Galambos Ravindra 1976

You might also like

Download as pdf or txt
Download as pdf or txt
You are on page 1of 103

WASHINGTON UNIVERSITY

SCHOOL OF ENGINEERING AND APPLIED SCIENCE


DEPARTMENT OF CIVIL ENGINEERING

LOAD AND RESISTANCE FACTOR DESIGN CRITERIA


FOR STEEL BEAMS

by

T. V. Galambos
and
M. K. Ravindra

Progress Report to the Advisory Committee


of AISI Project MS "Load Factor Design of Buildings"

Research Report No. 27 Structural Division February 1974


Revised February 1976
LOAD AND RESISTANCE FACTOR DESIGN CRITERIA

FOR STEEL BEAMS

by

T. V. Galambos

and

M. K. Ravindra

Progress Report to the Advisory Committee

of AISI Project 163 "Load Factor Design of Buildings"

February 1974

Revised February 1976

Research Report No. 27, Structural Division

Civil and Environmental Engineering Department

Washington University

St. Louis, Mo.


1

ABSTRACT

In this report Load and Resistance Factor Design Criteria are


developed for steel beams subjected to flexute, shear and combined
stress. Resistance factors and expressions for nominal resistance are
developed for solid web beams of wide-flange, channel and box shaped
cross sections. Plate girders, that is beams with transverse stiffeners
on slender webs, and composite beams are, however, not considered in
this report. The resistance factors are developed from the basis of
first-order probabilistic considerations. Nominal resistance expressions
are developed for the limit states of full plastification, lateral-
torsional buckling, local buckling and yielding.
DEPARTMENT OF CIVIL ENGINEERING
WASHINGTON UNIVERSITY
SCHOOL OF ENGINEERING AND APPLIED SCIENCE SAINT LOUIS, MISSOURI 63130

September 8, 1977

To: Messrs. I. M. Viest W. A. Milek, Jr.


W. C. Hansell C. W. Pinkham
L. S. Beedle M. K. Ravindra
C. A. Cornell G. Winter
E. H. Gaylord S. J. Fenves
J. A. Gilligan P. B. Cooper
I. M. Hooper

From: Ted Galambos

Subject: AISI Project 163, LRFD for Steel Structures, corrections to


Washington University Research Reports No. 29 and 45

In the course of preparing a paper on the LRFD of plate girders,


Pete Cooper discovered that I had committed an elementary arithmetical
error. On p. 36 of the Plate Girder Report (# 29) it appears that the
C.O.V. is not 0.20 but 0.11. As you may recall, the value of 0.20 is
responsible for the different values of used in shear ( = 0.78) and
flexure ( = 0.86) in the LRFD plate girder design criteria. The correct
value of the COV = 0.11 results now in the same for both flexure and
shear, and the enclosed errata sheets contain the repairs which must be
made to correct the error in the two reports which are affected.

I am very grateful to Pete Cooper for discovering this error. As


a result life has become just a grade more simple. At the same time I
am mortified to have let such a mistake happen in the first place.

TVG:ajb

Enclosure
September 7, 1977

ERRATA SHEET:

LRFD Criteria for Plate Girders: W.U. Research Report 29, Revised Feb. 1976

p. 17, 9th line from bottom, delete VP = 0.20 and replace by VP = 0.11

p. 18, second line from top:

replace by

p. 18, 6th line from top, replace whole line by

p. 18, 7th line from top, replace first sentence by

"The value of = 0.86 will be used, same as for flexure."

p. 18, 11th line from top, replace 0.78 by 0.86

p. 21, first line from top, replace 0.78 by 0.86

p. 21, first full sentence on page, delete the whole sentence

p. 21 through 23, delete all material starting with "This envelope...."

in the 8th line on p. 21 to Sec. 5 on p. 23, and insert intend the following:

"The equation of the boundary of the in the interaction range is

The interaction check can thus be expressed by

where = 0.86 and and "


2.

p. 26, 10th line from top, replace 0.78 by 0.86

p. 26, 13th line from top, delete "When 0.6 M u / V u " and capitalize

"An..."

p. 26, 15th line from top, replace 1.04 by 0.625 and replace 1.40 by 1.375

p. 26, 16th line from top, replace 0.78 by 0.86

p. 36, bottom line, replace 0.20 by 0.11

p. 42, replace 0.78 by 0.86 and adjust lines DB'A' accordingly.


September 7, 1977

ERRATA SHEET

Proposed LRFD Criteria, W.U. Research Report 45, May 1976

p. 2-6, last line, replace 0.78 by 0.86

p. 2-7, second and 5th line from top, replace 0.78 by 0.86

p. 2-7, last line, replace 260/(h/t) by

p. 2-8, 6th line from top, add at end of line "and = 0.86."

p. 2-8, Eq. 2.3.3.2-10, replace by

p. 2-10, 2nd to 4th line, delete part of sentence starting with "and

the ratio " in line 2 through "... V u /M u " in line 4.

p. 2-10, Eq. 2.3.3.2-15, replace 1.04 by 0.625 and 1.40 by 1.375

p. 2-10, 8th line from top, replace 0.78 by 0.86.


ii

TABLE OF CONTENTS PAGE NO.

1. Introduction 1

2. Maximum Strength of Beams 2

3. Bending Strength as Determined by Lateral-Torsional Buckling 5

3.1 Elastic Lateral-Torsional Buckling of Doubly Symmetric

Wide-Flange Beams 5

3.2 Experimental Verification of Elastic Buckling 8

3.3 Design Equations in the Elastic Range 9

3.4 Design Equations in the Plastic Range 14

3.5 Design Equations in the Inelastic Range 19

3.6 Selection of Resistance Factor 26

3.7 Design Equations for W-Beams 27

4. Bending Strength as Determined by Local Buckling 30

4.1 Flange Local Buckling in the Elastic Range 31

4.2 Flange Local Buckling in the Plastic Range 33

4.3 Flange Local Buckling in the Inelastic Range 35

4.4 Design Equations for Flange Local Buckling 35

4.5 Flange Local Buckling of Box Beams 37

4.6 Design Equations for Web Buckling 40

5. Hybrid Beams 43

6. Design for Shear Strength 46

7. Combined Stresses 49

8. Summary 49

9. Acknowledgements 50

10. References 51

11. Notation 57

Figures 60

Appendix: Design Rules 67


1.
1. INTRODUCTION
*
In Load and Resistance Factor Design (LRFD) the formula

(1)

represents the criterion which is to be satisfied for an acceptable design.


The right side of the criterion refers to the loading to be considered, and
are "load factors" by which the nominal load effects Qnk are multiplied
to account for the uncertainties of the loads and the structural analysis.
The left side of the design criterion refers to the resistance of the

structural member. Rn is the nominal resistance, and is the "strength"


or "resistance" factor* which accounts for the uncertainties inherent in the

determination of Rn.
The present report is concerned with the left side, i.e., with the term
Rn, only. Expressions for Rn and values for will be derived for solid
web beams whose flanges are not slender enough to require stiffeners when
the beam is subjected to pure shear and where webs will not buckle in flexure
prior to the full yielding of the flanges ("beams" in contrast to "plate
girders"). The report will deal firstly with beams subjected to bending
moments, and then with beams under shear force and torsion. Finally rules
will be discussed for combined loadings involving biaxial bending and torsion.
Special attention will be paid to the treatment of lateral-torsional buckling
and local buckling. The final derived rules are presented in the Appendix in
the form of a tentative suggested specification segment.

* The basis of LRFD as applied to steel structures is discussed, and illustra-


tions of its implementation are given, in Refs. 1 and 2, where further
references and discussion on the general subject of probabilistic design can
also be found.
2.

In order to proceed with the detailed derivations for the development


of Rn, a few concepts from Ref. 2 will first be reviewed. In Ref. 2 it
was demonstrated that

(2)

where Rn is the nominal resistance as computed by an equation given in a


design specification, Rm is the mean resistance, is a numerical constant
used in the development of the LRFD criteria so that a separation into a load
effect term and a resistance term is possible (see Ref. 2 for further details),

is the "safety index" obtained by calibration and VR is the coefficient of

variation of the resistance, given by

(3)

where V M , VF and VP are, respectively, the coefficients of variation charac-

terizing the uncertainties in the material strength, in the fabrication and


in the strength analysis, respectively.

The numerical value of was given as 0.55 in Ref. 2, and the safety
index was found by calibration against plastically designed braced beams
to be 3.0. In the following = 0.55 and = 3.0 will be used for all beams.
It should be realized that = 3.0 refers to the maximum strength as
2
the limit of structural usefulness , and thus the following portions of this

report deal only with strength criteria (strength limit states) and not with
serviceability criteria.
2. MAXIMUM STRENGTH OF BEAMS

The following discussion refers to singly or doubly symmetric beams


subjected to bending moment about the major axis (i.e., "uniaxial" as
3.

contrasted to "biaxial" bending). Such beams ultimately fail by lateral-


torsional buckling or by the local buckling of the compression flange,
provided the steel possesses the necessary ductility. Since the grades of
steel in the AISC Specification all possess sufficient ductility, failure
by tensile rupture will not be assumed to occur prior to a buckling type
failure associated with compression. Web buckling will be also considered.

Lateral-torsional buckling strength increases as the spacing of lateral


bracing becomes smaller, and local buckling strength increases as the width-
thickness ratio of the plate element decreases. The relationship between
bending capacity, represented by the moment M, and "slenderness", which is
related to either bracing spacing or plate width-thickness, and which is
denoted by the symbol , is shown schematically in Fig. 1. The solid con-
tinuous curve represents the actual ultimate strength, as might be obtained
from an analysis which takes "everything" into account. In later portions
of this report the symbol will be replaced by Lb for unbraced length and
b f / t f for local buckling, as appropriate.

Beam behavior, as characterized by a type of representation given in


Fig. 1, can be divided into three regions:
I. Plastic range, where the section is fully yielded or
strain-hardened;
II. Inelastic range, where buckling occurs after some portions
of the cross section have yielded but some portions are
still elastic;
III. Elastic range, where buckling takes place while the cross
section is still elastic.
4.

Practical beams fall usually in ranges I and II, while range III often
only becomes of importance during erection before all the bracing is in
place. In the plastic range not only the strength is counted on in design,
but also inelastic rotation capacity, and so plastic analysis may be used to
determine the bending moment distribution as a plastic mechanism forms.
In ranges II and III no appreciable rotation capacity exists, and so the
forces in the member must be obtained by an elastic analysis. The order of
importance, in terms of frequency of occurrence and optimum utilization of
the material is: ranges I, II and III.

While it is possible to determine analytically or numerically to


a satisfactorily accuracy the whole solid curve in Fig. 1, this is
not practical in a design specification. Usually strain-hardening is not
directly counted on, and so the maximum utilized moment in range I is MP,
the plastic moment. Various types of curves have been used to represent
the transition between range I and range II (e.g., parabolas), but a straight
line will be used here. In the elastic range the analytically exact equation
will be used. The transition between ranges II and III occurs at M = Mr,
where Mr is the yield moment which included the effect of residual stresses.

The strength relations between M and (where can be either unbraced


length or plate width-thickness ratio), representing the nominal resistance
Rn in Eq. 2, where Rn = Mu, the nominal maximum moment, can be represented
as follows:

(4)

(5)
5.

(6)

In Eq. 6, Me is the moment corresponding to elastic buckling.

In the following parts of this report the expressions for Mu and the

corresponding values of will be presented.

3. BENDING STRENGTH AS DETERMINED BY LATERAL-TORSIONAL BUCKLING

The relevant slenderness parameter in Fig. 1 for the case of lateral-

torsional buckling is the unbraced length Lb. In this section the term Mu

will be developed for the case of lateral-torsional buckling.

3.1 Elastic Lateral-Torsional Buckling of Doubly Symmetric Wide-Flange Beams.

The theoretical elastic lateral-torsional buckling moment of a laterally

and torsionally simply supported doubly symmetric beam under equal end moments

(uniform bending) is exactly represented by the following equation3:

(7)

where Me = elastic critical moment

Lb = length of beam

E = elastic modulus in tension or compression

G = elastic shear modulus

Iy = minor axis moment of inertia

J = St. Venant torsion constant

Cw = warping moment of inertia

For other loading and restraint conditions it is seldom possible to

achieve a closed form solution, and approximate or numerical schemes need

to be applied. Many such solutions are available, and all solutions for
6.

doubly, symmetric, uniaxially bent beams can be approximated by the formula3

(8)

The coefficient Cb accounts approximately for unequal end moments, and

a lower bound for Cb of the following form was suggested by Salvadori4:

(9)

In Eq. 9, M1 is the smaller and M2 the larger of the bending moments

at the ends of the unbraced beam segment, and M1/M2 is positive when the

moments bend the beam in reverse curvature. The Column Research Council

Guide5 lists values of Cb for other loading cases. The CRC Guide also

provides formulas to account for the application of loads above and below

the centroid of the section.

The coefficients Ky and Kz are "effective length" factors which account

for lateral and torsional end restraint. A table of these factors is given

in Ref. 3, (Table 3.2) for a variety of end conditions. For simply supported

ends, Ky = Kz = 1.0; when both ends are fixed, Ky = Kz = 0.492, etc. The

values, of Ky and Kz are not equal to each other if the boundary conditions

are mixed, e.g., for laterally simple ends with fully restrained warping

Ky = 0.883 and Kz = 0.492.


