- Email: [email protected]
Experimental study on rehabilitation of corrosion-damaged reinforced concrete beams with carbon fiber reinforced polymer
Construction and Building Materials 38 (2012) 708–716
Contents lists available at SciVerse ScienceDirect
Construction and Building Materials journal homepage: www.elsevier.com/locate/conbuildmat
Experimental study on rehabilitation of corrosion-damaged reinforced concrete beams with carbon fiber reinforced polymer Jian-he Xie a,b,⇑, Ruo-lin Hu b,c a
School of Civil and Transportation Engineering, Guangdong University of Technology, Guangzhou, China CCCC Forth Harbor Engineering Institute Co., Ltd., Guangzhou, China c Key Laboratory of Harbor and Marine Structure Durability Technology Ministry of Communications, Guangzhou, China b
h i g h l i g h t s " A modified retrofit method is developed for the corrosion-damaged RC beams. " This method provides better load carrying capacity for the corroded beams. " Optimizing CFRP amount to balance strength recovery with control of failure mode.
a r t i c l e
i n f o
Article history: Received 16 July 2012 Received in revised form 23 August 2012 Accepted 18 September 2012 Available online 23 October 2012 Keywords: Carbon fiber reinforced polymer (CFRP) Corrosion Retrofit Reinforced concrete (RC) Beam
a b s t r a c t The repair and rehabilitation of reinforced concrete (RC) structures in coastal areas is a challenging engineering problem. In order to retrofit the degraded RC structures, numerous studies have been devoted to the method of non-corroded RC components retrofitted with carbon fiber reinforced polymer (CFRP). There is, however, less research available on corroded, patched and CFRP-repaired RC specimens. The purpose of this paper is to investigate rehabilitation of corrosion-damaged RC beams with CFRP, which focus on the effectiveness of CFRP-repaired methods and the effects of CFRP amount on flexural behavior of the beams. In this study, a modified retrofit method based on substrate repairs was developed, which is bonding CFRP after replacing V-notch of substrate concrete with polymer mortar. To compare the modified method with two common retrofit methods, which are respectively bonding directly CFRP and bonding CFRP after replacing damaged concrete, four-point bending experiments were conducted on a series of corrosion-damaged RC beams with CFRP. Important factors were considered in the experimental study, including the number of CFRP layers and corrosion level denoted by the mass loss rate of tensile steel. There were totally 32 RC beams (250 mm 150 mm 1400 mm) constructed in these experiments, 27 of which were corroded by an accelerated aging approach. The results show that the modified retrofit method could provide better load carrying capacity for the beams having more than 15% mass loss of tensile steel. In addition, to improve the short-term performance, the simple method of directly bonding CFRP was suitable for the beams having less than 15% mass loss of tensile steel. It is noted that bonded CFRP could not work for the damaged beams which undertaken more than 50% mass loss of tensile steel. In particular, it is indicated that by optimizing the amount of CFRP, it is possible to balance strength recovery with control of failure mode. Ó 2012 Elsevier Ltd. All rights reserved.
1. Introduction Concrete highway bridges in coastal areas are continuously exposed to chloride corrosion and dry-wet cyclic attack, which lead to degradation of the concrete and the reinforcing steel. Corrosion of reinforcing steel leads to a reduction in the cross-sectional area, ⇑ Corresponding author at: School of Civil and Transportation Engineering, Guangdong University of Technology, Guangzhou, China. Tel.: +86 20 39322538, mobile: +86 13538832879; fax: +86 20 39322511. E-mail address: [email protected] (J.-h. Xie). 0950-0618/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.conbuildmat.2012.09.023
and produces corrosion products with higher volume than the original steel resulting in cracking of the concrete surrounding the bars. These deleterious environmental factors interacted with loads would aggravate the deterioration processes, leading to the reduction of the load carrying capacity and the safety of the structure. Nowadays, reinforced concrete (RC) structures deterioration has motivated the development of new and innovative materials and methods for structural rehabilitation, as replacement of structures would be very costly and nearly prohibitive. Traditionally, corroded RC structures such as bridges and marine structures have been repaired with steel plates. However, steel
709
J.-h. Xie, R.-l. Hu / Construction and Building Materials 38 (2013) 708–716
ness of CFRP-repaired methods and the effects of CFRP amount on flexural behavior of corroded RC beams. 2. Experimental program 2.1. Material properties There were totally 32 RC beams (250 mm 150 mm 1400 mm) constructed in this study. The typical geometry and reinforcement of the beams are shown in Fig. 1. The specimens were fabricated from ordinary Portland cement. The cube compressive strength and elastic modulus of concrete were 33.3 MPa and 30 GPa, respectively. The internal reinforcement was made of smooth bars, and the main tensile steel exhibited a yield strength of 310 MPa and an elastic modulus of 206 GPa. The elastic modulus of CFRP sheets was 205 MPa, with an ultimate tensile strength of 3500 MPa, and one layer of CFRP sheets had a nominal thickness of 0.11 mm. 2.2. Accelerated corrosion Since the objective of this research is to evaluate the effectiveness of applying CFRP to retrofit corrosion-damaged RC beams, 27 RC beams were firstly corroded by an accelerated corrosion system, and 5 specimens were un-corroded as the control beams. In this corrosion system, the bottom surface of beam was coated with a copper mesh, and then the beams were respectively connected by copper mesh to the cathode and by steel bar to the anode of the DC power supply, as shown in Fig. 2. In order to accelerate the corrosion process, the beams were placed in a corrosion pool which contained industrial salt solution of 3% concentration, and then subjected to wet-dry cycles (1 day wet and 2 days dry). To evaluate the corrosion level, the mass loss rate of tensile steel was regarded as corrosion rate, which can be easily calculated based on the assumption that the corrosion state is uniformly distributed within the steel material. In order to achieve the desired mass loss of tensile steel, the corrosion time was estimated using Faraday’s law, which is expressed by the following equation:
m¼
atI nF
ð1Þ
where m = mass loss; I = corrosion current; t = time of the corrosion process; a = atomic mass of iron (55.85 g); n = valence of the reacting electrode for the material (nsteel = 2); F = Faraday’s constant (96500 C/mol). Additionally, the maximum width of corrosion cracks as another index of the corrosion level was measured by crack width detector. In this study, according to the mass loss rate of tensile steel r and corrosion crack width w, the test beams were divided into five groups as listed below: (1) (2) (3) (4) (5)
Group Group Group Group Group
A, in which the beams had no corrosion damages. B (0 < r 6 5% and 0 < w 6 0:3 mm). C (5 < r 6 15% and 0:3 mm < w 6 1:0 mm). D (15 < r 6 50% and 1:0 mm < w 6 3:0 mm). E (50% < r and 3:0 mm < w).
2.3. Retrofit methods Traditionally, there are two common CFRP-retrofit methods for the corrosiondamaged RC members, which are directly bonding CFRP (retrofit method 1) and bonding CFRP after replacing corrosion-damaged concrete with polymer mortar (retrofit method 2), respectively. Comparing with method 2, retrofit method 1 is easier to handle, but does not prevent further steel corrosion. On the other hand, though retrofit method 2 could protect steel, it does not only increase the difficulty of retrofitting, but also is more costly. In addition, the interface between old concrete and new mortar is prone to crack. Thus, retrofit method 2 is usually just used for the beams undertaken high corrosion level. To overcome the main shortcomings of the above two methods, as described below, a modified retrofit method based on substrate repairs was developed in this study. Retrofit method 1 – Directly bonding CFRP (Fig. 3). For retrofit method 1, none of the old deteriorated concrete was removed prior to CFRP repair, CFRP were directly bonded on the tensile surface of the corrosion-damaged beams. During the retrofit
250mm
20mm
30mm
plates are prone to corrosion, and special equipments are needed to install these heavy plates. In recent years, carbon fiber reinforced polymer (CFRP) materials in the form of fabrics and laminates have been motivated to use for retrofitting these corrosion-damaged RC components. CFRP materials are lightweight, noncorrosive, and exhibit high tensile strength [1]. Additionally, they can be easily bonded to the concrete surface on-site, without the use of extensive scaffolding and jacks, requiring minimum amount of support equipment [2]. CFRP systems can also be used in areas with limited access where traditional techniques would be difficult to implement. Therefore, CFRP-repair technique not only increases the ease of the retrofitting, but also is more costefficient. The advantages of retrofit technology of using CFRP externally bonded on concrete attracted the attentions of numerous researchers. Some researches have been conducted on the durability of CFRP-repaired concrete beams. In these studies, some researchers used natural corrosion methods to account for deteriorate material properties, either obtained directly from the field [3] or conditioned in the laboratory [1,4,5]. Although natural corrosion methods could closely simulate field conditions, they require a long time to achieve the desired corrosion level. To accelerate the aging process, others researchers used the application of induced electrical current [2,6–15]. By using the accelerated corrosion methods, Wang et al. [2], Bonacci and Maalej [7], Masoud and Soudki [10] experimentally studied the short-term performance of corrosiondamaged RC beams retrofitted with CFRP, Masoud and Soudki [11], Al-Hammoud et al. [13] investigated the fatigue behavior of corroded RC beams with CFRP, and Al-Hammoud et al. [14] focused on the durability of CFRP-concrete interface. However, these researchers did not investigate the effects of concrete patching before bonding CFRP. In fact, substrate repairs are often carried out on corroded RC structures to replace the damaged concrete and to restore the bond between steel and concrete. During a substrate repair, some measures such as application of corrosion inhibitors, cathodic protection and chloride extraction could be undertaken to prevent further steel corrosion [6,15]. Although numerous studies have been devoted to the performance of non-corroded RC specimens retrofitted with CFRP, not much information is available on corroded and CFRP-repaired flexural RC components, especially on the effects of combining substrate repair and CFRP repair. To get better understanding of the performance of corroded RC structures with CFRP, an extensive experimental program was conducted by the authors to evaluate the post-repair structural behavior of corrosion-damaged beams repaired with three retrofit methods. In this study, a modified retrofit method was developed based on substrate repairs, which is bonding CFRP after replacing V-notch of substrate concrete with polymer mortar. To compare the modified method with two common retrofit methods, which are respectively bonding directly CFRP and bonding CFRP after replacing damaged concrete, fourpoint bending experiments were conducted on a series of corrosion-damaged RC beams with CFRP. Additionally, important factors were considered in the experimental study, including the number of CFRP layers and corrosion level which was denoted by the mass loss rate of tensile steel. The focus in this paper is on the effective-
1400mm Fig. 1. Concrete beam-geometry and reinforcement details.
