- Email: [email protected]
An experimental study of the mechanical properties of fibre reinforced polymer (FRP) and steel reinforcing bars at elevated temperatures
Composite Structures 80 (2007) 131–140 www.elsevier.com/locate/compstruct
An experimental study of the mechanical properties of fibre reinforced polymer (FRP) and steel reinforcing bars at elevated temperatures Y.C. Wang b
a,*
, P.M.H. Wong a, V. Kodur
b
a School of MACE, University of Manchester, Manchester M60 1QD, UK Department of Civil and Environmental Engineering, 3580 Engineering Building, Michigan State University, East Lansing, MI 488241226, USA
Available online 5 June 2006
Abstract This paper presents the results of an experimental study of the mechanical properties of FRP reinforcement bars, used as internal reinforcement in concrete structures, at elevated temperatures. Two types of FRP bars namely: carbon fibre reinforced polyester (CFRP) bars of 9.5 mm diameter and glass fibre reinforced polyester (GFRP) bars of 9.5 mm and 12.7 mm diameter were used in the study. For comparison, conventional steel reinforcement bars of 10 mm and 15 mm diameter were also tested. Results from the experimental study show that the stress–strain relationships of FRP bars remained almost linear at elevated temperatures until failure. However, there was a gradual reduction in the failure strength of FRP bars at elevated temperatures, at an almost linear rate to zero at about 500 C. Their elastic modulus remained almost unchanged until 300–400 C. After this temperature, there was a sharp drop in the elastic modulus. These properties can be used as input in computer programs for modelling the fire behaviour of concrete structural members reinforced with FRP bars. 2006 Elsevier Ltd. All rights reserved. Keywords: Fire resistance; Elevated temperature; Glass fibre reinforced composite; Carbon fire reinforced composite; Reinforcing bars; Mechanical properties; High temperature tests
1. Introduction An important feature of fibre reinforced polymer composites (FRP) is their extremely high corrosion resistance. This makes them suitable for use in structures subjected to severe environmental exposure. Applications for FRP bars as internal reinforcement in concrete structural members include parking garages, multi-storey buildings and industrial structures. In many of these applications provision of appropriate fire resistance is one of the major design requirements. Similar to other materials, the properties of Fiber Reinforced Polymer composite materials deteriorate when exposed to fire. One of the major concerns with using *
Corresponding author. E-mail address: [email protected] (Y.C. Wang).
0263-8223/$ - see front matter 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.compstruct.2006.04.069
FRP reinforcing bars in building construction is their early loss of strength and stiffness at elevated temperatures. There is very little information in literature on the variation of strength and stiffness of FRP with temperature [1,5,6,9]. To develop such information on strength and stiffness degradation with temperatures for FRP at elevated temperatures, a joint research program between the University of Manchester, UK, and the National Research Council of Canada (NRCC) was initiated by the authors. As part of this project, experimental studies on the tensile mechanical properties of steel and FRP bars at elevated temperatures were undertaken. Results from these experimental studies are presented in this paper to illustrate the comparative variations of tensile strength and stiffness of different types of FRP bars with traditional steel rebars. The objective of these tests is to develop material property data on the variation of strength and stiffness of FRP
132
Y.C. Wang et al. / Composite Structures 80 (2007) 131–140
reinforcement bars at elevated temperatures that can be used in computer models for evaluating fire resistance of concrete structural members reinforced with FRP rebars. 2. Test arrangement The tests were conducted in a purposely-built test rig in the structures and Fire Laboratory at the University of Manchester. This test rig consists of two parts: one for the preparation of test samples and one for loading test at elevated temperatures. 2.1. Preparation of test samples The test samples were obtained by NRCC from manufacturers [8] and were cut to 1.35 m in length and then shipped to the University of Manchester for undertaking strength tests. Each of the FRP test samples has a consistent rough surface made by embedding sand particles onto a layer of vinyl ester resin, as shown in Fig. 1. Note that the colours of the GFRP and CFRP rods are grey and black, respectively, due to the original colours of the fibres. Incorporating sand particles on the surface of the rebars is one of the methods to enhance bond between the rebars and concrete. An important consideration of testing FRP rods is to ensure that failure occurs in the test specimen, not at the anchorage. Because of the low gripping resistance of FRP rods on the surface, the design of an appropriate anchorage system is the most important issue. The ACI standard [1] does not have any specific recommendation on the anchorage system, other than giving information on how to use the test results to calculate the failure stress and the elastic modulus. Different types of anchorage systems have been tried by a large number of researchers. So far, the most reliable method is to use expansive cement confined by a circular steel tube [4,7]. When curing, chemical reactions take place in the expansive cement and pressure is generated due to confinement by the circular steel tube. If the FRP rod is
Fig. 1. Sand-incorporated surface treatment of FRP bars.
surrounded by expansive cement, the pressure in the cement will allow the FRP rod/cement interface to develop sufficient friction resistance to transfer failure from at the anchorage position at the end of the test specimen to within the specimen length. Important parameters in the design of an adequate anchorage system when using expansive cement are the length of the anchorage and the size of the steel tube. For this study, the Bristar 150 expansive cement was used. The nominal dimensions of the steel tubes were 48.3 mm in outer diameter, 3 mm in thickness and 350 mm length. The manufacturer’s literature for this expansive cement gives a maximum safe operation temperature of 20 C. However, it was found out that at temperatures of 18 and 19 C, this cement produced the blown-off phenomenon and preparation of the test samples had to be aborted until temperature in the laboratory was lower. At a cooler temperature of about 15 C, the expansive cement performed satisfactorily. Another important issue is to ensure that the FRP specimens are aligned vertically and centrally in the anchorage tubes. To do this, a wooden frame has been fabricated. Fig. 2 shows five test specimens being prepared in the wooden frame. The base of this wooden frame was fitted with 5 carefully machined steel housing caps for the anchorage steel tubes into which the expansive cement was poured. A hole with a diameter just bigger than the diameter of the test sample was drilled in the middle of each steel cap to allow the test bar to pass through. After cleaning, the anchorage steel tube was placed vertically into the steel cap. Silicone gel was then injected around the anchorage steel tube to secure it to the wooden base and to prevent leakage of the expansive cement. Once the silicone gel was hardened, a test specimen was pushed through the central hole in the steel cap. To ensure vertical alignment of the test sample, the top of the wooden frame had five holes with a diameter just slightly larger than the specimen diameters so that the specimens fitted snugly to the wooden frame. Once the expansive cement hardened in the anchorage tubes at the base of the wooden frame, the specimens were turned around for casting at the other ends. Typically, a batch of five specimens were cast on a Thursday or Friday and then tested in the following week.
