervals of 2 ft, the free bodies
shown in Figs. 5.4b through 5.4g are obtained. In eachtase the forces at the cut
are obtained by putting the free body on which they act into vertical and moment
Sec. 5.5 Relationships Between Load, Shear, and Bending-Moments 95
equilibrium. Since the same internal forces act on both the free body to the right
and to the left of a section, only one of the two need be considered. In each
instance we have used the free body that involves less numerical work in the
calculation of V and M. It should be noted that, at 4 ft and 6 ft from the left end
of the member, two separate sections are taken, one to the left of the
concentrated load and one to the right of it. The reason for this is that the shear
changes abruptly at a concentrated load and, while it is possible to calculate the
shear just before or just after such a change, it is not possible to obtain its value
at the point where it is changing. :
Using the shears and moments obtained from the free bodies in Figs. 5.4b
through 5.4g, we have plotted the diagrams in Figs. 5.4h and 5.41. Although shear
and moment diagrams can be obtained by cutting sections at regular intervals, as
was done above, this procedure is as a rule too time consuming. Instead, a more
efficient way of constructing shear and moment diagrams is to utilize other
techniques in addition to the method of sections. In the following sections these
techniques will be introduced and their use illustrated.
There exist simple mathematical relationships between the load, the shear, and
the bending moment in a beam, which can be extremely useful in the construction
of shear and bending-moment diagrams. For instance, let us consider the beam in
Fig. 5.5a and concentrate our attention on the differential element of the beam
shown in Fig. 5.5b. As indicated, the element has a shear V and a moment M
acting on its left face and a shear V + dV and a moment M + dM acting on its
right face. The quantities dV and dM represent the changes in the shear and
moment that occur over the distance dx. All forces are shown acting in positive
directions.
Writing an equation of equilibrium for forces in the vertical direction gives
V.= (V+ dV) —wdx=0
from which
dV = —wdx (5.1)
This expression states that the change in shear between two points along the
member is equal to the load between the points. The minus sign indicates that a
downward load, which we have assumed to be positive, results in an algebraic
decrease in the shear.
A second relationship is obtained by summing moments about an axis
through the left-hand face of the element. Thus
2
«( Oy]
ee (d)
Figure 5.5
dM
fa (5.2)
Or
dM = Vax (5.3)
The first of these two relations states that the slope of the moment diagram
at any
point is equal to the shear at that point, and the second relation
States that the
Sec. 5.6 Construction of Shear and Moment Diagrams 97
change in moment between two points is equal to the area of the shear diagram
between the points.
Let us now consider the element depicted in Fig. 5.5c and write an equation
of vertical equilibrium for it. Thus
dV =-P (5.4)
This expression states that the shear decreases by an amount equal to the
concentrated load whenever we pass from the left to the right of such a load.
Similarly, by writing an equation of moment equilibrium for the element in
Fig. 5.5d, we obtain
dM = M, (5.5)
which states that a couple results in an abrupt change of the bending moment,
similar to the abrupt change in the shear caused by a concentrated load.
We now have available to us all the principles necessary for the efficient
construction of shear and moment diagrams. There remains only the task of
becoming proficient in the application of these principles. Thus let us consider the
following illustrative examples.
Example 5.1
Construct the shear and bending-moment diagrams for the beam in Fig. 5.6a.
The first step is to calculate the reactions. The importance of carrying out this
step correctly warrants using two equations of equilibrium to solve for the reactions
and a third equation to check the results. Thus
SM, = 0
20R, — 10(14) — 20(8) = 0
R, = 15kips
SM, = 0
—20R, + 20(12) + 10(6) = 0
R, = 15 kips
DF,=0
15 + 15 = 10 + 20
To construct the shear diagram for a beam with concentrated loads, we simply
apply Eq. (5.4), which states that at each concentrated load the shear changes by an
amount equal to the load. Equation (5.4) also implies that the shear remains
s and Frames Chap.5
98 Shear and Moment Diagrams for Beam
Mesa eeSesar
= 1
Shear diagram
(d)
ii Moment diagram
(e)
Figure 5.6
constant between loads. If we start at the left end of the beam and work toward the
right, upward forces cause an increase in shear and downward forces a decrease in
the shear. Using this procedure, the shear diagram in Fig. 5.6d is obtained.
The bending-moment diagram can be constructed either by Passing sections
through the member and by considering the free bodies thus formed or by making
use of Eqs. (5.2) and (5.3). From the free body in Fig. 5.6b, it is evident that the
Sec. 5.6 Construction of Shear and Moment Diagrams 99
moment increases linearly from 0 to 90 kip-ft as x varies from 0 to 6 ft. The same
information can also be obtained using Eqs. (5.2) and (5.3). In accordance with Eq.
(5.2), the slope of the moment diagrams is equal to the shear. Since the shear is
positive and constant between points A and C, the slope of the moment diagram
must also be positive and constant between these points. In other words, the
moment increases linearly with x. Equation (5.3) states that the change in moment
between A and C is equal to the area of the shear diagram between A and C. The
change in moment between these points is thus equal to 90 kip-ft. Further use of
Eqs. (5.2) and (5.3) indicates that the shear and consequently the slope of the
moment diagram is smaller between C and D than it was between A and C. It also
indicates that the change in moment between C and D is equal to 30 kip-ft. The
moment at D is thus equal to 120 kip-ft. This checks with the value of the moment at
D obtained from the free body shown in Fig. 5.6c. Finally, the slope of the moment
diagram between D and B is negative because the shear is negative between these
points.
1. The shear changes abruptly in value at each concentrated load and remains
constant in value between loads.
2. The moment diagram consists of a series of straight lines whose slopes are
positive or negative depending on whether the shear is positive or negative
in the given region.
3. The maximum moment occurs where the shear changes sign.
Example 5.2
Construct the shear and moment diagrams for the beam in Fig. 5.7a.
The first step is to determine the reactions. Although the shear and moment
diagrams must be drawn using the actual distributed load, in calculating the
reactions we can replace the distributed load by a single concentrated force of
240 kN located 6 ft from the ends of the beam.
To draw the shear diagram, we make use of Eqs. (5.1) and (5.4), which state
that the change in shear between two points is equal to the load between the points.
The negative sign in the equations indicates that downward loads, which we consider
to be positive, cause a decrease in the shear. Making use of Eq. (5.1), we find that
the shear decreases by 80 kN between points A and B and by 160 kN between B and
C. At B and C the reactions cause the shear to increase by 180kN and 60kN,
respectively. Whereas concentrated loads, like the reactions at B and C, result in
abrupt changes in the shear, distributed loads cause the shear to change gradually.
In our case, the shear decreases by 20 kN per meter of member length because of
the distributed load.
To construct the bending-moment diagram, we use Eqs. (5.2) and (@=3)lorsiree
bodies like those in Figs. 5.7b and 5.7c. Equation (5.2) states that the slope of the
moment diagram is equal to the shear. Accordingly, the slope of the moment
Chap.5
100 Shear and Moment Diagrams for Beams and Frames
20 kKN/m
C
A = B Be
180 kN 60 kN
(a)
ET} ye “GO.
20kN/m 20x 90 20 kN/m
oo — 60
(b) (c)
100
—60
—80
-k¥—« ls pa i= ~
Shear
diagram
(d)
90
—160
Moment
diagram
(e)
Figure 5.7
diagram, in Fig. 5.7e, decreases gradually from a zero slope at A to a negative slope
at B. Then at point B there occurs an abrupt change in slope from negative to
positive. Finally, between B and C the slope changes gradually from a postive slope
at B to a negative slope at C.
To obtain values of the moment, we make use of Eq. (5.3), which states that
Sec. 5.6 Construction of Shear and Moment Diagrams 101
the change in moment between two points is equal to the area of the shear diagram
between the points. Thus the moment at B is equal to the area of the shear diagram
between A and B; that is, it is equal to —160kN-m. Since the shear is negative
between A and B, the moment decreases in magnitude as we go from A to B. By
comparison, the area of the shear diagram between B and D is positive, resulting in
an increase of the moment between these points. To determine the value of the
moment at D, we must first locate point D, the point where the shear is zero. This
can be accomplished using the similar triangles that make up the shear diagram
between B and C. Thus it can be determined that the distance between B and D is
5m and that the moment at D is 90 kN-m. ©
Most of the information obtained above by using Eqs. (5.2) and (5.3) can be
checked by considering free bodies such as those in Figs. 5.7b and 5.7c. For
example, the free body in Fig. 5.7b indicates that the moment between A and B is a
quadratic function and that its value at B is —160 kN-m. Similarly, the free body in
Fig. 5.7c verifies that the moment at D is 90 KN-m.
From the foregoing example, we can conclude that for uniformly distributed
loads the shear is a linear function of x and the bending moment is a quadratic
function of x. In addition, the maximum values of the moment occur either where
the shear is zero or where it changes sign.
Example 5.3
It is required to draw the shear and moment diagram for the beam in Fig. 5.8a.
First the reactions at A and B are determined. Then, using the principles
outlined in the preceding pages, the diagrams in Figs. 5.8b and 5.8c are obtained.
Example 5.4
Construct the shear and moment diagram for the beam in Fig. 5.9a.
Equations (5.1) and (5.3) indicate that the shear in a beam is obtained by
integrating the load and the moment by integrating the shear. Since the given load
varies linearly with x, the shear must be a quadratic function of x and the moment a
cubic function of x.
To construct the shear diagram, we proceed along the same lines as in the
previous examples. However, ordinates of the moment diagram are best obtained by
using free bodies and not by attempting to calculate the areas of portions of the
shear diagram. The latter procedure is helpful only when the shear diagram consists
of triangular or rectangular areas.
The shear and moment at B are obtained using the free body of segment AB,
shown in Fig. 5.9b. First, the intensity of the distributed load at B is calculated using
similar triangles. Then equations of vertical and moment equilibrium of element AB
are employed to determine the values of V and M at B.
The maximum positive moment occurs at D, where the shear is zero. To
obtain the magnitude of this moment, it is necessary to locate point D. This can be
' accomplished by equating the upward and downward loads acting on segment AD in
Fig. 5.9c. Thus
Chap. 5
102 Shear and Moment Diagrams for Beams and Frames
20 kN
10 kKN/m
(a)
34
Shear
diagram
(b)
Moment
diagram
(c)
Figure 5.8
from which
Lor
Knowing the value of x, we can determine both the magnitude and location of the
resultant of the distributed load acting on segment AD, and hence the moment at D.
Example 5.5
Construct the shear and the moment diagram for member ABCD of the structure in
Fig. 5.10a.
To obtain the reactions at A and D, as well as‘the forces applied to member
ABCD at B, the structure is broken down into the three free bodies shown in Figs.
Sec. 5.6 Construction of Shear and Moment Diagrams 103
ae M = 36 ‘
SokL t AB
Segment AD
(b) (c)
=)
[+ x = 19.6. + Se
Shear
diagram
(d)
25.7
—36
Moment
diagram
(e)
Figure 5.9
Chap.5
104 Shear and Moment Diagrams for Beams and Frames
Jaf to
15 26
(b) (c)
2 k/ft 90
26 - (e)
(d)
41
26
Shear ~26
(f)
169
—260
=350
—514 Moment
(g)
4
Figure 5.10
Sec. 5.7 Shear and Moment Diagrams of Determinate, Rigid Frames 105
5.10b, 5.10c, and 5.10d. From the free body in Fig. 5.10d, the reaction at D and the
shear at C are obtained. Then the free body in Fig. 5.10c is used to determine the
forces that act at B on member ABC. Finally, the free body in Fig. 5.10b is used to
calculate the reactions at A.
To determine the effect of the 90 kip-ft couple at B on the moment diagram of
member ABCD, the free body in Fig. 5.10e is drawn. Since the moment to the left
of B is —350kip-ft and the 90kip-ft couple is clockwise, it is evident that the
moment to the right of B is equal to —260 kip-ft.
Flexural members do not occur only as isolated beams. Often several of them are
connected to one another to form a frame. One of the most important of these is
the rigid frame, which consists of flexural members connected to one another in
such a manner that the members cannot rotate in relation to each other at the
joints. As a consequence, moments as well as forces can be transferred from one
member to another at their connection. By comparison, when two members are
hinged to each other at a joint, the members can rotate in relation to one another
and no moments can be transferred from one to the other.
Bending-moment and shear diagrams of rigid frames are easiest to construct
if the frame is broken down into individual members and each of these is
analyzed by itself. To see how this procedure is carried out, let us consider the
following examples.
Example 5.6
Construct the shear and the moment diagrams for each member of the frame in Fig.
Sella
First we determine the reactions at A and D by applying the equations of
equilibrium to the entire frame. Next we consider the three members out of which
the frame is constructed and draw for each a free body, a shear, and a moment
diagram. We begin with member AB, pictured in Fig. 5.11b. The free body of the
member, shown in the figure, extends from A to a section just below the joint at B.
Thus, the 40 kN force acting on joint B does not appear on the free body we are
considering. Knowing the magnitude of the forces acting at A and along the
member, we can use equations of vertical, horizontal, and moment equilibrium to
calculate the axial force, the shear, and the moment at the upper end of the
member. This gives us a complete free body of member AB, from which shear and
moment diagrams in Figs. 5.11c and 5.11d can readily be constructed.
Before we can turn our attention to member BC, it is necessary to consider
the free body of joint B. Starting with the forces at the lower end of the joint, which
are obtained from member AB by applying the principle of action and reaction, and
the external load of 40 kN applied directly to the joint, we obtain the forces at the
right end of the joint using equations of equilibrium. It is necessary to consider a
free body of a joint only when there are external loads applied directly to the joint.
Shear and Moment Diagrams for Beams and Frames Chap. 5
20 kKN/m
So fins
108.6 kN
- alan
(a)
108.6 171.4
108.6 171.4
#40 |
==>
80 -
—_—_—_
171.4
(b)
Figure 5.11
In the absence of such loads, we can proceed directly from one member to the next.
We are now ready to deal with member BC. First we obtain the forces acting
on the left end of the member from the free body of joint B. Then, using the
equations of equilibrium, we determine the forces at the right end of the member
needed to keep the member in equilibrium. From the free body thus obtained, shear
and moment diagrams are drawn.
Sec. 5.7 Shear and Moment Diagrams of Determinate, Rigid Frames 107
108.6
B C
Pa en,ble
171.4
80
Shear diagrams
140
(c)
735
440
B Cc
440 B
Moment diagrams
280
(d)
The analysis of the frame is completed by drawing the free body of joint C,
and a free body of member CD. Since member CD is subjected only to axial
compression, the shear and moment are zero for the entire length of the member.
Since the members in a frame are not all horizontal, it is necessary at this
time to replace the sign convention for bending that we have used in the past with
a new one. Instead of referring to a moment as positive when it produces
,
108 Shear and Moment Diagrams for Beams and Frames Chap. 5
compression on the upper fibers of the member and negative when it produces
compression on the bottom, we will from now on simply draw the moment
diagram on the compression side of the member. This new procedure, in addition
to being more versatile than the old one, produces the same moment diagram as
the old one when the member is horizontal.
3 k/ft
64.04 52.96
( B C
432 Cc
64.04 25.96 36 ‘ \aseee
25.96 Caer Janae
64.04
42.37
42.37
432 Ar
B <n
Sn 13.22
Ge
3 k/ft
——>' A 2
36 tt
31.78 \
64.04 42.37
(b) %
Figure 5.12
Sec. 5.7 Shear and Moment Diagrams of Determinate, Rigid Frames 109
64.04
B Cc
poe eee ee 13.22
: Cc
B
iE
%
36 Shear diagrams ~
A D
ihe
(c) aude
251.6
iiss
B
iC
432
139.3
B 432 C
168.4
Moment diagrams
A D
(d)
Example 5.7
Construct the shear and moment diagram for each member of the frame in Fig.
5.12a.
First the reactions at A and D are determined. Then a free body, a shear
diagram, and a moment diagram are drawn for each of the three members. To draw
the shear and moment diagram of member CD, it is necessary to resolve the forces
acting on the ends of the member into components along and normal to the axis of
the member. At C, this is accomplished by drawing a free body of the joint as shown
in the figure.
Example 5.8
Construct shear and moment diagrams for each member of the frame in Fig. 5.13a.
As in previous examples, the reactions are determined first. Next free-body
110 Shear and Moment Diagrams for Beams and Frames Chap. 5
30 kN/m
204.3 kN 335.7 kN
|}-—————14 ———— r “a
(a)
30 kKN/m
204.3 2 as a SO SE Me es Se —
B 80 40 D
Axno 240 3357
204.3 :
B 40 335.7
80
=—_—— c
—
40 40
A 80 40
| :
——
204.3
335.7
(b)
204.3
120
B D
- 6.81 m +|
B
21.547, Cc
40 40
80
A
Shear diagrams E
(c)
|
Figure 5.13
Problems 111
456
B D
240 240
320
B 240 80 @
160
A E
Moment diagrams
(d)
diagrams are drawn for each member, and, finally, shear and moment diagrams are
constructed.
PROBLEMS
5.3. 60 kN 30 kN
112 Shear and Moment Diagrams for Beams and Frames Chap.5
5.4. 40 kN 40 kN
A B
VY, Wa
pata te
Up
5.6. 3k/ft
A B
Lp
= a 8" si
6m 4m 4m
5.8. 4k 7k
5.9. 15 kN/m
A B
Ls re | .
Problems 113
5.10. 2k/ft
es B
pp
ml
5.11. a ~— 20KN
A B
i 12m +}. 4m |
A B
5.13. 20 kN
oe 10 kN/m
een Ie aaa
5.14. 1.5 k/ft
5.15. 40 kN 20 kN/m
> B
3m 3m 8m Eu
114 Shear and Moment Diagrams for Beams and Frames Chap. 5
5.16. 60 kN
40 kN
12 kN/m |
tm btm ees
5.17. 6 k/ft
5.18. 4 k/ft
5.19. 12kN/m
5.20. 6 k/ft
5.21. 20 kN/m
5.23: 30 kN
y 25 kN/m
A
Hinge )
‘aaa ee areas
5.24. Leet
A Hinge B
ps
f+ a pr
5.25. 15kN/m
| 12m ———+}+—3
m [+ —8 m —+
|20 kN
5.26.
10kN/m
A B
J 3 m +++ —5 m —+f+-2 mf —3 m —+
116 Shear and Moment Diagrams for Beams and Frames Chap. 5
———
|}_—12——+
pe 16" ——+.—s —|
5.28 15 kN/m
5.29.
5.30.
4m
4m
Problems 117
5.31. 16 k
5.32. 14kN/m
2m
2m
5.33 50 kN 50 kN
40 kN | | 100kN-m
5.34. 8 k/ft
4’
4'
IAAT
5.36. 20 kN
12 kN/m
L ue ae 7. ie -
eye 1.5k/ft
iS
eS a Hinge
—
10°
Problems 119
5.40. Hinges
2 k/ft
A B (e D E F
f 7
5.42. Find x so that the maximum moment in span ABC is equal to the maximum moment
in span CD.
Hinges kit
Deflections: Differential Equation
The structural engineer must be able not only to determine the internal forces
and stresses that exist in a structure but also to calculate the magnitude of the
deformations that occur. Just as it is necessary to ensure that allowable stresses
are not exceeded, it is sometimes important to limit the size of the deflections that
take place. Excessive deformations may prevent the structure from functioning
properly. For example, if the members used to frame windows and doors in a
building deform excessively, the free movement of the doors and windows will be
hampered.
In some instances excessive deformations are undesirable not because they
lead to malfunctioning of the structure but simply because they make people feel
uncomfortable. A bridge may, sag at its midpoint and still be structurally sound.
However, such deflections tend to give an impression of weakness, which is
undesirable. In the case of a bridge, the problem is dealt with by cambering the
structure; that is, the structure is given an initial upward curvature that masks the
downward deflection. A similar problem arises in the upper stories of very tall
buildings. Lateral deflections caused by strong winds are harmless from the point
of view of structural safety. However, because they make the people occupying
these spaces feel insecure, their magnitude must be limited.
A second and equally important reason for calculating deflections is that
they play a vital role in the behavior of indeterminate structures. Since internal
forces and deformations are coupled to one another in the equations that govern
the behavior of these structures, both the deformations and the internal forces
must be evaluated simultaneously. In other words, whereas we can obtain the
internal forces for a determinate structure without calculating the deformations,
the analysis of an indeterminate structure requires that both internal forces and
deformations be calculated simultaneously.
In this and the following chapter we will study several different methods for
calculating deflections. Since some of these are more suited to the analysis of one
type of structure than another, it is important not only to learn the various
methods but also to be able to choose the best method for a given task.
122 ‘
Sec. 6.2 Elastic Force-Deformation Relationships 123
for actual engineering structures. For example, steel having a working stress
of 24ksi and a modulus of 29 x 10° ksi will have at its working stress a strain
equal to
Bee
Sie oo tS
0r poner in./in.
ini
The elongation of a fiber of steel that is not stressed above its proportional limit
will thus be less than qg99 of its original length.
where 6 is the elongation, L the length of the member, A its cross-sectional area,
P the load, and E the modulus of elasticity.
Flexural Members
oh1 fae La
M
6.2
ip dL or)
where r is the radius of curvature at any section and M, E and / are the moment,
the modulus of elasticity, and the moment of inertia at that section. Since
1eeedd)ed-y.
erie sd)
Eq. (6.2) can also be written in the form
ath ue (6.4)
ds I
or :
d‘y M
See (6.5)
The quantities appearing in Eqs. (6.2), (6.4), and (6.5) are defined in Fig. 6.1.
Deflections: Differential Equations Chap. 6
124
a
je
Figure 6.1
Torsional Members
The relation between the torque and the angle of twist for a torsional member is
given by
See
am ey, (6.6)
where @ is the angle of twist, J is the torsional constant, G the shear modulus,
and T the torque. The form of J depends on the shape of the cross section.
Shear Panels
j _SeekVL
es (6.7)
AG
Figure 6.2
Sec. 6.3 Direct Integration 125
where V is the shear force, A is the cross-sectional area of the edge on which V is
acting, and L is the length of the side normal to the direction along which 6 is
measured, as shown in Fig. 6.2. The value of the factor k depends on the
cross-sectional shape. It is equal to 1.2 for a rectangular section and 1.0 for a wide
flange member.
If we multiply both sides of the equation by dx and take the indefinite integral, we
obtain
dy Px?
|EF Pi)
Pre meet AG (6.10)
6.10
Figure 6.3
Deflections: Differential Equations Chap. 6
126
integrating between
The constant of integration C, is present because we are not
A second integrat ion of the equatio n leads to
definite limits.
(6.11) .
3
Ely = * Cate as
gy = il) @x=L
dx
Substitution of the boundary conditions in Eqs. (6.10) and (6.11) gives
PL
— 5
and
(te Bi
3
The expression for the deflection of the beam can now be written as
EI oy res
ax ie
from which
d 2
Er ans Pe Pbx
iy o, 1 (6.15)
Sec. 6.3 Direct Integration 127
: Pbx _ ‘a
)aE Pix ai)
————
Figure 6.4
and
Pbx?
Ely = yniere sPuEle0 ae (5 (6.16)
Iikg A WEEN
Bly \sekmarys
ee a CHET i: Gi (6.18)
To evaluate the four constants C,, C,, C;, and C,, we make use of the
boundary conditions
y=0 @x=0, y=0 @x=L
and the conditions of continuity at x = a, which are that the deflection y and the
slope dy/dx given by Eqs. (6.16) and (6.15) must be equal to the corresponding
quantities given by Eqs. (6.18) and (6.17) at x = a.
Deflections: Differential Equations Chap. 6
128
and equating the slopes given by Eqs. (6.15) and (6.17) at x = a leads to
Ci =C;
Next we equate the deflection given by Eq. (6.16) to that given by Eq. (6.18) at
x = a, which results in
Cy = 0
Substitution of the above constants into Eqs. (6.16) and (6.18) leads to the
desired expressions for the deflection. Thus
y= oF - 6-2), Ox <a
1) Pbx P(x-— a)
4 =|ila (Pe bee oa+),6 area
ea
It is obvious from a comparison of Example 6.1 with Example 6.2 that the
presence of discontinuities in the moment expression increases the complexity of
the solution significantly. Thus, the solution for the deflection of a beam with
three concentrated loads would require four separate equations and eight
boundary conditions if we followed the procedure used in Example 6.2.
Fortunately, this is not necessary. There exists a mathematical device, called the
singularity function, whose use removes the difficulties introduced by discon-
tinuities in the moment expression. To see how singularity functions are used, let
us consider the following example.
Example 6.3
Obtain an expression for the deflection of the beam in Fig. 6.4 using singularity
functions.
As before, we choose the left end of the beam as the origin of the coordinate
system. The moment at any section along the beam can be given by the single
expression
Pbx
Me ~ Bae ad (6.19)
twice. Thus
d’y Pbx
EI— = —-— + P(x -
dx* IE, Ga
dy Poe (x — a)”
EIrE
— = —-——
aL + P———
5 +.G, (6.20)
Pbx°* (x — a)?
Elyly = —-——
6L + P P-———
p qe + CG;
eG
and
= Pb (L?- b)
C=—(L-
Finally, substituting these constants into Eq. (6.20) leads to
5 1 |Zwe- 2 Pie )
y= a
EI i ) 6
Comparison of this result with the one in Example 6.2 indicates that the two are
identical.
Loading
(a)
dx
Picea
ds Oh (DY ee
= —
Deflection
curve
(b)
M
EI
M oi
El Diagram ch faa
Figure 6.5
M
do = —ar : (6.21)
;
The change in slope between the ends of the element ds is thus equal to the area
of the M/EI diagram corresponding to ds. To obtain the change in slope between
two points such as C and D, which are a finite distance apart, we add the
incremental quantities d@ for all the elements between the points. Thus
D
M
It should be noted that this theorem gives only the change in slope between two
points along a beam and not the value of the slope at a specific point.
To derive the second moment-area theorem, let us consider again the
tangents drawn to the ends of element ds in Fig. 6.5b. The distance dA between
these tangents, measured along a vertical line through D, is given by
dA = xdé (6.23)
where x is the horizontal distance from D to the element ds. Substitution of Eq.
(6.21) into Eq. (6.23) gives
M
xe
dA =x—d. (6.24)
6.24
The vertical distance dA is thus equal to the moment of the M/EI diagram
corresponding to ds about point D.
To obtain Acp, the vertical distance from point D to a tangent drawn to C,
we sum the incremental quantities dA for all the elements between C and D.
Thus
D
M
Acp = i,xx (6.25)
1. The theorem does not give a beam deflection. Instead, it gives the vertical
distance from one point on a beam to a tangent drawn to some other point.
2. The moment of the M/EI diagram is always taken about the point on the
beam at which the vertical distance is measured.
Triangle t
bh =
2b
h Age ==
2 3
——
he b
Se RE
Xa] y ee
: Anebh x, =7b
= 3
h 5
2bh = 3
Az 7 Xo aap
o
yf
> iT] x| nN i
nN
|e ap
Oy
x
Figure 6.6
deflected shape of the member. Therefore, we will not develop a sign convention
for the method.
Before considering some examples, it is worthwhile to note that the
structures to which the moment-area method is usually applied can be subdivided
into two groups: those for which the location of a @orizontal tangent to the
deflection curve is known and those for which we do not know where a horizontal
tangent is located. Since the calculations are considerably easier to carry out
Sec. 6.5 Application of the Moment-Area Method 133
Pp
EI Constant
Beam
A B
Deflection samme ca:
curve Bok € B
eS
5 fa
|
4 Diagram = ee
ae 8
Figure 6.7
eT
_ PLLOme_ PL’
LET
To determine the deflection at B, we use the second theorem, which states that the
vertical distance from the deflection curve at B to the tangent drawn to A is equal to
the moment of the M/EI diagram between A and B about B. Hence
_PLL2L _ PL?
PROTONS Uy 3 a7
Example 6.5
For the simply supported beam in Fig. 6.8, determine the deflection at midspan and
the slope at A.
The deflection curve of a simply supported beam that is loaded symmetrically
’
Deflections: Differential Equations Chap. 6
134
EI Constant
Beam
Deflection
curve
Figure 6.8
has a horizontal tangent at its midpoint. Hence the slope at A, for the beam in Fig.
6.8, is equal to the change in slope between A and B, which is equal to the area of
the M/EI diagram between A and B. Thus
ee wie, (wit
6,=- —=
AP, SET 2 LTOARI
6, =d=-—
ea Swaz
* 38EI216 384EI
Example 6.6
It is required to determine the deflection at point B for the cantilever beam in Fig.
6.9. As indicated, the stiffness of the beam is equal to EJ between B and C and 2E/]
between A and B.
When dealing with a variable stiffness it is advisable to construct the M/EI
diagram in two steps. First the ordinary moment diagram is drawn, and then the
M/EI diagram is obtained by dividing the ordinates of the moment diagram by the
appropriate stiffness.
According to Theorem 2, 6, is equal to the moment of the M/EI diagram
between A and B about B. As shown in the figure, the area of the M/EI diagram
between A and B is subdivided into a triangle and a rectangle. Thus
Pld
=
aePla
aL -
es
B
4EFI24 2EI43 # 96EI
Sec. 6.5 Application of the Moment-Area Method 135
Beam
A ywB C
M-diagram PL
=e
ues
4EI ls
ty ated ag
Mae ee
si
—D iagram
pasa Teel
3PL 3
4El
Figure 6.9
Example 6.7
Determine the slope at A and the deflection at B for the simply supported beam in
Fig. 6.10.
Since the loading is not symmetric, neither is the deflection curve. As a
consequence, the location of the horizontal tangent is not known. Problems of this
type are dealt with as follows. Drawing a tangent to the deflection curve at A and
making use of similar triangles, we can write
6g +d, dy
Zlina Feei
from which
d
Suro da
The quantities d, and d,, which are the distances from C and B to the tangent drawn
to A, are determined using the second theorem. Thus
180 360 2700 KN-m?
Ge rer Wouter
SOaan ET
180 180 kN-m?
dais ep Ms mon ET
136 Deflections: Differential Equations Chap. 6
Deflection
curve
Ef
a, m
7m
Figure 6.10
and
_ 900 — 180 _ 720 kN-m?
5
e EI EI
Substitution of the given values for E and J gives
_ 720 kN-m?
B = 0.036m = 3.6cm
~ 20,000 KN-m?
To obtain the slope at A, we note that slopes as well as deformations are assumed to
be small. Hence
d
0, = tan 64 = >
2700 kKN-m?*
oo
7 '0(20,000) KNEE
A
Example 6.8
Determine the deflection at C for the beam in Fig. 6.1},
If a tangent is drawn to A, as shown in the figure, d, and d, can be obtained
using Theorem 2. Also, from similar triangles, d, = 1.5d,. Thus the desired
Sec. 6.5 Application of the Moment-Area Method 137
Beam
Deflection
curve
PL
2EI
Figure 6.11
bc = d,; = d,
_PLLL_ PL
eek -38 -12ET
PL?
d, = 1.5d, re ae
SEI
fad0ER} ogee)bygbp Sn idBy
°* 2612 6. wEI43. . 4EI
err
Oc ~ 8EI
Example 6.9
Determine the deflection at B for the beam in Fig. 6.12a.
If a beam is subjected to several loads, it is sometimes convenient to calculate
deflections by making use of the principle of superposition. This procedure is
applicable provided that the material obeys Hooke’s law and deformations are small,
that is, provided the behavior of the structure is linear.
In accordance with the principle of superposition the deflection at B due to
both the distributed and concentrated load acting simultaneously (Fig. 6.12a) is
equal to the sum of the deflections at B produced by the loads acting one at a time
138 Deflections: Differential Equations Chap. 6
E = 29 X 10° ksi
I = 150 in.4
12K
Figure 6.12
Sec. 6.5 Application of the Moment-Area Method 139
1 45 10 3750k-ft?
15)(10 — = ———_
q, al YO) 2 EI)? 3 EI
LA.
ee 0
10) =500 k-ft?
43 EI ‘ea
1375 k-ft?
One = 0.5d, Ca d, = Cpe
5542 k-ft?
Gs = Oni. + Ops =
Substitution of the given values for E and J and multiplication of the numerator by
1728 to obtain consistent units gives
ee5542(1728)
ee k-in°
00
® ~ 59 x 10°(150) k-in® 7
Example 6.10
Determine the deflection at C for the beam in Fig. 6.13a.
