----- - I

Unsymmetrical Plate Girders I

L HIG U I IT

UNSYMMETRICAL
PLATE GIRDERS
UNDER SHEAR AND MOMENT

by
Chingmiin Chern
and
Alexis Osta pen ko

October 1970--

Fritz Engineering Laboratory Report No. 328.9

 --1

<I. ,\

Unsymmetrical Plate Girders

UNSYMMETRICAL PLATE GIRDERS UNDER SHEAR AND MOMENT

by
Chingmiin Chern
and
Alexis Ostapenko

This work was conducted as part of the project

Unsymmetrical Plate Girders, sponsored by the American
Iron and Steel Institute, the Pennsylvania Department
of Transportation, the Federal Highway Administration
of the U.S. Department of Transportation, and the
Welding Research Council. The findings and conclusions
expressed in this report are those of the authors, and
not necessarily those of the sponsors.

Department of Civil Engineering

Fritz Engineering L'aboratory:.
Le~igh University .
Bethlehem, Pennsylvania
October 1970

Fritz ~ngineer1ng Laboratory Report No. 328.9

 TABLE OF CONTENTS

Page No.

ABSTRACT . . . .................... 1

1 c INTRODUCTION......... ~ 2

ANALYTICAL MODEL ANO INTERNAL FORCES .............. 6

3. ULTIMATE STRENGTH. (I
14

40 COMPARISON WITH TEST RESULTS . . . . . . . 23

CONCLUSIONS .......................... 26

6 '" ACKNOWLEDGEMENTS. 27

7. APPENDIX I. - REFERENCES . . . . . . . 28

8", APPENDIX II. - NOTATION .................... 31

9 Q T.i\BLES AND FIGURES " 35

 UNSYMMETRICAL PLATE GIRDERS UNDER SHEAR AND MOMENT

by

Chingmiin Chern l

and

Alexis Ostapenko 2

ABSTRACT

A method of determining the ultimate static strength of

transversely stiffened plate girders subjected to a combination of

shear and bending is presented. The method is applicable to homo-

geneous and hybrid girders with symmetrical or unsymmetrical cross

section. The ultimate strength is assumed to be given by the sum

of,three contributions: beam action, tension field action and frame

The behavior of a girder panel is described by a continuous inter-

action curve which is divided into three parts: web failure portion,

compression flange buckling portion and tension flange yield portion,

The theoretical ultimate loads compare well with the results of th,e

available fifty-three tests on symmetrical and unsymmetrical plate

girders. The average deviation is 5% with the extreme deviation

of 15%1l

lAssistant Professor of Civil Engineering, North Dakota State University,

Fargo, N.D., formerly Research Assistant at Lehigh University, Bethlehem,
Pennsylvania
2
Professor of Civil Engineering, Fritz Engineering Laboratory Department
of Civil Engineering, Lehigh University, Bethlehem, Pennsylv~nia
(1)
 l

328.9 ~2

1. INTRODUCTION

A plate girder in a building or in a bridge has a majority of its

panels subjected to Bome combination of shear and moment; only a few panels

would ordinarily be under pure moment or shear. Yet, most of the theoretical

and experimental research conducted so far on the ultimate strength of plate

gi"rdera has dealt with the simpler cases of pure moment or shear, and the

case of combined loads '* has been treated due to its complexity as some

plausible transition between these 'two strengths.

In 1961, Basler sugges~~d that the moment capacity of a piate

girder section be given by the yield strength of the flanges plus the yield

capacity of the web reduced by a shear stress assumed to be uniform. The

effect of web buckling was neglected and the approach is thus valid only

for webs with very low depth-thickness ratios. To overcome this difficulty

it was proposed that the shear capacity V (including the post-buckling

u
strength) be not reduced by bending up to M = 0.75 of the pure bending moment

M , causing yielding of the tension flange, and the moment capacity M be

y y
not reduced by shear up to V = 0.60 of the pure shear strength V.
u
The

transition between the two result~ng points was assumed to be a straight

line. If applicable, the moment from this interaction diagram should be

reduced to M which is primarily controlled by the strength of the compression

u
flange. (4) This interaction relationship was then adopted by the AISC

*Hereafter, whenever there is no possibility of confusion, "combination of

moment and shear" will be called "combined loads."
 328.9

Specification. (2) Reference 24 introduced a modification of the method by

replacing the tension flange yield moment M with the ultimate moment for
.Y
pure bending M .
u

In 1968, Akita and Fujii arrived at another interaction relation-

" cons1sting
Sh 1p " h'
0 f tree "h t I"1nes. (1) Teen
stra1g h d pOlnts
" were assume d to

be given by the ultimate strengths for pure shear V and pure moment M '
u v
One of the intermediate two points was defined by the shear causing pure

shear buckling of the web and by the moment produced by the yield~d flanges.

The other point was given by the ultimate shear of the web computed assuming

the flanges to be of zero rigidit~ and the moment produced by the flanges

yielded under the axial forces due to bending and the tension field action.

The method neglects the possibility that the compression flange may buckle

laterally.

In all the above described studies, only girder panels with

symmetrical cross sections were considered. However, in many practical

plate girders the cross section is unsymmetrical, that is, the centroidal

axis is not at the mid-depth of the web; most typically, this is the case

for composite and orthotropic deck girders. So far, the only consideration

given to unsymmetrical girders is an adaptation of Basler's interaction

relationship(4) in Reference 24.

The purpose of the present study is to describe a new method which

gives the ultimate strength of a plate girder panel directly for any com-

bination of moment and shear and is applicable to unsymmetrical, symmetrical,

homogeneous, and hybrid girders. The analytical model of the method and the

assumed pattern of the girder behavior are given next.

 Due to the complexity of the force interletlan in I plate girder
panel, an exact analysis of its behavior under load has been impossible,

and recourse had to be taken to represent the panel in the form of a model

as closely to the true state as possible and formulating the desired strength

equations on it. Defficiencies of the analytical models employed by previous

researchers have been pointed out. The model proposed here, although not

perfect, provides a means for explaining cases which could not be handled
before.

The model for combined loads represents an interaction between the

models which have been developed in the course of this research for the
(7) . (8)
case$of pure shear and pure bending. The web plate is assumed to be

flat until it buckles under the combined effect of increasing stresses due

to shear and moment. The Post-buckling strength of the web is aSSumed to

be in the form of the tension field action analogous to, but not the same

as, in References 1 and 3. The Contribution of it is limited by the yielding

of the web plate and both, shear and moment, are taken into accOunt.

The flanges together with an effective area of the web contribute

to the shear strength by forming a plastic mechanism with hinges at the

stiffeners (Frame Action). The etfect of the axial forces in the flanges

is included. The axial strength of the flanges in yielding or buckling

(Ia teral or torsional) controls the magni tude of the moment on:i:he panel.

The horizontal component of the tension field force reduces the flange

capacity available to carry the moment. When the flange bUckling capacity

is reached before the full capacity of the web or frame action is developed,

the reduction of the moment due to their presence is proportionately smaller.

 328.9
-s

The modes of behavior described above are equally valid whether

the larger portion of the web is in tension or compression, except that

a full plastification of the web may be possible as the web portion under

compression becomes smaller. Depending on whether or not the flanges fail

before the shear capacity of the panel is developed, two types of interaction

between shear and moment are possible; shear capacity r"educed(,by moment, and

moment capacity reduced by shear. One or the other will control the design.

A complete interaction relationship is thus established .

