158

THE BEHAVIOUR OF R E I N F O R C E D C O N C R E T E BEAMS

UNDER CYCLIC LOADING
R.C. F e n w i c k * and A. Fong **

SYNOPSIS
The behaviour of beams in which plastic hinges are formed under
cyclic loading is examined. The results are reported of five beam tests,
in which the shear stress level was varied. It is shown that even
relatively low shear stress levels have a significant influence on beam
behaviour. Two main effects of shear are to reduce the ability of the
hinge to dissipate energy and to reduce the stiffness of the beams at low
load levels.. The degradation in shear under cyclic loading is accompanied
by an appreciable growth in length of the beam.

1.0 INTRODUCTION found in frame buildings it was concluded

that a cyclic shear stress in excess of
To ensure ductile behaviour under seismic 0.29/f£ in the plastic hinge zone could be
conditions most modern multistorey frame expected to lead to a significant reduction in
structures in New Zealand are designed to form the energy dissipated in the hinge. This
plastic hinges in the beams. To prevent collap degradation arises as the deformation in
these hinges must be capable of withstanding shear is associated with only a small energy
high inelastic rotations, usually in both absorption.
directions, and of dissipating an appreciable
amount of energy without suffering appreciable From a review of previous tests Paulay
strength degradation. (* * ) proposed design limits for shear stress
4

levels in beams subjected to reversed hinging

Several series of tests have been made under seismic conditions. The most recent
to assess the performance of beams under of these proposals, which has been incorporated
seismic conditions. Under the reversed loading in the draft N.Z. code' ', limits the shear
4

involving inelastic extension of the rein- stress level "v " in conventionally reinforced
u

forcement (which is typical of seismic load- beams to:

ing) , failure has frequently been found to
occur in a different manner from that in a v < 0.3 (2 + r) ,/fI (1)
beam subjected to monotonic loading. Tests n — -
v

have shown that the shear resisted by the

concrete decreases under these such conditions, where r is the ratio of the shear forces
and a diagonal tension type of shear failure under positive and negative flexural hinging
may occur unless adequate web reinforcement at the section being considered, and its
is provided( / , ) . in assessing the area
1 2 3
limits are 0 and -1. The shear stress is
of stirrups required to prevent this form calculated assuming a capacity reduction
of failure the shear resisted by the concrete factor of 0.85, and allowance is made for
in the region of the beam where the flexural the likely overstrength and strain hardening
tension steel yields, should be taken as of the flexural reinforcement (2 5% for steel
zero. In addition, allowance should be made with a yield stress of 275 MPa).
for the greater shear which may be sustained
in the beam due to the yield point of The quantity and arrangement of web
flexural reinforcement being greater than the reinforcement in a beam has been found to
specified minimum value (overstrength) , and have a very significant influence on its
to its subsequent strain hardening. However, behaviour under reversed loading. As
even in beams in which the web reinforcement previously noted it has been recommended
meets these requirements a different form of that the web reinforcement is proportioned
shear failure may occur in the plastic hinge to resist all the shear thus preventing a
zone. This has been described as a sliding diagonal tension type of shear failure. In
shear failure^ ' ). 1 5
addition to this it has been found necessary
to detail the stirrups at close centres to
To prevent sliding shear failures hold the flexural reinforcement from buckling.
occurring in coupling beams between shear With yielding of the steel the deformed bars
walls Paulay^ ' recommended that the shear
1
are displaced through the concrete towards
stress level should not exceed 0.5/f unless c the adjacent flexural cracks. The deforma-
diagonal reinforcement was used in the web. tions move through the concrete breaking it
In assessing the shear stress level in these up. When this zone is subjected to compression
beams, allowance should be made for the over- the bars yield back through the concrete
strength and strain hardening of the flexural causing further disruption and accelerating
reinforcement. From a series of tests carried the tendency for the cover to spall off.
out at Berkeley ' ^ on beams typical of those The reinforcement softened by the reversed
loading (Bauschinger effect) and no longer
confined by the concrete is much more suscept-
* Senior Lecturer in Civil Engineering Dept., ible to buckling than in a monotonically
University of Auckland. loaded beam. The improved performance of
** Engineer with Auckland City Council. beams where the flexural reinforcement has