It is possible to apply the principle of superposition to the moment diagram of
a beam as well as to its loads. In other words, just as we are able to break down the
total loading into individual loads and consider them one at a time, it is possible to
subdivide the moment diagram into several parts and consider these separately. For
example, if we measure x from the left end of the member, the moment of the beam
in Fig. 6.13a is given by
M = 45x, Ope
M = 45x — 40(x — 3), 3<x <6
10(x — 6)?
M = 45x — 40(@@ — 3) - 5 , GSx <P
If the terms making up the above expressions are considered separately, then the
moment diagrams shown in Fig. 6.13c result. Between A and B the moment consists
of a single linear function, between B and C it is made up of the sum of two linear
functions, and between C and D it consists of the sum of two linear functions and a
quadratic function.
The deflection of C is now obtained by applying the moment-area theorem to
the diagrams in Fig. 6.13c.
1540 — ———
1180
— ——(6)-
7560
= —
Pa yO THI[OO 3 ET”)¥ EI
1270 1120 es
Gey Os Be 3)(1
2340
Oc oS 0.5d, aed ae —
ARQRETP
Deflections: Differential Equations Chap. 6
140
40 kN
10 kN/m
45 kN 55 kN
(b)
a
540EI
_ 360
EI
eRe d
EI
Figure 6.13
or
2340 kN-m?*
390,000 NE? RE ABS om
6.6 CONJUGATE-BEAM METHOD
In Section 6.3 it was demonstrated that the deflection of a beam can be obtained
by integrating the expression for the bending moment twice. Unfortunately, this
procedure becomes relatively involved whenever the bending moment cannot be
written as a single continuous function for the entire beam. The conjugate-beam
method, presented in this section, avoids this difficulty by substituting for the
sometimes cumbersome process of integration the relatively simple procedure of
constructing shear and moment diagrams. It is possible to do this because the
Sec. 6.6 Conjugate-Beam Method 141
drawing of the shear and moment diagrams is equivalent to the analytical process
of integration.
Consider the differential beam element depicted in Fig. 6.14, on which are
shown acting a distributed load w, shears V and V + dV, and moments M and
M + dM. Summing forces in the vertical direction leads to
dV = —wdkx
or
V= -| wae (6.26)
In addition, taking moments about the right end of the element and neglecting
higher-order terms gives
dM = Vdax
or
M= |V dx (6.27)
M= [J wax dx (6.28)
Equations (6.26) and (6.28) indicate that the shear in a beam can be
obtained by integrating the load once and the moment by integrating the load
twice. As a rule, structural engineers do not determine shears and moments by
integrating analytic expressions for the loading. Instead, they draw shear and
moment diagrams. In other words, the drawing of shear and moment diagrams
can be substituted for the analytical process of integration. Let us now see how a
similar substitution can be made in the calculation of slopes and deflections.
Since the curvature of a beam is proportional to the bending moment, the
dy
ial Vtav
Figure 6.14
142 Deflections: Differential Equations Chap. 6
slope and deflection of the beam can be obtained by successively integrating the
moment. Thus
dy _M
dx? EI
from which
dy [=
ates | tid 6.29
Ae Whi es
and
M
= — dx dx 6.30
4 ||fare C3
It is our intention to replace the integrations indicated in Eqs. (6.29) and
(6.30) by the drawing of shear and moment diagrams. To demonstrate how this is
accomplished, let us consider the beam in Fig. 6.15. We begin by constructing the
M/EI diagram of the beam. Next we introduce a new beam referred to as the
conjugate beam. This beam has the same length as the real beam. However,
instead of being loaded with the same load as the real beam, it is loaded with the
M/EI diagram of the real beam. In other words, the conjugate beam is subjected
to a distributed load whose intensity at any point is equal to the value of M/EI
for the real beam at that point. We now proceed to construct the shear and
moment diagrams of the conjugate beam.
According to Eqs. (6.26) and (6.28), the drawing of V and M diagrams for a
given loading is equivalent to two integrations of the loading. Thus the shear and
moment diagrams of the conjugate beam represent one and two integrations,
respectively, of the M/EIJ diagram of the real beam. But, in view of Eqs. (6.29)
and (6.30), integration of M/EI of the real beam gives the slope and deflection of
the real beam. We thus conclude that the shear and moment diagrams of the
conjugate beam represent the slope and deflection of the real beam.
The foregoing can be summarized as follows:
1. The shear at any section of the conjugate beam is equal to the slope at that
section in the real beam.
2. The moment at any section of the conjugate beam is equal to the deflection
of the corresponding section of the real beam.
In the foregoing it has been tacitly assumed that the conjugate beam has the
Same support conditions as the real beam. Although this is true for a simply
supported beam, it is not true for other boundary conditions. In general, the
boundary conditions of the conjugate beam correspending to a given real beam
can be determined by making use of the relationships between the two beams
Sec. 6.6 Conjugate-Beam Method 143
Real beam
(a)
Conjugate beam
Shear diagram of
conjugate beam
(d)
Moment diagram of
conjugate beam
(e)
Figure 6.15
stated in (1) and (2) above. If this is done, the results given in Table 6.1 are
obtained. For example, if the real beam has a fixed support, then y = y’ = 0 at
that support. At the corresponding support in the conjugate beam, M = V = 0,
which defines a free end. Thus a fixed support in the real beam becomes a free
end in the conjugate beam.
The conjugate-beam method, like the moment-area method, is primarily
intended for the calculation of slopes and deflections in simple beams.
Deflections: Differential Equations Chap. 6
144
TABLE 6.1
: Hinged
Hinges y' #0 v #0
=0 M= 0
s ; Lyn
Fixed Free
\——— ———
N
; , Fixed
y #0 \———
NN
Continuous
y’ = continuous V = continuous
_
== OS
yf =e M=0
ei
2. SIRI
i 4PL?
© 81EI
Since the slope in the real beam is equal to the shear in the conjugate beam, the
slope @, at the left end of the real beam is equal to the left reaction of the conjugate
beam. Thus
opr PL
A
81EI
Sec. 6.7 Application of the Conjugate-Beam Method 145
Real beam
ZR
Conjugate beam
_ aplie
81EI
Pu?
27EI
a ce 4PL 3
t 243EI
So:
5PL2
81EI
Figure 6.16
The deflection in the real beam is equal to the moment in the conjugate beam.
Hence the deflection at B in the real beam is obtained by determining the moment at
that point in the conjugate beam. Using the free body in Fig. 6.16c and summing
moments about B gives
SPLOL ee La
== ——M, =0
81EI3 27EI9 7
_ as
ee OAS ET
Thus
6, =
4PL?
“Se VASET
146 Deflections: Differential Equations Chap. 6
Note that both the slope and the deflection at specified points in the real
beam can be determined simply by considering free bodies of the conjugate
beam. It is unnecessary to draw either the shear or the moment diagram of the
conjugate beam for this purpose. However, should the entire deflection curve of
the real beam be required, then we would have to construct the bending-moment
diagram of the conjugate beam.
Example 6.12
Determine the deflections at points B and C for the beam in Fig. 6.17a.
The conjugate beam loaded with the M/EI/ diagram of the real beam is shown
in Fig. 6.17b. In accordance with Table 6.1, the conjugate beam is fixed at C and
free at A. To calculate the reactions of the conjugate beam, the loading is divided
GC Real beam
u
fe
sea Se
(a)
Ma _Hpa es
41wLl4
C 384EI
Conjugate beam
_ 7wl3
Cc 48EI
A B Mz = Tw 4
192EI
wL? wL?
16EI 16EI
Figure 6.17
Sec. 6.7 Application of the Conjugate-Beam Method 147
ee TwL>
Ke
48EI
and, from moment equilibrium,
_ wi 3L wL* SL | wL? 3L
Cc
16El 4 16EI6 =§ 48EI 8
_ 41wL*
Cc
~ 384EI
Since the deflection in the real beam is equal to the moment in the conjugate beam,
A 41wL*
~ 384EI
(oy
To obtain the deflection at B, we use the free body shown in Fig. 6.17c. Thus
R,= —
P EI
The reactions at C can now be determined from a free body of the entire conjugate
beam. Thus
466.7
Re=
EI
4000
bs Ae
and
_ 4000 kip-ft?
r)
e EI
Substituting for EJ and multiplying the numerator by 1728 to obtain consistent
Deflections: Differential Equations Chap. 6
148
E = 29 X 10° ksi
I = 120in.*
Figure 6.18
units gives
4000(1728)
eo 29 x 10°(120) = 1.99 in.
PROBLEMS
6.1 to 6.6. Use the method of double integration to obtain an expression for the deflection.
6.1. 7 *
pa
Problems 149
6.4. Wo
- ee
7 ff
6.5. P, P,
7 a Yi
ee
a
6.6. w
Yj Ue
-——. __|
6.7 to 6.35. Use the moment-area method to calculate the indicated quantities.
L L
ig 2 2
pt
6.10. Find 53, 5c, 9¢; E = 200 x 10°kN/m’, I = 20 x 10° mm‘.
10 kN
kK 2m aL 2m >
Boa pe
6.12. Find 5,, @,.
P
A : C
: sl
Y Yi
-E
6.13. Find 65, 04; E = 200 x 10°kN/m’, J = 50 x 10° mm‘.
20 kN 20 kN
Problems 151
A Cc
B
4k 4k
B c D
A E
Vp 4
A E
I I I
7B DY
ieee +h a a
wt tap
6.18. Find 6¢; E = 30 x 10’ ksi, J = 100in.*.
12k 12k
pepe pps
6.19. Find 63, 04.
> Q
p+ ofp a J
Deflections: Differential Equations Chap.6
152
10 kN 10 kN
6.21.
2P P
A Di
B C
Se
Gs 7
9a
6.23. Find 6¢; E = 200 x 10°kN/m?, J = 200 x 10° mm‘.
10 kKN/m
A k B ‘ aveper
12 kN 12 kN
A C E
B D
ep
6.27. Find 6c.
Ya a Sa:
Sess eee
6.28. Find 5,, 04; E = 30 x 10° ksi, J =" 100in~*.
1.5k/ft
A k B mt D
+ 10 1 10 +6 4
2k 6k 2k
B D
E
WZ Lp,
By 10’ 10’ a 5! +
12k
B C D
C, YZ
[ame Sta 3a —— oh —aol
10 kN/m
A
B
facieseee ns
6.34. Find 6,; E = 30 x 10° ksi, J = 120in.*.
|
Ok/tt
Problems 155
A C B
}-———12,———+}._ 12 ———
\
6.36 to 6.45. Use the conjugate-beam method to calculate the indicated quantities.
yi
a -
Y ce Y
20 KN
fame —m ——
he B ‘Le
Y
A B
+b} $+
6.41. Find 5,; E = 200 x 10°kN/m’, J = 40 x 10°mm”.
40 kN
SS
B C D
i 2m awecRres
Y, 2 k/ft
. 6' + 4' a
a aS
6.44. Find 6,; E = 30 x 10° ksi, J = 80in.*.
24k 12k
0p ate
7 Lie
10 kN 40 kN 10 kN
B D
2m —efe—2 ++ 2m te re +
Deflect ions:
Energy Methods
aE
Rh, SZ,"HU
Sh ch
Vom
ASAP EZ
oo
as RN WES. YE
ee |
SR
ROR WERE EES PU EE
ae
hh HR
SN a I
WS
ee
A
ST TS A ES
ss SD
ee
In the previous chapter we calculated deflections using methods whose basis is the
differential equation of the member. It will now be shown that deflections can
also be obtained by using the principle of conservation of energy. This principle
can be applied by using either the concept of real work or the concept of virtual
work. Of the two approaches, the latter leads to a method that is far more
versatile and consequently more useful than the former. However, to fully
understand the ideas involved, we will consider both procedures.
Since the work performed on the elements of a structure by the internal forces
causes the elements to become deformed, the internal work is sometimes referred
to as strain energy. The principle of conservation of energy is then stated in
the form
external work = strain energy
In a structural system the external work will be equal to the internal work
only if the loads are applied gradually and no acceleration occurs and if the
strains in the system remain elastic. If the system does accelerate, some of the
external work will be transformed into kinetic energy; and if inelastic behavior
occurs, energy will be lost in the form of heat. The above principle is thus
applicable only to elastic systems subjected to static loads,
Before we apply the principle of conservation of energy, let us consider the
concept of work. For the purposes of structural mechanics, work is defined as the
product of a force and the displacement, in the direction of the force, through
which the force acts. If the force remains constant in magnitude throughout the
entire displacement, the work is simply equal to the product of the force and the
total displacement.
W=PA (7.1)
However, in many instances the force does not remain constant. In this case the
total work is obtained by summing small increments of work for which the force
4
158
Sec. 7.2 Principles of Conservation of Energy 159
a Pe
ne)
8
A, 3
Ut
T date 4
Deflection A
Pe
(a) (b)
Figure 7.1
W= |PdA (7.2)
Let us now determine the work performed by force P; in stretching a bar an
amount A,;, as shown in Fig. 7.1a. If the load is applied in a static manner (that
is, both the load and the resulting deformation increase together from zero to
their final value) and if the material obeys Hooke’s law, then the load-deflection
curve will be as depicted in Fig. 7.1b. Since the applied load does not remain
constant, we use Eq. (7.2) to calculate the total work. Thus
As
W= [aw - | PdA
0
Substituting kA for P,
_ Ie
ita (7.5)
where @ is the angle of twist.
To see how the principle of conservation of energy can be used to determine the
deflection of a structure, let us consider the beam in Fig. 7.2a and calculate the
deflection 6 under the load P. It is assumed that P is applied gradually and that
the material out of which the beam is made remains elastic.
To apply the principle of conservation of energy, we must determine both
the external and the internal work. The external work consists of the total work
performed by all the external forces. Since the supports do not move in the
vertical direction as the structure deforms, the reactions at A and C perform no
work and only P contributes to the external work. Thus, in accordance with
Eqr(733)5
Wexr = ze (7.6)
2
Although the internal work theoretically consists of the total work due to all
the internal forces, it is sometimes possible in certain structures to neglect the
effects of some of the internal forces compared to that of others. In a truss only
(b) Q
Figure 7.2
Sec. 7.3 Method of Real Work 161
axial forces exist, and there is consequently no problem in deciding that they must
be used to calculate the internal work. However, in a beam the internal forces
consist of both shears and moments. Since it has been shown that the work
performed by the shear forces is negligible compared to that due to the moments,
provided the span of the beam is at least three times its depth, it is customary to
use only the moments to calculate the internal work for ordinary beams.
Figure 7.2b depicts an element dx of the beam we are considering. As the
beam bends, the cross sections of the element, on which the internal moments
act, rotate relative to one another by an angle d@. Since the internal moments,
like the external load, increase gradually from zero to their final value as the
beam is loaded, the work performed on the element by these moments is
dWixt = ——
Mdé
Dalene J 2/ 2EI
from which
sagod
~ 48EI
Although the method of real work led to a very simple and straightforward
solution in the foregoing example, it has some serious shortcomings. The method
can be applied only to a structure with a single, concentrated applied load.
Furthermore, it can then be used only to calculate the deflection under that load.
If we attempt to determine the deflection of some other point not under the load,
the deflection would not appear in the expression for the external work and
consequently could not be evaluated. On the other hand, the presence of more
Deflections: Energy Methods Chap. 7
162
than one external load would lead to an external work expression containing
several unknown deflections, one for each load. However, there would still be
only a single equation available to solve for these deflections.
The method of real work is thus not a very useful procedure. However, it
does serve the purpose of providing a good introduction to the method of virtual
work to be considered next.
In this section we will consider an energy method for determining deflections that
is not subject to the serious shortcomings of the procedure outlined in the
previous section. In fact, of all the methods for calculating deflections that we
have considered so far, the method of virtual work is the most versatile. Whereas
each of the procedures studied previously is applicable to some types of structures
but not to others, the principle of virtual work is well suited for dealing with all
types of structures, including beams, frames, and trusses.
To help us in the development of the method, let us consider a specific
problem, the calculation of the deflection 6, for the beam in Fig. 7.3a. A typical
element of the beam, showing the internal moment M that acts on any cross
section and the rotation d@ between two adjacent cross sections, is shown in
Fig. 7.3b.
Let us now consider a second beam, shown in Fig. 7.3c. To this beam we
apply, instead of the actual load, a unit load at the point where we are looking for
the deflection and in the direction of the desired deflection. In other words, we
apply a unit vertical load at A. This second beam is referred to as the dummy
structure. To differentiate between the moments and rotations in the dummy
beam and those in the real beam, we designate the moments and rotations in the
dummy structure by m and aq, as shown in Fig. 7.3d.
We will now subject the dummy beam, on which the 1k load is already
acting, to a virtual displacement and then apply the law of conservation of energy
to the resulting work. The term virtual displacement refers to a displacement that
is not actually occurring but is only imagined to take place. Unlike a real
displacement, a virtual displacement causes no change in either the external or
internal forces acting on the structure.
Thus we imagine the dummy beam with the 1k load already in place to be
subjected to the deformation that the actual load P produces in the real beam. In
other words, every point in the dummy beam is made to go through the
deformation that occurs at that point in the real beam. As a consequence, point A
moves down an amount 6, and the 1k load does an amount of work 1 - 5,4. Since
the 1k force is constant throughout the displacement, there is no 3 factor in the
work term. We refer to the term 1- 6, as the external virtual work.
The application of the virtual displacement to the dummy beam also gives
rise to internal virtual work. For each differential element, the existing moment m
Sec. 7.4 Virtual Work 163
y | Dummy structure
(c)
A da
i \re
ne
\
\
ea
m m
\
i \
L La
—| dx }-—
(d)
Figure 7.3
moves through a rotation dé. If these increments of work are summed up over
the entire beam, we obtain for the total internal virtual work
az “ Mmd
internal virtual work = I md@ = | eek
0 0 EI
equate the external virtual work to the internal virtual work. Thus
’ Mm dx
1-6, = aap (7.9)
Since all quantities appearing in this expression except 6, are known, the
expression can be used to evaluate 6,.
The principle of virtual work, embodied in Eq. (7.9), differs from the
method of real work in one very important way. In real work, work is the result
of forces acting through displacement that they themselves cause. By comparison,
in virtual work, work results when one force system acts through displacements
caused by another force system. In the preceding derivation, virtual work results
when the 1k load and internal moments m of the dummy structure are made to
act through the external and internal deformations produced by the actual load P
on the real beam.
a, Dummy beam
(b) @
Figure 7.4
Sec. 7.5 Application of Virtual Work to Beams and Frames 165
TABLE 7.1
Element x=0 M m
AB A 0 x
BC A P(x = 5) x
Fig. 7.4b. Next, the moments M and m needed to solve Eq. (7.9) are calculated and
listed in tabular form. To carry out the integration indicated in Eq. (7.9) requires
that the beam be subdivided into a sufficient number of sections so that both M and
m are continuous functions of x in each section. Thus we have divided the beam into
two sections, one from A to B and another from B to C. The origin for the moment
expression in any section should be chosen so that the resulting expression is as
simple as possible.
The moments listed in Table 7.1 are written employing point A as the origin
for both segments AB and BC. Substituting these expressions into Eq. (7.9) and
using the appropriate limits of integration, we obtain
i [ Pea eeee
6, = ——_— + |
0 EI Li2 EI
ise:
6. ~ 48EI
For comparison, the solution is repeated using the moments in Table 7.2. In
this instance the origin for segment AB is taken at A and the origin for BC at B. As
before, the limits of integration in Eq. (7.9) are chosen in accordance with the
location of the origins. Thus
Ne (ieO(x) dx i:(fePx{(L/2) + x] dx
EI EI
ai 5PL*
4 48EI
TABLE 7.2
Element x =0 M m
AB A 0 x
BC B Px 5 +x
Since use of the method of virtual work will not be restricted to simple
beams, whose deflected shape can be obtained by inspection, it is necessary to
adopt a sign convention. The following convention will be employed:
1. The same criterion must be used to define positive bending for both the real
and the dummy structure; that is, both M and m must be assumed positive
when they produce compression on the same side of the member.
2. A positive answer for the deflection indicates that the direction of the
deflection is the same as that of the dummy load. A negative answer means
that the deflection is opposite in direction to the dummy load.
2 k/ft
— —
E = 29 X 108 ksi
I = 100 in.4
(a)
(b)
Figure 7.5
Sec. 7.5 Application of Virtual Work to Beams and Frames 167
TABLE 7.3
Element x=0 M m
AB A —15x + x? 0.5x
BC Cc Me x
To calculate the required deflection, we place a unit load on the dummy beam
at C, as shown in Fig. 7.5b. Expressions for M and m are then obtained and listed in
Table 7.3. The moments for segment AB are written using an origin at A and the
moments for segment BC using an origin at C. The moments are assumed to be
positive when they produce compression on the lower edge of the beam. Substitu-
tion of these moments into Eq. (7.9) gives
A eae |9Pee
(—15x + x7)(0.5x) dx
ee (x )\(x)
| pee ac
i EI 0 EI
1+ be = Kip
OO kip*-ft?
2500 tt
EI
Finally, we substitute values for E and / and make use of a conversion factor to
obtain consistent dimensions. Thus
2500(1728) :
© ee
* 29 x 10°(100) = 149
zis in.
The positive sign of the result indicates that 6; is in the same direction as the dummy
load, that is, 6¢ is downward.
Example 7.3
Determine both the vertical and the horizontal deflection at A for the frame in Fig.
7.6a.
To calculate the vertical deflection, we apply a unit vertical load at A to the
dummy structure as indicated in Fig. 7.6b. A second dummy structure with a unit
horizontal load at A, as shown in Fig. 7.6c, is required to calculate the horizontal
deflection.
The expressions for M, the moment in the real structure, m,, the moment in
the dummy structure subjected to a unit vertical load, and m,,, the moment in the
dummy structure subjected to a unit horizontal load, are listed in Table 7.4. All
moments are assumed to be positive if they produce compression on the inside of the
frame as indicated below the table. It does not matter how positive bending is
defined. The only important factor is that the same convention is used for all
moments within a given structure segment.
Deflections: Energy Methods Chap.7
168
100 kN-m
50 kN
(a)
C
1kN—
D
7 <1 KN
4kN-m NS
5 kN-m
1kN
(b) (c)
Figure 7.6
eee Ndle*akMORO
(0)(x) dx
=4 o
(EI
7 (50x)(2 + x) dx
El 0
ANU
° (100)(4) dx
EI
2333 KN? - m°
3
p EI
:
0
oo
Lidar etic |? Oe
(50x)(0
0
s
EI hs
=
EI’
—1250 kN’ - m°
Le On == Pe Sao
EI
TABLE 7.4
M thy ae
Element x=0
B 50x 25a 0
BC 100 4 Pi
CD re
from which
Example 7.4
Determine the rotation of joint C for the frame in Fig. 7.7a.
The method of virtual work can be used to calculate rotations or changes in
3 k/ft
15)
aot ee 29
X 10? 3 ksiksi
oad I = 240 in.4
(b)
Figure 7.7
170 Deflections: Energy Methods Chap. 7
TABLE 7.5
Element a0 M m
AB A 0 0
-BC B 30x — 1.5x? —0.05x
CD D 0 0
This equation is very similar to Eq. (7.9). In fact, if we refer to both couples and
forces as generalized forces and to deflections and rotations as generalized
deflections, as is customary in mechanics, then the two equations are essentially
identical.
The moment expressions needed to solve Eq. (7.10) are listed in Table 7.5.
Moments are assumed to be positive if they produce compression on the outside of
the frame. Substitution of these expressions into Eq. (7.10) gives
mM dx iT dx
1-5,=
- EI
+ dete!
GJ (7.11)
Sec. 7.5 Application of Virtual Work to Beams and Frames 171
(a)
(b)
Figure 7.8
where T is the torque at any section in the real structure and ¢ is the torque at the
corresponding section in the dummy structure.
Substitution of the expressions for the moments and torques listed in Table 7.6
into Eq. (7.11) gives
s_ 1042(1728) 1250(1728)
= 1.04 in.
4 ~ 30 x 10°(144) | 12 x 10°(288) a
TABLE 7.6
Element x=0 M m T t
AB A 10x x 0 0
BC B 15x x 50 =)
172 Deflections: Energy Methods Chap. 7
Dummy
structure
|
1k
(b) &
Figure 7.9
Sec. 7.6 Deflection of Trusses using Virtual Work 173
E = 30 X 10° ksi
4 @ 20' = 80' 4
bf 1 ils) 20 0 0 2 0
cg eats | "50 0 0 1 0
4220.3
1.6=2 ase
_ 4220.3(12)
30 x 103
= 1.69 in.
(c)
Figure 7.10
174 Deflections: Energy Methods Chap. 7
the bars in the truss and equating it to the external work gives
uSL
f=) we 12)
Example 7.6
Determine the vertical deflection of point g for the truss in Fig. 7.10a.
The solution of Eq. (7.12) is best carried out by making use of a table, as
indicated in the figure. First the bar forces S in the real structure and the bar forces u
in the dummy structure are calculated and entered in the table. Next the quantities
uSL/A are determined for each bar and their sum obtained. Since E is constant, it is
easier to obtain the sum of all the uSL/A terms first, and then divide by E, than to
divide each of the terms by E before their sum is determined. Because of symmetry,
some bars contribute two identical uSL/A terms to the calculations.
Since the solution is positive, the truss deflects in the same direction as the
dummy load, that is, downward.
Mmdx SuL
ils iehe = ee
EI AE
where M and m are the real and dummy moments in the beam and S and wu are the
real and dummy axial forces in the cable.
Making use of the expressions for M and m given in the table in Fig<7,11¢
leads to
1 4
180= =| | 15x*dx + [60x?ae] +
2
90(1.5)(6)
EI Jo 0 AE
480 810
G
> 16:0000530,.000. “=
Sec. 7.7 Deflection of Structures 175
60 kN
0.5 kN 1.0k
Pee Tea Ts
ce
Note : Moments are assumed positive when they
produce compression on the lower side of the beam.
Figure 7.11
176 Deflections: Energy Methods Chap. 7
PROBLEMS
7.1 and 7.2. Use the method of real work to calculate the indicated quantities.
+434L 2L
7.3 to 7.37. Use the method of virtual work to calculate the indicated quantities.
po ee
7.5. Find 63; E = 200 x 10°kN/m’, J = 100 x 10° mm‘.
20 kN
4 10 kKN/m
$9 4
Problems 177
a ras Ue
+ — 19’ +} 6° rhe 6 +f — 19’ —+
7.8. Find 6,; E = 30 x 10° ksi, J = 50in.*.
20 k
rs
7.9. Find 6¢; E = 200 x 10°kN/m?, J = 40 x 10° mm‘.
10 kN/m
amendmen wine
7.10. Find 6p; E = 200 x 10°kN/m’, J = 20 x 10° mm‘.
40 kN
Deflections: Energy Methods Chap.7
178
i) 4
7.13. Find 6,; E = 30 X 10° ksi, J = 450in.*.
40k
+ — 19’ ——+}+——19’——+|.— 6—
20kN/m
10kN/m
A 4 B me D
ptm +p— am em
7.15. Find 6,; E = 30 < 10° ksi, J = 80in.*.
20 k 8k
A | |
B C D E
5k
A B 7; C > E
2 k/ft
0 safe
7.18. Find the vertical deflection at A; E = 30 x 10° ksi, J = 80 in’.
3 k/ft
20 k ~<A
ae
10k
20 kN/m
60 kN
180 Deflections: Energy Methods Chap. 7
7.20. Find the horizontal deflection and the rotation at C; E = 30 x 10° ksi, J = 60in.*.
4k/ft
7.21. Find the horizontal deflection at B; E = 200 x 10°kN/m7, J = 800 x 10° mm‘.
20kN/m
Tf aaa
50 kN
—pP
4m
3m
I ell
+ tom —_——,
aie 10k —— A 10
B fe
Problems 181
8m 10 KN
7.24. Find the vertical deflection at E; E = 200 x 10°kN/m’, J = 400 x 10° mm‘.
20 kN
10 kKN/m
toate 8m 4m
2 k/ft
182 Deflections: Energy Methods Chap. 7
7.26. Find the rotation and the vertical deflection at A; E = 200 x 10°kN/m/’, J = 120 x
10° mm‘.
Jet
afe 8m
7.27. Find the horizontal deflection at E; E = 30 x 10° ksi, J = 576in.*.
20 kN
ee C 2m
ee
7.32. Find the vertical deflection at E; E = 30 x 10° ksi; the area of each bar is noted in
the figure.
10k 10k
Deflections: Energy Methods Chap. 7
184
7.33. Find the vertical deflection at C; E = 200 x 10° KN/m’; the area of each bar is equal
to 500 mm’.
ra Lou
7,34. Find the vertical deflection at A; E = 30 x 10° ksi; the area of each bar is noted in
the figure.
8k 12k
Cable (A)
10 kKN/m
Beam (1)
eae a .
Problems 185
Cable (A)
10’ 1k/ft_ Cable (A) my
et
a ALS
a
7.37. Find the horizontal displacement at C; E = 200 x 10°kN/m’, J = 150 x 10°mm‘*,
A = 50mm’.
Cable
20kN/m
=r if
+ ssn i =7
eet ; 5 Rife 7
: ae ee a 7 “a >
‘t- ar:
re,
: _< 4 oa
a :
a,
Many structures must resist moving loads, in addition to loads such as their own
weight that remain fixed in place. For example, the force exerted by a truck on a
bridge may act anywhere along the span of the bridge. Since stresses caused by
moving loads will vary with the position of the load, and since structures must be
designed for the largest stresses that will occur, it is necessary to determine the
position of the load that produces the maximum stresses in the structure. A
convenient way of doing this is to make use of influence lines.
Influence lines provide us with a systematic procedure for determining how the
force in a given part of a structure varies as the applied load moves about on the
structure. For example, let us determine the value of the reaction R, for the
simply supported beam in Fig. 8.la, due to a unit load placed at different
distances x from the left-hand end of the member. Writing an equation of
moment equilibrium about A gives
Rp ole) (8.1)
(b)
Figure 8.1
188
Sec. 8.3 Influence Lines for Beams 189
It is evident from this equation that Rz = 0 when the load is acting directly over
the left support and that Rg = 1 when the load is at the extreme right end of the
member. Similarly, Eq. (8.1) can be used to obtain the magnitude of R, for any
intermediate position of the load. This has been done for several points, and the
results are given both in tabular form and graphically in Fig. 8.1b. The diagram in
the figure is referred to as an influence line. Its ordinate, corresponding to any
value of x, gives the magnitude of Rz due to a unit load acting at x. An influence
line can thus be said to depict graphically the effect, on some structural
component, of a unit load as it moves along the structure. In this case, the
influence line indicates that Rg varies linearly from 0 to 1 as a unit load moves
from the left end to the right end of the member, and that R, attains its
maximum value when the load is directly over the right reaction.
Using the procedure defined above, influence lines can be drawn for a
variety of internal and external forces, such as shears and moments at different
points in a beam, as well as forces in various bars of a truss.
In dealing with influence lines, it is important not to confuse them with
moment diagrams. Whereas a moment diagram gives the value of the bending
moment at different points in a beam for a load that is fixed in place, an influence
line gives the value of the bending moment at one point in the beam for different
locations of the load.
We will divide our study of influence lines into two parts. First we will
concentrate on the procedure used to construct influence lines, and then we will
make use of influence lines to determine the placement of loads for producing the
maximum forces in various parts of a structure.
A Cc 2
(a)
1.0
THE [anil Rertey57+sem enteaibpt snl
(b)
1.0
0.6
(c)
1 Mc
DCEome
Ra Re
(d)
M 1
[pei got on
Ra
=!) q
Re
(e)
2.4
0.4
IL for V,
0.6
(g)
Figure 8.2
Sec. 8.3 Influence Lines for Beams 191
1. When the load is directly over one of the supports, the reaction at that
support resists the entire load.
2. As the load moves away from the support, the value of the reaction at the
support decreases linearly until the load reaches the other support, at which
time the value of the reaction at the first support is zero.
3. From vertical equilibrium, the ordinates of the influence lines for the two
reactions must add up to unity for any point on the beam.
Let us now draw the influence line for the bending moment at C. To obtain
an expression for Mc, it is necessary to consider a free body of the beam either to
the right or to the left of C, as shown in Figs. 8.2d and 8.2e. When the unit load
is to the left of C (Fig. 8.2d), the right-hand free body is the simpler of the two to
use for determining M-. From this free body,
Me = 4Rz, Ox=x <6
According to this relation, the segment of the influence line for Mc between A
and C, shown in Fig. 8.2f, is obtained by taking the influence line for Rg between
A and C and multiplying it by 4. The influence line has been drawn with the
assumption that positive bending corresponds to compression on the upper part
of the beam.
