Sie sind auf Seite 1von 9

Engineering Structures 31 (2009) 22152223

Contents lists available at ScienceDirect


Engineering Structures
journal homepage: www.elsevier.com/locate/engstruct
Review article
Simplified model for damage in squat RC shear walls
Edward D. Thomson
a
, Mara E. Perdomo
b,
, Ricardo Picn
b
, Mara E. Marante
b
, Julio Flrez-Lpez
c
a
Structural Engineer, Fluor Canada Ltd, Suite 700, 1075 W Georgia St,Vancouver, Canada
b
Department of Structural Engineering, Lisandro Alvarado University, Barquisimeto, Venezuela
c
Department of Structural Engineering, University of Los Andes, Mrida, Venezuela
a r t i c l e i n f o
Article history:
Received 5 December 2008
Received in revised form
29 March 2009
Accepted 28 May 2009
Available online 21 June 2009
Keywords:
Shear walls
Reinforced concrete
Earthquake damage
Fracture mechanics
Finite elements
Lumped plasticity
Elastoplasticity
a b s t r a c t
In this paper, a newsimplified model for simulating damage of squat RC shear walls under lateral loads is
proposed. This simplified model is based on damage and fracture mechanics. It describes the reduction in
stiffness and strength due to diagonal cracking, permanent deformations due to yielding of transverse
reinforcement and sliding across shear cracks. First, the analytical expressions are developed for the
particular case of monotonic loading. Ayield function to describe permanent deformations due to yielding
of transverse reinforcement is proposed. Then, a crack resistance function, based on the Griffith criterion,
is introduced and experimentally identified. Finally, the necessary analytical expressions are developed
for hysteretic behavior. The proposed numerical model is implemented in a commercial finite element
program and validated against experimental results. It is shown that the model can predict well the
response of RC shear walls.
2009 Elsevier Ltd. All rights reserved.
Contents
1. Introduction.................................................................................................................................................................................................................... 2216
2. Model of monotonic behavior ....................................................................................................................................................................................... 2216
2.1. Element flexibility matrix ................................................................................................................................................................................. 2216
2.2. Evolution law of the permanent deformations................................................................................................................................................ 2217
2.3. Evolution law of the damage ............................................................................................................................................................................ 2217
2.4. Identification of the crack resistance function ................................................................................................................................................ 2217
2.5. Computation of the model parameters ............................................................................................................................................................ 2218
2.6. Numerical simulation........................................................................................................................................................................................ 2219
3. Model for hysteretic behavior ....................................................................................................................................................................................... 2219
3.1. Unilateral behavior ............................................................................................................................................................................................ 2219
3.2. Pinching effects in shear walls.......................................................................................................................................................................... 2220
3.2.1. Sliding function of a shear crack........................................................................................................................................................ 2220
3.2.2. Computation of sliding shear parameters......................................................................................................................................... 2221
4. Numerical implementation and model validation....................................................................................................................................................... 2221
4.1. A finite element for squat RC shear walls ........................................................................................................................................................ 2221
4.2. Numerical simulations ...................................................................................................................................................................................... 2222
5. Conclusions..................................................................................................................................................................................................................... 2222
Acknowledgements........................................................................................................................................................................................................ 2223
Appendix. Notations ................................................................................................................................................................................................. 2223
References....................................................................................................................................................................................................................... 2223