150mm
710
J.-h. Xie, R.-l. Hu / Construction and Building Materials 38 (2013) 708–716 Corrosion pool
high-water mark
Beam
Beam
Beam
Beam
Beam
Copper mesh
Pad
(a) Power supply
Steel
Copper mesh
(b) Fig. 2. Accelerated corrosion system: (a) schematic diagram of corrosion pool and (b) circuit of accelerated corrosion.
Fig. 4. Retrofit method 2: (a) removing damaged cover concrete and (b) replacing with polymer mortar.
All of the beams were subjected to four-point bending in this study. The loads were applied with a servo-hydraulic actuator controlled by force mode, and a load cell was installed on the loading actuator to record the applied loads. In addition, three dial indicators with an accuracy of 0.01 mm were used to measure the midspan deflection of the beam. Fig. 7 shows the test setup. Fig. 3. Beams retrofitted by method 1. process, the concrete surface was firstly cleaned and smoothed with the putty fillers followed by pasting the primer along the axial direction, and then the saturating resin was used to impregnate followed by bonding CFRP sheets. Retrofit method 2 – Bonding CFRP after replacing corrosion-damaged concrete with polymer mortar. For retrofit method 2, the corrosion-damaged concrete to the level of the reinforcement was removed followed by derusting, and replaced with polymer mortar (Fig. 4), then CFRP sheets were bonded on the bottom surface of new concrete cover after the polymer mortar was cured. Retrofit method 3 (A modified retrofit method) – Bonding CFRP after replacing Vnotch with polymer mortar. This modified retrofit method was based on substrate repairs. In this retrofit method, the bottom surface of concrete cover was firstly cutting by V-notch along the longitudinal corrosion cracks to the level of the main tensile steel, and then the V-notch was repaired with polymer mortar after derusting, as shown in Fig. 5. After the polymer mortar was cured, CFRP sheets were bonded on the tensile surface of the beam. Thus, comparing with method 1, the modified method prevents further steel corrosion, and comparing with method 2, it increases the ease of retrofitting. During the CFRP repair process, the test beams were bonded with 150 mm wide CFRP sheets on the tensile side, and then bonded with U-shaped CFRP strips of 80 mm wide at intervals of 80 mm near the CFRP end, as shown in Fig. 6. The main reason for using the U-shaped strip in the beams of Group C, D and E was to prevent CFRP end debonding due to serious corrosion damages. To more easily analyze the character of the test beams, all the specimens were coded according to corrosion level, retrofit method, and the number of CFRP layers, as listed in Table 1. The definitions of the code letters is that the first subscript represent retrofit method, and the second subscript represent the number of CFRP layers, and the third subscript denote the number of the beam under the same conditions. It is noted that the specimens with the subscript zero are the control beams without CFRP. 2.4. Testing procedure Prior to testing, all specimens were instrumented with electrical resistance strain gauges at the midspan. Six strain gauges were placed on the concrete surface at intervals of 50 mm along the height of the beam, two gauges were installed on the main steel bars and two gauges were installed on the surface of CFRP, as shown in Fig. 6.
3. Results and discussion The experimental results are summarized in Table 1, including yielding loads, ultimate loads and failure modes of the specimens. The results obtained from these tests are analyzed in the following. 3.1. Effects of retrofit methods 3.1.1. Failure mode A summary of the failure modes of the tested beams are listed in Table 1. The testing identified the following seven failure modes: (1) The tensile steel yields and concrete compression crushing occurs. (2) Concrete beam undergoes compressive crushing but the steel does not yield. (3) The tensile steel yields but concrete does not compression crush. (4) The tensile steel yields followed by concrete cover separating and CFRP sheets debonding. (5) The tensile steel yields followed by CFRP sheets breaking. (6) The tensile steel yields followed by CFRP sheets debonding. (7) Cracks between old concrete and new mortar followed by CFRP sheets debonding. As listed in Table 1, all the beams, which were directly bonded with one layer of CFRP sheets, underwent the process of the tensile steel yielding, with the exception of Beam A1-1-1 and Beam E1-1-1. Beam A1-1-1 experienced failure mode 2, which indicates that the amount of CFRP sheets for Beam A1-1-1 was excessive. It is noteworthy that failure mode 3, which is obvious brittle failure, occurred in
711
J.-h. Xie, R.-l. Hu / Construction and Building Materials 38 (2013) 708–716
Fig. 6. Schematic representation of corrosion-damaged beam strengthened with CFRP (units: mm).
Table 1 Summary of test conditions and results. Group Specimen Retrofit na wb rc(%) method (mm)
Yielding Ultimate Failure mode load load (kN) (kN)
A
68.6 88.0 67.2 92.5 83.0 101.1 105.5 113.5 114.3 66.0 90.8 92.3 92.5 95.3 112.1 98.2 84.6 56.2 72.0 71.3 81.3 80.0 112.3 100.9 89.8 102.2 65.1 68.2 54.5 41.3 59.8 42.2
B
C
D
E
a b c
Fig. 5. Modified retrofit method (retrofit method 3): (a) cutting V-notch, (b) part of V-notch, and (c) repairing V-notch with polymer mortar.
Beam E1-1-1. This shows that directly bonding CFRP is not suitable for the beams which had 50% mass loss of steel. The failure modes of beams retrofitted by method 1 are shown in Fig. 8. The failure mode 4 of the beams in Group B occurred near the longitudinal CFRP end, which can be explained by the lack of the end anchors, as shown in Fig. 8a. For the beams in Group C, CFRP sheets breaking
A0 A1-1-1 B0 B1-1-1 B1-1-2 B1-2-1 B1-2-2 B1-3-1 B1-3-2 C0 C1-1-1 C1-1-2 C1-2-1 C1-2-2 C1-3-1 C1-3-2 C3-1-1 D0 D1-1-1 D1-1-2 D1-2-1 D1-2-2 D1-3-1 D1-3-2 D3-1-1 D3-2-1 D2-1-2 D2-2-2 E0 E1-1-1 E2-2-1 E2-3-1
– 1 – 1 1 1 1 1 1 – 1 1 1 1 1 1 3 – 1 1 1 1 1 1 3 3 2 2 – 1 2 2
– 1 – 1 1 2 2 3 3 – 1 1 2 2 3 3 1 – 1 1 2 2 3 3 1 2 1 2 – 1 2 3
0 0 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.7 0.7 0.6 0.6 0.7 0.6 0.6 0.7 1.4 1.4 1.2 1.3 1.2 1.3 1.2 1.1 1.1 1.5 1.6 3.1 4.6 3.5 4.4
– – 3–4 3–4 3–4 3–4 3–4 3–4 3–4 8–11 8–11 8–11 8–11 8–11 8–11 8–11 8–11 18–22 18–22 18–22 18–22 18–22 18–22 18–22 18–22 18–22 18–22 18–22 50–55 70–78 50–55 70–78
102.7 150.6 99.4 140.5 118.8 150.6 138.8 154.6 160.6 93.7 135.5 145.3 145.6 135.5 152.3 145.6 132.2 82.0 125.5 132.2 128.8 135.5 148.9 138.9 145.6 172.3 92.0 98.7 78.7 55.2 77.0 58.6
1 2 3 4 4 4 4 4 4 3 5 5 4 4 4 6 5 3 6 6 6 6 6 6 5 6 7 7 3 3 3 3
n = The number of CFRP layers. w = The maximum width of corrosion cracks before loading. r = The mass loss rate of tensile steel.
occurred near the loading point, as shown in Fig. 8b. This failure mode got the maximum utilization of CFRP strength. However, for the beams in Group D, CFRP debonding occurred from midspan to end, as shown in Fig. 8c. This failure mode is mainly because that serious corrosion damages degraded the adhesion performance of CFRP-concrete interface, resulting in the initiation of interface cracks and then propagation. The failure modes of beams retrofitted by method 2 are shown in Fig. 9. As shown in Fig. 9a, the beams in Group D experienced failure mode 7. The process of failure mode 7 is that interface cracks initiate and propagate between old concrete and new polymer mortar, subsequently the concrete cover separates followed by CFRP debonding. Obviously, this failure mode decreased the efficiency of CFRP. For the beams in Group E retrofitted by method 2, the failure mode was similar as that of the beam retrofitted by method 1, as show in Fig. 9b. This can be explained by that the tensile steel of the beams in Group E had been seriously corrosiondamaged, subsequently the corrosion cracks generated, resulting in failure before CFRP sheets working. This experiment result also
712
J.-h. Xie, R.-l. Hu / Construction and Building Materials 38 (2013) 708–716
Fig. 7. Testing system. Fig. 9. Failure modes of beams retrofitted by method 2: (a) Beam D2-1-2 and (b) Beam E2-3-1.
