Documente Academic
Documente Profesional
Documente Cultură
net/publication/266143966
Bond stress behavior between concrete and steel rebar: Critical investigation
of pull-out test via Finite Element Modeling
Article in International Journal for Computational Civil and Structural Engineering · September 2014
DOI: 10.6088/ijcser.2014050008
CITATIONS READS
9 3,284
2 authors:
Some of the authors of this publication are also working on these related projects:
Seismic performance of beam-column joint reinforced with different shape memory alloy alternatives View project
All content following this page was uploaded by Md. Rashedul Kabir on 28 September 2014.
ABSTRACT
Bond performance of reinforced concrete is of prime importance in the study of load transfer
mechanism from concrete to inner reinforcing bar and vice versa. The bond-slip behavior and
ultimate pull-out strength for pull-out specimens are mostly covered in several researches.
Slip occurs in the interface is primarily due difference in stress between concrete and
reinforcement. Initiation of crack in surrounding concrete is also caused by this difference. In
this article, conscientious analysis of the transfer of bond stress in concrete surrounding the
reinforcing steel is done. Generated stress in the transition region between steel concrete is
detected by finite element modeling using ANSYS 11.0 program for different pull-out
specimens of varying concrete strength. Development of longitudinal and radial shear stress
in concrete due to applied load on a 20mm diameter steel rebar is observed. Considering
perfect bonding between concrete and steel, the most vulnerable zone which is prone to crack
formation in pull-out specimens is also identified regarding their stress distribution patterns
obtained from finite element modeling.
Keywords: Bond Stress, Pull-out Test, Finite Element (FE) Modeling and Analysis, Perfect
Bond, ANSYS.
1. Introduction
Concrete is a very strong material in compression and reinforcing steel in tension. But their
integrated action and good bonding is vital for structural strength and serviceability. This
integrated action is represented by bond stress developed between concrete and reinforcing
steel during loading. Whenever external load is applied on concrete, reinforcing bar receives
part of the load through load transfer mechanism from concrete to steel. When tensile force is
applied to the reinforcing bar, it develops stress components parallel and perpendicular to the
contact surface (Figure 1a). The stress parallel to bar is termed bond stress. Radial stress
generated perpendicular to contact surface is presented by shear stress in XY plane of concrete
(Figure 1b).
Interaction between concrete and reinforcing bar is due to chemical adhesion, friction,
mechanical interlock and shear along cylindrical concrete surface between adjacent ribs
(Belarbi et al. 2010). The first two properties contribute more in case of plain bar. For
deformed bar, the surface roughness and closely spaced ribs produce great interlocking with
bearing against the key formed between concrete and ribs. For a given concrete and a given
test apparatus, pullout strengths can be related to compressive strength test results. Such
strength relationships depend on the configuration of the embedded insert, bearing ring
dimensions, depth of embedment, and level of strength development in that concrete (ASTM
C 900 – 06). This is the major reason for superior bond performance and effective load
transfer. Transmitted force to the surrounding concrete has two components. Component
parallel to bar is responsible for the breakdown of adhesion and frictional resistance which
leads to pull-out failure. The perpendicular component creates circumferential or radial stress
which may lead to splitting type of failure.
Most of the experimental investigations done by the researchers are limited on finding
ultimate bond stress and bond-slip model for varying parameter. Finite element modeling
(FEM) is preferred by numerous researchers since it can evade experimental hurdle. Proper
modeling can resemble the experimental result. From beginning to the recent advancement,
FEM has been used by several researchers (Ngo and Scordeis, 1967; Bamonte et al, 2003;
Sezen and Mohle, 2003; Jendele and Cervenka, 2006, and Khalfallah and Ouchenane, 2007).
Shafaie et al. (2009) studied the stress distribution in reinforced concrete from finite element
modeling. Numerical prediction of bond-slip behavior in simple pullout test was investigated
by Al-Zuhairi and Al-Fatlawi (2013).