B U L L E T I N OF T H E NEW Z E A L A N D N A T I O N A L SOCIETY FOR E A R T H Q U A K E E N G I N E E R I N G , V O L . 12, NO. 3 J U N E 1979

 159

been adequately restrained by stirrups has leaving a clear gap round each stud.
been demonstrated by tests(5,6)^
Details of the test beams are shown
Closely spaced stirrups improve the in Fig. 1 and further information is given
confinement of the concrete. Comparative in Table 1.
tests have shown that where such reinforce-
ment is combined with some longitudinal 3.0 LOADING SEQUENCE AND MEASUREMENTS
bars and cross ties in the mid regions of
the web a further significant improvement In all the beams except 1A and 2B the
occurs in the shear performance of the hinge tests were initiated by loading to three-
zone due to the improved confinement of the quarters of the theoretical yield flexural
web(6). capacity in each direction for two complete
cycles (see Fig. 2 ) . In the first beam,
The use of diagonal reinforcement in a 1A, a malfunction of the load measuring
beam has been shown to be a very effective equipment resulted in the beam being over-
means of reducing degradation under high loaded to the extent that the steel
shear stresses( > ' > . However, the added
1 2 1
yielded in the first half cycle. Consequently
complexity involved in the steel placing the test was taken straight to the 2D cycles
detracts from this solution. In this paper described in the next paragraph. Due to
only the performance of conventionally rein- error beam 2B was initially cycled at load
forced beams is considered, where the rein- levels close to the theoretical ultimate.
forcing bars are restrained against buckling Some yielding occurred in the second of
at centres not exceeding the smaller of six these cycles. In all the beams the
bar diameters nor 100 mm centres, and the deflection was monitored at a point located
web reinforcement is proportioned to resist at a distance of 1100 mm from the beam
all the shear (making allowance for over- springing. In the first two (1A and IB)
strength and strain hardening of flexural the displacement at this reference point
reinforcement). was plotted against load and the deflection
corresponding to the three quarter theoretical
2.0 EXPERIMENTAL WORK yield load level was assessed. This was then
scaled, assuming linear behaviour, to
Early in 1977 a research project was give the theoretical displacement at the
started at the University of Auckland to calculated yield load. The scaled value,
study the behaviour of beams under cyclic which was assessed as 6 mm, was taken as
loading. The experimental work was in two the ductility one displacement (ID), and
parts. In the first part, five beam tests it was used for all beam tests. Reanalysis
were carried out to investigate the influence of the data indicates that this displacement
of shear stress level on the performance of was approximately 10 per cent too high.
plastic hinge zones. In the second part the
effect of a different reinforcement grade After the first load controlled cycles,
was studied as well as various methods of displacements were imposed on the beams.
improving the performance in shear. In this At each level the beam was displaced in
paper only the results from the first series both directions for two complete cycles,
of tests are presented. see Fig. 2. Displacement ductilities of
2, 4, 6 and 8 were used, corresponding to
In the beams the shear stress level was 12, 24, 36 and 48 mm deflections at the
systematically varied by changing the span reference point 1100 mm from the beam
to depth ratio. The flexural steel arrange- springing. These load stages are subsequently
ment was kept constant in all beams except the referred to as the 2D, 4D, 6D and 8D
last (3A) in which the area of bottom steel ductilities. The + and - signs used in
was reduced, see Fig. 1. the figures refer to the direction of loading,
and the i and ii symbols refer to the first
The test beams were made in units, each and second cycles. When the first two 8D
unit consisting of two cantilever beams cycles had been completed either two further
springing from a central concrete block cycles of 8D were applied followed by two
representing a stiff column. This block was 10D cycles or the tests went straight to
stressed down to the strong floor with the 10D level. The loading sequence is
Macalloy bars enabling the beams to be tested illustrated in Fig. 2.
independently. The reinforcement in the two
beams was continuous through the central block. Each beam test lasted about six days.
To prevent progressive yielding in this zone Numerous strain and displacement measurements
two 10 mm reinforcing bars were welded to were made to enable the following values to
each continuous bar of flexural beam rein- be determined:
forcement (see Fig. 1 ) .
(1) The load deflection curves for the beams
In all beams mild steel reinforcement at the reference point (1100 mm from springing),
was used which had a minimum specified yield and at the load point,
strength of 275 MPa. The stirrups were (2) The strain distribution in the flexural
designed to carry all the shear and to hold reinforcement at the peak beam displacements
each bar of flexural reinforcement at 100 mm (2D, 4D, etc.),
centres against buckling. The concrete was (3) The shear deformation in the beam, and
purchased from a local ready mix firm and (4) The changes in length of the beam.
was required to have a minimum strength of
30 MPa at 28 days. To enable flexural To relate the results of these tests
strains in the reinforcement to be measured to a typical structure it is necessary to
by a 100 mm mechanical strain gauge, 30 mm have some idea of the relationship between
long steel studs were welded to the rein- the ductility measured in these tests and
forcement. Before casting, plastic tubes the ductility displacement in a multistorey
were fitted over the studs, which were frame. Such a structure is illustrated in
withdrawn when the concrete had hardened Fig. 3. If a displacement ductility of 4
160

