Beruflich Dokumente
Kultur Dokumente
Composite Structures
journal homepage: www.elsevier.com/locate/compstruct
INSA-Lyon, Laboratoire de Gnie Civil en Ingnierie Environnementale, Bt. Coulomb, 34 avenue des Arts, 69621 Villeurbanne, France
Universit de Lyon 1, Laboratoire de Gnie Civil en Ingnierie Environnementale, 82 bd Niels Bohr, 69622 Villeurbanne, France
a r t i c l e
i n f o
Article history:
Available online 30 March 2014
Keywords:
RC walls
CFRP retrot
Pushover tests
FE analyses
Smeared xed crack approach
Coupled elastoplastic damage approach
a b s t r a c t
The efciency of shear walls to sustain the earthquake loads mainly depends on the aspect ratio, the
vertical load, and the ratios of longitudinal and transverse reinforcement. In this paper, two lightly
reinforced concrete walls have been experimentally and numerically investigate with an aspect ratio
equal to 0.67 (short wall) and 2.5 (slender wall). Different retrotting strategies have been considered
using Carbon-Fiber-Reinforced-Polymer (CFRP) materials. Pushover analyses have been carried out using
two advanced numerical approaches for the concrete modeling: the rst one adopts a 2D plane stress
approach with a local concrete model based on the smeared xed crack approach and a classical
regularization technique based on the fracture energy; the second investigates a coupled elastoplastic
damage model using local approach in 2D and 3D simulations. The numerical results in terms of the
loaddisplacements are very consistent with experimental data. The observed failure modes and crack
patterns are satisfactorily reproduced for both walls, without or with the CFRP strengthening.
2014 Elsevier Ltd. All rights reserved.
1. Introduction
Reinforced concrete (RC) shear walls represent one of the most
widespread bracing systems for buildings. Post-earthquake observations surprisingly report good seismic behavior of structural wall
buildings [1] while evaluation of existing buildings according to
modern standards often concludes that they provide insufcient
safety margins [2]. Although reinforced concrete walls are efcient
to dissipate the induced seismic energy, they exhibit signicant
post-earthquake damage. The main causes of damage are: occurrence of unpredictable high seismic activity, improper design and
construction aws. The RC wall response behavior depends to a
great extent on its height to length ratio. The RC walls characterized by H/L ratio greater than 2 are considered as slender or long
walls, whereas the walls characterized by H/L ratio less than 2
are considered as short walls. Based on experimental works, some
analytical models were developed to predict the inelastic strength
of walls subjected to reverse cyclic loading and to assess the natural frequencies of the walls [35].
During an earthquake, the shear walls play a major role in
resisting lateral loads for tall concrete buildings. They must be
carefully designed to provide not only adequate strength, but also
Corresponding author. Tel.: +33 680959568.
E-mail address: khuong.le-nguyen@insa-lyon.fr (K. Le Nguyen).
http://dx.doi.org/10.1016/j.compstruct.2014.02.026
0263-8223/ 2014 Elsevier Ltd. All rights reserved.
404
Table 1
CFRP material properties.
Properties
Value
Average thickness
Ultimate strength
Young modulus
Strength/1 cm (longitudinal direction)
Strength/1 cm (horizontal direction)
0.48 mm
1700 MPa
105,000 MPa
8.15 kN
3.5 kN
405
Table 2
Geometric characteristics of specimens.
Specimen
lw (m)
hw (m)
t (m)
N (kN)
VE1
VER1
VC1
VCR1
0.6
0.6
0.9
0.9
1.5
1.5
0.61
0.61
0.08
0.08
0.08
0.08
90
90
110
110
were used for the reinforcement of the top beam and the bottom
beam composing the footing of the specimen.
For the retrot of walls by CFRP materials, CFRP 50 mm wide
strips were employed. Before applying the CFRP strips, the wall
surfaces were sandblasted and cleaned using high-pressure air
gun. Holes into the footing were drilled and cleaned, before putting
the CFRP wicks used as anchors (Fig. 5a). Holes were also drilled in
the thickness of the walls for the anti-buckling anchors (Fig. 5b).
After the drilling and cleaning stage, the resin adhesive was applied
on the wall surface where the CFRP strips will be bonded, and injected into the holes.
Three strips over the total height of the wall were applied on
each side of the walls as shown in Fig. 4. The distance between
406
vertical strips is equal to 225 mm for the slender wall (VER1) and
230 mm for the short wall (VCR1). Seven horizontal strips and four
horizontal strips were bonded on each side of VER1 and VCR1,
respectively.
3. Test results
The test results are presented in the following: the behavior of
the short wall is essentially controlled by shear, while the slender
407
Fig. 5. (a) Injecting resin into the anchors. (b) Implementation of anti-buckling anchors.
