Analysis of FRP wrapped concrete columns under uniaxial compression
P Sangeetha*
Department of Mechanical Engineering, SCSVMV (Deemed University), Enathur, Kancheepuram 631 561 Received 27 June 2006; accepted 11 October 2006
Fiber wrapping or encasement of columns with fiber-reinforced plastic (FRP) sheets significantly enhances strength and ductility of concrete columns. To investigate the behavior of concrete columns confined by fiber reinforced polymer (FRP) sheets under uniaxial compression, analytical models were solved using Finite Element Method (FEM) against published experimental data. Cross sections of concrete columns in the analysis are categorized into circular, square and rectangular sections. Finite Element Analysis (FEA) can effectively simulate the behavior of concrete columns confined by FRP sheets when the proper numerical model is adopted. Results from a series of the analysis on small-scale specimens showed that confinement increased strength (20-25%) and ductility of concrete columns loaded axially. ANSYS (version 6.0) offers a series of very robust nonlinear capabilities for designs and analyses.
Keywords: Composite columns, FEM, Fiber wrapping, FRP
Introduction
Retrofitting of concrete columns (CCs) by wrapping and bonding of fiber-reinforced plastic (FRP) sheets, straps, belts, or precured shells around the columns has become increasingly popular. Studies of CCs, confined with glass, aramid, or carbon fibers, have been used successfully to retrofit building columns, bridge or expressway piers, and chimneys in Japan1. Many researchers2-7 have proposed models for concrete cylinders and square columns strengthened with FRP sheets. Rochette & Labossiere8 used anincremental finite element approach to evaluate response and behavior of square columns confined with carbon and aramid fiber sheets. Huang et al9 investigated the axial load behavior of concrete-filled tubular (CFT) columns with the width-to-thickness ratios. Hu et al10 proposed proper material constitutive models for CFT columns and verified by the nonlinear finite element program ABAQUS against experimental data. In this paper, finite element method is used to analyze the mechanism and the behavior with more rational model and failure criterion.
Materials and Methods
Column models were made of a normal strength concrete with uniaxial compression strength
(fc, 40 Mpa). Two prepreg composite materials, used to confine the column models fully, were unidirectional carbon fibers and bi-directional aramid-woven fabric, which were characterized by ASTM standard D 3039- M89 (Table 1). Stress-strain curve of the carbon lamina (Fig. 1a) and behavior of the aramid-woven fabric (Fig. 1b) illustrate clearly the difference in behavior between the two materials. Carbon fiber lamina is linearly elastic up to failure, whereas nonlinear relationship between stresses and strains in aramid- reinforced product is seen. The progressive tightening of fibers, as the tensile load is increased, leads to a progressive increase of stiffness explaining nonlinear behavior. Initial elastic modulus (Eo) defined as the slope of curve tangent at the origin, is less (50%) than the secant modulus (E), which is measured on a straight line between the origin of the system of axis and the failure point. Strength of carbon sheet is higher than that of aramid fabric.
Specimens
Identification of the specimens (Tables 2&3) is based on a two-part code, which provides condensed configuration details for each of them. The first portion of the code indicates shape of the section (cylinder, square or rectangular), its diam (100 or 150 mm) in the case of circular section, and the radius of the corners (5, 25, or 38 mm) for square and rectangular section (Fig. 2). The second part identifies the confining material (carbon or aramid) and the number plies.
_____________
*Tel: 044 27264301/ext 238 E-mail: p_sangeetha77@yahoo.co.in
J SCI IND RES VOL 66 MARCH 2007
236
Finite Element Analysis (FEA)
A series of models of various shapes (circular, square and rectangular) were wrapped with 2-5 plies of carbon fiber. Circular and square models were confined with 3, 6, 9, or 12 plies of aramid fiber-
woven around the columns axis, in the 0º orientation.
