Effect of corrosion-damaged RC circular columns enveloped with hybrid and non-hybrid FRP under eccentric loading

Abstract

The paper presents the effect of corrosion-damaged reinforced concrete (RC) columns with circular cross section enveloped (i.e. wrapped) with externally bonded hybrid and non-hybrid fibre-reinforced polymer (FRP) reinforcement. This study consisted of 18 RC circular columns with a diameter and height of 130 mm and 780 mm, respectively. These specimens were divided into three divisions, namely (i) no corrosion, (ii) mild corrosion and (iii) severe corrosion activities. All the mild and severe corroded columns, except control column, were exposed to corrosion activity, repaired with FRP and tested under eccentric loading. However, the columns without any corrosion activity were enveloped with FRP reinforcement and tested under eccentric loading. The corrosion process in RC columns was accelerated by impressing direct current on the steel reinforcement cage, adding 5% of NaCl in concrete mixture and wet/dry cycles. The variables investigated were different period of corrosion levels, plies of FRP (carbon and glass) reinforcement, hybrid (i.e. combination of carbon and glass FRP plies) and non-hybrid FRP (i.e. carbon or glass) reinforcement. From the experimental results, it was found that the external FRP confinement significantly increased the load-carrying capacity of the corrosion-damaged RC circular columns by 8% to 36% for mild corrosion level and 21% to 34% for severe corrosion level over the unconfined corrosion-damaged RC circular columns. Moreover, the results show that the performance of corrosion-damaged RC circular columns enveloped with hybrid FRP was better over the non-hybrid CFRP and GFRP RC circular columns. However, the results of CFRP enveloped circular columns were close to the hybrid FRP strengthened circular columns.

Keywords

Carbon FRP glass FRP corrosion column

Introduction

Reinforced concrete structures are generally exposed to environmental condition that leads to corrosion, which could seriously affect the steel rebars. Moreover, the corrosion of steel rebar affects the serviceability and durability of the reinforced concrete structures. The cause of initiation of corrosion on reinforcement steel bar is due to the ingress of chloride ions (Cl⁻) to the surface of the steel rebars even in a highly alkaline concrete environment. This ingress of chloride ions could destroy the passive film of oxide on the steel rebar and it could be a resource for active reinforcement corrosion.^1,2 The two main external sources for diffusion of chloride into reinforced concrete structures are saltwater in marine environment and de-icing of salts in cold environment. Moreover, other sources of chloride that attack the RC structures are airborne and groundwater salts.

Over three decades, fibre-reinforced polymer (FRP) has received attention from the civil engineering as an alternative material to repair, upgrade and rehabilitation of damaged structures. A number of experimental investigations have proved that the application of FRP system as an external reinforcement could enhance the structural performance of RC members.^2–12 Pantazopoulou et al.³ tested the corrosion-damaged circular columns under axial loading to failure. Different types of repairing schemes were studied which include the jacketing with glass fibre wraps, in combination with grouting the voids between the jacket and the original surface of specimen with either conventional or expansive grouts. The key finding was the external confinement in the form of jacketing could slow down the rate of corrosion reaction, and impart ductility and strength of the confined columns. Moreover, the investigation explored the performance and efficiency of FRP jacketing as an alternative to conventional repair methods for corrosion-damaged concrete columns. In 2003, Li and Hadi⁴ investigated the performance of high-strength concrete circular columns confined externally with carbon and E-glass FRP reinforcement and tested under concentric and eccentric loading. It was concluded that the external confinement significantly increased the strength of columns under concentric loading. However, for eccentric loading, the strength of confined columns was increased to some extent. This was due to the fact that eccentric loading was induced by both axial and bending actions. Subsequently, Hadi and Li⁵ have also investigated the performance of externally reinforced high strength-concrete circular columns under eccentric loading. Results have shown that the confinement of the concrete could prevent the concrete from expanding, and therefore the FRP confinement allowed the concrete to absorb more energy, and as a result, the load-carrying capacity has been significantly increased. El-Maaddawy⁶ studied the effect of the corrosion damaged square RC columns repaired with CFRP reinforcement under eccentric loading. The test variables were CFRP repair scheme and eccentricity to section height ratio. The key results were (i) the gain in strength of repaired RC columns was less pronounced at higher eccentricity ratio and (ii) the partial CFRP wrapping was less effective than the full CFRP wrapping scheme.

Sadeghian et al.⁷ studied the performance of rectangular RC columns confined with CFRP reinforcement under the combination of axial load and bending moment. The tested variables were different plies of CFRP, orientation of CFRP and two different eccentricities. The moment curvature behaviour results showed that the longitudinal layers improved the bending stiffness and moment capacities of the CFRP-confined RC columns. However, the behaviour of these columns with orientation of FRP had a little effect. From the literature review, it is very apparent that limited researchers have studied the behaviour of corrosion-damaged RC columns repaired with FRP under eccentric loading. Moreover, all researchers have used carbon and glass FRP reinforcement (i.e. non-hybrid FRP). However, the effect of hybrid FRP reinforcement (i.e. combination of CFRP and GFRP reinforcement) has not received much attention among the researchers. Therefore, the main objective of this experimental investigation was to study the effect of corrosion-damaged RC circular columns enveloped with externally bonded hybrid and non-hybrid FRP reinforcement.