Fig. 2. Wooden frame for the preparation of test samples.
Y.C. Wang et al. / Composite Structures 80 (2007) 131–140
This was necessary to allow the expansive cement to develop and maintain sufficient confinement pressure. Before commencing the loading test series, pull-out strength of the anchorage system was predicted to ensure that failure would occur in the specimen. To do this, strain gauges were attached to the anchorage steel tube to continuously record the circumferential tensile strains for one week. Fig. 3 shows the development of strains on the anchorage tube. It can be seen that the expansive cement performed consistently in different specimens and that the confinement pressure was maintained for a long period of time. Making use of Fig. 4, the confinement pressure of 1800 1600 12.7mm ∅ GFRP
Microstrain
1400 1200
15mm ∅ steel
10mm ∅ steel
1000 800 600 400
9.5mm ∅ CFRP
200 0 0
24
48
72
96
120
144
168
192
Time (hrs)
Fig. 3. Development of circumferential tensile strains in the anchorage tube.
133
Table 1 Strain of steel and confinement pressure of the expansive cement inside the anchorage tube After period of time (h)
Average microstrain (e)s
Pressure, P ¼ 2tEDs es ðN=mm2 Þ
24 48 72 96 120 144 168 192
693.52 907.09 1033.52 1128.62 1215.04 1308.03 1426.63 1549.81
18.83 24.63 28.06 30.64 32.99 35.52 38.74 42.08
the expansive cement can be calculated. Table 1 gives the average circumferential tensile strain in the anchorage steel tube and the confinement pressure. Table 2 compares the predicted pull-out strength of the anchorage and the maximum tensile strength of the various test bars. The predicted pull-out strength of the anchorage was calculated according to [4]. The predicted maximum tensile strength of the test bar was according to the manufacturer’s tensile strength. It can be seen that 48 h after casting, the expansive cement would develop sufficient confinement pressure to ensure that failure would occur in the specimen except for CFRP specimens. However, because the actual tensile strengths of the CFRP specimens were lower, the tests later revealed that there was no problem on the pull-out resistance of the anchorage system. 2.2. Test setup
PD
σs t
σs t
σ s = Es εs 2σ s t = PD 2tE s ε s P=
D
t= thickness of tube, D= inner diameter of tube Fig. 4. Calculation of confinement pressure in the expansive cement.
The elevated temperature strength tests were carried out in a specially-built test rig. This facility is shown in Fig. 5 and consists of a strong reaction frame, a hydraulic loading jack and an electrically heated kiln. Fig. 6 shows a schematic view of the test setup. The test specimen assembly, including the bar and the anchorage system, was passed through the electrically heated kiln and then clamped to the strong reaction frame at both ends through a pair of clamping brackets (shown in Fig. 5) at each end. Loads were applied from the hydraulic jack as shown in Fig. 6. According to the dimensions shown in Fig. 6, the tensile force in the specimen is twice the applied load through the hydraulic jack.
Table 2 Predicted strength of rod and anchorage pull-out Rebar
Ø (mm) b
CFRP GFRPc GFRPc Steel Steel a b c
9.5 9.5 12.7 10 15
Tensile strength (N/mm2) a
1596 689a 617a 460 460
Tensile strength (stress · pd2/4) (kN)
Pull out force, T = UL (s + lP) kN After 2 days
After 3 days
After 8 days
113.1 48.8 78.2 36.1 81.3
109.4 109.4 146.2 115.1 172.7
118.1 118.1 157.9 124.3 186.5
153.8 153.8 205.7 161.9 242.9
Ultimate tensile strength according to manufacturer (not design strength). Intermediate modulus carbon fibres vf = 0.58. Glass fibres vf = 0.75.
134
Y.C. Wang et al. / Composite Structures 80 (2007) 131–140
Fig. 5. Elevated temperature test facility.
950mm
950mm
Pin
Loading jack 200mm Kiln Displacement transducer
Specimen
Fig. 6. Schematics of test rig for high temperature strength tests.
All tests were conducted under the steady state condition. After a specimen was fixed as shown in Fig. 6, the kiln temperature was raised to the required temperature and then held constant for about half an hour to allow the test specimen to reach the same temperature. During the load test, there was no temperature measurement inside the test specimen to minimise damage to the test specimen. However, temperatures inside the kiln and on the specimen surface were continuously measured by thermocouples. In addition, a temperature monitoring study was conducted before the strength tests. Fig. 7 presents typical results and it indicates that after about half an hour, the specimen
and the kiln had reached thermal equilibrium and that the internal and surface temperatures of the specimen were very close to the target temperature. During the strength test, the target temperature was maintained and the test sample was loaded to failure. Four displacement transducers were attached to each test specimen as indicated in Fig. 6, two at each of the two locations (which were 200 mm apart) within the kiln to record the specimen axial deformations at each end. At each location, the two transducers, which were outside the kiln, were fixed, through a pair of ceramic rods, to a pair of metal brackets on the test specimen inside the kiln. Measurement of the displacement transducers started at the same time as loading and after the thermal expansion of the test specimen had occurred. Therefore, the displacement transducers measured the mechanical deformations of the test specimen, including creep but excluding the thermal expansion. However, because the loading stage was short, it is unlikely that creep was large. In all, 57 strength tests were carried out on FRP and steel specimens. Each type of specimen was tested at five different temperatures, except for the 12.7 mm diameter GFRP bars which were tested at six temperatures. These include tests on steel reinforcing bars of two different diameters for comparison. Because all materials have variability in their properties, each test was repeated at least once. To generate as much information as possible, test temperatures for the GFRP and steel bars of two different diameters were arranged so that they covered as many temperatures as possible for each type of material. Table 3 gives a summary of all the tests.
