( Reaffirmed 1995 )

Please purchase PDF Split-Merge on [Link] to remove this watermark.

HANDBOOK
FOR
STRUCTURAL
ENGINEERS
No. 6

Please purchase PDF Split-Merge on [Link] to remove this watermark.

As in the Original Standard, this Page is Intentionally Left Blank

Please purchase PDF Split-Merge on [Link] to remove this watermark.

SP : 6(6) - 1972

HANDBOOK
FOR
STRUCTURAL ENGINEERS
6. APPLICATION OF PLASTIC THEORY IN
DESIGN OF STEEL STRUCTURES

BUREAU
OF
INDIAN
STANDARDS
MANAK BHAVAN, 9 BAHADUR SHAH ZAFAR MARG
NEW DELHI 110002
Price Rs. 275.00

October 1973

Please purchase PDF Split-Merge on [Link] to remove this watermark.

BUREAU

OF INDIAN

STANDARDS

Edition 1 - 1972

Sirth Reprint DECEMBER 1998

UDC 624.014.2

: 624.04

SP : 6(6)-1972

Copyrtght 1973

BUREAU OF INDIAN STANDARDS

This publication is protected under the Indian Copyright Act (XIV
of 1957) and reproduction in whole or in part by auy means except
with written permission of the publisher shall be deemed to be an
infringement of copyright under the said Act.

Printed in India by Printsgraph, New Delhi and

Published by Bureau of Indian Standards, New Delhi 110002

Please purchase PDF Split-Merge on [Link] to remove this watermark.

CONTENTS
PACE

FOREWORD

. ..

. ..

..

. ..

...

SYMBOLS

...

. ..

. ..

...

s..

II

...

...

15

SECTION
1. SCOPE

...

2. GENERAL

...

INTRODUCTION

...

...

...

. ..

. ..

. ..

15

...

...

...

...

16

4. MECHANICALPROPERTIESOF STEEL . . .

. ..

...

19

5. MAXIMUM STRENGTHOF SOME ELEMENTS

. ..

...

21

6. HISTORICALDEVELOPMENT

. ..

. ..

23

3. STRUCTURALSTRENGTH

SECTION

...

JUSTIFICATION

7. WHY PLASTIC DESIGN

FOR PLASTIC DESIGN

...

...

. ..

8. INADEQUACYOF STRESSAS THE DESIGN CRITERION

...

25

. ..

26

9. EXPERIMENTALVERIFICATION

...

...

...

32

10. THE CASE FOR PLASTICDESIGN

...

...

...

33

SECTION

FLEXURE

OF BEAMS

11. ASSUMPTIONSAND CONDITIONS

...

...

...

35

12. BENDING OF RECTANGULARBEAM

...

. ..

..

35

13. BENDING OF WIDE FLANGE BEAM

...

. ..

. ..

40

. ..

...

. ..

43

...

...

. ..

45

14. PLASTIC HINGE

...

15. REDISTRIBUTIONOF MOMENT

SECTION

PLASTIC ANALYSIS

16. FUNDAMENTALPRINCIPLES

...

...

...

49

17. STATICALMETHOD OF ANALYSIS

...

...

. ..

53

18. MECHANISMMETHOD OF ANALYSIS

. ..

...

...

56

...

...

. ..

66

19. FURTHERCONSIDERATIONS
5

Please purchase PDF Split-Merge on [Link] to remove this watermark.

8F: 6(6) - 1972

SECTION
20. GENERAL
n.

...

APPLICATION
...

PRFUYINARYDESIGN

.. .

22. [Link]~AL DESIGNPROCEDURE

P.

TO DESIGN

...

...

. ..

77

. ..

. ..

. ..

77

...

. ..

. ..

78

. ..

. ..

81

SECORDARY
DESIGNCONSIDERATIONS
SECTlON

24. INTRODUCTION
.. .

DESIGN

.. .

EXAMPLES
.. .

. ..

25. DESIGNEXAMPLESON CONTINUOUS

BEAMS

125
.. .

125

26. DESIGNEXAMPLESON IKDUSTRIAL

BUILDINGFRAMES

..

136

STRUCTURES
27. DESIGNEXAMPLEON MULTI-STOREY

. ..

178

SECTION

. ..

SIMPLIFIED

PROCEDURES

. ..
2%. INTRODUCTION

. ..

.. .

. ..

. ..

187

BEAMS
29. CONTINUOUS

.. .

. ..

. ..

...

187

30. SINGLE-SPAN
FRAMES(SINGLESTOREY)

. ..

...

188

31. MULTI-SPANFRAMES

. ..

...

. ..

191

APPENDIXA

SELIXTEDREFERENCES . . .

. ..

...

195

APPENDIXB

SPACINGOF LATERALBRACING

. ..

. ..

198

APPENDIXC

CHARTSAND FORMULAS
FOR SIMPLEBEAMS

. ..

203

APPENDIXD COMPOSITION
OF STRUCTURAL
ENGINEERINGSEG
...
. ..
TIONALCOMMITTEE,
SMBDC 7
. ..

216

...

APPENDIXE INDIAN STANDARDSRELATING TO STRUCTURAL

. ..
...
. ..
.. .
ENGINEERING
218

Please purchase PDF Split-Merge on [Link] to remove this watermark.

SP: 6(6) - 1972

FOREWORD
This Handbook,
which has been processed
by the Structural
Eng!neering
Sectional Committee,
SMBDC 7, the composition
of which
is g;ven in Appendix
D, had been approved
for publication
by the
Struttural
and Metals Division
Council
and the Civil Engin&ring
Division Council of ISI.
Steel, which is a very important
basic raw materials
for industrialization,
had been r&iving
attention
from the Planning
Commission even from the very early stages of the countrys First Five Year
Plan period.
The Planning
Commission
not only envisaged
an increase
in production
capacity in the country,
but also considered the question
of even greater
importance,
namely,
taking
of urgent measures
for
the conservation
of available
resources.
Its expert committees
came
to the conclusion
that a good proportion
of the steel consumed by the
structural
steel industry
in India could be saved if higher efficiency
procedures
were adopted
in the production
and use of steel.
The
Planning
Commission,
therefore,
recommended
to the Government
of
India that the Indian
Standards
Institution
should take up a Steel
Economy Project and prepare a series of Indian Standard
specifications
and codes of practice in the field of steel production
and utilization.
Over fifteen. years of continuous
study in Tndia and abroad, and
the deliberations
at numerous
sittings of committees.
panels and study
groups resulted
in the formulation
of a number
of Indian
Standards
in the field of steel production,
design and use, a list of which is given
in Appendix E.
This Handbook
which relates to the application
of plastic theory
in design of steel structures
is intended
to present the important
principles and assumptions
involved
in the plastic method of structural
analysis, and to provide illustrative
examples for the guidance of the
designer in the analysis
of practical
design problems.
The subject
is introduced
by considering
the various
limits of
usefulness
of a steel structure,
the limits that are function
(in part)
of the mechanical
properties
of steel.
Knowledge
of these properties
is used in Section A to show how the maximum strength of some simple
The historical
development
of the
structures
may
be computed.
plastic theory of structures
is also dealt with in brief.
that

Section
stress

B answers the question Why plastic design . It is shown

is an inadequate
design criterion
for a large number
of
7

Please purchase PDF Split-Merge on [Link] to remove this watermark.

SP: 6(6) - 1972

The experimental
verificatiorr of tha
practical
engineering
structures.
plastic theory
(which bases the design of structures
on the maximum
strength)
has also been indicated.
The basic theoretical
work is deait
with in Sections C and D. The concepts of plastic bending and redlstribution
of moments
are described
and the methods
of analysis
has
been
indicated.
Section
E contains
general
comments
on design
procedures.
Although
this section covers a few examples
relating to
multistorey
frames, it is proposed
to deal with the subject in detail in
a supplement
in due course.
The limitations,
modifications
and design
details
have been described
under
the heading
Secondary
Design
Consideration
. Proper attention
should be given to the effect of shear
force, axial force, local and lateral buckling,
etc.
Further,
the beams,
columns
and connections
should be designed to meet the requirements
of plastic hinge formation.
The section on design examples treats a number of building
frames
of different
profiles.
The secondary
design considerations
are checked
throughout.
Section 7 describes simplified procedures
of solving design
problems
with the use of formulas,
charts and graphs.
In
detailed

Appendix
A is given a list of selected references
information
on plastic theory of structures.

for further

What will plastic design mean ? To the sidewalk superintendent

,
it will mean nothing.
The structure
will look just the same as a conventionally
designed structure.
To the engineer,
it will mean a more
rapid method of analysis.
To the owner, it will mean economy, because
plastic
design requires
less steel than
conventional
design.
For the
building
authority,
it would mean more efficient
operations
because
designs may be checked faster.
To steel industry,
it would mean more
efficient use of its products.
Finally,
to a nation,
it will mean better
use of her natural
resources.
This Handbook
is based on and requires
publications
issued by ISI:
for structural
IS: 226-1969 Specification
(fourt/z revision)

reference
steel

IS:

800-1962 Code of practice for use of structural

building construction
(revised)

IS:

875-1964 Code of practice

Loading standards
(revised)

for

structural

to the

following

(standard

quality)

safety

steel

in general

of buildings:

2062-1969
Specification
for structural
steel
(fusion
welding
quality)
(@St revision)
for assembly
of structural
joints
IS: 4000-1967 Code of practice
using high tensile friction grip fasteners

IS:

Please purchase PDF Split-Merge on [Link] to remove this watermark.