Experimental study

Description of RC circular columns

A series of 18 reinforced concrete circular columns of dimension 130 mm in diameter and 780 mm in height were cast with a concrete grade of 25 MPa at the Structural Engineering Laboratory, Universiti Sains Malaysia. These columns were reinforced with a longitudinal steel reinforcement of four rebars with 10 mm diameter, and the transverse steel stirrups at mid-section of columns were reinforced with 6 mm diameter at 180 mm centre to centre. However, to prevent any local failure at the ends of column, the steel stirrups were closely placed at 100 mm centre to centre. All these columns were provided with a concrete cover of 20 mm. The reduction in concrete cover was mainly to simulate the old RC structures and also to trigger the accelerated corrosion activity. A smooth stainless steel bar with a diameter of 6 mm was placed at the centre of cross section of the column (i.e. parallel to the longitudinal steel rebars) to serve as the cathode in the accelerated corrosion process. Figure 1 shows the dimension and internal steel reinforcement details of RC circular columns.

Preparation of RC circular columns

The formwork consisted of two components, namely PVC pipes and a triangular-shaped wooden platform. A series of six PVC pipes were placed in a triangular-shaped wooden platform. The supporting wooden platform was drilled with five holes for steel reinforcement cage and stainless steel rebar for each PVC pipe. The prepared reinforcement cage was placed in the formwork and all the rebars were protruded out of the wooden platform to provide ease of access points for accelerated corrosion activity. Moreover, the protruded steel rebar could support the reinforcement cage within the formwork and to ensure the reinforcement in the vertical direction without any dislocation. As shown in Figure 1, type 304 stainless steel rebar with a diameter of 6 mm was placed at the centre of the column to act as a cathode during the accelerated corrosion activity. To facilitate the connection of steel and stainless steel rebars with external power supply system, the longitudinal steel and the stainless steel rebars were extended about 100 mm at one end of the RC circular columns. All the PVC pipes were secured with both the supporting wooden platform and a series of steel hoop straps. Figure 2 portrays the arrangement of PVC pipes with wooden formwork.

Ordinary Portland cement, sand and coarse aggregate of nominal size of 10 mm were used to prepare a concrete grade of 25 MPa. Prior to adding the mix water, 5% of sodium chloride (NaCl) crystals by weight of cement were dissolved into the mix water to induce the corrosive environment for reinforcing steel rebars.^6,9–11 The added NaCl crystals were thoroughly mixed in the mix water for five minutes. A total of three batches of concrete mix were used to cast all the 18 reinforced concrete circular columns. After casting, these columns were cured for 28 days at the control room.

Accelerated corrosion aging

Different techniques are generally used to accelerate corrosion aging process such as wet/dry cycling, freeze/thaw cycling, applying electrical current and chemical soaking.¹² However, most of investigations have used impressed electric current technique with addition of NaCl in preparation of RC members.^{3,8–11,13–17} From the existing investigations, it was found that the application of impressed electric current with addition of chloride in concrete mix is one of the most effective methods for accelerated corrosion aging activity.

During the preparation of concrete mix, 5% of NaCl crystals by weight of cement were dissolved into the mix water to aggravate the corrosive environment for reinforcing steel rebars. Moreover, the process of corrosion activity on steel rebar was accelerated by applying a current density of 5.09 mA/mm². According to Faraday’s law, it was expected that the accelerated corrosion activity could significantly amplify the age of RC columns as it was exposed to different periods such as three (200 amp-hours) and six (400 amp-hours) weeks for mild and severe corrosion levels, respectively.

The power supply system with a maximum output values of 30 V and 300 mA was used to provide a constant electric current to all RC circular columns. This equipment is designed with dual outputs; therefore, the current and voltage can be easily adjusted. Prior to the application of current, the surface of steel rebars was grinded to provide good surface contact. Subsequently, to prevent any corrosion effect at the ends of columns, the steel rebars were wrapped with teflon tape. As the dual output power supply system was used in this investigation, three columns were wired to each output of power supply as shown in Figure 3. A special chamber was prepared to expose the columns for wet and dry cycles (see Figure 4). The wet condition with constant moisture level was maintained by covering the chamber with plastic sheet. The reason is because the concrete between the longitudinal steel and stainless steel would remain moist to allow for electrical conductivity. This would allow the access of oxygen needed for the expansive corrosion reactions to occur as desired. Moreover, the wet and dry cycles were carried out for four days per week by closing the water source and removing the plastic cover.