is imposed on the frame each bean column tension reinforcement during the elastic
unit (such as ABCDj, should be subject to load cycles are shown for the first four
approximately the same ductility demand. beam tests in Fig. 5. For beams 1A and IB
The elastic displacement of this unit arises which contained significant diagonal cracking
from deformation in both the beams and at this load level the measured steel strains
columns. However, the inelastic deformation are greater than those corresponding to
occurs only in the beams. Thus if a typical flexural theory alone, and they are in
beam is tested by itself a displacement reasonable agreement with the theoretical
ductility of 6 corresponds to a value of flexural value plus half the shear. In
approximately 4 in the equivalent beam column beams 2A and 2B, which did not contain major
unit, see Fig. 3. diagonal cracks the flexural strains are
generally just below the value from flexural
4.0 BEHAVIOUR OF BEAMS UNDER ELASTIC LOAD theory. This is due to the tension stiffening
CYCLES effect of the concrete, in which it carries
some of the tension force between the cracks,
In the elastic load cycles the cracks thus reducing the average strains. In
were nearly normal to the flexural reinforce- assessing the theoretical deflections an
ment in beams 2A and 2B. These beams had allowance for a shear lag tension force
shear stress levels of I.16 and 0.94 MPa increase of V/2 has been included for beams
respectively. However, in beams 1A and IB, 1A, IB and for the downward loading in beam
which had the higher shear stress levels of 3A. No allowance has been made in the other
2.21 and 1.56 MPa, the cracks were inclined beams.
at approximately 45 degrees. In beam 3A the
tension cracks were inclined on the upper The theoretical and experimentally
side of the beam which was reinforced with measured deflections are in reasonable
the 2 0 mm bars, but were near to vertical agreement with each other. Inspection of
on the side containing the 16 mm flexural the values for the reference point 1100 mm
reinforcement. from the beam springing illustrates the
importance of shear deformation on member
The deflections of the beams at the behaviour. For beam 1A with a theoretical
load and reference points at the 3/4D load ultimate shear stress level of 2.21 MPa,
levels, or the values extrapolated from the the calculated deformation due to shear
two sets of readings taken immediately below effects amounted to 37% of that due to
this level, are given in Table 2. The last flexure. However, in the beam with the
reading in the nominal elastic load cycles greatest span over depth ratio 2B (lowest
for beam 2B has been discarded as the measure- shear stress) the theoretical shear
ments show the flexural steel had yielded deformation is negligible and consequently
in the previous load application. The the total deflection is less than that of
measured deflections may be compared with the 1A in spite of the 18% greater theoretical
calculated values in the table. The flexural flexural deformation of 2B. The experimental
deformations were determined from section deflection measurements follow the theoretical
properties based on the cracked section. trends. The magnitude of the shear deforma-
The elastic modulus of the concrete was taken tion in the beams with the higher shear
as 25000 MPa (Draft N.Z. code( > value
4
stress levels makes it difficult to assess
25700 MPa). To allow for the deformation both the displacements corresponding to a
of the flexural reinforcement in the central ductility of one and its significance.
block the beam length was taken as the free
length plus 200 mm. This additional length 5.0 BEHAVIOUR OF BEAMS UNDER INELASTIC LOAD
was chosen as it was reasonably consistent CYCLES
with the additional length required to give
the same rotation as the pull out of the In the displacement controlled cycles,
reinforcement from the central block. The yield of the reinforcement caused the cracks
rotation of the beam supports and central to open up. After a few cycles they remained
column block was not monitored during the open over the full depth of the beam. In
test, and it cannot be accurately calculated. the beams with the higher shear stress
The value listed in the table for this levels the concrete spalled off the rein-
represents the best estimate the writers forcement in the 6D cycles. With the lower
can make. shear stress levels this was delayed to the
end of the 8D cycles. Spalling was followed .
Shear deformation in a beam consists by the breaking up of the concrete in the
of two components, the first of which is web of the beam. After a number of further
caused by the deformation of the stirrups and load cycles the concrete disintergrated and
the concrete in diagonal compression. A fell out of the web, and this was accompanied
method of calculating this deformation is by a sharp decrease in the load resisted by
given in reference 1. The second component the beam.
arises from shear lag in the tension zone
increasing the flexural tension forces. This The relative performance of the beams
occurs with diagonal cracking, and it is may be assessed visually from the load
illustrated in Fig. 4. It may be seen from deflection curves; the fatter the loops the
the free body diagram in this figure that the more energy is dissipated by the hinge and
increase in the flexural tension force the better is the performance. The load
depends upon the length of the diagonal crack deflection curves for the beams with the
and the magnitude of the stirrup forces. If highest and lowest shear stress levels are
the longitudinal projection of this crack is shown in Fig. 6. A comparison of these
taken as the flexural leverarm jd, then with shows that the higher shear stress level
no shear resisted by stirrups the increase resulted in a more pinched load deflection
in flexural tension force is equal to the characteristic and it reduced the number of
shear V. However, if all the shear is cycles sustained by the beam before serious
resisted by stirrups the increase is reduced stiffness degradation occurred at low load
to V/2. Strains measured in the flexural levels and appreciable strength degradation
 161

