Effect of Rebar Splice on the Seismic Response of Shear Wall

Boundary Element

M. Hoseini, S. Tariverdilo* , S. Gholizadeh

1

†
Department of Civil Engineering, Faculty of Engineering, Urmia University, Urmia, Iran.

Abstract

Due to the adverse effect of lap splice on the ductility of shear walls, ACI 318-19 does not allow the use of lap
splice in stories with probable nonlinear excursions. However, due to lack of capacity design in ACI design
approach for shear walls, in fact there is a high probability of nonlinear deformation in all stories. This means
possible nonlinear demand in stories with lap spliced bars. This paper investigates the performance of lap spliced
boundary elements of lightly reinforced shear walls, where boundary elements subjected to mainly tensile
loadings. Two specimens with continuous rebars and three specimens with lap splice of different lengths are
included in the experimental program. The specimens are subjected to asymmetric cyclic axial loading. Test
results reveal that lap splice increases deformation demand on the portions of the specimen outside of lap splice
length, leading to rebar fracture at lateral drifts much smaller than that except for specimens with continuous
rebars. The presence of lap splice could substantially decrease lateral drift capacity from 0.04 to about 0.013. Also
out of plane buckling for specimens with lap splice occurs at smaller tensile strains.

Kewords: Boundary element, Shear walls, Rebar fracture, Out of plane buckling, Lap splice.

1. Introduction
It is known that lap splice of longitudinal bars could adversely affect the ductility of reinforced concrete
shear walls. Accounting for this, ACI 318-19 [1] prohibits the use of lap splice in stories with possible
nonlinear deformation. However, due to lack of capacity design in the seismic design procedure adopted
by ACI for shear walls, nonlinear deformation could occur at different locations along the height of the
wall. Due to the inevitable existence of lap splice in stories with nonlinear deformation in the shear
wall, the question is how much nonlinear deformation capacity of the shear wall is affected by the
presence of lap splice .
There are limited experimental tests on the effect of lap splice on the seismic performance of shear
walls. Layssi and Mitchell [2] investigated possible retrofit schemes for shear walls designed using the
1960's detailing. They demonstrated that in addition to other factors related to poor detailing, lap splice
could significantly reduce the ductility of the shear wall. Lowes et al. [3] investigated how lap splice,
reinforcement layout, and shear demand could affect the seismic performance of shear walls. They
tested four specimens with and without lap splice with different shear demands and reinforcement
layouts. They found that splice shifts the location of nonlinear deformation to the top of a splice. They
also found that increase in the shear demand reduces maximum affordable lateral drift to less than 1%,
however lap splice does not have an appreciable impact on the ductility of the shear walls. It should be
noted that in their tests, longitudinal rebars in the splice length are confined adopting detailing
requirements of ACI for special structural walls. Contrary to this finding regarding the seismic effect
of lap splice, Aaleti et al. [4] found that lap splice significantly reduces the ductility of shear walls.
Testing three walls, with and without lap splice, they found that maximum lateral drift of 2% for wall

* 1
E-mail address of corresponding author: s.tariverdilo@urmia.ac.ir (S.Tariverdilo)
without splice reduces to about 1% for wall with splice. Also the failure mode from global buckling in
the specimen without splice changes to bar buckling and fracture for the wall with splice. Villalobos et
al. [5] studied the effect of confinement of boundary elements and lap splice on the seismic performance
of shear walls. Lap splice lengths were 40 and 60 times the rebar diameter. They reported no failure of
the splice, but significant reduction in ductility due to splice. They also reported a substantial increase
in the localized strain of longitudinal bars at the bottom end of lap splice. Almeida et al. [6] investigated
the cyclic response of asymmetric shear walls with poor detailing with and without lap splice. Lap splice
length was 35 times the bar diameter. They found that confinement and length of lap splice plays a
dominant role in the cyclic response of the shear walls.
As discussed in the proceeding paragraph, lap splice results in the localization of strain in the
longitudinal bar at the splice end sections, which could result in rebar fracture at low drift angles. This
could be aggravated for walls with a low ratio of longitudinal bars. Wood [7] analyzing tests on shear
walls concluded that in lightly reinforced walls, there are limited cracking with very large strain demand
on rebars that could lead to premature fracture of rebar. Lu et al. [8] conducting an experimental
program on lightly reinforced walls, found that a low ratio of longitudinal bars leads to a small number
of cracking and rebar fracture. Latter Lu et al. [9] using finite element models studied the effect of the
content of longitudinal bars under cyclic loading. They demonstrated that desirable response under
cyclic loading requires an increase in the ratio of longitudinal rebars in both web and boundary elements.
Rosso et al. [10] investigated the cyclic response of thin lightly reinforced boundary elements, with
main attention to out of plane buckling of the boundary element, rather than strain profile of the
longitudinal bars. They found that rebar ratio and wall thickness are the main parameters controlling
the out of plane buckling of thin and lightly reinforced boundary elements .
This study investigates the cyclic response of boundary elements of lightly reinforced shear wall, with
the main emphasis on the effect of lap splice on the seismic performance. For this purpose, specimens
without splice and with a splice of different lengths are considered in the program. As the specimens
represents lightly reinforced shear wall with low seismic demand, the specimens have nominal
confining reinforcement. In the following, the first experimental program including specimens
dimensions, reinforcement, and loading protocol are presented and then test results are reviewed.
2. Experimental Program