In applying Eq. 8 the most difficult aspect is the choice of the

effective length factors Ky and Kz. The increase of Me from simply


7.

supported to fully fixed ends may be as much as three-fold, thus any


uncertainty on the end conditions can have a very large effect. Usually
there is lateral restraint at the ends of a beam and at several Intermediate
locations. Furthermore, many beams are continuous in the plane of loading.
The lateral-torsional buckling load is a function of end-restraint in the
plane of bending, and of the end-restraint offered against lateral deflec-
tion and warping by the adjacent segments. If the adjacent segments are
unloaded, then the in-plane restraint can be determined using first order
elastic theory. However, if they are also loaded, as they usually are,
then the restraint is reduced, finally reaching zero when the adjacent
segments, too, are critically loaded. Thus there is an interaction which
should be considered 6,7 . Several problems involving interaction have been
studied. Trahair6,7 and Hartmann 8 , 9 , among several other investigators,
studied two and three span beams and single-story single-bay frames. Both
of these studies verified the theoretical predictions by tests on model beams.

One way out of this dilemma was shown by Salvadori10 who suggested
that each unbraced segment be assumed to be at the critical level of load-
ing and that, therefore, no interaction exists between adjacent segments
and thus Ky = Kz = 1 should be used. Since not all segments will be loaded

and/or braced in such a way that they will be critical it would seem that
this procedure gives a lower bound to the critical load. Both Trahair and
Hartmann, however, have shown cases where this is not so, although in most
cases this approach is conservative. For single-story single-bay frames
Hartmann9 has suggested that Ky = 1 and Kz = 0.5 will give satisfactory

results.
8.

In view of this complicated state of affairs it appears that at present

Salvadori's suggestion is a reasonable one for a specification, that is,

consider each segment as laterally and torsionally simply supported and

thus Ky = Kz = 1. Furthermore, Cb shall be as defined in Eq. 9 except that

Cb = 1.0 when no lateral bracing exists between the end supports of a beam.

This assumption will usually lead to conservative results, the ratio of the

computed critical moment by this method to the true elastic lateral-torsional

buckling moment being a variable of undetermined magnitude (i.e., from some-

what less than 1.0 to possibly as much as 3.0).

3.2 Experimental Verification of Elastic Buckling

The reliability of the predicted elastic critical load can be gaged by

examining lateral-torsional buckling tests and comparing the buckling loads

to the theoretical predictions. Many tests of the lateral-torsional buckling

of beams have been performed. Some of these have been evaluated by Lee and
11
Ueda but no attempt was made by them to compare test results with theoreti-

cal predictions. Major test programs on many specimens were performed by

Flint and his co-workers in England, but the examination of the original

papers showed that it would be impossible to pick off the test results from
7
very small graphs, and so no comparisons were made for these tests. Trahair
12
and Hartmann presented in tabular form both the test buckling loads and

the predicted theoretical buckling loads and so this data can be used directly.
13
Similar data are available for the elastic tests reported by Hechtman . The
14
report by Clark and Jombock only gave test results, but the theoretical

buckling loads were easily computed from the data given.

The tests by Trahair, Hartmann, Hechtman and Clark were used to examine

the elastic lateral-torsional buckling theory. Many of the tests were made
9.

on aluminum alloy beams; however, the material strength does not affect
the elastic buckling capacity, so these tests are also valid for checking
out the elastic theory. A variety of loading cases have been tested, and
shapes varying from about 1/2 in. to 18 in. in depth were used.
The histogram of the ratios of the test load to the predicted load is
shown in Fig. 2. The mean is 1.03 and the standard deviation is 0.10.
Thus the best estimate by elastic buckling theory, accounting for all
factors the theory is capable of handling, is an excellent predictor of
what happens in the laboratory.
3.3 Design Equations in the Elastic Range

The relationship from Eq. 2 is the basis for developing design criteria
and from this equation the resistance factor is equal to

(10)

where Mm is the mean elastic critical moment of a beam in place, i.e.,

(11)

The first parenthesis translates the beam from the ideal beam to a beam in the
laboratory, and the second bracket is the elastic critical moment computed by
the "best" theory for mean material and geometric properties. The Test
Capacity-to-Prediction ratio was shown to be 1.03 (Fig. 2 ) . With regard to
the elastic critical moment it will be assumed that Handbook values of the
2
cross-sectional properties are mean values , and so Iym = Iy, Jm = J, Cwm = Cw.
10.

Furthermore, it will be assumed that (G/E)m = 0.385, a constant for all


metals, and that Em = E = 29,000 Ksi. The coefficient of variation of the

Test Capacity-to-Prediction ratio (the "professional" factor P in Ref. 2)


was found to be VP = 0.09 (Fig. 2). The coefficient of variation of E was
determined to be VM = 0.06 (from Ref. 15, the subscript M representing the
2
"material" factor ). The coefficient of variation of the cross-sectional
properties is lumped together into one coefficient, VF = 0.05 (where the

subscript F denotes the "fabrication" factor 2 ), signifying the relatively


good control on dimensions characteristic of steel construction. With
these values then, from Eq. 3,

and, from Eq. 11,

(12)

The nominal moment Rn = Mu is the one which will be used in design.


It will be assumed here that design convenience requires that the exact
effective length factors will not be determined, and that Ky = Kz = 1.0
will be used. Thus

(13)
11.

Substitution of = 0.55, = 3.0, VR = 0.12 and Eqs. 12 and 13 into

Eq. 10 results in the following expression for :

(14)

In case of simply supported beams under uniform moment Cbm = Cb = 1.0,


Ky = Kz = 1.0, and so = 0.84. If, however, other loading and restraint
conditions exist, all three ratios in Eq. 14 will be greater than unity.
On the other hand, there will be a greater dispersion and

(15)

where VPA is the coefficient of variation of the ratio of the actual pre-
dicted critical moment to Mn from Eq. 13. Following are some estimated

numbers to gage the effect on : Cbm/Cb = 1.05, Ky = 0.9, and 1.05 for
the ratio of the square roots in Eq. 14. These numbers are assumed to be
the mean values of all the possible beams, and a relatively large value of

VPA = 0.2 indicates large spread. With these values VR = 0.23 and = 0.86
Thus = 0.84 represents a lower value, and a larger number might be justified.
The selection of a recommended value of will be postponed until later in
this report. The elastic nominal critical moment is given by Eq. 13 for
doubly symmetric beams. For wide-flange shapes, alternately, approximations
to this equation can also be used. Two will be suggested as perhaps offer-
ing a measure of simplification over Eq. 13. However, Eq. 13 is not partic-
ularly difficult since values of J and Cw are tabulated in the Seventh
12.

Edition of the AISC Manual. Both suggested equations simply involve some
numerical approximations of geometric and material ratios, and their
derivations from Eq. 13 are given in Ref. 3. The first approximate rela-
tionship is

(16)

where Sx is the elastic section modulus about the x - axis, ry is the minor

axis radius of gyration and

(17)

In Eq. 17, A is the cross-sectional area of the wide-flange shape and d is


its depth. From a study of the values of Mu from Eq. 16 for all W and S
shapes in the AISC Manual, it was determined that the mean ratio of Mu from
Eq. 16 (the approximation) to Mu from Eq. 13 (the formula which is replaced
by the approximation) was 0.97 and the coefficient of variation was 0.07.
The approximation modifies also the - factor, which now becomes equal to

In these computations the additional coefficient of variation, reflecting


the uncertainties of the approximation is included, i.e.,

, where VA = 0.07 is given above. The other


13.

alternative approximation is the one which is the basis for the 1969 AISC

Specification

(18)

where

(19)

(20)

In Eqs. 19 and 20 Af is the area of the compression flange and rT is the

radius of gyration of the compression flange and one sixth of the web about

the minor axis of the section. The mean ratio of Mu from Eq. 18 to Mu from

Eq. 13 was found to be 0.97, with a coefficient of variation of 0.10. By

a similar reasoning as above, it was determined that

Thus the following choices are available for selection of a design

criterion:

Mu from Eq. 13, = 0.84

Mu from Eq. 16, = 0.84

Mu from Eq. 18, = 0.82

All of these values are likely to be higher because of the conservative

approximations involving Cb and assuming the effective lengths to be unity.

This table does show, however, how the various equations chosen for the

nominal moment affect in a relative sense. The final choice of the specific
14.
equation to be selected will be made later in this report when the analyses
for the plastic and the inelastic ranges have been completed. For other
shapes the nominal elastic critical moment can be obtained from the litera-
ture. Formulas for some common shapes are discussed below.
Channels Bent in a Plane Parallel to the Web and Load Passing Through
the Shear Center

The critical elastic buckling load is the same as for wide-flange


3
shapes , and so Eq. 13 is to be used.
Rectangular Solid and Box Shaped Beams Bent About the Major Axis
The equation of elastic buckling is 5 :

(21)

Singly Symmetric Beams Bent in the Plane of Symmetry

Formulas for such beams are listed in the literature (e.g., CRC Guide 5 ),
and the elastic critical moment from these formulas should be used for the
nominal resistance Mu.

3.4 Design Equations in the Plastic Range

The plastic range (Region I in Fig. 1) represents the optimum capacity


of the beam. In this range the plastic moment Mp = Fy Z can be reached or
exceeded, and this moment-level can be maintained for a large enough rota-
tion so that inelastic force redistribution can take place and finally a
mechanism can form. While in the elastic range of lateral-torsional buckling
the situation is clear, i.e., the member is buckled or it is stable, the
factors affecting the behavior in the plastic range are complex and intri-
cately interrelated. Local and lateral-torsional distortions interact and
they tend to build up gradually rather than form suddenly. Strain-hardening
on the one side and instability on the other side work against each other and
they tend to balance out to give M = Mp at the critical length Lbp ( in Fig.
1). While many experiments have been performed and several theoretical models
15.

for relating unbraced length, flange and web width-thickness and rotation

capacity have been proposed, (Refs. 16 through 20), no generally satisfac-

tory analysis which fully recognizes these complex interrelationships has

as yet been presented. In design the requirement of comparing available

and required rotation capacities at plastic hinge locations is circumvented

by demanding conservative limits on spacing of beam bracing and on cross

section geometry.

The nominal resistance of a beam in the plastic range is the plastic

moment Mp, determined for the nominal dimensions of the cross section and

the specified yield stress of its components. For a shape made of a uniform

grade of steel,

(22)

where Zx is the plastic modulus and Fy is the specified yield stress of the
2
grade of steel. The mean resistance was shown to be equal to :

(23)

where (Fsy)m is the mean static yield stress of the flanges, and the mean

Test/Prediction ratio is obtained from tests on beams for which "simple

plastic theory" was the basis for analysis. The mean section modulus Zm

is assumed to be the Handbook value, i.e., Zm = Zx, and a coefficient of


variation VF = 0.05 will again be assumed *. In Ref. 2 it was shown that

for the yield stress of the flange material in rolled W-shapes (Fys)m = 1.05,

with VM = 0.10 is a reasonable estimate.

*In one recent test program involving 12 in., 16 in. and 24 in. deep beams 2 1 ,
the mean ratio of measured-to-Handbook values of Zx was 1.00, and the
coefficient of variation was 0.03.
16.

Several series of tests were considered for the determination of the

Test Capacity-to-Prediction ratio, and the results are listed in the table

below. The values are subsequently determined by the formula

(24)

where (Fys)m = 1.05 Fy, = 0.55, = 3.0 and VR is computed by Eq. 3 with

VM = 0.10 and VF = 0.05.

Type Member Source Number VP


of Tests

Stat. det. beams under


uniform moment 23 - 31 * 33 1.02 0.06
Stat. det. beams under
moment gradient 22 - 26 43 1.24 0.10
Stat. indet. beams
and simple frames 32 41 1.06 0.07

If it is assumed that the results from each type test are representa-

tive, that i s , the Test-to-Prediction ratios and the coefficients of variation

V P would not change when new test results are included, then an overall

factor for beams can be determined by also assuming that the three sets of

data are independent and that the total population of beams has a specified

distribution among simply supported beams under uniform moment, simply sup-

ported beams under moment gradient and continuous beams. The real distribu-

tion i s , of course unknown, but for the sake of arriving at a - factor the

reasonable assumption will be made that 5 0 % of all beams are continuous, and

that 2 5 % of the determinate beams are each under moment gradient and 2 5 % are
under uniform moment.
*Refers to References at end of this report.
17.

The mean Test-to-Prediction ratio is then

53
The coefficient of variation is equal t o :

The coefficient of variation of the resistance, VR, and the - factor

are:

This value of = 0.89 could be used in the plastic range. In the pre-

vious section of this report it was shown that in the elastic range = 0.84

but that a somewhat higher value could be justified. In Ref. 2 a value of

= 0.86 was found for short columns and simply supported beams. In order to

keep consistency as much as possible, a value of will not be recommended

until the inelastic range (Range II. in Fig. 1) has been examined.
The critical unbraced length ( = Lbp in Fig. 1) is the maximum

bracing distance which will permit the attainment of Mp and still maintain

sufficient rotation capacity. This critical unbraced length is given by


20
the following formulas :

(25)

for uniform moment and

(26)
18.

for beams under moment gradient. In these equations E is the elastic


modulus, Est is the strain-hardening modulus and ry is the minor axis

radius of gyration. Fy, E and Est are random variables for which the mean
values and the coefficients of variation are known. Thus these formulas
could be subjected to the same kind of an analysis as the strength equations.