Fig. 10. Failure mode of beams retrofitted by method 3: (a) Beam D3-1-1 and (b) Beam D3-2-1. Fig. 8. Failure modes of beams retrofitted by method 1: (a) Beam B1-1-1, (b) Beam C1-1-1, and (c) Beam D1-1-1.
indicates that the retrofit methods with CFRP are not suitable for the beams which had 50% mass loss of steel. So, in this study, retrofit method 3 did not be used for repairing the beams in Group E. The failure modes of beams retrofitted by the modified method are shown in Fig. 10. As shown in Fig. 10a, the beams with one layer of CFRP sheets in Group C and Group D experienced failure mode 5. Compared with method 1 and method 2, the modified retrofit method provided a more ideal failure mode for the beams, which could increase the utilization rate of CFRP strength. For the beams with two layers of CFRP sheets in Group D, CFRP debonding occurred followed by U-shaped strips broke, as shown in Fig. 10b. These results indicate that the CFRP thickness should be reasonably arranged to optimize to balance strength recovery with control of failure mode. 3.1.2. Load carrying capacity The ultimate loads of the retrofitted beams with one layer of CFRP sheets (except the beams in Group E) are shown in Fig. 11. It can be seen from Fig. 11, retrofit method 1 and method 3 could obviously improve the load carrying capacity of the beams in
Group B, C and D, but the strengthening effectiveness of retrofit method 2 is poor for the beams in Group D. To analyze the effects of retrofit methods on the load carrying capacity of the beams, the increase rate of the ultimate loads between the retrofitted beams with one layer CFRP and the control beam in same group is shown in Fig. 12. As can be seen in Fig. 12, retrofit method 1 averagely increased the ultimate loads of the beams in Group B by 30.4%. For the beams in Group C, retrofit method 1 and method 3 had averagely increased the ultimate loads by 49.8% and 44.1%, respectively. For the beams in Group D, the ultimate loads had an averagely increase of 57.1%, 16.3% and 93.8% by retrofit method 1, method 2 and method 3, respectively. However, for Beam E2-2-1 with two layers of CFRP in Group E, retrofit method 2 decreased the ultimate loads of the beams by 2.2%. These results indicate that, compaired with two common methods, the modified retrofit method could provide better load carrying capacity for the beams having more than 15% mass loss of steel. In addition, to improve the short-term performance, the simple method of directly bonding CFRP was suitable for the beams which had less than 15% mass loss of steel. It is noted that the beams undertaken more than 50% mass loss of steel have no need to be retrofitted by bonding CFRP sheets.
713
J.-h. Xie, R.-l. Hu / Construction and Building Materials 38 (2013) 708–716
150
160
140
140 120
120
Group B Group C Group D Group E
110 100 90
100
Load (kN)
Ultimate load (kN)
130
80 60
B0
80
B1-1-1
40
B1-1-2
70 20
60
0
50 None
1
2
3 0
Retrofit method
2
4
6
8
10
12
Midspan deflection (mm)
(a)
Fig. 11. Ultimate loads of beams retrofitted by different methods.
160 100%
120
Retrofit method 2
80%
Retrofit method 3
Load (kN)
Improvment percent
140
Retrofit method 1
60%
100 80 60
C-0 C1-1-1 C3-1-1
40% 40 20
20%
0 0%
0 B
-20%
C
D
2
E
4
6
8
10
12
14
16
18
Midspan deflection (mm)
Corrosion group
(b)
Fig. 12. Effects of retrofit methods on load carrying capacity.
160 140
3.1.4. Strain evolvement The strains variation of beam in Group B (Beam B1-3-1) is shown in Fig. 14. From Fig. 14a, it can be seen that the midspan strains of CFRP and steel have similar three-stage evolvement as the flexural
120
Load (kN)
3.1.3. Flexural stiffness The curves of loads versus midspan deflection for the retrofitted beams with one layer of CFRP sheets are shown in Fig. 13. As shown in the curves, the evolvement of flexural stiffness of the beams retrofitted with different methods had almost same trends, and the whole evolvement process during loading can be divided into three phases: (1) In the elastic phase, the relationship between the loads and the midspan deflection is linear; (2) In the cracked phase, the curves are approximately linear, but the slope of the curve is less than that in the first phase; (3) In the yielded phase, the deflection rapidly increased resulting in the failure of the beams. As shown in Fig. 13a, the flexural stiffness of the beams retrofitted by method 1 in Group B is larger than that of the non-strengthened beams during three phases, especially in the third phase. For the beams in Group C, the flexural stiffness of the beams retrofitted by method 1 has a similar increase as that of the beams retrofitted by method 3, as shown in Fig. 13b. For the beams in group D, as can be seen from Fig. 13c, retrofit method 3 obviously provided better flexural stiffness and ductility than retrofit method 1 and method 2.
100 80 60
D-0 D1-1-1 D2-1-2 D3-1-1
40 20 0 0
2
4
6
8
10
12
14
16
Midspan deflection (mm)
(c) Fig. 13. Midspan deflection of beams retrofitted with different methods: (a) Group B, (b) Group C, and (c) Group D.
stiffness, including elastic phase, cracked phase and yielded phase. In Fig. 14b, the strains distribution along the midspan cross section is shown during loading. In this figures, four strain values were measured including CFRP strains, steel strains and two concrete strains. As shown in Fig. 14b, the strains along the cross section are obviously linear distribution before cracking, and though CFRP
714
J.-h. Xie, R.-l. Hu / Construction and Building Materials 38 (2013) 708–716
3500
7000 6000
3000
Strain (με)
Strain ( με)
2000
Top Concrete Midspan CFRP 1 Midspan CFRP 2
5000
Midspan CFRP 1 Midspan CFRP 2 Steel
2500
1500
4000 3000 2000 1000
1000
0
500
-1000 0
-2000 0
20
40
60
80
100
120
140
50
80
(a)
(a)
100
120
8 kN 28 kN 48 kN 69 kN 82 kN 95 kN 109 kN 122 kN 142 kN
200 150 100 50
140
160
8 kN 28 kN 39 kN 48 kN 92 kN 98 kN 105 kN 112 kN 119 kN 125 kN 132 kN
250
0
0 -1000
60
Section height (mm)
100
40
Load (kN)
Section height (mm)
150
20
Load (kN)
250
200
0
160
-500
0
500
1000
1500
2000
2500
3000
-1000
0
1000
2000
Strain ( με)
Strain (με)
(b)
(b)
3000
4000
5000
Fig. 14. Strains variation of Beam B1-3-1: (a) load–strain curves and (b) strains distribution along midspan cross section.
Fig. 15. Strains variation of Beam C1-1-2: (a) load–strain curves and (b) strains distribution along midspan cross section.
strains lag behind other strains due to slipping between CFRP and concrete, the strains along the cross section approximately keep a linear distribution after cracking. For the beams in Group C, Fig. 15 shows the strains variation of Beam C1-1-2. It can be seen from Fig. 15a that the strain values of concrete have similar trends as that of CFRP. In Fig. 15b, the strains distribution along the midspan cross section is shown during loading, including CFRP strains, two concrete strains, and the part of steel strains. As shown in Fig. 15b, the strains along the cross section are obviously linear distribution before cracking, but after cracking, steel strains obviously lag behind other strains due to slipping between serious damaged steel and concrete, resulting in the strains along the cross section could not keep linear distribution. Thus, to simplify the calculation, it is recommended that plane section assumption is still suitable for the CFRP-retrofitted beams having less than 5% mass loss of tensile steel.
beam, the flexural stiffness of the retrofitted beams increases with the increase of CFRP layer number, but the yield point of the retrofitted beams is not easy to judge. This indicates that the increase of CFRP thickness would enhance the probability of brittle failure of the beams.