Corresponding author. Tel.: +58 251 2529279; fax: +58 251 2592173.
E-mail address: mariaperdomo@ucla.edu.ve (M.E. Perdomo).
0141-0296/$ see front matter 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.engstruct.2009.05.020
2216 E.D. Thomson et al. / Engineering Structures 31 (2009) 22152223
1. Introduction
Simulation models of shear wall nonlinear behavior can be
classified into three groups: lumped plasticity models, distributed
plasticity models, and multi-layer models.
Lumped plasticity models are easier to implement because
inelastic effects are considered concentrated on nonlinear springs
or plastic hinges of zero length. The nonlinear behavior of these
hinges is described by complicated rules. Most used typical
models are those reported by Riyadh et al. [1] Williams et al. [2],
Reinhorn et al. [3], Bazant and Bhat [4] and Ma et al. [5]. The
weakness of these models results from the difficulty in choosing
appropriate model parameters. These models usually represent
real behavior when applied to laboratory specimens and using
appropriate parameters. However, when they are used to simulate
real structure behavior, many uncertainties in the correct choice of
adequate parameters appear.
Distributed plasticity models are slightly more complicated, as
they take into account the distribution of inelastic effects along
a finite length as described by Kunnath et al. [6]. They are less
popular than the lumped plasticity models, because they have the
same shortcomings of these models with an added uncertainty
when estimating the length along which inelastic effects are
distributed.
Multi-layer models are based mainly on the finite element
method. These models use discretization of elements for structure
representation. Material behavior is represented by constitutive
relations that are usually well known. In general, the results ob-
tained withthese models are suitable; however, the computational
cost and the time consumed in the preparation of the necessary in-
put data make these models of limited use when large shear wall
structures are to be modeled. Vulcano [7] analyzes several models
which fall into this last category comparing analytical simulations
withexperimental results. Models based onVulcanos macroscopic
approach are more effective than those based on a microscopic ap-
proach. Other authors such as Colotti [8] and Ghobarah [9] report
multi-component models that include some refinements allowing
a better representation of the nonlinear behavior, but there is ba-
sically no improvement in computational cost.
In this paper, a newsimplified model for simulating the damage
of squat RC shear walls under lateral loads is proposed. This
simplified model is based on damage and fracture mechanics. It
can be classified in the group of lumped plasticity models that
describes the reduction in stiffness and strength due to diagonal
cracking, permanent deformations due to yielding of transverse
reinforcement and sliding across shear cracks.
This paper is organized as follows; in Section 2 a model of
the monotonic behavior of shear walls is proposed; in Section 3
the model is extended to the more general case of walls
subjected to cyclic loading; the numerical implementation of the
model is briefly described in Section 4 and some simulations of
experimental tests are presented in order to showthe performance
of the model.
2. Model of monotonic behavior
2.1. Element flexibility matrix
Consider a shear wall as a deep beam, the damage model of RC
frame members is adapted for members subjected to high shear
forces. The model is based on methods of continuum damage
mechanics and fracture mechanics; see Flrez-Lpez [10].
The generalized stresses and deformations matrices of a wall
member are: {M}
t
= (M
i
, M
j
, N) and {}
t
=
_

i
,
j
,
_
respec-
tively.
The mechanical interpretation of the components in those
matrices is present in Figs. 1 and 2.
Fig. 1. Generalized stresses.
Fig. 2. Generalized deformations.
In this paper, permanent deformation due to flexural effects is
neglected; only plastic deformations due the shear effects are con-
sidered. Therefore, a new variable denoted generalized plastic de-
formation matrix is introduced:
_

P
_
t
= (
p
s
,
p
s
, 0), where
p
s
represents the plastic deformations due tothe yielding of the trans-
verse reinforcement and is represented in Fig. 3. This assumption
restricts the use of the model to the case of squat elements.
Generalized stresses and deformations in an elastoplastic shear
wall are related by:
_

p
_
= [F
o
] {M} (1)
where [F
0
] is the flexibility matrix in local coordinates whose
expression is:
[F
0
] =
_
F
a
o
_
+
_
F
f
o
_
+
_
F
s
o
_
. (2)
The matrices
_
F
a
o
_
,
_
F
f
o
_
and
_
F
s
o
_
represent the flexibility due to
axial forces, flexure effects, and shears respectively. These matrices
have the following expressions:
_
F
a
o
_
=
l
EA
_
0 0 0
0 0 0
0 0 1
_
(3a)
_
F
f
o
_
=
l
3EI
_
1 1/2 0
1/2 1 0
0 0 0
_
(3b)
_
F
s
o
_
=
1
GA
v
l
_
1 1 0
1 1 0
0 0 0
_
(3c)
E.D. Thomson et al. / Engineering Structures 31 (2009) 22152223 2217
Fig. 3. Plastic rotation in a shear wall.
where E is the modulus of elasticity, A is the total area of cross
section, A
v
is the effective shear area, I is the moment of inertia,
G the shear modulus and l the length of the member. It can be seen
that for large values of l, the shear term becomes small while the
flexure termincreases. This is the case for slender members where
shear deflections can be neglected.
Another significant inelastic phenomenon is concrete cracking.
This effect produces a reduction of the element stiffness. The goal
of this paper is the inelastic analysis of RCshear walls; therefore the
latter termin expression (2) is modified by introducing the damage
variable (d
s
) that can take values between zero and one:
_
F
s
(d
s
)
_
=
1
GA
v
l(1 d
s
)
_
1 1 0
1 1 0
0 0 0
_
. (4)
Physically, the damage variable measures the degree of con-
crete cracking in the wall, i.e. d
s
= 0 indicates that there is no
concrete cracking, d
s
= 1 represents a cracked wall that has no
shear stiffness at all (see Fig. 4). The flexibility matrix of a degrad-
able shear wall has the following expressions:
[F(d
s
)] =
_
F
a
o
_
+
_
F
f
o
_
+
_
F
s
(d
s
)
_
. (5)
Therefore, the state law of a member with shear deformations,
damage and plastic rotations is:
_

p
_
= [F (d
s
)] {M} . (6)
The internal variables in the shear wall model as the permanent
deformations (
p
) and damage (d
s
) are obtained from evolution
laws.
2.2. Evolution law of the permanent deformations
When actions on the member exceed some critical value,
permanent or plastic deformations appear in the member. As
aforementioned only shear plastic effects are considered. In order
to compute the evolution of the plastic rotation, a yield function f
y
is introduced:
f
y
=