This was not done because these equations have already been manipulated so
much in their derivation, so that it would not be sound to operate on them.

For the present, Lbp will be considered as a non-random quantity. Further-

more it will be assumed that E and Est are not functions of the grade of
steel (this is actually a good assumption for E and also for Est - - - see

Ref. 19 where this latter point is demonstrated statistically). Equations


25 and 26 can be written in the following form:

(27)

where K is a coefficient dependent on E and Est.


23,27 16,17
Based on test results and on recommendations by Lay
Lbp/ry = 40 gives satisfactory rotation capacity for beams under uniform
moment and Lbp/ry = 65 is recommended for beams under moment gradient for
A36 steel. Using these values as base values,

(28)

for beams under uniform moment and

(29)

for beams under moment gradient.


As in the 1969 AISC Specification, uniform moment will be defined to
be in the range -0.5 > M/Mp -1.0 and moment gradient to be in the range
19.

+ 1 M/Mp -0.5, where M is the lesser of the moments at the ends of the
unbraced segment and M/Mp is positive when the segment is bent in reverse
curvature and negative when it is bent in single curvature.
The definition of Lbp as in Eqs. 28 and 29 is open to the criticism
that it involves a discontinuity when M = -0.5 Mp and that it is very con-
servative for beams under reverse bending. In order to alleviate these
shortcomings, two suggestions have recently been advanced for consideration
by the AISC Specification Advisory Committee: The first one, advanced by
Yura and based on the work reported in Ref. 21, uses only one formula:

(30)

When M/Mp = -1 (uniform bending), Eq. 30 gives more conservative values


than Eq. 28, but it gives very liberal values when M/Mp = +1. The second
formula, as suggested by Lu, reverts to a previously used (1963 AISC
Specification) continuous form:

(31)

Whether Eqs. 28 and 29, or Eqs. 30 or 31 should be used in LRFD will


not be decided at this point. Each approach has merits. For the purposes
of this report the use of Eqs. 28 and 30 will be retained. The other
formulas are given, however, to point the way to alternative approaches in
the future.
3.5 Design Equations in the Inelastic Range
According to Fig. 1 the nominal moment capacity in the inelastic range
will be specified as a straight line between the end of the plastic range
20.

(Mu = Mp and = Lb = Lbp) and the end of the elastic range (Mn = Mr and
= Lb = L b r ) . The "yield moment" Mr is the value of M where yielding just
commences due to applied elastic flexural stresses and compressive residual
stresses in the compression flange, that is,
(32)

where Sx is the elastic section modulus and Fr is the maximum compressive


*
residual stress in the compression flange .
Residual stresses due to uneven cooling after rolling and due to cool-
ing after welding have been measured extensively, and Refs. 33 through 47
contain the data from which the following results were estimated:
1) The mean maximum compressive residual stress in the flanges of
rolled wide-flange shapes was 9.5 ksi, with a coefficient of variation of
0.48 (this information was obtained from 38 tests)
2) The mean maximum compressive residual stress in the flanges of
welded H-shapes and in welded plates was 20 ksi, with V = 0.40 for 128 tests.

The scatter of the residual stresses is quite high and there appears to
be a substantial difference in the mean residual stress magnitude between
rolled and welded H-shapes. However, most of the residual stress data for
welded H-shapes were taken from sections primarily intended for columns and
not for members subjected mainly to flexure. There is a definite need to
determine the residual stresses in welded beam shapes, and until good data
becomes available which proves the contrary, it will be assumed that for
further use in this report welded and rolled beam shapes have the same resid-
ual stress data, i.e., (F r ) m = 10 Ksi and = 0.5, and that the residual
stress is independent of the grade of steel.

In the inelastic range part of the section will be yielded when buckling
occurs. The buckling load is not unique because it can be defined as the

*For W shapes F at the flange tips is more critical with respect to inelastic
r 48
buckling than the tensile residual stresses at the flange-to-web junction .
21.

tangent modulus load (T.M. in Fig. 3) or the reduced modulus load (R.M. in

Fig. 3) for an ideally perfect beam system. The former is a lower bound and

the latter is an upper bound to the true maximum strength (Max. in Fig. 3)

which is somewhere between the two. Imperfections of loading and geometry

tend to reduce the capacity. The true maximum capacity depends on the magni-

tude of the imperfections, on the residual stresses and on the yield stress

of the material. In contrast, elastic buckling load is an unambiguous criti-

cal state. Ideally it corresponds to the situation at which lateral deflection

and twist can occur for the first time. This, of course, is an idealization

which occurs rarely because of the unavoidable eccentricities of loading and

because of the presence of unavoidable imperfections. Despite these imper-

fections the load-deformation curve approaches the ideal curve asymptotically

and thus the buckling load has a definite meaning (Fig. 3 ) .

The determination of the tangent modulus, the maximum strength, and the

reduced modulus loads for inelastic beams is considerably more complicated

than the calculation of the elastic critical loads and recourse must be had

in numerical methods48,49,50,52. This is especially true when residual

stresses are present. Some calculations of the reduced modulus loads and

maximum moments (with and without initial geometric imperfections) have been

made to show that the problem can be solved, but major parametric studies of

an extent to reveal trends have only been made on the determination of the

critical tangent modulus moment (Refs. 48 through 52 and 58 through 62). These

studies have shown that the solution is sensitive to the magnitude and the

distribution of the residual stresses and to geometric properties of the

wide-flange shapes. For rolled shapes the straight-line approximation of

the moment appears to be very reasonable 59 , with sections having high values

of Dt (Eq. 17) lying above and those with low values falling somewhat below

the straight line. One set of studies, including theoretical work and

tests34,51,60, appears to indicate that the straight line approximation is


22.

unconservative for welded H-shapes.

Unfortunately not many tests exist for steel wide-flange beams failing

in the inelastic range. It is necessary that these tests be full sized and

of steel because initial imperfections and residual


13 stresses are difficult

to scale down. Hechtman and his associates tested 12 beams which failed

by inelastic buckling. In Ref. 3 a value of 0.3 Fy is used for the maximum


compressive residual stress for the interpretation of these tests and most

of the test points fall below the straight-line approximation. Furthermore,

the test results are quite scattered. It should also be realized that com-
parison was made to an ideal situation in which end moments are applied to
the wide-flange beam segment3. The actual loading was, however, a two-point

loading at the top of the compression flange and thus a more severe situation

existed than in the ideal case to which the tests were compared.

Tests for which more direct comparison can be made were performed by

Dibley at the British Industrial Steel Research Association (BISRA) labora-


33 24
tories in England and by Fukumoto in Nagoya, Japan . Dibley tested as-

rolled wide-flange beams under symmetric two-point loading (Fig. 4a) and

Fukumoto tested as-welded and annealed wide-flange shapes made from three

steel plates under equal end moments (Fig. 4b). Statically indeterminate

beams (Fig. 4c) were tested by Bansal on rolled wide-flange shapes.


60
Fukumoto's tests were analyzed by Nethercot . The tests by Dibley and

Bansal are shown in Fig. 5 where the ratio of the maximum test capacity to

the critical moment predicted from the straight line approximation is

plotted against the unbraced length Lb. This value is non-dimensionalized

by the limiting elastic unbraced length Lbr. The equation for the straight-

line approximation for the prediction is, from Eq. 5,

(33)
23.

where Mp = Fy Zx, Mr is defined by Eq. 32, Lbp is the limiting unbraced


length of the plastic range (Eqs. 28 and 29) and Lbr is determined from
setting M = Mr and Lb = Lbr in Eq. 8, i.e.,

(34)

For Dibley's tests Cb = 1.0 and the effective length factor was computed
and tabulated by the author. For Bansal's tests Cb = 2.3 and Ky = 0.78.
In the latter tests the moment distribution was determined by elastic
theory. In both test series Fr = 9.5 ksi, G/E = 0.385, E = 29,000 ksi and
Handbook geometric properties were used in the calculations. The reported
measured values of Fy were utilized.
The information in Fig. 5 shows that the tests on statically deter-
minate beams (Dibley's tests) are reasonably well predicted by the straight
line theory (mean 1.06, V = 0.09), but that the tests on indeterminate
beams were considerably stronger (mean 1.22, V = 0.12) and more dispersed.
The reason for this is that partial or almost complete moment redistribution
was able to occur. Thus the limiting unbraced lengths Lbp (Eqs. 28, 29)
appear to be conservative for indeterminate beams.

The following derivation will consider the development of an estimate


of the resistance factor for the inelastic range. This derivation will
consider statically determinate simply supported rolled steel wide-flange
beams as the basis for obtaining . For these beams it will be assumed
that the straight line formula represents the prediction of the capacity.

(35)
24.

It will be assumed that Dibley's tests are representative of the types of


results which can be obtained, and thus (from Fig. 5)

In the following analysis it will be assumed that Zx and Sx, the plastic
and the elastic section moduli, respectively, are mean values as listed in
the handbooks, and that their coefficient of variation is VF = 0.05 (as

assumed previously). Furthermore Lbp and Lbr will be assumed to be deter-


ministic quantities, representing the limits of applicability of Eq. 35.
The standard deviation of M, as reflected in the variability of Fy and
Fr, can be determined by the equation53:

(36)

where (from Eq. 35)

(37)

(38)

and (39)

(40)

The coefficient of variation of material properties is then

(41)
25.

The coefficient of variation of the resistance is equal to

(42)

where VP = 0.09 and VF = 0.05 and VM is determined from Eq. 41.

The resistance factor can now be obtained from Eq. 10 with = 0.55

and = 3.0 and the nominal resistance Mn:

(43)

In Eq. 43 Fy is the specified yield stress, and a nominal (rounded off)


value of Fr = 10 Ksi will be used; Lbp is determined from Eq. 28 and Lbr
is obtained from Eq. 34 with Ky = Cb = 1.0 and Mr = Sx (Fy - F r ) .
The resulting resistance factor is dependent on Lb and Fy, the cross-
sectional variations between wide-flange shapes having only a negligible
effect. The following table shows for a representative section, the
W24 x 76 shape:

Fy = 36 Ksi Fy = 50 Ksi

Fr = 10 Ksi Fr = 10 Ksi

Lb /ry Lb /ry

40 = L bp 0.88 34 = L bp 0.88

80 0.87 50 0.88

120 0.83 70 0.87

147 = L br 0.78 90 0.85

113 = L br 0.83
26.

It is evident that decreases with increased length because of the

greater effect of the residual stresses, but when Fy = 100 Ksi this effect

is almost completely wiped out so that = 0.88 for any length.

The use of these factors, and the selection of for the lateral

buckling capacity of beams, will be discussed next.

3.6 Selection of Resistance Factor

The curves in Fig. 6 show the nominal bending capacity Mu and the

factored bending capacity Mu for the W24 x 76 section, using the values

of = 0.89 in the plastic range, from the tables at the end of Sec. 3.6

for the inelastic range and = 0.84 in the elastic range. The information

in Fig. 6 reveals the fact that the ends of the factored curves do not match

at the boundaries of the regions. This is not surprising, since different

data, different assumptions and different degrees of approximation were used

in the background development.

A design rule, however, should not have discontinuities, and it would

be desirable to have a uniform resistance factor. The following is sugges-

ted:

In the plastic range a value of = 0.86 has been recommended in Ref. 2.

This is less than = 0.89 obtained in Sec. 3.4 and less than = 0.88 from

Sec. 3.5, but it corresponds to recommended previously for simple beams

and short columns2. This same value of is recommended for the whole length
range of steel beams. In the elastic range this compares with = 0.84 from

Sec. 3.3, but there it was shown that higher values could be justified. In

the inelastic range, when Lb = Lbr (see Fig. 6 ) , = 0.78 was determined

previously, and = 0.86 might appear to be too unconservative. However,

ends of unbraced segments are seldom without restraint, and while the

residual stress effect is highest at that point, so is the restraint effect.


27.

3.7 Design Equations for W-Beams

In the previous portions of this section nominal resisting bending

moment and factors were developed for W-Beams failing by lateral-torsional

buckling. The nominal maximum bending moments are summed up below:

(44)

(45)

(46)

where for doubly symmetric wide-flange beams

(47)

In these equations the section properties Zx (plastic section modulus),

Sx (elastic section modulus), J (torsion constant), I (moment of inertia

about y-axis) and Cw (warping constant) are Handbook values, and

(48)

(49)

(50)
28.

(51)

where
(52)

M1/M2 is the ratio of smaller to the larger of the end moments of the

unbraced length, positive when causing reverse curvature.

is the ratio the end moment at the end of the unbraced length away

from the plastic hinge to the plastic moment Mp, positive when causing

reverse curvature.