Purpose of this paper is to scrutinize the stress distribution in the concrete surrounding the
reinforcing bar. Change in stresses both in parallel and transverse to reinforcement along Y
direction are explored to find the basis of formation of primary and secondary crack in
concrete adjacent to it. Considering perfect bond between concrete and reinforcement, change
in stress distribution in relation to concrete strength is inspected. Most importantly, sudden
change in stress distribution in concrete parallel to reinforcement from the starting point of
contact to the end of contact due to formation of crack in high loading is depicted here. Zone
which is more vulnerable to formation of crack is located from the stress distribution
behavior along the depth of concrete.
(a) (b)
Figure 1: (a) Force components parallel and perpendicular to the steel concrete interface
(b) Shear stress distribution in XY plane of concrete.
2.1 Strategy
The concrete is modeled using SOLID65 eight-node brick element, which is capable of
simulating the cracking in tension and crushing in compression behavior (Figure 2a). It
simulates the elastic and plastic deformations that would happen in concrete and steel
reinforcement inclusive of cracking until ultimately concrete crushing as the load is stepwise
increased. It is a dedicated three-dimensional eight noded element with three degrees of
freedom at each node, i.e. translations in the x, y and z directions. Two noded LINK8 bar is
generally used in modeling steel reinforcement. It works like a line element with two nodes at
ends. But in this study, reinforcement is modeled using SOLID45 element (Figure 2b) which
has identical degrees of freedom to those for SOLID65 (ANSYS 2005). To relate with the
practical situation of pull-out test, SOLID45 is chosen over LINK8 to model 3D eight noded
elements. Perfect bonding between these two elements is considered during modeling.
(a) (b)
While modeling, a cylinder length equals to the required embedded length 240mm and
150mm diameter are considered. Extensive use of deformed bar has led us to consider perfect
bonding between concrete and reinforcing steel bar. To resemble the experimental scenario,
concentric steel bar of 20 mm diameter and 736mm (including 240mm embedded length)
length is modeled (Figure 3a). To obtain precise results in concrete vicinity of reinforcement,
meshing is done much finer close to the reinforcement (Figure 3a, 3b).
Displacement boundary condition is applied as load at top of the reinforcement (Figure 3c).
The boundary conditions are also applied to the concrete cylinder by restraining translation of
nodes at top surface of cylinder in three directions except the nodes adjacent and neighboring
to the reinforcement. Displacement at top of the reinforcement is given only in positive Y
direction (upward).
496mm
Finer mesh near steel
150mm
240mm
Figure 3: Geometry of Pull-out specimen (a) mesh along depth, (b) mesh along width,
(c) boundary conditions.
2.5.1 Concrete
Concrete shows different behavior in tension and compression. The ultimate compressive
strength can be used to predict the elastic modulus of concrete. Stress-strain plot based on
compressive strength on concrete is determined from the equation proposed by MacGregor,
1992, i.e., f= Ecε/1+(ε/εo)2 ; Here, f= stress at any strain ε, ε= strain at stress f psi, εo= strain at
ultimate compressive strength fc’ psi and E= elastic modulus of concrete 57000√ fc’ psi (ACI
318-11)). In each case, strain in concrete at ultimate strength is taken 0.003. Table 1 provides
the properties of concrete used to analyze the FE models.
Poisson’s ratio, density, elastic modulus, yield stress and tangent modulus are defined for
different grade of steel. Since the stress distribution in surrounding concrete mostly depends
on the concrete properties rather than the steel grade, comparisons are made over Grade 60
(yield stress 420 Mpa) steel reinforcement for different concrete strength. The properties of
steel rebar are tabulated in Table 1 with designated labels which are used in analysis.
The load is applied as displacement boundary condition and the applied displacement is
provided in 500 steps followed by 2 sub-steps for each step. Figure 4 shows the applied
displacement vs. time step curve for different models. It is clear that the concrete with higher
compressive strength is capable to carry higher displacement before failure.