Near the wall critical section, there is a negligible moment gradient and nearly constant axial loading
on the boundary elements. Accounting for this, test setup (Figure 1) designed to apply axial cyclic
loading on the specimens. The experimental program includes cyclic tests of specimens with and
without lap splice as given in Table 1. Tests are conducted using a universal jack of 1000 KN capacity
in infrastructure research center of Urmia University.
Table 1. Samples description, geometry and reinforcement.
Sample Descrption f' c Dim. (mm) Splice Long. Rein. Trans.
Designation (MPa) w1xw2xl Length Bar Ratio Rein.
(mm)
BC1 Sample without rebar splice 30 150x150x1000 - T10 0.00347 T6@150
BC2 Sample without rebar splice 30 " - T10 0.00347 T6@150
SC1 Sample with rebar splice 39 " 200 T10 0.00347 T6@150
SC2 Sample with rebar splice 34 " 300 T10 0.00347 T6@150
SC3 Sample with rebar splice 39 " 400 T10 0.00347 T6@150
 Figure 1. Specimens setup and instrumentation

(a) symmetric loading in small displacements (b) asymmetric loading in large displacements

Figure 2. Loading protocol used in the experiments

Boundary elements of lightly reinforced shear walls are anticipated to experience mainly tensile loading
with a brief excursion into compressive loadings. Considering this and following Rosso et al. [10], an
asymmetric loading protocol is adopted for cyclic loading. The protocol is symmetric up to an axial
strain of about 0.003 (Figure 2a), after that there is steady increase in tensile loading amplitude, while
compressive strain becomes capped at 0.003 (Figure 2b).
Table 2 gives materials property for rebars. Reinforcements satisfy requirements regarding tensile
strength and elongation of ISIRI 3132 for S400 rebars. In addition to the tensile test of rebars, the
correlation between rebar strain and hardness is also determined. This is done using interrupted tensile
tests of the rebars at desires strain and then evaluating its hardness. Deriving the correlation between
rebar strain and hardness, it is possible to evaluate strain experienced along the rebar length at the end
of the test. This provides a better estimate of the local and uniform strain of the rebar at the test end.
Hardness tests are done employing indent universal hardness test machine. Figure 3 shows the derived
correlation between Rockwell B hardness and strain of the rebars.
Table 2. Material properties for reinforcements
Designation Material Property Test ISIRI 3132
Results Requirements
For S400
fy Yield Stress (MPa) 433 ≥400
fsu Ultimate Strength (MPa) 622 ≥600
fsu/ fy Ratio of ultimate to yield strength 1.44 ≥1.25
Fracture Elongation strain in 5db 0.30 ≥0.16
εfc_test Fracture Elongation strain in 10db 0.27 ≥0.12
Fracture Elongation strain in 200 mm 0.21 -
 Figure 3. Correlation between rebar residual strain and its hardness