V
1 d
s
c
s

p
s

V
y
(7)
where V = (M
i
+M
j
)/l is the shear force on the member, c
s
and V
y
are parameters of the model that depend on the properties of the
element. There may be plastic rotation evolution only if the yield
function is equal to zero:

p
s
> 0 only if f
y
= 0. (8)
Fig. 4. Physic representation of damage variable by shear.
2.3. Evolution law of the damage
The Griffith criterion, which is the basis of Fracture Mechanics,
states that there may be crack propagation only if the energy
release rate equals the crack resistance of the wall:

d
s
> 0 only if G
s
= R(d
s
) (9)
where R = R(d
s
) is the crack resistance of the wall that is assumed
to be a function of the damage state of the wall.
The energy release rate of a damaged shear wall can be defined
as:
G
s
=
W

d
s
=
V
2
l
2GA
v
(1 d
s
)
2
(10)
where W

is the complementary strain energy of a damaged wall


that can be written as: W

= 1/2{M}
t
[F(d
s
)]{M}.
As in Fracture Mechanics, the crack resistance function has to
be identified from experimental results, as described in the next
section.
2.4. Identification of the crack resistance function
The model that describes the behavior of a shear wall is com-
posed by the state law (6), the plastic rotation evolution law (8)
with yield function (7), and the Griffith criterion (9). It can be no-
ticed that only the crack resistance termneeds experimental iden-
tification. In order to carry out this identification an experimental
program of shear walls was carried out at in Laboratory of Struc-
tural Mechanics at the Lisandro Alvarado University.
The shear walls were designed according to ACI Code 318-05
[11]. A relationship l/d (d is the effective depth) less than 2.5
was used in order to obtain a shear dominant failure mode.
Reinforcement of the specimens where chosen so that damage
or cracking due to bending are negligible. A RC non-slender
element with a high percentage of longitudinal reinforcement
and a low concrete resistance allows a further degradation of the
strength and stiffness to achieve shear failure [12]. The geometric
characteristics of shear walls are shown in Table 1. These walls
were tested under cyclic loading and zero axial force, see Fig. 5.
Fig. 6 shows a shear wall built in as a cantilever. The boundary
and kinematic conditions of the test are: M
i
= V l; M
j
= 0;

i
=
t
l
where t is the lateral displacement at the top of the wall.
2218 E.D. Thomson et al. / Engineering Structures 31 (2009) 22152223
Table 1
Geometry of shear wall specimens.
Specimen w (mm) e (mm) l (mm) d (mm) l/d
v
(%)
s
(%) f

c
(MPa) F
y
(Mpa) F
su
(Mpa)
SW-M01 500 125 600 475 1.26 6.25 0.50 16.7 389.3 630.0
SW-H02 600 100 850 575 1.48 3.29 0.73 16.5 461.0 630.0
SW-H03 585 100 700 560 1.25 0.33 0.26 37.0 607.8 759.7
w = wide of wall
e = thickness of wall
l = length of wall
d = effective depth

s
= percentage of transverse reinforcement

v
= percentage of longitudinal reinforcement
f

c
= nominal resistance of concrete
F
y
= yield stress of transverse steel
F
su
= ultimate stress of transverse steel
Fig. 5. Shear wall specimen geometry and loading.
The relationship between force and displacement can be obtained
from the state law (6) and those conditions.
(t t
p
) =
_
l
3
3EI
+
l
GA
v
(1 d
s
)
_
V (11)
where, t
p
=
p
s
.l is the plastic deflection.
The slope of an elastic unloading in the test (see Fig. 6), denoted
as Z, is:
Z =
V
(t t
p
)
. (12)
Therefore, after (11)(12) the following relationship between
shear damage d
s
and the slope Z is obtained:
Z =
1
l
3
3EI
+
l
GA
v(1d
s)
(13)
Then
d
s
= 1
l
GA
v
_
1
1
Z