0.6
Applied displacement (mm)
0.5
0.4
0.3
0.2
0.1 C3.5S60
C4.5S60
C2.5S60
0.0
0 10 20 30 40 50 60
Time step
The stress in steel outside the concrete is invariable throughout the length and the stress in
steel inside the concrete decreases gradually in a parabolic fashion and zero at the end. Sudden
fall in stress is detected at the top contact surface due to griping of concrete. Variations in
tensile stress distributions with length of steel rebar for different model are shown in Figure 5.
This pattern is similar in all models but concrete of higher strength creates higher tensile stress
in steel which indicates better bonding with steel rebar.
Figure 5: Tensile stress (Y direction) distribution along the length of steel (a)C2.5S60,
(b)C3.5S60, (c)C4.5S60
The Y-direction stress in steel rebar and surrounding concrete are shown in Figure 6, 7 and 8
for concrete with 31, 24 and 17 MPa compressive strength respectively bonded with 60 grade
steel rebar. The difference in the stress between steel and surrounding concrete at the upper
part of specimens is found significantly large which is an evidence of inter-facial stress
development and possibility of slip. This inter-facial stress causes the slip of steel rebar as a
result of localized cracking of concrete due to stress difference between steel rebar and
surrounding concrete. Change in stress from tensile to compressive also occurs in concrete due
formation of crack. This mechanism occurs dominantly at the top portion of the specimen and
similar in all specimens (Figure 6, 7 and 8). Concrete with higher compressive strength
develops more stress to generate crack in it.
250 250
200 200
150 150
Time step
0.5
3
3.5
7
100 100
8
Time step
8.5
0.5 14.5
3 15
7 18
50 15 50 21
21.5 21.5
26 25
32 31
36 36
0 0
0 20 40 60 80 100 120 140 -14 -12 -10 -8 -6 -4 -2 0 2
Stress along Y direction (MPa) Stress along Y direction (MPa)
(a) (b)
Figure 6: Stress distribution in (a) steel and (b) surrounding concrete along Y direction at
different level for different time step (C2.5S60)
250 250
200 200
Depth of cylinder (mm)
Steel length (mm)
150 150
Time step
0.5
2.5
100 100 3
Time step 5.5
0.5 7
5 9
10 11.5
50 15 50 12
20 14.5
25 22
30 29.5
34 34
0 0
0 20 40 60 80 100 120 140 -14 -12 -10 -8 -6 -4 -2 0 2
Stress along Y direction (MPa) Stress along Y direction (MPa)
(a) (b)
Figure 7: Stress in (a) steel and (b) surrounding concrete along Y direction at different level
for different time step (C3.5S60)
250 250
200 200
Figure 8: Stress in (a) steel and (b) surrounding concrete along Y direction at different level
for different time step (C4.5S60)
The shear stress distributions in concrete at XY Plane are shown in Figure 9. The shapes of
shear stress distribution are quite similar but higher shear capacity found for concrete with
higher compressive strength. The location of maximum shear stress was found at the top
portion of the concrete cylinders which is the major cause of crack generation in radial
direction.
Time step
100 100 0.5 100 Time step
Time step 2.5 0.5
0.5 5 4
3 7.5 7
6 12 10
50 9 50 16 50 20
14 22 30
20 26 40
30 50
30
34 56
36
0 0 0
-14 -12 -10 -8 -6 -4 -2 0 -14 -12 -10 -8 -6 -4 -2 0 -14 -12 -10 -8 -6 -4 -2 0
Shear Stress in XY plane (MPa) Shear Stress in XY plane (MPa) Shear Stress in XY plane (MPa)
(a) (b) (c)
Figure 9: Shear stress in surrounding concrete in XY plane at different level for different
time step (a)C2.5S60, (b)C3.5S60, (c)C4.5S60
Figure 10 represents stress distribution after FE analyses. Stress contours show that higher
intensity in the neighboring concrete which decrease with increasing distance from rebar. The
failure patterns are shown in Figure 11 at which the cracks (red dots) are clearly visible at the
location of higher intensity. Since both the shear and Y direction stress governs at the upper
portion of the cylinder sample; most of the cracks appear in this area.
