Finite Element Modelling of Wrapped Concrete Specimens Subjected To Localised Axial Compression
Finite Element Modelling of Wrapped Concrete Specimens Subjected To Localised Axial Compression
Finite Element Modelling of Wrapped Concrete Specimens Subjected To Localised Axial Compression
https://doi.org/10.17576/jkukm-2021-33(4)-35
University of Mosul, College of Engineering, Civil Engineering Dept. Mosul 41002, Iraq
ABSTRACT
The strengthening and rehabilitation of reinforced concrete structures using Carbon Fibre Reinforced Polymer (CFRP)
wrapping is most common. Recently, this method was used in the rehabilitation of deteriorated reinforced concrete
columns, the pier of girder bridges and corbels that are subjected to localised axial compression. The numerical
simulation which is verified by experimental works minimizes the time and cost to get the internal behaviours of
structures. This study is an attempt to numerically investigate the performance of CFRP wrapped concrete specimens
subjected to localised compressive loading conditions. The finite element modelling by ABAQUS software was used in
the simulation of adopted specimens in this study. The FEM results show good agreement with the experimental data.
The performance of adopted specimens and the behaviour of the concrete core and CFRP wrapping were extensively
studied. The FEM results indicated that the CFRP wrapping improves the load carrying capacity and increases the
dissipation of energy by increasing the deformation capacity and subsequently the wrapped specimens behave more
ductile. The fully wrapped cylinder subjected to a smaller area of loading exhibits higher capacity (119%) of exposed
specimen, while the wrapped cube subjected to area of loading (size:75 mm) exhibits higher load carrying capacity
about(52%) than the exposed cube.
Keywords: CFRP Wrapping; Finite Element Modelling; Localised Axial Compression; Strengthening and Rehabilitation;
Concrete Confinement.
strongly dependent on the resistance of concrete blocks time and cost to obtain initial estimates of the structure's
loaded through a steel bearing plate (Scheffers and performance. FEM is used to generate numerical data which
Ravindrarajah 2009). Most of the existing formulas in the then provides a basis for validating experimental analyses
American Concrete Institute (ACI) and the AASHTO (LFRD) (Liu et al. 2011, Abdullah 2010 and Hibbitt et al. 2006).
for predicting the concrete bearing capacity depend on In the last few decades, there is a growing number of
concrete compressive strength and ratio of the concrete- sophisticated FEM commercial software. ABAQUS is
to-steel area (Ravindrarajah and Reinaldy 2010). appropriate software used in structural modelling which
Many studies have been conducted to investigate the was adopted in the current study to simulate the
performance of exposed concrete parts subjected to experimental performance of CFRP wrapped concrete
localised axial compression (Hawkins 1968, Niyogi 1973 specimens subjected to different localised areas of axial
and Chen 1971). Few studies have been taken of the loading. In the present study, the performance of CFRP
confined concrete members subjected to this loading wrapped standard cylinders and cubes subjected to
condition. Han et al (2008) conducted experimental and localised axial compression is numerically investigated.
theoretical research to investigate the behaviour of concrete The main objectives of this research are folded into three
filled steel tubular stub columns subjected to axially local aims: first, to develop a numerical model using FEM to
compression. The results showed that the bearing resistance simulate the CFRP wrapped specimens under a localised
is affected by the thickness of the walled tube and the axial compression and verify it using recent experimental
bearing plate. data; second, to analyze the influences of several
Ravindrarajah and Reinaldy (2010) experimentally parameters, such as the shape of specimens, the ratio of
investigated the bearing behaviour of CFRP wrapped localised compressive area on the behaviour of CFRP
concrete. The parameters of the study were the bearing wrapped specimens; and third, to analyze the mechanism
strength ratio and bearing shape ratio. Results showed that of simulated specimens subjected to localised axial
the bearing strength of concrete is enhanced due to CFRP compression using stress distribution in constructional parts
wrapping, as well as, it was depicted that the shape of the of specimens (i.e. the concrete core and CFRP wrapping).
bearing area had a minor effect on the bearing capacity of
concrete.
Mai et al (2019) conducted an experimental investigation DATA COLLECTIONS
on square and circularised square RC columns intermittently
wrapped with CFRP under different load cases. The test This study took the performance of concrete elements
results proved that the intermittent confining enhances the strengthened using CFRP wrapping into considerations.
strength and ductility of square columns, as well as, the The experimental results of a recent study were taken to
circularisation and intermittent confining significantly verify the accuracy of the developed model herein
improve the ductility and load capacity of columns. (Abdullah et al. 2020). The details of adopted specimens
Shen and others (2019) numerically investigated the are summarised in Table 1; further, the testing procedure
performance of circular concrete filled steel tubular stub of the simulated specimens was conducted according to
columns partially wrapped using CFRP belts. Two design the scheme shown in Figure 1.