3.2. Effects of CFRP thickness 3.2.1. Flexural stiffness To analyze the effects of CFRP thickness on the flexural stiffness of the beams, the curves of loads versus midspan deflection of the beams retrofitted with different number of CFRP layer is shown in Fig. 16. As can be seen in Fig. 16, compared with that of the control
3.2.2. Load carrying capacity The ultimate loads of the corrosion-damaged RC beams retrofitted with different CFRP thickness are shown in Fig. 17. In this figure, it can be seen that the load carrying capacity of the retrofitted beams could increases with the increase of CFRP thickness. To discuss the effects of CFRP thickness on the load carrying capacity of the corrosion-damaged RC beams, the increase rate of the ultimate loads and yielding loads between the retrofitted beams and the control beam in same group are shown in Fig. 18. For the beams in Group B, as shown in Fig. 18a, the ultimate loads of the retrofitted beams with one layer of CFRP sheets have same increase rate as the yielding loads, but with the increase of CFRP thickness, the increase rate of yielding loads are greater than that of ultimate loads. In addition, it can be found that the relationship between the load carrying capacity of the beams and CFRP thickness is not linear, the increase rate of load carrying capacity decreases with the increase of CFRP thickness. For the beams retrofitted by method 1 in Group C, Fig. 18b shows that the yielding loads of beams with one layer of CFRP sheets had a 40% increase,
715
J.-h. Xie, R.-l. Hu / Construction and Building Materials 38 (2013) 708–716
160
None Corrosion Group B(Retrofit method 1)
180
140
170 160
Group D (Retrofit method 1) Group D (Retrofit method 2) Group D (Retrofit method 3)
150
100 80
Ultimate load (kN)
Load (kN)
120
Group C (Retrofit method 1)
B0 B1-1-1
60
B1-1-2 B1-2-1
40
B1-3-1
20 0 0
2
4
6
8
10
12
140 130 120 110 100 90 80 70
Midspan deflection (mm)
60
(a)
None1 160
2
3
CFRP layer
140
Fig. 17. Ultimate loads of beams with different CFRP thickness.
Load (kN)
120 100 80 60 C0 C1-1-1 C1-2-2 C1-3-1
40 20 0 0
2
4
6
8
10
12
14
16
Midspan deflection (mm)
(b) 180 160 140
those of beams with more than one layer of CFRP sheets were CFRP dedonding. For the retrofitted beams by the modified method in Group D, it can be seen from Fig. 18c that one layer of CFRP sheets increased respectively the yielding loads and the ultimate loads of the beams by 60% and 78%, and caused the failure mode of CFRP fracture, while two layers of CFRP increased respectively the yielding loads and the ultimate loads by 82% and 113%, but caused the failure mode of CFRP debonding. Thus, to enhance the utilization rate of CFRP and avoid debonding, it is important to optimize CFRP amount to balance strength recovery with control of failure mode. From Fig. 18b and c, it can be found that if using the modified method to retrofit the beams, CFRP thickness should not exceed the amount which could increase the yielding loads of the beams by more than 60%, while for the beams retrofitted by directly bonding CFRP, CFRP thickness should not exceed the amount which could increase the yielding loads of the beams by more than 40%.
Load (kN)
120
4. Conclusion
100 80 60 D0
40
D3-1-1 D3-2-1
20 0 0
2
4
6
8
10
12
14
Midspan deflection (mm)
(c) Fig. 16. Effects of CFRP thickness on midspan deflection: (a) Group B, (b) Group C (retrofit method 1), and (c) Group D (retrofit method 3).
is close to that of the beams with two layers of CFRP sheets, and is slightly less than that of the beams with three layers of CFRP sheets. This result indicates that the utilization rate of more than one layer of CFRP sheets is low, which is mainly related to the failure modes of the beams. As shown in Fig. 18b, the failure mode of beams with one layer of CFRP sheets was CFRP fracture, while
In this paper, an experimental investigation was conducted into the rehabilitation of corrosion-damaged beams with CFRP sheets. A modified retrofit method based on substrate repairs was developed to compare with two common retrofit methods in this study. In addition, important factors were also considered in these experiments, including CFRP thickness and corrosion level denoted by the mass loss rate of tensile steel. The results show that the modified retrofit method could provide better load carrying capacity and flexural stiffness for the corrosion-damaged beams having more than 15% mass loss of tensile steel. For the beams undertaken less than 15% mass loss of tensile steel, the simple retrofit method of directly bonding CFRP sheets could be suitable for improving their short-term performance. It is noted that bonded CFRP could not work for the damaged beams which had more than 50% mass loss of tensile steel. In addition, by optimizing the amount of CFRP, it is possible to balance strength recovery with control of failure mode, for example, the brittle failure mode of debonding can be avoided. Based on the experimental results, it is recommended that, when using the modified method to retrofit the corrosiondamaged RC beams, CFRP thickness should not exceed the amount which could increase the yielding loads of the beams by more than
716
J.-h. Xie, R.-l. Hu / Construction and Building Materials 38 (2013) 708–716
60%. In particular, the results indicate that plane section assumption is still suitable for the CFRP-retrofitted beams having less than 5% mass loss of tensile steel. This research is expected to contribute to guidelines for CFRP-repairing corrosion-damaged RC beams under service loads and environmental effects. As an extension of the present work, a rigorous theoretical model needs to be proposed for designers to predict the load carrying capacity of CFRP-retrofitted corroded beams.
80% Yielding load
Improvement percent
70%
Ultimate load
60% 50% 40% 30%
Acknowledgements
20% 10% 0%
1
2
3
CFRP layer
(a)
References
80% Yielding load
Improvement percent
70%
Ultimate load CFRP Debonding
60% 50%
CFRP fracture
CFRP Debonding
40% 30% 20% 10% 0%
1
2
3
CFRP layer
(b) 120%
CFRP Debonding
Improvement percent
100%
80%
CFRP fracture
60% Yielding load
40%
Ultimate load
20%
0%
The authors gratefully acknowledge the financial support provided by National Natural Science Found (No. 51208117), Foundation for Distinguished Young Talents in Higher Education of Guangdong (No. LYM10069), and Doctor Science Found of Guangdong Province (No. 10451009001004769).
1
2
CFRP layer
(c) Fig. 18. Effects of CFRP thickness on load carrying capacity: (a) Group B, (b) Group C (retrofit method 1), and (c) Group D (retrofit method 3).
[1] American Concrete Institute. Guide for the design and construction of externally bonded FRP systems for strengthening concrete structures (ACI 440). Farmington Hills: Michigan; 2008. [2] Wang CY, Shih CC, Hong SC, Hwang WC. Rehabilitation of cracked and corroded reinforced concrete beams with fiber-reinforced plastic patches. J Compos Constr 2004;8(3):219–28. [3] Aidoo J. Flexural retrofit of reinforced concrete bridge girders using three CFRP systems. Columbia: Univ. of South Carolina; 2004. [4] Toutanji HA, Gómez W. Durability characteristics of concrete beams externally bonded with FRP composite sheets. Cem Concr Compos 1997;19(4):351–8. [5] Grace NF. Concrete repair with CFRP. Concr Int 2004;26(5):45–52. [6] Ray I, Parish GC, Davalos JF, Chen An. Effect of concrete substrate repair methods for beams aged by accelerated corrosion and strengthened with CFRP. J Aerospace Eng 2011;24(2):227–39. [7] Bonacci JF, Maalej M. Externally bonded fiber-reinforced polymer for rehabilitation of corrosion damaged concrete beams. ACI Struct J 2000;97(5): 703–11. [8] Hearn N. Effect of mechanical restraint on the rate of corrosion in concrete. Can J Civil Eng 1998;25(1):81–8. [9] Maddaawy TE, Soudki K. Carbon-fiber-reinforced polymer repair to extend service life of corroded reinforced concrete beams. J Compos Constr 2005;9(2): 187–94. [10] Masoud S, Soudki K. Evaluation of corrosion activity in FRP repaired RC beams. Cem Concr Compos 2006;28(10):969–77. [11] Masoud S, Soudki K, Topper T. Postrepair fatigue performance of FRP-repaired corroded RC beams: experimental and analytical investigation. J Compos Constr 2005;9(5):441–9. [12] Soudki KA. Behavior of reinforced concrete beams strengthened with carbon fiber reinforced polymer laminates subjected to corrosion damage. Can J Civil Eng 2000;27(5):1005–10. [13] Al-Hammoud R, Soudki K, Topper TH. Fatigue flexural behavior of corroded reinforced concrete beams repaired with CFRP sheets. J Compos Constr 2011;15(1):42–51. [14] Al-Hammoud R, Soudki K, Topper TH. Bond analysis of corroded reinforced concrete beams under monotonic and fatigue loads. Cem Concr Compos 2010;32(3):194–203. [15] Malumbela G, Moyo P, Alexander M. Load-bearing capacity of corroded, patched and FRP-repaired RC beams. Mag Concr Res 2011;63(11):797–812.