Table 3 Test parameters for reinforcing bars Specimen type (mm)
Number of tests @ (·) temperature (C)
9.5 GFRP 12.7 GFRP 9.5 CFRP 10 steel 15 steel
3 · 20, 2 · 20, 2 · 20, 2 · 20, 2 · 20,
2 · 100, 2 · 100, 2 · 100, 2 · 100, 2 · 100,
2 · 250, 2 · 200, 2 · 200, 2 · 200, 2 · 300,
160 140
Kiln temp
Temperature (˚C)
120 100 80 60
Internal temp
40
Surface temp 20 0 0:00
0:05
0:10
0:16
0:21
0:27
0:32
0:38
0:43
0:49
0:54
1:00
Time (hr:min)
Fig. 7. Temperature monitoring study for 9.5 mm a GFRP bar for target temperature 100 C.
3 · 350, 2 · 300, 4 · 400, 2 · 400, 2 · 500,
2 · 500 2 · 400, 2 · 500 2 · 600 3 · 600 2 · 700
Y.C. Wang et al. / Composite Structures 80 (2007) 131–140
3. Observations and results 1. In all tests the specimen was at almost the same temperature as the kiln when loaded to failure. 2. As intended, all the test GFRP and CFRP bars failed by fracture within the specimen length and not in the
135
anchorage. Fig. 8 illustrates the failed specimens of GFRP and CFRP reinforcing bars at different temperatures. 3. There was no fracture of the steel bar. This was because all steel bars were ductile and the test setup could not
Fig. 9. Burning of resin at 500 C, diameter 12.7 mm GFRP sample.
Table 4 Measured tensile strength and modulus of elasticity of the test specimens Specimen type
Test temperature
Strength (N/mm2)
Modulus (N/mm2)
Glass fibre
9.5
20 20 100 100 250 250 350 350 350 500 500
561.95 560.49 559.04 511.85 398.59 410.93 286.78 257.01 215.63 114.71 61.71
41.02 41.11 40.97 40.91 37.77 40.21 N/A N/A 37.24 N/A N/A
Glass fibre
12.7
20 20 20 100 100 200 200 300 300 400 400 500 500
608.56 572.81 578.50 436.72 457.84 331.50 340.44 326.62 304.69 237.66 182.41 40.62 64.59
N/A N/A 42.68 34.32 40.03 34.39 39.24 35.47 50.14 35.66 39.53 11.75 N/A
9.5
20 20 100 100 200 200 400 400 600 600
1260 1280 993.21 1283.62 702.79 763.05 371.73 387.70 132.86 9.44
121.41 126.50 105.10 132.73 107.51 107.27 N/A N/A 44.68 N/A
Carbon fibre
Fig. 8. Illustration of FRP specimens after high temperature strength tests: (a) 9.5 mm diameter CFRP rebars, (b) 9.5 mm diameter GFRP rebars, (c) 12.7 mm diameter GFRP rebars.
Diameter (mm)
N/A: result not available.
136
Y.C. Wang et al. / Composite Structures 80 (2007) 131–140
allow for sufficient extension of the test bars to break them. Nevertheless, all steel bars were loaded to the effective yield range. The effective yield stress is defined as the steel stress at 0.2% proof strain, i.e., 0.2% permanent strain. 4. The displacement transducer system performed well for most tests. It failed to record data for the smaller diameter GFRP and CFRP bars at very high temperatures. This is because the resin in these reinforcing bars burnt and the displacement clamp system collapsed. 5. In one test of GFRP reinforcing bars the resin ignited when the temperatures were about 500 C (see Fig. 9).
Table 4 gives the recorded strength and modulus of elasticity for each test specimen. The nominal test temperature was used, this temperature being very close to the actual specimen temperature. All GFRP and CFRP bars fractured and the failure strengths based on the original cross-sectional area of the bar are calculated. Also, the stress–strain relationships of GFRP and CFRP bars are almost linear until specimen fracture. Figs. 10 and 11 show examples of recorded stress–strain relationships of FRP reinforcement bars. Figs. 12 and 13 compare the normalized average strength and modulus curves of all five types of specimens.
0.8 0.7
Stress (kN/mm2)
0.6 0.5 0.4
y = 107.51x + 0.1217 R2 = 0.9926
0.3 0.2 0.1 0
-0.002
0
0.002
0.004
0.006
0.008
0.01
Axial Strain Fig. 10. Stress–strain relationship of CFRP bar at 200 C.
0.50 0.45 0.40
Stress (kN/mm2)
0.35 0.30 0.25 0.20
y = 40.026x - 0.0183 R2 = 0.9975
0.15 0.10 0.05 0.00 0.0000
0.0020
0.0040
0.0060
0.0080
0.0100
Axial Strain Fig. 11. Stress–strain relationship of 12.7 mm diameter GFRP bar at 100 C.
0.0120
0.0140
Y.C. Wang et al. / Composite Structures 80 (2007) 131–140
137
1.2
Normalized modulus
1 0.8 0.6
9.5mm CFRP Reinforcement Bar Tensile Test 9.5mm GRP Reinforcement Bar Tensile Test
0.4
10mm Steel Reinforcement Bar Tensile Test 0.2
12.7mm GRP Reinforcement Bar Tensile Test 15mm Steel Reinforcement Bar Tensile Test
0 0
100
200
300
400
500
600
700
800
Test temperature (°C) Fig. 12. Comparison of elastic modulus as a function of temperature for FRP and steel bars.
1.2
9.5mm CFRP Reinforcement Bar Tensile Test 9.5mm GRP Reinforcement Bar Tensile Test
1
10mm Steel Reinforcement Bar Tensile Test
Normalized strength
12.7mm GRP Reinforcement Bar Tensile Test 0.8
15mm Steel Reinforcement Bar Tensile Test
0.6
0.4
0.2
0 0
100
200
300
400
500
600
700
800
Test temperature (°C) Fig. 13. Comparison of tensile strength as a function of temperature for FPR and steel bars.