. Since for a given pattern of loading, both shear and moment~ are

directly proportional to the intensity of loads. it is convenient to visualize.

any panel as if it were a panel in a simply supported beam under a mid-

span concentrated load as shown in Fig. 1. Then the maximum moment in the

p.anel 1.8 Mmax = Vx and the mid-panel moment is M = V(x - ~a) for "any intensity

of loading. ihie ttieali...za-tion- is --emp1-oyedlat-e:r-=.in- the-,le-rivat:iQ'Q o-f tbQ

,su@Flgtb e tim"rst h' eJ yew Ct,,,,,-f". 0 -K-r00 UuPL

a-ri
VV\,rorv"..e/V\trL tJ. ,. Jt~J1 -~~ i- aNi e..f ""I'll} h.e
fli\JLr"-~ - f'1 VI eat
A-eTI
'1 rV\ fe.1'.nt';>

o~ th~ Sh,~A,V' SpCtM. V'Q,ho /"'~

(8)
 328.9 -6

2. ANALYTICAL MODEL AND INTERNAL FORCES

The static ultimate ~tren8th of a girder panel under combined

shear and bending expressed by means of shear force is assumed to be

given by a sum of the following three contributions: (a) the beam action

shear V . (b)
'fC'
the tension field action shear V ; and (c) the frame
crc
action shear Vfc.

(1)

Each of the shears in Eq.l is computed considering the effect of the bend-

ing moment. *

Beam Action Strength . - Figure 2 shows a rectangular web pl~te

subjected to combined loads. The top fiber compression stress is positive,

and shear 1s positive when it acts down on the right hand face. The web

stresses at the buckling load may be evaluated with sufficient accuracy by

means of the following interaction equation: (19)

)2
( T:
:r
r
(2)

* Theeffect of the moment is not considered for the pure shear case de-
scribed in Reference 7. Subscript lie" in E'll.l and subsequent equations
designates "combined" to differentiate from~ analogous notation in Ref'er-
ences 7., and ~
 where

t
c ~ shear bucklins stress under combined loads

L cr = kv 12TI~l~V2)(~)2, buckling stress under pure shear

with the shear buckling coefficient
2.31 ~)A4 8.:)C}
o Lv = 5 ;J +;5 - 1./ _+,!..4<10 ex (3a)

for ex < 1.0

or (3b)

for ex > 1.0

ex = alb, panel aspect ratio, or the ratio of the panel length

to panel depth

abc bending buckling stress at the extreme compression

fiber ~hen the plate is subje~ted to combined load~

a cp .. ~ 12 ~2(l-v
E )
2
(-bt)~ , buckling stress at the extreme com-

pression fib~r of the web under pure bending

with the bending buckling coefficient ~ taken at its minimum value neglect-

ing the effect of ex (or setting ex 00)

-7' ~ = l3~54 - 15.64 C + 13.32 C2 + 3.38 C3 (4)

for -1.5 < C < 0.5

C ratio of maximum tensile stress to maximum compressive

s tress for the combined loading case as' shown ~n Fig'.2.

Then, the shear force carried by the web at the web buckling load is
VTC := TcAw L
C
bt ~ (5)
 328.9
-8

As a simplification of further analysis it is assumed that the

stress is const~h~!~~~~~_~!l!_!~~~!~_~~~~~~_
E the=we"b=,~slQ~!~~~L~~ Then, at the web buckling .load the stress in the
compression flange is

(6a)

and the m~nitude of the stress in the tension flange is

(6b)

Tension Field Action Strength. - In evaluating the ultimate

st engt of t e wee, thr e additiona as umpti nsreade (a) ~thee

buckling stresses remain constant after the plate buckles, (b) the linearly
- - - - - - ~_===__ _=-==_~~~~,~=~~!!:~~;;::-~-..;::~;;;~'~'Y.'"N~~~:r........
=_. """'---- .. ~.
""=:c.-.:--=_ =

values (this assumption introduces only a negligible inconsistency between

the total external and internal moments), and (c) the ultimate strength is
~ ~ - : t o

attained when the combination

=="'"" ..,. . .,
of the shear buckling stress under combined
~=n:.:==.
~~ ~~~~,-:;r.-~~:::::-o:=~~~,~~~~~.~..;'~~,~~_,;,:",~~.~,":,:,,:"~~~"'~~i;;:~J~;:.~;:r-;:'.'.'r.";
"~."':~::"".',~r~,7::::'~'-"''';:'$::-:;:-::~7/'''/::~~-::Y'::I'~~'_7.~~"7.''''
.'~)..~~

ave~age tensile stresses and the tension field stress

reaches the yiel~ ~~l.1.?ition. The idealized tension field model under com-

bined loads is shown in Fig.3 where angle c indicates the inclination of

the tension field.

The process of deriving the equation for the tension field action

shear V starts by transforming the stat~~ of stress defined by T and

crc c
abc to the coordinate axes inclined the yet Unknown angle
~. The normal stress in the direction of is then increased by a and
C c tc
the von Mises yield condition is imposed on the resulting combined state

of stress. The unkowns in the thus derived Eq. 7 are a and . ..Ns-te
tc c

(7)

where

s = - ~ Ob~J~<1~1
[

R = vt 2 + T :'
c

0=-
2
1
tan
-1'( s .)
"2-t.",,,:"
e'

of t9'
/

!I.
b;
///x - a/2
, = - _..........
b
 328.9 -10

The tension field shear is found from the stress diagram in

Fig o 4b to be

vac =!2 A
w
atc [Sin 2~ c (1 - p) a + (1 -~) a Cos 2 ~cJ (9)

As shown in Reference 6, a conservative estimate of the parameter

p for a plate girder is

p = 0.5 (10)

With parameter p being assumed constant and the panel having the

given geometrical and material properties, cr is a function of ~ only.

tc c
Equation 1 thus can be rewritten as

vcrc = V (rh)
qc 't'c

in which ep is the only variable. The maximum value of V is obtained

c Oc
by setting the derivative of the function with respect to ~equal to
c
zero,

This gives the following expression for . , the value of for which
co c
Vac is a maximum:
 328.9
-11

aa
[Sin 2~ + (1 - p) a Cos 2~ ] tc
co co a ~co

+ 2 [Cos 2~ co '- (1 - p)
-
a Sin 2~ co ] ate = 0 (11)

Equation 11 can be solved for by iteration~ (An iteration method is

co
explained in the Appendix of Ref.7). Substituting ~co back into Eqs. 7

and 9, the maximum tension field contribution becomes

1
=- A
2 w
atc [Sin 2 ~co - (1 - p) a + (1 - p) a Cos 2 ~coJ (12)

where

[1:.2 (~)
2
+ 1-
2
R Sin (2 ~ co + 20)J (13)

Flange stresses under the optimum tension field shear are shown in
(j)
Fig. 6b. There are two contributions: one comes from t~e hor_izontal.com-:-
(p
fonent oU.J:u,unsfon .Uej !: :., nd t2Lyth r=2~S"! n!?\Y2t :
1ibrium for the tension f~d action shear. Thus, the stress in the com-

pression flange is

V a Aw
02c = ~ Oc + tc [1 + Cos 2~ - (1 - p) a Sin 2~ J
Afe 4 Afc co co
328.9 -12

and the stress in the tension flange is

= ~ Vae _ ate ~ [1 + Cos 2~ - (1 - p) a Sin 2~ ]

Aft 4 Af~ co co

Let

1
H=-cr Aw [1 + Cos 2 - (1 - p) a Sin 2$ ] (14)
2 tc co co

represent the horizontal component of the tension field force which is

equally carried by the top and bottom flanges.