l
3
3EI
_
. (14)
It can be noted that this procedure to measure shear damage
is a modification of the stiffness variation method of continuum
damage mechanics, Lemaitre [13].
The energy release rate can be computed from Eq. (10) with
the experimental values of V and d
s
. Fig. 7 shows the energy
release values for specimen SW-M01. An expression for the crack
resistance function is:
R (d
s
) = G
crs
+ q
s
ln (1 d
s
)
(1 d
s
)
. (15)
Two member dependent parameters are necessary to define the
crack resistance of the wall: G
crs
and q
s
. A plot of this function with
appropriate values of the parameters can also be seen in Fig. 7.
A good correlation is observed between the experimental results
and the proposed crack resistance function. A similar analytical
expression was proposed by Cipollina et al. [14] for RC frames.
2.5. Computation of the model parameters
The proposed model has four parameters: G
crs
, q
s
, c
s
, V
y
; they
depend on the cross-section of the wall, the horizontal and
Fig. 6. Representation of the variable Z.
E.D. Thomson et al. / Engineering Structures 31 (2009) 22152223 2219
Fig. 7. Damage variable vs. Energy release rate for shear wall SW-M01.
vertical reinforcement, and the material properties. The direct
determination of these parameters is not convenient; instead they
can be computed by the resolution of the following system of
equations:
V = V
cr
implies d
s
= 0 (16a)
V = V
p
implies
p
s
= 0 (16b)
V = V
u
implies dV/dd
s
= 0 (16c)
V = V
u
implies
p
s
=
p
us
(16d)
where, V
cr
is the shear that produces the first diagonal crack, V
p
is the shear that lead to yielding of transverse reinforcement, V
u
is the ultimate shear resisted by the wall, and
p
us
is the ultimate
plastic rotation. All these wall properties can be computed from
conventional reinforced concrete theory.
The cracking shear of a RC member, when the member is sub-
jected to shear and axial loads can be obtained by the expression
of ACI 318-05 [11]:
V
cr
=
_
1 +
P
14A
g
_
_
_
f

c
6
_
0.8A
g
(17)
where A
g
is the total area of the wall cross section, f

c
is the nominal
resistance of the concrete inMPa, andP is the axial loadonthe wall.
The shear load that leads to yielding of transverse reinforce-
ment can be obtained by the expression of ACI 318-05 [11]:
V
p
=
_
A
v
F
y
Cot
s
_
d (18)
where A
v
is the transverse reinforcement area of the wall, d is the
effective depth of the wall, F
y
is the yielding stress of transverse
reinforcement in MPa, s is the separation between stirrups, and
is the angle between the compression strut and the longitudinal
axis of the shear wall.
The ultimate shear can be obtained by the expression proposed
in Sezen and Moehle [15]:
V
u
=
A
v
F
y
d
s
+
_
0.5
_
f

c
l/d
_
1 +
P
0.5
_
f

c
A
g
_
0.8A
g
. (19)
The ultimate plastic rotation for a shear load can be computed by
the expression proposed by Park and Paulay [16]:

P
us
=
(F
su
F
y
)A
v
E
s
ts
_
1

s
+
E
s
E
c
_
0.25l (20)
where the F
su
is the ultimate stress of the transverse reinforcement
in MPa, E
s
is the modulus of elasticity of the reinforcement in
MPa, E
c
is the modulus of elasticity of concrete in MPa,
s
is the
percentage of transverse reinforcement.
Table 2
Computed properties of the Specimen SW-M01.
Specimen V
cr
V
p
V
u

P
us
SW-M01 34.05 50.75 170.60 0.0066
V
cr
= shear force that produces the first diagonal crack (kN)
V
p
= shear force that yields the horizontal reinforcement (kN)
V
u
= the ultimate shear force resisted by the wall (kN)

P
us
= the ultimate plastic rotation in a member due to shear
Table 3
Model parameters of the Specimen SW-M01.
Specimen V
y
c
s
G
crs
q
s
SW-M01 51.72 61371 3.73 253.46
V
y
= parameter for yield function (kN)
c
s
= parameter for yield function (kN)
G
crs
= parameter for crack resistance function (kN mm)
q
s
= parameter for crack resistance function (kN mm)
2.6. Numerical simulation
A simulation of the SW-M01 test was carried out. The results
of this simulation are shown in Fig. 8(b). As it can be seen,
the proposed model represents adequately the evolution of the
damage due to shear and the accumulation of plastic deformations
in the wall. The wall properties used for the simulation are
presented in Table 2 and the corresponding model parameters are
shown in Table 3.
The envelope of the numerical result can be seen, together with
the experimental results, in Fig. 8(a). It can be observed that the
model represents correctly the experimental behavior of the wall.
3. Model for hysteretic behavior
3.1. Unilateral behavior
During cyclic loadings, two distinctive sets of shear cracks can
appear inthe wall (see Fig. 9). Eachset is mainly relatedto a specific
direction of the shear load. In continuum damage mechanics,
similar phenomena are represented by the introduction of two
damage variables. One of them is related to microcracking density
due to positive stress (positive damage) and the other represents
damage due to negative stress (negative damage), see [17]. When
the shear force changes sign, one set of cracks tends to close and its
presence has a reduced effect in the wall behavior while the other
set of cracks tends to open and became the dominant stiffness
reduction phenomenon. This class of behavior is called unilateral
in the damage mechanics literature.
The model described in this paper can include the concept to
unilateral damage as described in Flrez-Lpez [10]. There are now
two damage variables for shear: d
+
s
and d

s
, which characterize
the state of damage due to positive and negative shear forces,
respectively (see Fig. 9).
The elasticity law (1) can be generalized as:
_