4. Discussions From Table 4, it can be seen that different duplicate tests on the same type of specimen at the same temperature produced similar results (<10% difference) at low temperatures (<350 C). The difference becomes greater at higher temperatures. This is clearly a result of decomposition of the resin. The fibres used to make the specimens are long fibres, but not continuous. Therefore, when the resin decomposes, the mechanical behaviour of the fibre composite becomes more influenced by the variable bond behaviour between the fibres and the decomposing resin. At temperatures of above 350 C, Figs. 12 and 13 show that both carbon and glass fibre composites still have a high level of average tensile strength and elastic modulus. However, results of individual specimens in Table 4 indicate a large variability in strengths
at the same temperature. It is doubtful whether FRP reinforcement bars should be used at such high temperatures. Fig. 12 shows that, GFRP bars suffer very little loss in tensile modulus up to about before 400 C, retaining about 90% of its ambient temperature value. After this temperature, GFRP bars suffer a drastic reduction in tensile modulus, with Fig. 12 showing less than 30% of the ambient temperature value at 500 C. Results are not available for CFRP bars at the crucial temperature of 400 C due to collapse of the displacement measuring device. At temperatures up to 200 C, the results of CFRP bars are similar to those of GFRP bars. In contrast, steel reinforcement bars experience a steady decline in the tensile modulus. From results in Fig. 13, it is observed that there is considerable difference between results for the 9.5 mm and 12.7 mm GFRP bars at low temperatures, with the 9.5 mm
138
Y.C. Wang et al. / Composite Structures 80 (2007) 131–140
giving high strengths. The higher degradation of strength in the 12.7 mm diameter bars could partly be attributed to higher resin content in these bars because at high temperatures the resin gets damaged first. The nominal diameters of the reinforcement bars were used to calculate their strengths. Comparing results for GFRP and CFRP bars of the same diameter (9.5 mm), it can be seen that reductions in CFRP strengths are greater than those in GFRP strengths. This is expected because CFRP bars have much greater strengths at ambient temperature. At elevated temperatures, the resin becomes more influential so that the absolute strengths of GFRP and CFRP bars become closer. This means that the relative reductions of CFRP bars are greater. The review of Blontrock et al. [2] indi-
cates that CFRP and GFRP composites follow an almost linear reduction in their strengths with temperature, with the strength of CFRP bars reduced to zero at about 500 C and GFRP bars reduced to zero at about 550 C. This is in agreement with the present test results. In comparison, steel bars reduce in their strengths at a lower rate than FRP composite bars. Also the variations in strength of steel reinforcing bars of the two different diameters are very similar. Fig. 14 compares the normalized tensile strength (strength at elevated temperature divided by that at ambient temperature) and the normalized elastic modulus (modulus at elevated temperature divided by that at ambient temperature) between the authors’ tests and those pro-
1.20 9.5mm CFRP 9.5mm GFRP
1.00
Normalised Tensile Stress
12.7mm GFRP Saafi's CFRP
0.80
Saafi's GFRP
0.60
0.40
0.20
0.00 0
100
200
300
(a)
400
500
600
700
Temperature (°C)
1.20
Normalised Tensile Modulus
1.00
0.80
0.60
9.5mm CFRP
0.40
9.5mm GFRP 12.7mm GFRP 0.20
Saafi's CFRP Saafi's GFRP
0.00 0
(b)
100
200
300
400
500
600
Temperature (°C)
Fig. 14. Comparison of tensile strength and modulus for FRP bars with predictions of Saafi: (a) normalised tensile strength at increasing temperatures, (b) normalised tensile modulus at increasing temperatures.
Y.C. Wang et al. / Composite Structures 80 (2007) 131–140
139
1.20 10mm Steel
Normalised Tensile Stress
1.00
15mm Steel Eurocode 4 Part 1.2
0.80
0.60
0.40
0.20
0.00 0
100
200
300
(a)
400
500
600
700
800
Temperature (°C) 1.20 10mm Steel
Normalised Tensile Modulus
1.00
15mm Steel Eurocode 4 Part 1.2
0.80
0.60
0.40
0.20
0.00 0
(b)
100
200
300
400
500
600
700
800
Temperature (°C)
Fig. 15. Comparison of 0.2% proof stress and tensile modulus for steel bars with predictions of Eurocode 4 Part 1.2: (a) normalised 0.2% proof strength at increasing temperatures, (b) normalised tensile modulus at increasing temperatures.
posed by Saafi [9] for all the FRP specimens. It indicates that Saafi’s proposal gives reasonably close lower bound values for GFRP specimens. For CFRP specimens, Saafi’s proposal gives slightly higher values of tensile strength at around 200 C. Fig. 15 compares the normalized tensile strength and elastic modulus between the authors’ tests and predictions of Eurocode 4 Part 1.2 [3] for steel reinforcing bars. It suggests that the model in Eurocode 4 Part 1.2 gives reasonably good prediction of the test results. 5. Summary The experiments were carefully and successfully conducted to determine the variations of tensile strength and
elastic modulus at elevated temperatures of steel and FRP reinforcement bars. The main conclusion is that a temperature of about 350 C appears to be critical for FRP composite bars. Below this temperature, FRP composite bars retain a very high level (90%) of their original stiffness. Also the duplicate test results are similar, being less than 10% in difference, indicating high reliability of the FRP mechanical properties. Strength reductions of different FRP bars are almost linear, being about 45% and 35% of the original ambient temperature strengths of GFRP and CFRP composite bars respectively at 350 C. At higher temperatures, some FRP composite bars can still have high strength and stiffness. However, variability in the duplicate tests is very high and it is doubtful whether the FRP reinforcement bars will have sufficient reliability in
140
Y.C. Wang et al. / Composite Structures 80 (2007) 131–140
their tensile mechanical properties at such high temperatures for practical applications.
[3]
Acknowledgements This collaborative research study was conducted when Dr. Kodur was employed at the National Research Council of Canada. The authors would also like to thank Mr. Jim Gee and Mr. Jim Gorst of the Structures and Fire Test Laboratory of the University of Manchester for preparing the test specimens and assistance with the tests.