T A B L E 2: MEASURED AND COMPUTED DEFLECTIONS AT LOAD AND

TABLE 1: DETAILS OF TEST BEAMS REFERENCE POINTS D U R I N G ELASTIC LOAD CYCLES

Measured deflections at reference point.

I Stirrup b ar Theoretical Values'^
Beam
Beam MPa MPa diameters v a
P
u° v
v
Load
kN Cycle 1A IB 2A 2B 3A
a b c kN kNm KPa

+37 4 i 4.0 4.9 3.6 3.7 2.2

IA 290 30* 923 10 6 6 294 183.6 198.9 2.21 0.41
-3/4i
+3/4ii
- -4.2
4.9
-3.6
3.7
-3.5
3.8 2.4
-4.0

IB 290 30* 1 329 6 10 6 233 183.6 138.1 1.56 0.29 -4.4 -3.6 -3.6
-3/411 ~

2A 280 34.7 1 735 6 6 6 172 178.4 102.8 1.16 0.19 4.0 3.7
Average 4.6 3.6 2.3 3.80

2B 280 34.7 2 142 6 6 6 172 178.4 83.3 0.94 0.16

Measured deflections at load point.
3A top 280 27.7 1 329 6 10 6 233 174.9 131.6 1.49 0.28
+3/4i 3.50 5.9 5.5 8.8 2.8
bot 29B 122.9 92.4 1.05 0.20 -3/41
- -4.9 -5.2 -8.0 -4.4
+3/4ii
-3/411
- 5.8
-5.4
5.7
-5.8
8.5 2.9
-4.4

Average 3.50 5.5 5.6 8.4 2.9 4.4

$ Theoretical values using Whitney stress block for flexural calculations Calculated deflections at reference point

* Test cylinders incorrectly cured f^, taken as 30 M P a flexure 2.50 2.76 2.81 2 94 2.52 2.80

-- -
shear I 0.42 0.20 0.19
a Dead load of beam neglected shear 2
fixity
0.50.
0.36
0.35
0.32 0.30 0 28 -
0.22
0.33
0.31
V fl Shear capacity of w e b reinforcement
Total 3.78 3.63 3.11 3 22 2.74 3.63
f Yield of flexural reinforcement
y
Calculated deflections at load point
Cylinder strength of concrete.
flexure 2.03 3.55 5.35 7.70 3.25 3.60

- -
See Fig. 1 f o r further d e t a i l s . shear I 0.42 0.24 _ 0.19

- -
shear 2 0.36 0.51 0.48
fixity 0.32 0.37 0.42 0.47 0.25 0.36

Total 3.13 4.67 5.77 8.17 3.50 4.63

flexure » flexural deformation

shear 1 • shear deformation d u e to web members

shear 2 « deformation due to shear lag

LOADING FRAME fixity - deformation due to rotation of unit supports.

