Comparison of GFRP and CFRP confinement of normal and high strength concrete

An experimental and analytical study was conducted to quantify the increase in strength and ductility of GFRP and CFRP confined concrete cylinders made of normal strength concrete and high strength concrete. The test data are also compared with theoretical predictions from three North American and European FRP design guidelines. The test results show that the effectiveness of both GFRP and CFRP is more pronounced in normal strength concrete compared to high strength concrete.


Introduction
Structures currently in use are being subjected to increased service loads and severe environmental conditions. The rehabilitation and retrofitting of existing structures is traditionally done using steel or reinforced-concrete jacketing. But the relatively newer materials like fibre reinforced polymers are gaining in popularity because their properties are superior to those of conventional materials. Structural members reinforced with these materials have high strength to weight ratio, high corrosion resistance, longer life span, and enhanced ductility. Reinforced-concrete (RC) columns are critical members in a structure that resist vertical and lateral loads and are hence more vulnerable to failure during earthquakes [1]. Therefore, they must be retrofitted seismically, as has become clear after the collapse and damage of several RC structures during earthquakes. During an earthquake, good energy dissipation is facilitated by well-confined concrete resulting in structural safety. Structural fire is another hazard resulting in the deterioration of concrete strength [2] that can be retrofitted by FRP. Both normal strength concrete (NSC) and high strength concrete (HSC) exhibit different behaviour under fire conditions [3] and so an appropriate retrofit system should carefully be selected to meet strength requirements of such structures. This can be accomplished by selecting a suitable fibre reinforced polymer (FRP) confinement for structural concrete members susceptible to fire. External wrapping of concrete structures using FRP composites provides a significant amount of lateral confinement leading to an increased axial strength and energy absorption as reported in literature [4][5][6][7]. Researchers have used several parameters like unconfined concrete strength, shape of specimens, presence of internal reinforcement, types of FRP, fibre orientations of FRP wraps, and number of confining FRP layers [8][9][10][11][12][13]. The stress-strain response of FRP confined concrete specimens is illustrated in Figure 1, as indicated in various above mentioned studies. It can be seen that the stress-strain response presented by Rahai et al. [8] is significantly stiffer compared to others, which can be attributed to the Carbon Fibre Reinforced Polymer (CFRP) confined high strength concrete having strength of 45 MPa. Pessiki et al [11] used cylindrical specimens 610 mm in height and 152 mm in diameter having internal reinforcement and FRP confinement, with fibre oriented at 45° with regard to hoop direction, which increased the ductility of specimens. This behaviour closely matches that of square unreinforced specimens tested by Rochette and Labossiere [12], which were wrapped with confining carbon fibres oriented at 15° with respect to hoop direction. This indicates that the behaviour of FRP confined concrete depends on the presence of internal reinforcement as well as on the shape of the specimen. The confinement of circular concrete specimens 305 mm in height and 152.5 mm in diameter, tested by Parwin and Jamwal [10], shows quite a different response when compared to the results obtained by Mirmiran and Shahaway [5]. This is attributed to the number of layers used for confinement and their orientation, as well as to the type of confining material. It can be seen that with a smaller wrap thickness of the Glass Fibre Reinforced Polymer (GFRP), Parwin and Jamwal (2006) obtained stiffer stress-strain response, compared to the highly ductile response of 6, 10, and 14 plies of GFRP confinement used by Mirmiran and Shahaway The difference in the stress-strain response of FRP confined concrete can be attributed to the unconfined concrete strength, internal reinforcement, FRP (fibre) orientation of confinement, type of FRP used for confinement, and shape of the specimen. With such variation in parameters affecting behaviour of confined concrete, an accurate prediction of stress-strain response is quite difficult. Moreover, there is a need to validate the predictions by various design models and codes, as design is significantly affected by a number of variable parameters, as mentioned above. It is therefore desired to recognize and quantify a more effective and predictable confinement by studying the stress-strain response of two types of FRP systems, namely CFRP and GRRP.

Sample testing program and procedure
To study the comparative stress-strain response of GFRP and CFRP specimens, a comprehensive test program was conducted for 3 different strength concrete batches. The data obtained from tests was utilized to compare confinement characteristics provided by two types of FRPs. Moreover, the data was utilized to study and validate confinement predictions provided by different codes and guidelines.