0Ze = _1_ [11 V

A oc
+!2 H]
(15a)
fc

[11 V
crc
_!2 HJ
(ISb)

Frame Action Strength. - Analogously to the frame action strength

developed in Ref. 7 for the case of pur~ shear, it is ,assumed for the case

of combined loads that the maximum shear contribution of the flanges is

reached when the plastic hinges form at both ends of the flanges to develop

a panel mechanism shown in Fig.5.

(16)

where mel and mer are, respectively, the plastic moments of the compression

flange at the left and right sides of' the panel modified for the effect of
 --~-------- -----------------------

328,9 ~13

the axial force in the flange. m land m are the analogous plastic
t tr

moments in the tension flange~ Since the cross section of each flange

is unsymmetrical being assumed to consist of the flange proper and an

effective portion of the web (see Refs. 7 and 8) ,the axial force 1n-

fluences the plastic moments at the left and right sides of the panel

to a different degree and they are, therefore, not equal to each othero

The compression and tension flange stresses, 03c and 03t' are produced by this

action, as shown in Fig . 6'c,

(17a)

(17b)

Reference Stresses in the Flanges. - The following flange stresses

serve as reference stresses to describe the behavior of a girder panel under

combined shear and bending:

(a) aBe the sum of the compression flange stresses

contributed by the beam,tension field, and

frame actions.

(lJ) ast -- the sum of the tension flange stresses can-

tributed by the beam, tension field, and frame

actions.

(c) a -- the maximum stress that the compression flange

cf
can resist. It is taken to be equal to the
 328.9
-14

buckling stress of a col,umn formed by the

compression flange with a portion of the web

P1a t e.
(3,7,8)

(d) a the maximum stress that the tension flange can

yt
resist. If the strain hardening is neglected,

it is equal to the yield stress of the tension

flange.

3. ULTIMATE STRENGTH

The failure of a panel subjected to a combination of shear and

moment, may be due to the failure of the web, buckling of the compression
, _. ~~= ---=~- -~. -", ~ . ,. ~

other mode of failure is determined by comparing reference stresses in

the flanges and selecting the lower as the controlling one. The regions

for each of the three modes of failure are shown in Fig.7 where the she~r

is plotted against moment at the controlling ultimate condition. The

non-dimensionalizing values V and M are, respectively, the ultimate shear

u u
capacity.with the moment equal to zero (Ref.7) and the ultimate moment

capacity with the shear equal to zero (Ref. B). As seen in the figure, the

interaction curve represents a series of stress inequality relationships.

The interaction curve is separated by the ordinate V/V into two portions:
u
the larger portion of the web in compression and the larger portion of the

web in tension, as indicated in the figure by portions Q -Q -Q and Q _

1 2 3 1
Q -Q, respectively.
4 5
 -15

Larger Portion of the Web in Compression (QI-Q2-QS)' - The

ultimate strength equations are formulated from the conditions in

specific portions of the interaction curve.

(a) At point Ql in Fig. 7, the panel is under pure shear, that is, M = O.

The capacity will be limited by the web plate failure and V = V . (7)
th u

(b) Portion between Ql and Q2. The panel is under high shear and relatively

low moment. When the web is stressed to its ultimate capacity under

combined loads, the total stress a introduced to the compression flange

se
is still less than the flange ul~imate carrying capacity a fO At this
c '
stage, the flange is strong enough to resist buckli~g, and the failure

mechanism of the panel is the web plate failure. Therefore, the ultimate

strength equations are

(l8a)

(lOb)

where V ,V ,and Vfc are obtained respectively from Eqs. 5,12 and 16.
Tc crc
328.9

(c) At point Q2 the stress in the compression flange due to the ultimate

shear strength under combined loads is equal to the buckling stress

of the compression flange column, asc = acf' and failure may occur

simultaneously in the compression flange and in the web p~ate.

(d) Portion Q2 - Q3. If the web stress increased to the ultimate shear

strength, the compressive flange stress a would be greater than

sc
0cf' the ultimate stress the compression flange can resist. There-

fore, in this range, the web does not reach its ultimate shear strength,

and the panel fails due to the buckling of th'e, compression flange.

In order to calculate the panel strength, the following assumptions

sian field stress a ~

- . .. - --=--===-=,.~~~"=". '_~._,,,,,,,,,,,,1;,.Q,,~~,,"
The stresses introduced to the compression

flange due to the post-buckling behavior are then

(19a)

strength. (7) When the panel is under the combined loads, the shear
-----~~-

capacity of the flanges will be further reduced because of the axial

force~ acting on the flanges. Thus, the higher is the moment on

~ ,_ .. ~. =~""'~=~"""-''''''_''L",,'''''''' .,.~.,,~
328.9
~17

,~he p ~~:~"}?~r,~~~,c~e.~t:'~~~2yJ;~~2!ly."~Jl=.~,:~2'!~': ~''''2S~.y.~y.~~.Y.~.!.~.~y~. ::

Therefore, no significant error will be introduced if cr is assumed
3c
to be a known quantity given by Eq. 17a.* Thus, the reference flange

stress due to the horizontal component of the tension field force be-

comes

(19b)

Observing that Eqs. l5a and 19b give the same stress 0Zc' the follow-

ing equation is obtained:

from which the horizontal component of the tension field force H' is

found by utilizing Eqs. 12 and 14.

H.. . = S1 V'"
Oc

where

1 + Cos 2 - (1 - p) a Sin 2 ,
co co:
~ = Sin 2~ co - (1 - p) a+(l - p) a Cos 2~
co
(19c)

The incomplete tens,ion field shear Va~ is from the above

*A correcting refinement can be made by varying the frame action

contribution, for example linearly, from the full value of 03c at
point Q2 to zero at point Qa-
328.9 -18

v ,A
(19d)
ac

Then, the ultimate strength equations for a > a (region Q2- Q3)
sc cf
are:

(1ge)

and

Vth (x - ~ a) = ~b Vth (19f)

(g) At point Q3, the panel is under pure bending, that is, V = O,and
th
the failure mechanism will he due to the failure of the compression

flange acting as a column; Mt h = Mu .

Larger Portion of the Web in Tension (Q1-Q4-QS). --

(a) In the portion QI-Q4,(Ost < ayt ), the panel is under high shear

and relatively low moment. The panel behavior is similar to that

of the portion Q1-Q2, described above.

(b) At point Q4, the stress in the tension flange due to the ultimate

web shear strength under combined loads is equal to the tension

flange yield stress (a st = a ) and the panel will fail simultaneously

yt
due to yielding of the tension flange and of the web plate.

(c) In the portion between Q4 and Qs, the panel is under shear and

high moment. The tension flange starts yielding before the web
 328.9 -19

plate reaches its ultimate shear strength. Then, the yielding will
penetrate into the web and finally cause the plastification of the
cross section.

Based on the location of the neutral axis, the derivation of the

ultimate strength equations can be separated into two cases.

loads yields uniformly through the full depth, the average shear

stress, as shown in Fig. 9 , is

T = (20a)

where Vth is the shear force under investigation. Substituting T

from Eq. 20a into the von Mises yield condition, the web stress due

to bending is obtained.