During the corrosion process, the amount of steel loss can be measured to a certain degree of accuracy by means of the measurement of electrical current generated by the anodic reaction and consumed by the cathodic reaction. Faraday’s law has a direct relationship between the measured corrosion current and the mass of steel consumed. One Faraday (F) is equal to 96,486.7 Coulombs of charge transfer to oxidise or reduce 1 g equivalent weight of material involved in the electrochemical reaction. The gram equivalent weight of iron can be calculated by dividing of gram atomic weight to number of transferred electrons (z).¹⁸ Using Faraday’s law, the relationship between the corrosion current density and the weight of steel mass loss can be determined as follows:

Δ ω = \frac{t \cdot I \cdot M}{z \cdot F}

(1)

where t is the time (s), I is the current (A) and M is the atomic weight of iron (55.847 g/mol).

Repair of RC circular columns using FRP reinforcement

The unidirectional glass and carbon FRP reinforcement was typically applied to RC circular columns using manual wet lay-up technique. Two-component epoxy was used to apply FRP reinforcement on the RC circular columns.¹⁹ The FRP sheets were provided with a bond length of 50 mm. The wrapping schemes used for corrosion-damaged RC circular columns are depicted in Table 1. The wrapping schemes I and II consisted of one and two plies of CFRP reinforcement, respectively. Similarly, the wrapping schemes III and IV comprised of one and two plies of GFRP reinforcement, respectively. Eventually, the wrapping scheme V consisted of hybrid FRP confinement which is a combination of single ply of CFRP and GFRP reinforcement. The cross-sectional area of FRP reinforcement used for one and two plies of FRP reinforcement was of 450 mm × 780 mm (i.e. area = 351,000 mm²) and 867 mm × 780 mm (i.e. area = 676,260 mm²), respectively. To determine the tensile strength of GFRP and CFRP reinforcement, the tensile test was conducted on FRP coupons using test method described in ASTM D3039.²⁰ The observed tensile strength values were 4470 MPa and 3220 MPa for CFRP and GFRP reinforcement, respectively. However, the measured tensile strength of hybrid FRP was 5200 MPa. Table 2 presents the details of external FRP reinforcement and degree of corrosion of RC circular columns.

Table 1.

FRP wrapping schemes used for corrosion-damaged RC circular columns.

	Non-hybrid FRP				Hybrid FRP
Wrapping scheme	I	II	III	IV	V
Number of FRP plies	1	2	1	2	2
Nominal thickness of FRP (mm)	0.065 (one ply)	0.130 (two plies)	0.114 (one ply)	0.228 (two plies)	0.179 (one ply of carbon + one ply of glass)
FRP material	Carbon		Glass		(carbon + glass)

Table 2.

Details of external FRP reinforcement and degree of corrosion of RC circular circular columns.

Specimen	FRP	Number of layers	Thickness of FRP (mm)	Corrosion level
Ctrl	–	–		No corrosion
Ctrl-1-C	Non-hybrid CFRP	1	0.065
Ctrl-2-C	Non-hybrid CFRP	2	0.130
Ctrl-1-G	Non-hybrid GFRP	1	0.114
Ctrl-2-G	Non-hybrid GFRP	2	0.228
Ctrl-H	Hybrid FRP	2	0.179
M-Ctrl	–	–		Mild corrosion (3 weeks)
M-1-C	Non-hybrid CFRP	1	0.065
M-2-C	Non-hybrid CFRP	2	0.130
M-1-G	Non-hybrid GFRP	1	0.114
M-2-G	Non-hybrid GFRP	2	0.228
M-H	Hybrid FRP	2	0.179
S-Ctrl	–	–		Severe corrosion (6 weeks)
S-1-C	Non-hybrid CFRP	1	0.065
S-2-C	Non-hybrid CFRP	2	0.130
S-1-G	Non-hybrid GFRP	1	0.114
S-2-G	Non-hybrid GFRP	2	0.228
S-H	Hybrid FRP	2	0.179

Test methodology

Prior to the application of eccentric loading, all extended steel rebar were cut off and the top and bottom surfaces of columns were grinded flat without any irregularities. Each column was tested under 50 mm eccentric loading. To apply eccentric loading, a special set of steel cap with swivel steel plates was fabricated. As shown in Figure 5, this set consisted of two parts, namely top swivel plates and a steel cap. Top swivel plates composed of two same-sized steel plates with central groove and a smooth steel rod. The steel caps were fabricated with top steel plate, steel shells, stiffeners and bolts. The specially fabricated steel shoe was properly affixed at the top and bottom of RC circular columns. A load cell with a maximum capacity of 500 kN was placed at the top of RC column to measure the applied load. A series of six linear variable displacement transducers (LVDT) were mounted to measure the lateral and axial displacements. In order to measure the longitudinal and transverse strains in FRP reinforcement, two electrical strain gauges of size 10 mm were affixed at the mid-height of RC circular columns. Moreover, these strain gauges were placed on the tension face of the eccentrically loaded RC columns. The eccentric loading was applied at a constant load rate of 0.5 kN/min to failure. The load cell, strain gauges and LVDTs were connected to data logger to record all measurements. Figure 5 shows the experimental test set-up of circular RC columns under eccentric loading.

