Sei sulla pagina 1di 11

EARTHQUAKE ENGINEERING A N D STRUCTURAL DYNAMICS. VOL.

15,993-1003 (1987)

REINFORCED CONCRETE HYSTERESIS MODEL BASED ON THE


DAMAGE CONCEPT

MING-LIANG WANG
Department of Civil Engineering, University of New Mexico, Albuquerque, N M 87131, U.S.A

AND

SURENDRA P. SHAH
Department of Civil Engineering, Northwestern University, Evanston, I L 60201, U.S.A.

SUMMARY
Reinforced concrete members subjected to cyclic inelastic deformation may exhibit both stiffness and strength
degradation, depending on the maximum amplitude and the number of cycles experienced by the member. Many of the
currently available models do not simulate the cycle-dependent stiffness loss often observed during the experiments.An
analytical model based on a damage parameter which is a function of the cumulative cyclic inelastic deformation is
proposed. A small-scale reinforced concrete beam<olumn joint was constructed and subjected to cyclic loading to
calibrate the damage parameter. The model is able to approximate the complete hysteretic response and is simple to
implement. The sensitivity of the analyticalmodel was examined by comparing the model with several experimentalresults.

INTRODUCTION
During strong ground motions members of reinforced concrete structures undergo cyclic deformations and
experience permanent damage. Several hysteresis models have been proposed to predict the response of
reinforced concrete members subjected to cyclic loading.'-3 These range from relatively simple models
proposed by Clough' and Mehdi' to more complicated models such as those proposed by Takeda et aL3 These
models generally assume that stiffness degradation is a function of the maximum amplitude of cyclic
deformation experienced by the member but does not depend on the number of cycles. However, experiments
show that the stiffness ofa member may continue to decrease when subjected repeatedly to the same maximum
displacement. This is shown in Figure 1, taken from Townsend and Hansom4 It is likely that this cycle-
dependent reduction in stiffness of a member may significantly influence the response of the ~ t r u c t u r e . ~
Therefore in modelling hysteretic loops for dynamic inelastic analysis it may be important to simulate this
cycle-dependent stiffness degradation.
Many attempts have been made to develop theoretical models based on the knowledge of the cyclic
behaviour of concrete, steel and the interface between concrete and steel. An analytical model that is based on
this relationship has been proposed by Filippou et aL6*'. Damage was determined by the total energy
dissipated during each cycle of inelastic deformation. This model was shown to be capable of predicting the
hysteretic behaviour of interior and exterior R/C beam-column joints including the effects of strength and
stiffness deterioration. However, the extensive computational effort involved makes this method relatively
expensive.
In this paper, a hysteretic model based on the concept of accumulated damage is proposed. The concept of
damage has been used to predict response of reinforced concrete structures subjected to fatigue loading,', to
identify hysteretic behaviour of existing structures''-' and for analysing the behaviour of the bond between
reinforcing bars and concrete for structures subjected to seismic e ~ c i t a t i o n . ' ~

0098-8847/87/080993-11$05.50 Receiced 18 August 1986


((2 1987 by John Wiley & Sons, Ltd. Revised 10 February 1987
994 M.-L. WANG A N D S. P. SHAH

-75.0 -50.0 -25.0 0 25.0 50.0 71 0


displacement- mm

Figure 1. Hysteresis diagram for Specimen 10 of Reference 4

The predictive model presented here is simple in nature and as a result it is neither comprehensive nor
precise. However, the approach presented is promising and amenable to further refinement. Only external
beam-column joints are studied. The paper consists of (1) presentation of the basic theoretical approach, (2) a
description of the calibration of the damage parameters based on a small-scale model beam-olumn joint and
(3) an examination of the predictive capability of the model and an examination of the sensitivity of one of the
damage parameters.