All specimens were simulated in ANSYS (version 6.0), which offers a series of very robust nonlinear capabilities for designs and analyses. SOLID 186 is used for the three-dimensional modeling of solids,
Table 1Properties of composite materials4
E X
Fibers
Gpa KN/mm² Mpa N/mm²
Eulti "
%
t mm
W g/cm³
Carbon 82.7 24.8 1,265 380 1.50 0.30 1.80
Aramid 13.6 5.71 230 96.6 1.69 0.42 *
*Not available, X-ultimate stress, t-thickness of fibre, w-density of fibre
Table 2Experimental data for carbon fibers4
Test results Sl No. Specimen code No. of plies Concrete
strength fc’
Mpa fz ult /fc’ εz,ult
%
εt,ult
%
εf, ult
% 1
2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
S-NC S5-C3 S25-C3 S38-C3 C100-C2 S5-C5 S25-C4 S25-C5 S-NC S25-C4 S25-C4 S38-C4 S38-C5 R-NC R25-C3 R38-C3 R5-C5 R25-C4
0 3 3 3 2 5 4 5 0 4 4 4 5 0 3 3 5 4
42.0 42.0 42.0 42.0 42.0 43.9 43.9 43.9 35.8 35.8 35.8 35.8 35.8 42.0 42.0 42.0 43.9 43.9
0.99 0.94 0.99 1.13 1.75 1.00 1.16 1.09 1.01 1.46 1.61 1.66 1.92 1.00 1.00 1.04 1.01 1.01
- 0.69 0.94 1.08 1.60 1.02 1.35 0.90 - 2.04 2.12 1.92 2.39 - 0.79 0.85 0.98 0.93
- 3.11 2.18 1.99 0.89 3.05 2.49 2.05 - 2.91 2.57 1.87 2.29 - 2.88 2.04 2.17 3.10
- 0.23 0.56 0.71 - 0.44 0.59 0.51 - 0.70 0.65 0.89 0.86 - 0.74 0.68 0.43 0.53 Table 3Experimental data for aramid fibers4
Test results Sl No. Specimen code No. of plies Concrete
strength fc’
Mpa fz ult /fc’ εz,ult
%
εt,ult
%
εf, ult
% 1
2 3 4 5 6 7 8 9 10 11 12 13 14
S5-A3 S5-A6 S5-A12 S25-A3 S25-A6 S25-A9 S25-A12 S38-A6 S38-A9 C150-A3 C150-A6 C150-A9 C150-12 S5-A9
3 6 12
3 6 9 12
6 9 3 6 9 12
9
43.0 43.0 43.0 43.0 43.0 43.0 43.0 43.0 43.0 43.0 43.0 43.0 43.0 43.0
1.18 1.20 1.26 1.19 1.19 1.24 1.28 1.18 1.23 1.10 1.37 1.65 1.73 1.25
1.06 1.49 2.08 1.24 0.79 0.97 1.10 1.26 0.96 1.18 1.11 1.47 1.69 1.74
2.60 2.68 3.00 6.52 2.64 2.11 2.57 2.24 0.69 1.75 1.55 1.39 1.33 1.18
0.79 1.30 1.48 0.90 1.12 1.27 0.94 1.04 1.05 0.97 1.53 - - -
which are capable of cracking in tension, crushing in compression, creep nonlinearity and large deflection geometrical nonlinearity. The fiber sheets as anisotropic material, which adopted SHELL 93 element, is a 3-D element having membrane (in plane) stiffness.
Mesh and Boundary Conditions
Fig. 3 shows meshed columns and boundary conditions. All joints of elements satisfy displacement coordination, including intersections of fiber and concrete. Exploiting symmetry of columns, only a quarter models were used in calculations. One end of columns was fixed where there was no degree of freedom. Uniform compressive loading was adopted in the calculation.
Results and Discussion
FRP wrapped CCs under uniaxial compression were analyzed using ANSYS (6.0) from a series of 32 columns (carbon 18, aramid 14) of square, rectangular and circular cross-section. Parameters that are varied
include cross sectional shape, corner radius, and number and types of plies.
Effect of Confinement
Influence of confinement on the behavior of column models are studied by plotting stress-strain curve for various sections with corner radius of 5, 25 and 38 mm. Square column (corner radius, 25 and 38 mm) with varying carbon plies (0, 3, 4 and 5) shows increase in the maximum stress for change in plies from 0 to 3, 0 to 4 and 0 to 5 as 12%, 17% and 5% respectively (Fig. 4a). The decrease in strain for a change in plies from 3 to 4 is 12.5 % and from 4 to 5 is 10%. Square column (corner radius, 5, 25 and 38 mm) with 3, 6, 9, and 12 numbers of aramid fibers shows increase in the maximum stress for change in plies from 3 to 6, 3 to 9 and 3 to 12 as 7%, 8% and
Fig. 1Stress-strain curves of composite materials
Fig. 2Dimensions of the specimens
Fig. 3Meshed column model
J SCI IND RES VOL 66 MARCH 2007
238
15% respectively (Fig. 4b). The decrease in strain for change in plies from 3 to 6 is 6 %, 3 to 9 is 11.2 % and from 3 to 12 is 17%. Deformation (Fig. 5), stress (Fig. 6) and strain (Fig. 7) contour along Y- direction for square section with 4 plies of carbon fiber (S25-C4) are shown.
Effect of Corner Radius
Increase in maximum stress (3 %) and decrease in strain (5 %) are observed for a change in corner radius
from 25 to 38 (Fig. 8a). With varying corner radiuses (5, 25 and 38 mm), increase in maximum stress is as follows: 25 for 5, 13; and 38 for 5, 15% (Fig. 8b).
With a change in corner radius, decrease in strain is as follows: 38 for 5, 15; and 25 for 5, 37%. Corner smoothening do not increase elastic limit load (Fig.9); may be cause of the effect of loss of area due to the corner radius.