(a) Test Arrangement

2-10mm BARS WELDED TO

E A C H BAR O F F L E X U R A L '
STEEL _ ,, 50

FIGURE 2: LOADING SEQUENCE

200

} 5-20-mm. ALL BEAMS

SEE TABLE 1 FOR

FURTHER DETAILS

5-20mm IN 1A.1B.2A.2B
5-16 m m IN 3A

STR SET
SECTION 1-1
(b) Reinforcement Details

FIGURE 1: TEST ARRANGEMENTS A N D BEAM DETAILS

 EQU1L1 B R 1 U M REQUIREMENTS.
DUCTILITY

BEAM COLUMN UNIT

/////////////// //;s/
Tjd=M a + Ve-V /2 s
e

REGULAR FRAME
FOR V=Vs A N D e = jd

a = ELASTIC DISPL DUE TO COL. DEFORMATION + v T = M a / j d + V/2, C = M / j d - V / 2

b

b =ELASTIC DISPL. DUE TO BEAM DEFORMATION

c = INELASTIC DISPL. DUE TO BEAM DEFORMATION WHERE M = b . m . AT A
0

Mb=b.m. AT B

F I G U R E 3 : D I S P L A C E M E N T D U C T I L I T Y IN A F I G U R E 4 : SHEAR LAG DUE T O D I A G O N A L

BEAM AND A B E A M COLUMN UNIT CRACKING

FLEXURAL

1000 BEAM 1A
S H E A R - 1 6 5 k N 1st CYCLE

500
FLEXURAL ^ \ \
THEORY \

800

BEAM 1B
LOAD STAGE - ^ i i
SHEAR - 1 0 4 k N

< - 0 -

BEAM 2 A
LOAD STAGE* ^4 ir
SHEAR 71 kN

LOAD STAGE «• D1
1000 SHEAR 7 6 k N

(b) Beam 2B
2000
400 800 1200 1600
DISTANCE FROM BEAM SPRINGING FIGURE 6 : LOAD D E F L E C T I O N C U R V E FOR
R E F E R E N C E P O I N T FOR B E A M S 1A A N D 2B
F I G U R E 5: S T R A I N D I S T R I B U T I O N IN T E N S I O N
REINFORCEMENT IN ELASTIC LOAD C Y C L E S
 163

120

10
ENER6Y INDEXr-Ve
Ixi
_ 0 7 5

& 0-5
DC
LU
0-25
Ul
ACTUAL ELASTO-FLASTIC,
0 -
2D 4D 6D 8D 8D 10 D
(10 D}
1st CYCLE - LOAD CYCLES

1-0-
x
LU
o 075
z -2A
o 0-50
or
LU

5 0-25 L 1 A •
2D ID 6D 8D 10D
0- 80 120 160 200
2nd C Y C L E - LOAD CYCLES
ELAST0 P L A S T I C ENERGY (THEORETICAL) k N m

F I G U R E 7: V A R I A T I O N OF E N E R G Y INDEX F I G U R E 8: E N E R G Y D I S S I P A T E D IN B E A M S
W I T H LOAD CYCLE C O M P A R E D W I T H E Q U I V A L E N T ENERGY IN
ELASTO-PLASTIC BEAM

F I G U R E 10: F L E X U R A L S T R A I N S - B E A M 1A
164

occurred at the maximum displacements. The prematurely, increasing its susceptibility

