J. Construct.

Steel Research 30 (1994) 107 124

~' 1994 Elsevier Science Limited
Printed in Malta. All rights reserved
O143-974X,94/$7.00
ELSEVIER

Behaviour of Single Angle Beam-Column Element

Subjected to Load Reversal

G. M. S. Knight & A. R. Santhakumar

Department of Civil Engineering, College of Engineering, Anna University,
Madras 600 025, India

(Received 6 April 1992; revised version received 16 October 1992;

accepted 19 May 1993}

ABSTRACT

Multi-storeyed industrial structures subjected to seismic excitation are invariably

subjected to a slow cycle fatique type of load reversal. I f a structure has to pe(lbrm
satisJactorily its components should have sufficient ener.qy-dissipatin,q characteristics
and also be capable of bein(j repaired after a definite excursion into the post-elastic
region. Single angle members are extensively used in transmission and microwa~Je
towers and also in braced domes. The earthquake type of loading produees adverse
effects on single anole members which are unsymmetrical in nature. This investiqa-
tion examines the perJbrmance of beam and beam-column elements subjected to
reversed, static cyclic loading. As a first step, the performance of rolled steel equal
an.qle members subjected to shear centre and off-shear centre loading has been
studied under .flexure, .flexure shear and .flexure shear axial load combinations.
Based on the above experimental results the importance of reversal ~ Ioadin9 on
single anole members has been established.

NOTATION

C x x ~ Cyy Centroidal distances in x and y directions

e Eccentricity with respect to shear centre
L Span of the beam/beam-column
P Axial load through shear centre
I4/ Lateral load through shear centre/gusset plate
~y Yield strain

107
108 G. M. S. Kni,qht, A. R. Santhakumar

1 INTRODUCTION

The object of this investigation is to study the behaviour of single angle

beams and beam-columns under simulated slow cycle load reversals.
Generally single angle members have semi-rigid end connections. Their
performance will depend upon the type of end fixture. As a first step the
tests in this investigation have been on members with end connections
which permit full rotation. It may be pointed out that such a flexibility at
the two ends will generally lead to adverse effects, especially under
reversed cyclic loading. The members used for the test have been obtained
commercially and hence have the usual imperfections. However, it has
been ensured that the imperfections are within the prescribed code limits.
The object of this study is to quantify the energy-dissipating characteristics
of single angle beam and beam-column members. Angle specimens were
tested by Jain e t al. t under static and slow dynamic loading conditions. It
was concluded that the effective slenderness ratio of a member was the
most important parameter governing its hysteresis behaviour and that
the total energy dissipation through a hysteretic cycle was independent of
the direction of loading. Dux and Kitipornchai 2 have done a series of ex-
periments on the buckling of laterally simply supported continuous I-beams
in the inelastic range. They measured the geometric and material imper-
fections. The experimental results were compared with the theoretical
predictions.
Kitipornchai ~ has studied the torsional flexural buckling loads of angle
compression members. The cubic buckling equation, which is to be solved
for the required result (for the lowest roots), was solved by parametric
solution. Madugula and Kennedy'* studied the behaviour of the single
and compound angles under elastic and inelastic conditions. Kitipornchai
and Lee s compared the experimental results with the theoretical pre-
dictions on the inelastic behaviour of angles in compression; AISC,
AS-1250 and SSRC formulae were used. It was proposed to use an SSRC
curve for the design of single angle struts. Ray and Madugula 6 have
developed a Fortran 77 computer program on the basis of the theory
developed by Leigh, Thomas and Lay for the design of laterally unsup-
ported steel angle beams. Periyasamy 7 has studied the behaviour of rolled
steel equal angles in bending, in axial force and in axial compressive force
alone under monotonic loading. Two different thicknesses of equal angles
were tested for the above combination of loading with centrally concen-
trated loads and two point loads with shear centre and off-shear centre
loading.
Indumathi e t al. s have examined the performance of beam and beam-
column elements subjected to reversed, static cyclic loading. However,
 Sin(fle angle beam-column element sul?jected to load reversal 109

there is a need for experimental data on the fundamental behaviour of

beams and beam-columns subjected to a hysteretic type of loading under
hinged conditions. The object of this investigation is to attempt quantifica-
tion of the characteristics of members subjected to high strain rate slow
cycle load reversals.
A total of 18 specimens (see Table 1) were subjected to first monotonic
and then reversed cyclic loading under different eccentric loading condi-
tions realisable in practice.