....,

a 3 (0 Vt~
yw W
) 2 (20b)

By setting the sum of the normal forces, acting on the cross section

shown in Fig. 9b, equal to zero, the location of the neutral axis

of a fully yielded cross section is expressed in.terms of nondimensional

parameters nand p *

= 0.5 _ 0.5 (p P p p)
ng P7 1 4
(ZOe)
2 5

* p's and n's are defined in Appendix II (Notation).

328.9
-20

The shear force and the moment which act at the center of the panel
are then

(20d)

and

(20e)

where the plastic shear force of the web, V , is

P

1
V = T A = ~(J A (20f)
p y w y,j yw w

axis of a fully yielded cross section shown in Fig. 10 is

p) 1 (21a)
P2 5 P nh
4

The shear force Vth and the corresponding moment M acting at the
th
center of the panel are

+ P (0.5 + n ) + P P (1 + n + n )]
g (21b)
7 g 2 5 f
 328,9
-21

and

(2lc)

(d) At point Qs, the panel is under pure bending. As shown in Ref. 8,

the failure mechanism will be the plastification of the cross section,

M (22)
P

Yhere Mp is the plastic moment of the cross section.

Maximum Moment in Panel. - Since under combined loads the moment

at one end of the panel is greater than the mid-panel moment for which the

.analysis is performed according to the above described procedure, it may

happen that this maximum panel moment will control the panel strength.

This is especially true for panels with large aspect ratios.

The shear producing the maximum panel moment may not exceed, de-

pending on the case, one of the following values:

M a
V' < u-:i..S: (23a)
th - )lb + ~a 0' cf

p
V' < /3
th - ~ + ~ a
V [
PI
P
4
(
ng
+
n e
) + P (n 2
7 g
_ ng + 0.5>"

(23b)
'328.9 -22

or

+P (0.5 +n ) + P P (1 + n + n )] (23c)
7 g 2 5 g f

for the neutral axis in the web or in the compression flange, respectively.

These two equations (Eqs. 23b and 23c) are simply Eqs. 20d and 21b modified

for an increased shear span (from ~b to ~b + ~ a).

It seems reasonable and sufficiently accurate, mostly on the safe

side, to simplify the maximum panel moment limitation to keeping it below

the moment which would produce yielding according to the ordinary beam

theory.

I (Jyf
V' < --~-~--:-- (24)
th - y b (~ + ~ a)

where 0yf is the yield stress and y is the distance from the centroid to

the flange for either the compression or tension flange, whichever gives

the smaller V and thus controls.

th
328.9
-23

4. COMPARISON WITH TEST RESULTS

The ultimate strength theory was checked against the available

f 1f ty - tree
h tests on
symrnetr~ca
1 pate
1 ders, (1,5,9,10,17,18,21)
glr
(13 16) . (12 23)
hybrid girders, ----, and unsynnnetrical plate glrders. ' Tables 1

to 3 summarize the dimensions of the test panels, material properties,

the experimental ultimate loads, the ultimate loads fr,omthe theoretical

analysis, as well as a comparison between the experimental and predicted

ultimate loads. The test load is shown on the theoretical interaction

curves for each individual panel in Figs. 11 to 24.

Symmetrical girders with homogeneous material properties are

given in Table 1 and the interaction curves for each panel are shown in

Figs. 11 to 16. The mode of failure of all these panels except for one

is classified as the web failure (shear failure). The average deviation

of the available thirty-one test loads is 4% with the maximum deviation

of 12% (G4 in Reference 18).

The comparison of the theory with tests on unsymmetrical plate

girders is made in Table 2 and Figs. 17 to 20. The tests on panels with

the smaller flange and, thus, the larger portion of the web, in compression

and faili~g in the web are shown in Table 2(a). The interaction diagrams

are in Figs. 17 and 18 (UG2.2 and UG4.1). The theory gives an over-

estimate of 1%.

Table 2(b) gives the girder panels with the smaller flange in

compression and subjected to high shear and high moment. The panel strength

for these cases was limited by the failure of the compression flange.
328.9

rhe interaction curves for these tests are shown in Figs~ 17,
18 and 19 where the reduction of the panel bending carrying capacity due

to the effect of shear, is illustrated. An average of 6% underestimate

is obtained for the four tests with the extreme deviation of 10% under~

estimate.

The girder panels with the smaller flange in tension and sub-

jected primarily to shear are given in Table 2(c). The interaction

curves are in Figs. 18 and 19. The mode of failure is classified as

the web failure. An average deviation of 5% is obtained for the avail-

able three tests with the maximum deviation of 12%.

A comparison of the test results for symmetrical hybrid girders

with the predictions of the proposed approach is shown in Table 3. The

interaction curves fo~ the individual tested panels are given in Figs.

21 to 24. An average deviation from the available thirteen tests is 7%

with the maximum deviation of 15%.

A comparison for girder panels with the same geometrical and

material properties, but subjected respectively to pure shear, a com-

bination of shear and moment, and pure moment is shown in Fig. 20. A

good agreement between the test results and the computed values is ob-

served.

The overall average deviation of the available fifty - three

test results (Tables 1 to 3) is 5% with the extreme deviation of 15%

A comparison of some test results for unsymmetrical plate

girders with the proposed approach and with the approaches currently
32809
-25

available (1970) is shown in Figo 250 As in all previous plots, the

ordinate gives the shear force divided by the pure ultimate shear V
u
obtained by the proposed approach and the abscissa gives the moment

divided by the pure ultimate moment M 0 The test results for three
u
unsymmetrical panels, identical in all respects except for the type

of loading, .are shown by the heavy dots 0 The interaction diagram

according to the proposed approach is represented by curve (1)0 The

methods of References 2 and 4 were adjusted following the reasoning

of Reference 24 to make them applicable to unsymmetrical girders and

the corresponding interaction diagrams are given by curves (2) and

(3). The vertical line (4) represents a cut-off on the moment

capacity given by curve (3) and thus limits safe designs to the re-

gion indicated by cross hatching. The interaction diagram of Refer-

ence 24, curve (5), is the most conservative of those shown. The

method of Reference 1 is not included because it is not applicable

to unsymmetrical sections and cannot be readily modified as was done

here for the methods of References 2 and 40 It is seen in the figure

that the proposed approach gives the most consistent correlation with

the test results.

Compari.sons of interaction diagrams for p'anels of other

dimensions were also made and they consistently showed a greater

accuracy of the proposed method than of others.

 5. CONCLUSIONS

The conclusions drawn as a result of this invistigation are

the following:

1. The ultimate strength of an unsymmetrical plate girder

panel depends on the direction of the moment which acts

on the panel, that is, whether the larger portion of

the web is in tension or in compression. ,The panel

capacity is greater when the larger portion of the web

is in tension.

2. The ultimate strength of a plate girder panel under

combined loads may be of two types: the shear ca-

pacity is reduced by bending when the panel is sub-

jected primarily to shear, and the bending capacity

is reduced by shear when the panel is subjected pri-

marily to bendingo

3. The proposed approach gives a reliable means of de-

termining the ultimate strength of homogeneous or

hybrid girders with symmetrical or unsymmetrical

cross section.
 6 ACKNOWLEDGEMENTS

This report was prepared as part of a research project on

unsynunetrical plate girders conducted in the Departm~nt of Civil En.gi-

neering, Fritz Engineering Laboratory, Lehigh University, Bethlehem,

Pennsylvania. Dr. David A. VanHorn is Chairman of the Department and

Dr. Lynn S. Beedle is Director of the Laboratory.

Appreciation is due to Mrs. A. Brumbelow for typing and to

Mr. J. M. Gera for drafting.