Results and discussion

Discussion on accelerated corrosion aging

The unconfined RC column specimens M-Ctrl and S-Ctrl were exposed to mild (i.e. 3 weeks) and severe (i.e. 6 weeks) corrosion levels, whereas the columns M-1-C, M-2-C, M-1-G, M-2-G, S-1-C, S-2-C, S-1-G and S-2-G were exposed to corrosion activity and repaired using non-hybrid carbon and glass FRP reinforcement, and the columns M-H and S-H were repaired with hybrid FRP reinforcement.

In the early period of corrosion activity (i.e. at seven days (67 amp-hours)), a number of hairline cracks were observed due to the ingress of chloride ions in concrete. Subsequently, at 9-days (86 amp-hours) period of corrosion activity, the corrosion products began to ooze out and strain on concrete in the localised areas followed by secreting corrosion products on the surface of column (see Figure 6). As the period of corrosion increased, the early developed cracks, at the end of columns, were then propagated in inward direction and the width of cracks was also significantly increased.

For mild corrosion level, the cracks in RC columns (e.g. M-Ctrl and M-1-C) emerged in the longitudinal direction of steel rebar (i.e. at least for two sides of steel rebar), whereas for severe corrosion level, the longitudinal cracks in these columns (e.g. S-Ctrl and S-1-C) were also observed along the direction of longitudinal steel rebar. Similarly, the cracks were also observed along the shear stirrups due to the corrosion of steel stirrups.

After the completion of eccentric loading, the longitudinal steel rebar in RC columns (i.e. M-Ctrl and S-Ctrl) were removed and cleaned using a solution of hydrochloric acid and hexameythylene tetramine, as specified in ASTM G-1-03.²¹ It was found that the rebar mass loss was of 7% for mild corrosion level which is equivalent to a cross-sectional loss of about 5.5 mm², and 16.5% of mass loss for severe corrosion level which is equivalent to a cross-sectional loss of about 12.96 mm². Figure 7 shows longitudinal steel rebar extracted from RC columns M-Ctrl and S-Ctrl. It can be seen that the steel rebar, at 7% of mass loss, was almost in pristine state (i.e. ribs in steel rebars were intact) except some small localised pits. However, for severe corrosion level of 16.5% steel rebar mass loss, significant pitting was observed on the steel rebar and some were moderately deep. Moreover, the measured corrosion levels from the experimental investigation were relatively close to the level of simulated corrosion aging by Faraday’s law.

Crack width measurements

During the process of accelerated corrosion activity, the crack width of RC columns was measured at least one time in a week. Table 3 shows the summary of crack width measurements (i.e. before application of FRP reinforcement) for all RC columns exposed to corrosion activity at different periods of corrosion level. As this investigation has been focussed on the effect of FRP reinforcement on corrosion-damaged RC columns, the crack patterns and length of crack was not paid much attention; however, the crack width at different periods of corrosion level was addressed.

Table 3.

Summary of crack width measurements (i.e. before application of FRP reinforcement) for RC circular columns exposed to corrosion activity at different periods.

Specimen	Maximum Crack Width (mm)
Specimen	After one week	After two weeks	After three weeks	After four weeks	After five weeks	After six weeks
M-Ctrl	0.1	0.4	0.6	–	–	–
M-1-C	0.2	0.55	0.7	–	–	–
M-2-C	0.15	0.4	0.7	–	–	–
M-1-G	0.1	0.6	0.7	–	–	–
M-2-G	0.1	0.4	0.65	–	–	–
M-H	0.1	0.4	0.8	–	–	–
S-Ctrl	0.1	0.45	0.75	2.0	3.0	4.5
S-1-C	0.1	0.5	0.85	1.5	2.5	3.0
S-2-C	0.05	0.3	0.6	1.0	1.5	3.0
S-1-G	0.2	0.6	0.8	1.2	2.5	4.0
S-2-G	0.1	0.35	0.6	1.5	3.0	4.0
S-H	0.1	0.45	0.8	1.5	3.5	4.0

The maximum observed crack width in the mild corroded RC columns (i.e. after three weeks of corrosion activity) was varied between 0.6 mm and 0.85 mm. However, the maximum crack width in the severe-corroded RC columns after six weeks ranged from 3.0 mm to 4.0 mm. Moreover, as the period or corrosion increased, it was also found that the rate of crack width at three and six weeks (202–403 amp-hours) was noticeably greater than one and three weeks (67–202 amp-hours) by increasing from 1.0 mm to 4.5 mm and from 0.05 mm to 0.2 mm, respectively. The observed maximum average crack width in mild and severe-corroded RC columns was 0.71 mm and 4.5 mm, respectively. A trend of the crack width for all RC columns is presented in Figure 8. These results show that the width of crack increased as the period of corrosion activity increased. Analogous to crack width, it was observed that the mass loss in steel rebar was increased when the corrosion level increased. Figure 9 shows the cracking patterns of corrosion-damaged RC columns.