Contents lists available at SciVerse ScienceDirect
Construction and Building Materials journal homepage: www.elsevier.com/locate/conbuildmat
Experimental study on rehabilitation of corrosion-damaged reinforced concrete beams with carbon fiber reinforced polymer Jian-he Xie a,b,⇑, Ruo-lin Hu b,c a
School of Civil and Transportation Engineering, Guangdong University of Technology, Guangzhou, China CCCC Forth Harbor Engineering Institute Co., Ltd., Guangzhou, China c Key Laboratory of Harbor and Marine Structure Durability Technology Ministry of Communications, Guangzhou, China b
h i g h l i g h t s " A modified retrofit method is developed for the corrosion-damaged RC beams. " This method provides better load carrying capacity for the corroded beams. " Optimizing CFRP amount to balance strength recovery with control of failure mode.
a r t i c l e
i n f o
Article history: Received 16 July 2012 Received in revised form 23 August 2012 Accepted 18 September 2012 Available online 23 October 2012 Keywords: Carbon fiber reinforced polymer (CFRP) Corrosion Retrofit Reinforced concrete (RC) Beam
a b s t r a c t The repair and rehabilitation of reinforced concrete (RC) structures in coastal areas is a challenging engineering problem. In order to retrofit the degraded RC structures, numerous studies have been devoted to the method of non-corroded RC components retrofitted with carbon fiber reinforced polymer (CFRP). There is, however, less research available on corroded, patched and CFRP-repaired RC specimens. The purpose of this paper is to investigate rehabilitation of corrosion-damaged RC beams with CFRP, which focus on the effectiveness of CFRP-repaired methods and the effects of CFRP amount on flexural behavior of the beams. In this study, a modified retrofit method based on substrate repairs was developed, which is bonding CFRP after replacing V-notch of substrate concrete with polymer mortar. To compare the modified method with two common retrofit methods, which are respectively bonding directly CFRP and bonding CFRP after replacing damaged concrete, four-point bending experiments were conducted on a series of corrosion-damaged RC beams with CFRP. Important factors were considered in the experimental study, including the number of CFRP layers and corrosion level denoted by the mass loss rate of tensile steel. There were totally 32 RC beams (250 mm 150 mm 1400 mm) constructed in these experiments, 27 of which were corroded by an accelerated aging approach. The results show that the modified retrofit method could provide better load carrying capacity for the beams having more than 15% mass loss of tensile steel. In addition, to improve the short-term performance, the simple method of directly bonding CFRP was suitable for the beams having less than 15% mass loss of tensile steel. It is noted that bonded CFRP could not work for the damaged beams which undertaken more than 50% mass loss of tensile steel. In particular, it is indicated that by optimizing the amount of CFRP, it is possible to balance strength recovery with control of failure mode. Ó 2012 Elsevier Ltd. All rights reserved.
1. Introduction Concrete highway bridges in coastal areas are continuously exposed to chloride corrosion and dry-wet cyclic attack, which lead to degradation of the concrete and the reinforcing steel. Corrosion of reinforcing steel leads to a reduction in the cross-sectional area, ⇑ Corresponding author at: School of Civil and Transportation Engineering, Guangdong University of Technology, Guangzhou, China. Tel.: +86 20 39322538, mobile: +86 13538832879; fax: +86 20 39322511. E-mail address: [email protected] (J.-h. Xie). 0950-0618/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.conbuildmat.2012.09.023
and produces corrosion products with higher volume than the original steel resulting in cracking of the concrete surrounding the bars. These deleterious environmental factors interacted with loads would aggravate the deterioration processes, leading to the reduction of the load carrying capacity and the safety of the structure. Nowadays, reinforced concrete (RC) structures deterioration has motivated the development of new and innovative materials and methods for structural rehabilitation, as replacement of structures would be very costly and nearly prohibitive. Traditionally, corroded RC structures such as bridges and marine structures have been repaired with steel plates. However, steel
709
J.-h. Xie, R.-l. Hu / Construction and Building Materials 38 (2013) 708–716
ness of CFRP-repaired methods and the effects of CFRP amount on flexural behavior of corroded RC beams. 2. Experimental program 2.1. Material properties There were totally 32 RC beams (250 mm 150 mm 1400 mm) constructed in this study. The typical geometry and reinforcement of the beams are shown in Fig. 1. The specimens were fabricated from ordinary Portland cement. The cube compressive strength and elastic modulus of concrete were 33.3 MPa and 30 GPa, respectively. The internal reinforcement was made of smooth bars, and the main tensile steel exhibited a yield strength of 310 MPa and an elastic modulus of 206 GPa. The elastic modulus of CFRP sheets was 205 MPa, with an ultimate tensile strength of 3500 MPa, and one layer of CFRP sheets had a nominal thickness of 0.11 mm. 2.2. Accelerated corrosion Since the objective of this research is to evaluate the effectiveness of applying CFRP to retrofit corrosion-damaged RC beams, 27 RC beams were firstly corroded by an accelerated corrosion system, and 5 specimens were un-corroded as the control beams. In this corrosion system, the bottom surface of beam was coated with a copper mesh, and then the beams were respectively connected by copper mesh to the cathode and by steel bar to the anode of the DC power supply, as shown in Fig. 2. In order to accelerate the corrosion process, the beams were placed in a corrosion pool which contained industrial salt solution of 3% concentration, and then subjected to wet-dry cycles (1 day wet and 2 days dry). To evaluate the corrosion level, the mass loss rate of tensile steel was regarded as corrosion rate, which can be easily calculated based on the assumption that the corrosion state is uniformly distributed within the steel material. In order to achieve the desired mass loss of tensile steel, the corrosion time was estimated using Faraday’s law, which is expressed by the following equation:
m¼
atI nF
ð1Þ
where m = mass loss; I = corrosion current; t = time of the corrosion process; a = atomic mass of iron (55.85 g); n = valence of the reacting electrode for the material (nsteel = 2); F = Faraday’s constant (96500 C/mol). Additionally, the maximum width of corrosion cracks as another index of the corrosion level was measured by crack width detector. In this study, according to the mass loss rate of tensile steel r and corrosion crack width w, the test beams were divided into five groups as listed below: (1) (2) (3) (4) (5)
Group Group Group Group Group
A, in which the beams had no corrosion damages. B (0 < r 6 5% and 0 < w 6 0:3 mm). C (5 < r 6 15% and 0:3 mm < w 6 1:0 mm). D (15 < r 6 50% and 1:0 mm < w 6 3:0 mm). E (50% < r and 3:0 mm < w).
2.3. Retrofit methods Traditionally, there are two common CFRP-retrofit methods for the corrosiondamaged RC members, which are directly bonding CFRP (retrofit method 1) and bonding CFRP after replacing corrosion-damaged concrete with polymer mortar (retrofit method 2), respectively. Comparing with method 2, retrofit method 1 is easier to handle, but does not prevent further steel corrosion. On the other hand, though retrofit method 2 could protect steel, it does not only increase the difficulty of retrofitting, but also is more costly. In addition, the interface between old concrete and new mortar is prone to crack. Thus, retrofit method 2 is usually just used for the beams undertaken high corrosion level. To overcome the main shortcomings of the above two methods, as described below, a modified retrofit method based on substrate repairs was developed in this study. Retrofit method 1 – Directly bonding CFRP (Fig. 3). For retrofit method 1, none of the old deteriorated concrete was removed prior to CFRP repair, CFRP were directly bonded on the tensile surface of the corrosion-damaged beams. During the retrofit
250mm
20mm
30mm
plates are prone to corrosion, and special equipments are needed to install these heavy plates. In recent years, carbon fiber reinforced polymer (CFRP) materials in the form of fabrics and laminates have been motivated to use for retrofitting these corrosion-damaged RC components. CFRP materials are lightweight, noncorrosive, and exhibit high tensile strength [1]. Additionally, they can be easily bonded to the concrete surface on-site, without the use of extensive scaffolding and jacks, requiring minimum amount of support equipment [2]. CFRP systems can also be used in areas with limited access where traditional techniques would be difficult to implement. Therefore, CFRP-repair technique not only increases the ease of the retrofitting, but also is more costefficient. The advantages of retrofit technology of using CFRP externally bonded on concrete attracted the attentions of numerous researchers. Some researches have been conducted on the durability of CFRP-repaired concrete beams. In these studies, some researchers used natural corrosion methods to account for deteriorate material properties, either obtained directly from the field [3] or conditioned in the laboratory [1,4,5]. Although natural corrosion methods could closely simulate field conditions, they require a long time to achieve the desired corrosion level. To accelerate the aging process, others researchers used the application of induced electrical current [2,6–15]. By using the accelerated corrosion methods, Wang et al. [2], Bonacci and Maalej [7], Masoud and Soudki [10] experimentally studied the short-term performance of corrosiondamaged RC beams retrofitted with CFRP, Masoud and Soudki [11], Al-Hammoud et al. [13] investigated the fatigue behavior of corroded RC beams with CFRP, and Al-Hammoud et al. [14] focused on the durability of CFRP-concrete interface. However, these researchers did not investigate the effects of concrete patching before bonding CFRP. In fact, substrate repairs are often carried out on corroded RC structures to replace the damaged concrete and to restore the bond between steel and concrete. During a substrate repair, some measures such as application of corrosion inhibitors, cathodic protection and chloride extraction could be undertaken to prevent further steel corrosion [6,15]. Although numerous studies have been devoted to the performance of non-corroded RC specimens retrofitted with CFRP, not much information is available on corroded and CFRP-repaired flexural RC components, especially on the effects of combining substrate repair and CFRP repair. To get better understanding of the performance of corroded RC structures with CFRP, an extensive experimental program was conducted by the authors to evaluate the post-repair structural behavior of corrosion-damaged beams repaired with three retrofit methods. In this study, a modified retrofit method was developed based on substrate repairs, which is bonding CFRP after replacing V-notch of substrate concrete with polymer mortar. To compare the modified method with two common retrofit methods, which are respectively bonding directly CFRP and bonding CFRP after replacing damaged concrete, fourpoint bending experiments were conducted on a series of corrosion-damaged RC beams with CFRP. Additionally, important factors were considered in the experimental study, including the number of CFRP layers and corrosion level which was denoted by the mass loss rate of tensile steel. The focus in this paper is on the effective-
1400mm Fig. 1. Concrete beam-geometry and reinforcement details.