Mix proportions and test specimens
A total of 27 cylinder specimens measuring 150 mm x 300 mm were cast in 3 batches of normal, medium, and high strength concrete. The mix proportions of these concrete batches are given in Table 1. The specimens were immersed in water and cured in water tank for 28 days. The wet lay-up method of FRP wrapping was employed. Out of 27 specimens, 9 were wrapped with carbon fibre reinforced polymer Sikawrap Hex 230C, and 9 specimens were wrapped with 2 wraps of glass fibre reinforced polymer Sikawrap Hex 106G, with the Sikadur 330 adhesive epoxy featuring a tensile strength GRAĐEVINAR 72 (2020) 6, 475-484 Comparison of GFRP and CFRP confinement of normal and high strength concrete of 33.8 MPa and an elongation of 1.2% at breaking. The strength and physical properties of these materials are given in Table 2. The overlapping was kept at 10 cm according to the requirements of the system so as to avoid localized bond failure at overlap.

Test apparatus and procedure
The universal compression testing machine with a loading capacity of 2000 kN was utilized for axial compressive testing of the specimens. The Micro-Measurements P3 Strain Indicator and Recorder manufactured by Vishay Precision Group was used for strain measurements. This system is capable of recording strains through linear strain gauges using 4 input channels. The Vishay Micro-Measurements C2A-Series 062LW strain gauges with the resistance of 350 ohms were utilized for strain measurements. The strain gauges were attached to the specimens in horizontal, vertical and diagonal (45°) arrangement in order to measure longitudinal and lateral strains. Dial-type strain gauges were also used in every third test to cross check measurement accuracy of linear strain gauges. The test equipment, assembly, and arrangement are shown in Figure 2.

Test results and discussions
Tests were conducted on unconfined and confined concrete specimens to measure the compressive confined strength, and ultimate load carrying capacity, and to compare longitudinal, hoop and volumetric strains. The volumetric strain (ε v ) can be measured by adding the longitudinal strain (ε l ) and twice-hoop strain (ε h ).
The comparison was carried out in terms of confined to unconfined strength, and gain in the axial load carrying capacity. Visual observations were also made during experiments and the results were analysed to determine behaviour of these specimens.

Axial compression tests
Axial compression tests were performed to determine the effect of FRP confinement for normal, medium and high strength concrete. Maximum strength, strain, and stiffness were measured through these tests on plain (control), GFRP, and CFRP confined concrete specimens. The axial compressive strength of control and FRP confined specimens is shown in Table 3.
The results show that concrete specimens confined with CFRP exhibit a higher increase in axial compressive strength as compared to GFRP confined specimens, as shown in Figure  3. However, the confined strength increase decreases with an increase in the unconfined concrete strength. This indicates that the strength of confined concrete depends on the type of confinement provided as well as on the original unconfined concrete strength. This effect of unconfined concrete on confinement has also been reported by Pessiki, Harries [11] in his study of FRP jacketed concrete.