 328,9
-28

7 I) APPENDIX I =m REFERENCES

Jll' Akita, Yo, and Fujii 9 To

ON ULTIMATE STRENGTH OF PLATE GIRDERS, Japan Shipbuilding and
Marine May~ 1968.

2. 'American Institute of Steel Construction

SPECIFICATION FOR THE DESIGN, FABRICATION AND ERECTION OF
STRUCTURAL STEEL FOR BUILDING~, AISC, New York, 1969.

'/3. Basler, Ko
STRENGTH OF PLATE GIRDERS, PhDD. Dissertation, Lehigh. University,
1959, available from University Microfilms, Ann Arbor, Michigan 0

J 4 Basler, K.
STRENGTH OF PLATE GIRDERS IN COMBINED BENDING M~D SHEAR~ Proceedings
ASCE, Vol. 87, ST7~ Part 1, October) 1961.

J S Basler, K., Yen, B. T., Mueller, J. A., and Thurlimann, B.

WEB BUCKLING TESTS ON WELDED PLATE GIRDERS, Welding Research
Council, Bulletin Noo 64, Sept., 1960. .

6. Bergfelt, Ao~ and Hovik, Jo

THIN-WALLED DEEP PLATE GIRDERS UNDER STATIC LOADS, Final Report
of the 8th Congress of the International Association for Bridge
and Structural En.gineering, held in New York, Sept., 19.68, ETH,
Zurich Q

~7. Chern, C., and Ostapenko, A.

ULTIMATE STRENGTH OF PLATE GIRDERS UNDER SHEAR, Fritz Engineering
Laboratory Report NOQ 32807, Lehigh University, Sept. 1969.

J S. Chern, C., and Ostapenko, A,

BENDING STRENGTH OF UNSYMMETRICAL 'PLATE GIRDERS, F1;'itz Engineering
Laboratory Report NOQ 32808, Lehigh University, Sept., 19700

90 Cooper, P. Ba
BENDING AND SHEAR STRENGTH OF LONGITUDINALLY STIFFENED PLATE
GIRDERS~ PhoDo Dissertation, Lehigh University, 1965, available
from University ~1:tcrofilms, Ann Arbor, Michigan.

10. Cooper, Pa Bo fi Le\fl:;l Ho So ~ and Yen, BQ T.

WELDED CONSTRUCTIONAL ALLOY STEEL PLATE GIRDERS, Proceedings, ASCE,
V90, STIl, Part I, Feb9' 1964.

11. Cooper, Pe Bo
PLATE GIRDEItS 9 Cllapter 8 in STRUCTURAL STEEL DESIGN, by L. Tall,
et ale, Ronald Press Sl New York, 1964.
12. Dimitri, Jp R~ and O~tap~fiko, At,
PILOT TESTS ON THE ULTIMATE STATIC STRENGTH OF UNSYMMETRICAL PLATE
GIRDERS, Fritz Engineering Laboratory Report No. 328.5, Lehigh
University, June, 1968.

13. Fielding, DG Jo, and Toprac, A. A.

FATIGUE TESTS OF HYBRID PLATE GIRDERS UNDER COMBINED BENDING
AND SHEAR, Research Report No. 96-2, Center for Highway Research,
The University of Texas, Austin, Texas, July 1967.

Fujii, T"
ON AN IMPROVED THEORY FOR DR. BASLER'S THEORY, Final Report of
the 8th Congress of the International Association for Bridge and
Structural Engineering, held in New York, Sept., 1968, ETR, Zurich.

15. Kollbrunnzr, Co F" and Meister, M.

AUSBEULEN, Springer-Verlag, Berlin", 1958.

Lew, He So and Toprac, Ao A.

THE STATIC STRENGTH OF HYBRID PLATE GIRDERS, Report No. S.F.R.L.
RPT. P550~11, Structural Fatigue Research Laboratory, Department
of Civil Engineering, T~e University of Texas, Austin, Jan., 1968.

17. Lyse, I. and Godfrey, He Jo

INVESTIGATION OF WEB BUCKLING IN STEEL BEAMS, Trans. ASCE, Vol. 100,
1935.

18. Nishino, Fe and Okumura, T.

EXPERIMENTAL INVESTIGATION .OF STRENGTH OF PLATE GIRDERS IN SHEAR,
Final Report of the 8th Congress of the International Association
for Bridge and Structural Engineering, held in New York, Sept.,
1968, ETR, Zurich.

19. Ostapenko, Ao and Dimitri, J. R.

BUCKLING OF PLATE GIRDER WEBS, Fritz Engineering Laboratory Report
Noo 32803, Lehigh University (in preparation).

20. Owen, Do R. Jo, RockeYj K. C., and Skaloud, M.

BEHAVIOR OF LONGITUDINALLY REINFORCED PLATE GIRDERS, Final Report
of the 8th Congress of the International Association for Bridge
and Structural Engineering, held in New York, Sept., 1968, ETR,
Zurich 0

21. Patterson, p~ J0' and Yen, B. T.

PROOF-TESTS OF TWO PLATE GIRDERS FOR DESIGN RECOMMENDATIONS, Fritz
Engineering Laboraotry Report No. 327.7, Lehigh University, June,
19690

22. Rockey, Ko Co, and Skaloud, M.

INFLUENCE OF FLANGE STIFFNESS UPON THE LOAD CAPACITY OF WEBS IN
SHEARs Final Report of the 8th Congress of the International
Association for Bridge and Structural Engineering, held in New
York, Septa, 1968, ETR, Zurich.
 32809 -30

vl230 Schueller, We, and Ostapenko, A.

MAIN TESTS ON THE,ULTIMATE STATIC STRENGTH OF UNSYMMETRI~M~ PLATE
GIRDERS, Fritz Engineering Laboratory Report No. 328.6, Lehigh
University, Aug., 19680

240 Vincent, Go So
TENTATIVE CRITERIA FOR LOAD FACTOR DESIGN OF STEEL HIGHWAY BRIDGES,
American Iron and Steel Institute, Bulletin Noo IS, March 1969.
328.9 -31

8. APPENDIX II. - NOTATION

IG Lower Case Letters

a Panel width, that is, distance between transverse

stiffeners.

b Panel depth, that is, distance between flanges.

c Half width of the compression flangee

c
c Half width of the tension flange.
t

d Thickness of the compression flange

c
d Distance from the compression flange-web junction to
c
the compression flange.

Thickness of the tension flange.

Distance from the tension flange-web junction to the

centroid of the tension flange.

\ Plate buckling coefficient under pure bending.

k Plate buckling coefficient under pure shear.

v

mel' mer' Plastic moments developed in the compression and tension

flanges at the left and right sides of the panel due to
ro , m
tl tr
the frame action under combined loads.

t Web thickness
M
x = max
V Shear span, that is, the location of the far end of the

panel relative to the point of zero moment.

Distance from the neutral axis to the extreme compression

fiber of the web before buckling.

Distance from the neutral axis to the compression flange-

web junction for a fully yielded cross section.

Distance from the neutral axis to the extreme tensile fiber

328,9

2. .EJ~lr Case

A Arf:.'8 of the plate girder cross section.

A Area of the compression flange.
fc
.A
ft Area of the tension flange.

A Area of the web,

w
C The ratio of the maximum tensile stress (or minimum

compressive stress) to the maximum compressive stress

of the web plate (for a positive moment, C is the

ratio of the bottom fiber stress to the top fiber

stress). Note that C is negative when the bottom

fiber is in tension.

E Modulus of elasticity.