TABLE l
Comparison of Ultimate Moments for Tests Conducted

Test identification ISA Ultimate moment % error No. of

65 x 65 (kN m) cycle.~

Observed Predicted

Centrally concentrated load 6 2.79 2-52 10-71 4

through shear centre 8 3.26 310 5.16 8
Centrally concentrated load 6 3.03 2-52 2024 8
through gusset plate 8 338 310 9-03 8
Two-point load at one-third
span through shear 6 2-64 2.52 4.76 8
centre 8 3"19 3' 10 2.90 8
Two-point load at one-third 6 2.79 2-52 10-71 4
span through gusset plate 8 3-42 3-10 10-32 8
Uniaxial load through 6 1.98 1.91 3.66 8
major axis 8 2.45 2.35 425 8
Uniaxial load through 6 4-19 3.82 968 2
minor axis 8 4-89 470 404 3
Beam-column (Biaxial)
axial force (P) through 6 2-70 3-25 t6"92 8
shear centre ( P = 4 ' 9 kN) 8 3-59 4-07 11.80 8
Beam-column (Biaxia[)
axial force (P) through 6 2.20 2.96 25.70 5
shear centre (P=9-8 kN) 8 2.92 3.76 22.30 4
Beam-column (Uniaxial)
through major axis with
axial force (P) through 6 2.06 2.78 25-90 8
shear centre (P ~ 4-9 kN) 8 2-69 3.46 22.30 8
110 G. M. S. Knight, A. R. Santhakumar

2 DETAILS OF EXPERIMENTAL INVESTIGATION

The angles are held by means of 16 mm diameter bolts, one on either side.
Nuts on the bolt are kept snug so that the support acts as a simple
support. The variables considered while testing are axial load P, lateral
load W and eccentricity e with respect to the shear centre. The intention of
the test was to find the effect of these variables on the stiffness and energy
absorption characteristics of single angle beams and beam-columns.
Figure 1 shows the different types of loading adopted on the member

| /-- 8MM DIA ROD OF

~ LENGTH50NM

UNIAXIAL LOADING ALONG MAJOR AXIS

UNIAXIAL LOADING ALONG MINOR AXIS

t ~. it,/-GMM DIA ROD

(Jr'- LENGTH 50MMOF

~
LOADING THROUGH SHEAR CENTRE

GUSSET PLATE

,°I
LOADING THROUGH GUSSET PLATE

Fig. I. Types of loading.

 Sin,qle angle beam-column element subjected to load rerersal t 11

and Fig. 2 shows the details of the strain gauge positions for correspond-
ing loadings. The typical experimental set-ups for shear centre and gusset
plate loading are shown in Figs 3 and 4 respectively. The attachment of
the gusset plate is as shown in Fig. 2(b). The plate was welded to the angle
with a bearing of 65 x 50 mm. Welds of 6 mm were used for a length of
65 mm on either side. To enable reversal of loading, special end fixtures
were designed to accommodate this during testing.

Y
I
" I
4 5 /U 4
v // ' i
I
I
/ I
6MM WELD -- |

I / I
\ \ _ _ 1 ,'~ 4s ° I i
X . . . . . . ~_~L_ ~ X X--- - - X
6

!
/ 2 I X% I 2 I
/ I %

U V
Y
(a)CENTRAL C O N C E N T R A T E D LOAD (b) C E N T R A L CONCENTRATED LOAD
THROUGH SHEAR CENTRE THROUGH GUSSET PLATE

U V

W t

2 i
V U " -'-~" - - U
•

i
U
(c) SIMPLY SUPPORTED BEAM (dISIMPLY SUPPORTED BEAM
UNIAXIAL LOADING THROUGH UNIAXIAL LOADING THROUGH
MAJOR PRINCIPAL AXIS MINOR PRINCIPAL A X I S

Fig. 2. Details of strain gauge positions.