load deflection curves for the beams with to buckling.
intermediate shear stress levels follow these
trends. During the tests the beams were
observed to grow in length. This was
To allow a numerical comparison of the accompanied by an increase in the shear
different beams to be made a value which is deformation. The mechanism of growth is
referred to as the energy index was calculated. described below and illustrated in Fig.
This was defined as the area under the load 10, where the displacement of a section
deflection curve in a particular cycle located 250 mm away from the springing in
divided by the area sustained by a beam behaving beam 1A is plotted at different cycles.
in an elasto plastic mode. In Fig. 7 the Under the first half cycle of loading
variation of energy index with load cycle is (+2Di) the bottom steel yields in tension
shown. It may be seen that the higher the and small compressive strains develop in
shear stress level the smaller the energy the top steel because the compression force
index; that is the load deflection loop is is shared with the concrete. Under the next
more pinched in form. Fig. 8 shows the half cycle of loading (-2Di) the top steel
actual energy dissipated by the beam plotted yields in tension and the bottom steel goes
against the theoretical elasto plastic value. into compression. However, not all the
This clearly indicates the poorer energy tensile strain in the steel is recovered
dissipation in the beams with the higher shear and the cracks do not fully close. When
stress levels. the reinforcement yields, local slip occurs
close to the main cracks. The deformations
In Fig. 9 the variation in the load on the bar plough through the concrete pulling
sustained at peak displacement for the differ- broken material into the cracks. In the low
ent load cycles is shown. The higher shear load stages of the next half cycle the
stress levels may be seen to reduce the number cracks are open right through the beam and
of cycles sustained by the beam before strength all the shear is resisted by interface
degradation occurs. This figure shows that as friction. Hence a shearing force is applied
the test progressed the load sustained by to the broken concrete surrounding the
the beams increased above the theoretical reinforcement. The disruption caused by this
ultimate level based on the yield strength prevents the material from fitting back into
of the reinforcement. The peak increase was its initial location. Consequently the
between 18 and 24 per cent for all beams, cracks remain wedged open, and as the beam
and it was sustained in the D6 or D8 load is cycled it progressively grows in length
cycles. The average increase in strength with the two sides of the cracks being ground
above the theoretical load in all the D6 against each other. This produces more
cycles for all five beams was 15 per cent. debris to fill the crack.

Beam 3A differed from the other beams The flexural deformation of the hinge
in that the area of bottom steel was reduced zone in the presence of shear causes the
to 67% of the area of top steel. Except beam to grow in length. The opening of the
for this difference it was essentially the same cracks right through the beam reduces the
as beam IB. The different flexural strengths shear stiffness of the member, and consequently
in each direction lead to the beam being this growth is accompanied by an increase in
subject to different peak shear stresses shear deformation. The beam length may
depending on the direction of loading. stabilise for two different reasons. Firstly
Inspection of Figs. 7, 8 and 9 show that the the compression reinforcement may buckle,
performance of beams 3A and IB were almost leading to a shortening of the compression
identical. This was an unexpected result zone and hence the beam. Secondly, if the
as it is contrary to the suggested design shear deformation increases to a sufficient
rules in the draft N.Z. code as quoted in extent the required flexural deformation
equation 1. For the test condition this reduces, and hence the driving action for
expression gives an allowable shear stress growth is removed. Fig. 11 shows the way
level of 1.64 MPa (0.3/f£) for beam IB as the in which the beam increased in length with
value of r is equal to -1. The corresponding cycling. The maximum growth was reached in
value for beam 3A is 2.05 MPa (0.39/fJ) with the 6D to 8D cycles, and it varied from 13
r = -0.7. The actual shear stress levels to 19 mm in magnitude.
for the two beams were 1.56 and 1.49 MPa for
IB and 3A respectively. As beam IB was much Fig. 12 shows the proportions of
closer to the proposed limit than 3A its flexural and shear deformation sustained
performance could have been expected to be in the test beams at different load stages,
poorer than 3A. This was not the case. Shear deformation is important in all the
f 6) beams. At the end of the 6D cycles it
accounted for approximately 20 per cent of
Comparative beam tests at Berkeley the deformation in beam 2B which had the
have shown that reducing the flexural steel lowest shear stress level (0.94 MPa), while
on one side of the beam has two effects. for beam 1A which had the highest shear
Firstly the shear deformation in the member is stress level (2.21 MPa) it accounted for
slightly reduced due to the smaller shear 7 5 per cent of the deformation. The
forces which may be sustained in one direction, performance of beam 3A with the unequal top
and secondly the flexural performance of the and bottom steel was very similar to its
member is reduced as the smaller area of comparison beam IB which had equal flexural
steel is more susceptible to buckling. This steel areas.
area of flexural reinforcement tends to be
yielded in tension and compression much more Fig. 13 shows the relationship between
than the larger area, and consequently its the load and shear deformation for beams
stiffness is reduced by the Bauschinger 2B and 3A. These diagrams are similar in
effect. This combined with the requirement form to those reported in the literature
that part of the compression force must be (2,6,7) ^ Most of the shear deformation
carried by the concrete causes it to spall
 165