Failure modes and overall behaviour

All the unconfined and FRP-confined corrosion-damaged RC circular columns were tested under eccentric loading (e = 50 mm). In the early stages of loading, the flexural cracks were initially emerged on the tension face of the corroded and uncorroded–unconfined RC circular columns. As the applied load increased, the width of flexural cracks significantly increased and followed by spalling of concrete on the compression face of the unconfined RC column. In the case of severe-corroded RC circular columns (see Figure 10), the observed spalling of concrete was relatively more as compared to the mild corroded and uncorroded–unconfined RC circular columns due to the effect of exposure level of corrosion activity. Eventually, all these corroded and uncorroded–unconfined RC circular columns failed in buckling with crushing of concrete. However, the longitudinal steel rebar in uncorroded RC columns was not exposed or buckled as observed in corroded RC columns. The observed failure patterns of all unconfined RC columns occurred close to the mid-height of the column. To prevent any local failure, the unconfined columns were wrapped with FRP reinforcement at the ends.

In the case of FRP confined corrosion-damaged RC columns, the crackling sound with snapping of FRP reinforcement was observed in the early stages of loading. This was mainly due to the effect of transferring the stress from concrete to FRP reinforcement. However, at proximity to the failure, the FRP reinforcement was fractured (i.e. without any FRP rupture) at different locations along the tension face of the non-hybrid GFRP and CFRP, and hybrid FRP-reinforced RC corrosion-damaged RC columns. Moreover, these columns attained a failure of buckling with fracture of FRP reinforcement. Figure 11 shows the buckling with FRP fracture failure for non-hybrid GFRP confined corrosion-damaged RC column (M-1-G). No debonding of FRP reinforcement was observed in any of hybrid FRP and non-hybrid CFRP and GFRP confined corrosion-damaged RC circular columns.

Uncorroded FRP confined columns

Figure 12 shows the comparison of failure loads for uncorroded and corroded FRP-confined RC circular columns. From Figure 12, it is evident that FRP wrapping system enhanced the load-carrying capacity of uncorroded FRP confined RC circular columns under eccentric loading. The load-carrying capacity of uncorroded RC columns confined with one ply of CFRP (Ctrl-1-C), two plies of CFRP (Ctrl-2-C), one ply of GFRP (Ctrl-1-G) and two plies of GFRP (Ctrl-2-G), and hybrid FRP (Ctrl-H) reinforcement was 20%, 28%, 7%, 18% and 38% higher than the unconfined column (Ctrl), respectively. This experimental test results show that the percentage gain in strength of non-hybrid CFRP-confined columns (Ctrl-1-C and Ctrl-2-C) was greater than the non-hybrid GFRP-confined columns (Ctrl-1-G and Ctrl-2-G) by 12% and 8% for one and two plies of FRP reinforcement, respectively. The specimen with hybrid FRP reinforcement (Ctrl-H) attained a gain of 8% and 17% over the non-hybrid confined RC circular columns with two plies of CFRP and GFRP reinforcement (Ctrl-2-C and Ctrl-2-G). Thus, it was found that the performance of hybrid FRP confined circular columns (i.e. 1 ply CFRP + 1 ply GFRP) was relatively better than the columns confined with two plies of non-hybrid CFRP and GFRP reinforcement. However, the performance of hybrid FRP was close to two plies of CFRP reinforcement. It was also proved that the load-carrying capacity of non-hybrid CFRP and GFRP-confined RC circular columns was increased as the number of FRP plies increased. Figure 13 shows the buckling with FRP fracture failure pattern of uncorroded FRP-confined RC circular columns.

Mild and severe corroded FRP-confined columns

The ultimate load-carrying capacity of mild corrosion-damaged unconfined RC column (M-Ctrl) was approximately of 15% less than the unconfined RC column (Ctrl) due to the percentage mass loss in steel rebar of 7% for corrosion exposed RC column. The obtained experimental results have shown that the FRP confinement could enhance the load-carrying capacity of mild corroded FRP confined columns M-1-C (1 ply of CFRP), M-2-C (2 plies of CFRP), M-1-G (1 ply of GFRP), M-1-G (2 plies of GFRP) and M-H (1 ply of CFRP + 1 ply of GFRP) by 17%, 33%, 8%, 27% and 36% over the corrosion-damaged unconfined RC column (M-Ctrl). Moreover, the results show that increasing the thickness of non-hybrid CFRP and GFRP reinforcement from one ply (M-1-C & M-1-G) to two plies (M-2-C & M-2-G) significantly increased the flexural capacity of the confined RC circular columns up to 16% and 13%, respectively. Result shows that the performance of column with hybrid FRP reinforcement (M-H) was approximately 3% and 8% greater than the non-hybrid CFRP (M-2-C) and non-hybrid GFRP (M-2-G) specimens, respectively.