112 G. M. S. Kniyht, A. R. Santhakumar

Fig. 3. Centrally concentrated load through shear centre.

Fig. 4. Centrally concentrated load through gusset plate.

3 THEORETICAL ANALYSIS UNDER

MONOTONIC LOADING

Using basic structural mechanics principles and assuming stress and strain
to be linearly related when the deflections are small, strains were predicted
 Sm.qle an qle beam-column element sul!jected to load recersal 113

for both beams and beam-columns. For beam-columns the P - A effects

were also considered while predicting the strains. The purpose of this was
to investigate the deviation between the predicted and observed strains
assuming initial crookedness to be zero.

4 COMPARISON OF BEHAVIOUR UNDER

MONOTONIC LOADING

Beams

Figures 5 and 6 give a typical comparison of predicted and observed

strains at six locations for two different types of loading. It can be seen
that before the onset of nonlinearity the comparison between the observed
and predicted results is good. This validates the testing method adopted.

Beam-columns

Figures 7 and 8 show a typical comparison of predicted and observed

strains at six locations for two different types of loading, with axial force
P - 4 - 9 kN. In the case of biaxially loaded beam-columns the predicted
compressive strain is more than the observed value by less than 10% and
the predicted tensile strain is more than the observed value by 8%. In the

1
=l l 1.2 +

:,.o 1-

I ,i \ I

L : 1.87m i t I ;

-- 3000 2400 1800 (~ylZO0 600 0 600 1200 £V 1800 2400 3000
C O M P R E S S I V E S T R A I N MICRO ram/him T E N S I L E STRAIN M I C R O l l i m / m m

Fig. 5. Strain variation with load, centrally concentrated load, shear centre loading for 1SA
65x65×8.
 114 G. M. S. Knight, A. R. Santhakumar

~ o.s!
~\ : o.~
i e , I SAI~I$
3 041

I TO S STRAIN GAUGES
1 t/2 -I' L/2 ~'I o.2~ //~/, • i OBSERVED
I.=I. 8 nl ~ ~ ~ • p .... PREDICTED

i l I |, I I

3000 2400 1800 ey 1200 600 0 600 iZO0 (¥ 1800 2400 3000
COMPRESSIVE STRAIN MICRO ram/ram TENSILE STRAIN MICRO rain/ram
Fig. 6. Strain variation with load, centrally concentrated load, uniaxial bending abou;
major axis for ISA 65 x 65 × 8.

z ~ I.O-t-

,/: f / ;
.5 / / ',

" ~ " , , ~t/ //,6' ,~,

[ ~ "~,0141/ /~//~-x. i ISA ,$X65X S
wI : ~l~k\ ' III /L '/ '1 I TO 6 STRAIN ~IUJGES

"] ' ,i, \\zI Z/YI ~ I/ 1

L/: ~
- ~-<-'~i ~./v~" ....
o,,,v,o
P REDICTE 0

- - ~ . . . . . . . . . 1 P~49._on,.. w--_o__

3 0 0 0 2400 I l O 0 £1 1200 600 0 600 1200 £y 1800 2400 3000

COMPRESSIVE STRAIN MICRO ram/ram TENSILE STRAIN MICRO ram/ram
Fig. 7. Strain variation with load for beam-column ISA 65 x 65 x 8.

case of uniaxial loading, before the onset of nonlinearity the c o m p a r i s o n

between the observed and predicted results is in g o o d agreement. This also
validates the testing m e t h o d a d o p t e d .
 Single angle beam-column element subjected to load reversal 115

O.S,

0.5,

\ \ ~/F2S3 o 0.4. 4 /'