[4]
[5]
[6]
References [1] American Concrete Institute (ACI 2001), ACI 440.1R-40, Guide for the design and construction of concrete reinforced with FRP bars. American Concrete Institute, Farmington Hills, MI. [2] Blontrock H, Taerwe L, Matthys S. Properties of fiber reinforced plastics at elevated temperatures with regard to fire resistance of reinforced concrete members. In: Dolan CW, Rizkalla SH, Nanni A, editors. Proceedings of fourth international symposium on fiber
[7]
[8] [9]
reinforced polymer reinforcement for reinforced concrete structures, 188. ACI Special Publication; 1999. p. 44–54. Committee of European Standardization (CEN 2001), ENV 1994-1-2: 2001, Eurocode 4 Design of composite steel and concrete structures, Part 1.2: Structural rules – Structural fire design, British Standard Institution. Harade T, Idemitsu T, Watanabe A, Khin M, Soeda K. New FRP tendon anchorage system using highly expansive material for anchoring. Modern Prestressing Techniques and their Applications. Proceedings, FIP’93 Symposium, Vol. II. Japan Prestressed Concrete Engineering Association; 1993. p. 711–8. Kodur VKR, Baingo D. Fire resistance of FRP reinforced concrete slabs. RC Internal Report No. 758. National Research Council of Canada, Ottawa, ON, 37, 1998. Kodur VR, Harmathy TZ. ‘‘Properties of building materials’’ SFPE handbook of fire protection engineering, 3rd ed., P.J. DiNenno, Quincy, MA: National Fire Protection Association; 2002. pp. 1.155– 81. Lees JM, Gruffydd-Jones B, Burgoyne CJ. Expansive cement couplers, A means of pre-tensioning fibre-reinforced plastic tendons. Construction and Building Materials. 1995;9(6):413–23. Pultrall Inc. Isorod guide specification. A ADS Company; 2001. www.pultrall.com. Saafi M. Effect of fire on FRP reinforced concrete members. Journal of Composite Structures. 2002;58:11–20.
An experimental study of the mechanical properties of fibre reinforced polymer (FRP) and steel reinforcing bars at elevated temperatures Y.C. Wang b
a,*
, P.M.H. Wong a, V. Kodur
b
a School of MACE, University of Manchester, Manchester M60 1QD, UK Department of Civil and Environmental Engineering, 3580 Engineering Building, Michigan State University, East Lansing, MI 488241226, USA
Available online 5 June 2006
Abstract This paper presents the results of an experimental study of the mechanical properties of FRP reinforcement bars, used as internal reinforcement in concrete structures, at elevated temperatures. Two types of FRP bars namely: carbon fibre reinforced polyester (CFRP) bars of 9.5 mm diameter and glass fibre reinforced polyester (GFRP) bars of 9.5 mm and 12.7 mm diameter were used in the study. For comparison, conventional steel reinforcement bars of 10 mm and 15 mm diameter were also tested. Results from the experimental study show that the stress–strain relationships of FRP bars remained almost linear at elevated temperatures until failure. However, there was a gradual reduction in the failure strength of FRP bars at elevated temperatures, at an almost linear rate to zero at about 500 C. Their elastic modulus remained almost unchanged until 300–400 C. After this temperature, there was a sharp drop in the elastic modulus. These properties can be used as input in computer programs for modelling the fire behaviour of concrete structural members reinforced with FRP bars. 2006 Elsevier Ltd. All rights reserved. Keywords: Fire resistance; Elevated temperature; Glass fibre reinforced composite; Carbon fire reinforced composite; Reinforcing bars; Mechanical properties; High temperature tests
1. Introduction An important feature of fibre reinforced polymer composites (FRP) is their extremely high corrosion resistance. This makes them suitable for use in structures subjected to severe environmental exposure. Applications for FRP bars as internal reinforcement in concrete structural members include parking garages, multi-storey buildings and industrial structures. In many of these applications provision of appropriate fire resistance is one of the major design requirements. Similar to other materials, the properties of Fiber Reinforced Polymer composite materials deteriorate when exposed to fire. One of the major concerns with using *
Corresponding author. E-mail address: [email protected] (Y.C. Wang).
0263-8223/$ - see front matter 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.compstruct.2006.04.069
FRP reinforcing bars in building construction is their early loss of strength and stiffness at elevated temperatures. There is very little information in literature on the variation of strength and stiffness of FRP with temperature [1,5,6,9]. To develop such information on strength and stiffness degradation with temperatures for FRP at elevated temperatures, a joint research program between the University of Manchester, UK, and the National Research Council of Canada (NRCC) was initiated by the authors. As part of this project, experimental studies on the tensile mechanical properties of steel and FRP bars at elevated temperatures were undertaken. Results from these experimental studies are presented in this paper to illustrate the comparative variations of tensile strength and stiffness of different types of FRP bars with traditional steel rebars. The objective of these tests is to develop material property data on the variation of strength and stiffness of FRP
132
Y.C. Wang et al. / Composite Structures 80 (2007) 131–140
reinforcement bars at elevated temperatures that can be used in computer models for evaluating fire resistance of concrete structural members reinforced with FRP rebars. 2. Test arrangement The tests were conducted in a purposely-built test rig in the structures and Fire Laboratory at the University of Manchester. This test rig consists of two parts: one for the preparation of test samples and one for loading test at elevated temperatures. 2.1. Preparation of test samples The test samples were obtained by NRCC from manufacturers [8] and were cut to 1.35 m in length and then shipped to the University of Manchester for undertaking strength tests. Each of the FRP test samples has a consistent rough surface made by embedding sand particles onto a layer of vinyl ester resin, as shown in Fig. 1. Note that the colours of the GFRP and CFRP rods are grey and black, respectively, due to the original colours of the fibres. Incorporating sand particles on the surface of the rebars is one of the methods to enhance bond between the rebars and concrete. An important consideration of testing FRP rods is to ensure that failure occurs in the test specimen, not at the anchorage. Because of the low gripping resistance of FRP rods on the surface, the design of an appropriate anchorage system is the most important issue. The ACI standard [1] does not have any specific recommendation on the anchorage system, other than giving information on how to use the test results to calculate the failure stress and the elastic modulus. Different types of anchorage systems have been tried by a large number of researchers. So far, the most reliable method is to use expansive cement confined by a circular steel tube [4,7]. When curing, chemical reactions take place in the expansive cement and pressure is generated due to confinement by the circular steel tube. If the FRP rod is
Fig. 1. Sand-incorporated surface treatment of FRP bars.