P
_
=
_
F(d
+
s
)
_ _
M
+
_
+
_
F(d

s
)
_ _
M

_
(21)
where, {M
+
} represents the positive part of the elements of
matrix {M} and {M

} is the negative part of the elements of {M};


i.e.
M
i

+
= M
i
if M
i
> 0 and M
i

+
= 0 otherwise (22a)
M
i

= M
i
if M
i
< 0 and M
i

= 0 otherwise. (22b)
The flexibility matrices have the same basic form of Eq. (5)
substituting d
s
by d
+
s
andd

s
:
_
F(d
+
s
)
_
and
_
F(d

s
)
_
. It canbe noticed
that for a positive shear force the flexibility terms are increased
2220 E.D. Thomson et al. / Engineering Structures 31 (2009) 22152223
Fig. 8. Shear wall SW-M01 (a) Experimental results (b) Numerical simulation.
only by the damage variable d
+
s
and negative cracks are assumed
totally closed with no influence at all in the wall behavior.
The evolution of shear damage is described according again to
the Griffith criterion:

d
+
s
> 0 only if G
+
s
= R(d
+
s
) (23a)

s
> 0 only if G

s
= R(d

s
). (23b)
The plastic evolution law is similar to the one for the model of
monotonic behavior, but the yield function has now two expres-
sions: one for positive actions and another for negative ones.
f
y
=
V
(1 d
+
s
)
X Q; if
_
V
(1 d
+
s
)(1 d

s
)
c
s

p
s
_
0
f
y
=
V
(1 d

s
)
+ X Q otherwise (24)
where X is a kinematic hardening term, and Q is an isotropic hard-
ening term, which are defined as follows:
X = 0.60c
s

P
s
(25)
Q = 0.40c
s
p
s
+ V
y
. (26)
The variable p
s
is the maximum plastic rotation at any given time
of the entire plastic deformation history.
3.2. Pinching effects in shear walls
The so called pinching effect in the hysteretic behavior curves
was observed during the experimental analyses. This phenomenon
is due to some sliding between the cracked surfaces before they
come in full contact [18]. The basis for the modeling of this
phenomenon is explained below.
Consider an interface between two different continuum bodies
as is shown in Fig. 10(a) and let and be the normal and
shear stresses on the interface. If the surface is characterized by
a Coulomb friction criterion, the relative horizontal displacement
h between the blocks obeys the following law:
_

h > 0 if ||
s
() = 0

h = 0 if ||
s
() < 0
(27)
where the term
s
is the slide resistance that depends on the
normal stress. The non-slide domain, for an arbitrary resistance,
is represented in Fig. 10(b). It can be noted that slide occurs when
the shear stress reaches the slide resistance. The latter value is not
constant but depends on the normal stress. For higher values of the
compressive normal stress, higher values of the slide resistance are
obtained. A general presentation of interface behavior can be seen
in plasticity textbooks (see for instance Salenon [19]).
Fig. 9. Representation of positive and negative shear damage.
Fig. 10. (a) Interface between two media. (b) Non-slide domain.
3.2.1. Sliding function of a shear crack
The process of slide across a shear crack can also be explained
in terms of Coulomb friction criterion. Consider a shear crack
in a shear wall which has been formed under positive load.
As the load is reduced to zero, the crack remains open. Once
the load starts to be applied in the negative direction, friction
across the crack is small, but as the crack begins to close, friction
increases gradually, which can be seen as a gradual increase in
the normal stress and consequentially in the slide resistance.
Additionally, if reinforcement yielding has occurred as the crack
opens, it is evident that in order to close the crack completely, the
reinforcement must be yielded in compression. Therefore, there is
an interaction between two phenomena: slide across shear cracks
and yield of the reinforcement. Both phenomena generate plastic
rotations in the wall.
Ageneralizationof the concept of Coulomb frictioncriterioncan
be used to describe the behavior of an inelastic shear wall with
slide. Thus, the following slide function is introduced:
f
s
= |V| k
s
. (28)
Expression (28) allows one to define the evolution of plastic rota-
tions as follows: there will be increments of the plastic rotations
due to slide across shear cracks if the shear force reaches the crit-
ical value k
s
, otherwise these increments are null.
In the case of Coulomb friction criterion, it is accepted that
the slide critical value depends on the normal stresses on the
interface. For slide across shear cracks, it will be assumed that the
E.D. Thomson et al. / Engineering Structures 31 (2009) 22152223 2221
Fig. 11. Interaction between yield and slide functions.
critical value k
s
corresponds to a hardening function. The analytical
determination of the hardening function is a very complex
problem, therefore the following phenomenological expression is
proposed:
k
s
= V
o
e
sign(V)
p
s
. (29)
An exponential function of the plastic rotation has been chosen so
that the typical pinched curves are obtained when slide is present
in the wall. The term V
o
will be called slide resistance which is
a concept similar to the yield shear force in plasticity, i.e., V
o
is
the shear force that produces slide when no plastic rotations have
occurred yet. The computation of the parameters V
o
and will be
discussed in a following section.
To model sliding shear together with damage due to cracking,
a slide function due to sliding shear is proposed, similar to that
proposed by Picn et al. [20] for a similar phenomenon observed
in beams with bond failure. This slide function (f
s
) is defined as
follows:
if
_
V
(1 d
+
s
)(1 d