Fy is the specified yield stress for the grade of steel used,

Fr = 10 Ksi

E = 29,000 Ksi and G/E = 0.385.

The factors accompanying these nominal moments in the design equation,

Eq. 1, are equal to = 0.86.

The nominal equations given above (Eqs. 44, 45, 46) apply for all

shapes, but the subsequent formulas for Me, Lbp and Lbr apply strictly only

to doubly symmetric W-shapes. Following are the corresponding formulas for

other type beams. Since there is little experimental evidence available

for most of these beams, it is recommended that the same proposed above

be adopted throughout, that is, = 0.86.


29.

Rectangular Beams Bent About the Major Axis

For these beams the elastic critical moment is5

(53)

and in the plastic range

(54)

where Est and Gst are the tensile and shear moduli in the strain-hardening
range. If Est = 630 ksi and Gst = 2260 ksi (see the following Sec. 4 for
the basis of these values) are used, Lbp can be computed. Similarly, Lbr
can be determined from Eq. 53 by setting Me = Mr = S x F y . This latter

substitution ignores residual stresses in the rectangular bars. The follow-


ing equations can now be derived:

(55)

(56)

(57)

(58)

Box Beams Bent About the Major Axis


For box beams Eqs. 53 and 54 apply also, but residual stress must be
accounted for. The relevant equations thus are as follows:

(59)
30.

(60)

(61)

(62)

Singly Symmetric Wide-Flange Beams


The formula for Me is given in the Column Research Council Guide and
other references (e.g., Ref. 3 ) .

(63)

Lbr can be computed by substituting Mr = Me and solving for Lb, and

(64)

where ry is the radius of gyration of the compression flange and one half
of the web. The basis for Eq. 64 is that one of the models used for the
development of Eq. 25, on which Eq. 64 is based, was a column composed of
the compression flange and half of the web 1 6 . For beams under moment
gradient Eq. 64 is conservative.
Channels Bent About the Major Axis in the Shear Center Plane
The same formulas apply for channels as for doubly symmetric wide-
flange beams.
4. BENDING STRENGTH AS DETERMINED BY LOCAL BUCKLING
The relevant slenderness parameter in Fig. 1 for the case of local
buckling is the width-to-thickness ratio of the compressed plate elements
of the cross section. In the following section expressions for Mu, the
nominal moment capacity, and values for will be presented for beams
31.

subjected to bending moment.


The ultimate strength of a plate under compression, such as the com-
pression flange of a member in flexure, is governed by its post-buckling
56
strength . This means that after the formation of the buckling waves,
failure does not occur until the member can sustain no additional load due
to the development of relatively large deformations. Post-buckling strength
will be explicitly used in the LRFD criteria for webs and for compression
flanges of box beams. While post-buckling strength exists for unstiffened
elements, such as the compression flanges of wide-flange beams, the result-
ing deformations are so large that the usual practice is to base the strength
of these elements on the classical buckling theory 56 . This will also be done
here.

The first element to be considered will be the compression flange of


wide-flange beams, and the relevant width-to-thickness ratio is b f / t f , where
bf is the flange width and tf is its thickness.
4.1 Flange Local Buckling in the Elastic Range, Wide-Flange Shapes.

The critical buckling stress of a plate subjected to uniform compres-


sion along one of its axes is 5 :

(65)

where E and are the modulus of elasticity and Poisson's ratio, respectively,
k is the plate buckling coefficient and t/b is the ratio of plate thickness
to plate width perpendicular to the direction of the applied compressive
stress.
For the special case of the flange of a wide-flange or I-shaped beam,
t = tf and b = b f /2 in Eq. 66, and k is dependent on the rotational stiff-
ness of the web. The value of k can vary from 0.425 for zero restraint to
32.

1.277 for full fixity. There is no convenient formula to relate k to the


restraint, but the relationship is given in graphical form by Stowell and
54
Lundquist , who present curves relating the plate buckling coefficient to
the restraint factor which can vary from 0 to . If it is assumed that
the lower part of the web is fixed against rotation, then

(66)

In this equation t is the web thickness and d is the section depth. With
the aid of the charts from Ref. 54 and Eq. 66, the plate buckling coeffi-
cients of all W and S shapes in the AISC Manual were determined, and it was
found that the mean value of k was km = 0.821 and its coefficient of varia-

tion was 0.13.

The mean value of the critical stress is thus (Eq. 65)

(67)

where Em = E = 29,000 ksi, = 0.30 (assumed to be deterministic) and


km = 0.821.

In setting up the design moment Mu, it is recommended that the


nominal buckling stress (Fcr)n be equal to the mean buckling stress of
Eq. 67, i.e.,

(68)

The resistance factor (Eq. 10) is equal to

(69)
33.

The first ratio in Eq. 69 is unity. Since no extensive test results of

flange buckling in flexure exist, some assumptions will need to be made

about the Test-to-Prediction ratio. It will be assumed that the ratio is

arbitrarily but reasonably taken to be equal to 1.05, with a coefficient of


variation of 0.05. The coefficient of variation VR is determined from

(70)

where VM = 0.06 (for the modulus of elasticity) VF = 0.05 (using the same

value as previously) = 0.13 (for the plate buckling coefficient k) and


= 0.05 (the assumed value for the Test-to-Prediction ratio). With these
numbers VR = 0.16 from Eq. 71, and equals

The limit of applicability of Eq. 69 is when yielding commences in the


presence of residual stress, i.e.,

(71)

The limiting flange slenderness is thus

(72)

4.2 Flange Local Buckling in the Plastic Range

The nominal moment is the plastic moment, and so the same values of Mu
and , as given previously for the plastic range, apply, i.e., from Sec.
3.7,
(73)
34.

The critical flange slenderness ratio, beyond which Eq. 73 no longer is


18
valid, is given by Lay as

(74)

In this equation Gst and Est are the strain-hardening shear modulus and the
strain-hardening modulus in tension or compression, respectively. The mean
19
value of Est was reported by Doane for all data available to him, and it
18
is equal to ( E s t ) m = 630 ksi. According to Lay ,

(75)

where G = E/(1 + ) was substituted. Using E = 29,000 ksi, = 0.3 and

E s t = 630 ksi, (Gst)m = 2260 ksi. The mean value of

(76)

for all Wand I shapes in the AISC Manual, and thus the limiting flange
slenderness is equal to

(77)

The design conditions in the plastic range are thus given by Eq. 73
64
as long as , where Fy is in units of ksi. Based on tests a
more liberal limit of has been adopted in Suppl. 3 (1974) of the
AISC Specification for compact beams designed by the allowable stress method.
35.

4.3 Flange Buckling in the Inelastic Range


As in the case of lateral-torsional buckling the straight line form
will be used in the inelastic range also for flange local buckling (Fig. 1)
and thus, from Eq. 5:

(78)

where

(79)

4.4 Design Equations for Flange Local Buckling of Wide-Flange and Channel
Shapes

It was shown in the previous section of this report that the resistance
factor = 0.86 is applicable in the plastic range and = 0.81 is appli-
cable in the elastic range for the maximum moment governed by the local
buckling of the flanges of wide-flange shapes. For the designer it would
be convenient if the same could be used in both ranges. Since = 0.86
was used for beams under the limit state of lateral-torsional buckling and
for short columns, it would also be convenient if the same value of could
be used for all three ranges of beams failing by flange local buckling.
Considering (1) the assumptions underlying the development of for flange
local buckling, (2) the post-buckling strength which exists, and (3) the
fact that for shapes with hot-rolled elements with which these criteria are
concerned the b f /2t f ratio will not generally exceed the value of (b f /2t f ) p
(Eq. 77) by very much, it is quite reasonable to also take = 0.86 for
flange local buckling.
36.

The following equations of design moment Mn apply to the design of


beams controlled by the local buckling of the compression flange of doubly
or singly symmetric wide-flange beams and for channels bent in the shear
center plane. In all range = 0.86
(80)

(81)

(82)
Zx and Sx are the plastic and elastic section moduli, respectively,

Fy is the specified yield stress of the grade of steel used, and

(83)

where Fr is the residual stress equal to 10 ksi for rolled shapes and 20 ksi
for welded shapes. The material properties E = 29,000 ksi and = 0.3 are
to be used in these equations. The slenderness parameter is equal to

for wide-flange shapes

for channels

The plate buckling coefficient k is equal to


k = 0.821 for wide-flange and channel shapes
The limiting slenderness at the end of the plastic range is

(84)

depending on whether plastic or elastic analysis is used for the design of


the member.
37.

The limiting slenderness at the beginning of the elastic range is

(85)

In Eq. 84 the value of 52.2 is recommended instead of 48.5 which would

result from Eq. 77, so that the same limiting flange width-thickness ratio

is used as in the 1969 AISC Specification. The limiting value of

applies only if design is performed according to plastic theory. Based on

tests performed by Yura and his associates at the University of Texas64 a

more liberal limit of was adopted in Supplement 3 of the AISC Speci-

fication (June 12, 1974) for compact beams designed according to allowable

stress design. The value used for the limit for plastic design (i.e.

is probably conservative and further research should be performed

to arrive at a more realistic value. The higher limit of is probably

too liberal to permit the full moment redistribution implied in plastic

design 6 5 , although this should be further verified on experimental and theo-

retical studies on continuous beams and on rigid frames.

4.5 Flange Local Buckling of Box Beams

For the case of the compression flange of box beams the elastic buckling

theory underestimates the ultimate strength by a considerable amount, and so

post-buckling strength must be explicitly accounted for 5 6 . In lieu of using

the exact theory of plates with large deformations the compression flange is

replaced by a plate having an effective width be which is smaller than the

actual width when the stresses have exceeded the elastic critical buckling

s t r e s s 5 6 , 5 7 . This effective width has been determined experimentally from

many t e s t s 5 6 , 5 7 , and the following formula has been recommended for use

with edge stiffened cold-formed members:

(86)
38.

In this equation b and t are the width and thickness of the stiffened plate,

be is the effective width and fmax is the limiting or actual stress in the

equivalent plate.

In Appendix C of the 1969 AISC Specification, where provisions for

effective width calculations are given for stiffened elements from hot-

rolled plates, the coefficient 0.415 in the parenthesis of Eq. 90 has been

modified to a value of 0.37 for square and rectangular sections of uniform

thickness and to 0.32 for other uniformly compressed elements. The smaller

of these two coefficients gives a larger effective width and is thus more

liberal. The provision for the effective width in the AISC Specification

will be retained here as being appropriate for box beams fabricated from

hot-rolled plates. There is a question as to which of the two coefficients,

i.e., 0.37 or 0.32 should be used, or whether the distinction in the AISC

Specification should be retained. The distinction appears to be due to the

fact that with the larger value (i.e., 0.37) the buckling of the restraining

elements, which in this case are the webs, is also involved. This would be

more the case when the whole box is in compression as a column, and not so

much in the case of flexure. Thus the smaller value will be used here for

bending. The following formula is recommended for the effective width in

the LRFD criteria for the flange of a box beam subjected to flexure:

(87)

The limiting stress fmax in Eq. 86 has been replaced by the stress at the

proportional limit, Fy - Fr, since the post-buckling theory is based on

elastic assumptions, and thus the limit state according to this theory is
the start of yielding. Eq. 87 ceases to apply when be bf, the actual
flange width. If be = bf is substituted into Eq. 87, the limiting value
39.
(b f /t f ) is obtained:

(88)

Full plastic capacity can be expected when b f /t f is less than ( b f / t f ) p ,

where

(89)

This formula is the same as was recommended for the limit of web slenderness
under compression and flexure32 in plastic design. Between the limits

defined by Eqs. 88 and 89 the moment capacity will be represented by a


linear relationship.
The flexural capacity of box beams, as defined by the limit state of
flange local buckling, is thus given by the following formulas:

(90)

(91)

(92)
40.

In these equations Fr = 10 Ksi should be used, and (Sx)eff is the

effective section modulus computed for the section having a compression

flange with a width be, where be is defined by Eq. 87. The resistance

factor for box beams is assumed to be the same as for wide-flange beams,

that is, = 0.86.

4.6 Design Equations for Web Buckling

In the plastic range the plastic moment can be reached and maintained

for a sufficient rotation capacity as long as the web slenderness

d/t < (d/t)p where 3 2

(93)

More recent research has indicated that the limiting web slenderness given

by Eq. 93 is too conservative, and based on experimental and analytical

studies it has been recommended that revised limits as follows are appro-

priate:

For continuous beams and frames designed according to the plastic

method 6 6 :

(94)

(95)

For simple beams and for indeterminate structures in which the forces

are determined by elastic analysis 6 3 , 6 6 :

(96)

(97)
41.

Equations 94 through 97 are recommended for use in LRFD with a resis-

tance factor = 0.86 because they reflect the present (1976) state of

knowledge. These equations also apply to beam-columns where a factored

design axial force PD is present. For beams PD = 0. The term Py = AFy

in these equations, where A is the gross cross-sectional area of the member.

Thus the same dual situation exists for the web slenderness limit as

for the flanges. This is unfortunate because it complicates design. How-

ever, the higher limits are in the present (1976) AISC Specification for

use with compact beams in allowable stress design, and they are based on

good evidence. Research has not, however, proven yet that these same limits

are adequate for plastic design.