To obtain the influence line for Mc when the load is between C and B, we
consider the free body to the left of C in Fig. 8.2e. From this free body,
Mc = 6R,; 6<=x = 10
Thus the section of the influence line for Mc between C and B is obtained by
taking the influence line for R4 between C and B and multiplying it by 6.
The influence line for the shear at C is drawn using a procedure similar to
that employed for constructing the moment influence line. The right-hand free
body in Fig. 8.2d is used to obtain the influence line for V. when the load is
between A and C, leading to
Vo = —Rz, 0O=<=x =6
192 Influence Lines Chap. 8
and the left-hand free body in Fig. 8.2e is used when the load is between C and
B, giving
Vo i Ry, 6 ee 10
The above relations indicate that Vc is equal and opposite in sign to Rg when the
load is between A and C and that V< is equal to R4 when the load is between C
and B. The influence line for V., shown in Fig. 8.2g, is thus obtained directly
from the influence lines for Rg and R,.
It should be noted that we assumed a positive shear to be a force causing a
clockwise moment about any point in the free body on which the force is acting.
Example 8.1
For the beam in Fig. 8.3a, construct influence lines for the following: the reactions at
A and B, the moments at A and D, and the shears at D and just to the left of A.
a
(a)
25) 0.5
(b)
Ra Re
(c)
1
| Mp Vo
8 >| \~—12
Ra Re
(d) 4
Figure 8.3
Sec. 8.3 Influence Lines for Beams 193
Vp Mo
8>{ |e—12
Ra Re
(e)
1
Ma
Va
(f)
es)
1.0
0.6
IL for Ry
(g)
1.0
0.4
IL for Rg
=0:5
(h)
4.8
IL for Mp Eeceorernent
=6,0
(j)
0.5 0.6
IL for Vp
—0.4
(k)
IL for M,, Peetsteate oe mb ace
—10
(I)
pS ea
a tO
(m)
B
To begin, we place a unit load at C and determine the reactions at A and
R, and R, at point
(Fig. 8.3b). This gives us the ordinates of the influence lines for
can
C. Next we move the unit load from C to A (Fig. 8.3c). The effect of this on R,
be determined by considering moment equilibriu m about A. The moment of the unit
by
load, which varies from 10 to 0 as the load moves from C to A, must be balanced
20R,. Thus the magnitude of R, decreases from 0.5 to 0 (Fig. 8.3h). Similarly, it is
evident from vertical equilibrium that R, = 1 + Rp and that R, therefore decreases
from 1.5 to 1 as Rg decreases from 0.5 to 0 (Fig. 8.3g).
Once the load is to the right of R,, the overhang no longer has any effect on
the behavior of the beam. The structure now behaves as if it were a simply
supported beam. Thus R, varies from 1 to 0 and R, from 0 to 1, as the unit load
moves from A to B.
Having obtained the influence lines for R4, and Rg, we can construct the
influence lines for Mp and V, by writing equations that relate Mp and Vp to R, and
R,. This is accomplished by using Figs. 8.3d and 8.3e. When the load is between Cc
and D, we use the right-hand free body in Fig. 8.3d, and when the load is between D
and B, we employ the left-hand free body in Fig. 8.3e.
The influence lines for M, and for V, just to the left of A are obtained by
making use of the free body in Fig. 8.3f. If the unit load is on the free body, V, = 1
and M, is equal to the distance between the unit load and A. When the load is
between A and B, it is evident from a free body of the overhang that both V, and
M,, must be equal to zero.
Example 8.2
For the beam in Fig. 8.4a, construct influence lines for the reactions at A, B, and D,
the moment at B, and the shear at C.
To draw the influence lines for the reactions and V., we make use of the free
bodies to the right and to the left of the hinge shown in Figs. 8.4b and 8.4c. In both
figures the left-hand free body has three unknown forces acting on it. Consequently,
it is necessary, regardless of the position of the load, to analyze the right-hand free
body first.
Let us start by considering the case depicted in Fig. 8.4b, the unit load located
to the left of the hinge. For this condition, moment equilibrium of the right-hand
free body indicates that Rp = V- = 0. Once V¢~ is known to be zero, the left-hand
free body becomes a determinate structure that can be used to evaluate R, and R,z.
With the load between A and B, segment AB behaves like a simply supported beam;
that is, as the load moves from A to B, R, varies from 1 to 0 and R, from 0 to 1.
When the load is just to the left of the hinge, R, and R, take on the values shown in
Fig. 8.4d.
Next we consider the case where the load is to the right of the hinge as shown
in Fig. 8.4c. From the right-hand free body in this figure, it is evident that V. varies
from 1 to 0 and Rp from 0 to 1 as the unit load moves from C to D. Furthermore,
the free body in Fig. 8.4d indicates that R, and R, vary from 0.25 to 0 and from 1.25
to 0, respectively, as V. decreases from 1 to 0.
To construct the influence line for Mz, we proceed as follows. When the unit
load is to the left of the hinge (Fig. 8.4b), it has beea demonstrated that V. = 0.
Consequently, M, = 0 when the load is between A and B, and M, is equal to the
unit load multiplied by its distance to B when the load is between B and C. Finally,
Sec. 8.3 Influence Lines for Beams 195
Ve
Ra Re Rp
(c)
0.25 1.25
(d)
IL for Ra Rite en
ee EEE
0129
(f)
IL for Rp 2
(g)
1.0
IL for Rp
(h)
IL for Mg
(k)
a"
Figure 8.4
196 Influence Lines Chap. 8
when the load is to the right of the hinge (Fig. 8.4c), the magnitude of Mz is two
times that of Vo.
Influence lines for bar forces in trusses are constructed using the same general
procedure that was employed to obtain influence lines for shears and moments in
beams. A unit load is moved along the loaded chord of the truss, and the
resulting value of the bar force is plotted versus the location of the load. One
difference between beams and trusses is that the load can be applied anywhere
along the span of a beam, whereas loads are usually applied to trusses only at
their joints. As shown in Fig. 8.5, the steel or concrete deck of a truss bridge rests
on longitudinal members called stringers. These stringers, in turn, span between
transverse members called floor beams, and the floor beams frame into the joints
of the trusses. Loads applied to the deck of the bridge are thus transferred from
the deck to the stringers, from the stringers to the floor beams, and thence to the
joints of the trusses. As a result, a load applied anywhere along the deck of the
bridge is felt by the trusses only at their joints.
To see how influence lines for bar forces in a truss are constructed, let us
consider the truss in Fig. 8.6a. The load is assumed to be applied to the truss
along the lower chord.
We begin by drawing the influence lines for the reactions at A and G. These
influence lines, shown in Figs. 8.6c and 8.6d, are identical to those for a beam.
Next we construct the influence line for the force in member BC. We can obtain
an expression for Fgc by considering the free body of the truss either to the right
or to the left of a vertical section between joints B and C (Fig. 8.6b).
é Truss
Stringers
Floor beams
Tru§s
Figure 8.5
Sec. 8.4 Influence Lines for Trusses 197
IL for F
ae |
(f)
Figure 8.6
As long as the unit load is to the left of J, the right-hand free body is the
easier of the two to use. Taking moments about J gives
Or
F3c —5.33RG
198 Influence Lines Chap. 8
The negative sign indicates that Fg- is in compression when Rg is positive. Thus
the segment of the influence line for Fg- between A and J, shown in Fig. 8.6e,
is obtained by multiplying the corresponding influence line segment for Rg
Oh? 2 SRE
When the unit load is to the right of J, the left-hand free body in Fig. 8.6b is
used. Summing moments about J of the forces acting on the free body leads to
15 Fc ae 40R, = 0
Or
Fec = —2.6/R,
Using this expression, we construct the segment of the influence line for Fgc
between J and G.
The free bodies in Fig. 8.6b can also be used to draw the influence line for
Fz,. If the unit load is to the left of H, we write an equation of vertical
equilibrium for the right-hand free body. Thus
0.6F:; Ry 0
or
Fey = —1.67Re
This relation leads to the segment of the influence line between A and H shown in
Fig. 8.6f.
When the unit load is between H and J, we can continue to use the
right-hand free body in Fig. 8.6b. However, a vertical downward load that varies
from 0 to 1 as the unit load moves from H to J is now applied at J in addition to
the forces shown in Fig. 8.6b. If the unit load is located at a distance x from H,
vertical equilibrium of the free body gives
0.6Fp, ae Ro = athe = 0
20
Or
x
FpyBI = —1.67Rg GT + —15
The force Fg, thus varies linearly from —1.67(0.167) = —0.278 when the unit load
is at H, to —1.67(0.333) + 1.67 = 1.11 when the unit load is at J.
Finally, when the unit load is to the right of J, we write an equation of
vertical equilibrium for the left-hand free body in Fig. 8.6b and obtain
0.6Fx, = Ra
or
Fp, = 1.67R,
This expression is used to construct the influence line
ségment for Fz, between J
and G.
Sec. 8.4 Influence Lines for Trusses 199
Example 8.3
Construct influence lines for the forces in members AB, BF, and BG, for the truss in
Fig. 8.7a. The loads are applied to the truss at the joints of the lower chord.
As the first step, the influence lines for the reactions at A and E, shown in
Figs. 8.7d and 8.7e, are drawn. Next the influence line for F,, is obtained by
considering the free bodies in Fig. 8.7b. When the unit load is between A and F, F,,
can be determined by taking moments about F of the forces acting on the right-hand
free body. Thus
or
1a = —5.42R,
This expression leads to the influence line segment for F,, between A and F.
When the unit load is to the right of F, we take moments about F of the forces
acting on the left-hand free body in Fig. 8.7b. This leads to
0.83F,,(10) + 15R, = 0
or
Ea =a Sales aie,
Similarly, when the unit load is between G and E, we take moments about J of the
forces acting on the left-hand free body. Thus
0.55Fye(45) — R4(15) 0
or
Fgg = 0.61R,
Using the foregoing results, we construct the influence line segments between A and
F and between G and E as shown in Fig. 8.7h. The last segment of the influence
line, between F and G, is obtained by drawing a straight line between the previously
constructed segments.
The validity of the last step can be checked by placing the unit load halfway
between F and G and summing moments about J for the right-hand free body. This
Influence Lines Chap. 8
200
(h) q
Figure 8.7
Sec. 8.5 Uses of Influence Lines 201
leads to
which checks with the value at the midpoint of the influence line segment between F
and G in Fig. 8.7h.
Having learned how to construct influence lines, we are now ready to make use of
them.
Concentrated Loads
To see how influence lines can be employed to determine the effects of moving
concentrated loads, let us consider the beam in Fig. 8.8a and the accompanying
influence line for the moment at C. The ordinate of the influence line,
corresponding to any point on the beam, gives the value of Mc when a unit load
is placed at that point. Thus a 1k load placed 6ft to the right of A results in a
moment Mc. = 2.4k-ft. Similarly a 15k load placed in the same positions gives
rise to a moment 15 times as large: Mc = 36k-ft.
To obtain the value of Mc- due to several concentrated loads, we multiply
each load by the ordinate of the influence line corresponding to the position of
WA
IL for Mc agai Pee
(b)
10k 6k 8k
fae
fereferefer ey (c)
Figure 8.8
202 Influence Lines Chap. 8
the load and then add the resulting moments. For example, the three loads,
placed as indicated in Fig. 8.8c, give rise to a moment
Me = 10(1.6) + 6(2.4) + 8(1.2) = 40 k-ft
R,= | wy dx
R, = w| y dx (8.3)
|[-s
~~ a +
w kN/m
s
oo
es
Figure 8.9
Sec. 8.5 Uses of Influence Lines 203
If the same load acts over the entire beam, instead of simply from A to B, we
obtain
R, = 4(1.33)(16)(0.5) — 4(0.33)(4)(0.5) = 40 kN
The influence line for R, indicates that loads to the left of B cause an upward
force R,, whereas loads to the right of B result in a downward reaction at A. As
a consequence, the second quantity in the foregoing calculation is subtracted from
the first.
Example 8.4
For the beam in Fig. 8.10a, find the maximum value of M, due to a uniform load of
intensity 5kN/m, which can act over any part of the beam, and two concentrated
loads of 10 kN each, with a fixed distance of 4m between them.
A D B C
ane 7
(a)
g
IL for My
ae 4m 5 kN/m
7 iA
(e)
10 kN 10 kN
5 kN/m 4m
(f)
Figure 8.10
204 Influence Lines Chap. 8
First we construct the influence lines for the reactions and from these the
influence line for Mp shown in Fig. 8.10d. Next we decide where to place the loads
to produce the largest possible value of Mp.
Since loads acting in a region where the influence line is positive have an effect
that is opposite from that produced by loads acting where the influence line is
negative, we must place all the loads either to the left or to the right of B.
The maximum value that M, can attain if the loads are all acting to the left of
B is obtained when the loads are placed as indicated in Fig. 8.10e. This leads to
PROBLEMS
rae.
A Cc B
tlt
8.2. Reactions at A and B, moment and shear at C.
A C B
4 —___*___
1p
ptm fea
8.4. Reactions at A and B, moment and sheanatG-
A Cc B
i
a
Paeiet— te | :
Problems 205
8.5. Reactions at A and B, moment and shear at C, moment at B, shear to the left of B.
A C B
tm,
om
fetal tele tm
8.6. Reactions at A and B, moment and shear at C, moment at A, shear to the right of
A.
A C . B
p— 6 op fp J
8.7. Reactions at A and B, moment at B, C, and D, shear at C and D.
C A D B
BS
8° beg
f+ — 12”—+1-—18’——+}.—10°+
8.8. Reactions at A and B, moments at A, C, and D, shear at C and D.
A Cc B D
oe men
meee 6m SRR aES aay
8.9. Reactions at A, B, and C, as load moves from D to C along member DEC.
Ne
D
er ar
eee
6'
:4
(e
papa
7, 7
A Hinge B Cc
A D B E Cc
A Hinge B C_ Hinge D
ae ee
8.14. Reactions at A and B, moment at B and C, shear at D.
A B C D
ts Be OD OD
30m 15m ,;10m,10m Se 30m i
A
[+6 @ 20’ = 120’ —_—_+|
8.19. Determine the maximum value of Mc due to two 20k concentrated loads that must
remain 4 ft apart but can be placed anywhere along the span.
A Cc B
+ 10’ >}<—8' +. i
8.20. Determine the maximum value of the reaction at A and the moment at B due to a
uniformly distributed load of 10kN/m that can be placed over any portion or
portions of the span.
A B
— ack
on 8m >< 6m >
8.21. Determine the maximum value of the reaction at A and the moment and shear at C
due to a uniformly distributed load of 2 k/ft and a concentrated load of 20k, both of
which can be placed anywhere along the span.
A C B
Be Se ais
m4
soe
7 _ i _ ip 7 =
2, eb A dial ea sat
te
* a ~~—
—
ae:
_
_ ~
LP BR Mei lae
ke ts wiees
. | >
lh
A etn
me —— ay » i
3 fip 2 » “ *y > ee
Wt
<< “ye ¥ 9 | |
oe
—s . S58 :
: 34, MemSoah ARoh Uae Bs oa + Ae
7
ce Bie
' A” :
- ime : 7 _ 7
e'ol4 wi @ 18 narnia a! . s(el7 =) ee | maT.
ic me |
45. Gow tortie
RA ve wo ae
7 = ; : a
. — ae af i ae e : ey
Mines 7 rs ee
6 Vewls Due T-SrrNEAD)
bedot9 to bay Dee cao Wot tit0
i c , ae r . pies F R
fae dn -~! oe
| oe a -s *
oe, Ray nes AB, Lk BM a en hee |
yw
Arches and Cables
The arch is a structural system in which the primary internal force is axial
compression. Most arches do develop some bending in addition to the compres-
sion. However, the bending moments that a given set of loads produces in an arch
are much smaller than those produced in a beam of the same length. As a
consequence, arches are able to span much larger distances than beams.
Unlike a beam, which develops only vertical reactions at its supports, the
arch gives rise to both horizontal thrusts and vertical forces at its supports. In
fact, it is the horizontal reactions that are responsible for the relatively small
bending moments in an arch. Without these forces the arch would simply be a
curved beam whose moments would be identical to those of an equivalent straight
beam. Thus the moment at midspan for both the curved and straight beams in
Fig. 9.1 is equal to 500 k-ft. By comparison, the moment at the middle of the arch
is only 158 k-ft.
The arch in Fig. 9.2a has two hinges, one at each support, and is referred to
Straight beam Curved beam Arch
10 k 10k
5k
500 k-ft 500 k-ft 158 k-ft
Figure 9.1
U2
fe2
Two-hinged arch Three-hinged arch
(a) (b)
Figure 9.2
4
210
Sec. 9.1 Arches 211
as a two-hinged arch. This system has more unknown reactions than equations of
equilibrium and is consequently indeterminate. In Chapter 12 after methods for
analyzing indeterminate structures have been introduced, we will consider
two-hinged arches. However, for the present we will restrict our considerations to
three-hinged arches, which are determinate. As shown in Fig. 9.2b, a three-
hinged arch has a hinge at its crown in addition to those at the supports.
Example 9.1
Determine the reactions for the arch in Fig. 9.3a.
If the supports are at the same elevation, the vertical reactions are
determined using the same procedure as was employed for beams. Summing
moments about B for the free body in Fig. 9.3a gives
V,(40) — 100(30) — 60(12) = 0
V, = 93 KN
and from vertical equilibrium
pz = 160
— 93 = 67KN
To determine the horizontal reaction, we sum moments about C for the free body in
Fig. 9.3b. Thus
100 KN 60 KN 100 kN
Figure 9.3
212 Arches and Cables Chap. 9
Example 9.2
Determine the reactions and construct the moment diagram for the arch in Fig. 9.4.
Since the supports are not at the same elevation, the reactions cannot be
determined without solving simultaneous equations. Taking moments about B in
Fig. 9.4a and moments about C in Fig. 9.4b gives
V,(100) — H,(5) — 40(80) — 30(28) = 0
and
V,(50) — H,,(20) — 40(30) = 0
from which
H, = 46.9k, V, = 42.8k
30k
H Vv
A 20' ol. 30' 22' 28' s
Va
Kk 100’
(a)
Ho
20
(b)
7 201
C
E
D
B
A
(c) @
Figure 9.4
Sec. 9.2 Cables 213
The remaining two reactions are determined using equations of equilibrium for the
entire arch. Thus
Vz = 27.2k, Hz = 46.9k
The moment diagram for the arch is shown in Fig. 9.4c.
9.2 CABLES
The cable is a structural element that resists forces by developing tension stresses.
Because of its flexibility, it is unable to develop either compression or bending
stresses. When subjected to transverse loads, a cable adjusts its shape to the loads
in such a manner that tension forces are sufficient to resist the loads. Concen-
trated loads cause the cable to take on a shape consisting of a series of linear
segments, and when subjected to a distributed load, the cable becomes a curve. If
the load is uniformly distributed in the horizontal direction, the curve is a
parabola.
Example 9.3
Determine the reactions and the tension in each segment for the cable in Fig. 9.5a.
The reactions are determined using the same procedure as was employed in
analyzing three-hinged arches. Summing moments about E for the entire system
gives
60V,, — 12(50) + 12(25) + 8(15) 0
Va 17kN
H, = H, = 24.6kN
The tension in each segment of the cable is determined using free bodies of
the individual segments as shown in Fig. 9.5c. Thus
(b)
246
Problems 215
The preceding calculations indicate that the horizontal component of the cable
tension is constant throughout, and that the maximum cable tension therefore occurs
in the segment with the largest vertical component, that is, the one with the largest
slope.
PROBLEMS
9.1 to 9.3. Determine the reactions and sketch the moment diagram.
9.1. 40 kN 60 kN
At 40 kN Fines +
ay
9.2. 40 k 20 k
9.4 and 9.5. Determine the reactions and the indicated bar forces.
216 Arches and Cables Chap.9
30 k 40k
8 O10 = 50 a
9.6 and 9.7. Determine the reactions and the tension in each segment of the cable.
9.6. 4k
| 7 Y )
bes |e 10’ (A and D are at same level)
A B (C
D
K 16’ ah 26' oe 12>
9.7. 20 kN
50 kN i
ak :
Re r p 15m (D is 5m above A)
A
10
Indeterminate Structures:
Introduction
HRRNERSORURS
a=
a :
=
mtsy
DA =0
Fig sinaw = Fgcsina (10.1)
Fag = Fec
Fi =0
2 4 (10.2)
Fiz COS @ + Fec cos aw= P
Combining Egs. (10.1) and (10.2) gives
Figure 10.1
218
Indeterminate Structures: Introduction 219
Figure 10.2
It is obvious from these results that the two-bar system in Fig. 10.1 is a
determinate structure that can be analyzed completely using only the equations of
equilibrium.
We now turn our attention to the three-bar system in Fig. 10.2. As before,
we write equations of vertical and horizontal equilibrium for joint B. Thus
> E=0
Fiz sin @ = Fecsin a (10.3)
Fag = Fec
25 = 0 (10.4)
Fy, CoS @ + Fgc cosa + Fyp = P
For this structure the available equations of equilibrium are not sufficient to
obtain the force in each bar, and the structure is therefore said to be
indeterminate. To calculate the bar forces, we must consider the deformation of
the structure as well as the conditions of equilibrium.
If the bars deform as indicated by the dashed lines in the figure, then 6,¢,
the elongation of bar BC, is related to é6gp, the elongation of member BD, by
Onc = Opp COS @ (10.6)
220 Indeterminate Structures: Introduction Chap. 10
In writing this equation we have assumed that the angle w between bars BC and
BD does not change significantly as the structure deforms. This is true as long as
the elongations of the bars are small compared to the original lengths of the
members.
To solve Eqs. (10.5) and (10.6) simultaneously, it is necessary to relate the
bar elongations to the bar forces. To this end, we introduce the force-deformation
relation for axially loaded bars:
6= Hi (10.7)
AE
Since Eq. (10.7) is valid only as long as Hooke’s law is satisfied, we have now
introduced a second assumption, that the material obeys Hooke’s law. For most
engineering materials this is equivalent to stating that the material remains
elastic.
Substitution of Eq. (10.7) into Eq. (10.6) gives
Fgcl_ _ Fgpl cos’ w
ApgcEsgc AgpEsp
or if all bars are made of the same material and have equal areas,
Fe quasi COS. (10.8)
Combining this expression with Eq. (10.5), we obtain for the bar forces
te
Fa os ete
os
oe vail bcd COS a0?)
and
P cos* aw
(10.10)
Eoa = foc = 1+ 2cos'a
The preceding example demonstrates clearly that, whereas the equations of
equilibrium suffice to determine the internal forces and reactions of a determinate
structure, they do not suffice when calculating these quantities in an indetermin-
ate structure. For the latter it is necessary to consider the deformations of the
structure in addition to the equations of equilibrium.
11
Method of Consistent
Deformations
(a)
eal
=>
NIT
+.
TS
wy
(b)
Figure 11.1
222
Sec. 11.2 Basic Principles 223
4 = ih yy,
Y ee ge jr Y
othe Spt
| 16
Re {
Pp
Figure 11.2
224 Method of Consistent Deformations Chap. 11
necessary to support the structure in a stable manner. For example, the cantilever
beam that results if the vertical restraint R,, at the right end of the beam, is
removed suffices to support the load. Similarly, the simply supported beam
obtained by removing the moment restraint M, at the left end of the beam is also
adequate for supporting the load. Restraints that can be removed without
impairing the load-supporting capacity of the structure are referred to as
redundants. In general, the number of redundants that a structure possesses is
equal to the degree of indeterminacy. The beam in Fig. 11.2a can thus be said to
have one redundant or to be indeterminate to the first degree.
Having determined how many redundants a structure possesses, the next
step is to decide which reaction is to be considered the redundant and to remove
this restraint, thus forming a determinate structure. Any one of the reactions may
be chosen to be the redundant provided that a stable structure remains after the
removal of that reaction. For example, let us take the reaction Rg as the
redundant. The determinate structure obtained by removing this restraint is the
cantilever beam shown in Fig. 11.2b. We denote the deflection at end B of this
beam, due to P, by Agp. The first subscript indicates that the deflection is
measured at B and the second subscript that the deflection is due to the applied
load P. Using the moment-area method, it can be shown that Agp = 5PL?/48EI.
Next we apply the redundant R, to the determinate cantilever beam, as
shown in Fig. 11.2c. This gives rise to a deflection Agr at point B, whose
magnitude can be shown to be R,L?/3EI.
In the actual indeterminate structure, which is subjected to the combined
effects of the load P and the redundant Rz,, the deflection at B is zero. Hence the
algebraic sum of the deflection Agp in Fig. 11.2b and the deflection Age in
Fig. 11.2c must also vanish. Assuming downward deflections to be positive, we
can write
App = Apr = 0 (11.1)
or
SPL eR si x
48EI 3EI ~
from which
pers)
aceite
Equation (11.1), which is used to solve for the redundant, is referred to as an
equation of consistent deformations.
Once the redundant R, has been evaluated, we can determine the
remaining reactions by applying the equations of equilibrium to the structure
in
Fig. 11.2a. Thus ¥ F, = 0 leads to
5 Tree
Ra Pi pee
: 16 te
Sec. 11.3 Application of Consistent Deformations to Structures 225
and )) M, = 0 gives
(2) begets 3
Mie ee Pee PT
pee a tt
A free body of the beam, showing all the forces acting on it, is shown in Fig.
12d:
The method of consistent deformations, which has been presented here, can
be summarized as follows. In the indeterminate structure (Fig. 11.2a) the
deflection at point B is zero. When we remove the redundant R, to obtain a
determinate structure, we allow point B to deflect. We then ask: What value must
the redundant Rz have so that its application to the determinate structure will
cause point B to return to its original position of zero deflection? Viewed from a
slightly different perspective, the method of consistent deformations can be
considered to be an application of the principle of superposition. The structure in
Fig. 11.2a can be assumed to be a determinate cantilever beam loaded with a
known load P and an unknown load Rg. Since the deflection at B due to these
two loads acting simultaneously is zero, the sum of the deflections that these loads
produce at B when permitted to act one at a time must also vanish.
The individual steps in the method of consistent deformations can be
summarized as follows:
Example 11.1
Determine the reactions for the beam in Fig. 11.3a and draw its shear and moment
diagrams. _
As shown in the figure, the beam possesses three reactions, which is one more
226 Method of Consistent Deformations Chap. 11
(d)
Figure 11.3
The deflections that appear in Eq. (11.2) can be determined using one of the
methods presented in Chapters 6 and 7. In this instance we will use the moment-area
method. Thus
A = —
PES — 4+ —=
eRPIS
2 TPL? —- =
oe QE 2a RIO Shen o kh)
and
_ Rslk2L | Rgl?
seh? te4i) SP ee ae DO |
With these results, Eq. (11.2) becomes
gts GAO
ET SET —
from which
7P
Rz — Tr
The remaining reactions are obtained by applying the equations of vertical and
moment equilibrium to the free body in Fig. 11.3a. Thus
3 Lage
Ries — =P: M
<4 4 fh ee
A free body of the beam and its moment diagram are shown in Fig. 11.3d.
Example 11.2
When analyzing a structure by the method of consistent deformations, we usually
have a choice regarding the selection of the redundant. To illustrate this point, let us
reanalyze the beam considered in Example 11.1, this time choosing M, as the
redundant instead of Rz. The beam is shown in Fig. 11.4a.
The determinate structure obtained by removing the moment restraint at A is
a simply supported beam with an overhang on its right side. To this structure we
apply separately the applied load P, as shown in Fig. 11.4b, and the redundant M,,
as indicated in Fig. 11.4c. Since the slope at A is zero for the beam in Fig. 11.4a, the
algebraic sum of the slopes at A, for the beams in Figs. 11.4b and 11.4c, must also
vanish. Thus
—O4p + Oar = 9 (11.3)
Using the moment-area method to determine the angles, we obtain for the beam in
Fig. 11.4b
_ PELL J Ph
* 2EI23 125i
Oh dad
Bap =
DE
Method of Consistent Deformations Chap. 11
228
(c)
Figure 11.4
Example 11.3
It is required to determine the reactions and to draw the bending-moment diagram
for the beam in Fig. 11.5.
First we choose R, as the redundant. Then we apply separately to the
determinate beam, obtained by removing R,, the 15k load and the redundant R,
(Figs. 11.5b and 11.5c). Since the deflection at B is zero in the actual indeterminate
15k
37 k-ft
(d)
Figure 11.5
230 Method of Consistent Deformations Chap. 11
structure,
Ae -_ Nae = 0 (11.4)
The above deflections are determined using the conjugate beams shown in Figs.
11.5b and 11.5c. Thus
~ 5833
Bea ET
and
_ 444.4R,
BReaw EI
5 kN/m
Paes Fo ine
(a)
5 kN/m
Neal bes25 KN
(b)
Figure 11.6
Sec. 11.3 Application of Consistent Deformations to Structures 231
(c) (d)
Element x=0 Mp Mp m
AB A 0 —H, x ae
(e)
5 kN/m
8.34kN—>
43.4 kN-m
41.7 kN-m
(f)
5 kN/m
—~—-
6.24 KN
21.88 kN
6.24 kN—>—
28.12 kN
47.9 kN-m
31.2 kN-m
31.2 kN-m
(g)
Figure 11.6 (continued)
As shown in the figure, the frame has four reactions. The structure is thus
indeterminate to the first degree and has one redundant reaction. Let us choose H,
to be the redundant. To the determinate structure obtained by removing H,, we
apply separately the uniformly distributed load, as indicated in Fig. 11.6b, and the
redundant H,, as shown in Fig. 11.6c. Since the horizontal deflection at A is zero in
the actual structure, the sum of A,p and A,p must also vanish. Thus
[Nips + Ne = 0 (11.5)
The method of virtual work will be used to calculate both the magnitudes and the
directions of the deflections. If we designate Mp as the moment due to the
distributed load, Mg as the moment caused by the redundant H,, and m as the
moment due to a unit horizontal load at A (Fig. 11.6d), then
Mpm dx
A AP
= EI
and
_ { Mem dx
AR EI
&
Substitution of the moments listed in the table in Fig. 11.6e into the relations above
Sec. 11.3 Application of Consistent Deformations to Structures 233
leads to
—1042 + 125H, = 0
and
H, = 8.34kN
A moment diagram of the frame is shown in Fig. 11.6f.
In the foregoing analysis it was assumed that both members of the frame had
the same stiffness EJ. Let us now repeat the analysis of the frame assuming this time
that the stiffness of member BC remains equal to E/ but that the stiffness of member
AB is reduced to 0.5EI.
Since A,» depends only on the bending of member BC, whose stiffness is the
same as it was before,
_ -1042
AP EI
However, A,4z depends on the bending of both members AB and BC, and its value
will now be given by
UH ax [ee st OLS gX (= 5 U0 SX aN
NAR — S5 Min 2% kc LL... sae
po OSET I, EI
_ 167H,
AR, a EI
Substituting the new values of the deflections into Eq. (11.5) gives
—1042 + 167H, = 0
from which
H, = 6.24kN
The moment diagram corresponding to the new solution is given in Fig. 11.6g.
The only facts that mattered were that EJ had the same value for both members
in the first part of the analysis and that EJ of member BC was twice as large as
the EI of member AB in the second part of the analysis.
Example 11.5
In each of the previous problems we considered elastic structures restrained by rigid
supports. In other words, while the structure itself was able to deform, the supports
remained stationary. Let us now analyze the beam in Fig. 11.7a, which is fixed at its
left and elastically restrained at point B by a cable.
The beam is indeterminate to the first degree, and we will let T, the force in
the cable, be the redundant. Removal of the cable leads to a determinate cantilever
beam, to which we apply separately the 5kN force as shown in Fig. 11.7b and the
redundant T as shown in Fig. 11.7c. In Example 11.1 we considered a beam similar
“ie
Cable ————_______»
A = 20 mm?
E = 200 X 10® kN/m?