SP: 6(6) - 1972

In the preparation of this handbook, the technical committee has
derived vahlable assistance from Dr Lynn S. Bee&e, Professor of
Structural Engineering, Lehigh University, Bethlehem, USA. Dr Bcedle
prepared the preliminary
draft of this handbook.
This assistance
was made available to IS1 through Messrs Ramseyer & Miller, Inc, Iron,
and Steel Industry Consultants, New York, by the Technical Co-operation
Mission to India of the Government of India under their Technical
Assistance Programmc.
No handbook of this kind may be made complete for all times to
come at the very first attempt.
As designers and cnginecrs begin to
USC it, they will be able to suggest modifications and additions for
improving its utility.
They are requested to send such valuable
suggestions to ISI which will be received with appreciation and
gratitude.

Please purchase PDF Split-Merge on [Link] to remove this watermark.

As in the Original Standard, this Page is Intentionally Left Blank

Please purchase PDF Split-Merge on [Link] to remove this watermark.

SP : 6(6) - 1972

SYMBOLS
Symbols used in this handbook shall have the meaning assigned
to them as indicated below:
A
E
Area of cross-section
AP
=
Arca of both flanges of WF shape
Area
of split-tee
A,; =
Aw
= Area of web between flanges
= Flange width
b
Distance from neutral axis to the extreme fibre
Depth of section
E
= Youngs modulus of elasticity
5

ii

Es, = Strain-hardening
E,

Eccentricity
Load factor of safety

Shape factor = M,
M = s

Gt

=
=
=

H
HB
I
I

I*
J

rt

= Tangent modulus
s

.;

modulus = 2

z
=
=

KL
L
L,
iu

=
=
=
=

Fixity factor for :se in evaluating and restraint coefficient

Modulus of elasticity in shear
Modulus of elasticity in shear at onset of strain-hardening
Hinge rotation required at a plastic hinge
Portion of hinge rotation that occurs in critical (buckling) seg.
ment of beam
Moment of inertia
Moment of inertia of elastic part of cross-section
Moment of inertia of plastic part of cross-section
Number of remaining redundancies iu a structure that is
redundant at ultimate load
Euler length factor
Distance from flange face to end of f2lct
Effective (pin end) length of column
Span length; actual column length
Critical length for lateral buckling
Moment
11

Please purchase PDF Split-Merge on [Link] to remove this watermark.

SP: 6(6) - 1972

=

M,
M,

Number
of plastic
hinges developed
in a structure
that is
redundant
at ultimate
load
= Moment at the haunch point
a useful maximum
moment;
hinge moment
= End moment;
Plastic
moment
=
= Plastic moment capacity of a beam section
= Plastic hinge moment modified to include the effect of axial
compression
Plastic
hinge moment modified to include effect of shear force
=
Maximum
moment of a simply-supported
beam
=
Moment
at
which
yield
point
is
reached
in flcxure
=

i&c

= Moment

M,
N

=
=
=
=

Mb
MO
Mp
Mp

Mpc
Mps

at which initial outer fibre yield

thrust is present
Moment at the working load
Number of possible plastic hinges
Number of possible independent
mechanisms
Concentrated
load

occurs

when

axial

load.
A load used as the maximum
column
load
= Euler buckling load
= Reduced modulus load
= Stabilizing load
= Tangent modulus load
= Theoretical ultimate load
= Working load
to yield stress level; P = Aa,
= Axial load corresponding
= Rotation capacity
= Radius of gyration
= Section modulus, I/C
= Section modulus of elastic part of cross-section
= Force
= Flange thickness
= Stiffner thickness

PC, = Useful column

P
P,
P,
Pt
P"
P,

P,

R
;

s,
T
tf

ts
t.,

= Web t hickncss

Shear force
Shear
carrying capacity of a section
vc
=
_.
2c, v, w = Displacements
in x, y, and z directions
= Total distributed
Joad
W
12

Please purchase PDF Split-Merge on [Link] to remove this watermark.

SF: 6(6) - 1972

W EXT=
W

Wd

WU
X
x
x
Y
Y

=
=
=
=
=
=

External
work due to virtual displacement
Internal
work due to virtual displacement
Distributed
load per unit of length
Thickness of the wet doublers
Total uniformity
distributed
load
Number of redundancies
Longitudinal
coordinate
Distance
to position
of plastic hinge undrr distributed
Transverse
coordinate
Distance
from neutral
axis to centroid
of half-area

Plastic

2,

WINT

modulus

load

= MtJ
CY

2,
2
AL
6
d
Gt

=
=
=
=
=
=

CY

=
=
=
=

p
d
QY

9
0,
a,rt
QWY

0,
fJY

=
=
=
=

=
I

=
=
Y

Plastic modulus of elastic portion

Plastic modulus of plastic portion
Lateral co-ordinate
Equivalent
length of connection
DCflection
Strain
Strain at strain-hardening
Strain
corresponding
to first attainment
of yield stress level
Measured angle change; rotation.
Rotation
Poissons ratio
Radius of curvature
Nomial stress
Lower yield poiat
Proportion limit
Residual stress
Ultimate tensile strength of material
Upper yield point
Working stress
Yield stress level
Shear stress
Rotation per unit length, or average unit rotation; curvature
Curvature
corresponding
to first yield in flexure

Please purchase PDF Split-Merge on [Link] to remove this watermark.

As in the Original Standard, this Page is Intentionally Left Blank

Please purchase PDF Split-Merge on [Link] to remove this watermark.

SECTION A
INTRODUCTION
1. SCOPE
1.1 It is the purpose of this handbook to present the fundamental
concepts involved in plastic design and to justify its application to
structural steel frames. The methods of plastic analysis will be described together with the design procedures that have so far been
developed. Secondary design considerations are also included.

1.2 Specific application may be made to statically loaded frames of

structural steel to continuous beams, to single-storeyed industrial frames
and to such other structures whose condition of loading and geometry
are consistent with the assumptions involved in the theory. Numerous
applications will undoubtedly be made to other types of structures
such as rings and arches, but for the time being the scope of application
is limited to the indicated structural types.
2. GENERAL
2.1 Steel possesses ductility, a unique property that no other structural

material exhibits iu quite the same way. Through ductility structural

steel is able to absorb large deformations beyond the elastic limit without the danger of fracture.
2.2 Although there are a few instances where conscious use has been
made of this property, by and Large the engineer has not been able to
fully exploit this feature of ductility in structural steeL As a result of
these limitations it turns out that considerable sacrifice of economy is
involved in the so-called conventional design procedures.
2.3 Engineers have known of this ductility for years, and since the
19206 have been attempting to see if some conscious use could be
made of this property in design. Plastic design is the realization of that
goal. This goal has been achieved because two important conditions
have been satisfied. First, the theory concerning the plastic behaviour
of continuous steel frames has been systematized and reduced to simple
design procedures. Secondly, every conceivable factor that might tend
to Limit the load-carrying capacity to something less than that predicted
15

Please purchase PDF Split-Merge on [Link] to remove this watermark.

SP: 6(6) - 1972

by the simple plastic theory has been investigated
formulated to safeguard against such factors.
3. STRUCTURAL

and rules have been

STRENGTH

3.1 The design of any engineering structure, be it a bridge or building,

is satisfactory if it is possible to built it with the needed economy and
if throughout its useful life it carries its intended loads and otherwise
performs its intended function. As already mentioned, in the process
of selecting suitable members for such a structure, it is necessary to
make a general analysis of structural strength and secondly to examine
certain details to assure that local failure does not occur.
3.2 The ability to carry the load may be termed structural
strength.
Broadly speaking, the structural strength or design load of a steel frame
may be determined or controlled by a number of factors, factors that
have been called limits
of structural usefulness . These are: first
attainment of yield point stress (conventional design), brittle fracture,
fatigue, instability, deflections, and finally the attainment of maximum
plastic strength.
3.3 Strictly speaking, a design ba~sed on any one of the above-mentioned
six factors could be roferred to as a limit design , although the term
usually has been applied to the determination of ultimate load as limited
by buckling or maximum strengthl*.
Plastic design as an aspect of
limit design and as applied to continuous beams and frames embraces,
then, the last of the limits _ the attainment of maximum plastic strength.
3.4 Thus, plastic design is first a design on the basis of the maximum
load the structure will carry as determined from an analysis of strength
in the plastic range (that isi a plastic analysis). Secondly it consists of
a consideration by rules or formulas of certain factors that might otherwise tend to prevent the structure from attaining the computed maximum load. Some of these factors may be present in conventional
(elastic) design. Others are associated only with the plastic behaviour
of the structure.
But the unique feature of plastic design is that the
ultimaL
load
rather than the yield stress is regarded as the de&u
Criterion.
3.5 It ha; long been known that whenever members are rigidly connected, the structure has a much greater load-carrying capacity than
indicated by the elastic stress concept. Continuous or rigid frames
are able to carry increased loads above first yield because structural
steol has the capacity to yield in a ductile manner with no loss in strength;
*This
Appendix

number
A.

refers

to the serial number

16

of the

selected

references

given in

Please purchase PDF Split-Merge on [Link] to remove this watermark.

SP: 6(6) - 1972

indeed, with frequent increase in resistance. Although the phenomenon
will be described in complete detail later, in general terms what happens
is this:
As load is applied to the structure, the cross-section with the
greatest bending moment will eventually reach the yield moment.
Elsewhere the structure is elastic and the peak moment values
are less than yield. As !oad is added, a zone of yielding develops
at the first critical section; but due to the ductility of steel, the
moment at that section remains about constant.
The structure,
therefore, calls upon its less-heavily stressed portions to carry the
increase in load. Eventually,
zmes
of yielding are formed
at other sections until the moment capacity has been exhausted
at all necessary critical sections. After reaching the maximum load
value, the structure would simply deform at constant load.
3.6 At the outset it is essential to make a clear distinction between
elastic design and plastic design. In cmventional
elastic design practice, a member is selected such that the maximum allowable bending
stress is equal to 1 650 kgf/cm2 at the working load. As shown in
Fig. 1 such a beam has a reserve of strength of 1.65 if the yield point
stress is 2 400 kgf/cm*. Due to the ductility of steel there is an

OESIBNBASIS

CONVENTIONAL
FIG. 1

DESIGN

PLASTIC

DESIGN

PLASTIC DESIGN COMPARED WITH ELASTIC DESIGN

I_
.:

17

Please purchase PDF Split-Merge on [Link] to remove this watermark.