%, ~1 II tI

~, X .,, 0.3f t/ ' /,, / " "~ ISA 65 XG~X8

\ " I : ITO G STRAIN GAUGES

W . i / / t ~
I iB , \\~'
\L o.2f , [ . OBSERVED
f "~ ',P ~\ / ,, .... P.EDICTED
,/E , L/, "i "\°It
L=l.em .g ',
. .. ~ ,,, [ P = 4905 N W:O

i i i • i i T i i I ! e:o ,w=o
3000 2400 1800 £y 1200 600 O 600 1200 £y 1800 2400 3000
COMPRESSIVE STRAIN MICRO ram/ram TENSILE STRAIN MICRO ram/ram

Fig. 8. S t r a i n v a r i a t i o n with l o a d for b e a m - c o l u m n ISA 65 × 65 × 8, u n i a x i a l b e n d i n g .

5 BEHAVIOUR UNDER CYCLIC LOADING

Beams

Figure 9 shows a typical hysteresis curve obtained during testing and Fig.
l0 shows the variation of vertical deflection with the increase in cycles. It
may be observed from Fig. 9 that the stiffness has degraded after the fifth
cycle due to buckling of the compression flange. Figure 11 shows the
hysteresis for a typical gusset plate loading and Fig. 12 shows the
variation of vertical deflection with increase in cycles. It may be noticed
that, with gusset plate loading, the failure load level has decreased. Figure
13 shows the hysteresis curves for an unsymmetrically loaded angle subjec-
ted to cyclic loading. The test could not be continued after three cycles
because of excessive lateral deflection resulting in buckling failure as
indicated in Fig. 14. It may also be noticed that under unsymmetrical
loading the hysteresis shows degrading stiffness characteristics compared
to symmetrical loading as shown in Fig. 15. Figure 16 shows the
variation of vertical deflection with increase in cycles for symmetric
loading. Figure 17 shows the hysteresis curves for shear centre loading
with loads at one-third span and Fig. 18 shows the variation of vertical
deflection with increase in cycles.
116 G. M. S. Kniaht, A. R. Santhakumar

L/2 ~ L/2 •
1.2" -- L = I.Q7 m

0.8.

0.4"

O -0.4

-0.0

-I.2.
ISA 65165x8
200 I00 0 I00 200
VERTICAL DEFLECTION IN mm

Fig. 9. Hysteresis curves for ISA 65 × 65 × 8, ccntrall.~ concentrated load, shear centre
loading.

~
W

300

_
2OO
t
I00 -
• = bp'r m ,t , 11

0-
t /
-I00.

-200 -

-300-
~'~ ISA 65X65X8
LOAD CYCLE NUMBER

Fig. 10. Vertical deflection vs load cycle for ISA 65 x 65 x 8, centrally concentrated load,
shear centre loading.
 Single angle beam-column element subjected to load reversal 117

1.2"
"l' L=I.EI7m _~
0.8
Z
z 0,.4-

o
X
-~ W ---~) W
..z -0.4 -

-O.B"

-I.Z:
ISA 65 X6.'$ X8
-,~o o ,~o ~;o lio ,~o 2~o 2;,0 2ao s2o
VERTICAL DEFLECTION IN mm

Fig. 11. Hysteresis curves for ISA 65 x 65 x 8, centrally concentrated load, gusset plate
loading.

. . . . L : 1.87m -,t
E 200t

;,,o4 A I

i-4°1 V
- 1 2 0 ]

- 2 0 0 ~ 1~6SX65X8
LOAD CYCLE NUMBER

Fig. 12. Vertical deflection vs load cycle for ISA 65 × 65 x 8, centrally concentrated load,
gusset plate loading.
118 G. M. S. Kniqht, A. R. Sc~nthakutnctr

1.6

1,2]
0.6
[ 3

o1
-0.6-J
i
-i.2!
-I.6~
-eo -~O -4b -io o ~o ,/~ do 8'o
VERTICAl- DEFLECTION iN mm

Fig. 13. Hysteresis curves for ISA 65 × 65 × 8, centrally concentrated load, uniaxial bending
about minor axis.