Beam 5kN
(a)
5 kN
_— —
-— _ ~ _
(b)
(c) Q
Figure 11.7
Sec. 11.4 Support Settlement 235
to the one being analyzed here, except that the beam rested on an immovable
support at B. As a consequence, the sum of the deflections A,p and Apr Was equal
to zero. However, in the present case point B moves down by an amount equal to
the elongation of the cable. Hence
TL
Asp — Agr = AR (11.6)
630 pee
BP ise EI , BR EI
Substitution of these expressions together with values for A, E, and / into Eq. (11.6)
gives
Ge0 3 10°C. 6727 & 103 ie Lae 10° 117
$00 10% 100.5610° eae 20 aia)
from which
T = 5.16kN
If the support at B had been rigid, the right-hand side of Eq. (11.7) would
have been zero and the reaction at B would have had the value
Rg = 8.75 kN
In general, the problem of support settlement is far more serious for indetermin-
ate structures than it is for determinate ones. If one of the supports of a simply
supported beam settles by a small amount, no major changes occur in the
external or internal forces acting on the member. However, if one of the supports
of a multispan beam settles a small amount, significant changes will occur in both
the reactions and the bending moments. To see how the method of consistent
deformations can be used to analyze a structure with support settlements, let us
consider the following example.
Example 11.6
It is desired to determine the reactions and to draw the moment diagram for the
beam in Fig. 11.8 assuming that the center support settles 2 in., and to compare the
results thus obtained with those corresponding to zero settlement.
Let us choose the reaction at B as the redundant. To the determinate structure
obtained by removing the redundant, we apply first the distributed load as indicated
in Fig. 11.8b and then reaction Rg, as shown in Fig. 11.8c. In the absence of
settlement at B,
Asp — Apr = 0 (11.8)
236 Method of Consistent Deformations Chap. 11
1 k/ft
I = 500 in.4
E = 30 X 10° ksi
Zero
settlement
2” settlement
14k
Figure 11.8
Sec. 11.5 Structures with Several Redundants 237
The moment diagrams corresponding to the two solutions are given in Fig. 11.8d.
Comparison of the two diagrams indicates that the maximum moment increases from
50 to 98k-ft as a result of the support settlement. It is thus evident that support
settlement can give rise to significant increases in stress in an indeterminate
structure.
(e)
Figure 11.9
we let
Equation (11.12) states that the deflection at i due to a load at j is equal to the
deflection at 7 due to a unit load at j multiplied by the value of the load at j.
Making use of the above notation, Eqs. (11.10) and (11.11) can be rewritten
in the form
Aip = Oni aaa 6,.R> = 0 (11513)
In this form, the equations allow us to distinguish clearly the known coefficients
from the unknown variables.
Let us now consider some examples of structures with several redundants.
Example 11.7
It is required to find the reactions and to draw the moment diagram for the beam in
Fig. 11.10a.
Since the structure is indeterminate to the second degree, we must remove two
restraints to form a determinate structure. Let us choose the reactions at B and C as
the two redundants and designate them as R, and R,. We then apply to the
determinate structure the known loads and each of the redundants, one at a time, as
shown in Figs. 11.10b, 11.10c, and 11.10d. Since the deflections are zero at B and C
in the original beam,
0.264PL* 0.215PL?
TP he SS) Se
EI EI
0.444L? 0.389L*
On = Oy = 7 Op = db = EI
Substituting these values into Eqs. (11.15) and solving the resulting equations gives
R, = 0.725P, R, = —0.15P
A free body of the beam and its moment diagram are shown in Fig. 11.10e.
240 Method of Consistent Deformations Chap. 11
(c)
Beers
pete
542 Ro
a
by Ro Pies
Ve Y
Ro
(d)
0.2 PL
ial » 0.025 PL
OF
(e) q
Figure 11.10
Sec. 11.5 Structures with Several Redundants 241
Example 11.8
It is required to determine the reactions and to draw the bending-moment diagram
for the frame in Fig. 11.11a.
Let us choose the horizontal and vertical reactions at A as the redundants and
label them R, and R,, respectively. We then apply to the determinate structure,
obtained by removing the redundants, the known distributed load and the unknown
(c) (d)
Segment x=Oat Mp m, Mo
AB A 0 x 0
BC B xe 10 x
Note: Moments are positive when they
produce compression on the outside of
the frame
(e)
Figure 11.11
242 Method of Consistent Deformations Chap. 11
11.3 k-ft
2 k-ft
~< 0.72k
7.2 k-ft
“F942 k-ft
{ 21.2 k-ft
11.4k
(f)
redundants one at a time. The horizontal and vertical deflections at A due to these
three force systems must vanish. Thus
We will calculate the necessary deflections using the method of virtual work.
Although for this structure it is possible to determine the direction of the deflections
by visual inspection, we will not do this. Instead, we will apply a more general
procedure. We will initially assume all deflections to be in the same directions, that
is, the directions in which the redundants R, and R, have been assumed to act.
Accordingly, we apply the unit loads to the left and upward as indicated in Figs.
11.11c and 11.11d. Expressions for the bending moments due to the distributed load
and due to the unit loads are given in the table in Fig. 11.1le. They are designated
as Mp, m,, and mz, respectively. The moments are assumed to be positive when
they produce compression on the outside of the frame.
Making use of the expressions in the table, we obtain
sd dx | 3333
Aip = = | Mem: ax = iF———————_ = Er
a ae _ 2500
aN ys — = | Mma de = {= alee
1 10 2 dy 100dx 1333
2 alm se ae 2 EI
é one cet = | q“-—- =
MM sede at
on = epee: oe
| at
0» ws 2
Sec. 11.6 Structures with Internal Redundants 243
The negative signs of A,p and A, indicate that the deflections are to the right and
downward, respectively. These terms will therefore appear as negative quantities in
Eqs. (11.16). By comparison, the remaining deflections are to the left and upward
and appear as positive terms in the equations. Thus
Example 11.9
To see how an internally indeterminate structure is analyzed, let us consider the
truss in Fig. 11.12a. It is assumed that EA is constant for all bars.
The truss has three reactions, which can readily be calculated using equations
of equilibrium. However, if we then attempt to determine the bar forces, it soon
becomes apparent that these cannot be obtained using only conditions of equi-
librium. For example, if one applies the method of joints, it is evident that there
exist at every joint three unknowns but only two equations of equilibrium. Since
removal of any one of the bars changes the truss from an indeterminate structure to
a determinate one, we conclude that the structure is indeterminate to the first
degree.
Let us designate the force in bar 6 as the redundant and obtain a determinate
structure by cutting this bar. To the determinate structure thus formed, we apply the
known load of 40 kN as shown in Fig. 11.12b and the unknown redundant R, as
indicated in Fig. 11.12c. Since in the actual structure there exists neither a gap nor
an overlap in member 6, the sum of A¢p and A,z must vanish. Thus
Acp + Acre = 0
where 6,, is the deformation due to a unit force in bar 6 as shown in Fig. 11.12d.
Method of Consistent Deformations Chap. 11
244
2 40 kN
i
30 kN 30 kN 30 kN 30 KN
(b)
(c) (d)
Figure 11.12
To calculate the deformations appearing in the above equation, we will use the
method of virtual work. The data necessary to obtain the required deformations are
listed in Table 11.1. The symbols F and f represent the bar forces due to the applied
load and the unit load in bar 6, respectively. Using the data in the table, we obtain
F{L
as ys —= =i 1120
OP
Ale A aA
0) 34.56
be = ae +a
Substituting these values into Eq. (11.17) gives
—1120 + 34.56R = 0
from which
R = 32.41 kN
The positive sign of R indicates that the direction astintd for the redundant, that is,
tension, was correct.
Sec. 11.6 Structures with Internal Redundants 245
TABLE 11.1
F if ih
Member (KN) (KN) (m) FfL fPL
1 30 —0.6 6 —108 2.16
2 40 —0.8 8 250 2
3 0 —0.6 6 0 2.16
4 40 —0.8 8 —256 5.12
5 50 1.0 10 —500 10.0
6 0 1.0 10 0 10.0
—1120 34.56
Once the force in bar 6 is known, the method of joints allows us to calculate
the forces in the remaining bars. The results of these calculations are listed in
Table 11.2.
TABLE 11.2
Bar force
Member (KN)
1 10.55 (T)
a 14.07 (T)
3 19.45 (C)
4 14.07 (T)
5) 17.58 (C)
6 32.41 (T)
PROBLEMS
11.1 to 11.18. Solve for the reactions and draw the bending-moment diagram.
4
11.1. w
A G
B
$1 |
[le
= constant
11.2. 20 kN 10 kN
y 2m
A (&
I = constant
,
246 Method of Consistent Deformations Chap. 11
11.3. Analyze the beam in the figure letting (a) R, be the redundant (b) M, be the
redundant.
w
A G
B
2
I = constant
L
11.4. Analyze the beam in the figure letting (a) Rz be the redundant (b) M, be the
redundant.
10k 5k
|} 10°—+}——12.—+-|.— ¢—+|
I = constant
11.5. 1k/ft
A C
B
|-—— 2 ——+|.—.
I = constant.
11.6. Ww
pe
I = constant
11.7. 10 kN/m
+10 +} en — ‘
I = constant
Problems 247
11.8. 10 k 20 k
A B Cc
ela Sew
“if
I = constant
11.9. 40 kN
20 kN/m
A B C
I = constant
B C D
11.11. 2k/ft
11.12. 40 kN
Vi
I = constant
248 Method of Consistent Deformations Chap. 11
11.13. 10kN/m
——
wo =
pe ee Bs
I = constant
11.14. Analyze the frame in the figure assuming that (a) all members have the same J; (b)
Ins = Icp = 2c.
1.5 k/ft
11.15.
Cable 3m
pee
pm ote am ‘
Tyeam = 100 x 106 mm4, Acayie = 150 mm?, E = 200 X 10° kKN/m?
Problems 249
11.16.
" B
/-—10' ane 10’ ~
11.17.
20 k 20 k
}— 5’ +L — 5° +L — 5° +L. 5
Ioeam = 144inL*, Avo. = OT in?
11.18. 20kN/m
|} tom ———-
Ie 4= 400 x 10°mm*, A!,.. = 20 mm*
11.19. Find the reactions assuming that the support at C settles 0.4in. E = 30 x 10° ksi,
I = 1000in.".
250 Method of Consistent Deformations Chap. 11
11.20. Assuming that the stiffness of the support at C is 4k/in., find the reactions and
determine the settlement at C. E = 30 x 10°ksi, J = 1000 in.*.
20k
“ 10’ ——+|«——10'——->
11.21 to 11.27. Solve for the reactions and draw the bending-moment diagram.
11.21.
I = constant
11.22.
I = constant
11.23. 20 kN/m
A D
YZ, Me pu Ve
x 6m + 6m >t. 10m ~
I = constant
11.24. 5k 10k
“?
k—4m —+}. 6m - q
I = constant
Problems 251
11.26. 10 kKN/m
Pe 4 “/, /.
«10 m>{+10 m>t+10 m>{+10 splbw: m>
I = constant
11.27. 40 kN
fo) S)
b— 4m —t— 4m --
D V——» 30 kN
|
100 KN
A = constant
A = constant
252 Method of Consistent Deformations Chap. 11
VY 7
verge
We will derive Castigliano’s theorems in two stages. First we will formulate the
law of reciprocal deformations, and then, using this principle, we will derive
Castigliano’s theorems.
To begin, let us consider the beam in Fig. 12.1a. Because of the load F, the
Figure 12.1
254
Secii2:2 Derivation of Castigliano’s Theorems 255
A, = OF,
A = 6o1F,
where 6,, and 65, are the deflections at 1 and 2 due to a unit load at 1. If, instead
of applying a load at 1, we apply a load at 2 as shown in Fig. 12.1b, the
deflections at 1 and 2 can be written as
A, = 6p
Aa = OnF
Using this notation, let us now formulate an expression for the work due to
F, and &. It is assumed that F, and F are applied one at a time and that the
forces are applied gradually, so that both the forces and the deflections increase
simultaneously from zero to their final values. If F, is applied first, the amount of
work performed is
1
dW, = 0)F,(6,,/)
Next we apply & to the beam on which F, is already acting. The additional work
resulting from the application of F, consists of two parts. First F, does an amount
of work at point 2 equal to
1
dW, = 5 Fa S22F2)
In addition, the force F,, already on the beam, moves through the displacement
that F, causes at point 1. Thus
dW, = F,(6,2.F)
The 4 factor is absent in this expression because F, remains constant at its full
value during the entire displacement.
The total work due to F, and F, is obtained by adding the increments of
work due to the individual forces. Thus
1 1
WwW tS 5 uF i + 5 OnaF 2 te ODA EE (1251)
Let us now determine the work if the loading process is reversed. That is,
is applied first and F, is then applied to the beam on which & is already acting. In
this case the work due to F, is
1
dW, = 5 Fa OnaF2)
Method of Least Work Chap. 12
256
1
dW, = 5 F(OisFi) + F,(62,F)
Although it has not been demonstrated here, the theorem applies to moments
and rotations as well as to forces and displacements. Thus the rotation at 1 due to
a unit force at 2 would be equal to the deflection at 2 due to a unit couple at 1.
Having derived the law of reciprocal displacements, we are now in a
position to derive Castigliano’s theorems. With that end in mind, let us redirect
our attention to the beam we have been considering and this time apply the forces
F, and F, simultaneously as shown in Fig. 12.1c. The resulting work is given by
1
Wo, Ai +B As) (12.4)
where
A, = 61% + 62h
(12.5)
Az = 62%, + 62F
Substitution of the expressions in Eqs. (12.5) into Eq. (12.4) gives
1
W= 5 (uF i a Oph F =: Onk ar OoF3)
Since the strain energy U stored in a deformed structure is equal to the work
performed by the external loads, we can also write
1 &
U= 5 (OuFi + 62KF, + 62,45 + 62F5)
Sec. 12.2 Derivation of Castigliano’s Theorems 257
In view of the law of reciprocal deformations given by Eq. (12.3), this expression
for U can be simplified to
1
= 5 (OuFi + 26,F,5 + 5oF3)
ee A
re (12.6)
12.6
Theorem 5. MRedundants must have a value that will make the strain
energy in the structure a minimum.
i Py
Figure 12.2
258 Method of Least Work Chap. 12
vz
Pd PL
— = SS S= 5
2 2 2AE 229)
To see how the method of least work is applied to various structures, let us
consider several examples.
Example 12.1
It is required to determine the reactions for the beam shown in Fig. 12.3a.
Since the beam has three reactions but only two equations of equilibrium are
available to evaluate the reactions, the beam is indeterminate to the first degree and
possesses one redundant reaction. Let us choose R, as the redundant and determine
its value by minimizing the strain energy in the beam with respect to R,. That is.
aU oa;
(12.10)
OR,
Since the strain energy in a beam is usually assumed to consist only of bending
energy, it can be written as
Nii? dx
o 2ET
(12.11)
To solve Eq. (12.10), we would ordinarily carry out the integration indicated by Eq.
(12.11) first and then differentiate the resulting expression in accordance with
EG
(12.10). However, it is customary to reverse these procedures, that is to differentia
te
first and then integrate. Thus
It is evident from Eq. (12.12) that by differentiating first and then integrating we
avoid squaring the moment in the strain energy term and thus
reduce the numerical
work significantly. The process of reversing the order of
differentiation and
Sec. 12.3 Application of the Method of Least Work 259
(a)
EI ement =O
x= M “
aR,
AB A RaX x
L
BC A Rax-P (x- >} x
(b)
Pp
Sap
16
16 16
(c)
Figure 12.3
integration used here can only be applied when the limits of integration are constant,
as they are in our case.
Figure 12.3b lists expressions for the moment and its derivative needed in the
solution of Eq. (12.12). It should be noted that the moment must be expressed as a
function of the redundant R, and that this is most easily accomplished by taking the
origin of coordinates at A.
Substitution of the expressions in the table into Eq. (12.12) gives
ey 2 d pe PL
Example 12.2
Determine the reactions for the beam shown in Fig. 12.4a.
The beam is indeterminate to the first degree, and we will designate R, to be
the redundant. The condition for determining R, is
fe}
OU = i“M OR
M(OM/0Ra) |, = (2513)
OR, 0 EI
To solve Eq. (12.13), the moment in the beam must be expressed as a function of
R,. Accordingly, it is necessary to express either R- or Rp in terms of R,. Since we
will use only R, and Rp in formulating the moment, we will determine Rp as a
function of R,. Taking moments about C, we obtain
(a)
Ele men t CS 0 M aM
—
dR
AB A Rak x
BC A Rx = 10{x= 10) x
CD D 2Rioo xe ROY
(b)
wre 2 k/ft
3.33 k 20 k 6.67 k
(c) Q
Figure 12.4
Sec. 12.3 Application of the Method of Least Work 261
Figure 12.4b lists the expressions for the moment and its derivatives needed in
the solution of Eq. (12.13). The origins used in the table were chosen so that the
expressions for the moment are as simple as possible. Substitution of these
expressions into Eq. (12.13) gives
10 20 10
from which
Rae 3.335:k
Example 12.3
It is required to determine the reactions for the frame shown in Fig. 12.5a. Assume
that EJ is constant.
Since the structure is indeterminate to the second degree, it has two redundant
reactions. Let us choose R, and R,, the reactions at A, to be the redundants. The
conditions for evaluating R, and R, are
ls jSA a4 (12.14)
OR, EI
OR, EI
The expressions for the moment and its derivatives needed to solve Eqs. (12.14) and
(12.15) are listed in Fig. 12.56. Moments are considered positive if they cause
compression on the outside of the frame.
Substitution of the expressions in Fig. 12.5b into Eqs. (12.14) and (12.15)
gives
5 10
5 5
0 0
10
from which
A free body of the frame showing all the reactions appears in Fig. 12.5c.
’
262 Method of Least Work Chap. 12
(a)
AB A -Rx 0 -x
BC B R,x—5R, x —5
cD B R,x-5R,-40x+200 x -5
(b)
40 kN
60 kN-m
6 kN
23 KN
6 kN
17 KN
(c)
Figure 12.5
Example 12.4
Determine the tension in the wires that support the beam shown in Fig. 12.6a. Each
wire has a cross-sectional area of 0.1in.?, J = 288 in.“gfor the beam, and E =
30 x 10° psi for both the wires and the beam.
The structure is indeterminate to the first degree, and we will choose
T,, the
Sec. 12.3 Application of the Method of Least Work 263
10
(b)
aM
Element x=0 M ae
oT,
AB A ASS x
BC Cc aes x
(c)
Figure 12.6
force in the left-hand wire, to be the redundant. From equilibrium it is apparent that
The strain energy of the system consists of the energy of axial deformation in
the wires and the bending energy in the beam. Thus
Example 12.5
Determine the bar forces for the truss shown in Fig. 12.7a.
Since the truss possesses one more bar than is necessary to resist the applied
load in a stable manner, it is indeterminate to the first degree. Let us choose the
force in bar AB as the redundant and designate it as R. Accordingly,
a 0 (12.18)
OR
Bar areas
AB 20 cm?
AC 10 cm?
AD 10 cm?
E = constant
(a)
AB 20 10 R 1 — 0.50R
—235.1+5.31R
(b) @
Figure 12.7
Sec. 12.3 Application of the Method of Least Work 265
or (12.19)
where S is the force in any bar and the summation is carried out over all the bars.
Substitution of Eq. (12.19) into Eq. (12.18) gives
eM SIeE
= =} (12.20)
Just as it was necessary to express the moment in a beam in terms of the redundant,
we must now express the bar forces S in terms of the redundant R. Thus vertical
equilibrium of joint A gives
Sap = 1.34R — 89.4
and from horizontal equilibrium of joint A we obtain
Sac = 80.0 — 2.0R
To evaluate R,, using Eq. (12.20), the terms in the last column of Fig. 12.7b are
added. Thus
—239.1 + 5:31R = 0
from which
R = 44.3kN
and
Sac = —8.6kN, Sap = —30.0 kN
Example 12.6
Draw the bending-moment diagram for the ring in Fig. 12.8a.
A closed ring is an indeterminate structure. This becomes evident if we
consider the free body of half the ring, shown in Fig. 12.8b. From symmetry about a
horizontal line through A and C, we conclude that an axial force P/2 exists at both B
and D and that the moments at B and D are equal. Furthermore, the ring is
symmetric about a vertical line through B and D. As a consequence, shear, which is
an asymmetric quantity, cannot be present at B and D. Since the use of symmetry
is equivalent to employing equations of equilibrium, the above applications of
symmetry have exhausted the available conditions of equilibrium and we are left
with one unknown, M,. The given ring is thus indeterminate to the first degree, and
in accordance with Castigliano’s theorem
aU
=0
OM,
Provided the depth of the ring cross section is small in comparison with the
radius R of the ring, it is permissible to use the expression for the strain energy of a
straight beam. Thus
M’ ds
2EI
Method of Least Work Chap. 12
266
Mg
= |
N
= o
aM
(a) (b)
> B 0.18 PR
"Ce iT
ve 6
Moment diagram
22 Serr
(c) (d)
Figure 12.8
and
OU M(0M/0M,) ds
itll |amcor eat 12221
OM, J EI ( )
It should be noted that the integration is carried out around the circumference of the
ring.
When dealing with a circular member, polar coordinates are preferable to
Cartesian coordinates. If 0 is measured counterclockwise from line OB, as shown in
Fig. 12.8c, the moment at any section between B and A is given by
P
BAN a
PR
Te a ky PON) 0<a<
@
and
aky = TRANG
Problems 267
Substitution of the above expressions into Eq. (12.21) and making use of symmetry
gives
PR
M, = — (== )= 0.18PR
4a 4
from which
PROBLEMS
12.1 to 12.17. Determine the reactions and draw the bending-moment diagram.
12.1.
ae a
I = constant
12.2. 4 15 kN/m
I = constant
I = constant
268 Method of Least Work Chap. 12
I = constant
B C
gf
I = constant
pg
ne
12.6. 10 kN/m
I = constant
IPeTs 30 kN
I = constant
Problems 269
12.9. 10kN/m
I
~ ro) 3
| —
12.10. 4kN/m
10 kN
——
20 kN/m
ate
(oer: feed
4
D
+ 10 ++ 10 —
« Ieeam = 400 X 10° mm*, A... = 20 mm.
20kN/m
12.16.
Problems 271
12.17. 10k
Semicircular
arch with
50 ft radius
Zp
12.18. Draw the bending-moment diagram for the ring in the figure. J = constant.
20 k
10k 10k
20k
Semicircular
archwith
unifornvertical
load.
272 Method of Least Work Chap. 12
12.20. Draw the bending-moment diagram for the box frame in the figure. J = constant.
40 kN/m
kN/m
40
I = constant.
Hint: Make use of symmetry.
4
Hint: Make use of symmetry.
Problems 273
12.23. Find the bar forces for the truss in the figure. A = constant.
60 KN
onl
oi
12.24. Find the bar forces for the truss in the figure. A = constant.
15k
~ 20 ++ 29’ ———|
aay
TiAl. Dat-viniie | pact
> = Saeetan nn
His ‘conanike ON 1 GTN re
—
= aed
=
A)
ya
c?
fr Perm le|
Na
"= Ee a
r
« ie em
Lo tu oe oe
13
Slope-Deflection Metho
Figure 13.1
276
Sec. 13.2 Derivation of the Slope-Deflection Equation 277
will develop end moments and become deformed as indicated. The notation used
in the figure to describe moments and deformations will be followed throughout
the chapter.
1. The moments at the ends of the member are designated as M,, and M,,,
indicating that they act at ends A and B of member AB.
2. The rotations of ends A and B of the member are denoted by 6, and Oz.
Since the rotations of all members of a rigid frame meeting at a common
joint are equal, it is customary to refer to each of them as the joint rotation.
3. The term A,, represents the translation of one end of the member relative
to the other end in a direction normal to the axis of the member. Sometimes
the rotation of the axis of the member wag = A,z,/L is used in place of
Aap:
In dealing with the variables described above, the following sign convention is
used.
1. The moments acting on the ends of the members are positive when
clockwise.
2. The rotations of the ends of the members are also positive when clockwise.
3. The relative displacement of the ends of a member A is positive when the
axis rotation w is clockwise.
Thus the moments and deformations are all positive when in the directions
indicated in Fig. 13.1.
We will now derive an expression for the member moments M,, and Mz, in
terms of the deformations 6,, 0;, and A,, and the load P acting on the member.
It is easiest to carry out the proposed derivation by considering the effects of the
four variables on the moments one at a time.
ManLL
EI2 3
Moa 2b _
EIM2S 3
from which
1
Mpa = 5 Man (13.1)
From the first moment-area theorem, the change in slope between A and B is
’
Slope-Deflection Method Chap. 13
278
= BA
EI
Figure 13.2
Sire lend
24 Eg) CORT D
In view of Eq. (13.1) this reduces to
4EI0 2EI0
Mas = ia : , Mpa a 7 = (2322)
4EI6, 2E16,
L
3. End moments due to a relative joint displacement A; 0, = 03, = P = 0:
Figure 13.4 depicts a member with a relative joint displacement but no joint
rotations. From the first moment-area theorem, the change in slope between A
Figure 13.3
Sec. 13.2 Derivation of the Slope-Deflection Equation 279
Figure 13.4
and B is given by the area of the M/EI diagram between A and B. Thus
Mpa L Map L
jee
ain she
Or
Mpa = Map (13.4)
From the second moment-area theorem, the distance between B and a tangent
drawn to A is equal to the moment of the M/EI diagram about B. Hence
_MpaL Lb_Map
a Klatisanbel
L2L3
Combining this expression with Eq. (13.4) gives
6EIA
Map = Mga = — L2 C325)
The negative sign in Eq. (13.5) indicates that a positive translation leads to
negative end moments. The translation A is positive because w is clockwise.
4. End moments due to loads acting on the member; 6, = 0, = A = 0:
These conditions describe the fixed-end beam shown in Fig. 13.5. The moments
Mrap and Meg, produced at the ends of such a member by loads acting along its
Figure 13.5
280 Slope-Deflection Method Chap. 13
PL % ae
BS 8
B Yj
i L
2 2
Pab2 Pba?
L2 L?
y w k/ft
12 iY, 12
-——.—__j
ig
w k/ft
wl? wl?
“30 20
Pa
span are accordingly called fixed-end moments. Table 13.1 gives the fixed-end
moments for several common loading conditions. The fixed-end moments for
other loads, not included in the table, can be derived using the moment-area
method.
Combining Eqs. (13.2), (13.3), and (13.5) and the fixed-end moments, we
obtain
2EL 3A
Map Se (26, as Oz a =) Se Meas (13.6)
L L
2El 3A
Mpa = Tie (26, =F 0, — =) an Mepa (13.7)
Equations (13.6) and (13.7) give the moments at ends A and B of a member
AB in terms of the deformations at the ends of the shWee and in terms of any
loads acting along the span of the member. These equations are known as the
q
slope-deflection equations.
Sec. 13.3 Alternate Derivation of Slope-Deflection Equation 281
If we compare Eqs. (13.6) and (13.7), it becomes apparent that the two
equations are essentially identical. In other words, the moment at either end of a
member is a function of twice the rotation at that end, the rotation at the far end,
the relative joint displacement, and the fixed-end moment at the end being
considered. Either one of the equations is thus sufficient to determine the
moment at both ends of a member.
As was the case in the methods of consistent deformations and least work,
when they were applied to flexural members, the slope-deflection equations
consider only bending deformations. Deformations due to shear forces and axial
loads in bending members are ignored. Since the deformations caused by shear
forces and axial loads are very small compared to transverse bending deforma-
tions in most beams and frames, this is a reasonable assumption.
—_> VY
Figure 13.6
Slope-Deflection Method Chap. 13
282
differential equation
M Mpx
(13.8)
Ely! = at: —L)+ L
twice. Thus
The solution of this equation is obtained by integrating both sides
_ Mz 2 13.9
Ely! = (=> L
M, fe cal M,x°
Ely = —iw (— — —]+
5) 6L +). Gat 1x HAG 2 (13.10 )
ely, gives
Substitution of these conditions in Eqs. (13.10) and (13.9), respectiv
C,.= UF ands. Cy == E10,
M, = rhs
2EI
— (204 + 03 — =|
3A\ (13213)
Similarly, we can eliminate M, between eqs. (13.11) and (13.12) and obtain
_ 2El 3A
Since no load was included along the span of the member in Fig. 13.6, Eqs.
(13.13) and (13.14) give only the effect of joint deformations on the moments at
the end of a member. The total end moments are obtained by adding the
fixed-end moments, which are the effects of interfor loads, to the above
equations.
Sec. 13.4 Application of the Slope-Deflection Method 283
The primary aim of the slope-deflection method, like that of the methods of
consistent deformations and least work, is to determine unknown reactions and
internal forces. In the methods of consistent deformations and least work, this is
accomplished by setting up equations in terms of the unknown forces and then
solving the equations for these forces. By comparison, the slope-deflection
method employs a more indirect approach. Using the slope-deflection equation,
the desired moments are first expressed in terms of unknown joint displacements.
Equations of equilibrium are then written, and their solution gives the values of
the joint displacements. Finally, the values of the joint displacements are
substituted back into the slope-deflection equations, giving the desired moments.
This indirect approach is justified because it leads to far less complex equations
than those obtained using the more direct force methods.
Example 13.1
Determine the reactions and draw the moment diagram for the beam in Fig. 13.7a.
40 kN
(; ee »)
Mea Mce
(b)
40
Meag = 80 a Mega
=80
Figure 13.7
Slope-Deflection Method Chap. 13
284
5313
93.3
(e)
member moments.
JIL, 40(16)
Vi FAB eee
8 ee 8 80
Merrga = +80
2EI
Man = [= (On) ~ 80 = 0.125E105 — 80
2EI
Mpa = [= (202) + 80 = 0.25EI0, + 80
2EI
Msc — ~g (282) = 0.5EIO,
2EI
Mca = ~~ (On) = 0.25E10,
Since use of the slope-deflection equation implies that we initially assume all 6’s as
well as all moments to be clockwise, the negative sign indicates that 0, is
counterclockwise.
from which
Vee kN
Vag = 22.5KN
Voc = Von — 10 kN
The bending-moment diagram and the reactions are shown in Fig. 13.7e. The
reaction at B is the algebraic sum of Vz, and Vzc.
The foregoing solution involves two separate and independent sign conven-
tions, which should not be confused with one another. The first, employed in the
slope-deflection equations, states that moments acting on the ends of members
are positive if they are clockwise. This is a “rigid body” sign convention of the
type used when Newton’s law is used to write moment-equilibrium equations. By
comparison, the second sign convention, used to construct the bending-moment
diagram, states that moments producing compression on the upper side of a
member are positive. This is a “strength of materials” sign convention that deals
with stresses and deformations instead of external moments.
Example 13.2
Determine the reactions and draw the bending-moment diagram for the beam in Fig.
13.8a.
Equations of equilibrium
Fixed-end moments
20(12)7(8)
Mie
FAB eee (207° a ee 57.6
20(8)?(12)
Mec=
FBA a (20)?ee +38.4
2(20)?
a
FBC D =e 66.7 @
Meer = +66.7
Sec. 13.4 Application of the Slope-Deflection Method 287
20k
12'—>+ 2 k/ft
d 7B I €
|. 20' ol 20'—— — 20'—>|
(a)
20K
2 k/ft
31.0
75.8
(b)
Figure 13.8
Member moments
Aah
At this stage we could simply substitute the above expressions for the member
moments into the four equations of equilibrium and solve them for the unknown
joint rotations. However, in a problem such as this, where one or more ends of the
structure are simply supported, an alternative procedure, which simplifies the
numerical work, can be employed. Before solving the equations of equilibrium for
the unknown joint rotations, Eqs. (13.16) and (13.19) can be used to express 6, and
Op in terms of 6, and @., thereby eliminating two unknowns.
From Eq. (13.16) we obtain
EI0, = 288 — 0.5EI160, (13.20)
288 Slope-Deflection Method Chap. 13
Op = —0.56¢ (13:21)
from which
Example 13.3
Determine the reactions and draw the bending-moment diagram for the structure in
Fig. 13.9a.
Equations of equilibrium
Fixed-end moments.
M FAB See
12(10)?
a ae ee
12
Meza = +100
6(10)?