,\y

.L

SP: 6(6)-

1072

additional reserve which amounts to 12 to 14 percent for a wide flange

shape. Thus the total inherent overload factor of safety is equal to
1~65x1~12=1~85 as an average value.
3.7 In plastic design, on the other hand, the design commences with
the lcltimate load. (As will be evident later, it is much easier to analyze
an indeterminate structure for its ultimate load than to compute the
yield load.) Thus the working load, P,, is multiplied by the same load
factor (1.85) and a member is selected that will reach this factored
load.
3.8 The load v deflection curve for the restrained beam is shown in
Fig. 1. It has the same ultimate load as the conventional design of the
simple beam and the member is elastic at working load. The important
thing to note is that the factor of safety is the same in the plastic design
of the indeterminate structure as it is in the conventional design of the
simple beam.
3.9 While there are other features here, the important point to get in
mind at this stage is that in conventional procedures one computes the
maximum moment under the working load and selects a member such
that the maximum stress is not greater than 1650 kgf/cm2 on the
other hand in plastic design one multiples the working load by
F = 1.85 and selects a member which will just support the ultimate
load.

3.10 Terminology
- Plastic design naturally involves the use of some
new terms. Actually these are few in number, but for convenience are
listed below :

Limit Design-

A design based on any chosen limit of structural

usefulness.

Plastic Design - A design based upon the ultimate load-carrying

capacity (maximum strength) of the structure. The term plastic

is derived from the fact that the ultimate load is computed from
a knowledge of the strength of steel in the plastic range.

Ultimate Load (P,,) or Maximum Strength - The highest load a

structure will carry. (It is not to be confused with the term as
applied to the ultimate load carried by an ordinary tensile test
specimens.) In the design P, is determined by multiplying the
expected working load (PW) by the load factor (see below).

Plastijicath
section.

- The development of full plastic yield of the cross-

Plastic Moment (M&- Maximum moment of resistance of a fullyyielded cross-section.

18

Please purchase PDF Split-Merge on [Link] to remove this watermark.

SP: 6(6) - 1972

Plastic l
odzclus (Z)- Combined static moments about the neutral
axis of the cross-sectional areas above and below the neutral axis.
Plastic Hiqe e A yielded section of a beam which acts as if it
were hinged, except with a constant restraining moment.
Shape Factor (fl- The ratio of the maximum
of a cross-section (M,) to the yield moment
Mechanism - A hinge system , a system
move without an increase in load.

resisting
(M,).

of members

moment
that

can

Redistributiorc of Moment 4 A process which results in the successive formation of plastic hinges until the ultimate load is
reached. By it, the less-highly stressed portions of a structure
also may reach the (M&-value.
Load Factor (F)- A safety factor. The term is selected to emphasize the dependence upon load-carrying
capacity.
It is the
number by which *the working load is multiplied to obtain P,.
4. MECHANICAL

PROPERTIES

OF STEEL

4.1 An outstanding

property of steel, which (as already mentioned)

sets it apart from other structural materials, is the amazing ductility
which it possesses. This is characterized by Fig. 2 which shows in
somewhat idealized form the stress-strain properties of steel in the initial
portion of the curve. In Fig. 3 are shown partial tensile stress-strain
curves for a number of different steels. Note that when the elastic
limit is reached, elongations from 8 to 15 times the elastic limit take
place without any decrease in load. Afterwards some increase in strength
is exhibited as the material strain hardens.
4.2 Although th; first application of plastic design is to structures
fabricated of structural grade steel, it is not less applicable to steels of

STRAIN

STRESS-STRAIN
CURVE OF S~-42STEEL IDEALIZED
19

Please purchase PDF Split-Merge on [Link] to remove this watermark.

SP: 6(6) - 1972

LASTIC
0

RANGE --W&RAIN

HARDENINO

If0

$0

3dx10-2

STR4IN

FIG. 3 STRESS-STRAIN
CURVEOF ST&~ AND ST-%

STRUCTURAL

STEELS

higher strength as long as they possess the necessary ductility. Figure 3

attests to the ability of a wide range of steels to deform plastically
with characteristics similar to steel conforming to IS: 226-1969*.
4.3 It is important to bear in mind that the strains shown in Fig. 3 are
really very small. As shown in Fig. 4, for ordinary structural steel,
final failure by rupture occurs only after a specimen has stretched some

IMALIZED

.Od!

o-11 lb50

CURVE

25

15

PERCENTAGE

ELONGATION

____c

FIG. 4 COMPLETE
STRESS-STRAIN
CURVEOF STRUCTURAL
STEEL
-.__*Structural steel (standard quality) (fourth rcuision).

20

Please purchase PDF Split-Merge on [Link] to remove this watermark.

SP: 6(6) - 1972

15 to 25 times the maximum strain that is encountered in plastic design.
Even in plastic design, at ultimate load the critical strains will not have
exceeded percentage elongation of about 1.5. Thus, the use of ultimate
load as the design criterion still leaves available a major portion of the
reserve ductility of steel which may be used as an added margin of
safety. This maximum strain of 1.5 percent is a strain at ultimate
load >n the structure not at working load. In most cases under working
load the strains will still be below the elastic limit.

5. MAXIMUM

STRENGTH

OF SOME ELEMENTS

5.1 On the basis of the ductility

of steel (characterized by Fig. 2) it is

now possible to calculate quickly the maximum carrying capacity of
certain elementary structures.
As a first example take a tension member such as an eye bar (Fig.. 5).
The stress is Q = P/A.
The lo&d v deflection relationship will be\ elastic until the yield
point is reached. As shown in Fig. 5 deflection at the elastic limit is
given by 6, = P,L/AE.

STRESS:

Ti
i

I
I

6 =- p
A

DEFLECTION:

UNRESTRICTED
PLASTIC
FLOW

$1
: 7:

6,.EL

PUL

AE

P,,

$A

FIG. 5 MAXIMUMSTRENGTHOF AN EYE BAR (DETERMINATE

STRUCTURE)
21

Please purchase PDF Split-Merge on [Link] to remove this watermark.

SP:

6(6)

- 1972

Since the stress distribution is uniform across the section, unrestricted plastic flow will set in when the load reached the value given by

P,=

ayA

This is, therefore, the ultimate load. It is the maximum load the
structure will carry without the onset of unrestricted plastic flow.
As a second example consider the three-bar structure shown in
Fig. 6. It is not possible to consider the state of stress by statics alone
and thus it is indeterminate. Consider the elastic state. From the
equilibrium condition there is obtained:

. ..
.. .
.. .
. ..(l)
where T1 is the force in bars 1 and 3 and T, the force in the bar 2.
2Tl+TI=P

PARTIALLV

EWL~IUY:

PLASTIC

21, + 12-P

VLLO

FIG.6

[Link]

pr 212 *

LIMIT:

2byh

OEFLECTION

PLASTICAND ELASTICANALYSISOF AN INDETERMINATESYSTEM

The next condition to consider is continuity.

For a rigid cross bar,
the total displacement of Bar 1 will be equal to that of Bar 2. Therefore:
l-14

T&a

AE - AE
T,= 2

(as L,=

.**
2L.J

.*

***

. ..

...

. . . (2)
. ..(3)

With this relationship between T, and T, obtained by the continuity

condition, using Eq (1) it is found that:

T&.

.. .

. ..

. ..

. ..

. ..(4)

22

Please purchase PDF Split-Merge on [Link] to remove this watermark.

SP: 6(6) - 1972

The load at which the structure
will tirst yield may then be determined
by substituting
in Eq (4) the maximum
load which T, can reach,
namely, oYA .
Thus,
P,,=
The

2T,=

displacement
8=
Y

2a,A

at the

q2&

or;

...
yield

load

would

be determined
. . .

2E

Now, when the structure

is partially
two-bar structure except that a constant
Bar 2 (the member is in the plastic
until the load reaches the yield value
how easily it is possible to compute

...

. ..(5)
from:
. ..(6)

. . .

plastic it deforms as if it were a

force equal to CT,,Ais supplied by
range).
This situation
continues
Notice
in the two outer bars.
the ultimate
load:

P, = 3ayA

...

. ..(7)

...

The basic reason for t,his simplicity

is that the continuity
condition
need not be considered
when the ultimate
load in the plastic range is
being computed.
The load-deflection
relationship
for the structure
shown in Fig. 6
is indicated
at the bottom.
Not until the load reaches that value
computed by a plastic analysis (Eq 7) did the deflections commence to
increase rapidly.
The deflection when the ultimate
load is first reached
can be computed
from:
I.

The three

essential

a) Each portion
condition,

features

of this simple

of the structure

plastic

. ..(8)

...
analysis

(each bar) reached

are as follows:
a plastic

yield

b) The equilibrium
condition
was satisfied at ultimate
load, and
c) There was unrestricted
plastic flow at the ultimate
load.
These same features are all that are required to complete the plastie
analysis of an indeterminate
beam or frame, and in fact, this simple
example illustrates
all of the essential features of a plastic analysis.
6. HISTORICAL

DEVELOPMENT

6.1 The concept of design based on ultimate

load as the criterion
is
more than 40 years old ! The application
of plastic analysis to structural
design appears to have been initiated by Dr Gabor Kazinszy, a Hungarian, who published
results of his Tests2 of Clamped Girders as early as
23

Please purchase PDF Split-Merge on [Link] to remove this watermark.