s
)
c
s

p
s
_
0
then f
s
=

V
(1 d
+
s
)

V
o
e
sign(V)
p
s
else f
s
=

V
(1 d

s
)

V
o
e
sign(V)
p
s
. (30)
Now, there are two yield functions which interact, one due to
actual yielding of horizontal reinforcement and the other due to
sliding shear.
The function which controls the evolution of plastic deforma-
tions will be the one with the largest value at any given time as is
illustrated in Fig. 11.
This function takes into account the fact that on closure of the
shear cracks, there are two competing effects: friction between
crack faces and compression forces acting on the horizontal rein-
forcement.
3.2.2. Computation of sliding shear parameters
In expressions (29) and (30), two new parameters are intro-
duced: V
o
and . Where V
o
represents the value of shear force
which produces slide across a crack for zero plastic rotation and
is a parameter which can be calculated by solving the following
equations:
if f
y
or f
s
= 0 then
_
G
+
s
= R(d
+
s
) for positive actions
G

s
= R(d

s
) for negative actions.
(31)
As a result, the following expression is obtained for positive ac-
tions:
=
(1 d
+
s
) ln
_
2GA
v
R(d
+
s
)
lV
2
0
_
2
_
2GA
v
(1d
+
s
)
2
R(d
+
s
)
l
(1 d
+
s
)V
y
(1 )(1 d
+
s
)c
s
p
s
(32)
Fig. 12. Effect of parameter.
Fig. 13. Generalized displacements {q} and internal forces {Q}.
and, for negative actions:
=
(1 d

s
) ln
_
2GA
v
R(d

s
)
lV
2
0
_
2
_
2GA
v
(1d

s
)
2
R(d

s
)
l
(1 d

s
)V
y
(1 )(1 d

s
)c
s
p
s
. (33)
The effect of the parameter on the hysteretic curves can be seen
in Fig. 12.
4. Numerical implementation and model validation
4.1. A finite element for squat RC shear walls
The model can be included in conventional structural analysis
programs as a new finite element. The generalized displacements
(degrees of freedom) and internal forces of the element are
given, respectively, by {q}
t
= (q
1
, q
2
, . . . , q
6
) and {Q}
t
=
(Q
1
, Q
2
, . . . , Q
6
) as indicated in Fig. 13. A finite element is defined
as the set of equations that relate the generalizeddisplacements {q}
with the internal forces {Q}. A finite element for a RC shear wall is
composed by the proposed model and two additional equations.
The first one is denoted kinematic equation and relates the
generalized deformations {} with the generalized displacements
{q}. The second one is the element equilibrium equation that
relates the element internal forces {Q} with the generalized
stresses {M}.
The kinematic equation is
{} = [B] {q}
[B] =
_
_
_
_
sec
l

cos
l
1
sec
l
cos
l
0
sec
l

cos
l
0
sec
l
cos
l
1
cos sec 0 cos sec 0
_

_
(34)
where [B] is called transformation matrix and is the angle
betweenthe chordof the element andthe global axis X (see Fig. 13).
The member equilibrium equation can be expressed as
{Q} = [B]
t
{M} . (35)
2222 E.D. Thomson et al. / Engineering Structures 31 (2009) 22152223
Fig. 14. Specimen SW-H02 (a) Experimental results (b) Numerical simulation.
Fig. 15. Specimen SW-H03 (a) Experimental results (b) Numerical simulation.
Table 4
Computed properties of the specimens SW-H02 and SW-H03.
Specimen V
cr
V
p
V
u