150mm
710
J.-h. Xie, R.-l. Hu / Construction and Building Materials 38 (2013) 708–716 Corrosion pool
high-water mark
Beam
Beam
Beam
Beam
Beam
Copper mesh
Pad
(a) Power supply
Steel
Copper mesh
(b) Fig. 2. Accelerated corrosion system: (a) schematic diagram of corrosion pool and (b) circuit of accelerated corrosion.
Fig. 4. Retrofit method 2: (a) removing damaged cover concrete and (b) replacing with polymer mortar.
All of the beams were subjected to four-point bending in this study. The loads were applied with a servo-hydraulic actuator controlled by force mode, and a load cell was installed on the loading actuator to record the applied loads. In addition, three dial indicators with an accuracy of 0.01 mm were used to measure the midspan deflection of the beam. Fig. 7 shows the test setup. Fig. 3. Beams retrofitted by method 1. process, the concrete surface was firstly cleaned and smoothed with the putty fillers followed by pasting the primer along the axial direction, and then the saturating resin was used to impregnate followed by bonding CFRP sheets. Retrofit method 2 – Bonding CFRP after replacing corrosion-damaged concrete with polymer mortar. For retrofit method 2, the corrosion-damaged concrete to the level of the reinforcement was removed followed by derusting, and replaced with polymer mortar (Fig. 4), then CFRP sheets were bonded on the bottom surface of new concrete cover after the polymer mortar was cured. Retrofit method 3 (A modified retrofit method) – Bonding CFRP after replacing Vnotch with polymer mortar. This modified retrofit method was based on substrate repairs. In this retrofit method, the bottom surface of concrete cover was firstly cutting by V-notch along the longitudinal corrosion cracks to the level of the main tensile steel, and then the V-notch was repaired with polymer mortar after derusting, as shown in Fig. 5. After the polymer mortar was cured, CFRP sheets were bonded on the tensile surface of the beam. Thus, comparing with method 1, the modified method prevents further steel corrosion, and comparing with method 2, it increases the ease of retrofitting. During the CFRP repair process, the test beams were bonded with 150 mm wide CFRP sheets on the tensile side, and then bonded with U-shaped CFRP strips of 80 mm wide at intervals of 80 mm near the CFRP end, as shown in Fig. 6. The main reason for using the U-shaped strip in the beams of Group C, D and E was to prevent CFRP end debonding due to serious corrosion damages. To more easily analyze the character of the test beams, all the specimens were coded according to corrosion level, retrofit method, and the number of CFRP layers, as listed in Table 1. The definitions of the code letters is that the first subscript represent retrofit method, and the second subscript represent the number of CFRP layers, and the third subscript denote the number of the beam under the same conditions. It is noted that the specimens with the subscript zero are the control beams without CFRP. 2.4. Testing procedure Prior to testing, all specimens were instrumented with electrical resistance strain gauges at the midspan. Six strain gauges were placed on the concrete surface at intervals of 50 mm along the height of the beam, two gauges were installed on the main steel bars and two gauges were installed on the surface of CFRP, as shown in Fig. 6.
3. Results and discussion The experimental results are summarized in Table 1, including yielding loads, ultimate loads and failure modes of the specimens. The results obtained from these tests are analyzed in the following. 3.1. Effects of retrofit methods 3.1.1. Failure mode A summary of the failure modes of the tested beams are listed in Table 1. The testing identified the following seven failure modes: (1) The tensile steel yields and concrete compression crushing occurs. (2) Concrete beam undergoes compressive crushing but the steel does not yield. (3) The tensile steel yields but concrete does not compression crush. (4) The tensile steel yields followed by concrete cover separating and CFRP sheets debonding. (5) The tensile steel yields followed by CFRP sheets breaking. (6) The tensile steel yields followed by CFRP sheets debonding. (7) Cracks between old concrete and new mortar followed by CFRP sheets debonding. As listed in Table 1, all the beams, which were directly bonded with one layer of CFRP sheets, underwent the process of the tensile steel yielding, with the exception of Beam A1-1-1 and Beam E1-1-1. Beam A1-1-1 experienced failure mode 2, which indicates that the amount of CFRP sheets for Beam A1-1-1 was excessive. It is noteworthy that failure mode 3, which is obvious brittle failure, occurred in
711
J.-h. Xie, R.-l. Hu / Construction and Building Materials 38 (2013) 708–716
Fig. 6. Schematic representation of corrosion-damaged beam strengthened with CFRP (units: mm).
Table 1 Summary of test conditions and results. Group Specimen Retrofit na wb rc(%) method (mm)
Yielding Ultimate Failure mode load load (kN) (kN)
A
68.6 88.0 67.2 92.5 83.0 101.1 105.5 113.5 114.3 66.0 90.8 92.3 92.5 95.3 112.1 98.2 84.6 56.2 72.0 71.3 81.3 80.0 112.3 100.9 89.8 102.2 65.1 68.2 54.5 41.3 59.8 42.2
B
C
D
E
a b c
Fig. 5. Modified retrofit method (retrofit method 3): (a) cutting V-notch, (b) part of V-notch, and (c) repairing V-notch with polymer mortar.
Beam E1-1-1. This shows that directly bonding CFRP is not suitable for the beams which had 50% mass loss of steel. The failure modes of beams retrofitted by method 1 are shown in Fig. 8. The failure mode 4 of the beams in Group B occurred near the longitudinal CFRP end, which can be explained by the lack of the end anchors, as shown in Fig. 8a. For the beams in Group C, CFRP sheets breaking
A0 A1-1-1 B0 B1-1-1 B1-1-2 B1-2-1 B1-2-2 B1-3-1 B1-3-2 C0 C1-1-1 C1-1-2 C1-2-1 C1-2-2 C1-3-1 C1-3-2 C3-1-1 D0 D1-1-1 D1-1-2 D1-2-1 D1-2-2 D1-3-1 D1-3-2 D3-1-1 D3-2-1 D2-1-2 D2-2-2 E0 E1-1-1 E2-2-1 E2-3-1
– 1 – 1 1 1 1 1 1 – 1 1 1 1 1 1 3 – 1 1 1 1 1 1 3 3 2 2 – 1 2 2
– 1 – 1 1 2 2 3 3 – 1 1 2 2 3 3 1 – 1 1 2 2 3 3 1 2 1 2 – 1 2 3
0 0 0.2 0.2 0.2 0.2 0.2 0.2 0.2 0.7 0.7 0.6 0.6 0.7 0.6 0.6 0.7 1.4 1.4 1.2 1.3 1.2 1.3 1.2 1.1 1.1 1.5 1.6 3.1 4.6 3.5 4.4
– – 3–4 3–4 3–4 3–4 3–4 3–4 3–4 8–11 8–11 8–11 8–11 8–11 8–11 8–11 8–11 18–22 18–22 18–22 18–22 18–22 18–22 18–22 18–22 18–22 18–22 18–22 50–55 70–78 50–55 70–78
102.7 150.6 99.4 140.5 118.8 150.6 138.8 154.6 160.6 93.7 135.5 145.3 145.6 135.5 152.3 145.6 132.2 82.0 125.5 132.2 128.8 135.5 148.9 138.9 145.6 172.3 92.0 98.7 78.7 55.2 77.0 58.6
1 2 3 4 4 4 4 4 4 3 5 5 4 4 4 6 5 3 6 6 6 6 6 6 5 6 7 7 3 3 3 3
n = The number of CFRP layers. w = The maximum width of corrosion cracks before loading. r = The mass loss rate of tensile steel.
occurred near the loading point, as shown in Fig. 8b. This failure mode got the maximum utilization of CFRP strength. However, for the beams in Group D, CFRP debonding occurred from midspan to end, as shown in Fig. 8c. This failure mode is mainly because that serious corrosion damages degraded the adhesion performance of CFRP-concrete interface, resulting in the initiation of interface cracks and then propagation. The failure modes of beams retrofitted by method 2 are shown in Fig. 9. As shown in Fig. 9a, the beams in Group D experienced failure mode 7. The process of failure mode 7 is that interface cracks initiate and propagate between old concrete and new polymer mortar, subsequently the concrete cover separates followed by CFRP debonding. Obviously, this failure mode decreased the efficiency of CFRP. For the beams in Group E retrofitted by method 2, the failure mode was similar as that of the beam retrofitted by method 1, as show in Fig. 9b. This can be explained by that the tensile steel of the beams in Group E had been seriously corrosiondamaged, subsequently the corrosion cracks generated, resulting in failure before CFRP sheets working. This experiment result also
712
J.-h. Xie, R.-l. Hu / Construction and Building Materials 38 (2013) 708–716
Fig. 7. Testing system. Fig. 9. Failure modes of beams retrofitted by method 2: (a) Beam D2-1-2 and (b) Beam E2-3-1.