The rules (Eqs. 94 through 97) for the web slenderness limits are not

the same as in the AISC Specification when axial load is present, and they

reflect the more liberal requirements resulting from the research performed

by Kulak 6 6 . The various AISC and LRFD requirements for the web slenderness

limit are compared in Fig. 7, where it is evident that the LRFD rules permit

a considerable liberalization of the requirements when the axial load exceeds

about 20 percent of the yield load Py.

In the elastic range the scheme of using the critical elastic buckling

stress no longer applies because of post-buckling strength. Furthermore,

the corresponding width-thickness ratios are such that the beam will be a

plate girder and stiffeners are required. The maximum bending moment for

plate girders is

(98)
42.

In this equation Mcr is the maximum moment as determined by lateral-torsional

and local buckling considerations, Aw is the web area and Af is the flange

area. If Mcr = My = Sx Fy is substituted into Eq. 98, no reduction because

of web deformation is required if the square bracket becomes zero, i.e.,

(99)

where h is the web depth. Since h/t d/t, the maximum d/t ratio, (d/t) y ;
for which Mmax = My, is equal to

(100)

As long as d/t (d/t) r , the section is considered to be a beam and not a


plate girder.
In the 1969 AISC Specification the underlying philosophy is that as
long as d/t (d/t) p , the plastic moment can be reached, and in the range
(d/t)p d/t (d/t)r the capacity is My = Sx Fy. Thus there is a jump at

(d/t)p so that up to (d/t)p the allowable stress is 0.66 Fy, and for d/t
just above (d/t)p the allowable stress is 0.60 Fy. In reality some kind
of a transition exists. In this report it is suggested that a linear
transition be used from M = Mp at d/t = (d/t)p to M = My at (d/t) r . Thus,
the following formulas are suggested for the determination of the nominal
bending strength of doubly symmetric wide-flange beams:

(101)

(102)
43.

In Eq. 102, My is the moment causing web yielding, i.e.,

which can be taken to be equal to Fy Sx for all practical purposes. Should

the value of d/t exceed (d/t) r , the rules pertaining to plate girders apply.

The same rules (i.e., Eqs. 101 and 102) may be applied to singly sym-

metric wide-flange shapes, channels and box girders, the webs being assumed

to act in a similar manner to the webs in doubly symmetric wide-flange shapes.

It is suggested that = 0.86 be used for both the plastic and the inelastic

ranges.

5. HYBRID BEAMS

Hybrid beams are wide-flange sections made from two grades of steel:

the flanges are usually fabricated from material having a higher yield stress

than the web. For example, the flanges may be of material with Fy = 50 or

100 ksi, while the web may have Fy = 36 ksi. Research has shown that pre-

mature yielding of the web does not affect the total bending behavior of

the beam and so it is often advantageous to use a less expensive grade of

steel for the web. Considerable research has been performed on hybrid

girders, and this work was summarized by a. subcommittee of the ASCE55 in a

Committee Report which also contained design recommendations which were used

as the basis for the rules on hybrid girders in the 1969 AISC Specifications.

In the following the recommendations in Ref. 55 will be used in a modified

way to the LFD criteria of this report.

Plastic design was not recommended by the ASCE subcommittee for hybrid

girders, and the criterion for design was the attainment of yielding in the

flanges. There appears to be, however, no reason why plastic design should

not be used for these members, as long as local and lateral-torsional buck-

ling is prevented to permit the development of plastic hinges.


44.

The plastic moment of a doubly symmetric hybrid wide-flange beam is

equal to

(103)

where Fyf = yield stress of the flange material

Fyw = yield stress of the web material

Af = area of one flange, i.e., Af = bf tf

Aw = area of web, i.e., Aw = t (d - 2 t f )

d, t, t f , bf = the dimensions of the section.

In lieu of Eq. 103 an approximate equation may be used which will give

nearly identical results for doubly symmetric hybrid beams in which the area

of the compression flange is greater than the area of the tension flange by

not more than the factor 1.25. This equation i s 5 5 :

(104)

where

(105)

Since the value of the limiting slenderness ratios (Fig. 1) are

dependent principally on the elastic and strain hardening moduli, the

previously used limits are applicable, i.e.,

(106)

(107)

(108)
45.

(109)

In the inelastic range the moment capacity will again be assumed to be

a straight-line transition (Fig. 1) between Mp at and My at

The beginning of the elastic range occurs at M = Mr, but there will

be two such yield moments: the flange yield moment MHR which is equal to

(from Ref. 55)

(110)

and the web yield moment Myw, where

(111)

An approximate expression may be used, with similar restrictions as to

the ratio of compression and tension flange areas implied by Eq. 105, as

follows 55 :

(112)

where (113)

In Eq. 113 is the ratio: (the distance from the outer edge of the tension

flange to the neutral axis)/depth of the section.

In the elastic range the critical moment as controlled by flange local

buckling is (Eq. 65)

(114)

where k = 0.425 is to be used 5 5 . This local buckling coefficient assumes


no web restraint because the web is already yielded when the flange is still
46.

elastic. The term RHE in Eq. 113 accounts for the increased stress in the
flange due to web yielding. At the limit where Me = MHR, Fcr = Fyf - Fcr,
and thus

(115)

(116)

The lateral-torsional buckling moment in the elastic range is deter-


mined by Eq. 7, or, alternatively, by Eq. 18. Because of the fact that the
web is yielded while the flange is still elastic, the ASCE subcommittee55
recommends that, conservatively, only one of the two stresses under the
radical in Eq. 18 be used, i.e.,

(117)

At the limit between elastic and inelastic behavior Me = MHr, or


Fw = Fyf - Fr and thus

(118)

It is recommended that = 0.86 be used in the plastic, the inelastic


and the elastic ranges, respectively.
6. DESIGN FOR SHEAR STRENGTH
Bending moment and shear force occur together in a beam, and only at
special locations within the span and at special boundaries is one of these
maximum while the other is zero. In usual situations, however, the effect
of the shear force is small compared to the effect of the bending moment.
47.

The exception is at rigid supports where both high shear and high moment
occur simultaneously. While a variety of theoretical models have been pro-
posed for the development of interaction equations between shear force and
bending moment at the plastic limit, experiments tend to show that the
plastic moment Mp is not diminished unless the shear force exceeds Vp, the
plastic shear force corresponding to the shear yielding of the complete
32
web . This is so because high shear force and high bending moment occur
at points of high moment gradient where the effect of strain hardening is
most pronounced. Thus the full bending capacity can be counted on if
V Vp where
(119)

In this equation Aw is the web area


(120)

and is the yield stress in shear.


Various theories exist which relate the shear yield stress to the
tensile or compressive yield stress Fy, and the one which most closely
approximates this relationship for structural steel is the von Mises yield
condition, according to which

(121)

Thus the shear force Vp, is equal to

(122)

If a mean value of Fy = (Fsy)m is used

(123)
48.

where (Fsy)m is the mean static yield stress of the web material. In Ref. 2

it was shown that (Fsy)m = 1.1 Fy, where Fy is the specified yield stress of
the grade of steel. The coefficient of variation was given as 0.11. These
values apply to the webs of rolled wide-flange shapes and (Fsy)m is higher

than the value of 1.05 Fy used in previous sections of this report for the
flange material.
The nominal shear force is expressed by Eq. 122, where Fy is the speci-
fied yield stress. The coefficient of variation is (Eq. 3)

In the calculation of VR, VM = 0.11 was used for the uncertainty of the yield
stress, the small value of VF = 0.02 was assumed for the coefficient of vari-

ation of the uncertainties of fabrication since web dimensions are likely to


vary only a small amount from nominal values, and VP = 0.10 was assumed for
the coefficient of variation of the ratio of the actual shear yield stress
to the theoretical yield stress according to the von Mises yield condition.
The resistance factor is thus equal to

The resistance side of the design criterion for the shear strength
(Eq. 1) is thus

(124)

This shear force may not be exceeded, but as long as the shear force is less
than or equal to it, MP may be also attained at the same section.
49.

The relationship of Eq. 124 applies for webs in beams which are sub-

jected to bending moment and shear force in the plane of the web. In the

case of shear stresses due to biaxial loading or combined loading from

transverse forces and torsional moments, the nominal maximum shear stress

is the shear yield stress, or

(125)

and = 0.86 by the same reasoning as used above.

7. COMBINED STRESSES

In the case of biaxial bending, or combined bending and torsion, the

normal stresses are determined by the usual elastic theory using either

first order or second order calculations, as appropriate. The maximum

normal stress should not exceed Fy, where is the resistance factor. In

determining this factor it will be assumed that (Fys)m = 1.05 Fy, VM = 0.10,

VF = 0.05 and that the mean prediction from theory is equal to the experi-

mentally measured stress and that VP = 0.05. This latter assumption implies

that the calculations of stresses, by assuming elastic theory, are quite

reliable. Thus

(126)

and
(127)

8. SUMMARY

The resistance side of the load factor design criterion (Eq. 1) consists

of the product Rn, where is a "resistance factor" accounting for the

uncertainties of the "nominal resistance" Rn. In this report values for

and expressions for Rn were developed for beams under bending moment,
50.

shear and combined biaxial and torsional forces for elastic, partially
yielded and plastic beams. These are summarized in the Appendix in the
form of a tentative set of design criteria.
9. ACKNOWLEDGEMENTS

The work in this report was sponsored by the American Iron and Steel
Institute as AISI Project 163 "Load Factor Design of Buildings". The
advice and the discussion of the Project Advisory Committee is gratefully
acknowledged, and their critical remarks proved to be of great assistance
in arriving at the design criteria. This committee consists of Messrs.
L. S. Beedle, C. A. Cornell, J. A. Gilligan, W. C. Hansell, I. Hooper,
W. A. Milek, Jr., C. W. Pinkham, G. Winter and I. M. Viest.
51.

10. REFERENCES

1. T. V. Galambos
"Load Factor Design of Steel Structures"
AISC Engineering Journal, July 1972
2. T. V. Galambos, M. K. Ravindra
"Load Factor Design Criteria for Steel Buildings"
Research Report No. 18, Structural Division, Civil and Environmental
Engineering Department, Washington University, St. Louis, Sept. 1973.

3. T. V. Galambos
"Structural Members and Frames"
Prentice-Hall, Inc., Englewood Cliffs, N. J. 1968.

4. M. G. Salvadori
"Lateral Buckling of I-Beams"
Transactions, ASCE, Vol. 120, 1955.

5. Column Research Council


"Guide to Design Criteria for Metal Compression Members"
Editor: B. G. Johnston
John Wiley and Sons, Inc., New York, 1976, Third Edition.

6. N. S. Trahair
"The Elastic Stability of Frame Structures"
A thesis submitted for the Ph.D., University of Sydney,
Sydney, Australia.

7. N. S. Trahair
"Elastic Stability of Continuous Beams"
Journal of the Structural Division, ASCE, Vol. 95, No. ST6, June 1969.

8. A. J. Hartmann
"Elastic Lateral Buckling of Continuous Beams"
Journal of the Structural Division, ASCE, Vol. 93, No. ST4, Aug. 1967.

9. A. J. Hartmann, W. H. Munse
"Flexural-Torsional Buckling of Planar Frames"
Journal of the Engineering Mechanics Division, ASCE, Vol. 92, No. EM2.
April 1966.

10. M. G. Salvadori
"Lateral Buckling of Beams of Rectangular Cross Section Under Bending
and Shear"
Proceedings, 1st U. S. Congress of Applied Mechanics, 1951.

11. G. C. Lee, Y. Ueda


"Survey of Lateral Buckling Experiments"
Fritz Engineering Laboratory Report No. 205H.4, Lehigh University
July 1961.
52.

12. A. J. Hartmann
"Experimental Study of the Flexural-Torsional Buckling Behavior of
Elastic Planar Structures"
Civil Engineering Department, Marquette University, Milwaukee,
Wisconsin, July 1969.

13. R. A. Hechtman, J. M. Hattrup, E. F. Styer, J. L. Tiedermann


"Lateral Buckling of Rolled Steel Beams"
Transactions of the ASCE, Vol. 122, 1957.

14. J. W. Clark, J. R. Jambock, Jr.


"Lateral Buckling of I-Beams Subjected to Unequal End Moments"
Journal of the Engineering Mechanics Division, ASCE, No. EM3,
July 1957.

15. S. Desai
"Tension Testing Procedure"
Fritz Engineering Laboratory Report No. 237.44,
Lehigh University, Feb. 1969.

16. M. G. Lay, T. V. Galambos


"Inelastic Steel Beams Under Uniform Moment"
Journal of the Structural Division, ASCE, Vol. 91, No. ST6, Dec. 1965.

17. M. G. Lay, T. V. Galambos


"Inelastic Behavior of Beams Under Moment Gradient"
Journal of the Structural Division, ASCE, Vol. 93, No. ST1, Feb. 1967.

18. M. G. Lay
"Flange Local Buckling in Wide-Flange Shapes"
Journal of the Structural Division, ASCE, Vol. 91, ST6, Dec. 1965.

19. J. F. Doane
"Inelastic Instability of Wide-Flange Steel Beams"
M.S. Thesis, University of Texas at Austin, Aug. 1969.