SP: 6(6) - 1972

1914.
He
also
now current, and
carried out.

suggested
designs of

analytical procedures similar to those

apartment-type
buildings were actually

6.2 In his Strength of Materialsa, Timoshenko refers to early suggestions

to utilize ultimate load capacity in the plastic range and states as follows:
Such a procedure appears logical in the case of steel structures
submitted to the action of stationary loads, since in such cases a
failure owing to the fatigue of metal is excluded and only failure
due to the yielding of metals has to be considered.
Early tests in Germany were made by Maier-Leibnitz who showed that
the ultimate capacity was not affected by settlement of supports of
continuous beams. In so doing he corroborated the procedures previously developed by others for the calculation of maximum load capacity. The efforts of Van den Broekl in USA and J. F. Baker6310 and
his associates in Great Britain to utilize actually the plastic reserve
strength as a design criterion are well known. Progress in the theory
of plastic structural analysis (particularly that at Brown University)
has been summarized by Symonds and Neal.
6.3 For more than ten years the American Institute of Steel Constnmtion, the Welding Research Council, the Navy Department and the
American Iron and Steel Institute have sponsored studies at Lehigh
University6**se. These studies have featured not only the verification
of this method of analysis through appropriate tests on large structures,
but have given particular attention to the conditions that should be
met to satisfy important secondary design requirements.
6.4 Plastic design has now come of age . It is already a part of the
British Standard specifications and numerous structures both in Europe
and North America have been constructed to designs based upon the
plastic method.
IS: 800-1962* permits the use of Plastic Theory in
the design of steel structures (see 13.5.1 of IS: 800-1962*).

*Code of practice
-

for use of structural

steel in general

building

construction

(rcuised).

24

Please purchase PDF Split-Merge on [Link] to remove this watermark.

>T.$

SECTION
JUSTIFICATION

FOR PLASTIC DESIGN

7. WHY PLASTIC DESIGN

7.1 What is the justification for plastic design ? One could reverse the
question by asking, why use elastic design ? If the structure will
support
the load and otherwise meet its intended function, are the
magnitudes of the stresses really important ?
7.2 It is true that in simple structures the concept of the hypothetical
yield point as a limit of usefulness is rational. This is because the
ultimate load capacity of a simple beam is but 10 to 15 percent greater
than the hypothetical yield point, and deflections start increasing verv
rapidly at such a load. While it would seem logical to extend elastic
stress analysis to indeterminate structures, such procedures have tended
to overemphasize the importance of stress rather than strength as a
guide in engineering design and have introduced a complexity that now
seems unnecessary for a large number of structures.

7.3 Actually the idea of design on the basis of ultimate load rather than
allowable stress is a return to the realistic point of view that had to be
adopted by our forefathers in a very crude way because they did not
possess knowledge of mathematics and statics that would allow them
to compute stresses.
7.4 The introduction of welding, of
to studies of the ultimate strength
to achieve complete continuity at
it economically.
The full strength
mitted to another.

course, has been a very real stimulus

of frames. By welding it is possible
joints and connectionsand to do
of one member may thus be trans-

7.5 It has often been demonstrated that elastic stress analysis cannot
predict the real stress-distribution
in a building frame with anything
like the degree of accuracy that is assumed in the design. The work
done in England by Prof. Baker and his associates as a forerunner to
their ultimate strength studies clearly indicated this.
in
7.6 Examples of imperfections that cause severe irregularity
measured stresses are: differences in beam-column connection fit-up and
flexibility, spreading of supports, sinking of supports, residual stresses,
25

Please purchase PDF Split-Merge on [Link] to remove this watermark.

SP: 6(6) - 1972

flexibility
assumed
where actually
and points of stress concentration.
not influence the maximum
plastic
7.7 Assuming
to justify our
criteria there
economy and

there is rigidity
(and
Such factors, however,
strength.

vice-versa),
usually do

that stress is not the most rational design criteria, in order

further consideration
of maximum
strength
as the design
must be other advantages.
There are two such advantages:
simplicity.

7.8 Since there is considerable

reserve of strength
beyond the elastic
limit and since the corresponding
ultimate
load may be computed quite
accurately,
then structural
members
of smaller
size will adequately
support the working loads when design is based on maximum
strength.
Numerous
demonstrations
of this will be made later in this bandbook.
7.9 The second
feature
was simplicity . An analysis
based upon
ultimate
load possesses an inherent
simplicitv
because the elastic condition of continuity
need no longer be considered.
This was evident
from a consideration
of the three-bar truss in Section A (Fig. 6) and the
examples of Section D will demonstrate
this further.
Also the imperfections mentioned
above usually
may be disregarded.
7.10 As already
mentioned
the concept is more rational.
By plastic
analysis the engineer can determine
with an accuracy that far exceeds
his presently
available
techniques
the real maxirnum
strength
of a
Thereby the factor of safety has more real meaning than at
structure.
present.
It is not unusual for the factor of safety to vary from I.65 up
to 3 or more for structures
designed according to conventional
elastic
methods.
7.11 Thus the application
of plastic
analysis
should
be considered
seriously because it provides a less-expensive
structure,
it is a similar
design office technique,
and it constitutes
a rational design basis.
Further, these concepts are verified by tests and (as we shall now see) they
have been used consciously
or unconsciously
in conventional
design
practice.

8. [Link]

OF

STRESS

AS THE

DESIGN

CRITERION

8.1 The question

immediately
arises, will it not be possible simply to
change the allowable stress and retain the present stress concept ? While
theoretically
possible, the practical
answer is no . It would mean a
different working stress for every type of structure
and would vary for
different loading conditions.
8.2 To a greater
a structure
has

extent
always

than we may realize, the maxnnum

been the dominant
design criteria.

strength of
When the

26

Please purchase PDF Split-Merge on [Link] to remove this watermark.

SP: 6(6) - 1972

permissible
working stress of 1 400 k&ma has led to designs that were
consistently
too conservative,
then that stress has been changed.
Thus
the benefits of plasticity
have been used consciously
or unconsciously
in design.
It is also evident to most engineers that present design procedures completely
disregard
local over-stressing
at points
of stressconcentration
like bolt
holes,
notches,
etc.
Long experience
with
Thus,
similar structures
so designed shows that this is a safe procedure.
the stresses
that
are calculated
for design purposes
are not true
maximum
stresses at all, they simply provide an index for structura
design.
8.3 A number of examples will now be given in which the ductility
of
steel has been counted upon (knowingly
or not) in elastic design.
It
should be borne in mind that plastic analysis has not generally
been
used as a basis for determining
these particular
design rules and as a
result the so-called elastic stress formulas have been devised in a rather
haphazard
fashion.
A rational
basis for the design of a complete
steel frame
(as well as its details)
can only be attained
when the
maximum
strength
in the plastic
range is adopted.
as the design
criterion.
8.4 Such examples
are the following and are listed in two categories:
(a) factors that are neglected .because
of the compensating
effect of
ductility;
and (b) instances
in which the working stresses have been
revised
because
the normal value
was too conservative.
Several
examples of each are given:
a) F&tom

that are neglected:

1) Residual

stresses

2)

stresses

rolling) ;
Residual

(in the

case of flexure

resulting

stresses;
3) Erection
settlements;
4) Foundation
at points
5) Over-stress

of

from

the

due to cooling

cambering

stress-concentration

of

(holes,

after

beams;

etc) ;

6) Bending stresses in angles connected in tension by one leg only;

at points of bearing:
7) Over-stress
stress-distribution
in splices, leading to design
8) Non-uniform
of connections
on tbe assumption
of a uniform
distribution
of stresses among the rivets, bolts, or welds;
arising from the cantilever
9) Difference in stress-distribution
as compared
with
the
portal
method
of wind stress
analysis ;
27

Please purchase PDF Split-Merge on [Link] to remove this watermark.

SP:6(6)- 1972

10) IS: 800-1962specifies the following values for bending stresses:

1 650 kgf/cm for rolled sections,
1 575 kgf/cm* for plate girders, and
1 890 kgf/cm2 in flat bases.
b) Revisions

in working

stress due to reserve plastic

strength:

1) Bending stress of 2 109 kgf/cm2 (or 30 ksi) in round pins (in

AISC specification) ;

2) Bearing stress of 2 812 kgf/cms (or 40 ksi) on pins in double

shear ;
3) Bending
tures at
4) Bending
sae$ons

stress of 1 687 kgf/cm2 (or 24 ksi) in framed strucpoints of interior supports;

stress of 1 65? kgf/cm2 and. 1 575 .kgf/cm2 for rol!ed
and plate girders respectively
(m IS: 800-1962 ),

5) Bending stress of 1 890 kgf/cms in slab bases (iu IS: 800-1962*).

Consider Item (a) (1) for example. All rolled members contain
residual stresses that are formed due to cooling after rolling or due to
cold-straightening.
A typical wide flange shape with a typical residual
stress pattern shown in Fig. 7. When load carrying bending stresses
are applied, the resulting strains are additive to the residual strains
already present. As a result, the final stress could easily involve
yielding at working load. In the example of Fig. 7, such yielding has

RESIDUAL

(b)

(a)

FIG. 7

STRESS

APPLIED

STRESS

(cl

FINAL

STRESS
(d)

(0)

RESIDUAL STRESSESIN A ROLLED BEAM SECTION

*Code of practice for use of structural steel in general building construction

(mviscd).

28

Please purchase PDF Split-Merge on [Link] to remove this watermark.

SP: 6(6) - 1972

occurred both at the

tension flange. Thus,
influence is neglected
cause yielding in the

compression flange tips and at the centre of the

it is seen that cooling residual stresses (whose
and yet which are present in all rolled beams)
flange tips even below the working load.