formulas were suggested to assess the axial compressive
capacity of partially wrapped columns in compliance with
the tubed concrete method and the unified theory. FINITE ELEMENT MODELLING
Zheng et al. (2018) presented an experimental
investigation on the axial compressive behaviour of
A nonlinear three-dimensional (3D) FEM using ABAQUS
partially FRP wrapped circular reinforced concrete
software (Abdullah 2010 and Hibbitt et al. 2006) was
columns. The test results were compared with existing
utilized in simulating the performance of localised axial
representative stress-strain models to check the reliability
compression of CFRP wrapped concrete specimens. The
and accuracy of these models.
types of elements were carefully selected to simulate the
Abdullah and others (2020) conducted an experimental
parts of the adopted specimens to get an accurate model.
study to investigate the effect of CFRP wrapping length on
The elements' types and descriptions are summarised in
the bearing resistance of confined concrete specimens. The
Table 2 and the scheme of Figure 2. The concrete core
results appeared to show that the full wrapping has little
was solid which was confined by lateral CFRP wrapping
effect on the bearing resistance of specimens applied to
consisting of one layer with overlap length (100 mm).
small areas of loading.
The mesh size of elements was taken 25 mm for all
Finite element modelling (FEM) is a familiar tool
specimens.
within the field of structural engineering. It reduces the
1125
...continued
CYH-150-C 200
CYH-150-D 300
Group-6 CUN-50-A Cube 150 mm 36.8 50 0
CUN-50-B 150
Group-7 CUN-75-A Cube 150 mm 36.8 75 0
CUN-75-B 150
Group-8 CUN-100-A Cube 150 mm 36.8 100 0
CUN-100-B 150
Group-9 CUN-150-A Cube 150 mm 36.8 150 0
CUN-150-B 150
The boundary conditions and the applied loading bearing plate was modelled as an elastic material to
process of the developed model were simulated by following implement the rigid behaviour of loading applied to the
the testing procedure as depicted in Figure 1. The localised concrete core. The materials' modelling is summarised in
axial load was applied incrementally using the General Table 2 and Figure 3. For concrete, the compressive and
Solution Method (GSM) available in the software. ABAQUS tensile behavior is shown in Figure 3-a. The compressive
generally uses Newton's method as a numerical technique strength (fc'), tensile strength (ft), the modulus of Elasticity
for solving the nonlinear equilibrium equations. Monotonic (Ec) and the Poisson's ratio (µ) were taken from experimental
localised axial load similar to that in the tests was subjected results. The tensile behaviour of CFRP is depicted in Figure
using the displacement control at nodes of the bearing plate 3-b. The modulus of Elasticity (ECFRP), the Poisson's ratio
surface. The time increment of loading did not exceed 10% (µCFRP) and thickness (t) of CFRP were taken from a data
of the total displacement. sheet provided by the manufacturer company. The modulus
In the developed model, the damaged plasticity model of Elasticity (Es) and Poison's ratio (µs) of the bearing plate
was used to simulate the concrete behaviour. The CFRP was inputted as depicted in Table 2.
wrapping behaves Elastic-Plastic in tension only. The steel
VERIFICATION OF FEM AND RESULTS' DISCUSSION others obtained from the developed model. The comparison
between the experimental and FEM load carrying capacity
Table 3 and Figure 4 show the verification of the developed is depicted in Figure 4. The figure also shows great
model of simulated specimens. As depicted in the table, matching between them. Figure 5 shows the comparison
the mean value of Experimental to FEM load carrying of experimental failure modes and numerical plastic strain
capacity (PEXP/PFEM) ratio of simulated specimens is 1.07 of concrete cores which boosts the good correlation
and the correlation factor (R) is 0.97 while the standard between the experimental and the FEM results. The ratio
deviation (SDV) is 0.11. The signed values by a star were of load carrying capacity of wrapped specimen to that
excluded from the test of the correlation due to differences exposed specimen (PWrp./PExp.) is presented in Table 3. The
between the experimental results and reasonable FEM ratio proves the improvement of load carrying capacity of
results. The statistical values indicate that an excellent specimens in the presence of wrapping. The wrapped
matching is achieved between the experimental results and cylinders subjected to a smaller area of loading exhibited
1128
higher capacity (119%) for specimens with a full length of In addition, the wrapped cubes subjected to bearing area
wrapping. While the wrapped specimens subjected to the of loading (size: 75 mm) exhibited higher load carrying
full area of loading exhibited lower capacity about (23%). capacity about (52%) than others.
FIGURE 4. Comparison between experimental and numerical load capacity of tested specimens.
FIGURE 5. Comparison of experimental failure shape and numerical plastic strain (Max. Principal) of concrete.
1130
Table 4 and Figure 6 depict the comparison of the increases. While the wrapped cylindrical specimen which
localised axial load capacity of both cubic and cylindrical was subjected to a smaller loading area presented better
specimens for exposed and full wrapped cases. The table performance and higher localised axial load capacity than
and figure indicate that the strength of wrapped specimens cubic due to the better confinement of specimen against a
progresses better than those exposed as bearing area small area of loading.