3. Test Results
Figures 4 gives cracking pattern of the specimens at the end of test. Except for SC2, no cracking is
observed along splice length for the specimens with lap splice. Significant bar slip occurred in SC2,
probably due to smaller concrete strength of this specimen compared to SC1 and SC3. Reduced cracking
for SC1 and SC3 compared to BC1 and BC2, is the main reason for significant reduction in the axial
deformation capacity of these specimens. Except of SC2, where splice failure terminates test, the tests
are terminated due to rebar fracture. In most cases rebar fracture occur at element-foundation interface.
Figure 5 shows strain profile of the specimens evaluated using correlation between hardness and strain.
While there are significant residual strain (an indication of plastic deformation) along the rebar length
for BC1 and BC2, there is no residual strain along the significant portion of the rebar length for SC1
and SC3. In fact for SC1 and SC3, there are no plastic deformation except at crack proximities.
Increased stiffness and strength in the splice length, shifts crack locations to length outside of the splice.
Smaller strain along splice length, increases demand on the specimen length outside of splice, leading
to rebar fracture at much smaller deformations compared to the specimens without splice. Even for SC2,
where there is significant bar slip along splice length, the residual strain is much smaller than that for
BC1 and BC2. Strain profile could be used to find maximum local strain along rebar length and also
uniform strains. Uniform strain is calculated by dividing total elongation, evaluated by integrating strain
along the rebar length, by its length. Uniform and maximum strains for different specimens are reported
in Table 4. It should be noted that maximum strain values evaluated by this method have contributions
from flexural deformations and are not only due to axial deformation.
(a) BC1 (b) BC2 (c) SC1 (d) SC2 (e) SC3
Figure 4. Cracking pattern of specimens

(a) BC1 (b) BC2

(c) SC1 (d) SC2

 (e) SC3
Figure 5. Hardness and strain distribution along the rebars
To compare axial deformation capacity of the specimens with anticipated axial deformation demand, in
this study an estimation proposed by Dezhdar and Adebar [11] is used. Their proposed estimate of
mean (μφ) and mean plus one standard deviation (μφ+σφ) of curvature demand are as follow

 hw   roof
(  ) ,demand lw = 1.8 − 0.017  (1)
 Rd  hw
 hw   roof
(  +  ) ,demand lw =  2.8 − 0.022  (2)
 Rd  hw

where hw, lw are wall height and length, Δroof is roof displacement and Rd is ductility related force
modification factor (usually between 2 and 4.5). In lightly reinforced walls, depth of neutral axis in
comparison to the wall length is small, consequently tensile strain in the BE could be approximated by
φlw. Ignoring the second term in the parenthesis, a conservative estimate of axial deformation demand
will vary with drift as depicted in Figure 6. Also shown in this figure is axial deformation capacities of
the specimens. Considering deformation estimate provided by mean plus one standard deviation,
maximum affordable drift for different specimens are also depicted in this figure and reported in Table
4. Table 4 also compares axial deformation of the specimens evaluated by average strain and uniform
and maximum strains calculated using the correlation between hardness and strain. Significant
reduction of axial deformation capacity for spliced specimens is evident in the Figure 6 and Table 4.
While the specimens without splice could take drifts as large as 0.040, the specimens with splice length
of 40db could only withstand drifts as large as 0.013.
Figure 6. Comparison of anticipated demand and observed capacity of axial deformation for different specimens

Table 4. Elongation and deformation characteristics of different specimens.

Average
Average Uniform Maximum Permissible εcr
Specimen Elongation εcr‡
Strain* Strain** Strain** Drift† Equation 3
(mm)
BC1 111 0.111 0.144 0.381 0.040 0.032
BC2 84 0.084 0.087 0.357 0.031 0.032
SC1 13 0.013 0.013 0.047 0.005 0.011 0.0176
SC2 68 0.068 0.046†† 0.102 0.025 0.018
SC3 35 0.035 0.025 0.135 0.013 0.023
* Evaluated by dividing elongation by the specimen length
** Evaluated by employing correlation between hardness and strain
† Estimated using Dezhdar and Adebar proposed method (Equation 2)
†† Large contribution from bar slip
‡
Observed tensile strain triggering out of plane buckling