8.5 Structural members experience yield while being straightened in

the mill, fabricated in a shop or forced into position during erection.
Actually, it is during these three operations that ductility of steel
beyond the yield point is called upon to the greatest degree. Having
permitted such yielding in the mill, shop and field, there is no valid
basis to prohibit it thereafter, provided such yield has no adverse effect
upon the structure.
As an illustration of item (a) (3) in the list in 8.4,
Fig. 8 shows how erection forces will introduce bending moment into a
[Link] prior to the application of external load (see first line for P = 0).
Although the yield-point load is reduced as a result of these erection
moments (in the second line of the figure, the yield-point load has been
reached for case 2). them is #o efect whatever on the maximum strength.
The reason for this is that redistribution of moment followed the onset
of yielding at the corners (case 2) until the plastic moment was reached
at the beam centre; therefore, the ultimate load moment diagrams for
cases 1 and 2 are identical.
8.6 Consider. next. the desien of a riveted or bolted joint [Item (a) (8)
is made that each fastener carries the

in 8.41. The common ass&ption

p-0

P=Py

(21

P=P

FIG. 8 DEMONSTRATION
THAT ERECTIONSTRESSESDO NOT INFLUENCE
ULTIMATE LOAD

29

Please purchase PDF Split-Merge on [Link] to remove this watermark.

SP: 6(6) - 1972

When
same shear force.
This is true only in the case of two fasteners.
more are added (Fig. 9), then as long as the joint remains elastic, the
For example,
outer fastners should carry the greater portion of the load.
with four rivets, if each rivet transmitted
the same load, then, between
rivets C and D one plate would carry perhaps three times the force in
the other.
Therefore, it would stretch three times as much and would
necessarily
force the outer rivet D to carry more load.
The actual
forces would look something like these shown under the heading Elastic .
What eventually
happens is that the outer rivets yield, redistributing
forces to the inner rivets until all forces are about equal.
Therefore,
the basis for design of a rivet joint is really its ultimate load and not the
attainment
of first yield.
8.7 A revised working stress example [see Item (b) (1) in 8.41 is shown
In a simple
in Fig. 10 and is concerned with the design of a round pin.
beam with wide flanges, when the maximum stress due to bending reaches
the yield point, most of the usable strength
has been exhausted.
However, for some cross-sectional
shapes,
much additional
load may be
carried
without
excessive
deflections.
The relation
between
bending
moment and curvature for wide flange and round beams is shown in Fig. 10.
The upper curve is for the pin, the lower for a typical wide flange beam, the
non-dimensional
plot being such that the two curves coincide in the
elastic range.
The maximum
bending strength
of the wide flange beam
is l-14 MY, whereas that of the pin is I.70 My. The permissible
design
stresses (for steel with yield stress 36 ksi) according
to specifications
of the American
Institute
of Steel Construction
are 1 550 kgf/cmz (or
22 ksi) for the wide flange beam and 2 320 kgf/cm2 (or 30 ksi) for the
round pin.

FIG. 9

REDISTRIBUTION

OF SHEAR

IN

THE FASTENERS
OF A LAP JOINT

30

Please purchase PDF Split-Merge on [Link] to remove this watermark.

SP: 6(6) - 1972

lo

A*Y

FIG. 10 MAXIMUM STRENGTHOF A ROUND PIN COMPAREDWITH THAT

OF A WIDE FLANGE BEAM
Expressing

these stresses as ratios of yield point s+- :

Wide flange: 0 = F5g = 0.61

QY
2 320
=*
Pin:
-=20=
0.91
=Y

For a simplv-supported beam the stresses, moments, and load all bear
a linear relationship to one another in the elastic range and thus:

P
p,=

M
6 = My

Therefore, the moment at allowable working stress (M,) in the wide

flange beam is O-61 My; for the pin, on the other hand, M,= 0.91 My.
What is the true load factor of safety for each case 1
Wide

M
?_!%
Range:F=p~=M*=

0.61 M,

Pin :

~&_2!%=
0.91 My

= 1.87

1.87

The exact agreement between the true fktors of safety with respect to
ultimate load in the two cases, while somewhat of a coincidence, is indicative of the influence of long years of experience on the part of engineers
31

Please purchase PDF Split-Merge on [Link] to remove this watermark.

SP: 6(6) - 1972

which has resulted in different permissible working stresses for various
Probably no such analysis as the
conditions resulting in practice.
foregoing influenced the choice of different unit stresses that give identical
factors of safety with various sections; nevertheless, the choice of such
stresses is fully justified on this basis. When years of experience and
common sense have led to certain empirical practices these practices are
usually justified on a scientific basis.
8.8 Permitting a 20 percent increase in the allowable working stress at
points of interior support in continuous beams represents another case
in which both experience ad a plastic analysis justify a revision in
working stresses.
9. EXPERIMENTAL

VERIFICATION

9.1 In the previous clauses some of the important concepts of the plastic
theory are described. How well does structural behaviour bear out the
theory ? Do structures reaily contain the ductility assumed ? If we
test a full size structure with rolled members will it actually carry the
load predicted by plastic analysis 1
9.2 The important assumptions made with regard to the plastic behaviour of structures
are recapitulated
in Fig. 11). In Lecture 4 of
Ref 12 (see Appendix A), the experimental confirmation of these assumptions is given, demonstrating the ductility of steel, the development of
plastic hinges in beams and %oiinections, and redistribution of moment.
In the last analysis, the most important verification of plastic theory is
that given by the results of full-scale tests and some of these will now
be presented.
9.3 Typical structures were tested both in USA and other countries.
The structure carried the predicted ultimate load, the load-deflection
curve being shown in Fig. 12.
9.4 Further tests conducted on frames fabricated from rolled sections
have shown that the actual strength of even the weakest structure was
within 5 percent of its predicted ultimate load an agreement much better
than obtained at the so-called elastic limit ~J6*20*a0@*u. In tests on
beams with three supports, applying the vertical load, the central support
was raised until the yield point was: first reached, with the result that
application of the first increment of external load caused the structure
to yield. In spite of this, the computed ultimate load was attained.
In the tests conducted on pinned and fixed basis and with flat, saw
tooth ant1 gabled roofs, the ultimate load computed by the plastic
theory was reached and in numerous cases it exceededa14s*44.
32

Please purchase PDF Split-Merge on [Link] to remove this watermark.

SP: 6(6) - 1972

-G

MOMENT CAtXtTY=

FlASTC MOMENT

-9
PlASTC

HINGE

UTIMATEWAD HING&--i&ANISM

FIG. 11 ASSUMPTIONS
MADE IN REGARD TO PLASTIC
BEHAVIOUROF STRUCTURES

10. THE

CASE

FOR PLASTIC

DESIGN

10.1 As summarized in the preceding paragraphs the results of tests have

verified the theory of plastic analysis. Is the engineer now justified in
giving further attention to the method of plastic analysis, in studying
it, and in applying it to the appropriate design problems 7 The answer
is yes.
The case for plastic design is illustrated by the following observations :
4 The reserve in strength above conventional working loads is
considerable in indeterminate steel structures.
Indeed, in some
instances as much load-carrying capacity is disregarded as is
used in conventional design.
..
33

Please purchase PDF Split-Merge on [Link] to remove this watermark.

k..

SP: 6(6) - 1972

016
0

25

50

75

100
125
DEFLECTION IN cm

150

175

200

225

FIG. 12 LOAD-DEFLECTIONCURVE OF A TEST FRAME

b) Use
d
4

of ultimate load as the design criterion provides at least

the same margin of safety as is presently afforded in the elastic
design of simple beams (Fig. 1).
At working load the structure is still in the so-called elastic range
(Fig. 1).
In most cases, a structure designed by the plastic method will
deflect no more at working load than will a simply-supported
beam designed by conventional methods to support the same
load (Fig. 1).
Plastic design gives promise of economy in the use of steel, of
savings in the design office by virtue of its simplicity, and of
building frames more logically designed for greater over-all
strength.

It is important
upon the maximum
to details.
These
earlier and treated

to bear in mind that dependence may be placed

plastic strength only when proper attention is given
are the secondary design considerations mentioned
in Section E.

34

Please purchase PDF Split-Merge on [Link] to remove this watermark.

SECTION
FLEXURE
11. ASSUMPTIONS

OF BEAMS

AND CONDITIONS

11.1 Certain of the fundamental concepts of plastic analysis were prs

sented in Section A (see 3 and 4). The examples there, however, were
limited to cases of simple tension and compression. The next objective
is to determine how a beam deforms beyond the elastic limit under the
action of bending moments, that is, to determine the moment-curvature
(M-9) relationship.
The assumptions and conditions used in the following development are:
a) strains are proportional to the distance from the neutral axis
(plane sections under bending remain plane after deformation).
b) the stress-strain relationship is idealized to consist of two straight lines:
u = E
u = 0,

(O<e<eJ

Normal force:
Moment :

. ..(9)

**-. ..
(ey<c<co) >
The complete stress-strain
diagram is shown in Ffg. 4 and
is shown in an idealized form in Fig. 2. The properties in
compression are assumed to be the same as those in tension.
Also, the behaviour of fibres in bending is the same as in tension.
The equilibrium conditions are as given by Eq (10):
P =
M=

I AlCiZ
I Af&Z

adA
. ..(lO)
tsdA.y

where o is the stress at distance j from the neutral axis.

Deformations are sufficiently small so that + = tan 4 (+ =
curvature).

12. BENDING

OF RECTANGULAR

BEAM

12.1 The moment-curvature

relationship in the plastic range and the
magnitude of the maximum plastic moment are developed by following
the same processes as in elastic analysis, that is, consider the deformed
structure and obtain the corresponding cur$-ature and moment. The
development of strain and stress distribution as a rectangular beam is
bent in. successive stages beyond the elastic limit (Stage 1) and up to
35

Please purchase PDF Split-Merge on [Link] to remove this watermark.