CUMULATIVE DAMAGE CONCEPT


Damage to a reinforced concrete member or an assemblage subjected to cyclic loading is assumed to depend on
the maximum displacements experienced during each cycle. This cyclic-loading parameter is termed p and is
defined as

where X i= maximum displacement in the ith cycle, X , is failure displacement under monotonically increasing
loading and C is a constant less than one. Note that only the displacements occurring in one direction are
accumulated. For unsymmetrical loading or cross-sections, a separate value of p for each direction must be
used.
A non-dimensional and non-negative scalar quantity D is introduced and is termed a damage parameter
whose value varies between 0 and I. The damage parameter is a function of loading parameter p such that
D =f(B);f(O) =O;f(l)= 1 (2)
Note that D is a state variable which changes with 8. Assuming for simplicity that the rate of change of
damage (dD/dp) varies linearly with D, the following damage accumulation relation can be developed:
dD
- = nD (3)
dB
where n is a constant to be determined from experimental data. The solution of the differential equation ( 3 )
REINFORCED CONCRETE HYSTERESIS MODEL 995

P
Figure 2. Damage parameter D vs. cumulative damage variable /? with variable n values

with the initial conditions specified in equation (2) is


enp- 1
D = f ( j ) = ___ (4)
en- 1
A plot of damage parameter D versus the cumulative variable for various values of n is shown in Figure 2.

ANALYTICAL MODEL
It is assumed that a single-degree-of-freedom reinforced concrete structure or a component has a unique
monotonic load-displacement curve in each direction of loading before any damage has occurred. For
simplicity a bilinear skeleton curve is assumed (Figure 3). Note that FYand Xy are the load and displacement
corresponding to the yield point, K , = F , / X , is the initial stiffness and aK, is the post-yield stiffness.

Figure 3. Proposed model loadclisplacement curve


996 M.-L. WANG A N D S. P. SHAH

The damage due to cyclic loading is expressed in terms of the decrease in the yield-load carrying capacity.
The reduction in the yield strength (from point B to G in Figure 3) can be written as
F,(Xi)= F ( X i ) ( l - D ) (5)
where Xi is the maximum displacement the structure has experienced during the ith cycle, and F (Xi)is the load
corresponding to the displacement Xi on the skeleton loadtiisplacement curve. Thus the yield-load carrying
capacity of the structure is reduced from F (Xi)
(point B) to F,(Xi) (point G) as shown in Figure 3, after it has
experienced damage D which will depend on the cyclic loading parameter 8.

HYSTERESIS RESPONSE
Based on equations (l), (4) and (5) a hysteresis model that can account for stiffness degradation and yield-
strength deterioration is proposed. The procedure for symmetrical loading and symmetrical cross-sections is
outlined below.
1. The model operates on a bilinear skeleton curve defined by F,, K, and CI.Once the section has yielded,
unloading is associated with the smallest stiffness observed in the same loading direction. Before the
completion of the first cycle the smallest stiffness observed is equal to K , . After that, the smallest stiffness is
defined by K ( X , ) ,which is the line connecting point G and C in Figure 3. The point G is determined using
equation (5) and for subsequent loading is the current value of the yield point, whereas the point C is
determined from Xi and K,.
2. When the load is reversed, the stiffness is determined by considering the slope of the line connecting the
load reversal point (point C or point F ) to the current value of the yield point (point D or point G).
3. When the loading is continued beyond the maximum displacement previously experienced in that
direction (such as Xi),the load-displacement response is determined by assuming the same post-yield stithess
as that for the skeleton curve (line G H in Figure 3).
4. If during any cycle, the maximum displacement is less than the current yield point, no damage is
accumulated during that cycle (point F' in Figure 3).
The proposed model depends on the following quantities: F,, K, and a needed for evaluating the skeleton
curves, and the quantities X,, C and to define the relative extent of damage.

CALIBRATION O F THE ANALYTICAL MODEL


The parameters needed for the proposed hysteresis model were obtained from the results of a small-scale
model of a reinforced concrete beamxolumn joint; this model specimen was about 1/4 scale of the prototype
member. The dimensions of the specimen are shown in Figure 4. The specimen was designed according to the
seismic design requirements of ACI 318-83 and recommendations of the ACI-ASCE Joint Committee 352.
The reliability of the modelling technique was established by comparison of model results with full-scale
tests.' The details of the experiment and the detailed results are given elsewhere.' One set of the response
results of the specimen as indicated by the load vs. load-point displacement relationship of the beam is shown
in Figure 4. The assemblage was subjected to the loading history shown in Figure 5. The test configuration is
shown in Figure 6. The required parameters for the analytical model were evaluated based on the experimental
data of the test specimens.
For the skeleton curve, the yield load ( F , ) and the corresponding deflection ( X , ) were deduced from the
experiments. The slope of the post-yield curve was taken as 10 per cent of the initial stiffness (a = 0.1). In order
to determine the parameter n, the drop in the yield load from the bilinear skeleton curve after each cycle of
loading is plotted in Figure 7. From the figure it can be observed that the value of n = 1 approximately
represents the relationship between the damage parameter D and the loading parameter j as postulated in
equation (1). A better representation of the damage parameter D can be obtained by assuming a different
damage accumulation relationship than that assumed here [equation (3)]. Note that the specimen failed ( D
= 1.0) at the end of 14 cycles as shown in Figure 4.
REINFORCED CONCRETE HYSTERESIS M O D E L