Stress-strain response
The compressive tests conducted were also utilized on FRP confined concrete to study the stress-strain response. The stress-strain response of GFRP and CFRP confinement was also compared with previous studies. The comparison of stress-strain response of GFRP and CFRP confined specimens is presented in Figures 4, 5, and 6.   The volumetric strain response for the three concrete types is shown in Figures 7, 8, and 9. The volumetric stress-strain response indicates that the specimens undergo contraction in longitudinal direction under axial compressive loading and then, just before failure, the specimens start expanding laterally exhibiting higher strain in lateral direction. The volumetric stress-strain response of GFRP confined specimens follows the pattern similar to that of unconfined specimens, where the specimens undergo contraction when subjected to axial compressive loading. GFRP confined specimens absorb energy up to their capacity and then start expanding in lateral direction but, due to confinement, the specimens absorb more energy compared to unconfined specimens. This can be explained by the fact that the confinement resists lateral expansion, which enables confined specimens to withstand higher loads before expanding. GRAĐEVINAR 72 (2020) 6, 475-484 Comparison of GFRP and CFRP confinement of normal and high strength concrete  The CFRP confined concrete displays enhanced capacity to withstand load without much change in volume. The specimens resist both contraction and expansion as compared to GFRP confinement. The CFRP wrap takes part in absorbing energy along with concrete, resulting in more ductile behaviour as seen in 30 and 42 MPa concrete specimens, where the specimens fail at higher lateral strains. The behaviour of the 64 MPa concrete is more brittle owing to its higher strength as well as increased stiffness due to FRP. Therefore, FRPs show only an increased strength in HSC.
The test data is also compared to several previous works of this kind. Rahai and Sadeghian [8] used CFRP wraps measuring 0.9, 1.8, 2.7, and 3.6 mm in thickness and Bisby and Take [14] used CFRP wraps on 100 mm by 200 mm concrete cylinders to plot stress-strain response of confined concrete. Similarly Pessiki and Harries [11] used FRP wraps having different fibre strengths and densities on concrete specimens measuring 610 mm in height and 152 mm in diameter. The stress-strain response follows similar pattern as in the above studies. However, the values of corresponding stresses and strains are different, which is attributed to the difference in specimen size, fibre strength, and density of confining FRP wraps.

Stress and stiffness response of FRP confined concrete
The stress in the plain and wrapped concrete cylinder specimens was calculated by dividing the load applied with the original concrete cross sectional area. While calculating stress, the thickness of the wrap was not taken into account, as it is negligible as compared to the dimensions of the specimens. The stiffness was evaluated using the stress-strain response by calculating the slopes m 1 and m 2 at the bifurcation points as proposed by Samaan and Mirmiran [15]. The initial slope of the curve is m 1 and the slope after the bifurcation point is m 2 , as shown in Figure 10. The enhancement of stiffness is shown in Figure 11 where the ratio of stiffness of confined specimens (m' 1 ) to the stiffness of unconfined specimens (m 1 ) is plotted against the unconfined strength of specimens both in the axial and hoop directions.
With an increase in unconfined strength of concrete, the stiffness (m' 1 ) in longitudinal direction increases from 1.02 to 1.11 times in case of GFRP, and from 1.03 to 1.16 times in case of CFRP, as shown in Figure 11. The stiffness also increases in the hoop direction, but the increase in stiffness decreases as the unconfined strength of concrete increases. Only the 64 MPa concrete specimens show stiffer behaviour as compared to previous researches because of the difference in mix proportion GRAĐEVINAR 72 (2020) 6, 475-484 Muhammad Rameez Sohail, Naveed Ahmad, Muhammad Usman Rashid, Rana Waqas, Usman Muhammad and test conditions whereas, on GFRP and CFRP confined concrete, the 30 MPa and 42 MPa specimens exhibit behaviour similar to the previously published one [8,9].

Strain response
The strain at specimen failure was recorded using strain gauges attached both in the vertical and horizontal directions and, hence, both the axial and hoop strain values were recorded.
Maximum strain values are tabulated in Table 4.

Table 4. Average longitudinal and lateral strain at failure
The ratio of failure strain of confined specimens (ε' cc ) to that of unconfined specimens (ε c ) is illustrated in Figure 12. It can be seen that hoop strains in unconfined plain concrete specimens are much lower compared to axial strains. The ratio between the lateral and longitudinal strains varies from 0.15 to 0.20 as the Poisson's ratio for concrete lies in this range. The increase in hoop strain of wrapped specimens was noted to be much higher than the axial strain increase because of the enhancement in ductility provided by FRP wrap confinement along the hoop direction. The CFRP confined specimens exhibit an increase of 5 times the strain of non-wrapped specimens, while GFRP wrapped specimens show an increase of up to 2 times the strain of unconfined specimens, which means that the CFRP confinement enhances the strain capacity in concrete much more as compared to GFRP confinement.