20. G. C. Driscoll, et a l .
" P l a s t i c Design of Multi-Story Frames", Chap. 3 "Beams"
Lehigh University, Aug. 1965.
21. J. P. Bansal
"The Lateral Instability of Continuous Steel Beams"
CESRL Dissertation No. 71-1, Aug. 1971, Dept. of Civil Engrs.,
The University of Texas at Austin.

22. E. P. Popov, J. A. Willis


"Plastic Design of Cover-Plated Continuous Beams"
Journal of the Engineering Mechanics Division, ASCE, Vol. 84, No. EM1,
Jan. 1958.

23. T. V. Galambos, M. G. Lay, P. F. Adams


"Experiments on High Strength Steel Members"
WRC Bulletin No. 110, Nov. 1965.
53.

24. J. Janes, C. Massonnet


"The Extension of Plastic Design to Steel A52"
Publications, IABSE, Vol. 27, 1967.
25. "Experiments on A572 (Grad 65) Steel Beams"
Fritz Engineering Laboratory Report No. 343.4,
Lehigh University, Aug. 1969.
26. T. Kusuda, R. G. Sarubbi, B. Thurlimann
"Lateral Bracing of Beams in Plastic Design"
Fritz Engineering Laboratory Report No. 205E.11,
Lehigh University, 1960.
27. T. V. Galambos, G. C. Lee
"Post-Buckling Strength of Wide-Flange Beams"
Proc. of the ASCE, Vol. 88, EM 1, Feb. 1962.

28. J. Prasad, T. V. Galambos


"The Influence of Adjacent Spans on the Rotation Capacity of Beams"
Fritz Engineering Laboratory Report No. 205H.12,
Lehigh University, June 1963.
29. G. C. Lee, A. T. Ferrara, T. V. Galambos
"Experiments on Wide-Flange Beams"
Welding Research Council Bulletin No. 99, Sept. 1964.
30. H. A. Sawyer
"Post-Elastic Behavior of Wide-Flange Steel Beams"
Journal of the Structural Division, ASCE, Vol. 87, No. ST8, Dec. 1961.
31. R. J. Smith, et.al.
"Experiments on Wide-Flange Beams Under Moment Gradient"
Welding Research Council Bulletin No. 142, July 1969.
32. ASCE Manual No. 41
"Plastic Design in Steel, A Guide and Commentary"
ASCE and WRC, 1971.
33. J. E. Dibley
"Lateral-Torsional Buckling of I-Sections in Grade 55 Steel"
BISRA Industry Report, 1968.
34. Y. Fukumoto
A Japanese report of the Department of Civil Engineering,
Nagoya University, Nagoya, Japan.
35. L. S. Beedle, L. Tall
"Basic Column Strength"
Journal of the Structural Division, ASCE, Vol. 86, ST-7, July 1960.
36. L. Tall, G. A. Alpsten
"On the Scatter in Yield Strength and Residual Stresses in Steel
Members"
Final Report, IABSE Symposium on the "Concepts of Safety of Structures
and Methods of Design" London, Sept. 1969.
54.

37. D. K. Feder, G. C. Lee


"Residual Stresses in High Strength Steel"
Fritz Engineering Lab. Report No. 269:2, April 1959.
38. P. F. Adams, M. G. Lay, T. V. Galambos
"Experiments on High Strength Steel Members"
Fritz Engineering Laboratory Report 297.8, Lehigh University,
July 1964.
39. M. G. Lay, R. Ward
"Residual Stresses in Steel Sections"
Melbourne Research Lab Report, April 1969.
40. G. A. Alpsten
"Thermal Residual Stresses in Hot-Rolled Steel Members"
Fritz Engineering Lab. Report No. 337.3, Dec. 1968.
41. R. A. Scheid
"Stub Column Test and Residual Stress Measurement",
(A572, Grad 65)
Fritz Engineering Laboratory Report No. 343.3, May 1968.
42. N. R. N. Rao
"Columns Reinforced Under Load"
Fritz Engineering Laboratory Report No. 286.1, April 1962.
43. N. R. N. Rao, P. Marek, L. Tall
"Hybrid Steel Columns"
Fritz Engineering Laboratory Report No. 305.2, May 1969.
44. Y. Kishima, G. A. Alpsten, L. Tall
"Residual Stresses in Welded Shapes of Flane-Cut Plates in ASTM
A572 (50) Steel"
Fritz Engineering Laboratory Report No. 321.2, June 1969.
45. R. McFalls, L. Tall
"A Study of Welded Columns Manufactured from Flame-Cut Plates"
Fritz Engineering Laboratory Report No. 321.1, June 1967.
46. N. R. N. Rao, L. Tall
"Residual Stresses in Welded Plates"
Fritz Engineering Laboratory Report No. 249.7, Oct. 1960.
47. F. Nishino
"Buckling Strength of Columns and their Component Plates"
F r i t z Engineering Laboratory Report No. 290.10, 1964.
48. T. V. Galambos
"Inelastic Lateral Buckling of Beams"
Journal of the Structural Division, ASCE, Vol. 89, No. ST 5, Oct. 1963.
55.

49. Y. Fukumoto, T. V. Galambos


"Inelastic Lateral-Torsional Buckling of Beam-Columns"
Journal of the Structural Division, ASCE, Vol. 92, No. ST2,
April 1966.
50. N. S. Trahair, S. Kitipornchai
"Buckling of Inelastic I-Beams Under Uniform Moment"
Research Report No. R187, The University of Sydney, School of Civil
Engineering, Jan. 1972.

51. Y. Fukumoto, M. Kubo


"Lateral Buckling Strength of Girders with Bracing Systems"
Preliminary Report, Ninth Congress, IABSE, Amsterdam, May, 1972.
52. S. Rajasekaran, D. W. Murray
"Inelastic Buckling of Thin-Walled Members"
Draft of paper presented to ASCE for publication, 1972.
53. J. R. Benjamin, C. A. Cornell
"probability Statistics and Decision for Civil Engineers"
McGraw-Hill, 1970.
54. E. Z. Stowell, G. J. Heimerl, C. Liboe, E. E. Lundquist
"Buckling Stresses for Flat Plates and Sections"
ASCE Transactions, Vol. 117, 1952.
55. Subcommittee on Hybrid Beams and Girders
"Design of Hybrid Steel Beams"
Journal of the Structural Division, ASCE, Vol. 94, No. ST6, June 1968.

56. G. Winter
"Commentary on the 1968 Edition of the Specification for the Design of
Cold-Formed Steel Structural Members"
AISI, 1970.

57. W.-W. Yu
"Cold-Formed Steel Structures"
McGraw-Hill Book Co., New York, 1973.
58. D. A. Nethercot
"Residual Stresses and Their Influence Upon the Lateral Buckling of
Rolled Steel Beams", Research Report, Dept. of Civil and Structural
Engineering, University of Sheffield (undated).
59. D. A. Nethercot
"Factors Affecting the Buckling Stability of Partially Plastic Beams"
Proc. ICE, 1972, Vol. 53, pp. 285-304.
60. D. A. Nethercot
"Buckling of Welded Beams and Girders"
Research Report No. 30, Dept. of Civil and Structural Engineering,
University of Sheffield, Jan. 1973.
56.

61. D. A. Nethercot
"Imperfections and the Design of Steel Beams"
Research Report No. 33, Dept. of Civil and Structural Engineering,
University of Sheffield, Sept. 1973.
62. D. A. Nethercot
"Inelastic Buckling of Steel Beams Under Non-Uniform Moment"
Research Report No. 39, Dept. of Civil and Structural Engineering,
University of Sheffield, Nov. 1973.

63. A. D. Croce
"The Strength of Continuous Welded Girders with Unstiffened Webs"
CESRL Thesis No. 70-2, Jan. 1970, University of Texas at Austin.

64. H. E. Costley
"Lateral and Local Instability of Continuous Beams"
CESRL Thesis No. 70-1, Jan. 1970, University of Texas at Austin.

65. A. F. Lukey, P. F. Adams


"The Rotation Capacity of Beams Under Moment Gradient"
ASCE Journal of the Structural Division, ST6, June 1969.

66. M. J. Peolynn, G. L. Kulak


"Web Slenderness Limits for Compact Beam-Columns"
Structural Engineering Report No. 50, Sept. 1974, University of
Alberta.
57.

11. NOTATION

A : Area of cross section

Af : Area of one flange

Aw : Area of web

Cb, : Equivalent moment factors

Cw : Warping moment of inertia

Dt : Cross-sectional coefficient defined by Eq. 17

E : Elastic modulus

Est : Strain-hardening modulus


Fcr : Critical plate buckling stress

Fr : Residual stress

Fsy : Static yield stress

(Fsy)m : Mean static yield stress

Fv, Fw : Terms defined by Eqs. 18 and 19

Fy : Yield stress

Fyf : Flange yield stress

Fyw : Web yield stress

G : Shear modulus

Gst : Strain-hardening shear modulus

Iy : Minor axis moment of inertia

J : St. Venant torsion constant

K : Coefficient

Ky, Kz : Effective length factors

Lb : Unbraced length

Lbp : Critical unbraced length for plastic range

Lby : Unbraced length at beginning of elastic range


56.

M : Bending moment

M1, M2 : Moments at ends of unbraced segment

Me : Elastic critical moment

Mm : Mean moment

Mu : Ultimate moment

Mp : Plastic moment

Mr : Yield moment, when the elastic stress is Fy - Fr

MHR : Flange yield moment

Myw : Web yield moment

Qn : Nominal load effect

RHP : Plastic factor for hybrid beams

RHE : Elastic factor for hybrid beams


Rm : Mean resistance

Rn : Nominal resistance

Sx : Section modulus about major axis

VF : Coefficient of variation of fabrication uncertainties

VM : Coefficient of variation of material properties

VP : Coefficient of variation of theoretical assumptions


VR : Coefficient of variation of resistance

Zx : Plastic modulus about major axis

b : plate width

be : effective width

bf : flange width

d : section depth

h : section depth

h : web depth
59.

k : plate buckling coefficient


rT : radius of gyration of compression flange plus one sixth
of the adjacent web area
ry : minor axis radius of gyration

t : web thickness
t : plate thickness

: safety index
: load factors
: strength factor
: web restraint factor
: slenderness, as defined in Fig. 1
: slenderness at end of plastic range
: slenderness at beginning of elastic range
: Poisson's ratio
: standard deviation
: shear stress
: factor in Eq. 108
Subscripts:
e : elastic
f : flange
m : mean
n : nominal
p : plastic
r : yield, including effect of residual stress
w : web
60.

Fig. 1. Schematic Representation of Beam Bending Moment Capacity


61.

Fig. 2. Histogram of 185 Elastic Lateral-Torsional


Buckling Tests of Beams
62.

Fig. 3. Schematic Description of Elastic and Inelastic Buckling


63.

Fig. 4. Lateral-Torsional Buckling Test Setups


64.
Fig. 5. Beam Tests in the Inelastic Range
65.

Fig. 6. Variation of Nominal and Factored Moment for Rolled Shape


66.

Eqs. 94 and 95

Eqs. 96 and 97

AISC Specification, compact shapes, allowable stress design

AISC Specification, plastic design.

Fig. 7. Web-slenderness Requirements According to LRFD and AISC


67.