"° 1
z 1
; L=l.S.

W W

°j -BO

- 120
ISA 65x65x8

LOAD CYCLE NUMBER

Fig. 14. Vertical deflection vs load cycle for ISA 65 x 65 x 8, uniaxial bending about minor
axis.
 Single am jle beam-column element sut~jected to load re~,ersal 119

t
0.6

0.3 +___ L=I.B,,, ,t"

W
W

ISA 65 X 65X8

-&o -,oo -~ o ~'o 100 150

DEFLECTION IN mm

Fig. 15. Hysteresis curves for ISA 65 x 65 x 8, centrally concentrated load, uniaxial bending
about major axis.

~(_ L/2--~f-- L/2

1- L=l.Bm

z 2'° 1
A
5 __ _

~-'ol v V V

-250t_ -
ISA 65x 65X8

LOAD CYCLE NUMBER

Fig. 16. Vertical deflection vs load cycle for ISA 65 x 65 × 8, uniaxial bending about major
axis.
120 G. M. S. Knight, A. R. Santhakumar

/_~ w~; ;w~

1.2
I

o; o.4
x 0

-0.4

-I,2

- 2 .O

ISA 6 5 X 6 5 X 8

-21o -~o -9o -~o do ~ iso 330

VERTICAL DEFLECTION IN mm

F i g . 17. H y s t e r e s i s c u r v e s for I S A 65 x 65 x 8. s h e a r c e n t r e l o a d i n g , l o a d s a t o n e - t h i r d s p a n .

w/2 w/2

L= I.Blm

/ /
250-

AAA
E
E 150.
_z

O 50-
u
e
e.,.
..l
-150

> - 25O.
vvVI i LOAD CYCLE NUMBER
ISA 6 5 X 6 5 X B

F i g . 18. V e r t i c a l d e f l e c t i o n ~s l o a d c y c l e for I S A 65 x 65 x 8, s h e a r c e n t r e l o a d i n g , l o a d s at
one-third span.

Beam-columns

Figure 19 shows a typical hysteresis curve for centrally concentrated shear

centre loading. The lateral load-carrying capacity decreases by 20% with
 Sinyle an qle beam-column element subjected to load reversal 121

0.8 .~ L - I.Sm

x
o] __
! _ ®

-0.8 ~

-"2t. ,sA 8~x.~x8

-8o -~o -;o -io o 20 ~o 6'o ~o ,~o
VERTICAL DEFLECTION IN mm
Fig. 19. Hysteresis curves for ISA 65 x 65 x 8, centrally concentrated load, shear centre
loading.

the introduction of an axial force and the deflections remained constant

after three cycles of loading. Figure 20 shows the variation of vertical
deflection with increase in cycles. Figure 21 shows the hysteresis curve for
uniaxial loading (symmetric loading). Here also the lateral load-carrying
W

,ooJ L = l . Sm .~,,
P = 4 9 0 5 ~1

-- 60:

~ -20

~-IOI
ISA 65X65 X 8
LOAD CYCLE NUMBER

Fig. 20. Vertical deflection vs load cycle for ISA 65 x 65 x 8, centrally concentrated load,
shear centre loadimz.
122 G. M. S. Knight, A. R. Santhakumar

0.6

0.4

o2I
o; o
x

,( W
0 -0.2
_1

-0.4

-0,6.

-150 -i:~o
L
-9o -go -~o
ISA

3b
+
65X65X8

~--
VEWnCAL D E F L E C T ' O . ~N . .
Fig. 21. Hysteresis curves for ISA 65 × 65 x 8, centrally concentrated load, uniaxial bending
about major axis.

W
120 P ~ P

E
E 80 L= I.Bm .
z
P = 4905 N

,~ A 1

i
o

U
Ig
Iii
-40.