4.0-

3.0-

1
2.0-

a . D-2
b 9mm at lOmm
D-29mm 15mm stir&
1.0-
c D - 2 9 m m a t 23mm
Z
Y I"= 25 4mm

B- O-

-1.0-

-2.0-

-3.0-

A -A

-401 I 1 I I I I I I I
-200 -150 -100 -50 0 50 00
1 150 200
displacem~nt m m

Figure 4. Experimental loadtiisplacement curve

4.UI * 750.0!

-4.01
I I
-20.0
I
-150
I
-10.0
I
-5.0
tI
0
- 7500

5.0
I
I

10.0
I I
15.0
I
20.0
displacement - m m

Figure 5. Analytical load-displacement curve


998 M.-L. WANG AND S. P.SHAH

10.511'' i

Figure 6. Test configuration of the model specimen

Figure 7. Damage vs. cumulative damage variable /lfor the data shown in Figure 4 and determined based on parameters listed in Table I

There appears to be no exact method for determining the failure displacement X , of reinforced concrete
components, especially when shear deformation and bond slippage are dominant. However, since the yield
deformation X can be predicted with good accuracy, a practical means to determine the failure displacement is
Y.
to use the ductility factor p, X , = p X , . To calculate the failure deformation, 142 monotonically loaded beams
and columns were evaluated by Park and Anglo and a mathematical formulation was proposed. This approach
can also be used to calculate the values of X,. For simplicity, the value of X , was assumed to be five times the
yield displacement ( X , = 5X,). The predicted response of the model-joint specimen using the above values (n
= 1.0) and the value of C = 0.1 in equation (1) is shown in Figure 5.
Using these same values of the damage related parameters (XF, C and n), it was attempted to predict the
response of the comparable prototype specimen. The prototype specimen is shown in Figure 8 and was taken
from the study reported by Scribner and Wight.I6 Both the prototype and the small-scale model specimen had
comparable material properties of concrete and steel, design parameters, design details and casting and testing
procedure.' Based on the similitude requirements a force ratio of approximately 19.1 and a deflection ratio of
REINFORCED CONCRETE HYSTERESIS MODEL 999

Table I. Parameters used in the analysis for both model and


prototype specimens
Fy-kN Xy-mm X,-mm
Specimen Section (kips) (in) (in) n

TOP 2.89 5.08 25.4 1.0


Model (0.65)* (0-2) (1.0)
Bottom 2.89 5.08 25.4 1.0
(0.65) (0.2) (1.0)
TOP 53.31 1016 50.8 1.0
Prototype (12.0) (04) (2.0)
Bottom 44.38 1016 508 1.0
(10.0) (04) (2.0)

* Number in the parentheses is in U S . Customary Unit.

-63.0 -42.0 -21.0 0 21.0 42.0 63.0


disolocernent-mm

Figure 8. Experimental load-displacement curve of the prototype specimen (Reference 16)

approximately 2.3 were obtained. The parameters used for the prototype and for the model are listed in
Table I. The load history for the prototype, which was different from that used for the model, is shown in
Figure 9. The predicted response compared favourably with the observed data, as can be seen in Figures 8 and
9. The validity of the proposed model was further tested by comparison with the results of other full-scale tests.

COMPARISON WITH TEST RESULTS


To test the effectiveness of the proposed model in predicting the hysteresis response, data from several
laboratory tests of external beam-column joints were analysed. Seven sets of data from full-scale tests on
beam<olumn joints from three different investigations' 6 - 1 * were examined. The dimensions of the beams and
columns and relevant reinforcing details are given in the References (Table 11).The response of these specimens
lo00 M.-L. WANG A N D S. P. SHAH