SP: 6(6) - 1972

STRESS
vs
STRAIN

t
6

STRAINS

STRESSES

OISTRBUTIONS

FIG. 13 PLASTIC BENDING OF RECTANGULARBEAMS

the plastic limit (Stage 4) is shown in Fig. 13. The strain distribution
is first selected or assumed and this fixes the stress-distribution.
12.2 Let us now trace the stages of yield stress penetration in a rect,angular beam subjected to, a progressive increase in bending moment.
At the top of Fig. 13 is replotted for reference purposes the stress-strain
relationship. At Stage 1, as shown in the next line of Fig. 13, the strains
have reached the yield strain. When more moment is applied (say to
Stage 2), the extreme fibre strains are twice the elastic limit value. The
situation is similar for Stage 3 (ran = 4~~). Finally; at Stage 4 the
extreme fibre has strained to Ed:.
12.3 What are the stress distributions that correspond to these strain
diagrams 1 These are shown in the next line of Fig. 13. As long as the

Please purchase PDF Split-Merge on [Link] to remove this watermark.

SP: 6(6) - 1972

strain is greater than the yield value G,,, as could be noticed from the stressstrain curve that the stress remains constant
at o
The stress distributions, therefore, follow directly from the assume 8 strain distributions.
As a limit we obtain the stress block - a rectangular
is very close to the stress distribution
at Stage 4.
A new term introduced
in Fig. 13 is the curvature.
relative rotation
of two sections a unit distance
apart.
the first assumption
(as in elastic bending):

pattern

which

This is the
According to

.(11)
where P - radius of curvature
and l, the strain at a distance y from the,
J us t as it is basic to the fundamentals
of elastic analysis,
neutral axis.
of bending
moment
to this curvature,
4, is a basic
the relationship
concept in plastic analysis.
The expressions for curvature
and moment (and, thus, the resulting
M-4
curve) follow directly from Fig. 13. Curvature
at a given stage
is obtained- from a particular
stress-distribution*.
The corresponding
moment-value
is obtained
by integration
of stress-areas.
The derivation
of expressions
for curvature
and moment now follow.
Eq

Stage 2 of the example

11 the curvature
thus

..

=$-

shown in Fig. 13 is shown in Fig. 14.

becomes:
. ..

. ..

From
. ..(12)

where y,, is the ordinate to the neutral axis to the farthest still elastic
fibre.
To compute the bending moment for this same Stage 2, the stress
distribution
of Fig. 14 is divided into parts in Fig. 15. The moment of

(jq_~
$+
YIELD ZONE

FIG. 14

MOMENT

STRAIN

$iG$
'QY

STRESS

STRESS DISTRIBUTIONIN A PARTIALLY PLASTIC RECTANGULAR

CROSS-SECTION

*Even though curvature is a measure of strain distribution, the stress-distribution

diagram is used, since in the elastic range, the stress varies linearly with strain.
37

Please purchase PDF Split-Merge on [Link] to remove this watermark.

SP: 6(6)

- 1972

t
$

4=

SY

FIG. 15

6Y

6Yh

6YG

STRESSELEMENTSOF A PARTIALLYPLASTICDESIGN

resistance
may thus
and a plastic part

be considered
(u&),
or:

as being

made up of an elastic

...

M = crJ,+c&, = a,S,+uJ--a&e

where the subscripts

of the cross-section,
Fig.

Equation
14:

M=

YO

0
2

13 may also

be derived

directly

and plastic

from Eq $0.

. ..(13)
portions

Referring

to

crdA.y

B and p refer to the elastic

respectively.

(u$,)

uy

42

=2
[Link].y+2 u,.bdy.y

YO

Yy?bay

o-

Yo

+=,2 j:r

[Link]

YO

5
0

d/2

c$ibdy.y+2 u,.bdy.y
0
I YO
...

...

. ..(13a)

uys,
+uyz,

=
The quantity
Z is a property
of a cross-section
that corresponds
in
importance
to the section modulus, S. It is called the plastic modulus,
and (for symmetric
sections)
represents
twice the statical
moment
(taken about the neutral axis) of the plastic section area above or below
that axis.
General methods for computing
2 will be discussed later.
For the rectangular
section, necessary
values for section modulus,
S, and plastic modulus,
Z for use in Eq 13 are:
Z; =

2by,$ = by:

S = b(2YJ2
2
22
=jby;=3
6

z* = z-z,

z=4

bd2

...

. ..

. ..(14)

38

Please purchase PDF Split-Merge on [Link] to remove this watermark.

SP: 6(6) - 1972

Thus,

the bending

moment

in terms

of Z is given

by;

...
The maximum
zero or:

moment

is obtained

M,,=a,Z

when

...

Mp is called the plastic

moment

. ..

the elastic

...

part

. ..(15)
is reduced

...

to

. ..(16)

From the equations just derived for curvature

and moment, we are
now in a position to write the desired moment-curvature
relationship.
In terms of y,,, then:
I..

In terms

of c$, using

Eq

The following non-dimensional

sides of Eq 18 by My = 0,s:
3

wY=2-

1
c

. ..(17)

. ..

..

12:

M=o,(Z-&)
M

...

($y<$<co)

relationship

1
3

is obtained

MY)

by dividing

. ..

($4 I

both

. ..(19)

***

The resulting
non-dimensional
M-4 curve for a rectangle is shown in
Fig. 16. The numbers
in circles in Fig. 16 correspond
to stages of

5
*@Y

FIG.

16

10

NON-DIMENSIONAL MOMENT-CURVATURE RELATIONSHIP FOR

RECTANGULAR BEAM

39

pt-,

ii
w.

Please purchase PDF Split-Merge on [Link] to remove this watermark.

SI: 6(6) - 1972

Fig. 13. Stage 4, approached as a limit, represents complete plastic yield
of the cross-section, where Mp = uJ.
Note that there is a 50 percent
increase in strength above the computed elastic limit (Stage 1) due to
this plastification of the cross-section. This represents one of the
sources of reserve strength beyond the elastic limit of a rigid frame.
The ratio of the plastic moment (Mfi) to the yield moment (M,),
representing the increase of strength due to plastic action, will be a
function of the cross-section form or shape. Thus the shape factor
is given by:
. ..(20)
For the rectangular
indicated in Fig. 16.
13. BENDING

beam

OF WIDE

being considered, .f = y

FLANGE

+ -F = 1.50

as

BEAM

13.1 The action of a wide flange beam under bending moment is diagrammatically shown in Fig. 17. If it is assumed that all of the material
in a wide flange shape is concentrated in the flanges then (when the
elastic limit is reached) the compression flange shortens at constant load
and the tension flange lengthens at constant load. The resulting
moment is, therefore, constant; the member acts just like a hinge except
that deformation occurs under constant moment (the plastic-hinge
moment).
13.2 A more realistic picture of the moment-curvature relationship of
a wide flange shape is shown in Fig. 18. point 1 is the elastic limit; at
point 2 the..member is partially plastic and at point 3 the cross-section
approaches a condition of full plastic yield.

PLASTIC HINGE ( M=Mp)

FIG.

17

IDEALIZED MOMENT-CWRVATURERELATIONSHIP FOR WIDE

FLANGE BEAM
40

Please purchase PDF Split-Merge on [Link] to remove this watermark.

SP: 6(6) - 1972

FIG. 18 TYPICAL THEORETICAL

MOMENT-CURVATURE
RELATIONSHIP
OF
WIDE FLANGEBEAM

The magnitude of the moment, Mp, may be computed directlp from

the stress distribution shown for Point 3. As shown in Fig. 19 it is
equal to the couple created by the tensile and compressive forces. The
moment due to each of these forces is equal to the product of the yield
stress, a,, and the area above the neutral axis (A/Z) multiplied by the
distance ji measured to the centre of gravity of that area.
ELASTIC

PLASTIC

up=

WY 03s

FIG. 19

e,i.*
3

7.2

= uyz
Z-g_

$,F
,f
,\
I
. c <.f

ELASTICAND PLASTICLIMITMOMENTS
41

_:

Please purchase PDF Split-Merge on [Link] to remove this watermark.

SP: 6(6) - 1972

Thus :
Mp=2

CF,,; jj = ay.4g

The quantity
Ajj is called
therefore, as before,

the plastic

...
modulus

...
and

...

M* = a,Z

. ..(21)

is denoted
...

by 2;
. ..(16)

The plastic modulus, 2, is thus equal to the combined statical

of the cross-sectional
areas above and below the neutral axis,
stress. at every point on these areas is the same.

moments
since the

The moment-curvature
relationship
may be developed
for wide
flange sh&pes by the same procedure as outlined for a rectangular
crossDue to variation
of width of section with depth,
separate
section.
when yielding is limited to the flanges and
expressions
arc necessary
when yielding
has pcnctratcd
to the web.
the

For Case 1, in which the yield zone has pcnetratcd

part way through
flanges (Fig. 20), the non-dimensional
M-4
curve becomes:

~y=:(1_~~)+~[1__3(~.)al

.. .

. ..(22)

(b
l<&<($$
>
For Case 2, in which yielc!ing has penetrated
through the flanges and into
the web (Fig. 21), the non-dlmensioual
M-4 curve becomes:

,..(23)

FIG. 20

PLASTICSTRESSDISTRIBUTION
IN WIDE FLANGEBEAM CASE 1: PARTIAL YIELDING

42

Please purchase PDF Split-Merge on [Link] to remove this watermark.