surrounded by expansive cement, the pressure in the cement will allow the FRP rod/cement interface to develop sufficient friction resistance to transfer failure from at the anchorage position at the end of the test specimen to within the specimen length. Important parameters in the design of an adequate anchorage system when using expansive cement are the length of the anchorage and the size of the steel tube. For this study, the Bristar 150 expansive cement was used. The nominal dimensions of the steel tubes were 48.3 mm in outer diameter, 3 mm in thickness and 350 mm length. The manufacturer’s literature for this expansive cement gives a maximum safe operation temperature of 20 C. However, it was found out that at temperatures of 18 and 19 C, this cement produced the blown-off phenomenon and preparation of the test samples had to be aborted until temperature in the laboratory was lower. At a cooler temperature of about 15 C, the expansive cement performed satisfactorily. Another important issue is to ensure that the FRP specimens are aligned vertically and centrally in the anchorage tubes. To do this, a wooden frame has been fabricated. Fig. 2 shows five test specimens being prepared in the wooden frame. The base of this wooden frame was fitted with 5 carefully machined steel housing caps for the anchorage steel tubes into which the expansive cement was poured. A hole with a diameter just bigger than the diameter of the test sample was drilled in the middle of each steel cap to allow the test bar to pass through. After cleaning, the anchorage steel tube was placed vertically into the steel cap. Silicone gel was then injected around the anchorage steel tube to secure it to the wooden base and to prevent leakage of the expansive cement. Once the silicone gel was hardened, a test specimen was pushed through the central hole in the steel cap. To ensure vertical alignment of the test sample, the top of the wooden frame had five holes with a diameter just slightly larger than the specimen diameters so that the specimens fitted snugly to the wooden frame. Once the expansive cement hardened in the anchorage tubes at the base of the wooden frame, the specimens were turned around for casting at the other ends. Typically, a batch of five specimens were cast on a Thursday or Friday and then tested in the following week.
Fig. 2. Wooden frame for the preparation of test samples.
Y.C. Wang et al. / Composite Structures 80 (2007) 131–140
This was necessary to allow the expansive cement to develop and maintain sufficient confinement pressure. Before commencing the loading test series, pull-out strength of the anchorage system was predicted to ensure that failure would occur in the specimen. To do this, strain gauges were attached to the anchorage steel tube to continuously record the circumferential tensile strains for one week. Fig. 3 shows the development of strains on the anchorage tube. It can be seen that the expansive cement performed consistently in different specimens and that the confinement pressure was maintained for a long period of time. Making use of Fig. 4, the confinement pressure of 1800 1600 12.7mm ∅ GFRP
Microstrain
1400 1200
15mm ∅ steel
10mm ∅ steel
1000 800 600 400
9.5mm ∅ CFRP
200 0 0
24
48
72
96
120
144
168
192
Time (hrs)
Fig. 3. Development of circumferential tensile strains in the anchorage tube.
133
Table 1 Strain of steel and confinement pressure of the expansive cement inside the anchorage tube After period of time (h)
Average microstrain (e)s
Pressure, P ¼ 2tEDs es ðN=mm2 Þ
24 48 72 96 120 144 168 192
693.52 907.09 1033.52 1128.62 1215.04 1308.03 1426.63 1549.81
18.83 24.63 28.06 30.64 32.99 35.52 38.74 42.08
the expansive cement can be calculated. Table 1 gives the average circumferential tensile strain in the anchorage steel tube and the confinement pressure. Table 2 compares the predicted pull-out strength of the anchorage and the maximum tensile strength of the various test bars. The predicted pull-out strength of the anchorage was calculated according to [4]. The predicted maximum tensile strength of the test bar was according to the manufacturer’s tensile strength. It can be seen that 48 h after casting, the expansive cement would develop sufficient confinement pressure to ensure that failure would occur in the specimen except for CFRP specimens. However, because the actual tensile strengths of the CFRP specimens were lower, the tests later revealed that there was no problem on the pull-out resistance of the anchorage system. 2.2. Test setup
PD
σs t
σs t
σ s = Es εs 2σ s t = PD 2tE s ε s P=
D
t= thickness of tube, D= inner diameter of tube Fig. 4. Calculation of confinement pressure in the expansive cement.
The elevated temperature strength tests were carried out in a specially-built test rig. This facility is shown in Fig. 5 and consists of a strong reaction frame, a hydraulic loading jack and an electrically heated kiln. Fig. 6 shows a schematic view of the test setup. The test specimen assembly, including the bar and the anchorage system, was passed through the electrically heated kiln and then clamped to the strong reaction frame at both ends through a pair of clamping brackets (shown in Fig. 5) at each end. Loads were applied from the hydraulic jack as shown in Fig. 6. According to the dimensions shown in Fig. 6, the tensile force in the specimen is twice the applied load through the hydraulic jack.
Table 2 Predicted strength of rod and anchorage pull-out Rebar
Ø (mm) b
CFRP GFRPc GFRPc Steel Steel a b c
9.5 9.5 12.7 10 15
Tensile strength (N/mm2) a
1596 689a 617a 460 460
Tensile strength (stress · pd2/4) (kN)
Pull out force, T = UL (s + lP) kN After 2 days
After 3 days
After 8 days
113.1 48.8 78.2 36.1 81.3
109.4 109.4 146.2 115.1 172.7
118.1 118.1 157.9 124.3 186.5
153.8 153.8 205.7 161.9 242.9
Ultimate tensile strength according to manufacturer (not design strength). Intermediate modulus carbon fibres vf = 0.58. Glass fibres vf = 0.75.
134
Y.C. Wang et al. / Composite Structures 80 (2007) 131–140
Fig. 5. Elevated temperature test facility.
950mm
950mm
Pin
Loading jack 200mm Kiln Displacement transducer
Specimen
Fig. 6. Schematics of test rig for high temperature strength tests.