408
Fig. 9. Pushover curves for short walls: VC1 and VCR1 (CFRP-retrotted).
409
410
compression. The standard values employed for concrete are between 60 and 150 J/m2.
In the case of plain concrete, the failure strain in uniaxial traction is dened by the following relationship:
etm k
2Gf
ft h
h ah Ae 1=2
Fig. 12. Cracking Nadai criterion in traction and tractioncompression and isotropic hardening in compression.
h6k
Gf E0
ft2
r
Gf E0
ft k
h
h
Grc
G
1
f
f
ls
The average spacing of the cracks ls depends on the bar diameter, the reinforcement ratio into the volume, the concrete cover,
and the minimum bond length. For different types of structural
elements, empirical formulae for the calculation of average spacing
of the cracks ls can be found in the CEB-FIP Model Code 1990 [28].
4.1.2. Ultimate strain in uniaxial compression based on the
compressive fracture energy
The identication of the plastic strain in compression erupt requires a complete crushing test of a concrete sample in uniaxial
compression, under imposed and controlled displacement. The
uniaxial compressive fracture energy Gc is dened by the area under the stress-displacement curve.From Feenstra [30], a parabolic
curve compression hardening can be adopted as follows:
rc
8
>
2
>
> f3c 1 4 e e 2 ee2 ! if e < epic
<
pic
pic
2
>
e
e
>
pic
>
: fc 1 e e 2 ! if epic < e < erupt
rupt
pic
411
epic
4f c
3E0
Stress
su
Gc 11
epic
hfc 48
sy
su
sh
Strain
Fig. 13. MenegottoPinto law under monotonic loading for steels and CFRP strips.
failure at the ultimate strain esu. Phases of the monotonic curve are
given by:
erupt 1:5
erupt P 1:75epic
Concerning the compressive fracture energy, standard concrete values between 5 Nmm/mm2 and 10 Nmm/mm2, that is
500010,000 J/m2, are generally adopted [28].
4.2. Coupled elasto-plastic damage (RICRAG) model for concrete
The coupled elastoplastic damage model for concrete material
proposed by Richard and Ragueneau [25], labeled in the following
as RICRAG model, is in good agreement with experimental data on
beams tested in laboratory up to failure. It enables to conduct 2D
analyses and 3D analyses of RC structures under monotonic and
cyclic loads. The detailed features of the concrete model can be
found in [25]. Here, the focus will be on the identication of the
model parameters governing the damage mechanism. First, for
describing the undamaged behavior, two elastic parameters need
to be provided: the Youngs modulus and the Poisson ratio. The
main concept of the model lies in the decomposition of the thermodynamic state potential into spherical and deviatoric parts.
The damage mechanism concerning the spherical part of the thermodynamic state potential is controlled by the choice of four
parameters Y0, b, ADir and AInd. Y0 is an energy-like initial threshold
enabling the activation of the damage evolution. Its nominal value
can be assessed by the recommended values for the fracture energy Gf of the concrete material in traction as previously discussed
for the xed smeared crack approach. The b parameter introduces a
dissymmetry in traction with respect to the compression for driving the damage evolution. It can be identied by tting the experimental data available in bi-compression and bi-traction. The
remaining two parameters for describing the damage mechanism,
noted as ADir and AInd, can be viewed as brittleness parameters in
traction and compression, respectively. These parameters are
introduced in the denition of the isotropic hardening ow rules
and permit to reproduce the non-symmetry between traction
and compression. La Borderie proposed an identied procedure
through an inverse method [31]. Finally, the sliding mechanism
characterized by a non-associated kinematic hardening, requires
two additional parameters, noted as c0 and i0. These parameters
were identied by cyclic compressive tests for which hysteretic effects dominate in comparison to tensile test.
4.3. 1D cyclic elastoplastic model for steels and CFRP strips
The steel reinforcement was modeled using the uniaxial cyclic
law proposed by Menegotto and Pinto [32]. The cyclic law is related to bar elements in 2D and 3D analyses. The monotonic curve
of this law depicted in Fig. 13, is characterized by three phases: linear elastic dened by the Youngs modulus Ea until the plastic
stress rsy, followed by a plateau stress beginning at the plastic
strain ey until the nal strain esh, and nally a hardening curve until
8
r Ea e
>
>
<
r rsy
4
>
>
: r r r r esu e
su
su
sy
esu e
sh
0 < e < ey
10
e P etm
412
Table 3 resumes the type and number of nite elements employed in the 2D and 3D pushover analyses.
The parameters of the smeared xed crack model and the coupled elastoplastic damage model for reproducing the concrete
behavior are reported in Tables 4 and 5.