F I G U R E 13: SHEAR D E F O R M A T I O N V E R S U S LOAD

FOR B E A M S 2B A N D 3A
166

occurs at low load levels, which gives rise the lever arm of the flexural forces, k is
to the characteristic shear pinching of the the distance from the centroid of the tension
hysteresis loops, and reduces the energy steel to the neutral axis and f is the
which may be dissipated by the hinges. yield point of the reinforcement. The super-
scripted values apply to the reinforcement
Stirrup strains were not measured on the other side of the beam. The values
during the tests, but the distance between of jd and k are found from standard flexural
the studs welded to the top and bottom theory at ultimate load conditions.
reinforcement was monitored. Changes in
this distance may arise from the yield of These regression equations were adjusted
the stirrups and local buckling of the tc pass through the common ordinate of 1.30,
flexural reinforcement. The growth in the and a further regression analysis was carried
depth of the beams 2A, 2B and 3A is shown out for the coefficient of E against the
in Fig. 14. The stirrups appear to have average maximum shear stress without regard
yielded after the plastic hinge was well to sign. The resulting equation for the
established at the location being examined. shear deformation in this series of beams
The magnitude of the growth in the beam reduces to:
depth is even more surprising than the
growth in length of the beams. 1.15 v
S = — E + 1.30 (3)
6.0 ANALYSIS OF RESULTS FOR INELASTIC
CYCLING OF BEAMS . *Z
The magnitude of shear deformation Before this expression may be applied to
sustained by a beam could be expected to be other beams it should be adjusted for the
a function of the damage sustained and the effects of beam depth.
shear stress level. The damage may be
assessed by the dissipated energy. As most The expression given above is a tentative
of the shear deformation occurs at low equation which can only be expected to
load levels a regression analysis was carried indicate the order of shear deformation which
out for each beam in which the shear deform- may be expected. In its derivation many
ation in the plastic hinge zone was compared factors have not been considered. In part-
with the dissipated energy to the end of icular the quantity of web reinforcements
the previous cycle. The hinge shear deform- (which appears to yield in these tests) could
ation was taken as the total measured shear be expected to exert a considerable influence
deformation minus the extrapolated shear on the results, as a change in the area of
deformation corresponding to the ID displace- these would change the magnitude of the con-
ment. The regression analyses for the five fining force. The degree of confinement of
beams gave the following equations: the web concrete could also be expected to
have an appreciable effect(6).
= 0.41/F "
1
Beam 1A V r = 0.94
c 7.0 CONCLUSIONS
u
S = 0.46E + 2. 50 (2i)
(i) The test results show that shear in
Beam IB V = 0* 29/F " 1
r = 0.98 plastic hinge zones formed under reversed
u c loading conditions has a marked effect on
S = 0.30E 1. 32 (2ii) member behaviour. In the test beam 2B,
which had the lowest shear stress level
Beam 2A V = 0. 19/fJ r = 0.99 (0.16/rj") , the shear deformation at the
u local point amounted to 20% of the total
at the end of the 6D cycles. For test
S = 0.25E + 1. 15 (2iii beam IB (0.29/fj)the equivalent proportion
of shear deformation was 4 5%.
Beam 2B V = 0.16/f " 7
r = 0.99
c
u (ii) The experimental results show that as
S - 0.18E + 1. 06 (2iv) the shear stress level increases the perform-
ance decreases. To find a rational cut off
= 0. 28/F" 0. 20/r " r = 0.98
1
Beam 3A V point for design further work is required.
c c This would involve more detailed analysis
u
S = 0.28E 0. 40 (2v) of other published test results together
with inelastic analyses of structures in
where S is the shear displacement in ram, which the shear degradation effects are
r is the coefficient of determination which modelled. However, pending such a study
is a measure of the closeness of fit of the some indication of a suitable cut off shear
points to the equation, and E is the dissipated stress level may be gained from the results.
energy in energy units as defined below. An Inspection of Figures 7 and 8, which show
energy unit (IE) is the energy dissipated by the energy performance, and Figure 9, which
an effective length of beam at theoretical gives the strength degradation, show a
;