Similarly, in the case of severe corroded unconfined RC column (S-Ctrl), the load-carrying capacity was 57% less than the uncorroded RC column (Ctrl). This drastic drop in failure load was due to the increased mass loss in steel rebar by 16.5%. For severe-corroded non-hybrid CFRP and GFRP and hybrid FRP-confined RC circular columns, the percentage gain in flexural capacity was varied between 21% and 34% over the severe corroded unconfined RC column. These results indicate that the performance of severe corroded RC circular columns repaired with FRP reinforcement was significantly increased similar to that of mild corroded FRP-confined RC circular columns. As the thickness of CFRP and GFRP reinforcement increased from one to two plies, the specimens (S-2-C and S-2-G) with two plies of CFRP and GFRP reinforcement attained a gain of approximately 7.6% and 7.4% over the specimens (S-1-C and S-1-G) with one ply of CFRP and GFRP reinforcement, respectively. From this result, it is very difficult to conclude the effect of confinement on severe corroded RC circular columns with more number of layers of GFRP reinforcement. Therefore, more experimental investigations are required on the effect of GFRP confinement on severe corroded RC circular columns. Furthermore, the experimental test results confirm that the performance of specimen with hybrid FRP confinement (S-H) was 3% and 2% over the non-hybrid CFRP and GFRP RC circular columns (S-2-C and S-2-G), respectively. Figures 14 and 15 show the buckling with FRP fracture failure of mild and severe corroded FRP-confined RC circular columns, respectively.

Load–displacement behaviour

Various sets of load against lateral and longitudinal displacement profiles of uncorroded–unconfined and FRP-confined, and corroded–unconfined and FRP-confined column specimens are shown in Figures 16(a) through (d). From the Figures, it can be generally seen that before attaining the maximum peak load, the rate of displacement was steadily increased as the applied increased at constant rate. This was probably due to the effect of the FRP confinement on FRP confined RC circular columns had not been fully achieved at this stage. After the peak load, the displacement began to increase at a faster rate as the application of load increased. This indicates the FRP confinement activated at the later stages of loading. Moreover, there was a sudden drop at peak load for uncorroded–unconfined, mild and severe corroded unconfined RC circular columns (Ctrl, M-Ctrl and S-Ctrl). However, the ductility of uncorroded FRP confined, and mild and severe corroded FRP confined RC circular columns were significantly increased due to the effect of FRP confinement.

From Figure 16(a), it was obvious that the stiffness of lateral and longitudinal displacement curves for uncorroded–unconfined RC circular columns was relatively higher as compared to mild and severe corroded unconfined RC circular columns. The stiffness of RC circular columns decreased as exposed to different levels of corrosion activity. However, no significant difference in stiffness was observed between the uncorroded–unconfined and mild corroded RC circular columns before attaining the peak load. Moreover, it was observed that the ultimate failure load was decreased as level of corrosion increased.

The displacement profiles of uncorroded–unconfined and FRP-confined RC circular columns are shown in Figure 16(b). The lateral displacement profile of FRP-confined RC circular columns indicates that the specimen with two plies of CFRP reinforcement (Ctrl-2-C) was relatively stiffer than the specimen with two plies of GFRP reinforcement (Ctrl-2-G). It was also observed that the stiffness of longitudinal displacement profile of hybrid FRP confined column (Ctrl-H) was less than the CFRP and GFRP-confined RC circular columns. However, in case of lateral displacement profile, the stiffness of specimen with hybrid FRP reinforcement was close to that of specimen (Ctrl-2-C) with two plies of CFRP reinforcement.

From Figure 16(c) it was observed that the stiffness of mild corroded hybrid-FRP confined specimen (M-H) was slightly greater than the specimens with two plies of non-hybrid CFRP (M-2-C) and GFRP (M-2-G) reinforcement. It was generally observed that the stiffness of displacement curves of mild corroded FRP confined RC circular columns was marginally varied with respect to type of FRP, number of FRP layers and hybrid FRP reinforcement. However, for severe corroded FRP-confined RC circular columns (see Figure 16(d)), no significant difference in the stiffness of displacement curve was observed with different test variables due to the effect of increase in percentage of corrosion level. Moreover, the ductility of mild and severe corroded non-hybrid and hybrid FRP confined RC columns was relatively greater than corrosion-damaged unconfined RC columns.

Load–strain behaviour

The axial and hoop strain results for uncorroded–unconfined and FRP-confined, and corrosion-damaged unconfined and FRP-confined RC circular columns are shown in Figures 17(a) through (c). The strain values of some FRP confined specimens did not record any strain measurement due to the malfunction or damages in strain gauges during the application of load. It was generally observed that the FRP hoop strain values in all uncorroded FRP-confined, mild and severe corroded FRP-confined RC circular columns were relatively low as compared to the longitudinal FRP strain values. Prior to the peak load, the longitudinal strain in FRP confined RC circular columns was steadily increased as the applied load increased, whereas the longitudinal FRP strain values were significantly increased after the peak load was reached. This behaviour shows that a significant enhancement was achieved in terms of strength and ductility of FRP confined RC circular columns. Results also show that the longitudinal GFRP strain in uncorroded FRP-confined, and mild and severe corroded FRP-confined RC circular columns were greater than the strain in CFRP and hybrid FRP reinforcement. Moreover, it was found that the strain values at failure in severe corroded FRP confined columns were relatively less as compared to the mild corroded FRP confined columns.