Linear elastic behavior is assumed for both bottom and top
beams with the same Youngs modulus and Poisson coefcient as
given in previous tables.
The steel reinforcement is modeled using PintoMenegotto
model whose parameters identied by prior tests on the steel bars
are reported in Table 6.
5.2. Meshes and material properties for CFRP-reinforced RC walls
The CFRP strips are numerically introduced into the previous
models of RC walls as shown in Fig. 16. Each CFRP strip is modeled
by three truss-bars, 2.5 cm distant from each other corresponding
to the chosen nite element size, and having a total section equal
to the section of the CFRP strip, that is 50 0.48 = 24 mm2. Perfect
bond assumption between the CFRP bars and the concrete is
adopted.
As discussed in Section 4.3, we assume an elastoplastic model
for the CFRP material, characterized by a yield stress rsy equal to
472.5 MPa calculated by:
11
413
Fig. 15. Mesh of the short wall in 2D analysis and boundary conditions.
Table 3
Parameters of the 2D and 3D models for the slender and short walls.
Approach
Material
Element type
Table 6
Properties of the steel reinforcement.
Quantity
QUA4
SEG2
CUB8
SEG2
1800
864
7200
1728
QUA4
SEG2
CUB8
SEG2
1800
1100
7200
1968
QUA4
SEG2
CUB8
SEG2
4224
1242
18,000
2484
QUA4
SEG2
CUB8
SEG2
4224
2124
18,000
4968
Steel
E (GPa)
rsy (MPa)
ey
esh
rsu (MPa)
esu
210
500
2.38E3
3.5E3
550
0.05
Table 4
Numerical parameters of the model concrete INSA.
E0
fc
ft
etm
erupt
Denition
Short wall
Slender wall
Youngs modulus
Uniaxial compressive strength
Uniaxial tensile strength
Cracking strain
Fracture strain in compression
22..E3 MPa
32 MPa
1.6 MPa
5.E03
1.51E2
22.E3 MPa
32 MPa
1.6 MPa
5.E03
1.51E2
Table 5
Numerical parameters of the model concrete RICRAG.
E0
ft
ADir
AInd
c0
i0
Denition
Short wall
Slender wall
Youngs modulus
Ultimate stress in simple tension
Brittleness in tension
Brittleness in compression
Kinematics hardening
Nonlinear hardening
22.E3 MPa
1.6 MPa
1.E2 J m3
5.E04 J m3
7.E8 Pa
7.E07 Pa1
22.E3 MPa
1.6 MPa
2.E2 J m3
10.E04 J m3
7.E8 Pa
7.E07 Pa1
414
Table 7
Properties of the CFRP strips.
CFRP
E (GPa)
rsy (MPa)
esy
esh
rsu (MPa)
esu
105
472.5
4.5E3
6E3
500
0.05
415
Fig. 21. Pushover curves unretrotted and CFRP-retrotted short walls (VC1 and
VCR1).
416
Fig. 22. Pushover curves unretrotted and CFRP-retrotted short walls (VC1 and
VCR1) with Youngs modulus equal to 11 GPa.
Fig. 25. Failure modes of the unretrotted and CFRP-retrotted short walls (VC1 and VCR1) with Youngs modulus equal to 11 GPa.
Fig. 26. Percentage of Gauss points in concrete whose strain exceeds the threshold
of 1.5E3 (VE1 and VER1).
417
Fig. 29. Percentage of vertical (AV) and horizontal (AH) yielding steels for slender
walls (VE1 and VER1).
Fig. 27. Percentage of Gauss points in steels whose tensile strains exceed the
threshold of 2.6E3 (VE1 and VER1).
Fig. 30. Percentage of vertical (AV) and horizontal (AH) yielding steels for short
walls (VC1 and VCR1).
caught up and more yielding steels at the vicinity of failure are detected due to the distributive effect of the CFRP strips.
For the short walls, the percentages of yielding steels are compared for the unretrotted and CFRP-reinforced walls in Fig. 28.
The distributive effect of the CFRP strips is again highlighted.
In order to distinguish the roles of longitudinal (vertical) and
transverse (horizontal) reinforcement with or without CFRP strips,
the percentages of yielding steels are computed independently for
vertical and horizontal reinforcement. For slender walls, these percentages are plotted in Fig. 29: it is interesting to note that the
CFRP strips result in yielding of horizontal steels whereas no horizontal steel yields without CFRP strips. It can be explained by the
fact that the CFRP strips result in more important shear effects.
For the short walls in Fig. 30, the behavior is a combined case of
shear and bending effects: both vertical and horizontal steel bars
yield for the unretrotted case, with a strong importance of the
Fig. 28. Percentage of Gauss points for steel material whose tensile strains exceed
the threshold of 2.6E3 (VC1 and VCR1).