J".V

. 16.0

60.0-

30.0-

/
?z

D 0-
0
- J ~ l s p l a c e r n e n t - I n

-300-

900 1 I
420 -210 0 210 420
displacement- mm

Figure 9 Analytical load-displacement curve of the prototype specimen

Table 11. Flexural capacity of the specimen

Calculated Experiment Analysis


_. -~

Specimen MY XY F
Y MY XY FY F
Y XY Reference
Number (kN-m) (mm) (kN) (kN-m) (mm) (kN) (kN) (mm) n number

22.6 129.4 190 28.0 124.5 25.40


(0.89) (29.10) (140) (1.1) (28) (1.0) - 1.0 17
17.8 131.7 203 22.9 133.4 25.4
(07) (29.6) (150) (0.9) (30) (1.0) 1.0 17
12.0 158.0 263 25.4 169.0 25.4
(0.47) (35.5) (194) (1.0) (38) (1.0) % O 17
14.5 133.4 235 25.4 151.2 25.4
(0.57) (30.0) (173) (1.0) (34) (1.0) 1.0 17
12.7 64.5 71 12.7 66.7 12.7
(0.5) ( 13.6) (50) (0.5) ( 14.4) (0.5) 1.0 16
12.7 64.5 71 12.7 66.7 12.7
(0.5) (14.5) (52) (0.5) (15.0) (0.5) 3 16
30.5 89.0 287 28.0 94.3 30.5
(1.2) (20) (212) (1.1) (21.2) (1.2) 1.0
25.4 60.0 194 25.4 63.6 25.4
( 1.O) (13.5) (143) (1.0) (14.3) (1.0) 1.0 18

My = yielding moment, (T) = positive, (B) = negative.


F , = yielding strength, XY = yielding displacement.
* values in the parentheses are in U S . customary units: My = kip-ft, F , = kips, X, = in.

as measured by the cyclic load applied to the beam versus the load-point displacement was analysed. All these
specimens were designed according to the strong column-weak beam concept.
The proposed model has three damage based parameters ( X , , C and n).The value of these three parameters
will depend on the member properties such as shear span to depth ratio, joint shear reinforcement and beam
REINFORCED CONCRETE HYSTERESIS MODEL loo1
shear reinforcement, among others. A large set of experimental data is required to evaluate the dependence of
the three damage parameters on member properties. In the present study, only one parameter (n)was adjusted
to match the experimental data of the full-scale tests. The other two parameters, X , and C , were kept the same
as for the test specimens discussed earlier. The idea was to examine the sensitivity of the value of n to some key
design parameters.
For the skeleton curve, the value of the post-yield stiffness was always assumed to be 5 per cent of the initial
stiffness. The value of the load at the yield (F, )was taken as that reported by the authors as the calculated value.
The value of X , was taken from the experiments (using the reported calculated value of F,) so that the
displacement due to cracking and debonding can be accounted for. The reported calculated values of the yield
displacement were often lower than the experimental values, as seen in Table I1 (see also Reference 19).This is
partly because the calculated values do not include deformations due to cracking and slippage.
One of the primary goals of a hysteresis model is to be able to predict the response of a structure subjected to
random excitation. To examine the capability of the proposed model in predicting reinforced concrete
response when subjected to randomly applied cyclic loading, Specimen 7 from Uzumeri18 was examined. In
this specimen the plastic hinge occurred in the beam. Other relevant information and the parameters chosen
for the analytical model are listed in Table 11. The predicted response compared favourably with the
experimental load4isplacement curves (Figures 10, 11). Note that this specimen had an unequal area of top
and bottom reinforcing steel and showed non-symmetrical hysteretic behaviour.

I0W.U

120.0
c
32.0 i

-60.0
- -8.0 -6.0

-120.0

-180.0 - - 200 0
I
- 100.0
I
0
I
displacement -
I

mm
' I
100.0
I I
200.0

Figure 10. Measured load vs. deflection response of Specimen 7 (Reference 18)

180.0

1200

60.0
z
Y

D O
-
0

-60.0

-1200

-180.0
- 200.0 -100.0 0 100.0 200.0
displocement - m m

Figure 11. Predicted load vs. deflection of Specimen 7 (Reference 18)


1002 WANG A N D S. P. SHAH
M.-L.

It is believed that the inelastic flexural deformation causes a progressive decrease in member shear strength,
while the axial load and close spacing of ties slow the rate of strength loss.2o The ability of a member to
withstand many inelastic repeated loading cycles is also closely linked to the maximum shear stress experienced
by the member. In the following examples emphasis will be placed on examining the variation of the damage
parameter n to the change in the design properties. However, it is expected that the model will not be able to
simulate the severe “pinching” of the hysteresis loop, since there is no separate mechanism in the analytical
model to account for the shear and bond deterioration in the plastic region of the beamxolumn
su bassemblages.