Moment–curvature behaviour

The moment (M) at mid-height of RC columns is reported in Table 4. It consisted of primary and secondary moments. The primary moment is computed using the initial (nominal) eccentricity, and the secondary moment is calculated by the mid-height lateral displacement at failure (U_Lat)⁷. The computed total moment capacities were varied between 6 kNm and 11 kNm. The actual value of e/h at failure can be computed by dividing the total moment by the failure load. It should be noted that the designed e/h for this study was 0.1.

Table 4.

Results of primary and secondary moments of uncorroded–unconfined and FRP-confined and corroded–unconfined and FRP-confined RC circular columns.

Specimen	Failure load (kN)	Gain (%)	Maximum displacement (mm)	Primary moment (kNm)	Secondary moment (kNm)	Total moment (kNm)	Actual e/h at failure
Ctrl	132.30	–	9.44	6.62	1.25	7.86	0.08
Ctrl-1-C	158.87	20.08	14.76	7.94	2.35	10.29	0.08
Ctrl-2-C	168.75	27.55	10.43	8.44	1.76	10.20	0.08
Ctrl-1-G	141.48	6.94	9.97	7.07	1.41	8.48	0.08
Ctrl-2-G	144.18	8.98	13.28	7.21	1.92	9.12	0.08
Ctrl-H	182.20	37.72	11.17	9.11	2.04	11.15	0.08
M-Ctrl	115.13	–	6.75	5.76	0.78	6.53	0.07
M-1-C	135.00	17.26	12.34	6.75	1.67	8.42	0.08
M-2-C	153.04	32.93	18.67	7.65	2.86	10.51	0.09
M-1-G	125.17	8.72	13.95	6.26	1.75	8.00	0.08
M-2-G	146.02	26.83	20.81	7.30	3.04	10.34	0.09
M-H	157.03	36.39	11.41	7.85	1.79	9.64	0.08
S-Ctrl	84.24	–	7.27	4.21	0.61	4.82	0.07
S-1-C	101.65	20.67	13.30	5.08	1.35	6.43	0.08
S-2-C	109.34	29.79	9.79	5.47	1.07	6.54	0.08
S-1-G	103.29	22.61	9.89	5.16	1.02	6.19	0.08
S-2-G	110.89	31.64	15.05	5.54	1.67	7.21	0.08
S-H	112.84	33.95	17.37	5.64	1.96	7.60	0.09

Figures 18(a) and (b) show the moment–curvature profile at mid-height of FRP-confined RC circular columns. Based on the plane section assumption, the moment–curvature can be obtained using the differential longitudinal strain on the tension and compression faces of mid-height section.⁷

M = P (e + U_{Lat})

(2)

φ = \frac{ɛ_{L, top} - ɛ_{L, bot}}{h}

(3)

where

ɛ_{L, top}

is the longitudinal strain on tension face of column at mid-height section,

ɛ_{L, bot}

is the longitudinal strain of compression face of column at mid-height section and h is the height of section (i.e. 500 mm).

As seen in results, there was no significant difference in bending stiffness, moment capacity except specimen M-2-C and curvature capacities between non-hybrid CFRP and GFRP, and hybrid FRP-confined specimens in term of mild corrosion damages; however, the bending stiffness and moment capacity of severe corroded non-hybrid CFRP and hybrid FRP specimens substantially dropped down by 36% and 31%, respectively.

Conclusions

A series of 18 reinforced concrete circular columns were cast with a concrete grade of 25 MPa. These specimens were exposed to different corrosion levels. Among all, 12 RC circular columns were exposed to two different levels of corrosion in terms of rebar mass loss, namely mild and severe corrosion levels and the remaining six specimens were not exposed to any corrosion activity. The specimens were exposed to mild corrosion level by 7% rebar mass loss which is equivalent to a cross-sectional loss of about 5.5 mm², and severe corrosion level by 16.5% of mass loss which is equivalent to a cross-sectional loss of about 12.96 mm². All, except the unconfined specimens, 15 reinforced concrete specimens were wrapped with non-hybrid CFRP and GFRP and hybrid FRP reinforcement. The following conclusions can be deduced from this experimental investigation:

The gain in strength of corrosion-damaged RC circular columns repaired with hybrid and non-hybrid FRP reinforcement was varied between 8% and 36% for mild corrosion level with a steel rebar mass loss of 7%, and 21% and 34% for severe corrosion level with rebar mass loss of 16.5% over the unconfined corrosion-damaged RC circular columns. Moreover, the performance of FRP-confined severe corroded RC columns with a rebar mass loss of 16.5% was also attained a maximum strength similar to the FRP-confined mild corroded RC columns with a rebar mass loss of 8%.

The load-carrying capacity of uncorroded FRP confined RC circular columns ranged between 7% and 38% over the uncorroded–unconfined column. Moreover, the load-carrying capacity of mild and severe corroded FRP confined RC circular columns was less than uncorroded FRP-confined RC circular columns without any corrosion activity.