Relating curvature at mid span to strain in the longitudinal reinforcement and using moment area
theorem, Paulay and Preistley [12] found that average critical strain triggering out of plane buckling for
specimen with one layer of reinforcement could be estimated as
2
b
 cr = 1.6   (3)
 lo 
Table 4 compares observed tensile strain triggering out of plane buckling with its estimate obtained
using Equation 3. While estimates using this equation for spliced samples have good accuracy, it
underestimate strain triggering buckling for samples without splice.
Figure 7 depicts load-displacement diagram of the specimens. Significant reduction in ductility is
observed for all of the specimens with lap splice. Smaller splice length for SC1 results in significant
strength degradation and test termination at smal axial deformations. Onset of out of plane buckling is
also depicted by asterisk on the load-displacement diagram. Before and after buckling, there are
essential changes in the shape of load-deflection diagram. After initiation of buckling, significant
pinching in tensile and compressive reloading is evident in the figures for all of the specimens, which
is due to relatively zero stiffness of buckled specimen at start of tensile reloading and slip of the rebars
at start of compressive reloading. Due to buckling induced pinching, energy dissipation of the
specimens is significantly reduced after buckling (Figure 7f).

(a) BC1 (b) BC2

 (c) SC1 (d) SC2

(e) SC3 (f) BC2, before and after buckling

Figure 7. Load-displacement of specimens

4. Conclusion
Effect of lap splice on the seismic response of boundary elements of lightly reinforced shear walls is
experimentally investigated. Experimental program includes two specimens with continuous rebars and
three specimens with different lap splice lengths. Cyclic loading protocol is asymmetric with capped
compressive strains. Tests lead to the following results
• Reduced number of cracking occur in lap spliced specimens. This results in higher localization of
strain and rebar fracture at lateral drifts much smaller than that for the specimens with continuous
bars.
• Deriving strain profile of the rebars using correlation between hardness and strain, shows that
plastic strains are mainly limited to cracking locations in the spliced specimens, while for specimens
with continuous rebars, significant plastic strain occurs along rebar length other than crack points.
This further increases localized strain at limited crack locations for spliced samples.
• Out of plane buckling for lap spliced specimens occurs at much smaller axial strains.
• After buckling, there is significant reduction in the energy absorption capacity of the element.
Reference
1. ACI (American Concrete Institute) Building code requirements for structural concrete and commentary,
ACI318-19, Farmington Hills, MI, 2019.
2. Layssi H, Mitchell, D Experiments on seismic retrofit and repair of reinforced concrete shear walls,
Proceeding of 6th International Conference on FRP Composites in Civil Engineering, Rome, (2012).
3. Lowes LN, Lehman DE, Birely AC, Kuchma, DA, Marley, KP, Hart, CR Earthquake response of slender
planar concrete walls with modern detailing, Engineering Structures, 43(2012) 31-47.
4. Aaleti S, Brueggen BL, Johnson B, French CE, Sritharan S Cyclic response of reinforced concrete walls
with Different anchorage details: Experimental investigation, ASCE Journal of Structural Engineering,
139(2013), 1181-1191.
5. Villalobos E, Escolano-Margarit D, Ramìrez-Màrquez AL, Pujol S Seismic response of reinforced concrete
walls with lap splices, Bulletin of Earthquake Engineering, 15(2017), 2079–2100.
6. Almeida P.,Prodan O, Tarquini D, Beyer K Influence of lap splices on the deformation capacity of RC
walls. I: Database assembly, recent experimental data, and findings for model development, ASCE Journal
of Structural Engineering, 143(2017).
7. Wood SL Minimum tensile reinforcement requirements in walls, ACI Structural Journal, 86(1989), 582-
591.
8. Lu Y, Henri RS, Ma QT Numerical modelling and testing of concrete walls with minimum vertical
reinforcement, NZSEE conference, (2014).
9. Lu Y, Henry RS, Gultom R, Ma QT, 2017, Cyclic testing of reinforced concrete walls with distributed
minimum vertical reinforcement, ASCE Journal of Structural Engineering, 143(2017).
10. Rosso A, Jimenez-Roa LA, Almeida JP, Blando CA, Bonett RL, Beyer K Cyclic tensile-compressive tests
on thin concrete boundary elements with a single layer of reinforcement prone to out-of-plane instability,
Bulletin of Earthquake Engineering, 16(2018), 859-887.
11. Dezhdar E, Adebar P Estimating seismic demand on concrete shear wall buildings, 11th Canadian
Conference on Earthquake Engineering, (2012).
12. Paulay T, Priestley MJN, 1992, Seismic design of reinforced concrete and masonry building, John Wiley
and Sons, (1992).