All tests were conducted under the steady state condition. After a specimen was fixed as shown in Fig. 6, the kiln temperature was raised to the required temperature and then held constant for about half an hour to allow the test specimen to reach the same temperature. During the load test, there was no temperature measurement inside the test specimen to minimise damage to the test specimen. However, temperatures inside the kiln and on the specimen surface were continuously measured by thermocouples. In addition, a temperature monitoring study was conducted before the strength tests. Fig. 7 presents typical results and it indicates that after about half an hour, the specimen
and the kiln had reached thermal equilibrium and that the internal and surface temperatures of the specimen were very close to the target temperature. During the strength test, the target temperature was maintained and the test sample was loaded to failure. Four displacement transducers were attached to each test specimen as indicated in Fig. 6, two at each of the two locations (which were 200 mm apart) within the kiln to record the specimen axial deformations at each end. At each location, the two transducers, which were outside the kiln, were fixed, through a pair of ceramic rods, to a pair of metal brackets on the test specimen inside the kiln. Measurement of the displacement transducers started at the same time as loading and after the thermal expansion of the test specimen had occurred. Therefore, the displacement transducers measured the mechanical deformations of the test specimen, including creep but excluding the thermal expansion. However, because the loading stage was short, it is unlikely that creep was large. In all, 57 strength tests were carried out on FRP and steel specimens. Each type of specimen was tested at five different temperatures, except for the 12.7 mm diameter GFRP bars which were tested at six temperatures. These include tests on steel reinforcing bars of two different diameters for comparison. Because all materials have variability in their properties, each test was repeated at least once. To generate as much information as possible, test temperatures for the GFRP and steel bars of two different diameters were arranged so that they covered as many temperatures as possible for each type of material. Table 3 gives a summary of all the tests.
Table 3 Test parameters for reinforcing bars Specimen type (mm)
Number of tests @ (·) temperature (C)
9.5 GFRP 12.7 GFRP 9.5 CFRP 10 steel 15 steel
3 · 20, 2 · 20, 2 · 20, 2 · 20, 2 · 20,
2 · 100, 2 · 100, 2 · 100, 2 · 100, 2 · 100,
2 · 250, 2 · 200, 2 · 200, 2 · 200, 2 · 300,
160 140
Kiln temp
Temperature (˚C)
120 100 80 60
Internal temp
40
Surface temp 20 0 0:00
0:05
0:10
0:16
0:21
0:27
0:32
0:38
0:43
0:49
0:54
1:00
Time (hr:min)
Fig. 7. Temperature monitoring study for 9.5 mm a GFRP bar for target temperature 100 C.
3 · 350, 2 · 300, 4 · 400, 2 · 400, 2 · 500,
2 · 500 2 · 400, 2 · 500 2 · 600 3 · 600 2 · 700
Y.C. Wang et al. / Composite Structures 80 (2007) 131–140
3. Observations and results 1. In all tests the specimen was at almost the same temperature as the kiln when loaded to failure. 2. As intended, all the test GFRP and CFRP bars failed by fracture within the specimen length and not in the
135
anchorage. Fig. 8 illustrates the failed specimens of GFRP and CFRP reinforcing bars at different temperatures. 3. There was no fracture of the steel bar. This was because all steel bars were ductile and the test setup could not
Fig. 9. Burning of resin at 500 C, diameter 12.7 mm GFRP sample.
Table 4 Measured tensile strength and modulus of elasticity of the test specimens Specimen type
Test temperature
Strength (N/mm2)
Modulus (N/mm2)
Glass fibre
9.5
20 20 100 100 250 250 350 350 350 500 500
561.95 560.49 559.04 511.85 398.59 410.93 286.78 257.01 215.63 114.71 61.71
41.02 41.11 40.97 40.91 37.77 40.21 N/A N/A 37.24 N/A N/A
Glass fibre
12.7
20 20 20 100 100 200 200 300 300 400 400 500 500
608.56 572.81 578.50 436.72 457.84 331.50 340.44 326.62 304.69 237.66 182.41 40.62 64.59
N/A N/A 42.68 34.32 40.03 34.39 39.24 35.47 50.14 35.66 39.53 11.75 N/A
9.5
20 20 100 100 200 200 400 400 600 600
1260 1280 993.21 1283.62 702.79 763.05 371.73 387.70 132.86 9.44
121.41 126.50 105.10 132.73 107.51 107.27 N/A N/A 44.68 N/A
Carbon fibre
Fig. 8. Illustration of FRP specimens after high temperature strength tests: (a) 9.5 mm diameter CFRP rebars, (b) 9.5 mm diameter GFRP rebars, (c) 12.7 mm diameter GFRP rebars.
Diameter (mm)
N/A: result not available.
136
Y.C. Wang et al. / Composite Structures 80 (2007) 131–140
allow for sufficient extension of the test bars to break them. Nevertheless, all steel bars were loaded to the effective yield range. The effective yield stress is defined as the steel stress at 0.2% proof strain, i.e., 0.2% permanent strain. 4. The displacement transducer system performed well for most tests. It failed to record data for the smaller diameter GFRP and CFRP bars at very high temperatures. This is because the resin in these reinforcing bars burnt and the displacement clamp system collapsed. 5. In one test of GFRP reinforcing bars the resin ignited when the temperatures were about 500 C (see Fig. 9).
Table 4 gives the recorded strength and modulus of elasticity for each test specimen. The nominal test temperature was used, this temperature being very close to the actual specimen temperature. All GFRP and CFRP bars fractured and the failure strengths based on the original cross-sectional area of the bar are calculated. Also, the stress–strain relationships of GFRP and CFRP bars are almost linear until specimen fracture. Figs. 10 and 11 show examples of recorded stress–strain relationships of FRP reinforcement bars. Figs. 12 and 13 compare the normalized average strength and modulus curves of all five types of specimens.
0.8 0.7
Stress (kN/mm2)
0.6 0.5 0.4
y = 107.51x + 0.1217 R2 = 0.9926
0.3 0.2 0.1 0
-0.002
0
0.002
0.004
0.006
0.008
0.01
Axial Strain Fig. 10. Stress–strain relationship of CFRP bar at 200 C.
0.50 0.45 0.40
Stress (kN/mm2)
0.35 0.30 0.25 0.20
y = 40.026x - 0.0183 R2 = 0.9975
0.15 0.10 0.05 0.00 0.0000
0.0020
0.0040
0.0060
0.0080
0.0100
Axial Strain Fig. 11. Stress–strain relationship of 12.7 mm diameter GFRP bar at 100 C.
0.0120
0.0140
Y.C. Wang et al. / Composite Structures 80 (2007) 131–140
137
1.2
Normalized modulus
1 0.8 0.6
9.5mm CFRP Reinforcement Bar Tensile Test 9.5mm GRP Reinforcement Bar Tensile Test
0.4
10mm Steel Reinforcement Bar Tensile Test 0.2
12.7mm GRP Reinforcement Bar Tensile Test 15mm Steel Reinforcement Bar Tensile Test
0 0
100
200
300
400
500
600
700
800
Test temperature (°C) Fig. 12. Comparison of elastic modulus as a function of temperature for FRP and steel bars.