Ductility response
Ductility of a material is its ability to absorb energy. Ductile materials allow better stress distribution and give warning with regard to impending failure. In the case of FRP wrapped concrete specimens, specimen ductility is expressed in terms of deformability, which is defined as the ratio of energy absorption (or area under load-deflection curve) at ultimate strength to energy absorption at limiting curvature, as shown in the Figure 13. In this study, deformability was calculated by finding the total energy under the curve up to failure and by calculating the ratio between the total energy and the energy at limiting strains of 0.0008, 0.002, and 0.003, as tabulated in Tables 5 and 6 for longitudinal and hoop directions, respectively. Comparison of GFRP and CFRP confinement of normal and high strength concrete Energy absorbed by confined specimens (Δ cc ) are compared in Figure 14 by considering the ratio of total energy absorbed by confined specimens to the total energy absorbed by unconfined specimens Δ c . It can be seen in Figure 13 that energy absorbed by GFRP confined specimens is 1.33 to 1.99 times the energy absorbed by control specimens in longitudinal direction, whereas the energy absorbed in the hoop direction is 1.05 to 1.3 times the energy absorbed by unconfined concrete specimens. CFRP confined specimens show 1.34 to 2.87 times the energy absorbed by unconfined concrete specimens, while in the hoop direction this increment in energy absorption is 1.07 to 2.4 times. Moreover, as reported by Rahai and Sadeghian [8], the increment in energy absorption decreases as the unconfined strength of specimens increases in the longitudinal as well as in the hoop directions for both types of FRP confinements provided. It shows that the effectiveness of FRP confinement is more pronounced at lower strengths of concrete, and that it becomes less effective with an increase in concrete strength.

Strength predictions of FRP confined concrete based on various codes and standards
Various design codes and guidelines have been developed for the prediction of confinement pressure, confined compressive strength, axial load carrying capacity, and other parameters. The most prominent design guidelines are those provided by American Concrete Institute ACI, Canadian Standards Association CSA, Canadian Intelligent Sensing for Innovative Structures ISIS, and European CEB/FIP Model Code that uses the guidelines provided in Technical Report by the Fédération Internationale du Béton, fib Bulletin 14 [16][17][18][19]. Experimental results obtained in the test program are compared and correlated with theoretical values predicted in these FRP design guidelines and codes. Empirical formulae are shown in Table 7.

Comparison of theoretical guidelines with experimental results
Compressive strength values obtained through experimentation and theoretical guidelines are shown in Table 8. It can be noted that fib exact guidelines give the closest predictions for concrete confined with both FRP types. However, these guidelines overestimate the strength of CFRP confined specimens as the unconfined concrete strength increases from the domain of normal strength to high strength. The comparison of theoretical guidelines with experimental results is illustrated in Figures 15 and 16.

Failure modes and patterns
The surface of wrapped cylinder specimens was carefully observed after the achievement of failure loads. Failure modes of specimens wrapped with GFRP and CFRP are shown in Figure 19. All CFRP-wrapped specimens failed by the FRP wrap rupture due to hoop tension. Similar case was observed for GFRP wrapped specimens. The collapse occurred almost without any advance warning by sudden rupture of the composite wrap. For all confined specimens, delamination was not observed at the overlap location of the jacket.

Conclusions
Based on the research presented in this paper, the following conclusions can be made: -The increase of longitudinal (axial) strain in the CFRP confined concrete is by 1.99 to 2.22 times higher compared to the GFRP confined concrete, whereas the hoop strain in the CFRP confined concrete is by 1.08 to 2.04 times higher than that of the GFRP confined concrete in case of 30 to 64 MPa unconfined concrete strengths. Both FRP confinements enhance the hoop strain far more than longitudinal strain, but this enhancement effect decreases with an increase in the unconfined strength of concrete.
-Stiffness of the FRP confined concrete is also affected by unconfined strength of concrete as it decreases with an increase in unconfined concrete strength and vice versa.
-The increase in total energy absorption is more pronounced in the CFRP confined concrete as compared to GFRP confined concrete; however, this increment in total energy absorption decreases with an increase in the unconfined strength of concrete. This is attributed to the safety factor of 0.85 used by CSA-S806-02 for reduction of the unconfined concrete strength, which leads to underestimation of the effectiveness of CFRP.
-For GFRP confinement, CSA-S806-02 and fib approximate guidelines give lower strength of confined specimens compared to unconfined specimens.