Mrgc aad ( ) =
12 @
Merce = +50
Sec. 13.4 Application of the Slope-Deflection Method 289
12 kKN/m
6 kKN/m
ack
ro>)
e=)
, 12 kKN/m
3.65 i 6 kN/m
————= |
108.7
62.6
One)
108.7 14.6
Has
(b)
Figure 13.9
Slope-Deflection Method Chap. 13
290
Member moments
Mp = 0.4EI0;, — 100
Mya = 0.8EI0, + 100
Mac = 0.4EI0, + 0.2EI0- — 50
Men = 0.4E10¢ + 0.2EI05 + 50
Mapp = 0.667EI05
Mpp = 0.333E10,
Solution for unknowns. Using Eq. (13.23), we can express 6¢ in terms of
0,. Thus
EI@¢ = —57.50 — O5EI@;
Example 13.4
Determine the member moments and reactions and draw the bending-moment
diagram for the frame in Fig. 13.10a.
Unknowns. 95, 6c, A.
A frame, whose upper part is not prevented from translating laterally, will
sway either to the right or to the left a distance A, as shown in Fig. 13.10b. Since we
ignore axial shortening of flexural members, both that due to axial stress and that
due to bending, ends B and C of member BC are assumed to move laterally the
same amount; that is, Az, = Acp = A.
34,58
12.89 13.78
12.89 13.78
TEA 6.22
(d)
Figure 13.10
292 Slope-Deflection Method Chap. 13
Since the slope-deflection equation relates only member moments and not shears to
the unknown joint rotations and displacements, it is necessary to express H, and Hp
in terms of the member moments. From Fig. 13.10c it is evident that
At 6 a a
= Mas + Mpa +7 Moc + Meco
Example 13.5
Determine the member moments and the reactions and draw the bending moment
diagram for the frame in Fig. 13.11a.
Unknowns. 63, 9¢, 9p, 9c, A;, A>: In this frame there are, in addition to
four unknown joint rotations, two unknown joint displacements. Remembering that
1 k/ft
Figure 13.11
294 Slope-Deflection Method Chap. 13
iN ax
— i Mbe
H He
Mpc — Mep —
KT) KT)
(d) Vp Ve
(e)
1k/ft
4t-==—
34.07k-ft
13.43
13.43 41.85 41.85
15.35 26.93
4.05
34.07 73.83
(9)
Figure 13.11 (continued)
joint displacements are defined as the distance that one end of the member
translates relative to the other end, the displacement of B relative to A is designated
as A, and the displacement of C relative to B as A, (see Fig. 13.11b). Since we
neglect axial deformations of flexural members, the lateral translation of E is equal
to that of B, and the displacement of D is the same as that of C.
H, = —“—*
—Maps ae Mpa
H, = ————=
—Mrr a, Mer
¥ 10 4 10
which makes it possible to rewrite Eq. (13.32) in the form
In a similar manner, using the free bodies for the upper story shown in Figs. 13.11d
and 13.1le, we can write
Hz a lily => 4
and
Te
B
Cee Se,
—Mosc = Mes
10 Hy;
SE
—Mep Dee
10
DE
= Moe
which leads to
Member moments
PROBLEMS
13.1 to 13.21. Determine the member moments and the reactions and draw the
bending-moment diagram.
13.1. 2 k/ft
4
A C
B
oe
ZI = constant
Slope-Deflection Method Chap. 13
298
13.2. 2m ,100KN
30 kKN/m
J = constant
13.3. 20 k
1.5 k/ft
120 kN
13.4.
5m>
20kN/m 20kN/m
13.5.
13.6. 20kN/m
13.7.
3.8. as
B C
}+-—— 20” es 20’ ——.— 10 as
I = constant
13.9. 50kN
| 40kN/m |
Cc
:
A B
3m Alla 12m 10m |
I = constant
13.10. 40 k 5k
a5
B Cc
Y Y Zi Uf,
ZI = constant
13.11. 16 k-ft
I = constant
300 Slope-Deflection Method Chap. 13
13.12.
I = constant
Problems 301
13.16. 20kN/m
13.18.
7 =)
13.19. 20 kN/m
I = constant
302 Slope-Deflection Method Chap. 13
10kN/m
Moment-Distribution Method
1. The fixed-end moments, which are the moments caused by the loads acting
on the member, with the ends of the member assumed to be fixed.
2. The moments due to the rotations that actually take place at the ends of the
member.
3. The moments caused by the translation of one end of the member relative
to the other.
Actual
structure
(a)
8 Locking moment
Unlocking moment
Unlocking of
joint B
(c)
(b)
Figure 14.2
Moment-Distribution Method Chap. 14
308
AB
M Se
4EL,c0
ee
AC lege
4EI, p90
MapAD = ———
lips
stiffness and to
It is customary to refer to the quantity / /L of a member as its
denote this ratio by k. Thus the moment s can be rewritte n as
Map = 4EOk ap
4E0 = a
YK
and this expression substituted back into Eqs. (14.1), we obtain
_ Mkaz
AB ‘3 ka
Map = AAD
uk
It is evident from Eqs. (14.3) that an external moment applied to a joint will be
resisted by the members framing into the joint in proportion to their stiffnesses
: The ene eaXk, for a given member, is referred to as the distribution
actor of that member. It gives the fraction of the total m i jOi
that is resisted by that member. ie
Sec. 14.3 Structures Without Joint Translations 309
1
Mpa = 5 Map
The carry-over factor, which is defined as the ratio of the moment induced at the
fixed end to the moment applied at the end that is able to rotate, is thus equal
10 4.
It should be kept in mind that the foregoing values of the carry-over factor,
the distribution factors, and the fixed-end moments listed in Table 13.1 assume a
constant stiffness for the entire length of the member.
1. The first step is to calculate the distribution factors at all joints that are free to
rotate. In this instance there is only one such joint, joint B. The distribution
factor of a member at a joint was shown in Section 14.2 to be equal to the
stiffness k of the member divided by the sum of the k’s of all the members
framing into the joint. Thus
These factors are recorded in a table below the structure, as shown in the
figure.
310 Moment-Distribution Method Chap. 14
20k
1.5 k/ft
Distribution
factors
Fixed-end
moments
Balancing of
+16.67 |+8.33 ~_
joint B i
Carry over
moments
Final
moments eo!
+41.67 | —41.67
20
Saal Sickie.
1.5 k/ft
20.8 27.2
41.7
54.2
Figure 14.3
3. We are now ready to begin the actual moment-distribution process. Since joint
B is the only joint that was artificially locked in the preceding step, it is the
only joint that must now be unlocked. In the locked position an artificial
external moment of 25 kip-ft was needed to keep the two member moments
acting on joint B in equilibrium. When we release the joint, by removing the
artificial locking moment, the moment necessary to produce equilibrium of the
joint will be supplied by the members framing into the joint in proportion to
their stiffnesses. In other words, the moment of +25 kip-ft needed to produce
equilibrium is distributed among members AB and BC in accordance with the
distribution factors. Thus member AB develops a moment of +16.67 kip-ft
and member BC a moment of +8.33 kip-ft. these moments are recorded as
shown in the figure, and a line is drawn below them to indicate that the joint is
now in equilibrium.
4. According to the theory developed in Section 14.2, moments one-half as large
as those produced at end B of members AB and BC by the rotation of joint B
are induced at ends A and C of these members. Thus moments of 8.33 and
4.17 kip-ft are recorded at A and C in the table.
5. At this stage in the analysis, joints A and C are fixed and joint B is externally
unrestrained. Since this corresponds to the support conditions in the actual
structure, the moment-distribution procedure is complete.
6. The final moment at the end of each member is obtained by adding the
moments developed during each of the preceding steps. Thus
May = —16.67 kip-ft
Mpa = +41.67 kip-ft
Myc = —41.67 kip-ft
Men = +54.17 kip-ft
It should be evident from the foregoing analysis that whereas the moment-
distribution process involves both members and joints, it is necessary in the actual
calculations to record only the member moments.
Example 14.2
Determine the member moments and construct a moment diagram for the structure
in Fig. 14.4.
The main difference between this structure and the one considered in the
previous example is that there are now two joints (B and C) that must be artificially
locked at the outset and then released, whereas there was only one such joint in
Example 14.1.
1. As before, the first step consists of calculating the distribution factors, locking
all the joints, and determining the fixed-end moments. Since member AB is
not subjected to any loads, there are no fixed-end moments at its ends.
2. We now proceed to unlock joints B and C one at a time. In other words, we
unlock either one of these joints while holding the other fixed. Let us begin by
unlocking joint B and distribute the resistance to the unbalanced moment of
—53.3kN-m to members AB and BC in accordance with the distribution
31 2 Moment-Distribution Method Chap. 14
30 kN
30 kN
4 Pat ?.
4.5 4.5 19.0 11.0 20.5 19.5
51.6
12.0 15.8
31.6
ae 36.8
Figure 14.4
B and D. As before, the joint is relocked once these operations have been
completed.
4. From here onward, we continue to carry out the process of unlocking,
balancing, carrying over, and relocking first at one joint and then at the other.
The procedure is stopped when the unbalanced moment remaining at both
joints B and C is negligible. Unlike joints B and C, joints A and D are never
released. As a consequence, no unbalanced moments are ever created at these
joints and no carry over takes place.
5. The analysis is completed by adding the moments at the end of each member
and thus obtaining the final member moments.
Example 14.3
Determine the member moments for the two-span beam in Fig. 14.5. The member is
fixed at A, continuous over the support at B, and simply supported at C.
6.0
Figure 14.5
Moment-Distribution Method Chap. 14
314
When there exists a simple support at the end of a structure, certain modifications
can be introduced in the moment-distribution procedure to reduce the numerical
work. In Section 14.2 it was shown that a moment applied to a joint at which
several members meet, each of which is fixed at its far end, is distributed among
these members in proportion to their stiffnesses. Let us now determine how this
distribution of moments is affected if one of the members is hinged at its far end.
Consider the three members meeting at joint A shown in Fig. 14.6. Members AB
and AC are fixed at their far ends, while member AD is hinged at D. In
accordance with Eq. (13.6), the member moments at A resulting from the
application of an external moment M to joint A are
Figure 14.6
Sec. 14.3 Structures Without Joint Translations 315
it follows that
6
On = =o
3
Map = 4E04(5 kav) (14.5)
or simply
Equations (14.7) indicate that the part of the total joint moment that is resisted
by any one member framing into the joint is equal to k’/).k'. This result is
identical to the one obtained in Section 14.2 except that effective stiffnesses are
now employed in place of ordinary stiffnesses.
It should also be noted that, whereas moments equal to 0.5 M4, and 0.5Mi4c
are induced at B and C by the application of M, no moment is induced at D.
Thus the carry-over moment to a hinged end is zero.
To illustrate the concepts introduced above, we will make use of effective
stiffnesses to reanalyze the beam considered in Example 14.3.
Example 14.4
Determine the member moments for the beam in Fig. 14.7 using effective stiffnesses.
The effective stiffnesses for members AB and BC are
I ami
kin = — = 0.051, kho = Sa = Ue!
20 4 20
316 Moment-Distribution Method Chap. 14
Figure 14.7
From these quantities the distribution factors shown in the figure are calculated.
Next all joints are locked and the fixed-end moments determined.
We are now ready to unlock the joints one at a time. Whenever effective
stiffnesses are used, it is necessary to unlock all joints at simple supports first. Thus
we begin by unlocking joint C, balancing the joint, and carrying over one-half the
balancing moment to B. If we were using ordinary stiffnesses, as we did in Example
14.3, we would now relock joint C and proceed to joint B. However, since
the
distribution factors at B are based on the assumption that joint C is hinged, we do
not relock joint C prior to unlocking joint B. Furthermore, since joint C remains
free to rotate during the remainder of the analysis, no moment is carried
over to C
when joint B is unlocked and balanced. '
Example 14.5
Determine the member moments and draw the bendin g-moment diagram for the
structure in Fig. 14.8.
The analysis of a frame like the one in Fig. 14.8,
none of whose joints are
permitted to translate, is very similar to that of
a continuous beam.However, it is
necessary to record the moments in such a way that they
do not interfere with one
another. As will be demonstrated, this can be
accomplished by recording some of
the columns of moments vertically and some
horizontally.
As usual, we begin by determinin
& member stiffnesses, distribution factors,
and fixed-end moments. Since simple su
pports exist at A and F, effective stiffnesses
are used. The overhang to the right of
F can offer no restraint to the rotation of joint
F. Hence both the effective stiffness
of member FG and its distribution factor at F
Sec. 14.3 Structures Without Joint Translations 317
40 kN 40 kN 25 kN
EI = constant
[Smetana
Sn abe
= 3 (5) fo
[Ae e eee kee 3 ft Ps
k= ={ —
A 4 \10 B 10 C Fae a
+81.8 =50:9
+0.1 =0:9
+6.7 =12.5
0 a ne +25.0 +12.5 +75.0 |—75.0
+50 oa +25.0
roo]
Jee
Z oi o[ | 8-—
+
,
+ |O}-100_
gale]
Bis} iis
+100
JO) 7] | Hi
12 YD
Z
+9.0 +0.2+8.8
Figure 14.8
Figure 14.9
Sec. 14.4 Structures With Joint Translations 319
Example 14.6
Determine the member moments for the structure in Fig. 14.10a.
In the first stage of the analysis the structure is artificially restrained against
lateral translation by the addition of a support at C. A routine moment distribution
is then carried out for the artificially restrained structure. This analysis is presented
in Fig. 14.10b.
The next step is to determine the magnitude of the artificial restraining force
at C. Using the member moments determined in the preceding analysis and writing
equations of equilibrium for the individual members of the frame (Fig. 14.10c), we
ie
I = constant
Actual structure
(a)
Sup
= ena,
Analysis of artificially
restrained structure
(b)
Figure 14.10
Moment-Distribution Method Chap. 14
320
F = 3.98
=p
ed
————
Hp = 6.82
H,
=2.84
Artificial
restraining force
(c)
= wm O12
Ell ero eS
olyol;rno— 8.4
+
TE TI ah Se lene! he
eis
Stine =
Gal iS +1 + +15.4
alj- Oo
+], io +
YZ
(d)
3.36
oS
Hp = 2.38
H, = 0.98
(e)
obtain
F=—H, 14 20 28.4 oe 22 Ea
+H, = =—————-_ + = 3.98k
Pps ti 15 10
Thus, the artificial restraint at C consists of a 3.98 k fotce acting to the
right.
We are now ready to proceed with the second stage of the analysis,
whose
purpose is to cancel out the effects of the artificial restraining force
F. Accordingly,
Sec. 14.4 Structures With Joint Translations 321
we apply a force F’, which is equal and opposite to F, to the structure as shown in
Fig. 14.10d and determine the corresponding member moments. Since it is not
possible to obtain a direct solution for member moments due to a given lateral force
such as F'’, we proceed in an indirect manner as follows. If we assume all joints to be
initially locked against rotation but free to translate, the lateral force F’ gives rise to
a lateral displacement A at B and C as indicated in Fig. 14.10d. The member
moments that are produced when the ends of a member translate a distance A in
relation to each other but do not rotate are, according to Eq. (13.5), given by
6EIA
Mas = Mga = (15)?
6EIA
Mep = Mpc = (oy?
Not knowing the magnitude of A at this time, we arbitrarily assign values to the
member moments that are proportional to the quantities //L*. Thus we let
Mas = Mg, = 10k-ft
Mep = Moc = 22.5 k-ft
These moments are fixed-end moments not unlike those used in the ordinary
moment-distribution procedure. The difference is that the present moments are due
to a joint translation, whereas the previous ones were caused by loads acting normal
to the members. However, in both instances they correspond to joints that have
been prevented from rotating. As in the ordinary moment-distribution procedure,
the next step therefore consists of releasing joints B and C, one at a time, balancing
the joints, and recording the carry-over moments. During this procedure the
translations of joints B and C are kept fixed at their original value, A. The details of
the moment distribution are recorded in Fig. 14.10d.
As before, we now determine the value of F’ that corresponds to the member
moments obtained in the analysis. Using the free body in Fig. 14.10e leads to
8.3+64 15.4 + 8.4
F' = H, + Hp = = 3.36k
15 a 10
It was our objective, in the second stage of the analysis, to determine
moments corresponding to F’ = 3.98k. Instead, the moments we calculated
correspond to a force of 3.36k. However, member moments are proportional to the
applied load, and the desired moments can therefore be obtained by multiplying the
calculated ones by the ratio 3.98/3.36 = 1.18.
The final member moments in the structure are now obtained by adding the
corrected moments from the second analysis to the moments determined in the first
analysis. Thus
Map = 14.2 + 1.18(8.3) = 24.0k-ft
Mpa = 28.4 + 1.18(6.4) = 36.0k-ft
Mac = —28.4 + 1.18(—6.4) = —36.0 k-ft
Men = 45.5 + 1.18(—8.4) = 35.6 k-ft
Mep = —45.5 + 1.18(8.4) = —35.6k-ft
Mpc = —22.7 + 1.18(15.4) = —4.5 k-ft
Chap. 14
Moment-Distribution Method
322
PROBLEMS
14.1.
2 k/ft
Cc
A
zB U,
K ee > 4! a4 +]
I = constant
A D
B Cc
<6 m sa 8m +} 8m >|
14.3. 120 kN
20kN/m 20kN/m
A D
B C
14.4. 40k
1.8k/ft
D
:
14.5.
Problems 323
14.6. 20k
jee,
| 1.5k/ft
B C
}-— 15° —+}—15°—+}-—15°—}—20" 4
I = constant
14.7. 4kitt
14.8.
1.5 k/ft
I = constant
14.9. 50 kN
| 10kN/m
A B C
JM
1.6m -t-— 12m ——}-—10m —+|
I = constant
14.10. 12 kN/m
|x —10m fe 10m -
I = constant
Moment-Distribution Method Chap. 14
324
14.11.
1.5k/ft
“sp
aZ
Th
14.14. 20kN/m
4m
mea
14.15.
I = constant
Problems 325
I = constant
14.17. 20 kN/m
I = constant
Matrix- Flex ty Method
The forces and the displacements that exist in a structure can be related to
one
another either by using flexibility or stiffness influence coefficients. The method
of
analysis to be considered in this chapter uses flexibility influence coefficien
ts and
is therefore referred to as the flexibility method.
Let us consider the beam shown in Fig. 15.1, along which we
have chosen
three points, designated as 1, 2, and 3. If we apply a load
W, at point 2, the
deflection A, at point 1 can be expressed in the form
A, = FW,
The term F, is a flexibility influence coefficient.
It is defined as the deflection at 1
due to a unit load at 2.
Figure 15.1
328
Sec. 15.2 Flexibility Matrix 329
If, instead of a single load at 2, we apply loads at all three points, the
deflection at 1 will be given by
A, = FW, + F2W, + Fi3W;
and the deflections at 2 and 3 by
A, = FiW, + RoW, + BW;
A; = iW, + FW, + BW;
These expressions can be rewritten in matrix form as
Ay Fi Fo Kz W,
A,|=|Fi & &#; WwW,
A; Fay, 14524 153_) |W;
or simply
[A] = [FI[W]
The matrix [F] that contains the flexibility influence coefficients and relates the
deflections [A] to the loads [W] is referred to as the flexibility matrix. Maxwell’s
law of reciprocal deformations states that F; = F;. Thus the flexibility matrix is a
symmetric matrix.
As an illustration, let us construct the flexibility matrix that relates the
forces and displacements shown in Fig. 15.2a to one another. The vectors in the
figure define not only the forces and deformations that we are considering but
also their positive directions. Thus the vertical force W, and the corresponding
vertical deflection A, are positive when downward, and the moment W, and its
corresponding rotation A, are positive when clockwise. It is customary to
generalize the term force to include moments as well as forces and the term
deflection to designate both rotations and translations.
The matrix equation that relates A, and A, to W, and W, is
puleigell
A, Fy, FoitW,
According to the definition of a flexibility influence coefficient, F,, and F, are the
es
vertical deflection and rotation at the free end of the member due to a unit
vertical load at that point. The first column of the flexibility matrix can thus be
obtained by considering the beam in Fig. 15.2b. Using the moment-area method,
we obtain
1Bea L?
hea, F,=—
ORT ees Fi
Similarly, the second column of the flexibility matrix is obtained by considering
the beam in Fig. 15.2c. Thus
Is L
E =——, 2= =
? nO ET EI
Matrix-Flexibility Method Chap. 15
330
(b)
1
pea iter S ae hz
ve
7, Licercaniarteae
gt]
=
Fo
m
(c)
Figure 15.2
|
Baad be
A ee
| BENE SET ||. 3
A
aye
SS
ae
ase
5 2hlae aL ae
Example 15.1
Construct the flexibility matrix for the beam in Figs5i3a:
To obtain the first column in the matrix, we apply a unit vertical
load at the
free end of the member, as shown in Fig. 15.3b, and make
use of the moment-area
method. Thus
F, 8L*
=—, i= 217
—. SL}
= NSS
Ve
=> =
a aya <n Ey Bay 6EI’ Fa 2EI
In a similar manner, columns 2, 3, and 4 are
obtained by applying a unit moment at
the free end and a unit force and a unit moment
at midspan as shown in Figs. 15.3c,
Sec. 15.2 Flexibility Matrix 331
Figure 15.3
15.3d, and 15.3e. The results, when combined, lead to the following flexibility
matrix:
8L° Se eal
A —| 20% ——=—|1 Ww,
: 3 Con :
Palio
- ek,
Ae gee
ereeek ||
Hee ee
Ay —; je =e” al |W,(
332 Matrix-Flexibility Method Chap. 15
Although the method used to form the flexibility matrix in the previous section
could theoretically be used on any structure, the difficulty of obtaining deflections
in large and complex structures would make it extremely impractical. As a
consequence, we will now introduce an alternative method for determining the
flexibility matrix. In this method the structure is subdivided into several elements,
and a flexibility matrix is formed for each of the elements. The flexibility matrix
for the entire structure is then obtained by combining the flexibility matrices of
the individual elements. This process, whereby the behavior of a large system is
synthesized from the behavior of the individual elements into which the system
has been subdivided, is the basis of matrix analysis.
Let us consider the beam in Fig. 15.4a. The four vectors in the figure define what
we will refer to as the structure forces and deformations. They include two forces
and two moments and the corresponding deflections and rotations. It is our aim
to obtain the flexibility matrix that relates these quantities to each other, that is,
the matrix [F] in the equation [A] = [F][W]. To this end, we subdivide the beam
W3, 3 W,, A,
Structure forces
and displacements
(a)
[++ [/-+-———_+
Element forces
and displacements
(b)
Figure 15.4
Sec. 15.3 Formation of the Structure-Flexibility Matrix 333
into two elements, as indicated in Fig. 15.4b, and define two moments and their
corresponding rotations for each element. A flexural element of a plane structure
actually has six forces acting on it: an axial force, a shear, and a moment at each
end. However, axial deformations are usually negligible in a flexural element, and
only two of the remaining four forces are independent. Thus we have chosen the
moments and the corresponding rotations at the ends of the elements as the
forces and deformations we will use to describe the behavior of the elements. To
differentiate these quantities from the structure forces and deformations, we refer
to them as element forces and deformations.
The juncture of two elements is called a node. For the time being we assume
all structure loads to be applied only at nodes. In the flexibility method we may
include as many or as few structure forces as we wish. The only restriction is that
the forces be independent of one another. Thus reactions that are dependent on
the applied loads cannot be included in the analysis. As far as element forces are
concerned, they too must be independent. Thus we can use either the moments at
both ends, as we have chosen to do, or the shear and the moment at one end.
Element-Flexibility Matrix
We will now construct the flexibility matrix, which relates the deformations to the
forces, for the typical flexural element shown in Fig. 15.5a. To obtain the first
column of the matrix, we apply a unit moment q,; = 1 as shown in Fig. 15.5b.
Using the moment-area method, we obtain
#diyL.
ay he ey A
G+)
qi, qj, 8;
ees) ANG)
(a)
(b)
Figure 15.5
Matrix-Flexibility Method Chap. 15
334
and 7 Sly 1
Ii L 6EI
Similar results are obtained if we apply a unit moment to the right end of the
element. The flexibility matrix thus takes the form
sl seili alle
0; OHI LL 21tg; ce
i
Equation i
(15.3) gives ibili
the flexibility matrixi for an element o f a planar structure for
which only flexural deformations are considered. It should be noted that
lowercase letters are used when dealing with elements and uppercase letters when
the entire structure is being considered.
A great advantage of subdividing a structure into a series of elements is that
the same element-flexibility matrix can be used for all the elements of the
structure. ad ;
We will not form a matrix containing a flexibility matrix of the type given by
Eq. (15.3) for every element in the structure. Thus
a 2 lOO q1
Orbs al v2) 0 SO.) a2
Cae OE] |Om Ome ates
O4 O ROTIg 2 qa
Or
Force-Transformation Matrix
or
W,=1
a 2L —-
or mary
q,=—2L
ba
qj=-L q3=-L
reagan
q4= 0
os
(c)
W,=1
(d)
Figure 15.6
Matrix-Flexibility Method Chap. 15
336
Similarly, the remaining three columns of [B] are obtained using the loadings
shown in Figures 15.6b, 15.6c, and 15.6d. The element forces produced by each
of the unit structure loads are shown in the figures. The negative signs indicate
that positive structure loads cause negative element forces in the system we are
considering.
The preceding results, when assembled, produce the following [B] matrix:
a1 “0D t 1 Sa ay
q2 Sie ty ald
qa |et | ee OO |
qs Oo) = 190 One
Transformation of [f,] into [F]
Ag b4
or
5(WILAl = 5Lal")
Substitution of Eq. (15.4) for [6] gives
[W]"[A] = [W]"[B]"[£IBIW)
from which it follows that
[A] = [F][W]
indicates that
Equation (15.6) shows that the structure-flexibility matrix [F] can be obtained
from the composite element-flexibility matrix if the latter is premultiplied by [B]”
and postmultiplied by [B]. Since the synthesis of structure properties from those
of the individual elements is the very essence of matrix analysis, we cannot
overemphasize the importance of Eq. (15.6).
We will now use Eq. (15.6) to obtain the structure-flexibility matrix for the
beam in Fig. 15.4.
> alent) 0 =2L aA eedBp
ay
are
2
2
2
L?
Example 15.2
Construct the structure-flexibility matrix for the frame in Fig. 15.7a. .
Both structure and element forces are defined in the figure. On the basis of
ibili
ix-Flexibi
Matrix- hod
lity Metho Chap. p 15
338
q3, 53 G4, O4
G2, 52 C » ds,55
Fy a
bed gy, 94
v
G6, 56
ie 6
(b)
Figure 15.7
ee
ee e
ee
Pe
ee N
and
0
iL,
» -iflw
q3 L Ol Ww,
0
46 0 0
The free bodies used to construct [B] are shown in
Pipes .7bi
Sec. 15.3 Formation of the Structure-Flexibility Matrix 339
(a) eet
A, 6EI.-2L al
4 Ww,
Example 15.3
Construct the structure-flexibility matrix for the truss in Fig. 15.8a.
The structure forces are defined in Fig. 15.8a, and the element forces are
assumed to consist of a single tension force for each bar. The element-flexibility
matrix for an axially loaded bar of the type shown in Fig. 15.8b is given by
(1 = | la
oo
|0.667 f,
667 a f0.5
(c)
Figure 15.8
Matrix-Flexibility Method Chap. 15
340
6
matrix, we must determine
To obtain the two columns of the force-transformation
unit loads W, and W, as indicate d in Fig. 15.8c. This leads to
the bar forces due to
qd 0.833 —0.625
q2 —0.833 —0.625
: 1
q3| = 0 1.0 W,
4 0.5 0.375
qs 0.5 0.375
The structure-flexibility matrix can now be constructed using the relation [F] =
[B]’[f.][B]. Thus we obtain
Pips ae
Ay| AE I225 17.501 LW,
To see how this procedure is carried out using thé flexibility method, let us
consider the beam in Fig. 15.9a. Choosing the moments at A and C to be the
redundants, we obtain for the determinate structure the simply supported beam
Sec. 15.4 Analysis of Indeterminate Structures 341
A C
Ro, A, ? B 3)
ry
Structure loads
and deformations
(b)
q,, 5, Q2,52 3,53 G4,54
SiGe ae
Element loads
and deformations
(c)
Figure 15.9
in Fig. 15.9b. The structure forces, which consist of the known load W, and the
unknown redundants R, and R3, are shown acting on the determinate structure.
The four element forces corresponding to the two elements into which the
structure has been subdivided are defined in Fig. 15.9c.
We now proceed to form the structure flexibility matrix [F] for the
determinate structure in Fig. 15.9b. The composite element-flexibility matrix is
Dil BOA
Taal al OMe O
Le] = Gry 0 208402
0 00 74
and the force-transformation matrix is given by
ai Dial pao
Mh 9)
ge |)= algotei aga | |P-h
R, (15.7)
Oi ee tt]
ee eee || R
42 a Bre
da oe yee
342 Matrix-Flexibility Method Chap. 15
lol=[cacl+('s sella
for which
ze =— ie | Sie abel
Es 9 18 —8L [M1
Wil ==
oNeD [Wi]
W,
Having evaluated the redundants, we can now obtain the member forces
and the structure deformations. Substitution of the values for R, and R; in Eq.
(15.7) gives
1 5 cs j 0 wW, —12L
1
q2 33 3a 3 4LW, 1 8L
Ww,
4LW, 1
Pacer
L
———
ol 2 aw
aloe a
814 a
|G =
_ 8WL
3
2LW,
9 *
summarized as follows:
The
22] [ES IE
A
Ar
Fy: Kolfw
Fy, | F2itR osm
horizontal partitioning of [F] corresponds to the separation of [A] into
the unknown deflections [Ay] and the known deflections [Ag]. Similarly,
the vertical partitioning of [F] corresponds to the separation of the loads
into [W] and [R].
[Fo] = [Bif.Bo]
[Fa] = [Bzf.B,]
[Fo] = [B2f-Bo]
Matrix-Flexibility Method Chap. 15
344
ed = Fn Fale
[Aw] a [Fu = FF x7 F,)[W]
(15.15)
Equations (15.14) and (15.15) give the member forces and the structure
deformations in terms of the known loads [W]. We can think of the matrices
to the right of the equality sign in these equations as the force-
transformation and flexibility matrices for the indeterminate structures.
Example 15.4
For the frame in Fig. 15.10a, find the member forces and the unknown structure
deformations.
Let the vertical and horizontal reactions at A be the redundants. The
determinate structure obtained by removing the redundants is subjected to three
forces consisting of W,, R, and R; as shown in Fig. 15.10b. The element forces are
defined in Fig. 15.10c.
To form the structure-flexibility matrix requires that we construct [f.] and [B].
Corresponding to the two elements in Fig. 15.10c,
2) 1
and if we use the free bodies in Fig. 15.10d, [B] takes the form
q1 Mea 0 7
qzj{._|.0 ! 0 —-E\l@e’
Foal labseman meg nalts. (15.16)
qs te el oe
Sec. 15.4 Analysis of Indeterminate Structures 345
(b) (c)
Figure 15.10
Using the relation [F] = [B]’[f.][B], we obtain [F] and write the equation
a 3W,
ee a 8L? Soa il 6L Jima x 10L
Ral lol ke =—6L le 3M,
5L
Matrix-Flexibility Method Chap. 15
346
Substitution of the redundants into Eq. (15.16) gives the member loads:
qa C0) 0
1 6
qo Ome ; Tae
q3
~ il @ =
~ 10L}(W,]
3
=| 104 |W]
5L 2
qs 1 2L -L 0
The unknown structure deformations are obtained by Substituting the redundants
into Eq. (15.17).
_3W
[O11 =555 9(MM
Hk,
+T6L —OL) ay, 10L
=
W,L
oy L
Example 15.5
For the structure in Fig. 15.11a, find the member forces and the unknown structure
deformations.
If we let the force in cable CD be the redundant, the determinate structure
subjected to the structure loads is as shown in Fig. 15.11b. The element forces are
defined in Fig. 15.11c.
To form [F], we need [f.] and [B]. Thus
2; it
i)
ieee
L Te Saar
6EI 1 45
! |6
1 A
and
qq eels kis
-1 01
as|=] 0 0: 2L]) w, (15.18)
qs 0 0: OTR.
qs 0 0
Using the relation [F] = [B]"[f.][B], we obtain
A, On 3 ra W,
Ao: |teee eal Rel Gree
2 ie a eee eee | (eee (15.19)
A =120) 605 a0
Sec. 15.4 Analysis of Indeterminate Structures 347
L =10ft
I = 1440 in.4
A =0.05 in.?
Ee = S0)X 1102 ksi
W, = 200k-ft
ik W, = 20k
Structure loads
and displacements
(b)
poe
TG >)|.