P
us
SW-H02 32.50 145.47 193.04 0.0162
SW-H03 47.44 74.15 154.84 0.0053
V
cr
= shear force that produces first diagonal crack (kN)
V
p
= shear force that yields horizontal reinforcement (kN)
V
u
= ultimate shear force resisted by a shear wall (kN)

P
us
= ultimate plastic rotation in a member due to shear
The finite element for squat RC walls was included in the library of
a commercial FE program [21]. The numerical implementation of
the model was carried out in a similar way as is described in [22].
4.2. Numerical simulations
In order to validate the model some additional tests were
carried out. The specimens are similar to the one described
in Section 2. They were called SW-H02 and SW-H03 and its
geometry is presented in Table 1. The specimens were subjected
to cyclic lateral loading of increasing amplitude and zero axial
force. Figs. 14 and 15 show the experimental results and the
numerical simulations of those tests. The shear wall properties
for the simulation are presented in Table 4 and the corresponding
model parameters are shown in Table 5.
It can be noticed that two degrees of pinching can be observed
in these tests. This difference might be related to the percentage
of transversal reinforcement in both specimens (0.73 % for SW-
H02 and 0.26 % for SW-H03). In the model the degree of pinching
is controlled by the parameter V
o
in Eq. (30). So far there is no
validated procedure to compute this parameter as a function of the
wall characteristics and this is a limitation of the model.
Table 5
Model parameters of the specimens WW-H02 and SW-H03.
Specimen V
y
c
s
G
crs
q
s
V
o
SW-H02 189.70 20427 6.99 668.18 103.52
SW-H03 78.53 63076 11.27 322.18 45.95
V
y
= parameter for yield function (kN)
c
s
= parameter for yield function (kN)
G
crs
= parameter for shear damage function (kN mm)
q
s
= parameter for shear damage function (kN mm)
V
o
= parameter for shear slide function (kN)
The envelope of the numerical results can be seen, together
with the experimental results, in Fig. 14(a) and Fig. 15(a). It can
be observed that the model represents correctly the experimental
behavior of squat RC shear walls.
5. Conclusions
A model for the simulation of damage in squat RC shear walls
under cyclic lateral loads has beenproposed. It is basedonconcepts
and methods of damage and fracture mechanics. It allows, at
least in a qualitative manner, a representation of the following
effects: stiffness and strength degradation due, mainly, to diagonal
cracking of the concrete; plastic deformations due to yield of the
horizontal reinforcement; and sliding shear across diagonal cracks
(pinching effect).
A good correlation between experiment and model can be
appreciated. Most parameters of the model can be determined
from conventional reinforced concrete theory.
In its present state, the model does not account for the
combined damage due to shear and bending, as in tall shear walls,
where cracking due to bending may be more significant than
cracking due to shear.
E.D. Thomson et al. / Engineering Structures 31 (2009) 22152223 2223
Acknowledgements
The experimental investigation presented in this paper was
carried out in the Laboratory of Structural Mechanics at the
Lisandro Alvarado University. The research work was sponsored by
FONACIT and CDCHT Lisandro Alvarado University, Venezuela.
Appendix. Notations
The following symbols are used in this paper:
A total cross section area
A
g
gross area of concrete section
A
v
effective shear area of cross section
c
s
parameter for yield function
d effective depth
d
s
shear damage variable
d
+
s
shear damage variable for positive actions
d

s
shear damage variable for negative actions
E
c
modulus of elasticity of concrete
E modulus of elasticity
E
s
modulus of elasticity of steel
f

c
nominal resistance of the concrete
f
s
shear slide function
F
su
ultimate stress of transverse steel
f
y
yield function
F
y
yield stress of transverse steel
[F
o
] flexibility matrix of member
[F
a
o
] flexibility matrix due to axial forces
[F
f
o
] flexibility matrix due to flexure effects
[F
s
o
] flexibility matrix due to shear effects
[F
s
(d
s
)] shear flexibility matrix of a damaged wall
[F(d
s
)] flexibility matrix of a damaged wall
G shear modulus
G
crs
parameter for shear damage function
G
s
energy release rate of damaged shear wall
G
+
s
energy release rate of a damaged shear wall for positive
actions
G

s
energy release rate of a damaged shear wall for negative
actions
h relative displacement between two blocks of an interface
k
s
critical shear force that produces slide
l length of wall
M
i
, M
j
flexural moments at nodes i and j of a member
M
+
positive part of the elements of matrix {M}
M

negative part of the elements of matrix {M}


N axial force in a member
P axial load
p
s
maximum plastic rotation achieved
q
s
parameter for shear damage function
Q isotropic hardening term for yield function