ultimate load conditions for moment under gradual decrease in performance as the
monotonic loading when a moment is applied shear stress increases from 0.16/fJ to
first in one direction and then in the other 0.29/fJ, but there is a marked decrease
direction to give one additional yield strain from this level to the test at 0.4l/fJ.
in the flexural tension reinforcements. The This suggests a cut off level of about
value of one unit of energy is given by: 0.29/f^. The proposed limit in the draft
N.Z. code( ) of 0.3/f for the case of equal
4
c

f . , f 1
shears in each direction fits in with this
observation.
Vy J -(1-k)- E s +
Vy J °(l-k)E s
(iii) The performance of beam 3A, with the
where A g is the area of tension steel, jd is unequal flexural steel areas, was not as
 167

good as was expected from the draft code 6. Ma, S. M., Bertero, V. V. and Popov, E.P.,
expression which gives the limiting shear "Experimental and Analytical Studies
stress levels in beams without diagonal on the Hysteretic Behaviour of Rein-
reinforcement (equation 1 ) . However, this forced Concrete Rectangular and T-Beams",
was only one test. Further tests where the Report No. EERC 76-2 May 1976, University
shear force in each direction is different of California.
is required to clarify this aspect. 7. Bull, I. N., "The Shear Strength of
Relocated Plastic Hinges", Research
(iv) All beams grew appreciably in length Report 7 8/11 Feb. 197 8, Department of
under cyclic loading (see Fig. 1 1 ) . This Civil Engineering, Canterbury University.
could have important implications for the
performance of frame structures under seismic
conditions. This growth must be resisted This paper was presented at the South Pacific
by slabs and columns. The magnitude of the Regional Conference on Earthquake Engineering
axial load induced in the beams due to this held in Wellington on 8, 9 and 10 May, 197 9.
restrain depends upon its axial stiffness.
The expansion of the beams appears to be
largely due to the wedging action of dis-
placed aggregate particles in the cracks.
If the expanded beam is not stiff then there
would be little axial load induced and little
growth in an actual frame structure. Under
these conditions the small axial load would
give a small increase in the moment capacity
of the section. However, as the growth is
limited wide cracks could not form right
through the beams as in these tests, and
consequently the sliding shear strength
could well be considerably greater than was
observed in the work (such as described in
this paper) where there is no restraint to
growth.

If the aggregate wedging action gives

a stiff beam for axial load as it grows then
the restraint would not prevent the growth.
Consequently the columns would be pushed
out from the centre of the building increasing
the P-A effect in these. In addition as
high axial loads would be induced in the beam
they could be expected to give a considerable
increase in the moment capacity of the beams.
This would increase the likelihood of a column
sway failure mechanism.

ACKNOWLEDGEMENTS

The writers are grateful to the MWD

who provided financial assistance for the
project. M r . Ian Armstrong of the MWD
Auckland District Office gave valuable
assistance in setting up this project, and
Mr. Neil Metcalf and his staff from the
same office very kindly assisted by tracing
the diagrams in this paper.

REFERENCES

1. Paulay, T., "Design Aspects of Shear Walls

for Seismic Areas", Research Report 74-11,
Civil Engineering Department, University
of Canterbury, (Or R. Park and T. Paulay,
"Reinforced Concrete Structures", Chapter
12. Published John Wiley and Sons, 1975.)
2. Bertero, V. V., and Popov, E. P.,
"Hysteretic Behaviour of Ductile Moment-
Resisting Reinforced Concrete Frame
Components", EERC 7 5-16 April 197 5,
University of California, Berkeley.
3. Paulay, T., "Shear Strength Requirements",
Bulletin of the N . Z . Soc. for Earthquake
Engineering, Vol. 10, N o . 2, June 1977.
4. S.A.N.Z. Draft N . Z . Standard - "Code of
Practice for the Design of Reinforced
Concrete Structures", DZ 3101 Pt. 1,
1978 - Standards Assoc. of N . Z .
5. Brown, R. H. and Jirsa, J. P., "Rein-
forced Concrete Beams Under Load
Reversals", Proceedings ACI V o l . 68,
No. 5, May 1971.