Results show that the corrosion-damaged unconfined RC columns were failed in buckling with crushing of concrete. However, all the corrosion-damaged non-hybrid CFRP and GFRP and hybrid FRP confined RC circular columns were failed in FRP fracture with buckling. Moreover, no debonding of FRP reinforcement was observed in any of the hybrid FRP and non-hybrid CFRP and GFRP repaired RC circular columns.

The results show that the performance of hybrid FRP strengthened corrosion-damaged RC circular columns was better over the non-hybrid CFRP and GFRP strengthened RC circular columns with two plies of FRP reinforcement. However, the results of CFRP strengthened RC circular columns were close to the hybrid FRP-strengthened circular columns.

It was generally observed that the stiffness of displacement curves of mild corroded FRP-confined RC circular columns was marginally varied with respect to type of FRP, number of FRP layers and hybrid FRP reinforcement. However, for severe-corroded FRP-confined RC circular columns, no significant difference in the stiffness of displacement curve was observed with different test variables due to the effect of increase in percentage of corrosion level. Moreover, the ductility of mild and severe corroded non-hybrid and hybrid FRP confined RC columns was relatively greater than corrosion-damaged unconfined RC columns.

The FRP hoop strain values in all uncorroded FRP confined, mild and severe-corroded FRP-confined RC circular columns were relatively low as compared to the longitudinal FRP strain values. However, the longitudinal strain in FRP-confined RC circular columns was significantly increased after the peak load.

Footnotes

Funding

The authors are grateful for the funding support provided by Research University (RU) and Postgraduate Research Grant Scheme (PRGS) (grants Ref. No. 811064 and 8043022) from Universiti Sains Malaysia, Malaysia.

Conflict of interest

None declared.

References

ACI 222R-01. Protection of Metals in Concrete Against Corrosion. American Concrete Institute, Farmington Hills, MI: American Concrete Institute, 2005.

El-Maaddawy

Chahrour

Soudki

. Effect of fiber-reinforced polymer wraps on corrosion activity and concrete cracking in chloride-contaminated concrete cylinders. J Compos Construct 2006; 10: 139–147.

Pantazopoulou

Bonacci

Sheikh

. Repair of corrosion-damaged columns with FRP wraps. J Compos Construct 2001; 5: 3–11.

Hadi

MNS

. Behaviour of externally confined high strength concrete columns under eccentric loading. Compos Struct 2003; 62: 145–153.

Hadi

MNS

. External reinforcement of high strength concrete columns. Compos Struct 2004; 65: 279–287.

El-Maaddawy

. Behaviour of corrosion-damaged RC columns wrapped with FRP under combined flexural and axial loading. Cement Concrete Compos 2008; 30: 524–534.

Sadeghian

Rahai

Ehsani

. Experimental study of rectangular RC columns strengthened with CFRP composites under eccentric loading. J Compos Construct 2010; 14: 443–450.

Hadi

MNS

. Behaviour of FRP wrapped normal strength concrete columns under eccentric loading. Compos Struct 2006; 72: 503–511.

Lee

Bonacci

Thomas

MDA

. Accelerated corrosion and repair of reinforced concrete columns using carbon fibre reinforced polymer sheets. Can J Civil Eng 2000; 27: 941–948.

10.

Masoud

Soudki

. Evaluation of corrosion activity in FRP repaired RC beams. Cement Concrete Compos 2006; 28: 969–977.

11.

Soudki KA, Sherwood T and Masoud S. FRP repair of corrosion-damaged reinforced concrete beams. Report 2002, Department of Civil Engineering, University of Waterloo, Canada.

12.

Belarbi

Bae

. An experimental study on the effect of environmental exposures and corrosion on RC columns with FRP composite jackets. Compos Part B 2007; 38: 674–684.

13.

Austin

Lyons

Ing

. Electrochemical behaviour of steel reinforced concrete during accelerated corrosion testing. Corrosion 2004; 60: 203–212.

14.

Wang

Shih

Hong

. Rehabilitation of cracked and corroded reinforced concrete beams with fiber-reinforced plastic patches. J Compos Construct 2004; 8: 219–228.

15.

Raupach

. Patch repairs on reinforced concrete structures – Model investigations on the required size and practical consequences. Cement Concrete Compos 2006; 28: 679–684.

16.

Sangeeta

Mukherjee

Malhotra

. Corrosion of steel reinforcements embedded in FRP wrapped concrete. Construct Build Materi 2009; 23: 153–161.

17.

Revathy

Suguna

Raghunath

. Effect of corrosion damage on the ductility performance of concrete columns. Am J Eng Appl Sci 2009; 2: 324–327.

18.

Spainhour

Wootton

. Corrosion process and abatement in reinforced concrete wrapped by fiber reinforced polymer. Cement Concrete Compos 2008; 30: 535–543.

19.

BASF manufacturer’s manual 2009.

20.

ASTM Standard D3039/D3039M - 14. Standard Test Method for Tensile Properties of Polymer Matrix Composite Materials. West Conshohocken, PA: ASTM International, 2005.

21.

ASTM Standard G1-03. Standard Practice for Preparing, cleaning, and evaluating corrosion test specimens. West Conshohocken, PA: ASTM International, 2003.