Fig. 4Stress-strain curve for square column with varying plies
Fig. 5Deformed shape of square column (S25-C4)
Fig. 6Stress along Y-direction of square column (S25-C4)
Fig. 7Strain along Y-direction of square column (S25-C4)
J SCI IND RES VOL 66 MARCH 2007
240
Fig. 8Stress-strain curve for rectangular and square column
Fig. 9Corner radius - % difference in load for square and rectangular column
Fig. 10Stress-strain curve for square column with varying ply material (carbon & aramid)
Effect of FRP Materials
Stress vs strain plot (Fig. 10a) for square column having corner radius of 5 mm with 3 plies of carbon and aramid fibers shows an increase in maximum stress (2 %) and in strain (36%) for a change in fiber from carbon to aramid. Stress vs strain plot (Fig. 10b) for square column having corner radius of 25 mm with 3 plies of carbon and aramid fibers shows an increase in maximum stress (4%) and decrease in strain (45%)for a change in fiber from carbon to aramid. The experimental and analytical stresses for
carbon and aramid fibers were compared (Tables 4&5).
Conclusions
Multiple placements of FRP plies improve the overall performance for square, rectangular and circular sections. The most effective confinements are obtained for circular section. Carbon fibers are more effective than aramid fibers at increasing the ultimate axial strength and ultimate axial strain of various reinforced concrete columns. Corner radius plays an
Table 4Comparison between experimental and analytical axial stresses – carbon fibre
Sl No. Model Experimental axial stress
KN/mm2
Analytical axial stress KN/mm2
1 S-NC 0.0416 0.0401
2 S5-C3 0.0395 0.0421
3 S25-C3 0.0416 0.0450
4 S38-C3 0.0475 0.0450
5 C100-C2 0.0734 0.0744
6 S5-C5 0.0439 0.0461
7 S25-C4 0.0509 0.0473
8 S25-C5 0.0479 0.0427
9 S-NC 0.0363 0.0367
10 S25-C4 0.0523 0.0529
11 S25-C4 0.0576 0.0595
12 S38-C4 0.0594 0.0502
13 S38-C5 0.0687 0.0580
14 R-NC 0.0420 0.0437
15 R25-C3 0.0420 0.0706
16 R38-C3 0.0437 0.0690
17 R5-C5 0.0443 0.0584
18 R25-C4 0.0443 0.0449
Table 5Comparison between experimental and analytical axial stresses – aramid fibre
Sl No. Model Experimental axial stress
KN/mm2
Analytical axial stress KN/mm2
1 S5-A3 0.0507 0.0462
2 S5-A6 0.0516 0.0430
3 S5-A12 0.0542 0.0714
4 S25-A3 0.0512 0.0466
5 S25-A6 0.0512 0.0495
6 S25-A9 0.0533 0.0470
7 S25-A12 0.0533 0.0463
8 S38-A6 0.0540 0.0495
9 S38-A9 0.0529 0.0469
10 C150-A3 0.0473 0.0728
11 C150-A6 0.0589 0.0737
12 C150-A9 0.0709 0.0706
13 C150-A12 0.0844 0.0714
14 S5-A9 0.0538 0.0517
J SCI IND RES VOL 66 MARCH 2007
242
important role on the mechanical properties of carbon and aramid fiber. As corner radius increases, the efficiency of the FRP wrapping increases. The experimental and analytical stresses for carbon and aramid fibers were compared and the percentage between them is within 10%.
References
1 Mirmiran A & Shahawy M, Behavior of concrete columns confined by fiber composites, J Struct Engg, 123 (1997) 583- 590.
2 Samaan M, Mirmiran A & Shahawy M, Model of concrete confined by fiber composites, J Struct Engg, 124 (1998) 1025-1031.
3 Marijn R S & Giorgio M, FRP-confined concrete model, J Composites Const, 3 (1999) 143-150.
4 Rochettee P & Labossiere P, Axial testing of rectangular column model confined with composites, J Comp Const, 4
(2000) 129-136.
5 Parvin A & Wang W, Behavior of FRP jacketed concrete columns under eccentric loading, J Comp Const, 4 (2000) 146-152.
6 Yang X, Nanni A & Chen G, Effect of corner radius of performance of externally bonded FRP reinforcement, Proc 5th Int Symp on Fiber-Reinforced Polymer for Reinforced Concrete Structures (FRPRCS-5) 16-18 July 2001, 197-204.
7 Cole C & Belarbi A, Coefficient Characteristics of Rectangular FRP Jacketed RC Columns, Proc 5th Int Symp on Fiber-Reinforced Polymer for Reinforced Concrete Structures (FRPRCS-5) 16-18 July 2001, 828-832.
8 Huang C S, Yeh Y K, Liu G Y, Hu H T, Tsai K C, Weng Y T, Wang S H & Wu M H, Axial load behaviour of stiffened concrete-filled steel columns, J Struct Engg, 128 (2002) 1222-1230.
9 Hsuan-Teh Hu, Chiung-Shiann Huang, Ming-Hsien Wu &
Yih-Min Wu, Nonlinear analysis of axially loaded concrete- filled tube columns with confinement effect, J Struct Engg, 129 (2003) 1322-1329.