SP: 6(6) - 1972

LU
FIG. 21

PLASTIC STRESSDISTRIBUTION IN WIDE FLANGE BEAM CASE 2: PARTIAL

YIELDING

The curve resulting

from Eq 22 and 23 is shown in Fig. 18 for a
(The stress-distributions
correspond
to the
typical
wide flange shape.
numbered
points on the M-4
curve.)
It will be noted that the shape
factor is smaller than for the rectangle
(compare Fig. 16), the average
value of f for all wide flange beams being 1.14. Correspondingly
there is a more rapid approach
to kfP when compared with rectangle.
As a matter of fact when the curvature
is twice the elastic limit value
(Stage 2 of Fig. 18) the moment has reached to within 2 percent of the
full MP -value.
14. PLASTIC

HINGE

14.1 The reason a structure

will support the computed
ultimate
load is
that plastic hinges are formed at certain critical sections,
What is the
plastic
hinge ? What
factors
influence
its formation ? What
is its
importance
?
The M-4
curve is characteristic
of the plastic hinge (Fig. 18). Two
features
are particularly
important :
a) the rapid
approach
to M = MP = o,Z;
and
b) the indefinite
increase
in 9 at constant
M.
An idealized M-4
curve is obtained
by assuming (for a wide flange
shape) that all of the material is concentrated
in the flanges as shown
in Fig. 17. The behaviour
shown there is of basic importance
to plastic
analysis.
According to it, the member remains elastic until the moment
Thereafter,
rotation occurs at constant
moment;
that is,
reaches Mp
the member acts as if it were hinged except with a constant
restraining
This, then, is the plastic
hinge.
moment,
Mp
According
discrete points

to the idealization
of Fig. 17, plastic
hinges form at
Actually the hinge extends over that part
of zero length.
43

Please purchase PDF Split-Merge on [Link] to remove this watermark.

i*,*......

.__., _*+...._.
.

SP:6(6)- 1972
of the beam whose bending moment is greater than M
is dependent
on the loading and geometry.
It is justifie 8'to
distribution
, however, and the length of hinge is assumed
Closely related to the plastic hinge is the plastic modulus,
already been defined for the symmetrical
sections as twice
moment
about the neutral
axis, of the half sectional
area.
earlier, 2 = Ay.
For wide flange beam shapes, the quantity
determined
directly
from the properties
of split tees and
. ..
where y1 is the distance

from the flange

That length
neglect this
to be zero.
2. It has
the statical
As noted
y may be
thus:

...

to the centre

of gravity.

The shape factors, already defined as f = Z/S, varies for wide flange
shapes from 1.09 to 1.22. The mode is 1.12 and the average is 1.14 for
I shapes.
Examples of the ratio of Z/S = f for symmetrical
shapes other
than the wide flange are shown in Fig. 22 and 23.
For sections ~with symmetry
only about an axis in the plane of
bending, the neutral axis at the plastic moment condition follows directly
from Eq 10. The general definition
for Z is The cbmbined
statical
moments of the cross-sectional
areas above and below the neutral axis .
Since P = 0, and B = I+, for equilibrium
the area above the neutral
axis should equal that below.
Thus, for a triangular-section
in Fig. 22
the elastic neutral axis is at a distance of 2/3d from the toe, while the
plastic neutral axis is at a distance of dl/Z.

FIG. 22

NEUTRAL AXIS OF A TRIANGULARSECTION

SECTION

FIG. 23

SHAPE FACTORS OF COMMON SYMMETRICALSECTIONS

44

Please purchase PDF Split-Merge on [Link] to remove this watermark.

,I

__

_.

_.

,..,^_._.,_.

__I....

SP: 6(6) - 1972

In addition to the shape factor whose influence on strength has
already been described, several other factors influence the ability of members to form plastic hinges. Some of these factors are important from
the design point of view (such as shear, axial force, instability) and are
treated in Section E. Others are primarily of academic interest in so far
as rout&e design applications are concerned. Factors affecting the
bending strength and stiffness of beams have been listed in Chapter 2 of
Ref 9 (see Appendix A) with further reference being made to other
sources of information on each factor.
The following definitions or principles briefly summarize this clause
and are ir;lportant to a later understanding of plastic analysis:
a) A @last& hiqe is a zo%e of yielding dzle to jlexure in a strzrctural
memberAlthough its length depends on the geometry and
loading, in most of the analytical work it is assumed that all plastic
rotation occurs at a point. At those sections where plastic hinges
are located, tb member acts as if it were hinged, except with a constant restraining moment Mp, (Fig. 17).
b) Plastic hinges form at points of maximum moment - Thus in a
framed structure with prismatic members, it would be possible
for plastic hinges to form at points of concentrated load, at the
end of each member meeting at a connection involving a change
in geometry, and at the point of zero shear in a span under distributed load.
c) The plastic moment, M,, equals aJ.
d) The shape factor

f = Z/S=

AY
-.- is one source of reserve strength
I/c >

beydnd the elastic limit.

Application of the plastic hinge concept
in 15.
15. REDISTRIBUTION

to analysis is illustrated

OF MOMENT

15.1 The second factor contributing to the reserve of strength is called

redistribution of moment and is due to the action of the plastic hinges.
As load is added to a structure eventually the plastic moment is reached
at a critical section -the
section that is most highly stressed in the
elastic range. As further load is added, this plastic moment value is
maintained while the section rotates. Other less highly-stressed sections
maintain equilibrium with the increased load by a proportionate increase in
moment. This process of redistribution of moment due to the successive
formation of plastic hinges continues until the ultimate load is reached.
15.2 This is exactly the process that took place in the case of the threebar truss of Fig. 6 except that tensile forces instead of moments were

45

Please purchase PDF Split-Merge on [Link] to remove this watermark.

SP: 6(6) - 1972

involved. When the force in Bar 2 reached the yield condition it remained constant there while the forcescontinued to increase in Bars 1
and 3. The ultimate load was reached when all critical bars became
plastic.
15.3 The phenomenon of redistribution of moment will now be illustrated
with the case shown in Fig. 24, a fixed ended beam with a concentrated
load off-centre.
As the load P is increased the beam reaches its elastic
limit at the left end (Stage 1). The moments at sections B and C are
less than the maximum moment. Note that in this example WCwill consider the idealized M-4
relationship as shown in the lower left-hand
portion.
(The dotted curve shows the more precise behaviour).

As the load is further increased, a plastic hinge eventually forms at

Section B. The formation of the plastic hinge at A will permit the
beam to rotate there without absorbing any more moment. Referring
to the load-dcgection cut-w*: immediately below the moment diagrams
the deflection is increasing at a greater rate.

STAGE 2.

STAGE 1

STAGE 3

t
M

DEFLECTION

FIG. 24

RE-DISTRIBUTIONOF MOMENTS

46

Please purchase PDF Split-Merge on [Link] to remove this watermark.

SP: 6(6) - 1972

Eventually, at Stage 3, when the load is increased sufficiently to

!orm a plastic hinge at Point C, all of the available moment capacity of
the beam will have been exhausted and the ultimate load reached.
It is evident from the load-deflection curve shown in the lower part
of the figure that the formation of each plastic hinge acts to remove one
of the indeterminates in the problem, and the subsequent load-deflection
In the elastic
relationship will be that of a new (and simpler) structure.
range, the deflection under load can be determined for the completely
elastic beam. Starting from Point 1 the Segment l-2 represents the
load-deflection curve of the beam in sketch b loaded within the elastic
range. Likewise the load-deflection curve of the beam in sketch c looks
similar to the portion 2-3.
Beyond Stage 3 the beam will continue to deform without an increase
inload just like a link mechanism if the plastic hinges were replaced by
real hinges. This situation is called a mechanism in the somewhat
special condition that motion is possible without an increase in load.
Two further fundamental concepts in addition to the four listed
in 14 are in summary of this clause and arc demonstrated by Fig. 24:
a) The formation of plastic hinges allows a subsequent redistribution
of moment until Mp is reached at each critical (maximum)
section.
b) The maximum load will be reached when a sufficient number of
plastic hinges have formed to create a mechanism.
On the basis of the principles just discussed one may readily visualize. how to compute the ultimate load: Simply sketch a moment diagram
such that plastic hinges are formed at a sufficient number of sections to
allow mechanism motion . Thus in Fig. 25, the bending moment diagram for the uniformly-loaded, fixed-ended beam would be drawn such

FIG. 25 MOMENTDIAGRAMAT VARIOUSSTAGESFORFIXED-ENDEDBEAM

WITH UNIFORMLY
DISTRIBUTED
LOAD
47

Please purchase PDF Split-Merge on [Link] to remove this watermark.

..z:
,i
; :
9.

L_

SP : 6(6) - 1972
that Mp is reached

mechamsm is formed.
WL

at the two ends and the centre.

By equilibrium :

In this way a

-!L=2M*
8

How does this compare with the load at first yield ? At the elastic
limit (see dotted moment-diagram in Fig. 25) we know from a consideration of continuity that the centre moment is one-half the end moment.
Thus:
lvyL =Mv$T=TMY
8

3MY

12M
WY5TY
Therefoie,

the reserve strength due to redistribution of moment is:

Wy 163$/L _ 4 M,
W; -- 12MyjL - 3 pY
Considering the average shape factor of wide flange beams, the total
reserve strength due to redistribuiion and shape factor (plastification)
is:
w,
4
w, = f X 1.14 = 1.52
For this pa&c&r
problem, then, the ultimate load was 52 percent
greater than the load at first yield, representing a considerable margin
that is disregarded in conventional design.
There are other methods for analyzing a stticture for its ultimate
load, in particular the Mechanism Method (to be described later) which
starts out with an assumed mechanism instead of an assumed moment
diagram. But in every method, there are always these two important
features :
a) the formation of plastic hinges, and
b) the development of a mechanism.
With these fundamental concepts regarding the mechanical properties of steel and the flexure of beams we are now in a position to
examine the methods of plastic analysis.

48

Please purchase PDF Split-Merge on [Link] to remove this watermark.