H Horizontal component of the tension field force; H',

for incomplete tension field.

I Moment of inertia of the' cross section about the

horizontal centroidal axis,

M Moment.

M Maximum moment in panel.

ma.x
M Plastic moment f
P
M Moment acting on the panel under V .
th th
M Ultimate moment controlled by the capacity of the com-
u

pression flange, pure bending.

R Parameter used in Eq.7.

s Average stress in the tension portion of the web at

buckling.

v Shear.
32-8.9 ~33

Plastic shear of'the web 0

Ultimate shear strength of the panel under combined

loads.

V Beam action shear; with subscript "Te" , under combined

T

loads.

Frame action shear; with subscript life", under combined

loads.

V
a Tension field action shear; with subscript "ac", under

combined loads.

v'a Incomplete tension field shear under combined loads.

3. Greek Letters

a = alb Aspect ratio, that is, panel length to depth ratio.

S = bit Web slenderness ratio, that is, web depth to thickness

ratio.

Parameter defined for Eqe7.

Strain; with subscript lty", yield strain.

Coefficient of effective web depth.

ne , "'If'G' Non-dimensional parameters: n =

e
d Ib, n
c f y /b1'
o
2c
c
nh = t
M
11 = Shear span ratio.
bV
v Poisson's ratio.

p Coefficient of equivalent tension field stress in the elastic

triangular portion.

p , p ,. It Non-dimensional parameters: , A /A , P
1 2 f C W 2
1

p = a / a ,P = a /a ,p = a/a
1.+ YC yw 5 yt yw 7 yw
328.9 -34

Ole' 0'2c' Reference stresses in the compression flange.

3c'
cr sc

It' Zt' Reference stresses in the tension flange.

a
3t' st

abc Bending buckling stress under combined loads.

Cf Critical stress of the compression flange column.

0
cp Plate buckling stress under pure bending.

t Tension field stress in the fully yielded zone; with sub-

script "e", under combined loads.

a Yield stress of the compression flange.

yc
a Yield stress of the tension flange.
yt
a Yield stress of the web.
yw
T
C
Shear buckling stress under combined loads.

T
cr Theoretical shear buckling stress under pure shear.

Inclination of the tension field; with subscript "e",

under combined loads;with subscript Irolr, the optimum

inclination of the tension field.

Ratio of the horizontal component to the vertical component

of the tension field force.

In general, subscript "ex" (experimental) refers to ultimate loads)moments)

and shears observed in tests.

328.9 -35

9. TABLES AND FIGURES

 Table 1. Comparison with Tests on Symmetrical Plate Girders W
i ; ; i --- I ;
N
ItO>
s Test Web Compression Flange Tension Flange l..l = ~

o V
No. M ex
u ex
r
a s V b
V
e "l:r
~}r",
V
th
v
th
ex
c b x t 0"
yw
2c
c
x d
c
I 0"
yc
2c
t
xd
t
1 cryt
e

(1) (2) (3) l (4) (5) (6) (7) (8) (9) (10) I (11) I (12) (13)1 (14)

in. x in. Ksi in. x in. Ksi in. x in. Ksi Kips i Kips

. G8- Tl I 3 O 1254150.0 x .1971 38.211Z.O x .752 41.3112.0 x .747 41.3 1.5 85 87.2 .98
Ref. 51 G8-T3 1.5" " " II " " " 2.25 116.5 117.2 .99
G9-T3 I " 382150.0 x .131144.5112.0 x .755 141.8112.0 x .745 I 41.8 2.25 79 83.5 .95
Hl-TI 3.0 127150.0 x .3931108.11*i~:g~ ~ :~~; 102.0 *i~:g~ ~ :~:~ '102.0 1.25 630 610 1.03
HI-T2 1.5 " " " 1 8 . 0 6 x .977 " 18.06 x .983 " 0.75 769 770 1.00
Ref. 10 I
HZ-II 1.0 1281 50.0 x .390 1110.21*i~:g~ ~ i:gg: 108.8 *i~:g~ ~ i:gg: I 108.8 2.5 917 971 .95
H2-TZ 0.5 " " " " " " II
2.75 1125 1235 .91
Ref~ gILS1-Tlll.O 2561 50.0 x .195 46.81 14.12 x 1.4981 30.5 114.12 x.I.498 30.5 2.5 1182 1181 11.00
F10-TIll.S 197150.0 x .254 38.7116.0 x 1.0181 30.2 116.0 x 1.018 30.2 2.43 170 174.6 .97
Ref .21\ F10-T2 1 " I"I It I" I It " " " 1.23 184 184.0 1.00
F10-T3 1 . 2 " "I "I " " "
H
1.08 190 202 .94
WE-I 12.64 56 14.0 x .248 43.3 10.0 x 1.55 33.0 10.0 x 1.55 33.0 1.32 109 109 1.00
WB-2 " 55 14.0 x .255 47.8 10.0 x 1.56 " 10.0 x 1.56 " 1.32 128 119.5 1.07
WB-3 2.56 59 16.03 x .273 49.6 10.06 x 1.50 " 10.06 x 1.50 " 1.28 139 140.5 .99
WB-6 2.45 70 17.56 x .251 33.1 10.02 x 1.51 VI 10.02 x 1.51 " 1.23 96 100 .96
Ref-Ill WB-7 2.51 61 15.34 x .253 33.7 10.07 x 1.50 " 10.07 x 1.50 " 1.26 95 98 .97
WE-8 2.46 60 15065 x 0262 2907 10.07 x 1.51 " 10 . 07 x.1.51 " 1.23 100 97.5 1.03
If
WB-9 2.68 50 12.50 x .250 30.3 10.04 x 1.50 " 10 . 04 x 1 . 50 VI
1.34 92 92 1 .. 00 ~
~
u
WB-IO II " I 12.50 x $252! H I 10 01 x ls51
0 tv 10.01 x 1.51 1.34 94 92.5 1 .. 01

* Cover plate welded to the flange. Continued

------.J
 Table I. Comparison with Tests on Symmetrical Plate Girders (Continuation) W
N-
00
S Test Web Compression Flange Tension Flange 11 = \.0

0 ./ V
No~ M ex
U e~
a S V V V r

r V b ex th th
ex
c b x t (J 2c x d (J 2c x d a
yw c c yc t t yt
e

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14)

kg/ mm2 mm x mm kg/ mm2 mm x mm kg Irrrrn 2 Tons Tons

mm x mID

Gl I 2.61 55 440 x 8.0 44.0 160 x 30 42.0 160 x 30 4280 1.31 82 91 .90
11
ZOO x 30 " 200 x 30
if
1.31 84 93.5 .92
G2 " " "
-

G3 2063 70 560 x 800 " 160 x 30 rr

160 x 30 " 1.31 99 107 .6_ .92
n
" 250 x 30 Ii'
1.84 97 101.2 .Q5
Ref.l G4 3.68 tv
" 250 x 30
" fI Ii'
" " 1.34 107 III 097
G5 2.68 " "
" " "
II
" 1.88 120 123.5 .97
G6 1.25 fI
'"
.97
':~ 2.68
,/'"
VI
" " " " " " 1.34 107 111

" " " 1.39 118 125 .95

G9 2.78 90 720 x 8.0 " " \

Gl 2.67 59.7 543 x 9,,1 38.0 301 x 22.4 44.0 301 x 22.4 44.0 1.34 110.5 110 1.00 "--v'

" 220 x 22.4 " 220.x 22.4 " 1.34 104 108 .96
Ref~18 G2 " " "
-G3 2.63 7608 722 x 904 " 302 x 22.2 " 302 x 22.2 " l~~t 124.5 137 .91
G4 " jl
" " 243 x 22.2 " 243 x 22.2 " . ** 131
131 114.5 .88

* Cover plate welded to the flange.