Fig. 10. Failure mode of beams retrofitted by method 3: (a) Beam D3-1-1 and (b) Beam D3-2-1. Fig. 8. Failure modes of beams retrofitted by method 1: (a) Beam B1-1-1, (b) Beam C1-1-1, and (c) Beam D1-1-1.
indicates that the retrofit methods with CFRP are not suitable for the beams which had 50% mass loss of steel. So, in this study, retrofit method 3 did not be used for repairing the beams in Group E. The failure modes of beams retrofitted by the modified method are shown in Fig. 10. As shown in Fig. 10a, the beams with one layer of CFRP sheets in Group C and Group D experienced failure mode 5. Compared with method 1 and method 2, the modified retrofit method provided a more ideal failure mode for the beams, which could increase the utilization rate of CFRP strength. For the beams with two layers of CFRP sheets in Group D, CFRP debonding occurred followed by U-shaped strips broke, as shown in Fig. 10b. These results indicate that the CFRP thickness should be reasonably arranged to optimize to balance strength recovery with control of failure mode. 3.1.2. Load carrying capacity The ultimate loads of the retrofitted beams with one layer of CFRP sheets (except the beams in Group E) are shown in Fig. 11. It can be seen from Fig. 11, retrofit method 1 and method 3 could obviously improve the load carrying capacity of the beams in
Group B, C and D, but the strengthening effectiveness of retrofit method 2 is poor for the beams in Group D. To analyze the effects of retrofit methods on the load carrying capacity of the beams, the increase rate of the ultimate loads between the retrofitted beams with one layer CFRP and the control beam in same group is shown in Fig. 12. As can be seen in Fig. 12, retrofit method 1 averagely increased the ultimate loads of the beams in Group B by 30.4%. For the beams in Group C, retrofit method 1 and method 3 had averagely increased the ultimate loads by 49.8% and 44.1%, respectively. For the beams in Group D, the ultimate loads had an averagely increase of 57.1%, 16.3% and 93.8% by retrofit method 1, method 2 and method 3, respectively. However, for Beam E2-2-1 with two layers of CFRP in Group E, retrofit method 2 decreased the ultimate loads of the beams by 2.2%. These results indicate that, compaired with two common methods, the modified retrofit method could provide better load carrying capacity for the beams having more than 15% mass loss of steel. In addition, to improve the short-term performance, the simple method of directly bonding CFRP was suitable for the beams which had less than 15% mass loss of steel. It is noted that the beams undertaken more than 50% mass loss of steel have no need to be retrofitted by bonding CFRP sheets.
713
J.-h. Xie, R.-l. Hu / Construction and Building Materials 38 (2013) 708–716
150
160
140
140 120
120
Group B Group C Group D Group E
110 100 90
100
Load (kN)
Ultimate load (kN)
130
80 60
B0
80
B1-1-1
40
B1-1-2
70 20
60
0
50 None
1
2
3 0
Retrofit method
2
4
6
8
10
12
Midspan deflection (mm)
(a)
Fig. 11. Ultimate loads of beams retrofitted by different methods.
160 100%
120
Retrofit method 2
80%
Retrofit method 3
Load (kN)
Improvment percent
140
Retrofit method 1
60%
100 80 60
C-0 C1-1-1 C3-1-1
40% 40 20
20%
0 0%
0 B
-20%
C
D
2
E
4
6
8
10
12
14
16
18
Midspan deflection (mm)
Corrosion group
(b)
Fig. 12. Effects of retrofit methods on load carrying capacity.
160 140
3.1.4. Strain evolvement The strains variation of beam in Group B (Beam B1-3-1) is shown in Fig. 14. From Fig. 14a, it can be seen that the midspan strains of CFRP and steel have similar three-stage evolvement as the flexural
120
Load (kN)
3.1.3. Flexural stiffness The curves of loads versus midspan deflection for the retrofitted beams with one layer of CFRP sheets are shown in Fig. 13. As shown in the curves, the evolvement of flexural stiffness of the beams retrofitted with different methods had almost same trends, and the whole evolvement process during loading can be divided into three phases: (1) In the elastic phase, the relationship between the loads and the midspan deflection is linear; (2) In the cracked phase, the curves are approximately linear, but the slope of the curve is less than that in the first phase; (3) In the yielded phase, the deflection rapidly increased resulting in the failure of the beams. As shown in Fig. 13a, the flexural stiffness of the beams retrofitted by method 1 in Group B is larger than that of the non-strengthened beams during three phases, especially in the third phase. For the beams in Group C, the flexural stiffness of the beams retrofitted by method 1 has a similar increase as that of the beams retrofitted by method 3, as shown in Fig. 13b. For the beams in group D, as can be seen from Fig. 13c, retrofit method 3 obviously provided better flexural stiffness and ductility than retrofit method 1 and method 2.
100 80 60
D-0 D1-1-1 D2-1-2 D3-1-1
40 20 0 0
2
4
6
8
10
12
14
16
Midspan deflection (mm)
(c) Fig. 13. Midspan deflection of beams retrofitted with different methods: (a) Group B, (b) Group C, and (c) Group D.
stiffness, including elastic phase, cracked phase and yielded phase. In Fig. 14b, the strains distribution along the midspan cross section is shown during loading. In this figures, four strain values were measured including CFRP strains, steel strains and two concrete strains. As shown in Fig. 14b, the strains along the cross section are obviously linear distribution before cracking, and though CFRP
714
J.-h. Xie, R.-l. Hu / Construction and Building Materials 38 (2013) 708–716
3500
7000 6000
3000
Strain (με)
Strain ( με)
2000
Top Concrete Midspan CFRP 1 Midspan CFRP 2
5000
Midspan CFRP 1 Midspan CFRP 2 Steel
2500
1500
4000 3000 2000 1000
1000
0
500
-1000 0
-2000 0
20
40
60
80
100
120
140
50
80
(a)
(a)
100
120
8 kN 28 kN 48 kN 69 kN 82 kN 95 kN 109 kN 122 kN 142 kN
200 150 100 50
140
160
8 kN 28 kN 39 kN 48 kN 92 kN 98 kN 105 kN 112 kN 119 kN 125 kN 132 kN
250
0
0 -1000
60
Section height (mm)
100
40
Load (kN)
Section height (mm)
150
20
Load (kN)
250
200
0
160
-500
0
500
1000
1500
2000
2500
3000
-1000
0
1000
2000
Strain ( με)
Strain (με)
(b)
(b)
3000
4000
5000
Fig. 14. Strains variation of Beam B1-3-1: (a) load–strain curves and (b) strains distribution along midspan cross section.
Fig. 15. Strains variation of Beam C1-1-2: (a) load–strain curves and (b) strains distribution along midspan cross section.
strains lag behind other strains due to slipping between CFRP and concrete, the strains along the cross section approximately keep a linear distribution after cracking. For the beams in Group C, Fig. 15 shows the strains variation of Beam C1-1-2. It can be seen from Fig. 15a that the strain values of concrete have similar trends as that of CFRP. In Fig. 15b, the strains distribution along the midspan cross section is shown during loading, including CFRP strains, two concrete strains, and the part of steel strains. As shown in Fig. 15b, the strains along the cross section are obviously linear distribution before cracking, but after cracking, steel strains obviously lag behind other strains due to slipping between serious damaged steel and concrete, resulting in the strains along the cross section could not keep linear distribution. Thus, to simplify the calculation, it is recommended that plane section assumption is still suitable for the CFRP-retrofitted beams having less than 5% mass loss of tensile steel.
beam, the flexural stiffness of the retrofitted beams increases with the increase of CFRP layer number, but the yield point of the retrofitted beams is not easy to judge. This indicates that the increase of CFRP thickness would enhance the probability of brittle failure of the beams.