418
[10] Antoniades KK, Salonikios TN, Kappos AJ. Tests on seismically damaged
reinforced concrete walls repaired and strengthened using ber-reinforced
polymers. J Compos Constr 2005;9:23646.
[11] Demeter I, Nagy-Gyrgy T, Stoian V, Dan D, Descu C. FRP composites for
seismic retrotting of RC wall panels with cut-out openings. In: Cruz P, editor.
Structures & architecture; 2010. p. 5412.
[12] Dan D. Experimental tests on seismically damaged composite steel concrete
walls retrotted with CFRP composites. Eng Struct 2012;45:33848.
[13] El-Sokkary H. Seismic retrot of reinforced concrete shear walls using bre
reinforced polymer composites. Ph.D thesis, Concordia University, Qubec,
Canada; 2012.
[14] Brun M, Labbe P, Bertrand D, Courtois A. Pseudo-dynamic tests on low-rise
shear walls and simplied model based on the structural frequency drift. Eng
Struct 2011;33:796812.
[15] Ile N, Reynouard JM. Behaviour of U-shaped walls subjected to uniaxial and
biaxial cyclic lateral loading. J Earthquake Eng 2005;9(1):6794.
[16] Ile N, Reynouard JM. Lightly reinforced walls subjected to multidirectional
seismic excitations: interpretation of CAMUS 2000-1 dynamic tests. ISET J
Earthquake Technol 2003;40:11735.
[17] Lu XZ, Teng JG, Ye LP, Jiang JJ. Meso-scale nite-element model for FRP sheets/
plates externally bonded to concrete. Eng Struct 2005;27(4):56475.
[18] Lu XZ, Teng JG, Ye LP, Jiang JJ. Bondslip models for FRP sheets/plates bonded
to concrete. Eng Struct 2005;27(6):92037.
[19] Lu XZ, Teng JG, Ye L, Jiang JJ. Intermediate crack debonding in FRPstrengthened RC beams: FE analysis and strength model. J Compos Constr
2007;11(2):16174.
[20] Matallah M, La Borderie C, Maurel O. A practical method to estimate crack
openings in concrete structures. Int J Numer Anal Meth Geomech
2010;34(15):161533.
[21] Le Fichoux E. Prsentation Et Utilisation De Cast3m. Support of CEA; 2011.
<http://www-cast3m.cea.fr>.
[22] Ile N, Reynouard J. Nonlinear analysis of reinforced concrete shear wall under
earthquake loading. J Earthquake Eng 2000;4(2):183213.
[23] Hillerborg A, Moder M, Petersson P-E. Analysis of crack formation and crack
growth in concrete by means of fracture mechanics and nite elements. Cem
Concr Res 1976;6:77381.
[24] Bazant ZP, Oh BH. Crack band theory for fracture of concrete. Matriaux et
Construction 1983;16(3):15577.
[25] Richard B, Ragueneau F. C. Cremona, and L. Adelaide, Isotropic continuum
damage mechanics for concrete under cyclic loading: stiffness recovery,
inelastic strains and frictional sliding. Eng Fract Mech 2010;77(8):120323.
[26] Richard B, Adelaide L, Cremona C, Orcesi A. A methodology for robust updating
of nonlinear structural models. Eng Struct 2012;41:35672.
[27] Merabet O. Modlisation des structures planes en bton arm sous
chargement monotone et cyclique. Ph.D thesis, INSA de Lyon, LyonFrance;
1990.
[28] Comite Euro-International du Beton. CEB-FIP Model Code 1990: design code.
Thomas Telford Limited; 1993.
[29] Rots J. Computational modeling of concrete fracture. Ph.D thesis, Delft
University of Technology, Netherlands; 1988.
[30] Feenstra PH. Computational aspects of biaxial stress in plain and reinforced
concrete. Ph.D thesis, Delft University Press; 1993.
[31] La Borderie C. Unilateral effects for damage-like materials: modelling and
application to concrete structures. Ph.D thesis, University Paris VI, Paris,
French; 1991.
[32] Menegotto M, Pinto P. Method of analysis for cyclically loaded reinforced
concrete plane frames including changes in geometry and non-elastic
behaviour of elements under combined normal force and bending. In: Proc,
IABSE symp of resistance and ultimate deformability of structures acted on by
well dened repeated loads, International Association of Bridge and Structural
Engineering, Libson, Portugal, vol. 13; 1973. p. 1522.
[33] Charras T, Di Paola F. La procdure PASAPAS. Support of CEA; 2011. <http://
www-cast3m.cea.fr>.
[34] Greifenhagen C, Lestuzzi P. Static cyclic tests on lightly reinforced concrete
shear walls. Eng Struct 2005;27(11):170312.