** Failure of the flange.

:1
\"J-.:J
-..J
 Table 2 Q Comparison with Tests on Unsymmetrical Plate Girders . W
- - '
N
00'
S Test Web Compression Flange Tension Flange 11 = \0..
0 V
No .. M ex
u ex - -
r
a S V V v
V b ex th t.h
c a a ex
b x t
yw
2c x d 2c x d a
e c c YC' t t yt

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) i

in .. x in. Ksi in. x in. ~(si in .. x in . Ksi Kips Kips

(a) Smaller Flange in Compression -- Failure of the We~

/
\\.,/
8.0 x .625
Ref(l~ UG2 . 2 1.2 295 36.0 x 0122 43.2 8 .. 0 x .625 36.7 36.7 1.39 70 70.5 0,.99
~'clO .. 5 x .750
Refa13 UG4 . 1 1.77 414 48~07 x .116 56.1 10.0 x .. 750 34.1 13.0 x 1.384' 34 . 1 0.88 81.6 82.5 0 .. 99
/1' \"'-'-"<'

(b) Smaller Flange in Compression -- Failure of the 'Compression Flange

/ 8.0 x .625
1.6 . 295 36.0 x .122 43.5 8. 0 x .625 33.3 33.3 3.86 43,,8 39.4 1.~ 20
Ref.",12 /UG3.2
!
i~lO.5 x .750
'\. i// UG3.3 IJ tr II it tr
" If H
3 . 86 42.5 tJ
i 1~O7
/\

Ref/23 \"
~' \, UG4e3 1~46 414 48.07 x .116 5611 10.0 x 9750 34.1 13.0 x 1.384 34.1 3.77 63.2 60.0 1~t()5

\;
II
"""".,UG4 " 4 le77 263 48 .. 07 x 0183 35.5 " " " 3.62 70 68,,0 l.()]

(c) ,Smaller Flange in Tension -- Failure of the Web

..
UG4.2 1.14 414 48.07 x 0116 56.1 13.0 x 10384 34.1 10.0 x .750 34.1 1.93 119.2 106.0 1 .. 12
Ref.23 UG405 (> 83 263 48.07 x .183 35.5 " .. " VI
2.19 136 133 .. 5 1.{j)1
n
UG406 1.77 " u Vi Of tr
" a 88
I' 9888 10000 0.99
-ll:.}'J'

DC

* c over .,
pJ...ate 1 "f
wecj]~de d'1:0 tJ.JLe
h r~
J:-LaA."'1ge IJ
 Table 3. Comparison with Tests on Sxmmetrica1 Hybrid Plate Girders.
!W
N
S Test Web Compr. Flange Tension Flange l-l = tOO
'\0
0
No. M V
u ex ex
V V
r a S V b ex th V
ex th
c
e
b x t ayw 2c
c
x d
c
0
yc
2c
t
x d
t yt

(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14)

in. x in. Ksi in. x in. Ksi in. x in. Ksi Kips Kips
HSl-T1 2.0 185.6 36.0 x .194 65.6 8.0 x .517 107.6 8.0 x .51.7 107.6 1,0 140 135 1.04
HSI-T2 1.0 " " " " " " -ff
1.5 190 173.5 1.09
HSI-T3 .5 " " VI
" " " " 1.75 226 214 1.05
HSIA-TI 2.0 192 35.87 x .187 49.0 7.99 x .533 104.2 7.99 x .533 104.2 1.0 98 104 .94
HSIA-T2 1.0 " " " " " ff
" 1.5 125 137 .96
Ref.16
HSIA-T3 .5 " " " II
" " fI
1.75 167 158.5 1.05 ,@
HS2-Tl 2.0 182 36.0 x .198 54.6 8.0 x .517 107.6 8.0 x .517 107.6 1.0 131 122 1.07
HS2-T2 1.0 " " ff
" " " " 1.5 177 153.5 '1.15
HS2-T3 .5 " " " " " " " 1.75 205 186.5 1-.10
31020 .83 190 36.0 x .189 40.8 8.02 x .522 105.0 8.02 x .522 105.0 2.08 115.8 115.8 1.00
31530 " " "
If
7.99 x .528 " 7.99 x .528 " 2.08 108.1 115.5 .94
32550-C2 1.0 176.5 36.0 x .204 51.2 7.98 x .532 106.2 7.98 x .532 106.2 2.5 153 139.5 1.09
Ref.13
rr
32550-C2R
I( I(
" fI
" Jf
" 2.5 148 139.5 1.06

:]
~
-D

1
_ J
 Panel p
~

a
x
)( - at 2

max

Moment ~------ptSo--+--~-+------------......--...

/ Mmid- panel

..---_ _--+-+--+--I--,v
,
Shear t

I I

.Fig.l -Loading Diagram of the Panel Under Investigation

328~9

010-----------0

1
c

Fig. 2 Beam Action Model Under Combined Loads

N.A. j r N.A. --
TC Te

cubc
- CTbc
ii4liliii
TC
2 C 2
First Yield
Zone

Idealized Tension Field Model Under Combined Loads

328. 9

~ Vcrc1

Ofc Ivcrc b 1 "tc

--j ~pcrtc

a a
~ ..! ~ --I
(a) (b)
Expected Simplified

Fig. 4 Tension Field Stress Distribution Under CombinedALoads

Fig. 5 FrameAction Model Under .Combined Loads

(a) Beam Action

1~/2:1'
---x---a-V-2---~-IiIIIIliIIIIa.
t
c .....

(b) Tension Field Action

...
(r3C

...

Fig. 6 Reference Stresses in the Flanges

 Wi
N

.
00

'-0

Vth
Vu
IO"IC+ U2C+0"3C S O"cf ] I O"lt + 0"2t+0"3t S 0"yt I
O"lt +U2t+0"3t SUyf7 / '\ 0"IC+0"2C+ 0"3C S 0"cf

0"It+02t+ U 3t :::uyt 4
-
, 0,1.0 ,

I
02 .......----U1C + 0"2C + 0"3C= O"cf

0"1t +"2t +u3f >0"yt ~- - -Web Failure--~

0-, C + 0-2C +u 3C >O"cf
/ ~ // \
Tens. " Compo
Flange ~ / Flange
Failure " / Failure
M \al5 I " / \ "C ~ M
-Mthu 1.0 0 1.0
-
Mu
th

Fig. 7 Schematic Interaction Curve - Stress Inequality Relationships

~
~
 (!Jl
~

m
..
\ijO

\ ~O~ I +~~~+=[
Frame Action 'iD~1
+.
~

.,.
~
.L
-y
+--44-

M Q5
-l--J-
M
u 1.0 o
r
1.0
3
M
Mu
1_ Mp -1_ Mb -1_ M-t -J

II

Fig. 8 Schematic Interaction Curve - Failure Mechanisms

....
Ul
 -,

32 ae 9

CTyc
a
..-..--.. . ""'--o-""''''''----'T c
N.A.

b t
T <T y

cryt
(a) Cross-Section (b) Bending Stress (c) Shear Stress
Fig. 9 Stress Distribution of Plastified Cross Section -
Neutral Axis in the Web

de Afe
-+--+--_-==-L- _ N.A.