APPENDIX

Following are excerpts from "Tentative Load and Resistance Factor


Design Criteria for Steel Buildings." These excerpts concern the rules
for steel beam design. At the beginning of the Appendix is given the
Table of Contents of the criteria, and the relevant sections, which are
reproduced later in the Appendix, are boxed in this Table of Contents.
Reproduced sections not pertaining to beams but reproduced in the Appendix
are crossed out. The page numbers are those which are used in the LRFD
Criteria, and they correspond to the Table of Contents.
i

TABLE OF CONTENTS

Part 1: Criteria 1-1

Sec. 1 General Provisions 1-1

1.1 Scope 1-1

1.2 Definition of LRFD 1-1

1.2.1 Limit State: Strength 1-2


1.2.2 Limit State: Serviceability 1-3

1.3 Loads and Load Combinations 1-3

1.3.1 Load Types 1-3


1.3.2 Load Combinations 1-4
1.3.3 Load Factors for Strength Design 1-5
Sec. 2 Design Criteria for the Limit State of Strength 2-1
2.1 Types of Structures 2-1

2.1.1 Material 2-1

2.1.2 Framing 2-1


2.2 Structural Analysis 2-1
2.3 Design of Members 2-2
2.3.1 Tension Members 2-2
2.3.1.1 Factored Maximum Strength 2-2
2.3.1.2 Limiting Slenderness Ratios 2-3
2.3.2 Compression Members 2-3

2.3.2.1 Factored Maximum Strength 2-3

2.3.2.2 Effective Length Factor 2-4

2.3.2.3 Effective Slenderness Ratio 2-4


2.3.2.4 Flexural-Torsional Buckling 2-4

2.3.2.5 Tapered Members 2-5


2.3.2.6 Details of Built-Up Compression Members 2-5
ii

2.3.3 Flexural Members 2-5

2.3.3.1 Scope 2-5

2.3.3.2 Factored Maximum Strength of Webs in Shear 2-6

2.3.3.2.1 Factored Maximum Strength of Beam Webs in Shear 2-6


2.3.3.2.2 Factored Maximum Strength of Plate-Girder Webs in Shear 2-6

2.3.3.2.3 Stiffener Requirements 2-8

2.3.3.2.4 Interaction Between Bending Moment and Shear Force 2-9

2.3.3.3 Factored Maximum Moment Capacity 2-10

2.3.3.3.1 Maximum Moment Capacity for Beams 2-10

2.3.3.3.2 Maximum Moment Capacity for Plate-Girder 2-12

2.3.3.3.3 Maximum Moment Capacity for Composite Beams 2-14

2.3.3.3.3.1 Definition 2-14

2.3.3.3.3.2 Factored Maximum Moment Capacities 2-14

2.3.3.3.3.3 Shear Connectors 2-18

2.3.3.3.3.4 Concrete Slabs on Formed Metal Deck 2-21

2.3.3.3.3.5 Vertical Shear Capacity 2-22

2.3.3.3.3.6 Special Cases 2-22

2.3.4 Members Under Combined Flexure and Axial Force 2-22

2.3.4.1 Members in Flexure and Tension 2-22

2.3.4.2 Members in Flexure and Compression 2-23

2.3.4.3 Definition of Terms in Sec. 2.3.4.1 and 2.3.4.2 2-23

2.3.4.4 Tapered Beam-Columns 2-25

2.3.5 Members Under Combined Stress 2-25

2.4 The Design of Connections 2-27

2.4.1 Definition 2-27

2.4.2 Design of Connecting Elements 2-27

2.4.3 Connectors 2-27

2.4.3.1 Welds 2-27


iii

2.4.3.2 Bolts, Rivets and High Strength Bolts 2-29

2.4.3.2.1 Tension 2-29

2.4.3.2.2 Shear 2-29

2.4.3.2.3 Combined Tension and Shear 2-30

2.4.3.2.4 Simplified Treatment of 2-30

2.4.3.2.5 Long Joints 2-30

2.4.3.2.6 Bearing Capacity of Bolt and Rivet Holes 2-30

2.4.3.2.7 Bolt and Rivet Hole Details 2-31

2.4.3.2.8 High-Strength Bolt Friction-Grip Joints 2-31

2.4.4 Bearing Stresses on Contact Area 2-32

2.4.4.1 Milled Surfaces 2-32

2.4.4.2 Expansion Rollers and Rockers 2-32

2.4.4.3 Masonry Bearing 2-32

2.5 Fatigue 2-33

Table 2.3.3.3 Formulas for the Maximum Moment Capacity of Beams 2-34

1. Doubly Symmetric Wide-Flange Beams Loaded in the Plane of

Symmetry 2-34

2. Channels Loaded Through the Shear Center Plane and Bent About

the Major Axis 2-36

3. Doubly Symmetric Wide-Flange Beams and Channels Bent About

the Minor Axis 2-36

4. Singly Symmetric Wide-Flange Shapes Loaded in the Plane

of Symmetry 2-36

5. Tee-Shaped and Double-Angle Beams Loaded Through the Plane

of Symmetry 2-37

6. Solid Symmetric Shapes 2-37

7. Symmetric Box Shapes Loaded in the Plane of Symmetry 2-38


iv

8. Doubly and Singly Symmetric Hybrid Beams 2-39

Notation for Use with Table 2.3.3.3 2-41

References 2-45

Nomenclature 2-48

Part 2 Commentary C-1

Sec. C.1 General Provisions C-1

C1.1 Scope C-1

C1.2 Definition of LRFD C-2

C1.2.1 Limit State: Strength C-7

C1.2.2 Limit State: Serviceability C-8

C1.3 Loads and Load Combinations C-11

C1.3.1 Load Types C-13

C1.3.2 Load Combinations C-20

C1.3.3 Load Factors C-21

Table C1.3.1-1 Mean Ground Snow Load Intensities for Various

U. S. Cities C-22

Fig. C1.2-1 Frequency Distributions of Load Effect Q and

Resistance R C-26

Fig. C1.2-2 Definition of Safety Index C-27

Fig. C1.2-3 Variation of for Beams C-28

Fig. C1.2-4 Variation of for Columns C-29

Fig. C1.3-1 Magnification Factors for Maximum Mid-Span Bending

Moments for Simply Supported Primary and Secondary

Roof Members Under Ponding Loads C-30

Fig. C1.3-2a Mean of the Logarithms of the Water Equivalent of

Ground Snow C-33

Fig. C1.3-2b Standard Deviation of the Logarithms of the Water

Equivalent of Ground Snow C-33


v

Sec. C.2 Design Criteria for the Limit State of Strength C-34

C2.1 Types of Structures C-34

C2.1.1 Material C-34

C2.1.2 Framing C-34

C2.2 Structural Analysis C-35

C2.3 Design of Members C-37

C2.3.1 Tension Members C-39

C2.3.2 Compression Members C-40

C2.3.2.1 Factored Maximum Strength C-40

C2.3.2.2 Effective Length Factor C-40

C2.3.2.3 Flexural-Torsional Buckling C-41

C2.3.2.4 Tapered Members C-43

C2.3.3 Flexural Members C-43

C2.3.3.2 Factored Maximum Strength of Webs in Shear C-44

C2.3.3.3 Factored Maximum Moment Capacity C-45

C2.3.3.3.3 Maximum Moment Capacity for Composite Beams C-49

C2.3.4 Members Under Combined Flexure and Axial Force C-49

C2.3.5 Members Under Combined Stress C-51

C2.4 The Design of Connections C-51

Table C2.3.2.1-1 Factored Column Stress Ratios C-52

Table C2.3.3.3-1 Coefficients X1 and X2 for Eq. A-2.3.3.3-6 C-55


2-5

2.3.3 Flexural Members


2.3.3.1 Scope
This section concerns the design of singly or doubly symmetric beam
and girder type members which are loaded in the plane of symmetry, and of
channel section beams loaded in a plane passing through the shear center
parallel to the web*** .

Flexural members are subjected to shear force and bending moment.


Design for the limit state of shear capacity is treated in Sec. 2.3.3.2,
while design for the limit state of bending moment capacity is considered
in Sec. 2.3.3.3. In plate girders it is necessary to consider interaction
between shear force and bending moment for certain combinations of the two

* See Commentary Sec. C.2.3.2 for methods of computing Fcre for flexural-
torsional buckling and for tapered or stepped members.
** See Commentary Sec. C.2.3.2
*** Unsymmetric section beams, and beams subjected to biaxial bending and/or
torsion are treated in Sec. 2.3.5. Members under combined bending and
axial force are considered in Sec. 2.3.4.
2-6

effects, and the requirements are given in Sec. 2.3.3.2.4.

2.3.3.2 Factored Maximum Strength of Webs in Shear

The maximum strength of singly or doubly symmetric members subjected

to a shear force in the plane of symmetry is provided by the ultimate

shear capacity of the web (or webs in case of multiple web members). The

factored maximum strength of webs in shear is , where the shear

resistance factor and the nominal maximum shear strength

(2.3.3.2-1)

is given in Sec. 2.3.3.2.1 for beams (no transverse stiffeners) and in

Sec. 2.3.3.2.2 for plate girders (transverse stiffeners required).

Webs of composite beams must be able to support the total vertical

factored design shear on the section.

2.3.3.2.1 Factored Maximum Strength of Beam Webs in Shear

(2.3.3.2-2)

where Fyw = specified yield stress of the steel in the web

Aw = web area

No transverse stiffeners are required, and no interaction check for

combined flexure and shear is necessary. Equation 2.3.3.2-2 applies if

, where h is the web height and t is its thickness.


2-10

2.3.3.3 Factored Maximum Moment Capacity

The factored maximum moment capacity of singly and doubly symmetric

beams and plate girders is , where the resistance factor and the

nominal resistance

(2.3.3.3-1)

is given for beams, plate-girders and composite beams in the following

section.

2.3.3.3.1 Maximum Moment Capacity for Beams

This section applies to

1a) Doubly or singly symmetric wide-flange beams loaded in the plane

of symmetry;

1b) Doubly or singly symmetric box-beams loaded in the plane of

symmetry;

1c) Doubly or singly symmetric hybrid wide-flange beams loaded in

the plane of symmetry

1d) Channels loaded through the shear center plane and bent about

the major axis;

provided that for these sections the web slenderness

and to

2) Symmetric wide-flange beams and channels bent about their minor

axis.

3) Doubly symmetric solid sections (solid round, square or

rectangular bars, etc.).

The resistance factor = 0.86 for these sections.


2-11

The maximum moment capacity for these sections is determined by the

following formulas:
(2.3.3.3-2)

(2.3.3.3-3)

(2.3.3.3-4)

where Mp = plastic moment

Mr = moment at elastic limit, including the effect of residual

stress

S = elastic section modulus

Fcr = elastic buckling stress

= slenderness parameter defined as

1) Lb/ry, the minor axis slenderness-ratio of the laterally unsup-

ported length Lb for the limit state of lateral-torsional buckling (LTB)

2) the flange-plate width-thickness ratio when the limit state is

flange local buckling (FLB)

3) the web-plate width-thickness ratio when the limit state is web

local buckling (WLB)

= slenderness parameter up to which the maximum moment

capacity is equal to Mp

= slenderness parameter below which elastic buckling no

longer will take place.

Mu must be determined for all appropriate limit states (LTB, FLB, WLB),

and the smallest Mu controls. Table 2.3.3.3 gives the relevant formulas

for the appropriate cross-sections.


2-34

Table 2.3.3.3 Formulas for the Maximum Moment Capacity of Beams

This table gives the formulas for determining the maximum moment

capacities for the beam sections in Sec. 2.3.3.3-1. The factored maximum

moment capacity is Mu, where

(2.3.3.3-2)

(2.3.3.3-3)

(2.3.3.3-4)

1) Doubly Symmetric Wide-Flange Beams Loaded in the Plane of Symmetry

(A-2.3.3.3-1)

Limit State LTB

(A-2.3.3.3-2)

(A-2.3.3.3-3)

(A-2.3.3.3-4)

(A-2.3.3.3-5)

(A-2.3.3.3-6)

is determined from Eq. A-2.3.3.3-6 by setting Fcr = Fy - 10 and

solving for .
2-35

Limit State: FLB

(A-2.3.3.3-7)

(A-2.3.3.3-8)

(A-2.3.3.3-9)

in indeterminate beams if the moments are determined by plastic analysis

(A-2.3.3.3-10)

in indeterminate beams if the moments are determined by elastic analysis

and in determinate beams.

(A-2.3.3.3-11)

(A-2.3.3.3-12)

(A-2.3.3.3-13)

Limit State: WLB

(A-2.3.3.3-14)

(A-2.3.3.3-15)

(A-2.3.3.3-16)

in indeterminate beams if the moments are determined by plastic analysis


2-36

(A-2.3.3.3-17)

in indeterminate beams if the moments are determined by elastic analysis

and in determinate beams;

(A-2.3.3.3-18)

When , the plate-girder formulas in Sec. 2.3.3.3.2 must be used.


For tapered members with a single web taper determine Mu as 5/3 of the
allowable moment obtained from Appendix D of the AISC Specification.

2. Channels Loaded Through the Shear Center Plane and Bent About the Major
Axis.

All the same formulas apply as for the doubly symmetric wide-flange
shape except that for the limit-state LTB.
3. Doubly Symmetric Wide-Flange Beams and Channels Bent About the Minor
Axis.
(A-2.3.3.3-19)

(A-2.3.3.3-20)

Limit-states LTB and WLB do not apply, i.e., Mu = Mp. For the limit-
state FLB, S ( F c r ) b = S y ( F c r ) b , and Eqs. A-2.3.3.3-9, 10, 12 and 13 apply
in calculating Mu from Eqs. 2.3.3.3-2 through 4.

4. Singly Symmetric Wide-Flange Shapes Loaded in the Plane of Symmetry.


All equations given for the doubly symmetric wide-flange shapes apply,
except that is to be used for the radius of gyration of the
compression flange in computing for the limit-state LTB (Eq. A-2.3.3.3-2)
and F c r for the limit-state LTB (Eq. A-2.3.3.3-6) is to be determined by
analysis*.

* Formulas are provided in Chap. 6 in the Column Research Council Guide


(Ref. 17) or in Refs. 27 through 29.
2-37

5. Tee-Shaped and Double-Angle Beams Loaded Through the Plane of Symmetry.

(A-2.3.3.3-21)

(A-2.3.3.3-22)

Limit State: LTB

(A-2.3.3.3-23)

(A-2.3.3.3-24)

where + applies when the flange is in compression and - applies when the

flange is in tension is determined by setting Fcr = Fy and solving for

from Eq. A-2.3.3.3-24).

Limit States: FLB and WLB


(A-2.3.3.3-25)

where Q is determined by Appendix C of the AISC Specification if b f /t f of

the flange, when it is in compression, or d/t of the web, when the flange

is in tension, exceeds the limiting ratios of Sec. 1.9 of the AISC Speci-

fication. Otherwise Q = 1.0.