Element loads
kU, A and displacements
q,,54
(c)
Figure 15.11
which leads to
12LW, — 6L’W,
te da 4018 017A
Matrix-Flexibility Method Chap. 15
348
L, A, and J, we obtain
Using the numerical values of W,, W2,
Rego lak
qs 0 O 2.31k
into
The unknown structure deformations are obtained by substituting the redundant
Eq. (15.19). Thus
FORTRAN Program
The following FORTRAN program will determine the redundants, the member
forces, and the structure deformations for either a plane frame or a plane truss.
The illustrative problem at the end of the program explains what input data are
required and how these data are to be entered into the program.
01300 RETURN
01310 END
01320 SUBROUTINE MATINV(A,N)
01330C MATRIX INVERSION BY AUGMENTED MATRIX METHOD
01340 DIMENSION S(100,100),A(20, 20)
01350 DO 100 I=1,N
01360 DO 100 J=1,N
01370 100 S(1,3)=A(I,J)
01380 [a1
01390 NX=N+1
01400 NY=2*N
01410 DO 80 J=NX,NY
01420 SC tiie 1s0
01430 T=I+1
01440 80 CONTINUE
01450 bint
01460 K=2
01470 110 XM=S(L,L)
01480 DO 140 J=L,NY
01490 S(L,J)=S(L,J)/XM
01500 140 CONTINUE
01510 DO 190 I=K,N
01520 XeSri)
01530 DO 190 J=L,NY
01540 S(1, J) =S(Tyl)2S(
bs99*%
01550 190 CONTINUE
01560 E-Lel
01570 K=K+1
01580 IF(L.LE.N) GO TO 110
01590 L=N
01600 235 LZ=L-1
01610 DO 290 K=1,LZ
01620 Tsien
01630 weectet)
01640 DO 290 J=L,NY
01650 Sis StS (lee
01660 290 CONTINUE
01670 Tale)
01680 TF(LGT.1) GO 10' 285
01690 DO 400 I=1,N
01700 DO 400 J=1,N
01710 NJ=NX+J-1
01720 400 A(I,J)=S(I,NJ)
01730C PRINT 330,((S(I,J), J=NX,NY), I=1,N)
01740C WRITEC 27330) ((S(1,J), JeNX,NY)) lel ,N)
01750 330 FORMAT(15H INVERSE MATRIX/(3X,1P3E10.3))
01760 RETURN
01770 END
Example 15.6
Using the preceding FORTRAN program, analyze the frame in Fig. 15.12a.
If we let the three reactions at D be the redundants, the determinate structure
subjected to the structure loads is as shown in Fig. 15.12b. The member forces are
defined in Fig. 15.12c.
352 Matrix-Flexibility Method Chap. 15
|
20 k
10 k B C
E = 30 x 10° ksi
| = 150 in.4
12°
(a)
|"
Y—
Ry<r
Ap,
[.
Structure forces
(b)
€ 2) C 5)
q3 N a,
Element forces
q, (c) F a,
Figure 15.12
Sec. 15.5 Computer Programs 353
Since the [f.] and [B] matrices make up a part of the input data, let us begin
by forming these matrices.
eh Gia
Re Ae 8 cee 1eee ace
1i5y 75g
el eth ie
2 ee anne anne a feaenceceecee qeneeeenee 15.20
CEI 7 7s oe
Bites Mee oes
lam et
24-12
‘24 24
[B,] [B,]
qi =—12 -15'30° 0 i
q2: Qu=15 (30-12
30, 2= ry,
qs (15200121)
7; ae ae Oe (Oa =o Bee
qs O- ONS-12 1 o
6 (i WEIN Se a =
qo 0 0:0 -12 1;+%
Qs 0 0; 0 O i
In the program, the four submatrices of [F] are formed separately, as outlined on
page 343. Accordingly, it is necessary to separate [B] into its two component parts
[B,] and [B,] and to enter these two submatrices into the program individually.
The input data are read into the program from a data file whose formation will
now be considered. The data file, shown later, is entitled DATA1.
Initially, the computer must be informed whether we are planning to analyze a
truss or a frame and whether we intend to use U.S. or SI units. Accordingly, the first
line in the data file contains the values of two indexes: IS, which indicates the type of
structure being analyzed, and IU, which indicates the type of units being used. If we
are analyzing a truss, we set IS equal to 1 and for a frame we let IS equal 2. If we are
using U.S. units, we let IU equal 3 and for SI units IU is set equal to 4. Thus the first
line in the data file for the sample problem being analyzed contains the numbers 2
and 3, indicating that we are analyzing a frame and using U.S. units. These two
numbers are read into the computer from the data file by the READ statement in
program line 00200.
The second line in the data file consists of the values of the modulus of
elasticity E and the section property AJ. If the structure being analyzed is a truss, Al
represents the cross-sectional area of the bars. For a beam or a frame, A/ represents
the moment of inertia of the members. If A/ is constant for all members, that value
of Al appears in the data file. On the other hand, if AJ varies from member to
member, the constant that is factored out in forming the [f.] matrix (see Eq. 15.20)
appears in the data file. Thus E = 30,000 and AJ = 150 for the problem under
consideration, and these two numbers make up the second line of the data file.
The third line of the data file contains the values of /FC, JB1, and JB2. IFC,
the number of rows in [f.] is equal to the number of element forces; J/B1, the
Matrix-Flexibility Method Chap. 15
354
number of columns in [B,], is equal to the number of applied loads; and JB2, the
number of columns in [B,], is equal to the number of redundants. Thus the third line
of the data file contains the numbers 8, 2, and 3.
The next eight lines of the data file contain the 64 elements of the [f.] matrix.
They are arranged so that each line of data corresponds to a row of [f.].
Using a similar arrangement, that is, one line of data corresponding to one
row of a matrix, the next eight lines of the data file contain the [B,] matrix and the
following eight lines contain the [B,] matrix.
Finally, the last two lines of the data file contain the applied loads, the [W]
matrix.
The following units must be employed when creating the data file:
E ksi kN/m?
Al in.* or in.4 mm’? or mm*
L ft 8 m
Ww kips or kip-ft KN or kKN-m
See
ee eee
Sec. 15.5 Computer Programs 355
The numeral 1, inside the first pair of parentheses of the READ statements
,
indicates that the data are to be read into the computer from TAPE1. Accordingly,
it is necessary to transfer the data file DATA1, which we have just formed, on to
TAPE1 by means of the appropriate system command before the program can be
run.
In the solution, forces are in kips or kilonewtons, moments in kip-feet or
kilonewton-meters, and deformations in inches or millimeters.
The output for the sample problem is as follows:
OUTPUT
Redundants
R(3) = 11.66
R(4) = 11.70
R(5) = 61.96
Member Forces
Q(1) = -8.39
Q(2) = —28.74
Q(3) = —28.74
Q(4) = 96.43
Q(5) = 96.43
Q(6) = —78.40
Q(7) = —78.40
Q(8) = 61.96
Structure Deformations
D(1) = 0.42
D(2) = 1.00
GWBASIC Program
The following GWBASIC program will determine the redundants, the member
forces, and the structure deformations for either a plane frame or a plane truss.
An illustrative problem is employed to explain what input data are required and
how these data are to be entered into the program.
Example 15.7
Using the GWBASIC program given on the following pages analyze the truss in Fig.
15.13a. The element and structure forces are defined in Fig. 15.13b.
Since the [f.] and [B] matrices make up part of the input data, let us begin by
forming these matrices.
36.06
36.06
1
2) 20.0 (15.22)
356 Matrix-Flexibility Method Chap. 15
Pu |, E = 30 x 10° ksi
A = 0.5 in?
it (constant for
20° all bars)
(b)
Figure 15.13
heap ig ae (15.23)
q4 0.50 0.75 —0.75%
The input data are read into the computer using DATA statements located at
the end of the program. The formation of these statements will now be considered.
Initially, the computer must be informed whether we are planning to analyze a
truss or a frame and whether we intend to use U.S. or SI units. Accordingly, the first
DATA statement (line 2520) contains the values of two indexes: IS, which indicates
which type of structure is being analyzed, and IU, which indicates the type of units
being used. If we are analyzing a truss, IS is set equal to 1 and for a frame we let IS
equal 2. If we are using U.S. units, IU is set equal to 3 and for SI units we let IU
equal 4. Since we are analyzing a truss ‘and using U.S. units, the first DATA
statement contains the numbers 1 and 3.
The second DATA statement (line 2530) contains the values of the modulus of
elasticity E and the section property AJ. If the structure being analyzed is a truss, Al
represents the cross-sectional area of the bars. For a beam or a frame, A/ represents
the moment of inertia. If AJ is constant for all members, that value of AJ appears in
the DATA statement. On the other hand, if AJ varies from member to member, the
constant that is factored out in forming the [f.] matrix (see Eq. 15.22) appears in the
DATA statement. Thus the second DATA statement contains the numbers 30,000
and 0.5.
The third DATA statement (line 2540) contains the values of JFC, JB, JR,
and JW. IFC, the number of rows in [f.], is equal to the number of element forces;
JB, the number of columns in [B], is equal to the sum of the number of applied
loads and redundants; JR is equal to the number of redundants, and JW is equal to
the number of applied loads. Thus the third DATA statement contains the numbers
Sy, Sethe Bieta ly
The next five DATA statements (lines 2550 to 2590) contain the 25 elements
of the [f.] matrix. They are arranged so that each DATA statement corresponds to a
row of [f.].
The following five DATA statements (lines 2630 to 2670) contain the [B]
matrix, and the last DATA statement (line 2710) contains the applied loads.
When creating the DATA statements, the following units must be employed.
E ksi kN/m?
Al in.” or in.* mm? or mm*
L ft m
Ww kips or kip-ft KN or kN-m
Since the program uses PRINT statements to produce output information, the
output will be displayed on the screen of the computer. If we desire to have the
output printed on a printer, it is necessary to replace all PRINT statements in the
program with LPRINT statements.
In the output, forces are in kips or kilonewtons, moments are in kip-feet or
kilonewton-meters, and deformations in inches or millimeters.
,
Matrix-Flexibility Method Chap. 15
358
OUTPUT
Redundants
R(3) = 13.93
Member Forces
Q(1) = 3.84
Q(2) = —5.76
Q(3) = —13.93
Q(4) = 4.80
Q(5)= 4.80
Structure Deformations
D(1) = 0.281
D(2) = 0.223
1150 F1(1,J)=F(I,J)
1160 NEXT J
1170 NEXT I
1180 FOR I EL OmoR
1190 FOR J iy ih, Se0) ay
1200 F21(1,J)=F(I+JW,J)
1210 NEXT J
1220 NEXT I
1230 FOR I = 1 TO JR
1240 FOR J = 1 TO JR
1250 F22(I,J)=F(1+JW, J+JW)
1260 NEXT J
1270 NEXT I
1370
1380 REM MATRIX INV.
1390 REM
1400 N=JR
1410 FOR I=1 TO N
1420 FOR J=1 TO N
1430 S(i4 J )=F22(150)
1440 NEXT J
1450 NEXT I
1460 I=1
1470 NX=N+1
1480 NY=2*N
1490 FOR J=NX TO NY
1500 S(I,J)=1.0
1510 T=I+1
1520 NEXT J
1530 L=1
1540 K=2
1550 XM=S(L,L)
1560 FOR J=L TO NY
1570 S(L,J)=S(L,J)/XM
1580 NEXT J
1590 FOR I=K TO N
1600 X=S(I,L)
1610 FOR J=L TO NY
1620 S(I,J)=S(I,J)-S(L,J)*x
1630 NEXT J
1640 NEXT I
1650 L=L+1
1660 K=K+1
1670 IF (L-N) < O THEN 1550
1671 IF (L-N) = 0 THEN 1550
1673 IF (L-N) > 0 THEN 1680
1680 L=N
1690 LZ=L-1
1700 FOR K=1 TO LZ
1710 I=L-K
1720 Y=S(I,L)
1730 FOR J=L TO NY
1740 S(1,J)=S(I,J)-S(L,J)*Y
1750 NEXT J
1760 NEXT K
1770 L=L-1
1780 IF (L-1) > 0 THEN 1690
Sec. 15.5 Computer Programs 361
2350 D(I)=D(I)+F1(1I,J)*P(J)
2360 NEXT J
2370 D(1I)=D(I)/CONST
2460 NEXT I
2470 PRINT
2480 PRINT " STRUCTURAL DEFORMATIONS "
2485 PRINT
2490 FOR I=1 TO JW
2500 PRINT "D (";I;") = "3D(I)
2510 NEXT I
2520 + «ODATA 1, 3
2530 DATA 30000.0, 0.5
OSLO MEDATAGS 03 7 15.2
2550 DATA 36.06, 0.0, 0.0, 0.0, 0.0
2560 DATA 0.0, 36.06, 0.0, 0.0, 0.0
2570 DATA,O.0.9 0.0),) 2020, 0058020
2580 DATAWOROm Ol OnmO Opus 0). Om O10,
2590 DATAW OH O;mOROnMmO nO) Or0),mes. 000)
2630 DATA 0.60, -0.90, 0.90
2640 DATA -0.60, -0.90, 0.90
2650 DATAS OF Om Ol Ome10.0
2660 DATAR ORSimeOfer Oeuf)
2670 SDATAW O}5 hanOl 2iun— Ona
2710 DATA 8.0, 15.0
2720 END
PROBLEMS
15.1 to 15.3. Construct the flexibility matrix [F] without subdividing the structure into
elements. EJ = constant.
15.1. Wi. 4; Wy, Ag W3, A
wpb bee
W,, A,
15.2
W2, A,
Problems 363
15.4 to 15.9. Subdivide the structure into elements as indicated, and use the trans-
formation [F] = [B]"[f.][B] to form [F]. EZ = constant.
W,,4, W3, 43
gh dyes,
pate ti
eS
15.5. W,,4,
W2, A,
aL Ve
Hy
pe ae Cpa
15.6 W,,A Q3 4
i —— aya am q ¢—— :
L an) fines
aL oftes
W, A, UA a,
hy)
q,
Ne
Ve
364 Matrix-Flexibility Method Chap. 15
15.9. W,, A, G4
A = constant
15.10 to 15.19. Using the indicated element and structure forces, solve for the redundants,
the member forces, and the unknown structure deformations.
W, A,
|
Problems 365
ee aie se
W,, 4, W2, A.
pip
7 7
Aa
{i D _——
2 q G2 3 Q4
R3
ee
EI = constant
15:12;
8k
E = 30 X 103 ksi
I = 72 in4
a. ean Cc
q, 92 3
e—>
94
R, R,
EI = constant
Matrix-Flexibility Method Chap.15
366
15.14. W,
20 KN
——
—~<~—— FR,
qT
4m
Be _
a
q3 G4
2 ale Os
q, (ors
15.15.
20k
12k ———e T
15°
Gable——
}
~
——
| Qo (ES
ee sc
>ON
Rs
EO ey eS
E = 30 X 10° ksi, I,eam = 100 in} Asse
= 0.12 in)
Problems 367
15.16.
45
q) G2 q3 G4
W, ies)
E = 30 x 103 ksi, Toeam = 72 in, Acabie = 0.1 in?
15.17.
a8 <2
_— Cables
R; R,
mh at
q, G2 93 G4 O5 G6
W, W,
15.18. g
ot 2
125
R >
-t = —> 10k nd
8! 93 Y
ERY, 20k
pe
E = 30 x 103 ksi, A = constant = 1 in?
Matrix-Flexibility Method Chap. 15
368
15.19.
Sree
15.20 through 15.24. Use computer program to determine member forces and structure
displacements.
15.20. 60 KN
)
E = 200 x 10°kN/m?, A = constant
= 300 mm?.
Problems 369
5
Ry ‘Ww Us Og
Ry 4%
15.22. = : Pr
oF Ze SO 46 Ai
y MOS ee
}+—— 20°—+|+— 20 ——+}.— 2 — |
:| |"
y
ZAIN
E = 30 X 10° ksi, A = constant = 0.5 in.?.
370 Matrix-Flexibility Method Chap. 15
15.23. 12 | 12 |
oe
kN
rign—p—sn 4
ie
$e
iam =) :
ae, io
E = 200 x 10°kKN/m?, J = 200 x 10° mm‘.
Problems 371
15.24.
OY
a
OY O44
E = 30 X 10°ksi, J = 100in.*.
16
Matrix-Stiffness Method
ANI OVaD: ae
AIR AL AX X TY
MN
The stiffness method, like the flexibility method described in the previous
chapter, is a matrix method for analyzing complex structures with the aid of an
electronic computer. Whereas the flexibility method uses forces as the independ-
ent variables in formulating the load-deformation characteristics of a system, the
stiffness method uses joint deformations as the independent variables. The
stiffness method is consequently also referred to as the deformation method.
Like the slope-deflection and moment-distribution method, the stiffness
method does not involve the concepts of indeterminancy and redundants. As a
consequence, it is far easier to program the stiffness method than the flexibility
method, and the former is the matrix method used almost exclusively today.
In the stiffness method, the forces and deformations in a structure are related to
one another by means of stiffness influence coefficients. For example, the moment
W, acting on the beam in Fig. 16.1a is related to the rotation A, by the expression
W, = K,,A,
W, = KyA, + Ky,A,
Figure 16.1
374
Sec. 16.3 Element-Stiffness Matrix 375
W, = K2,A, + KA,
Rewriting these equations in matrix form, we obtain
lee a ES abel
Ws Ky, Ky ILA,
or simply
[A] = [K]"'[W]
Comparison of this expression with the equation [A] = [F][W] employed in
Chapter 15 indicates that
Lal sly
In other words, the flexibility matrix of a structure is the inverse of the stiffness
matrix, and vice versa.
Let us construct the stiffness matrix for the member shown in Fig. 16.2a. The
vectors in the figure define the forces gq, and q, and the corresponding
displacements 6, and 6, at the ends of the member. They also define the positive
directions of these quantities. We refer to these vectors as element forces and
displacements.
The matrix equation that relates the element forces [q] to the corresponding
Matrix-Stiffness Method Chap. 16
376
a
es oe
(a)
(b)
Figure 16.2
(| * Ee fale
q2 Kx, ka ILd5
To obtain the influence coefficients in the first column of the stiffness matrix, we
note that these terms are the forces q, and q> due to a unit 6,, with 6, equal to
zero. Thus k,, and k3, are the axial forces corresponding to a unit contraction at
the left end of the member, as shown in Fig. 16.2b. These forces have the values
EA EA
ky = ie ky, = Dupe
According to the sign convention defined in Fig. 16.2a,-q, and q> are positive
when they act to the right. Thus k,, is positive and k,, is negative.
Similarly, the second column of the stiffness matrix is obtained by setting
62 = 1 and 6, = 0. As indicated in Fig. 16.2c, the forces corresponding to these
displacements are
kp = © ih , ky = Be
Flexural Element
Let us now construct the stiffness matrix for the flexural element in Fig. 16.3a.
The four forces and the corresponding displacements that will be used to describe
the behavior of the element are defined in the figure. They include the moments,
the shears, and the corresponding rotations and translations at the ends of the
member.
The matrix equation that relates these forces and displacements to one
Flexural element
EI
(d) (e)
Figure 16.3
378 Matrix-Stiffness Method Chap. 16
Kab kyb is
ye)
and
ky, = kay (16.3)
Furthermore, the moment of the M/EI diagram about the left end of the member
is equal to unity. Hence
ka L2L — ky LL
OFT Gea ET Gen
and, in view of Eq. (16.3),
,. Katka _ 12EI
Lia
L ie
and from equilibrium in the vertical direction
we obtain
12EI
ks) = ky = L:
To obtain the second column of the stiffness matrix, we let 6, = 1 and set
the remaining three displacements equal to zero as indicated in Fig. 16.3c. For
this case the area of the M/EI diagram between the ends of the member is equal
to unity, so that epee L :
ori OR Tee
and the moment of the M/EI diagram about the left end is zero, so that
KyLL Kab 2h _
ANRN PT ee
Combining these relations gives
4EI 2EI
Kate ity te Kear
From vertical equilibrium of the member,
Kix = kp
and moment equilibrium about the right end of the member leads to
Comparison of the forces in Fig. 16.3c with the positive directions defined in Fig.
16.3a indicates that all the influence coefficients except k,. are positive. Thus
6EI 4EI] 6EI 2hI
k 12 = L2
Ser 22 St 7
aes: k Be L?
iEOa Kya42 = L
By calculations similar to those above, the influence coefficients for the third
and fourth columns can be obtained using Figs. 16.3d and 16.3e. The results of
these calculations can then be combined with the preceding ones to give the
following element-stiffness matrix:
Equation (16.4) defines the element-stiffness matrix for a flexural member with
constant stiffness EV.
In some flexural structures, such as continuous beams, the ends of an
individual member can rotate but are not able to translate. As a consequence, the
translations 6, and 63 together with their corresponding forces q, and q3 can be
removed from the element-stiffness matrix. The resulting matrix takes the form
q1 1 —1 0 0 6,
qo Ars f 0 0 |16,
AE 1 1
G3] = 4-4 Oe °0 yo ea 53 (16.6)
Oe () ! anes 5
ie V2 ° ayes
Or
(e
54,44
B
WO2%
—_— ——
5, ,Qy A B 55, q2
Element displacements
and forces
(a) (b)
Structure displacements
and forces
(c)
Figure 16.4
Ay
6, 0 0 jy 0 0 A>
b> 1 On woe ( 0 0 Ae
= ’ (16.9)
03 =—cosw sina 0 0 0 0 A,
b,4 0 0 0 0 -cosa@ sina||A-,
382 Matrix-Stiffness Method Chap. 16
or
NK
ne 64=Ag sina
laAn
Fa
Figure 16.5
Sec. 16.5 Direct-Stiffness Method of Forming the Structure-Stiffness Matrix 383
Substitution of Eqs. (16.8) and (16.7) for [W] and [gq] into the above relation
gives
The procedure presented in the previous section, for obtaining the structure-
stiffness matrix, is very similar to the one used in the flexibility method. In both
instances, the force-deflection relationship of an entire structure is obtained from
a matrix containing the force-deflection relationships of all the individual
elements into which the structure has been subdivided. The disadvantage of this
procedure is that it does not readily lend itself to automatic computation. As a
consequence, we will now introduce a slightly different method for obtaining the
structure-stiffness matrix, one that can be entirely automated. This alternative
procedure, which will be used from here on, is called the direct-stiffness method.
In the stiffness method, deformations are used as the independent variables,
384 Matrix-Stiffness Method Chap. 16
Ba oy aie
Sec. 16.5 Direct-Stiffness Method of Forming the Structure-Stiffness Matrix 385
[W], = [Thlk]
AT] LA]n
from which it is evident that
A;
a lhe 0 0 | AG
Oil 0. 0. lae0 LeAG
A,
A,
Wd bnUM Se ps TH
Ae
Using the above transformation matrices and carrying out the operations
Matrix-Stiffness Method Chap. 16
386
[K] into which any term from [K],, goes is determined by the row and column
designation of that term. For example, the four terms in the upper-left corner of
[K]., go into rows 3 and 4 and columns 3 and 4 of [K]. When more than one
element-stiffness term goes into a single slot in the structure-stiffness matrix, the
terms are added.
Combining [K],, and [K],. in the manner described above leads to the
following structure-stiffness matrix:
1 Meetes} 4
5 6
Adair Vad, =4) 0 leea/ Dap W204
VON ER OM V2 rN 2012
[K) = 42 —4 0 40 0 (ni) (16.19)
4L 0 Op 0n0 0 014
= V2 eV 2) OO) m2 V2.5
VO= 20 0 FOE ave 216
It should be noted that, whereas the above procedure for forming [K] is
tedious when carried out by hand, the operation becomes relatively simple when
performed by a computer.
The physical significance of adding terms of the element-stiffness matrices to
obtain terms in the structure-stiffness matrix can be explained as follows. An
external load W applied to any joint of a structure must be resisted by all the
members framing into that joint. Thus each term in an individual [K],, matrix
represents the resistance of member 7 to one of the external forces at its ends,
and the process of combining terms of [K],, matrices to obtain terms in [K] is an
application of the laws of equilibrium at the joints of the structure.
To avoid confusing the various stiffness matrices introduced in this section,
it may help to remember that a lowercase symbol refers to element coordinates
and an uppercase symbol to structure coordinates. In addition, a subscript
denotes that the matrix describes a single member, whereas the absence of a
subscript means that it refers to the entire structure. Thus [Kk], is an element-
stiffness matrix in element coordinates and [K],, is an element-stiffness matrix in
structure coordinates. By comparison, [K] is the stiffness matrix for the entire
structure.
The great advantage of the direct-stiffness method is that the formation of the
structure-stiffness matrix can be completely automated. In this section we will
Matrix-Stiffness Method Chap. 16
388
Member
direction
Member
direction
(b)
Figure 16.6
demonstrate how this is accomplished for a truss. Not only will we construct a
transformation matrix [T],, that is applicable to any axially loaded bar, but we will
also form a member-stiffness matrix [K],, in structure coordinates, that is
likewise generally applicable. j
Let us consider the truss members in Fig. 16.6. To systematize the
formation of the [K],, matrix for members such as these, we adopt the following
procedure:
1. The member is assigned a direction, and 5, and 6,, the member deforma-
tions at the tail end and front end of the member, are assumed to be in this
direction.
2. The x and y structure deformations, at the tail and front end of the member,
are given the designations A;, A;, A,, and A,, respectively. The positive
directions for these vectors are always taken to besto the right and up.
3. The angles between the vectors representing the member deformations and
those representing the x and y structure deformations are designated as
w
Sec. 16.6 Analysis of Trusses by the Direct-Stiffness Method 389
A = cosa, yu = cosB
A;
Iolo oa alfa ey
6, A uw 0 "| Aj
= 16.20
A,
Equation (16.20) gives the transformation matrix [T],, for any axially loaded
bar, regardless of the direction of the member deformations relative to the
structure deformations. The fact that positive structure deformations give rise to
positive or negative element deformations, depending on whether the angle
between the two vectors is smaller or larger than 90°, is taken care of in Eq.
(16.20) by the signs of A and w. For example, the angles a and f are both less
than 90° for the bar in Fig. 16.6a. As a consequence, positive structure
deformations produce positive member deformations, and this is accounted for in
Eq. (16.20) by A and yu both being positive. By comparison, «@ is less than 90° and
B is greater than 90° for the bar in Fig. 16.6b. Accordingly, positive structure
deformations in the x-direction produce positive member deformations, whereas
positive structure deformations in the y-direction lead to negative member
deformations. Again this is taken care of in Eq. (16.20) by A being positive and u
negative.
We have now available to us, in general form, both an element-stiffness
matrix [kK], and a transformation matrix [T],. Using these matrices and the
relation [K], =[T]A[k],[T],, we can derive a member-stiffness matrix in
structure coordinates that is applicable to any axially loaded bar. Thus
eo Penal cot
mae el eee A AS "|
F(T = EB illo OWA
i oj k l
(16.21)
Matrix-Stiffness Method Chap. 16
390
ae eee pomer a
fe ee 0 lL Lah Bal
1.LO p
from which
i j k l
2 Aw A? -Ap] i
VideAE} A he'sSy wm.E aaeteek
eartt ay (16.22)
je eer le rhea} Au k
=u =e Aw pe
Equation (16.22) gives the element-stiffness matrix in structure coordinates
for any axially loaded bar. To construct [K], for a given member, we need only
know the values of A and w for the member and the structure coordinates 1, j, k,
and / at its extremities. The values of A and mw determine those of the terms in
[K],, and i, j, k, and / give the row and column designations of [K],,. The latter
are used to locate the individual terms of [K],, in the stiffness matrix for the entire
structure.
Example 16.1
As an illustration of the direct-stiffness method, let us employ Eq. (16.22) to
construct the structure-stiffness matrix for the truss in Fig. 16.7a.
We begin by numbering the members, choosing directions for them, and
defining the structure coordinates. Next we list in tabular form the data necessary to
form the individual [K],, matrices. Using these data and Eq. (16.22) gives
5 6 1 2
0.16 0:12,- —0:16" —0A27 5
[Kh = AE 0.12 0.09 —0.12 —0.09 | 6
5 —0516) 5-012 0.16 0.12 | 1
SW SOO" OZ 0.09 _| 2
il 2 B) 4
0.12 -—0.16 —0.12 Os1Guleal
[K] = AE —0.16 0.21 0.165 —O0.215172
c 5 (0), 11? 0.16 012° =0.:167}7 3
0.16 -—0.21 —0.16 0.21 _] 4
and
3 4 5 6
0.20 0 3
[K]; = ae : : 5
5 020 aRO 5
0 0 6
Sec. 16.6 Analysis of Trusses by the Direct-Stiffness Method 391
pte fo
Member Length a B r m i j k |
1 20 9936:87-/53.13-910:8 06 5 6 1 2
3 25' ieloye ei f 3 4 5 6
(c)
Figure 16.7
Finally, by placing the individual terms of the above matrices into their correct
positions in the structure-stiffness matrix, we obtain
1 z 3 4 5 6
0:28, .—-0/04 2-012. 0.164 —0.165 0.120111
=<(0):04595 50.30 Hat 031648 =-0.21 G—0. 12 —0:09" | "2
[K] = AE S02 50:16" 0.32" =0.16° 7407207 "*0 3
5 ON6Fi—0.2151-—0.16 ~ .0.21 0 0 4
—0.16 —0.12 —-0.20. 0 (36 lI 2. fe
—0.12,_ —0.09. 0 0 0.12 0.09 6
Matrix-Stiffness Method Chap. 16
392
The stiffness method, like the flexibility method, consists of two parts. In the first,
the
the structure-stiffness matrix is formed out of the stiffness matrices of
individual elements into which the structure has been subdivided, and in the
second part the structure-stiffness matrix is used to determine the deflections,
reactions, and internal forces of the structure. We have dealt with the first of
these two steps in the preceding pages and are now ready to consider the second.
The equation [W] = [K][A] relates the structure loads to the corresponding
structure displacements. To solve this equation for the unknown structure
displacements and for the reactions, it is useful to partition each of the matrices
involved. Thus we write
[6], = [T].[A],
Sec. 16.6 Analysis of Trusses by the Direct-Stiffness Method 393
we obtain
|-S[
ds L
7A wu
aisA alt
7}
% (16.28)
A,
Since q, is always numerically equal to q,, it is only necessary to solve for one of
these forces. Thus K
AE .
[a] =—>-[-A —# A gy] i (16.29)
A,
Because g, was assumed to be a tension force, a positive answer means that the
bar is in tension.
Example 16.2
Determine the reactions and the bar forces for the truss in Fig. 16.8a.
We begin by numbering the members, choosing directions for them, and
Aveaiiinica
E = 30X 109 ksi
pe (a) (b)
(c)
Figure 16.8
394 Matrix-Stiffness Method Chap. 16
3 4 1 2
0.09 =012 —0.99°mr0129%3
AE | —0:12) (0.16) '0,12-7-0.169| 4
Oa sa -0.09 0.12 0.09 -0.12 | 1
0.120 16m Onl e ROTOR a?
1 2 5 6
O21" 016 0 21 = 0 baled
GAE | 0:16" 9012 016 m0 ae
XL<5 |'=021' 016 — oie oneules
—0.16 -0.12 0.16 0.12) 6
Combining the element-stiffness matrices, we obtain the structure-stiffness matrix.
AE : ae :
= |Saem0. see 0 8)ee eG.
q1 = 59 | dir h Pracnd ba
~364.08
1 2 5 6
48.54
AE 1 | —364.08
Pe 8 4) 60.8 ee DG] AAV
q2 = F5 VE 0
0
An alternative way of evaluating the bar forces, which is especially convenient when
using the computer, is to combine the above matrices and eliminate the zero terms.
Thus
010 il »
011 rae¥. Res Sault are 2 is
q2 0.053 0.040 JL —364.08 SYA)
To begin, let us restrict ourselves to structures whose joints can rotate but do not
translate. The continuous beam in Fig. 16.9a is an example of such a structure.
Each joint is free to rotate, but none of the joints is able to translate. Since
G7 Z, gi
Structure deformations
and loads
ta) wm
Element deformations
OE
and loads
(b)
Figure 16.9
396 Matrix-Stiffness Method Chap. 16
(G——5)):
ony ds 86, Ab
ee Figure 16.10
rotations are the only deformations that can take place at the ends of the
individual members as well as at the structure joints, we can limit ourselves to the
element and structure coordinates shown in Fig. 16.9b.