TABLE 4. Comparison of cubic and cylindrical exposed and full depth CFRP wrapped specimens.
Bearing B/b Pcube/Pcy
PFEM (kN) PFEM (kN) Pcube/Pcy PFEM (kN) PFEM (kN)
Plate Size or Full
(Exposed cube) (Exposed cylinder) Exposed (Full rapped cube) (Full wrapped cylinder)
(mm) D/d wrapped
50 3 304 218 1.4 366 477 0.8
100 1.5 656 384 1.7 840 530 1.6
150 1 1072 900 1.2 1220 778 1.6
FIGURE 6. Comparison of load caring capacity for cubic and cylindrical specimens.
Depending on the obtained results from FEM, Figure (e) shows the axial behaviour of specimens subjected to
7 shows the localised axial load versus the penetration of the full area of loading. The figure proves that the presence
the bearing plate in the concrete core. Figure 7(a, b and c) of CFRP wrapping and the increasing of its length enhance
depict the axial behaviours of specimens prepared using the load carrying capacity of the simulated specimens
normal concrete. Each figure shows the axial behaviour of which behave more ductile.
one of the group specimens that were tested in the same Figure 8 depicts the relation of a localised load of
conditions except the length of wrapping. As depicted in simulated cubes versus the penetration of the bearing plate
the figures, the group subjected to a smaller area of axial in the concrete core. All the specimens were prepared using
loading behaves more ductile than others due to the large normal strength concrete. The figures reveal that the
dissipation of energy of specimens at the test, as well as, presence of CFRP wrapping enhances the load carrying
the figures indicate that the CFRP wrapping improves the capacity and increases the dissipation of energy by
localised loading capacity of the confined specimens while increasing the deformation capacity and subsequently the
the length of the wrapping is ineffective especially with wrapped specimens behave more ductile.
specimens tested using the small bearing area. Figures 7(d Figures (9-11) show the relationships of Von Mises
and e) show the axial bearing behaviour of specimens (VM) stress of CFRP wrapping versus the applied load of
prepared of high strength concrete. Figure 7 (d) depicts the wrapped specimens. Figure 9 depicts the VM stress
localised axial behaviour of the specimens subjected to the behaviour of cylindrical wrapping for specimens prepared
small area of loading. The figure proves that the load and using normal concrete and different wrapping lengths.
deformation capacities of the specimens improved due to These relationships indicate that the CFRP wrapping is
the presence of the CFRP wrapping, while the length of ruptured at peak load only for the specimens applied to the
wrapping doesn't give any indication, which confirms that localised loading area with a size equal to (50 mm) and
the full wrapping of specimens isn't necessary. Figure 7 length of wrapping (200 and 300 mm). Figure 10 shows
1131
the VM stress behaviours of cylindrical wrapping of prepared using normal concrete. The figure shows that the
specimens prepared using high strength concrete. These specimens applied to larger areas of loading (100 and 150
curves reveal that the CFRP wrapping isn't ruptured at peak mm) show high VM stress at the peak load point and still
load, while it tends to fail after peak load, which can be oscillate around the half value of the tensile strength of the
attributed to the smaller Poisson's ratio compared to that CFRP (4300 MPa), while, the specimens applied to smaller
of normal concrete. Figure 11 depicts the VM stress loading areas (50 and 75 mm) show less than a quarter of
behaviour of the CFRP wrapping of cubic specimens the tensile strength of the CFRP at peak load point.
FIGURE 7. Localised Axial Load-Penetration curves of tested cylinders for normal and high strength concrete.
1132
FIGURE 9. Von-Mises stress of CFRP Vs. Applied load (Cylinders with Normal Concrete).
FIGURE 10. Von-Mises stress of CFRP Vs. Applied load (cylinders with High Strength Concrete).
FIGURE 11. Von-Mises stress of CFRP Vs. Applied load (Cubes with Normal Concrete).
1134
Figures 12 and 13 show respectively the lateral tensile small area of loading. In addition, Figures 12 & 13 (b &
stress of CFRP wrapping and VM stress of concrete core at c) show the lateral tension of wrapping and concrete stress
peak load for specimens subjected to different sizes of the of specimens' applied to the bearing plate with size (100
loading area. As well known, the lateral tension of CFRP mm) and the full area of loading, respectively. As shown
indicates the efficiency of confinement. Figures 12 & 13 in the figures, the higher lateral tension of CFRP wrapping
(a) depict that the higher lateral tension of CFRP wrapping and concrete stress lie near the mid height of specimens
and concrete stress lie at the upper third of specimens, which indicates that the full wrapping provides better
revealing that the full wrapping is not necessary with the confinement.
CONCLUSIONS ACKNOWLEDGEMENT