(b) (c)
Figure 10: FE outcomes (a) shear stress distribution in XY plane, (b) Y direction stress
distribution.
Figure 11: Cracking failure of concrete due to pull out loading (a) top view, (b) isometric
view (c) elevation.
4. Conclusions
i. Tensile stress developed in steel during pull out reduces drastically at contact surface
with concrete and reaches to almost zero in a parabolic fashion. Higher strength
concrete results in higher tensile stress.
ii. Inter-facial stress difference between steel and surrounding concrete along Y-direction
at the upper part of specimen was notably high, which may cause slip and develop
localized crack in concrete.
iii. Maximum shear stress in concrete appears at the top portion of the concrete cylinders.
Shear stress is higher at the concrete close to steel rebar.
iv. Stress in concrete both in Y-direction and in XY plane increases with the increase in
concrete strength. High strength concrete generates more stress before failing occurs.
v. Most of the cracks appear at the upper portion of concrete cylinder since stresses in
this zone is significantly higher in both directions.
vi. After formation of crack, sudden change in stress direction occurs in concrete.
5. References
2. ACI 318-11, (2011), Building Code Requirements for Structural Concrete and
Commentry, An ACI Standard, Reported by ACI Committee 318, American Concrete
Institute, August.
3. ANSYS 2005, ANSYS Structural Analysis Guide, Version 11.0, ANSYS, Inc.,
Canonsburg, Pennsylvania.
4. ASTM C 900 – 06, Standard Test Method for Pullout Strength of Hardened Concrete,
American Society for Testing and Materials, ASTM Committee C09.
6. Bamonte, P., Coreonelli, D., and Gambarova, P.G., (2003) Smooth Anchored Bars in
NSC and HPC: A Study on Size Effect, Journal of Advanced Concrete Technology, Vol.
1, No.1, January, pp. 42-53.
7. Belarbi, A., Richardson, N.D., Swenty, M.K. and Taber, L.H., (2010), Effect of
Combination on Reinforcing Bar-Concrete Bond, Journal of Performance of Constructed
Facilities, ASCE, Vol. 24, No. 3, May-June.
8. Jendele, L. and Cervenka, V., (2006). A General Form of Dirichlet Boundary Conditions
Used in Finite Element Analysis. Paper 165, Proc. Eighth International Conference on
Computational Structures Technology, ed. B.H.V. Topping, G. Montero, and
R.Montenegro., Civil CompPress: Las Palmas. pp. 373-374.
9. Khalfallah, S., and Ouchenane, M., (2007), A Numerical Simulation of Bond for Pull-
out Tests the Direct Problem”, Asian Journal of Civil Engineering (Building and
Housing), Vol.8, No.5, pp. 491-505.
10. MacGregor, J.G., (1992), Reinforced Concrete Mechanics and Design, Prentice-Hall,
Inc., Englewood Cliffs, NJ.
11. Ngo, D. and Scordelis, A. C., (1967), Finite Element Analysis of Reinforced Concrete
Beams, ACI Journal, Vol. 64, No. 3, March, pp. 152- 163.
12. Shafaie, J., Hosseini, A. and Marefat, M.S., (2009), 3D Finite Element Modeling of
Bond-Slip Between Rebar and Concrete in Pull-Out Test, The Third International
Conference on Concrete and Development, Tehran, I.R. Iran.
13. Sezen, H., and Moehle, J.P., (2003), Bond-Slip Behavior of Reinforced Concrete
Members, The Earth Engineering Research Centers Program of the National Science
Foundation at the University of California, Berkeley.