3.2. Effects of CFRP thickness 3.2.1. Flexural stiffness To analyze the effects of CFRP thickness on the flexural stiffness of the beams, the curves of loads versus midspan deflection of the beams retrofitted with different number of CFRP layer is shown in Fig. 16. As can be seen in Fig. 16, compared with that of the control
3.2.2. Load carrying capacity The ultimate loads of the corrosion-damaged RC beams retrofitted with different CFRP thickness are shown in Fig. 17. In this figure, it can be seen that the load carrying capacity of the retrofitted beams could increases with the increase of CFRP thickness. To discuss the effects of CFRP thickness on the load carrying capacity of the corrosion-damaged RC beams, the increase rate of the ultimate loads and yielding loads between the retrofitted beams and the control beam in same group are shown in Fig. 18. For the beams in Group B, as shown in Fig. 18a, the ultimate loads of the retrofitted beams with one layer of CFRP sheets have same increase rate as the yielding loads, but with the increase of CFRP thickness, the increase rate of yielding loads are greater than that of ultimate loads. In addition, it can be found that the relationship between the load carrying capacity of the beams and CFRP thickness is not linear, the increase rate of load carrying capacity decreases with the increase of CFRP thickness. For the beams retrofitted by method 1 in Group C, Fig. 18b shows that the yielding loads of beams with one layer of CFRP sheets had a 40% increase,
715
J.-h. Xie, R.-l. Hu / Construction and Building Materials 38 (2013) 708–716
160
None Corrosion Group B(Retrofit method 1)
180
140
170 160
Group D (Retrofit method 1) Group D (Retrofit method 2) Group D (Retrofit method 3)
150
100 80
Ultimate load (kN)
Load (kN)
120
Group C (Retrofit method 1)
B0 B1-1-1
60
B1-1-2 B1-2-1
40
B1-3-1
20 0 0
2
4
6
8
10
12
140 130 120 110 100 90 80 70
Midspan deflection (mm)
60
(a)
None1 160
2
3
CFRP layer
140
Fig. 17. Ultimate loads of beams with different CFRP thickness.
Load (kN)
120 100 80 60 C0 C1-1-1 C1-2-2 C1-3-1
40 20 0 0
2
4
6
8
10
12
14
16
Midspan deflection (mm)
(b) 180 160 140
those of beams with more than one layer of CFRP sheets were CFRP dedonding. For the retrofitted beams by the modified method in Group D, it can be seen from Fig. 18c that one layer of CFRP sheets increased respectively the yielding loads and the ultimate loads of the beams by 60% and 78%, and caused the failure mode of CFRP fracture, while two layers of CFRP increased respectively the yielding loads and the ultimate loads by 82% and 113%, but caused the failure mode of CFRP debonding. Thus, to enhance the utilization rate of CFRP and avoid debonding, it is important to optimize CFRP amount to balance strength recovery with control of failure mode. From Fig. 18b and c, it can be found that if using the modified method to retrofit the beams, CFRP thickness should not exceed the amount which could increase the yielding loads of the beams by more than 60%, while for the beams retrofitted by directly bonding CFRP, CFRP thickness should not exceed the amount which could increase the yielding loads of the beams by more than 40%.
Load (kN)
120
4. Conclusion
100 80 60 D0
40
D3-1-1 D3-2-1
20 0 0
2
4
6
8
10
12
14
Midspan deflection (mm)
(c) Fig. 16. Effects of CFRP thickness on midspan deflection: (a) Group B, (b) Group C (retrofit method 1), and (c) Group D (retrofit method 3).
is close to that of the beams with two layers of CFRP sheets, and is slightly less than that of the beams with three layers of CFRP sheets. This result indicates that the utilization rate of more than one layer of CFRP sheets is low, which is mainly related to the failure modes of the beams. As shown in Fig. 18b, the failure mode of beams with one layer of CFRP sheets was CFRP fracture, while
In this paper, an experimental investigation was conducted into the rehabilitation of corrosion-damaged beams with CFRP sheets. A modified retrofit method based on substrate repairs was developed to compare with two common retrofit methods in this study. In addition, important factors were also considered in these experiments, including CFRP thickness and corrosion level denoted by the mass loss rate of tensile steel. The results show that the modified retrofit method could provide better load carrying capacity and flexural stiffness for the corrosion-damaged beams having more than 15% mass loss of tensile steel. For the beams undertaken less than 15% mass loss of tensile steel, the simple retrofit method of directly bonding CFRP sheets could be suitable for improving their short-term performance. It is noted that bonded CFRP could not work for the damaged beams which had more than 50% mass loss of tensile steel. In addition, by optimizing the amount of CFRP, it is possible to balance strength recovery with control of failure mode, for example, the brittle failure mode of debonding can be avoided. Based on the experimental results, it is recommended that, when using the modified method to retrofit the corrosiondamaged RC beams, CFRP thickness should not exceed the amount which could increase the yielding loads of the beams by more than
716
J.-h. Xie, R.-l. Hu / Construction and Building Materials 38 (2013) 708–716
60%. In particular, the results indicate that plane section assumption is still suitable for the CFRP-retrofitted beams having less than 5% mass loss of tensile steel. This research is expected to contribute to guidelines for CFRP-repairing corrosion-damaged RC beams under service loads and environmental effects. As an extension of the present work, a rigorous theoretical model needs to be proposed for designers to predict the load carrying capacity of CFRP-retrofitted corroded beams.
80% Yielding load
Improvement percent
70%
Ultimate load
60% 50% 40% 30%
Acknowledgements
20% 10% 0%
1
2
3
CFRP layer
(a)
References
80% Yielding load
Improvement percent
70%
Ultimate load CFRP Debonding
60% 50%
CFRP fracture
CFRP Debonding
40% 30% 20% 10% 0%
1
2
3
CFRP layer
(b) 120%
CFRP Debonding
Improvement percent
100%
80%
CFRP fracture
60% Yielding load
40%
Ultimate load
20%
0%
The authors gratefully acknowledge the financial support provided by National Natural Science Found (No. 51208117), Foundation for Distinguished Young Talents in Higher Education of Guangdong (No. LYM10069), and Doctor Science Found of Guangdong Province (No. 10451009001004769).
1
2
CFRP layer
(c) Fig. 18. Effects of CFRP thickness on load carrying capacity: (a) Group B, (b) Group C (retrofit method 1), and (c) Group D (retrofit method 3).
[1] American Concrete Institute. Guide for the design and construction of externally bonded FRP systems for strengthening concrete structures (ACI 440). Farmington Hills: Michigan; 2008. [2] Wang CY, Shih CC, Hong SC, Hwang WC. Rehabilitation of cracked and corroded reinforced concrete beams with fiber-reinforced plastic patches. J Compos Constr 2004;8(3):219–28. [3] Aidoo J. Flexural retrofit of reinforced concrete bridge girders using three CFRP systems. Columbia: Univ. of South Carolina; 2004. [4] Toutanji HA, Gómez W. Durability characteristics of concrete beams externally bonded with FRP composite sheets. Cem Concr Compos 1997;19(4):351–8. [5] Grace NF. Concrete repair with CFRP. Concr Int 2004;26(5):45–52. [6] Ray I, Parish GC, Davalos JF, Chen An. Effect of concrete substrate repair methods for beams aged by accelerated corrosion and strengthened with CFRP. J Aerospace Eng 2011;24(2):227–39. [7] Bonacci JF, Maalej M. Externally bonded fiber-reinforced polymer for rehabilitation of corrosion damaged concrete beams. ACI Struct J 2000;97(5): 703–11. [8] Hearn N. Effect of mechanical restraint on the rate of corrosion in concrete. Can J Civil Eng 1998;25(1):81–8. [9] Maddaawy TE, Soudki K. Carbon-fiber-reinforced polymer repair to extend service life of corroded reinforced concrete beams. J Compos Constr 2005;9(2): 187–94. [10] Masoud S, Soudki K. Evaluation of corrosion activity in FRP repaired RC beams. Cem Concr Compos 2006;28(10):969–77. [11] Masoud S, Soudki K, Topper T. Postrepair fatigue performance of FRP-repaired corroded RC beams: experimental and analytical investigation. J Compos Constr 2005;9(5):441–9. [12] Soudki KA. Behavior of reinforced concrete beams strengthened with carbon fiber reinforced polymer laminates subjected to corrosion damage. Can J Civil Eng 2000;27(5):1005–10. [13] Al-Hammoud R, Soudki K, Topper TH. Fatigue flexural behavior of corroded reinforced concrete beams repaired with CFRP sheets. J Compos Constr 2011;15(1):42–51. [14] Al-Hammoud R, Soudki K, Topper TH. Bond analysis of corroded reinforced concrete beams under monotonic and fatigue loads. Cem Concr Compos 2010;32(3):194–203. [15] Malumbela G, Moyo P, Alexander M. Load-bearing capacity of corroded, patched and FRP-repaired RC beams. Mag Concr Res 2011;63(11):797–812.