As a first step, let us construct a generally applicable element-stiffness
matrix in structure coordinates. For a flexural element whose end rotations are 6,
and 6, and whose end moments are q, and q,, as shown in Fig. 16.10, the
element-stiffness matrix in element coordinates, according to Eq. (16.5), is given
) ik Aw) al
ka
“|= E alli
16.31
( )
The above relation states that the element rotation at either end
of the member is
equal in magnitude and direction to the joint rotation at the same
end.
Applying the relation [K],, = [T]7 [k],.[7],, which transforms
the element-
stiffness matrix from element to structure coordinates, to
the matrices in Egs.
(16.30) and (16.31) leads to
ij
Ki -=——a F aete
[K],
ZEIT Qari
(16.32)
This is the element-stiffness matrix in structure coordinates for althe flexur
element in Fig. 16.10.
To illustrate the use of Eq. (16.32), let us
form the Structure-stiffness matrix
for the beam in Fig. 16.9. The element-st
iffness matrices for the two members
are
ie |=
NASA Nh a
[K], = — | |
L ie 2a?
Sec. 16.7 Analysis of Flexural Structures by the Direct-Stiffness Method 397
and
2 3
2EI {1 ONSa ee
[K], aes
IL, \URS) al 3
Placing each term of these element-stiffness matrices into its appropriate position
in the structure-stiffness matrix, as determined by the row and column designa-
tion of the term, we obtain
dion 8
pall 2enth WO 6th
Be or5| Tes O55) 2
0-05 14-1 3
To obtain a relation for calculating the member forces in a flexural
structure, we start with Eq. (16.27),
[a], = [KT],tAl,
and substitute the matrices in Eqs. (16.30) and (16.31) for [kK], and [T],,. This
leads to
Heele elt
LeeLee alla|
AS eon).
Example 16.3
Determine the member forces for the structure in Fig. 16.11a.
We begin by numbering the members, choosing directions for them, and
defining element and structure coordinates (Fig. 16.11b). Next we construct a table,
as shown in Fig. 16.11c, that contains the data needed to form the [K],, matrices.
Using these data and Eq. (16.32), we obtain
a5 2 Sea
0.8 0.4] 3 0.4 0.27 2
[K}; = E19, a D IK}, = Ei), wi 4
‘hs Woe
0.8 0.47 1
K], = EI
[K]; a, se 2
{)) @
5
Structure coordinates Element coordinates
(b)
Member Length i j
1 5m S 22
2 10-m 2 4
3 5m 1 2
(c)
Figure 16.11
a 0 04 08 0 4.44
q2 0 08 04 0 | [—s5.56 8.89
qs 0 04 0 o2|| ata 4.44
AP upuatee a vibe Pees
qs| [08 04 0 Oo 0 4 0
de | mld £0.88 OED 6.67
Sec. 16.7 Analysis of Flexural Structures by the Direct-Stiffness Method 399
Up to now we have assumed all structure loads to be applied only at the joints or
nodal points. Although this is satisfactory for trusses, it is obviously not the case
for beams and frames. The latter are usually subjected to concentrated or
distributed loads acting along the spans of the individual members. One way of
dealing with this situation would be to subdivide each of the members on which
intermediate loads are acting into a number of smaller elements and then
approximate the given loading by a series of concentrated loads at the newly
formed nodal points. However, this solution to the problem is very inefficient,
and a more direct way of dealing with the situation will consequently be
introduced.
The procedure that we will employ for dealing with loads acting between
nodal points is basically identical to the one we have already used in both the
slope-deflections and moment-distribution methods. The element forces [q] are
V4 Vi
a. aS
Se a Srp
q
Actual structure
(a)
OF C ) GF
Artificially
restrained
structure
(b)
Equivalent
joint loads
(c)
Figure 16.12
400 Matrix-Stiffness Method Chap. 16
1. The element forces that are caused by the applied loads acting on a
structure all of whose joints are assumed to be fixed.
2. The element forces due to the joint rotations and translations that actually
take place.
For example, let us consider the continuous beam in Fig. 16. 12a, which has
a uniformly distributed load acting along one of its members. First, we artificially
fix all the joints of the structure as shown in Fig. 16.12b. This necessitates the
application of fictitious structure forces [W;]. The element forces existing in this
structure are what we used to refer to as fixed-end moments and will now call
[qr]. Since the artificial forces [W,] do not exist in the actual structure, we now
introduce a second system, shown in Fig. 16.12c, to which we apply structure
loads [W,] that are equal and opposite to the forces [W;]. We use the symbol [q<]
to refer to the member forces produced by [Wz].
Since the actual structure, in Fig. 16.12a, can be obtained by adding the
structures in Figs. 16.12b and 16.12c, it follows that the member forces [q] in the
actual structure are given by the relation
—25
= ee —25
@el=) aos | Wel = EE
1255
2. The structure loads [W,] are equal and opposi
te to [W,]. Hence
Wl |tas]
To calculate the member forces [ge] corre
sponding t6 these loads, we follow the
procedure developed in the first part
of this section.
Sec. 16.7 Analysis of Flexural Structures by the Direct-Stiffness Method 401
20 kN 20 kN
A B
LG;
“74 ra 7
Ie 5m ar 5m 2.5m 2.5m
(a)
A,,W, 4,, W2
Bie | ey oe Ms
me 59, q2
ore
yee ie
Element and structure
coordinates
(b)
Ge)
ap, =-25 @ de> = 25 semana) ey 2s
He)
(0) 25 Ay 12.5
Wr, = 25
Weo= 12.5
Fixed-end member forces
and structure loads
(c)
1% We, = 25 BEN We2= 712.5 y
0, a .
(d)
Figure 16.13
we obtain
Solving for the unknown joint displacements,
A, 1 [0.4 ele 25 put al
A ~ El ie 1.2) |=105)" Br =22773
The member forces [gz] are calculated next.
e by adding [ar]
Finally, we determine the member forces [q] in the actual structur
and [q,]. Thus
q1 SSL aia b 25.0 0
q |_| 25.0), 5.68 |_ | 30.68
q, | | —12.5 —18.18 —30.68
qs +1245 —9.09 3.41
Let us now consider flexural structures that contain joints that can translate as
well as rotate. The frame in Fig. 16.14a is an example of such a structure. Joints
B and C are free to translate as well as rotate.
To analyze this frame, we can use the element stiffness matrix given by Eq.
(16.4). However, if we employ this matrix, we must limit ourselves to the nine
joint displacements shown in Fig. 16.14b. Since axial deformations were
neglected in the development of the stiffness matrix, the lateral displacement of
joint C must always be equal to the lateral displacement of B. Hence only one
independent horizontal joint displacement, A,, can be defined for member BC.
Likewise, the vertical displacements of B and C must always be equal to the
vertical displacements of A and D, and only a single vertical joint displacement
can be defined for each of members AB and DC. If we attempted to define an
additional horizontal joint displacement at C or additional vertical joint displace-
ments at B and C, they would not be independent variables and could therefore
not be included.
Whereas the frame in Fig. 16.14a can be analyzed using the stiffness matrix
given by Eq. (16.4) and the structure deformations defined in Fig. 16.14b, such an
analysis would suffer from a significant disadvantage; the direct stiffness method
could not be used. The cornerstone of the direct-stiffness method is the presence
of an element-stiffness matrix in structure coordinates that is applicable to each
member of the structure. For example, Eq. (16.22) gives the generally applicable
element-stiffness matrix in structure coordinates for truss members. If we were to
Sec. 16.8 Analysis of Flexural Structures With Joint Translations 403
P
B C
(b)
Figure 16.14
review the derivation of this matrix, it would be apparent that its formation
depends on each member of the truss having the same number and types of
structure deformations at its ends. This requirement is satisfied for all trusses.
There always exist independent vertical and horizontal joint displacements at the
extremities of each member. However, this criterion is not satisfied for the
structure in Fig. 16.14b. Here some joints have three degrees of freedom, some
have two, and some only one. It would thus not be possible to construct a
generally applicable element-stiffness matrix in structure coordinates for this
frame, and we could consequently not employ the direct-stiffness method.
Instead, we would have to develop separate transformation matrices for each
member and then use them to transform the element-stiffness matrix given by Eq.
(16.4) from element to structure coordinates for each member, one at a time. The
great disadvantage of doing this is that the formation of the structure-stiffness
matrix cannot be automated, as is the case when a generally applicable
element-stiffness matrix in structure coordinates exists. For hand calculations the
404 Matrix-Stiffness Method Chap. 16
There exist two types of axial deformations: (1) axial deformations caused by
axial loads and (2) axial displacements of the ends of a member resulting from the
bending deformation of the member. Of these two, we will only consider the
former. This is sufficient for our purpose and avoids the difficulties that would
arise from the use of the latter deformation, which involves nonlinearities.
We have already derived the element-stiffness matrix that relates axial loads
to axial displacements. It is given by Eq. (16.2). Since these axial deformations
are completely independent of the bending deformations in the flexural stiffness
matrix given by Eq. (16.4), the matrix we are attempting to form can be
constructed by simply combining the aforementioned two matrices. Thus we
obtain, for the six element forces and deformations depicted in Fig. 16.15, the
following relation.
AE AE
_ ae 0 ee
q1 L 0 - Pas 0 0;
12EI 6EI 12
q2 0 5
Ls
sea
is
(One ue L
eosLl? 2
gf || Boge
a
SEL neeSET:
6EI
L
4EI
ib,
suo, tow CELE la Hm
L2 ‘- 3
Z AE . : AE (16.35)
q4 if Te reals os Wiha
ds 0 I2ZEion ae
eae OEE 0 12EI 6EI i
L L La L? 5
PPA in ee6EI
L
aE
2EI
ih,
apes L?
Fee dh|.
L 6
Sec. 16.8 Analysis of Flexural Structures With Joint Translations 405
8), G5 Eee
, :
Figure 16.15
member
direction
’ Figure 16.16
406 Matrix-Stiffness Method Chap. 16
i J k l m n
Detar or(aces Ad seat a
D L ii di
Pigtiei20 6f-—1274 127k eee
ES Le) 12 “7a eee
a! (Te 77 ey,
L? ji Ve ie .- Ta ae
[aah Sys tee yy : a Ail (16.37)
ees yh 5 oe
Ile 1A 12ius aie es
i Le Oe oj
6lu OIA 6lu* 61 ay
Le EL) eae ee
Sec. 16.8 Analysis of Flexural Structures With Joint Translations 407
and
l J k l m on
GC; i
Ca CG; j
Pes | ee Cae Ce k (16.38)
=i —=Cs —C, a C, l
C, —C; C; —C, Cs Cs n
where
AEX 12EI
Ci oo L a je wu
AE 12EI
OE lee areal
AEw* 12EI
Cr + ——j?
4 if iP
6Elu 6EIA
C4 = L2 , Sti) [2
4EI 2EI
Co = —,; C———
agile lb ey
Equation (16.38) gives the element-stiffness matrix in structure coordinates
for a flexural member with axial deformation. To determine the individual terms
in [K],, it is only necessary to know the values of E, J, A, A, and w for the
member. The structure coordinates i, j, k, 1, m, and n at the extremities of the
member give the row and column designations of [K],. When the individual
element-stiffness matrices are combined to form the stiffness matrix [K] for the
entire structure, these row and column designations are used to determine the
location of each term in the structure-stiffness matrix. Once the structure-stiffness
matrix has been formed, the procedure for determining the unknown structure
deformations, reactions, and member forces is identical to that employed
previously in the analysis of trusses and beams.
Example 16.5
Determine the unknown structure deformations, reactions, and member forces for
the structure in Fig. 16.17a.
Since none of the joints in the given structure can translate, we could analyze
the structure using the 2 X 2 stiffness matrix given by Eq. (16.32). However, our
purpose is to illustrate the use of the more general stiffness matrix given by Eq.
(16.38), and we will therefore use it to solve the problem.
Element deformations
(b)
Figure 16.17
408
Sec. 16.8 Analysis of Flexural Structures With Joint Translations 409
First we number the members, choose directions for them, and define the
element and structure coordinates. These are shown in Figs. 16.17b and 16.17c. Next
we construct the table, given in Fig. 16.17d, that contains the data needed to form
the [K], matrix for each member. Using these data and Eq. (16.38), we obtain the
following matrices. The units we will use are kips and inches.
4 5 6 1 2 3
20.833 0 1250 20.833 0 1250 | 4
0 125(egy oe c0 Our @1250 $04 | 5
1250 0 100,000 -1250 0 50,000] 6
eT i evo chao Raye Orit Paes FL Pe
Ds a, +1250. 0 0 bie Oe
1250 0 50,000 -1250 0 100,000] 3
1 2 3 7 8 9
625 0 betas at me ae
0 2.604 -312.5 0 2.604 —3125| 2
O31 5 50 N00 es 31S 25,000 |3
[Kh =| _ és 0 0 625 0 Gaela7
0 =2,604 3125 06 2.604 312.5] 8
0 -312.5 25,000 0 312.5 50,000| 9
Combining the two element-stiffness matrices, we obtain the following structure-
stiffness matrix.
1 2 3 = 5 6 4 8 9
645.8 0 —1250 + —20.833 0 — 1250) 3-625 0 0 1
0 1252.6 Sk Wa, 0 =1250 0 0 = 2.004) 9—312:3° |2
This section contains FORTRAN and BASIC computer programs for analyzing
two-dimensional trusses and frames by the stiffness method. The programs are
written in FORTRAN-77 and GWBASIC. Each program is accompanied by an
illustrative example, which explains how the program is to be used. The diskette
enclosed with the book can be used to run the BASIC programs on a personal
computer.
The following programs determine the reactions, member forces, and structure
deformations for a plane truss. In their present form the programs are limited to
Sec. 16.9 Computer Programs 411
trusses all of whose members have the same cross-sectional area. However, a
simple alteration of the programs can remove this limitation. The illustrative
problem following the programs explains what input data are required and how
these data are to be entered into the programs.
00610 AK(L,1)=AK(L,1)-D
00620 AK(L,J)=AK(L,J)-E
00630 AK(L,K)=AK(L,K)+D
00640 AK(L, L)=AK(L,L)+E
00650 AKT (IJ, 1)=B*(-SLAM)
00660 AKT(IJ,J)=B*(-SMU)
00670 AKT (IJ ,K)=B*SLAM
00680 AKT (IJ, L)=B*SMU
00690 TJzlJe1
00700 20 CONTINUE
00710 DO 30 I=1,NK
00720 DO 30 J=1,NK
00730 30 AK11(I,J)=AK(1,J)
00750 DO 35 I=1,NU
00760 DO 35 J=1,NK
00770 35 AK21(1, J)=AK(1+NK, J)
00780 CALL MATINV(AK11, NK)
00790 CALL MATMULT(AK11,W,DEF,NK, NK)
00800 CALL MATMULT(AK21,DEF,R,NU,NK)
00810 DO 22 I=1,NM
00820 DO 22 J=1,NK
00830 22 AKT1(I,J)=AKT(I,J)
00840 CALL MATMULT(AKT1,DEF,Q,NM,NK)
00850 PRINT 100
00860 DO 210 I=1,NU
00870 210 PRINT 880, I+NK,R(I)
00880 PRINT 300
00890 DO 220 I=1,NM
00900 220 PRINT 890,1,Q(1)
00910 DO 66 I=1,NK
00920 DO 66 J=1,1
00930 DEF(1)=X*DEF(I)
00940 66 CONTINUE
00950 PRINT 500
00960 DOw230 0-1 NK
00970 230 PRINT 900,1,DEF(I)
00980 100 FORMAT(///,5X, ‘OUTPUT’ ,//,5X, REACTIONS’ )
00990 200 FORMAT (5X, F7.2)
01000 300 FORMAT(/,5X, MEMBER FORCES’)
01010 500 FORMAT(/,5X,’STRUCTURAL DEFORMATIONS’)
01020 800 FORMAT(/,10X, INPUT’ ,//,5X,’TYPE OF UNITS -IU = ’,11,/)
01030 810 FORMAT(/,5X,’A = ',F6.2,’, E o> FSO
01040 820 FORMAT(/,5X,’NM = ',12,’, ND Sa Satz ae NK =
01050+’, NU = ',12,//,5X,'W MATRIX’)
01060 830 FORMAT(/,5X, NODAL DISPLACEMENTS I > J; K, Ea 6 Ds
01070+' AND NODAL COORDINATES CI, CJ, CK, CL’)
01080 840 FORMAT(5X,’MEMBER /,1I2)
01090 850 FORMAT(5X,’I = ’,12,’, aJ°at Ay e7 ane
? K
’,12)
01100 860 FORMAT(5X,’CI = ',F6.1,", CS. SEM POILEts CK
01110+’, Cleared)
01120 880 FORMAT(5X,’R(',12,’) rE? oD)
01130 890 FORMAT(5X,’Q(',12,’) = ',F7.2)
01140 900 FORMAT(5X,'D(',12,’) = ',F7.2)
01150 STOP *
01160 END
01170 SUBROUTINE MATMULT(A,B,C,IA,JA)
Sec. 16.9 Computer Programs 413
1520 NEXT J
1530 1 eal
1540 Keae2
1550 XM = S(L, 15)
1560 KORG mm emOlin
1570 SChod)) = SGb5s) Psu
1580 NEXT J
1590 FOR Gea miko Om
1600 oe SCaIb)
1610 HORS mem lel OMNY,
1620 Eos) Ss SCE pM SO bash 2
1630
1640
1650 L
1660 K
1670 IF L < N THEN GOTO 1550
1672 IF L = N THEN GOTO 1550
1674 IF L > N THEN GOTO 1680
1680 Leo N
1690 2 =) Jee
1700 FOR ewe TO) bz
1710 TP ontK
1720 Veeescle ts)
1730 FOR J = L TO NY
1740 SCL e s (Is) ESC yey
1750 NEXT J
1760 NEXT K
1770 pes ties
1780 IF L > 1 THEN GOTO 1690
1782 IF L = 1 THEN GOTO 1790
1794 IF L < 1 THEN GOTO 1790
1790 FOR I = 1 TON
1800 FOR J = 1 TON
1810 NJ = NX+J-1
1820 AK11(1I,J) = S(I,NJ)
1830 NEXT J
1840 NEXT I
1850 FOR I=1 TO NK
1855 D(I)=0.0
1860 FOR J=1 TO NK
1870 D(T)=D(I)+AK11(1,J)*W(J)
1880 NEXT J
1890 NEXT I
1900 FOR I=1 TO NU
1905 R(I)=0.0
1910 FOR J=1 TO NK
1920 R(T)=R(1)+AK21(1,J)*D(J)
1930 NEXT J
1940 NEXT I
1950 FOR I=1 TO NM
1960 FOR J=1 TO NK
1970 AKT1(I,J)=AKT(I,J)
1980 NEXT J
1990 NEXT I
2000 FOR I=1 TO NM x
2005 Q(I)=0.0
2010 FOR J=1 TO NK
Sec. 16.9 Computer Programs 417
2020 Q(1)=Q(1)+AKT1(I,J)*D(J)
2030 NEXT J
2040 NEXT I
2050 PRINT
2051 PRINT
2052 PRINT
2054 PRINT " OUTPUT"
2056 PRINT
2060 PRINT "REACTIONS"
2070 FOR I=1 TO NU
2080 PRINT USING "R(##"3;T4NK;
2083 PRINT USING ") = ###.##";R(I)
2090 NEXT I
2100 PRINT
2110 PRINT "MEMBER FORCES"
2120 FOR I=1 TO NM
2130 PRINT USING "0(##";T;
2133 PRINT USING ") = ###.##";Q(I)
2140 NEXT I
2142 FOR I=1 TO NK
2144 D(I)=D(I)/CONST
2146 NEXT I
2150 PRINT
2160 PRINT "STRUCTURAL DEFORMATIONS"
2170 FOR I=1 TO NK
2180 PRINT USING "D(##";I1;
2184 PRINT USING ") = ###.##";D(I)
2190 NEXT I
2500 DATA 3
ZL DATA 0.5, 30000.0
2520 DATASS? 8 Ghy 2h
2525 DAT 6 005-15 aD NOLOsa 0.0
2530 DATAs). 072102, 7020.00,0, a30.407 920.0
2540 DATA. 16 2) 3.) Soar 00. 6000! 0
2550 DATAVE, 6; 17, 30.0, 0,0." 30-055 2050
2325 DATAGT APR), [4s 86050 805 050.50) 0:0
2560 DATAT 47: (613 5055 53000, 020860..09-020
2690 END
Example 16.6
Using one of the two preceding programs, analyze the truss in Fig. 16.18a.
We begin by numbering the members, choosing directions for them, and
defining the structure deformations, as shown in Figs. 16.18b and c. As was done in
the earlier sections of the chapter, it is necessary when numbering the structure
deformations to begin with the unknown deformations.
Next we list in tabular form, as shown in Fig. 16.18d, the four structure
displacements
J, J, K, andL associated with each member, and the x- and
y-coordinates of the joints at the extremities of each member.
The input data needed to run the FORTRAN program are identical to those
required for the BASIC program. Both sets of data have the same number of lines,
and corresponding lines contain identical information. The only difference is that the
FORTRAN program requires the formation of a separate data file, whereas the
input information for the BASIC program is stored in the DATA statements located
at the end of the program.
Matrix-Stiffness Method Chap. 16
418
15 k
8k al E = 30 x 10°ksi
A=05 in.
(constant for
00’ all bars)
(b) (c)
_ Structure Joint
ope coordinates
3
0.5, 30000.0
5,8,4,4
8.0, —15.0, 0.0, 0.0
7,8, 1,2, 0.0, 0.0, 30.0, 20.0
1,2, 3,5, 30.0, 20.0, 60.0, 0.0
4,6, 1, 2,30.0, 0.0, 30.0, 20.0
7,8, 4, 6, 0.0, 0.0, 30.0, 0.0
4, 6, 3, 5, 30.0, 0.0, 60.0, 0.0
Reactions
R(5) = 3.20
R(6) = 13.93
R(7) = —8.0
R(8) = —2.13
Member Forces
Q(1) = 3.84
Q(2) = -5.77
Q(3) = —13.93
Q(4) = 4.80
Q(5) = 4.80
Deflections
A, = 0.282
A, = —0.223
A, = 0.230
A, = 0.115
The following programs determine the reactions, member forces, and structure
deformations for a plane frame. The illustrative problem following the program
explains what input data are required and how these data are to be entered into
the programs.
Sec. 16.9 Computer Programs 421
00650 AK(J,K)=AK(J,K)-C5
00660 AK(J,L)=AK(J,L)-C2
00670 AK(J,M)=AK(J,M)-C3
00680 AK(J,N)=AK(J,N)-C5
00690 AK(K,1)=AK(K,1I)+C4
00700 AK(K, J)=AK(K,J)-C5
00710 AK(K,K)=AK(K,K) +C6
00720 AK(K,L)=AK(K, L)-C4
00730 AK(K,M)=AK(K,M)+C5
00740 AK(K,N)=AK(K,N)+C7
00750 AK(L,1I)=AK(L,I)-C1
00760 AK(L, J)=AR(L, J)=C2
00770 AK(L,K)=AK(L,K)-C4
00780 AK(L, L)=AK(L, 1) +C)
00790 AK(L,M)=AK(L,M)+C2
00800 AK(L,N)=AK(L,N)-C4
00810 AK(M,1)=AK(M,1)-C2
00820 AK(M,J)=AK(M,J)-C3
00830 AK(M,K)=AK(M,K)+C5
00840 AK(M, L)=AK(M, L)+C2
00850 AK(M,M)=AK(M,M)+C3
00860 AK(M,N)=AK(M,N)+C5
00870 AK(N,I)=AK(N,I)+C4
00880 AK(N,J)=AK(N,J)-C5
00890 AK(N,K)=AK(N,K)+C7
00900 AK(N,L)=AK(N, L)-C4
00910 AK(N,M)=AK(N,M)+C5
00920 AK(N,N)=AK(N,N)+C6
00930 T1=A*E*ALA/ALG
00940 T2=A*E*AMU/ALG
00950 T3=(12.0*EXAI*ALA) /ALG**3
00960 T4=(12.0*E*AI*AMU) /ALG**3
00970 T5=(6.0*E*AI*ALA)/ALG/ALG
00980 T6=(6.0*E*AI*AMU)/ALG/ALG
00990 T7=(6.0*EXAI ) /ALG/ALG
01000 T8=(4.0*E*AI)/ALG
01010 T9=(2.0*E*AI)/ALG
01020 AKT(IX+1,1)=AKT(IX+1,1)+T1
01030 AKT(IX+1,J)=AKT(IX+1,J)+T2
01040 AKT(IX+1,K)=AKT(IX+1,K)
01050 AKT(IX+1,L)=AKT(IX+1,L)-T1
01060 AKT(IX+1,M)=AKT(IX+1,M)-T2
01070 AKT (IX+1,N)=AKT(1IX+1,N)
01080 AKT(IX+2,1)=AKT(IX+2,1)-T4
01090 AKT(IX+2,J)=AKT(IX+2,J)+T3
01100 AKT (1X+2,K)=AKT(1X+2,K)-T7
01110 AKT (IX+2,L)=AKT(IX+2,L)+T4
01120 AKT (IX+2,M)=AKT(IX+2,M)-T3
01130 AKT(IX+2,N)=AKT(IX+2,N)-T7
01140 AKT(IX+3,1)=AKT(IX+3,1)+T6
01150 AKT (IX+3, J) =AKT(IX+3,J)-T5
01160 AKT(IX+3,K)=AKT(IX+3,K)+T8
01170 AKT(IX+3, L)=AKT(IX+3,L)-T6
01180 AKT(1X+3,M)=AKT(IX+3,M)+T5
01190 AKT(1X+3,N)=AKT(IX+3,N)+T9
01200 AKT(IX+4,1)=AKT(IX+4,1)-T1
01210 AKT(1X+4,J)=AKT(IX+4,J)-T2
Sec. 16.9 Computer Programs 423
02340 RETURN
02350 END
1 REM ** kkRk kk kk kkk
850 AK(N,J)=AK(N,J)-CS
860 AK(N,K)=AK(N,K)+C7
870 AK(N,L)=AK(N, L)-C4
880 AK(N,M)=AK(N,M)+C5
890 AK(N,N)=AK(N,N)+C6
900 T1=A*E*ALA/ALG
910 T2=A*EXAMU/ALG
920 T3=12.0*E*AI*ALA/ALG/ALG/ALG
930 T4=12.0*E*AI*AMU/ALG/ALG/ALG
940 T5=6.0*E*AI*ALA/ALG/ALG
950 T6=6.0*E*xAI*AMU/ALG/ALG
960 T7=6.0*E*AI/ALG/ALG
970 T8=4.0*E*xAI/ALG
980 T9=T8/2.0
990 AKT (1IX+1,1)=AKT(1X+1,1)+T1
1000 AKT (IX+1,J)=AKT(IX+1,J)+T2
1010 AKT ( 1IX+1,K)=AKT(IX+1,K)
1020 AKT (IX+1, L) =AKT(IX+1,L)-T1
1030 AKT (IX+1,M)=AKT(1X+1,M)-T2
1040 AKT (IX+1,N)=AKT(IX+1,N)
1050 AKT (IX+2,1)=AKT(1X+2,1)-T4
1060 AKT (IX+2, J) =AKT(IX+2,J)+T3
1065 AKT (1X+2,K)=AKT(IX+2,K)-T7
1070 AKT (1X+2,L)=AKT(1X+2,L)+T4
1075 AKT (IX+2,M)=AKT(1IX+2,M)-T3
1080 AKT(IX+2,N)=AKT(IX+2,N)-T7
1085 AKT (IX+3, 1) =AKT(IX+3,1)+T6
1090 AKT(IX+3, J)=AKT(IX+3,J)-T5
1095 AKT (IX+3,K)=AKT(IX+3,K)+T8
1100 AKT (1X+3, L)=AKT(1X+3,L)-T6
1105 AKT (IX+3,M)=AKT(1X+3,M)+T5
LO AKT (IX+3,N)=AKT(IX+3,N)+T9
ial; AKT (1X+4,1)=AKT(1X+4,I)-T1l
1120 AKT (1X+4,J)=AKT(1X+4,J)-T2
1125 AKT(1X+4,K)=AKT(1X+4,K)
1130 AKT (IX+4, L) =AKT(IX+4,L)+T1
135 AKT (1X+4,M)=AKT(IX+4,M)+T2
1140 AKT (IX+4,N)=AKT(IX+4,N)
1145 AKT (1IX+5,1)=AKT(1X+5,1)+T4
1150 AKT(1X+5,J)=AKT(1X+5,J)-T3
1155 AKT (1X+5,K)=AKT(1X+5,K)+T7
1160 AKT(1IX+5,L)=AKT(1IX+5,L)-T4
1165 AKT (IX+5,M)=AKT(1X+5,M)+T3
1170 AKT(1IX+5,N)=AKT(1X+5,N)+T7
1175 AKT (IX+6, I) =AKT(IX+6,1)+T6
1180 AKT (1X+6,J)=AKT(IX+6,J)-TS
1185 AKT (IX+6,K)=AKT(1X+6,K)+T9
1190 AKT (1X+6, L)=AKT(1X+6,L)-T6
TOS AKT (IX+6,M)=AKT(IX+6,M)+T5
1200 AKT (1X+6,N)=AKT(1X+6,N)+T8
1205 NEXT NI
1210 FOR I=1 TO NK
1220 FOR J=1 TO NK
1230 AKPT(1I,J)=AK(I,J)
1240 NEXT J
1250 NEXT I
1370 REM
428 Matrix-Stiffness Method Chap. 16
2690 END
430 Matrix-Stiffness Method Chap. 16
Example 16.7
Using one of the two preceding programs, analyze the frame in Fig. 16.19a.
We begin by choosing directions for the elements, numbering the element
displacements, and numbering the structure displacements, as shown in Figs. 16.19b
and c. As was done in the earlier sections of the chapter, it is necessary when
numbering the structure deformations to begin with the unknown deformations.
Next we list in tabular form, as shown in Fig. 16.19d, the six structure
displacements J, J, K, L, M, and WN associated with each member, and the x- and
y-coordinates of the joints at the extremities of each member.
The input data needed to run the FORTRAN program are identical to those
required for the BASIC program. The only difference is that the FORTRAN
program requires the formation of a separate data file, whereas the input
information for the BASIC program is stored in the DATA statements located at
the end of the program.
6k
—— E = 30 x 10°ksi
| = 130 in.*
A=5 in?
; (I and A are
12 constant for
all members)
; 30° —|
(a)
dg
544
84
6, 65
;
: 840 O16
5, :
5 12
— 5;
. ~—— 5,7
18
~— 5,
ry)15 beg 814
Element directions
and displacements
6,
(b)
r)13
Figure 16.19
Sec. 16.9 Computer Programs 431
Ag
| Ag Ato
Structure displacements
(c)
_ Structure Joint
displacements coordinates
Peed
verse [Ta [xe [ow] [on[on[onyom|
Sa a fea eed bea AT
ete WT PA PE Pe | 0 | 12 | 30 | 12|
taHie a4 O.Ladtoboteafat beBulefit158Qe|ley bin8 [212
Structure displacements and joint coordinates
associated with each member
(d)
Figure 16.19 (continued)
5
30000.0
3,12, 0.10
6.0, 0.0, 0.0, 0.0, 0.0, 0.0
IB0/0PS10 7705.95 1253
0.0, 0.0, 0.0, 12.0
EXO) 0S RP Se Be)
0.0, 12.0, 30.0, 12.0
130.0, 5.0, 10, 11, 12, 4, 5, 6
30.0, 0.0, 30.0, 12.0
432 Matrix-Stiffness Method Chap. 16
eee
4s
In the solution, forces are in kips, inch-k
ips, kilonewtons, or kilonewton-
meters and deformations are in inches
or millimeters.
Problems
433
Reactions
R(7) = —3.02
R(8) = —0.85
R(9) = —281.97
R(10) = —2.98
R(11) = 0.85
R(12) = —277.26
Member Forces
Q(1) = —0.85
Q(2) = 3.02
Q(3) = —281.97
Q(4) = 0.85
Q(5) = —3.02
Q(6) = —153.06
Q(7) = 2.98
Q(8) = —0.85
Q(9) = 153.06
Q(10) = —2.98
Q(11) = 0.85
Q(12) = 151.71
Q(13) = 0.85
Q(14) = 2.98
Q(15) = —277.26
Q(16) = —0.85
Q(17) = —2.98
Q(18) = —151.71
Defiections
D(1) = 0.3641
D(2) = 0.0008
D(3) = 0.0024
D(4) = 0.3570
D(5) = —0.0008
D(6) = 0.0023
PROBLEMS
16.1. Using the method outlined in Section 16.4, form the [K] matrix for the truss shown
in the figure. Employ Eq. (16.11), [K] = [T]’[k.][T], where [7] is the transforma-
tion matrix for the entire structure and [k,] is the composite element-stiffness
matrix.