s
percentage of transverse reinforcement

v
percentage longitudinal reinforcement
R, R(d
s
) crack resistance function
e thickness of wall
V shear force in a member
V
cr
shear force that produces first diagonal crack
V
o
parameter for shear slide function
V
p
shear force that yields horizontal reinforcement
V
u
ultimate shear force resisted by a shear wall
V
y
parameter for yield function
w wide of wall
W

complementary strain energy of a damaged wall


X kinematic hardening term for yield function
Z slope of elastic unloading
parameter of yield function
axial elongation of the member cord
t horizontal displacement at top of shear wall
t
p
plastic horizontal displacement at top of wall

i
,
j
total rotation at nodes i and j of member

P
s
plastic rotation in a member due to shear

P
us
ultimate plastic rotation in a member due to shear
parameter for shear slide function
normal stress across an interface
shear stress across an interface

s
shear slide resistance
References
[1] Riyadh H, Mohamad M, Murat D. Prediction of damage in R/C shear panels
subjected to reversed cyclic loading. J Earth Eng 2005;9(1):4166.
[2] Williams MS, Villernure I, Sexsmith RG. Evaluation of seismic damage indices
for concrete elements loaded in combined shear and flexure. ACI Struct J 1997;
94(3):31522.
[3] Reinhorn AM, Kunnath SK, Mander JB. In: Fafjar P, Krawinkler yH, editors.
Seismic design of structures for damage control in nonlinear seismic analysis
and design of reinforced concrete buildings. London: Elsevier Applied Science;
1992. p. 6376.
[4] Bazant ZD, Bhat PD. Prediction of hysteresis of reinforced concrete members.
J Struct Eng, ASCE 1977;103(1):15380.
[5] Ma SM, Bertero VV, Popov EP. Experimental and analytical studies on the
hysteretic behavior of reinforced concrete rectangular and T-beams. Report
No. EERC 76-2, Berkeley: Earthquake Engineering Research Center, University
of California, 1976.
[6] Kunnath SK, Reinhorn A, Park YJ. Analytical modeling of inelastic seismic
response of R/C structures. J Struct Eng, ASCE 1990;116(4):9961017.
[7] Vulcano A. In: Fafjar P, Krawinkler yH, editors. Macroscopic modeling for
nonlinear analysis of rc structural walls in nonlinear seismic analysis and
design of reinforced concrete buildings. London: Elsevier Applied Science;
1992. p. 81202.
[8] Colotti V. Shear behavior of RCstructural walls. J Struct Eng, ASCE 1993;119(3):
72846.
[9] Ghobarah A, Youssef M. Modelling of reinforced concrete structural walls. Eng
Struct, Elsevier Science 1999;21(10):91223.
[10] Flrez-Lpez J. Simplified model of unilateral damage for RC frames. J Struct
Eng, ASCE 1995;121(12):176572.
[11] American Concrete Institute (ACI). Building code requirement for structural
concrete. ACI Committee 318, Farmington Hills, Mich, 2005.
[12] Woodward KA, Jirsa JO. Influence of reinforcement on RC short columns
resistance. J Struct Eng, ASCE 1984;110(1):90104.
[13] Lemaitre J. A course on damage mechanics. Germany: Springer-Verlag; 1992.
[14] Cipollina A, Lpez-Inojosa A, Flrez-Lpez J. A simplified damage mechanics
approach to nonlinear analysis of frames. Comput Struct 1995;54(6):111326.
[15] Sezen H, Moehle J. Shear strength model for lightly reinforced concrete
columns. J Struct Eng, ASCE 2004;130(11):1692703.
[16] Park R, Paulay T. Reinforced concrete structures. New York: John Wiley and
Sons; 1975.
[17] Ladeveze P. On an anisotropic damage theory. In: Proc. of the CNRS
international colloquium of failure criteria of structural media. France: Villard
de Lans; 1983.
[18] Saatcioglu M, Humar JM. Dynamic analysis of buildings for earthquake-
resistant design. J Civ Eng Can 2003;30:33859.
[19] Salenon J. Calcul la rupture et analyse limite. Presses de lcole nationale des
ponts et chausses, Paris, France, 1983.
[20] Picn-Rodrguez R, Quintero-Febres C, Flrez-Lpez J. Modeling of cyclic bond
deterioration in RC beam-column connections. Struct Eng Mech 2007;26(5):
56989.
[21] Abaqus users manual Version 6.2. Pawtucket, RI: Hibbitt, Karlson &
Sorensen, Inc; 2001.
[22] Marante ME, Flrez-Lpez J. Three dimensional analysis of reinforced concrete
frames based on lumped damage mechanics. Int J Solids and Struct 2003;
40(19):510923.

Das könnte Ihnen auch gefallen