Effect of joint shear reinforcement


To examine the effect of varying joint stirrup reinforcement, data of two specimens tested by Ehsani and
Wight” (Specimens 2 and 4 in Table 11) were analysed. These two specimens were identical except for the
lower amount of the joint transverse reinforcement in Specimen 2 . As a result of this, Specimen 2 suffered more
severe joint damage than Specimen 4. The proposed model can predict this observed response if a lower value
of n = - 1 is selected for Specimen 2 and n = 1 for the adequately reinforced Specimen 4. It was observed that
the proposed model closely predicted the overall hysteretic response for both specimens. For the sake of
brevity, the experimental and the theoretical curves are not reproduced here.

Effect of joint shear stress


The effect of joint shear stress on the damage accumulation parameter of the proposed model was examined
by predicting the experimental results of Specimens 9 and 11 from Ehsani and Wight. Both of these specimens
showed little distress in the joint. However, since Specimen 9 had a joint shear stress which was about twice that
for Specimen 11 (15Jf; vs 9Jf i),it showed a larger stiffness reduction than Specimen 11 for the identical
loading history. It was possible to predict these differing responses by assuming that n was equal to 0 for
Specimen 9 and equal to 1.0 for Specimen 11.

Effect of beam-shear reinforcement


To assess the influence of the beam shear reinforcement on the parameter n, Specimens 3 and 7 from Scribner
and Wight16 were analysed. These two specimens were otherwise identical except for the amount of the beam-
stirrup reinforcement. Specimen 3 had beam web reinforcement which satisfied the ACI code requirement.
Specimen 7 had twice as much beam web reinforcement and the yield strength of that reinforcement was higher
(60 ksi vs 40 ksi). As a result of this Specimen 7 showed less incremental damage and was able to sustain a
higher number of load reversals. With the proposed model it was possible to simulate this behavior when the
values of n equal to 1 for Specimen 3 and 3 for Specimen 7 were chosen. Note that both of these specimens
showed distress primarily in the beams near the column face..
The effect ofcyclic loading on structural damage can be represented by the parameter n. Based on the current
study, the rate of damage was strongly dependent on the member design properties. The value of n increased by
increasing the amount of joint shear reinforcement and beam shear reinforcement and decreased by the
increase of joint shear stress. A mathematical formulation can be developed by considering parameter n as a
function of shear span to depth ratio, joint shear reinforcement and beam shear reinforcement. A large set of
experimental data is required to determine this relationship accurately. However, unless characteristic values
of the damage parameters are established, it will be difficult to predict the hysteretic behaviour of R/C members
using the approach described here.

CONCLUSION
1. A hysteretic model based on the concept of accumulated damage is proposed. The model operates on a
bilinear skeleton curve and requires a prior knowledge of three damage based parameters. The damage
parameters were calibrated from smatl-scale tests on reinforced concrete beamxolumn joints. Using
identical values of damage parameters it was possible to predict the response of the comparable
prototype specimens subjected to a different loading history.
REINFORCED CONCRETE HYSTERESIS MODEL 1003

2. The accuracy and sensitivity of the analytical model were examined by comparison with the experimental
results of several laboratory tests of external beam-column joints. It was observed that the damage
parameter n was sensitive to the amount of joint shear reinforcement, beam shear reinforcement and the
joint shear stress.
3. The model is capable of predicting the hysteretic behaviour of reinforced concrete beam-column joints,
especially the strength and stiffness deterioration and the damage state of the member. The damage state
is measured in terms of an index D which is a function of cumulative maximum deformation.
4. Parameter n, which determines the rate of damage, appears to be correlated to the member design
properties. Unless a relationship is established between n and these design properties, it will be difficult to
apply the method in general. However, with n equal to 1.0, the proposed approach predicted satisfactorily
the hysteretic behaviour of specimens designed according to the requirements of ACI 318-83 Building
Code and the recommendations of the ACI-ASCE Joint Committee 352.
5. The proposed model isa damage model that can be used to identify the damage state of the member. If the
structural response is known, the model can be used to identify the damage in conjunction with a system
identification approach (inverse problem).
6. The proposed model is compared only with the external beam-column specimens designed according to
the strong column-weak beam concept. A further study must be conducted to simulate the differing
damage accumulation behaviour of interior beam-column specimens and to simulate the pinching effects
due to severe cyclic-bond deterioration.