Matrix-Stiffness Method Chap. 16
434
pat os
~~ Element
deformations
54
~
55, —<—____ ~<—
56 55
15’
Ay
Structure
—> A, deformations
a+ fo
— A,
16.2. Form the [K] matrix for the truss in Problem 16.1, using the procedure outlined in
Section 16.5. A separate transformation matrix [T],, is formed for each member,
and Eq. (16.16), [K],, = [T];[k],[T],, is used to construct a [K],, matrix for each
element. [K] is obtained by combining the separate [K],, matrices.
16.3. Form the [K] matrix for the truss in Problem 16.1, using the procedure outlined in
Section 16.6 and illustrated in Example 16.1. Equation (16.22) is used to construct
the element-stiffness matrices, which are combined to form [K].
16.4 through 16.8. Use the direct-stiffness method, outlined in Section 16.6 and illustrated
in Example 16.2, to determine the unknown structure displacements and reactions and the
element forces.
16.4.
ee Element
deformations Structure
deformations
A = constant = 1 in?
E = 30 x 103 ksi
435
Problems
60 kN
16.5.
B |
Structure
Element deformations
A
|
deformations
ie
Ze A, ee 41
8 Ag
5,
ete
\
E = 200 x 10°kN/m’, A = constant = 200 mm?”
16.6.
5,
|
~ Element
> aeformations
5,
7,A™s 5
Structure
deformations
A = constant = 2 in?
E = 30 x 103 ksi
436 Matrix-Stiffness Method Chap. 16
16.7.
Element
deformations
ES,
a
Structure
deformations
ares
54
x | WA 5
81. 5, °8s
As 4s A,
fa fa fa
—p A,
Structure Ceteepsuore
16.9 through 16.11. Use the 2 x 2 element-stiffness matrix given by Eq. (16.32) to
determine the element forces.
10°
8, 5, 8, 8, i rh s
Casa ) EG _)
ON; 5
oe ay)
Element deformations Structure deformations
I = constant
16.10.
Ca) WN os
eee A,
56 Wy) 43
. Structure deformations
Element deformations
I = constant
Matrix-Stiffness Method Chap. 16
7, Yi
A D
|+—— 30 ——+« — 20’
ee (ean
ND,
ry wy) 5g Elemen t
deformations
A, A> As
2)
A; A,
Structure
deformations
I = constant
16.12 through 16.15. Use equivalent joint loads to determine the member forces.
16.12. 20 k
@ C
5, 8 8, 34 A, Ay Ag
—-—-——) Oe)
Element deformations Structure win,
I = constant
Problems 439
16.13. 30k
B
|~«—— 1g +112’ 20! +
55 53 54 A3 A, A>
by
€ Ga 2) (aed)
Element deformations Structure deformations
I = constant
5, 5,
80 kN €
~—— S
53
|
p =]
40 kN (1) 54
cane
5 3
bs
Structure deformations
I = constant
16.15. 1.5 Kitt
A,
16.16 through 16.20. Use computer program to determine member forces, reactions, and
structure displacements.
16.16. 80 kN 50 kN
| |
i
——>30 KN
i
/}—— 8m ——+.__ 8m —-|
@ @)
(4) Element
7) a (3) directions
Ag ()
5 A2 Ay
Ayo Structure
[ deformations
16.17 10 k
8k —— | ¢
15°
ny.
ae a
© 3
Se Pay
directions
@) @) @
x Seen Structure
{ t deformations
ease enn
or ame ae Ag
E = 30 X 10°ksi, A = constant = 0.5 in.”
442 Matrix-Stiffness Method Chap. 16
16.18. 60 |
30 kN —> oS
(oe) =
+ pa
Element
directions
Structure
deformations
16.19. 20 k
ate eee oe
545
b, ae h 840 ( h 516 859
:: iot TS 8 12 8 13 18 a4 8oq
Element deformations
niko! A
Structure deformations
16.20. 2k— a
15°
wiyin.
|—— 2. 40 E pee |
55 or
8, 57 . | 519 516
3g 849 |
543
—— 6 17
bs 5g
Element directions
and displacements 5
— 515 ———— 14
5, 3,
5, 543
Ag Ay»
Structure displacements
E = 30 X 10°ksi, J = 150 in.*, A = 8in.? (J and A are constant for all members. )
APPENDIX:
Matrix Algebra
A.1 DEFINITIONS
A matrix is a rectangular array of numbers, with m rows and n columns, for which
certain arithmetic operations have been defined. It is customary to enclose the
array within a set of brackets:
44, 42 443
[A] = | 421 422 43
43, 432 433
The numbers aj, @,2,... that make up the array are referred to as the elements
of the matrix. The first subscript of any element denotes the row and the second
subscript the column in which the element is located. Thus the element in the ith
row and jth column is represented by a, Either the capital letter used to name
the matrix or the array itself can be used to represent the matrix.
The order or size of a matrix refers to the number of rows and columns that
the matrix possesses. A matrix with m rows and n columns is said to be of order
m by n.
A.2 TYPES OF MATRICES
[B)=[1 2.7 4]
Column matrix. A column matrix consists of a single column.
5
Cd
—4
2 ONO
[E|= 1071 0
OF Ome,
3 2 =p
[F] = TNS 4
= a 1@
00 0
CSO oe
000 r
446
Sec. A.4 Multiplication 447
Li0en0
[Y=]; 0 1 0
LU) ol nag
[A]
=[B]
Addition and subtraction. Addition or subtraction of matrices is pos-
sible only if the matrices are of the same order. The addition or subtraction is
carried out by adding or subtracting corresponding elements term by term.
i oe 4pated pe mel
6 -7 2 Goel ge ele 2m 5d
It should be noted that the matrix obtained by adding or subtracting two matrices
is of the same order as the original matrices.
A.4 MULTIPLICATION
Ay4X1 + Ay2X2 = Cy
(A.1)
AX, + Ap2X2 = C2
448 Matrix Algebra Appendix
Here x, and x, are the unknowns and the a’s and c’s are constants. Next we
define the three matrices
[A][B] = [C]
then any element c,; in [C] is obtained by multiplying, term by term, the ith row
in [A] by the jth column in [B] and adding the resulting products. This rule can be
expressed analytically as
of = » Gin dy;
k=1 (A.3)
where n is the number of columns in [A] and the number of rows in [B].
For example,
1 3 2 > 1 aly
We 0 4;=|]-5 2
4 i SS) — eee 1-2
The rule for multiplication given by Eq. (A.3) immediately
establishes the
fact that two matrices can be multiplied only if the number
of columns in the first
is equal to the number of rows in the second. Two
matrices that satisfy this
criterion are said to be conformable for multiplication
. There-are no restrictions
regarding the number of rows in the first matrix
and the number of columns in the
second. The product of two matrices will have
the same number of rows as the
first matrix and the same number of column
s as the second.
A simple way of checking the conformabi
lity of two matrices and to
determine the size of their product is to
write the order of the matrices below
them, as follows:
[A][BC] = [AB][C]
and
[A][B + C] = [AB] + [AC]
However, matrix multiplication is not generally commutative. Thus
[A][B] + [B][A]
An exception to the last rule occurs when one of the two matrices to be
multiplied is the identity matrix. For example,
ek | Les O ahi 3
5 slo i r 5 |
Te || eed ee eS
lo lL 4 e B A
This example also illustrates the fact that multiplying by the identity matrix in
matrix algebra is analogous to multiplying by one in ordinary algebra.
fr d,
[Fl=|f |, [D] =| 4,
f d,
450 Matrix Algebra Appendix
Although it seems natural that the work should be the product of [F] and [D],
these matrices as defined above are not conformable and thus cannot be
multiplied. However, if we form the transpose of [F], it can be multiplied by [D]
and the product is indeed equal to the work. Thus
d,
[WI =[F)IDI=[f f fl) % | = (hd + hd + fd]
d,
A.6 DETERMINANTS
€44X1 + Ay2X2 = Cy
AX + An2X2 = C2
To solve for x;, we multiply the first equation by a2, and the secorid equation by
—a,2 and add the resulting equations. Thus
Q11A72X1 + A12Ax2X2 = C1 Ar
C142 — C2442
Ay SS Gee oe (A.4)
411422 — A242
The result given by Eq. (A.4) can also be expressed as the quotient of two
determinants. That is, :
Cy ay
Cz Az
Lite 44, Ay2
| (A.5)
43, An
The first fact to note about determinants is that they consist
of square arrays of
numbers bounded by two vertical lines. Next we define
the value of a determinant
in such a way that the solution given by Eq. (A.5)
is identical to that given by Eq.
(A.4). This can be accomplished if the value of a
2 X 2 determinant is equal to
the product of the elements along its major diagona
l minus the product of the
elements along the minor diagonal. Finally, we
note that the determinant in the
denominator of Eq. (A.5) consists of the array
of coefficients a, of the original
Sec. A.6 Determinants 451
simultaneous equations and that the determinant in the numerator contains the
same array except that the coefficients of the unknown x, have been replaced by
the constants c;.
In working with determinants and matrices, it is essential to keep in mind
the basic differences between the two. A determinant is always square and can be
reduced to a single value, whereas a matrix is simply an array of numbers, not
necessarily square, which cannot be reduced to a single value.
The form of solution given by Eq. (A.5) is known as Cramer’s rule and is
applicable to any number of simultaneous equations. However, the method of
evaluating a determinant described above can be applied only to a 2 x 2
determinant. Therefore, let us now consider a more general procedure for
evaluating determinants, which is applicable to determinants of any size. The
method is called the Laplace expansion of determinants, and it makes use of the
concepts of minors and cofactors.
Qi, Qy2_ 43
|A| = |@ 1 a2 ao3 (A.6)
43; 432 433
The minor of element aj, is
Ar2 23
M,, =
a32 33
Az, a3
M, =
a3; 433
Az, 43
Cy = (—1)°'M,2
a3, 433
452 Matrix Algebra Appendix
As an illustration, let us expand the following determinant along its first row:
i ea
Al =H 36 bel—2
Ar 3
We obtain
Matrix algebra does not include the operation of division. Instead of dividing
a
matrix [A] by a matrix [B], we multiply [A] by the inverse of [Bj. The
inverse of a
matrix [B] is defined as the matrix which when multiplied by [B] produces
the
identity matrix. That is,
[B[B)* = [B)[B] = 7]
where [B]~' denotes the inverse of [B]. The inverse of a matrix
can be likened to
the reciprocal of a number in ordinary algebra. Only
a nonsingular matrix has an
inverse.
Since the operation of inverting a matrix is closel
y related to the solution of
simultaneous equations, let us consider the follow
ing set of equations:
@11X1 + Ay2Xy + Ay3X3 = Cy
Ay1X1 + AxX7 + Ay3X3 = Cp
(A.7)
431X1 + Azx9Xo + 433X3 = C3
The equations can be written in matri
x form as
411 Gyn 43 xy Cy
421 Az2, An; X22} =] co (A.8)
G31 @32_ A33 || X, cz |%
or simply as
[A][X] = [C] (A.9)
Sec. A.7 Inverse of a Matrix 453
where [A] is the matrix of coefficients, [X] the matrix containing the unknowns,
and [C] the matrix made up of the constants on the right-hand sides of the
equations.
To solve for [X], we premultiply both sides of Eq. (A.9) by [A]~!. Thus
1
13 = A 18 + CC x3 + €3C33)
where C;; is the cofactor of the element in the ith row and jth column. In the
problem we are considering, C;, happens to be not only the cofactor of c, in the
numerator of x, but also the cofactor of a,, in |A|. Thus C;; is in general equal to
the cofactor of a, in |A|.
Rewriting (A.11) in matrix form gives
xy Ca Coy (Ch Cy
i
X2 = |A| Cy Cry Cr C2 (A.12)
X3 Ciz3 Cx Cs; C3
454 Matrix Algebra Appendix
The square matrix on the right-hand side of Eq. (A.12) is called the adjoint
matrix. It can be obtained in the following way. First we form the matrix [A]
containing the coefficients of the simultaneous equations.
44, 42 443
[A] = | aa. 422 a3
43; 432 433
Next we replace every element in [A] by its cofactor. This new matrix is referred
to as the cofactor matrix.
Cy Cr C3
[Cof. A] =|] Ca Cx Cr
C31 Cxr Cr3
Finally, we form the adjoint matrix by taking the transpose of the cofactor
matrix.
Cy Cy Cri
[Adj. A] =1Cyp Cy Cr
C3 Cy C33
Comparison of Eqs. (A.12) and (A. 10) indicates that the inverse of a matrix
is equal to its adjoint matrix divided by its determinant.
To illustrate the foregoing ideas, let us solve the following set of
simultaneous equations, using the inverse of the coefficient matrix:
tpt Xo Xs =
244 =) oe ar x3, = 5
X41 ae 2X AF 2X3 = 3
fom] Paid
(Al) 1
1 ota
4. 3
[Cof. AJ=| 0 1 -1
2 ed
Aire *Oring
[ed }a4 fe meen tn
press
|A] = 1(—4) + 1(—3) + 1(5) = -2
ag) py ee OT Pee | 2s a es
X3 5 —-1] -—3 3 3)
Sec. A.8 Partitioning of Matrices 455
ae eer Aa]
714 1 An, Ax
2
ya ad te
where A,;, Aj, Az, and A, are submatrices of [A].
To illustrate how we carry out the basic operations of matrix algebra on
partitioned matrices, let us partition two matrices, [A] and [B], and obtain their
product [C].
cep il at ay
[AJ=|]1 4:2]= be aa
4 2 1 21 ZZ
4 2 &
ial-|3.6]=[5 1 2) 21
Ax Ax By, Cr
Ax»Br, = [1)[1 2] = [1 2]
Finally, we combine C,, and C,, to obtain [C], the product of [A] and [B].
19 20
[ele ae
23° 22
To carry out an algebraic operation involving partitioned matrices requires that
two conditions be satisfied. First, the matrices to be partitioned must be
conformable, and, second, the partitioning must be done in such a way that the
operations involving the submatrices can be carried out. For example, if two
matrices are to be added or subtracted, the vertical and horizontal partitioning of
both must be identical. If two matrices are to be multiplied, the vertical
partitioning of the first must be similar to the horizontal partitioning of the
second. Thus if the vertical partitioning of the first matrix divides it into two parts
having three and two columns, respectively, then the horizontal partitioning of
the second matrix must divide it into two parts having three and two rows,
respectively. For multiplication, the horizontal partitioning of the first matrix and
the vertical partitioning of the second matrix are independent of one another.
Answers to Even-Numbered
Problems
Chapter2
2.2. Ray = 65 KN; Rax = 30KN; Ray = 95 KN
2.4. Ray = 11.43k; Rey = 38.75k
2.6. Ray = 6.96k; Ray = 2.28k; Rp = 3.8k
2.8. Ray = 52.2k; Rey = 16.8k; Rex = 18k
2.10. Ray = 78.8kN; Ray = 78.0KN; Rey = 71.2kN
2.12. Ruy = 6k; Rax = 40k; Rey = 44k
2.14. Ray = 75 KN; Rey = 85 KN; Rey = 100kN
2.16. Te = 32k; Ray = 216k; Rax = 10.4k;. Rp = 27k
2.18. Ray = 27.4k; Rax = 9.5k; Ray = 20.6k; Rex = 22.5k
2.20. Ray = 9.59k; Rax = 4.8k; Ray = 22.7k; Rey = 12.7k; Rex = 7.2k
2.22. Ray = 30KN; Ruy = 30 KN; Ray = 110kKN; Rey = 20 KN; Rex = 30KN
2.24. Ray = 29.2k; Ray = 86k; Rey = 54.8k; Roy = 15k
Chapter3
3.2. AB =12.5k(C); AE =20k(1); BE=0; BC =12.5k(T); CD= Pik (CG):
BD = 25k(C); DE = 20k(T)
3.4. AB = 8k(C); AE = 16k(T); BE = 0; BC = 17.9k(C); CE = 16k(C); CD =
16k (C); DE = 22.6k(T)
3.6. AB = 43.3kN(T); AF = 10kKN(T); BF = 16.7kN(C); EF = 60kN(C); BE
45kKN(C); CE =60kN(C); BC =54.1kN(T); DE =45kN(C); CD
54.1 kN (T)
3.8. AB =40k(C); AE =3k(C); AF =1.67k(T); CB =13.3k(C); BE =
13.3k (C); DB = 40k (C); CD = 24k (T); DE = 24k(T); EF = 15k (T)
3.10. ED =0; EF =5k(C); DC =3.61k(T); DF =8.77k(C); CF =2.13k(T);
FG = 13.5k(C); CB = 14.8k(T); CG = 12.1k(C); GB = 5.42k(T); GH =
24.4k(C); BH = 14.3k(C); BA = 26.6k(T); HA = 8.57k(T)
3.12. BC = 178.9kN (C); BG = 89.4kN (C); HG = 240 kN (T)
3.14. BC = 183.3kN(T); BD = 83.3kN(C); ED = 80kN(C)
3.16. BD = 20k(C); FD = 33.6k(C); GD = 14.1k(T)
3.18. AH = 33.5k(T); BC = 30.9k (C); CH = 5.34k(T); DE = 22.4k(C)
3.20. GH = 3.33k(C); DC = 3.33k(T);/ MH = 18k(C); MC = 18k(T); CN =
35 k (C); NH = 15k(T)
3.22. AB = 81.6KN(T); BC = 15.5kN(C); CD = 108.8kN(C); DE = 81.6kN(C);
EF =70kN(T); FG =73.1KN(T); GH = 135.7kN(T);_ DF = 21kN(C):
DG = 24.3KN(T); CG = 64kN(T); BG = 38.3kN(C); BH =39kN(C):
AH = 130kN (T)
3.24. AB = 44.7k(C); AG =22.4k(T); BC=33k(C); BG =17k(T): GC =
2.4k (C); GF = 33.3k (T)
3.26. AB = 3kN(T); AF = 8kN(T); BF = 2.57kN(C); BC = 1.67kN(T); CF =
14.3 kN (C); CD = 2.02 kN(T); DF = 17.7kKN(T); DE = 15kN(C); EF = 0
3.28. AB = 133.3 KN (C); AF = 67.1kN(T); BF = 28.6kN(C); BC = 101.0kN(Q);
CF = 72.8kN(T); CD = 86.9kKN(C); DF = 28.6kN(C); DE = 90.9kN(C):
EF = 67.1kN(T)
Chapter 4
4.2. (Bar forces only) AB = 83.4kN (T); AC = 228kN (C);
151 kN (T); BD = 387kN(C); CD = 89.0kN (T)
AD 172 kN (T); BC =
4.4, (Bar forces only) AB = 27.3kN (C); AD = 180 kN
(C); AE = 419kN (C); BG
43.6kN(T); BE = 101kN(T); CD = 27.1kN (C);
CE =141kN(T); DE
141 kN (T) :
Chapter 5
peas Vinx = 12k; Max = 120 k-ft
5.4. Vinax = 40 KN; Mua, = —160 kN-m
5.6. = 32k; Mnax = 171 k-ft
2*
SwL* wL?
6.36. 5
® = 3e4n7° 94 = D4R]
,
460 Answers to Even-Numbered Problems
23PL° ee
6.38. On 648EI’ A 9EI
Wake syab:
Op et ee
6.42. 5, = 1.89in.; 6, = 0.0205 rad
6.44. dc = 0.93 in.
Chapter7
=
4PL? 1:4... 07 =
TPL?
BLES * 486EI
7.6. dc = 1.62 in. 7.8. dg = 2.5in.
7.10. 5p = 67.5mm 7.12. d, = 47.3mm
7.14. 6, = 36mm TAKS Cys = S260.
Wks, One S 21k iin, 1.20500 c78—7o 2 ie
W494: Ore S Leer, Je24e On eee
7.26. dsy = 18.7cm; 0, = 0.0067 rad
7.28. dc = 9.07 cm 7.302051 — 03mm
1.32.0 pe 10.914 ins 1.34 On 104308
7.36. Od, = 1.58in.
Chapter9
92 Rag = 90K hay — 32.5 ke Rey = 21.0k 1; Rey = 32.5k<—
Chapter 11
11.2. R, = 6.2kN; M, = 5kN-m; Rz = 23.8kN
11.4. R, = 10.47k; M, = 74.1 k-ft; Rp = 4.53k
3wL SwL 3wL
11.6. ee es PS year SF a
13.16. M4, = 70.7 kKN-m; Mg, = 141.4kN-m; Mgc = —141.4kN-m; Mog = 295.6 kKN-m;
Mep = 12.8 kN-m; Mpc = 6.4 KN-m; Mere = —308.5 KN-m; Mec = 90.2 KN-m
13.18. Map = —12.9 kN-m; Maa a —7.06 kN-m; Msc = 7.06 kN-m; Mer = 7.06 kN-m;
Chapter 14
14.2. Map = —6.9 kN-m; Mpa = 92.9 kN-m; Mac te —92.9 kN-m; Mer — 45.4 kN-m;
14.8. M,, = 138 k-ft; Mac = —138k-ft; Mer = 77.5 k-ft; Mep = —77.5 k-ft;
Moc = 18.8 k-ft; Mor = —18.8 k-ft
14.10. M,, = —92.7kN-m; Mg, = 114.6kN-m; Mgc = —139kKN-m; Mgp = 24.4 kN-m;
Mop ns VD kN-m
Moa = 9.2 k-ft; Mac — 18.5 k-ft; Msp <= IPA k-ft; Mops = 6.2 k-ft
14.12. Map = 4.6 k-ft;
14.14. M4, = 70.8kN-m; Mz, = 141.5kN-m; Mgc = —141.5kN-m; Mcg = 295.7 KN-m;
Mcp — 12.8 kN-m; Mer = —308.6 kN-m; Moc = 6.4 kN-m; Mec — 90 kN-m
14.16. M,, = 12.9k-ft; Mineo alekatt: Mac = —32.1k-ft; Men. = 27.9 k-ft:
Mcp = —27.9 k-ft; Moc = —17.1 k-ft
Chapter 15
L? L
S20Pa 47) a
El 8
Ae 16
Vie
15.4. [F] = ik, WAL 18
a 1 > Say r= 15h 24Ee
—15L -9 -12L 18
1 30L?
15.6. [F] =——| -7L’ 4L?
as yh Sh &
: |73.4
15.8. [F] = FE ili AI
[se 24.0 24.0
i, 5W, L* WL WL
tO Atal SIAR AR ET actt ae Ames 5
WL
q3 = Scot i! a 0
15.12. R; =5.5k; A, = 0.024rad; A, = 0.384 ft; q:= 90; q2 = 30k-ft; qg3; = 78k-ft;
qa = —66k-ft; gq; = —66k-ft; gq, = 0
15.14. R, = 4.3kN; R; = 4.94kN; A, = 7.83 mm; q: = —40.5kKN-m; gq, = 22.3
kN-m;
93 = 22.3kN-m; q, = —17.2kN-m; q; = —17.2 kN-m; gq, = 0
15.16. R, = 6.16k; A, = 1.23in.; q, = —76.8 k-ft; q. = 61.6k-ft; 43 =
61.6k-ft; gq, = 0;
qs = 6.16k
Chapter 16
0.0256> 0.0192" 0.0256 0.0192 0 0
=U.0192) 7 0.0817 0.0192 —0.0144 0 —0.0667
—0.0256 0.0192 0.0756 —0.0192 —0.05 0
16.2. K = AE
0.0192 -—0.0144 —0.0192 0.0144 0 0
0 0 —0.05 0 0.05 0
0 —0.0667 0 0 0 0.0667
16.4. q,; = —7.14k; qo = —22.2k
16.6. q, = 22.2k; q. = —27.8k; q3 = 32k
16.8. gq, = 26.5 k; q2. = 42.9k; q3 = 22.5k
16.10. g, = 20kN-m; 42 = —0.97 kN-m; q3 = —19.5kN-m; qa = —9.75kN-m
qs = —19.5 kN-m; gq, = —9.75 kN-m
16.12. gq, = 9; q2 = 42.9k-ft; qg3 = —42.9k-ft; g, = —21.4k-ft
16.14. q,; = —85.7kN-m; q2 = 68.6 kN-m; q; = —68.6kN-m; q,4 = 0
16.16. gq, = 81.4KN; q2. = 96.7kKN; q3 = —83.3KN; qa = —91.5kN; qs; = —44.1kN;
do = —28.2 KN; q7 = 19.1 kN
16.18. gq, = —12.1KN; q2. = 5.36KN; q3 = —36.7KN; q4 = —38.9KN; qs = —36.7kN;
4o = 36.7KN; gz = —21.1 KN; gg = —52.3 KN; go = —20.0 KN
16.20. g, = -1.56k; q>=6.02k; q3 = —709.12in.-k; q,=1.56k; qs = —6.02k;
de = —374.55in.-k; q7 = 5.98k; gg = —1.56k; qo = 374.55 in.-k; qi = —5.98k;
Qu = 1.56k; qx.= 372.98in.-k; q13 = 1.56k; dis = 5.98k; Gis = —703.35 in.-k;
dis = —1.56k; gry = —5.98k; Gig = —372.98 in.-k
adie Psdthe iy 08 New os
Haval- «ig fate - oe MAP ol ee HIT A
> . Chk op !
1 ° bd 6
ij ah ag OS Rt
wwihe willy wale
‘ | i Hite > Lye; eH ob
’ rn hie vii
a ; ij gan a) t] ug
foar j fi a
ao?
| gi¢u But = ig | _
Mi eT¥-~ = vw jee AUD? > Me LOL = oy af
aia h- «© —« wees & ao” oe
sA.j0-—~— is £4 5 & © Beh OS<Q
S sD @ 2h
he Sy imi Pehe « ib genta ee = » .aeigt m= » ay ’
iwi kB i Le e eee ee) ee ee
el LY Sep wo
Ape ai ou 7 ee w epi 2 nee
fo whe -
ig
ip
¢ 222 Ge = 7) Wed tae * off
» . rs
j j . : . J a, 7 a a oes Ace
Pe
o8 ? . hy j P are : dma y _ * | a 7 «* 4 ' = > EAE ’
7 7 : ' A ame ray f 62 Mh ~ yy ~ 7
. tsp ee mm ry P| ddl ¢ gi ine “ if, oa! % 4 7 - 7
oe I Vast 1S is - We aX i= Sarbe * ;
- = a
: >
—s
‘+
a eas
™~,
oe : 4% .
an bi6 ‘
.
Instructions for Use
of Enclosed Diskette
The diskette enclosed with the book contains programs for analyzing plane
trusses and frames using IBM or IBM-compatible personal computers. The
diskette contains three programs, FLEX, STIFFT, and STIFFF, written using
GW/BASIC.
The first program, FLEX, uses the flexibility method to analyze trusses and
frames. A listing of the program together with an illustrative problem, which
explains what input data is required and how this data is to be entered into the
program is given on pages 355 to 362. The second and third programs, STIFFT
and STIFFF, use the stiffness method. STIFFT is intended for analyzing trusses
and STIFFF for analyzing frames. The listings of these programs, as well as the
illustrative problems with instructions regarding the entering of data can be found
on pages 414 to 420 and pages 425 to 433 respectively.
The following procedure should be employed when using the diskette:
1. Load BASIC
2. Load the desired program FLEX, STIFFT or STIFFF.
3. In accordance with the instructions given in the illustrative problem,
enter the correct data for the problem being solved into the DATA
statements at the end of the program. Initially these DATA statements
contain the data corresponding to the illustrative problem.
465
466 Instructions for Use of Enclosed Diskette
Arches:
Element flexibility matrix:
analysis, 210-213
axially loaded element, 339-40
definition, 4, 210 flexural element, 333-34
Axial deformation, 123 Element stiffness matrix:
axially loaded element, 375-76
flexural element, 377-80, 404-07
Beams: Energy methods:
definition, 6, 90 analysis, 254—67
deflections, 125-129, 131-140, 144-148, 164-167 deflection calculation, 158-75
differential equation, 123 Equilibrium equations, 22-23
influence lines, 189-96 External work, 158, 160
internal forces, 91
moment diagrams, 93-105
shear diagrams, 93-105 Finite-element method:
types, 90 flexibility method, 328-62
Bending moment diagrams: stiffness method, 374-433
beams, 93-105 Fixed-end moment, 279, 307
frames, 105-111 Flexibility matrix:
sign convention, 92 definition, 328-29
Bridge live loads, 14 element, 333-34
Building live loads, 14 structure, 336-37
Flexibility method, 328-62
Force method (See Flexibility method)
Cables, 3, 213-15 Force transformation matrix, 334-36
Cantilever structures, 32-33 Form, structural, 2-10
Carry-over factor, 309 Frames:
Castigliano’s theorems, 254—57 moment diagram, 105-111
Computer programs: shear diagram, 105-111
flexibility method, 348-62 sidesway, 290-297, 318-321
stiffness method, 410-33
Conjugate-beam method, 140-148 Hinges, internal, 30-33
Conservation of energy principle, 158 History of structural engineering, 16-20
467
Index
468
Sign convention:
Least-work method, 254-67
bending moment, 92
Live Icads, 13-16 moment-distribution method, 306
Loads:
shear, 92
bridge, 14
slope-deflection method, 277
building, 14
trusses, 45
dead, 13
Singularity functions, 128-29
live, 13-16
snow, 15
Slope-deflection equation, 276-81
Slope-deflection method, 276-97
wind, 16
Snow loads, 15
Space trusses, 78-86
Material properties, 10-13 Stability and determinacy, 24-27
Matrix algebra: Stiffness matrix:
addition, 447 definition, 374-75
inversion, 452-54 element, 375—80
multiplication, 447—49 structure, 380-87
partitioning, 455-56 Stiffness method, 374-433
subtraction, 447 Strain energy, 158
transposing, 449 Superposition, principle of, 137-39
Matrix method: Support settlement, 235-37
flexibility method, 328-62 Support types, 23-24
stiffness method, 374-433 Symmetry, 265
Maxwell’s law of reciprocal deflections, 256
Moment-area method, 129-40
Moment-area theorems, 129-31
Moment diagrams (See Bending-moment diagrams) Torsional deformation, 124
Moment-distribution method, 304-21 Trusses:
analysis, 42-68, 78-86
assumptions, simplifying, 42—44
Nodes, 333 definition, 5, 42
deflections, 172-74
determinate, 42-68
Reactions, 22-33
Reciprocal deformations, law of, 256 indeterminate, 243-45, 264-65
Redundants, 224
influence lines, 196-201
Rigid frames (See Frames) matrix analysis, 387—93
Rings, analysis of, 265—67 method of joints, 45-51
method of sections, 51—54
Sections, method of, 51—54 sign convention, 45
Settlement of supports, 235-37 space trusses, 78-86
Shear deformation, 124 stability and determinacy, 58—60
Shear diagrams:
beams, 93-105
frames, 105-111 Virtual work, 162-64
sign convention, 92
Sidesway:
moment-distribution method, 318-21 Wind loads, 16
slope-deflection method, 290-97 Work, 158-60
=
‘3 ‘Aus
t
ot Pa vs
i
=, tte! AT
go
DATE DUE
HIGHSMITH ~— # 45220
WO
IRIE VI
and Compatibles
for the IBMt™ PC
Student Diskette
S2EOIEMAN LIBRARY
rare
nyBlea fens: Acid AASUNIVERT7SI44TY6
.,
ic) ga BRAMRIR & AMSchIEustWe
TE
son er, Inc.
a Division of Sig
3- 5
ISBN Q-13-89507
/ NOTICE TO CONSUMERS
) THIS BOOK CANNOT BE RETURNED FOR CREDIT OR REFUND IF THE
| PERFORATION ON THE VINYL DISK HOLDER IS BROKEN OR TAMPERED
| WITH
RAPSERMIGES
Structural
ALEXANDER CHAJES
The new edition retains all the popular coverage of the first edition and
has been updated to include easy-to-follow computer programs in Fortran
and Basic to supplement written presentations of analyses.
e =
ISBN 0-13-855073 Ss
~
9°780138"550738 /