-80 -
V
-120- ISA 65X65X8

LOAD CYCLE NUMm[R

Fig. 22. Vertical deflection vs load cycle for ISA 65 x 65 × 8, uniaxial bending about major axis.

capacity decreases by 30% as compared to the beam, and the variation of

vertical deflection with increase in cycles is shown in Fig. 22.
The specimens were subjected to approximately 3 to 8 cycles depending
upon their ability to withstand the load cycling. Of the 18 specimens
 Sin.qle angle beam-column element sul~iected to load reversal 123

tested, seven typical test results have been reported in Figs 9-22. Figures
10, 12, 14, 16, 18, 20 and 22 indicate the cycle number versus the
maximum vertical deflection reached for the seven specimens whose
hysteresis curves are shown in Figs 9, 11, 13, 15, t7, 19 and 21 respective-
ly. The cycle numbers have also been indicated on the load deflection
hysteresis. Details of all the 18 test results have been reported in Refs 7
and 8.

6 CONCLUSIONS

From the tests conducted the following conclusions are drawn:

1. The behaviour of all 18 test specimens showed consistent load
capacity in that the predicted toad was always more than the
observed, because of initial imperfections. The overestimation of the
load by the theory was within 20%.
2. The observed and predicted strain values correlated well (within
1(~20%),
3. The load reversal causes stiffness degradations irrespective of the
type of load transfer (through shear centre or through gusset plate).
However, the gusset plate loading causes a reduction in load-carry-
ing capacity after three cycles in the case of beams.
4. Two-point loading generally appears to be more severe than a single
concentrated load. There is a load reduction as well as instability of
the hysteresis loops.
5. Loading through the unsymmetrical axis (minor axis) causes severe
buckling and limits the energy absorption capacity.
6. Compared to beams, beam-columns have stable hysteresis loops. The
load reduction with cycling witnessed in beams is not apparent in
beam-columns.
7. In the case of uniaxially loaded beams and beam-columns the energy
loss calculated based on the area enclosed by the hysteresis loop is
less than 10% for cycles after the first three cycles.

ACKNOWLEDGEMENTS

The work was carried out in the Strength of Materials Laboratory,

College of Engineering, Guindy, Anna University. The authors ack-
nowledge the facilities provided by the authorities of Anna Uni-
versity.
124 G. M. S. Knight, A. R. Santhakumar

REFERENCES

1. Jain, A. K., Goel, S. C. & Hanson, R. D., Hysteretic cycles of axially loaded
steel members. J. Struct. Div. ASCE, ST8 (1980) 1777 95.
2. Dux, P. F. & Kitipornchai, S., Inelastic beam buckling experiments. J.
Construct. Steel Res., 3(1) (1983) 3 9.
3. Kitipornchai, S., Torsional flexural buckling of angles: a parametric study. J.
Construct. Steel Res., 3(1) (1983) 27-31.
4. Madugula, M. K. S. & Kennedy, J. B., Single and Compound An.qle Members
Structural Analysis and Desiqn. Elsevier Applied Science Publishers, London,
1985.
5. Kitipornchai, S. & Lee, H. W., Inelastic experiments of angle and tee struts. J.
Construct. Steel Res., 6(1) (1986) 219 36.
6. Ray, S. K. & Madugula, M. K. S., Computer program for the design of
laterally unsupported angle beams. J. Construct. Steel Res., 12(1) (1989) 55 68.
7. Periyasamy, P., Investigation on the behaviour of single angle as beam and
beam-column. Thesis submitted for the partial fulfilment of the degree of
Master of Engineering in Structural Engineering, College ofEngineering, Anna
University, Madras, Jan. 1991.
8. tndumathi, M., Nirmala, P., Sudha, V. & Sujatha, B.~ Behaviour of single angle
beam-column elements subjected to fatigue. Thesis submitted for the partial
fulfilment of the degree of Bachelor of Engineering in Civil Engineering,
College of Engineering, Anna University, Madras, Apr. 1991.