SECTION

PLASTIC ANALYSIS

16. FUNDAMENTAL

PRINCIPLES

16.1 With the evidence presented in Section B that full-size structures

behave as predicted by plastic theory and having considered in Section C
the plastic behaviour of beams, we may now proceed to a consideration
of the methods of plastic analysis. The objective of this Section is to
describe brie5y the fundamental principles upon which plastic analysis
rests and then to describe how these principles are used in analyzing
continuous beams and frames.
The basis for computing the ultimate load (or maximum. plastic
strength) is the strength of steel in the plastic range. As shown in 3,
structural steel has the ability to deform plastically after the yield-point
is reached. The resulting Aat stress-strain characteristic assures dependable plastic strength, on the one hand, and provides an effective limit
to the strength of a given cross-section making it independent of further
deformation.
Thus, when certain parts of a structure reach the yield
stress, they maintain that stress while other less-highly-stressed parts
Since all critical secdeform until they, too, reach the yield condition.
tions eventually reach the yield condition, the analysis is considerably
It is not of interest
simpli5ed because only this fact need be considered.
how the stresses are redistributed;
we should only ascertain that they
did. We are thus freed from the often laborious calculations that result
from the necessity of considering the continuity conditions that are
essential to elastic analysis.
While elastic and plastic analysis were compared at the outset in 2
from the design point of view, it is of interest now, to compare them
as regards to the fundamental conditions satisfied by each.
Whatever method of plastic analysis is used, it should satisfy the
following three conditions that may be deduced from what has been said
in 15:
a) Mechanism condition (ultimate load is reached when a mechanism forms),
b) .Equilibrium condition (the structure must be in equilibrium),
and
c) Plastic moment condition (the moment may nowhere be greater
than Mp).
49

Please purchase PDF Split-Merge on [Link] to remove this watermark.

._

,__*_. .

..-.

.,

_,

._e-...

x_ ,._,. .__*-.

. .._
-

.^

.-

I. . _

SP: 6(6) - 1972

Actually these conditions are similar to those in elastic analysis which

requires a consideration of the continzlity, the .q&Mwium and t5e limiting
stress conditions.
The similarity is demonstrated
in Fig. 26: With
regard to continuity, in plastic analysis, the situation is just the reverse.
TheoretMly,
plastic hinges interrupt continuity, so the requirement
is that sufficient plastic hinges form to allow the structure (or part of it)
to deform as a mechanism. This could be termed a me&nism condi,tion. The equdibritim condition is the same, namely, the load should
be supported.
Instead of initial yield, the limit of usefulness is the
attainment of plastic hinge moments, notonly at one cross-section but
at each of the critical sections; this will be termed a plastic moment
condition.
As will be discussed further, two useful methods of analysis take
their name from the particular conditions being satisfied:
a) Mechanical
b) Statical

Method

Mechanical

condition

Equilibrium

condition

satisfies

(Equilibrium)Method-satisfies

Plastic moment
condition

In the first method, a mechanism is assumed and the resulting equilibrium

equations are solved for the ultimate load. This value is only correct
if the plastic moment condition is also satisfied. On the other hand, in
the statical or equilibrium method, an equilibrium moment diagram
is drawn such, that M < Mfi. The resulting ultimate load is only
the Correct .value if sufficient plastic hinges were assumed to create a
meahanism.
1

FIG. 26

ELASTIC

ANALYSIS

PLASTIC

CONDITIONS FOR ELASTIC AND

ANALYSIS

PLASTIC

ANALYSIS

50

Please purchase PDF Split-Merge on [Link] to remove this watermark.

SP: 6(6) - 1972

Having considered these three necessary and sufficient conditions,
it will next be of interest to examine certain additional principles and
assumptions upon which the plastic methods rest. Although the plastic
design procedures do not require a direct use of these principles (or
assumptions) they will be stated for background purposes.
a) Virtual Displacements
The principle
If a system
displacement,
work done by

of
of
the
the

virtual displacements is as follows*:

forces in equilibrium is subjected to a virtual
work done by the external forces equals the
internal forces.

This is simply a means of expressing an equilibrium condition. If the

internal work is called WI and the external work is called WE, we may
write :
. ..(25)
WE=WI
. ..
..*
Application of this equation will be demonstrated in 18.
b) Uppel and Lower Bound

Theorems

It is not generally possible to solve all three of the necessary oonditions (mechanism, equilibrium and plastic moment) in one operation.
Although the Equilibrium condition will always be satisfied, a solution
arrived at on the basis of an assumed mechanism will give a load-carrying
capacity that is either correct or too high. On the other hand, one that
is arrived at by drawing a statical moment diagram that does not violate
the plastic moment condition will either be correct or too low. Thus,
depending on how the problem is solved, we will obtain an upper limit
or bound below which the correct answer should certainly lie, or we
will determine a lower limit or bound which is certainly less than
the true load capacity.

The important upper and lower bound theorems or principles were

proved by Greenberg and Prager. When both theorems have been
satisfied in any given problem, then the solution is in fact the correct
one. The two principles will now be stated and illustrated.
Upper Bound Theorem -A
load computed on the basis of an
assumed mechanism will always be greater than or at best equal
to the true ultimate load.
Consider the fixed-ended beam in Fig. 27(A). If we assume a me&anism on the basis of a guess that the plastic hinge in the beam forms at
*Reference 21 contains m excellent discussion of the principle of virtual dia-

placements,

51

Please purchase PDF Split-Merge on [Link] to remove this watermark.

SP: 6(6) - 1972

27A

Upper Bound Theorem

FIG. 27

UPPER AND

278

Lower Bound Theorem

LOWER BOUND THEOREMS

B, then the equilibrium moment diagram would be as shown by the solia

line in Fii. 27(A). The beam would have to be reinforced over the length
BB' to carry the trial load, Wt; the load is too great. Only when the
mechanism is selected such that the plastic moment value is nowhere
exceeded (see the dotted lines) is the correct (lowest) value ohtained.
Lower Round Theorem -A
load computed on the basis of an
assumed equilibrium moment diagram in which the moments are
not greater than M,, is less than or at best equal to the true
ultimate load.
Illustrating with the fixed-ended beam of Fig. 27(B) j if we select the
redundants such that the moment is never greater than Mp, then the
corresponding trial load, Wt. may be less than W,, [Fig. 27(B)]. We
have not used the full load capacity of the beam because the centre line
moment is less than M,. Only when the load is increased to the stage
where a mechanism is formed (dotted) will the correct value be obtained.
Thus, if the problem is approached from the point of view of assuming a mechanism, an upper bound to the correct load will be obtained.
But this could violate the plastic moment condition. On the other hand,
if we approach it from the aspect of making arbitrary assumptions as
to the moment diagram, then the load might not be sufficiently great to
create a mechanism.
Incidentally, Fig. 27(B) demonstrates that conventional (elastic)
design is a lower bound solution. This is the explanation as to why

1.
j

Please purchase PDF Split-Merge on [Link] to remove this watermark.

L.,.

SP: 6(6) - 1972

the local yielding involved in many

has not resulted in unsafe structures.

of our current

It is seen, then, that the statical

is based on the lower bound principle.
other hand, represents an upper limit

the

design

asumptions

(equilibrium)
method of analysis
The mechanism
method, on the
to the true ultimate
load.

c) Further Assumptions - In addition to the assumptions

following further assumptions
are necessary:

of 11

a) The theory considers only first order deformations.

The deformations are assumed to be sufficiently small such that equilibrium
conditions
can be formulated
for the undeformed
structure
(just
as in the case of elastic analysis).
b) Instability
of the structure
will not occur prior to the attainment
of the ultimate
load (this is assured through attention
to secondary design considerations).
c) The connections
provide
full continuity
such that the plastic
(see Section E).
moment,
M*, can be transmitted
d) The influence of normal
and shearing
forces on the plastic
Mfi, are
neglected
(see Section
E for necessary
moment,
modifications).
e) The loading is proportional,
that is, all loads are such that they
increase in fixed proportions
to one another.
However,
independent increase can be allowed, provided no local failure occurs
(see Section E for repeated loading).

With the Principles of Virtual Displacements, the Upper and Lower

Bound Theorems, and the additional
assumptions
noted
possible to consider the various methods of analysis.
17. STATICAL

METHOD

it is now

OF ANALYSIS

17.1 As noted in 16, the statical

method of analysis
Lower Bound Principle.
The procedure is first described
examples are solved.
17.2 Method
of Analysis
by Statical
procedure
16nd an equilibrium
moment
that a mechanism
is formed:
a)
b)
c)
d)

above,

is based on the
and then several

Method -By
the following
diagram in which M6 Mp such

Select redundant(s),
Draw moment diagram for determinate
structure,
Draw moment
diagram
for structure
loaded by redundant(s),
Sketch composite moment diagram in such a way that a mechanism is formed (sketch mechanism),
53

Please purchase PDF Split-Merge on [Link] to remove this watermark.

SP: 6(6) - 1972

e) Compute value of ultimate load by solving equilibrium equation,
f) gtck

to see that M< Mp.

Fixed-ended,

uniformly

loaded beam, Fig.

25

(ifldetermimate to secorcd degree)

The problem (already treated in 15) is to find the ultimate load,
that a beam of moment capacity M, will support. For redundants, one could select the end moments. The resulting moment
diagram for the determinate structure would be the solid parabola
in Fig. 25, with:
W,,

The moment diagram for the structure loaded by the redundants

would be a uniform moment along the beam.
The composite moment diagram is actually what has been
sketched in Fig. 25 since, we can immediately see that a hinge must
also form at point 2. Notice that if the fixing line had been drawn
in any other

W,L

position than that which divides M, = 7

.
ln half,

then no mechanism would have been formed. The correct mechanism is sketched in the lower portion and M = Mp at the locations
of maximum moment.
The equilibrium
W,L
8

Example

from Fig. 25 (at location 2), is:

= M,+M,

and the ultimate

W =y

equation,

load is given by:

.. .

. ..(27)

. ..

2:

Two-span continuous beam, Fig. 28

(indeterminated

The
resultant

to j;rst degree)

redundant is selected as the moment at C(M,).

loadings are shown in Fig. 28(a) and 28(b).

Moment diagrams due to loads and redundants

Fig. 28(c) and 28(d).

The

are shown in

54

Please purchase PDF Split-Merge on [Link] to remove this watermark.