b t
T<Ty

Afe

C)tt
(0) Cross Section (b) Bendino Stress (c) Shear Stress

Fig 0 10 Stress Distribution of Plastified Cross Section -

Neutral Axis in the Flange
 I.O~ /,..n ~I 1.0
-T3
O.8~ / ~
I 1.0
0.8

~ 0.6
1/ \ Y-
V u
o.st
0.4 / \ ~U
0.8~ /
O.S \ ~:~:~/ \ UJ
N
0-
-
dt

o 0.2 0.4 0.6 0.8 1.0 o 0.2 0.4 0.6 OB LO

M M o 0.2 0.4 0.6 0.8 1..0 o 0.2 04 0.6 0 . 8 1.0
Mu Mu M M
Mu Mu

1.0 LO~ /LSI-TI

1.0
08 0.8
0.8 0.8
:::L 0.6 .::L 0.6
::L 0.6 .::L 0.6
Vu 0.4 V u 0.4 Vu Vu
0.4 0.4
0.2 0.2
o 0.2 0.4 0.6 0.8 1.0 o 0.2 0.4 0.6 0.8 La
M M
o 0.2 0.4 0.6 0.8 o 0.2 0.4 0.6 0.8 1.0
Mu Mu
..At. l!t
Mu Mu
i
Fig. 11 Interaction Curv~and Test Results _ Fig. l~ Interaction Curv~and Test Results - ~
88,89 and LSI-Tl (Ref. 5 & 9) HI and H2 Series (Ref. 10) '-..J

I
J
 loO~ / f'IO- T3

0.8
I.0r- I
IG3-----....,
L

as
0.6
Y- v 0.6
V
u 0.4 V
u 0.4

0.2
o 0.2
0
L 042 0.4 046 048 1.0 o 0:2 0.4 0.6
M M
Mu Mu
~
ext
Fig4 J.3 Interaction Curve~and Test Results - FlO Series
(Ref 21) Fig4 14 Interaction Curv~and Test Results - Gl to G4
(Ref. 1)

j
 1.0
1.01 , \,;1.
1.0
0.8

V 0.6
o.sr I \ O.8~ rI dI
\f'!I

Vu
0.4
~ 0.6 V 0.6
1 / ~ ...
0'

'"u 0.4 Vu 0.4

0.2
o 0.2 0.4 0.6 0.8 1.0
M
0 0.2 0.4 0.6 0.8 1.0 0 0.2 0.4 0.6 '0.8 LO
Mu
l!t 14..
My Mu

1"""""----
1.0..
I.OT'.-- - I.u....
r - - -__
o. 0.8
0.8 as
.:i- V 06
Vu Vu :y.... 0.6 V 0.6
0.4 Vu Vu
0.4 0.4
0.2 0.2
o 0.2 0.4 0.6 0.8 1.0 o 0.2 0.4 0.6 0.8 1.0
.M. 14.. o 0.2 0.4 0.6 0.8 1.0 o 0.2 0.4 0.6 0.8 ID
Mu Mu l!l .M.-
MU Mu I
~
Fig. 15 Interaction Curveaand Test Results - G5 to G9
(Ref. 1) -G
Fig. 16 Interaction Curv~and Test Results - G1 to G4 s~
(Ref. l8)
 1.01
'J

I
il

.J~U(~2-?
1.0

v :::t / '\ ~ V'0r- / y..

0.8r-

0.61- /
I l

\
fJ3
N
~
V
u
0.4 1
,I f \

\ Vu
0.4
w

!
0 0.2 0.4 0.6 0.8 100 0 0.2 0.4 0.6 0.8 to 0 0.2 0.4 0.6 0.8 to
M M M
My Mu Mu

to L I.

0.8 I \, /' 0.8 Q

UG3-2
UG3-3 UG4-3
V O.6k- ./\
0.6 0.6
Vu Y..

VU:"~I
0.4 "U Q4

0.2

o 0.2 0.4 0.6 0.8

M
1.0 0 0.2 0.4 0.6 0.8
.M-
1\ 1.0
0.2

0 0.2 04 0.6 0 ..8

.M-
to
M\IJ Mu Mu
j

rig~ 17 Interaction Curv~and Test Results -

Fig. 18 Interaction Curv~and Test Results - UG4 Series 01
(Ref. 23) Q
UG2 and UG3 Series (Ref" 12 )
 W
tV
co

(U I~ t) I v
{~ 1 ~~~) I \0

Vu
l.O

~, I I' I I I I I I _ . , I , , _
M
Mu L8 1.6 i.4 1.2 1.0 0.8 0.6 0.4 0.2 0 0.2 0.4 0,6 0,8 1.0 Mu

Fig. 19 Complete Interaction Curve and Test Results - UG4.4 and Ui;4. 6
(Ref. 23)

1
t.-u

_ _ _ _ _"""""' """""' _''._. .. _.. _. "-.0 .__ ..

".~._~._ _====== _
J
 ~
r">J
(OC

'0

{~. i ~ ~~ I x
Vu
{~i ~} I
1.0 ~UG2-1

UG2-3 ..M...
1.5 1.0 o ro Mu

Fig. 20 Complete Interaction Curve and rre s t Results -

UG2 Series CRef.12)

n
rf.F\1
(f'G:

~
j
 LO
---
I
~ 1.0~ .,
w

:~:t
t.)

0.8
Y- \ Ui
V
u 0.4
\ .::L
Vu

o 0.2 0.4 0.6 0.8 1.0

M
Mu
o 0.2 0.4 0.6 1.0
M
Mu

HSI .. T2 1.0
1.0 1.0

0.8 0.8
:L .::L
Vu Vu V 0.6 ::!... 0.6
0.4 Vu Vu
0.4 0.4
0.2 0.2
0.2
o 0.2 0.4 0.6 0.8 1.0 o 0.2 0.4 0.6 0.8 1.0
M M
Mu Mu o 0.2 0.4 0.6 1.0 o 0.2 0.4 0 . 6
-M.. -M..
Mu Mu E
Fig. 21 Interaction Curv$and Test Results - HSI Series
(Ref. 16) 0]
()J
Fig. 22 Interaction Curves and Test Results - HSlA Series
(Ref. 16)

J
 1.0
W
1"WI
0.8 0.8 .!JP
.0)

V 0.6 :i.. 0.6 :y- 0.6

Vu Vu Vu
0.4 0.4 0.4

0.2 0.2

0
_l_' o
0.2 0.4 0.6 0.8 1.0 0.2 0.4 0.6 0.8 1.0 o 0.2 0.4 0.6 0.8 La
M M M
Mu Mu Mu

HS2-T2

V 0.6 V 0.6 V 0.6

Vu Vu Vu
0.4 O. 0.4

0.2 0.2 0.2

0 0.2 0.4 0.6 0.8 1.0 0 0.2 0.4 0.6 0.8 1.0 o 0.2 0.4 0.6 0.8 1.0
M M.- M
Mu Mu Mu
Fig. 23 Interaction Curv~and Test Results - ES2 Series
(Ref. 16) 01
Fig. 24 Interaction Curves and Test Results - (Refs.. 13 and 16) ~

J
328.9

1.0 ( I) I UG2 - 2
(4)

VU (2,3,5)
Vu

-VV 0.5
u

UG2-3
( !5)
0.5
M/M u
0 MU (4, 5) J I. 0 MY(3) Mp
-M
Mu Mu U

Fig. 25 Comparison of UG2 Series with Various Theories