REFERENCES
1, R. W. Clough, ‘Effect of stiffness degradation on earthquake ductility requirements’; Structures and Materials Report No. 66-16,
Department of Civil Engineering, University of California, Berkeley, CA, 1966.
2. Saiidi Mehdi, ‘Hysteresis models for reinforced concrete’, J . struct. diu. ASCE 108, 1077-1086 (1982).
3. T. Takeda., M. A. Sozen and N. N. Nielsen, ‘Reinforced concrete response to simulated earthquakes’, J . struct. diu. ASCE 96,
2557-2573 (1970).
4. W. H. Townsend and R. D. Hanson, ‘Reinforced concrete connection hysteresis loops’, ACI Publication S P 53-1 3, 1977, pp. 351-370.
5. M. Saatcioglu, A. T. Derecho and W. G . Corley, ‘Modeling hysteretic behaviour of coupled walls for dynamic analysis’, Earthquake
eny. struct. dyh. 1 I, 71 1-726 (1983).
6. F. C. Filippou, E. P. Popov and V. V. Bertero, ‘Modeling of R/C joints under cyclic excitations’, J . struct. din ASCE 109,2666-2683
(1983).
7. F. C. Filippou, E. P. Popov and V. V. Bertero, ‘Analytical studies of hysteretic behavior of R/C joints’, J . struct. diu. ASCE 112,
1605-1622’ ( 1986).
8. S. P. Shah. ‘Prediction ofcumulative damaee for concrete and reinforced concrete’. Marer. struct. (RILEMI 17. No. 97.65-68 (1984).
9. P. Balaguru and S. P. Shah, ‘A method of iredicting crack width and deflection’, in Fatigue ofcdncrete Structures (Ed. S. P. Shahj,
Publication SP-75, American Concrete Institute, 1982.
10. Y. J. Park and A. H. S. Ang, ‘Mechanistic seismic damage model for reinforced concrete’, J . struct. div. ASCE 1 1 1, 722-739 (1985).
1 1 . S. Toussi, J. T. P. Yao and W. F. Chen, ‘A damage indicator for reinforced concrete frames’, ACI J . 81, 26@-267 (1984).
12. Y. J. Park and A. H. S. Ang., ‘Seismic damage analysis of reinforced concrete buildings’, J . struct. diu. ASCE 111, 74&757 (1985).
13. H. Banon, J. M. Biggs and H. M. Irvine, ‘Seismic damage in reinforced concrete frames’, J . struct. diu. ASCE 107, 1713-1729 (1981).
14. V. Ciampi, R. Eligehausen, V. Bertero and E. Popov, ‘Analytical model for concrete anchorage of reinforcing bars under generalized
excitation’, Report No. EERC 82/23, Earthquake Engineering Research Center, University of California, Berkeley, CA, 1982.
15. S. P. Shah, M. L. Wang and L. Chung, ‘Model concrete beam-column joints subjected to cyclic loading at two rates’, Accepted for
publication, Mazer. struct. (1987).
16. C. F. Scribner and 1.K. Wight, ‘Delayingshear strength decay in reinforced concrete flexural members and large load reversals’, Report
No. U M E E 78R2, Department of Civil Engineering, University of Michigan, Ann Arbor, Michigan, 1978.
17. M. R. Ehsani and J. K. Wight, ‘Behavior of external reinforced concrete beam to column connections subjected t o earthquake type
loading’, Report No. U M E E 82R5, Department of Civil Engineering, University of Michigan, Ann Arbor, Michigan, 1982.
18. S. M. Uzumeri, ‘Strength and ductility of cast-in-place beamxolumn joints’, ACI Publication SP53-13, 1977, pp. 283-350.
19. S. P. Shah, H. L. Chenand J. G . Sun, ‘A hystereticmodel ofmoment-rotation relationship for reinforcedconcretecolumns’, Submitted
for Publication, Pacific conJ earthquake eng. Wairakel, New Zealand (1987).
20. J. K . Wight and M. A. Sozen, ‘Strength decay of RC columns under shear reversals’, J . struct. diu. ASCE 101, 1053-1065, (1975).

Potrebbero piacerti anche