6. Solid Symmetric Shapes

(A-2.3.3.3-26)

Limit states FLB and WLB do no apply, nor does LTB except for rectangular

bars bent about their major axis; for these sections

(A-2.3.3.3-27)

(A-2.3.3.3-28)
2-38

(A-2.3.3.3-29)

(A-2.3.3.3-30)

(A-2.3.3.3-31)

(A-2.3.3.3-32)

7. Symmetric Box-Shapes Loaded in the Plane of Symmetry.

(A-2.3.3.3-33)

Limit State: LTB

Applies only if d > bf

(A-2.3.3.3-34)

(A-2.3.3.3-35)

(A-2.3.3.3-36)

(A-2.3.3.3-37)

(A-2.3.3.3-38)

Limit State: FLB

(A-2.3.3.3-39)

(A-2.3.3.3-40)
2-39

(A-2.3.3.3-41)

(A-2.3.3.3-42)

(A-2.3.3.3-43)

where (Sx)eff is an effective section modulus determined for a section with

a reduced compression flange width beff if beff < bf , where

(A-2.3.3.3-44)

Limit State: WLB


Use the same formulas as those given for the web of the symmetric wide-
flange shape.
8. Doubly and Singly Symmetric Hybrid Beams.

(A-2.3.3.3-45)

where (A-2.3.3.3-46)

Limit State: LTB

(A-2.3.3.3-47)

(A-2.3.3.3-48)
2-40

(A-2.3.3.3-49)

where (A-2.3.3.3-50)

(A-2.3.3.3-51)

(A-2.3.3.3-52)

Limit State: FLB

of compression flange (A-2.3.3.3-53)

is same as Eq. A-2.3.3.3-49,

(A-2.3.3.3-54)

(A-2.3.3.3-55)

(A-2.3.3.3-56)

Limit State: WLB

is determined by Eqs. A-2.3.3.3-16


2-41

(A-2.3.3.3-58)

(A-2.3.3.3-59)

When , Sec. 2.3.3.3.2 must be used for determining the maximum

moment capacity.

The formulas presented herein for RHP and RHE apply for the usual case

when F y f > F yw and (A f t A fc 1.25 A f t .

Notation for Use with Table 2.3.3.3

A = Cross-sectional area; subscripts f, and w refer to

flange, and web, respectively.

A1,A2 = coefficients in Eq. A-2.3.3.3-25

where

ar = ratio of web area to compression flange area, Aw/Afc

b = width of rectangular section

bf = flange width, subscripts fc and ft refer to compression

and tension flange, respectively.

beff = effective flange width of box section (Eq. A-2.3.3.3-44)


2-41a

Cb = equivalent moment factor

where M1 is the smaller and M2 the larger end-moment

in the unbraced segment of the beam; M 1 /M 2 is positive

when the moments cause reverse curvature

Cw = warping constant

d = depth of a section

E = modulus of elasticity (E = 29,000 Ksi)

(Fcr)b = critical elastic buckling stress of beam

Fr = compressive residual stress in flange (Fr = 10 Ksi)

Fy = specified yield stress, subscripts yf and yw refer to

flange and web, respectively

G = shear modulus (G/E = 0.385)

Ix = second moment of area about x-axis

J = torsion constant;

for solid rectangle

for box shape

for singly symmetric


W-shape
Lb = unbraced length

M/Mp = ratio of moment M at the end of the unbraced section of

a beam to the plastic moment at the other end; M/Mp is

positive when the moments cause reverse curvature.


2-42

Mp = plastic moment

Mr = moment when yielding commences

Mu = maximum moment capacity

m = ratio of web to flange yield stress in hybrid beams

PD = factored design axial load

Py = Ag Fy, where Ag is the gross area

Q = plate buckling reduction factor from Appendix C of

the AISC Specification

RHE,RHP = coefficients defined by Eqs. A-2.3.3.3-50 and 46,

respectively

rT = radius of gyration of compression flange plus one-

sixth of the web

ry = minor axis radius of gyration

S = elastic section modulus, subscripts x and y refer

to major and minor axis, respectively

(Sx)eff = effective section modulus for box shapes

t = web thickness; twice the angle thickness for double

angles

tf = flange thickness; fc and ft refer to compression

and tension flange, respectively


X1, X2 = coefficients in Eq. A-2.3.3.3-6; X1 and X2 are

tabulated for all rolled shapes in the AISC Manual

in Table C-2.3.3.3-1.

y = distance from the top of the flange to the centroid

for Tee and double-angle shapes


2-43

= distance from bottom of tension flange to centroid

for hybrid W-shapes


Z = plastic section modulus, subscripts x and y refer to

major and minor axis, respectively.


= the ratio to d for hybrid shapes.
C-55

TABLE C-2.3.3.3-1: Coefficients X1 and X2 for Eq. A-2.3.3.3-6

SHAPE X1 X2

W36x300 3,834 10,820


280 3,613 12,250
260 3,350 14,270
245 3,163 16,040
230 2,972 18,020
W36x194 3,036 20,390
182 2,854 23,050
170 2,678 26,085
160 2,520 29,790
150 2,370 34,160
135 2,149 43,830
W33x240 3,534 12,740
220 3,257 15,060
200 2,970 18,148

W33x152 2,736 24,370


141 2,535 28,760
130 2,339 34,630
118 2,143 43,260
W30x210 3,647 11,760
190 3,318 14,170
172 3,009 17,180

W30x132 2,893 22,020


124 2,719 25,014
116 2,548 28,960
108 2,380 34,280
99 2,196 41,410
W27x177 3,707 11,360
160 3,358 13,700
145 3,060 16,400
W27x114 2,963 20,310
102 2,662 25,130
94 2,467 29,930
84 2,182 38,670
W24x160 3,806 10,180
145 3,459 12,360
130 3,103 15,440
W24x120 3,224 14,900
110 2,965 17,540
100 2,710 20,940
C-56

TABLE C-2.3.3.3-1 (CONTINUED)

SHAPE X1 X2

W24x94 3,079 18,610


84 2,751 23,140
76 2,498 28,690
68 2,253 36,860

W24x61 2,378 37,180


55 2,168 42,770

W21x142 4,283 8,050


127 3,849 9,833
112 3,405 12,480

W21x96 3,875 11,370


82 3,327 15,270

W21x73 3,017 19,174


68 2,826 21,870
62 2,576 26,610
55 2,303 34,430

W21x49 2,400 34,850


44 2,189 44,300

W18x114 4,494 7,192


105 4,160 8,394
96 3,799 9,904

W18x85 4,222 8,828


77 3,869 10,500
70 3,519 12,550
64 3,233 14,840

W18x60 3,244 16,110


55 2,979 19,130
50 2,720 22,900
45 2,451 28,680

W18x40 2,552 28,260


35 2,240 38,580

W16x96 4,488 6,998


88 4,140 8,206

W16x78 4,645 7,310


71 4,256 8,664
64 3,837 10,620
58 3,484 12,810
C-57

TABLE C-2.3.3.3-1 (CONTINUED)

SHAPE X1 X2

W16x50 3,296 15,230


45 2,996 18,580
40 2,672 22,990
36 2,405 29,650

W16x31 2,465 29,900


26 2,088 44,120

W14x730 24,660 292.0


665 23,030 329.8
605 21,380 369.8
550 19,960 418.9
500 18,500 476.8
455 17,180 543.9

W14x426 16,240 595.1


398 15,350 658.5
370 14,460 734.8
342 13,560 828.3
314 12,550 938.9
287 11,610 1,085
264 10,770 1,236
246 10,110 1,386

W14x237 9,775 1,474


228 9,433 1,567
219 9,078 1,679
211 8,795 1,788
202 8,436 1,919
193 8,088 2,071
184 7,732 2,243
176 7,420 2,429
167 7,084 2,656
158 6,728 2,915
150 6,392 3,201
142 6,067 3,534
W14x320 13,020 953.4

W14x136 6,081 3,557


127 5,694 3,996
119 5,367 4,484
111 5,023 5,104
103 4,655 5,824
95 4,305 6,745
87 3,976 7,901
C-58

TABLE C-2.3.3.3-1 (CONTINUED)

SHAPE X1 X2

W14x84 4,504 6,447


78 4,194 7,422

W14x74 4,630 6,506


68 4,258 7,580
61 3,839 9,220

W14x53 3,987 9,138


48 3,628 10,960
43 3,280 13,300

W14x38 3,086 16,700


34 2,770 20,740
30 2,458 27,340

W14x26 2,658 25,130


22 2,273 35,830

W12x190 11,240 1,171


161 9,698 1,525
133 8,103 2,865
120 7,400 2,531
106 6,570 3,108
99 6,165 3,503
92 5,771 3,999
85 5,338 4,572
79 4,993 5,236
72 4,577 6,196
65 4,155 7,410

W12x58 4,337 6,960


53 3,982

W12x50 4,482 8,293


45 4,054 7,020
40 3,658 8,452

W12x36 3,645 10,320


31 3,166 11,200
27 2,761 14,750

W12x22 3,071 19,400


19 2,670 20,060
16.5 2,372 27,610
14 2,036 38,350
C-59

TABLE C-2.3.3.3-1 (CONTINUED)

SHAPE X1 X2

W10x112 9,966 1,443


100 8,994 1,745
89 8,074 2,118
77 7,071 2,708
72 6,636 3,061
66 6,109 3,527
60 5,593 4,170
54 5,062 5,000
49 4,615 5,962

W10x45 5,123 5,092


39 4,476 6,724
33 3,833 9,314

W10x29 4,081 8,931


25 3,534 11,780
21 3,000 17,290

W10x19 3,430 15,050


17 3,069 19,780
15 2,774 25,730
11.5 2,194 41,426

W8x67 9,335 1,627


58 8,191 2,082
48 6,879 2,817
40 5,789 3,995
35 5,112 5,000
31 4,553 6,267

W8x28 4,872 5,711


24 4,229 7,484
W8x20 3,995 9,283
17 3,441 12,920

W8x15 3,718 12,720


13 3,296 16,930
10 2,555 26,750
W6x25 4,524 3,560
20 5,043 5,240
15,5 4,022 8,647
W6x16 5,673 4,657
12 4,401 8,251
8.5 3,203 15,480
C-60

TABLE C-2.3.3.3-1 (CONTINUED)

SHAPE X1 X2

W5x18.5 7,198 2,606


16 6,295 3,378

W4x13 7,955 2,314

M14x17.2 1,997 55,450

M12x11.8 2,020 58,180

M10x29.1 4,764 8,419


22.9 3,639 11,330

M10x9 2,154 49,370

M8x34.3 5,336 4,928


32.6 5,054 5,188

M8x22.5 5,194 6,613


18.5 4,193 8,390

M8x6.5 2,337 39,510

M7x5.5 2,562 32,190

M6x22.5 6,587 3,482


20 5,732 4,043

M6x4.4 2,675 28,090

M5x18.9 8,111 2,178

M4x13.8 9,767 1,586


13 9,179 1,692
S24x120 4,805 9,124
105.9 4,308 10,050

S24x100 4,255 13,210


90 3,828 15,020
79.9 3,466 16,440
S20x95 5,394 7,697
85 4,788 8,801
S20x75 4,453 11,360
65.4 3,927 12,980
S18x70 5,074 9,531
54.7 3,906 13,040
C-61

TABLE C-2.3.3.3-1 (CONTINUED)

SHAPE X1 X2

S15x50 4,870 9,187


42.9 4,178 10,840

S12x50 7,165 4,297


40.8 5,722 5,620

S12x35 4,936 8,022


31.8 4,508 8,728

S10x35 7,009 4,627


25.4 4,858 7,157

S8x23 6,739 4,518


x18.4 5,293 5,921

S7x20 7,608 3,649


15.3 5,607 5,225

S6x17.25 8,876 2,765


12.5 6,023 4,455

S5x14.75 10,990 1,881


10 6,652 3,611

S4x9.5 9,649 2,048


7.7 7,553 2,727

S3x7.5 13,000 1,168


5.7 9,121 1,826

HP14-X117 5,427 4,954


102 4,788 6,346
89 4,202 8,136
73 3,458 11,710

HP12x74 4,870 6,235


53 3,548 11,386

HP10x57 5,517 4,901


42 4,131 8,479

HP8x36 5,422 5,035


C-62

TABLE C-2.3.3.3-1 (CONTINUED)

SHAPE X1 X2

C15x50 6,570 7,293


40 5,046 10,666
33.9 4,269 13,010

C12x30 5,783 8,640


25 4,677 11,510
20.7 3,952 13,910

C10x30 8,958 4,238


25 6,995 6,244
20 5,277 9,316
15.3 4,076 12,474

C9x20 6,650 6,477


15 4,803 9,971
13.4 4,352 11,020

C8x18.75 7,965 4,690


13.75 5,441 8,018
11.5 4,621 9,576

C7x14.75 7,827 4,557


12.25 6,210 6,307
9.8 4,981 8,095

C6x13 9,364 3,259


10.5 7,108 4,838
8.2 5,476 6,563

C5x 9 8,518 3,399


6.7 6,202 5,028

C4x 7.25 10,320 2,284


5.4 7,387 3,495
C3x 6 14,683 1,149
5 11,460 1,629
4.1 9,289 2,136

You might also like