1.2
9.5mm CFRP Reinforcement Bar Tensile Test 9.5mm GRP Reinforcement Bar Tensile Test
1
10mm Steel Reinforcement Bar Tensile Test
Normalized strength
12.7mm GRP Reinforcement Bar Tensile Test 0.8
15mm Steel Reinforcement Bar Tensile Test
0.6
0.4
0.2
0 0
100
200
300
400
500
600
700
800
Test temperature (°C) Fig. 13. Comparison of tensile strength as a function of temperature for FPR and steel bars.
4. Discussions From Table 4, it can be seen that different duplicate tests on the same type of specimen at the same temperature produced similar results (<10% difference) at low temperatures (<350 C). The difference becomes greater at higher temperatures. This is clearly a result of decomposition of the resin. The fibres used to make the specimens are long fibres, but not continuous. Therefore, when the resin decomposes, the mechanical behaviour of the fibre composite becomes more influenced by the variable bond behaviour between the fibres and the decomposing resin. At temperatures of above 350 C, Figs. 12 and 13 show that both carbon and glass fibre composites still have a high level of average tensile strength and elastic modulus. However, results of individual specimens in Table 4 indicate a large variability in strengths
at the same temperature. It is doubtful whether FRP reinforcement bars should be used at such high temperatures. Fig. 12 shows that, GFRP bars suffer very little loss in tensile modulus up to about before 400 C, retaining about 90% of its ambient temperature value. After this temperature, GFRP bars suffer a drastic reduction in tensile modulus, with Fig. 12 showing less than 30% of the ambient temperature value at 500 C. Results are not available for CFRP bars at the crucial temperature of 400 C due to collapse of the displacement measuring device. At temperatures up to 200 C, the results of CFRP bars are similar to those of GFRP bars. In contrast, steel reinforcement bars experience a steady decline in the tensile modulus. From results in Fig. 13, it is observed that there is considerable difference between results for the 9.5 mm and 12.7 mm GFRP bars at low temperatures, with the 9.5 mm
138
Y.C. Wang et al. / Composite Structures 80 (2007) 131–140
giving high strengths. The higher degradation of strength in the 12.7 mm diameter bars could partly be attributed to higher resin content in these bars because at high temperatures the resin gets damaged first. The nominal diameters of the reinforcement bars were used to calculate their strengths. Comparing results for GFRP and CFRP bars of the same diameter (9.5 mm), it can be seen that reductions in CFRP strengths are greater than those in GFRP strengths. This is expected because CFRP bars have much greater strengths at ambient temperature. At elevated temperatures, the resin becomes more influential so that the absolute strengths of GFRP and CFRP bars become closer. This means that the relative reductions of CFRP bars are greater. The review of Blontrock et al. [2] indi-
cates that CFRP and GFRP composites follow an almost linear reduction in their strengths with temperature, with the strength of CFRP bars reduced to zero at about 500 C and GFRP bars reduced to zero at about 550 C. This is in agreement with the present test results. In comparison, steel bars reduce in their strengths at a lower rate than FRP composite bars. Also the variations in strength of steel reinforcing bars of the two different diameters are very similar. Fig. 14 compares the normalized tensile strength (strength at elevated temperature divided by that at ambient temperature) and the normalized elastic modulus (modulus at elevated temperature divided by that at ambient temperature) between the authors’ tests and those pro-
1.20 9.5mm CFRP 9.5mm GFRP
1.00
Normalised Tensile Stress
12.7mm GFRP Saafi's CFRP
0.80
Saafi's GFRP
0.60
0.40
0.20
0.00 0
100
200
300
(a)
400
500
600
700
Temperature (°C)
1.20
Normalised Tensile Modulus
1.00
0.80
0.60
9.5mm CFRP
0.40
9.5mm GFRP 12.7mm GFRP 0.20
Saafi's CFRP Saafi's GFRP
0.00 0
(b)
100
200
300
400
500
600
Temperature (°C)
Fig. 14. Comparison of tensile strength and modulus for FRP bars with predictions of Saafi: (a) normalised tensile strength at increasing temperatures, (b) normalised tensile modulus at increasing temperatures.
Y.C. Wang et al. / Composite Structures 80 (2007) 131–140
139
1.20 10mm Steel
Normalised Tensile Stress
1.00
15mm Steel Eurocode 4 Part 1.2
0.80
0.60
0.40
0.20
0.00 0
100
200
300
(a)
400
500
600
700
800
Temperature (°C) 1.20 10mm Steel
Normalised Tensile Modulus
1.00
15mm Steel Eurocode 4 Part 1.2
0.80
0.60
0.40
0.20
0.00 0
(b)
100
200
300
400
500
600
700
800
Temperature (°C)
Fig. 15. Comparison of 0.2% proof stress and tensile modulus for steel bars with predictions of Eurocode 4 Part 1.2: (a) normalised 0.2% proof strength at increasing temperatures, (b) normalised tensile modulus at increasing temperatures.
posed by Saafi [9] for all the FRP specimens. It indicates that Saafi’s proposal gives reasonably close lower bound values for GFRP specimens. For CFRP specimens, Saafi’s proposal gives slightly higher values of tensile strength at around 200 C. Fig. 15 compares the normalized tensile strength and elastic modulus between the authors’ tests and predictions of Eurocode 4 Part 1.2 [3] for steel reinforcing bars. It suggests that the model in Eurocode 4 Part 1.2 gives reasonably good prediction of the test results. 5. Summary The experiments were carefully and successfully conducted to determine the variations of tensile strength and
elastic modulus at elevated temperatures of steel and FRP reinforcement bars. The main conclusion is that a temperature of about 350 C appears to be critical for FRP composite bars. Below this temperature, FRP composite bars retain a very high level (90%) of their original stiffness. Also the duplicate test results are similar, being less than 10% in difference, indicating high reliability of the FRP mechanical properties. Strength reductions of different FRP bars are almost linear, being about 45% and 35% of the original ambient temperature strengths of GFRP and CFRP composite bars respectively at 350 C. At higher temperatures, some FRP composite bars can still have high strength and stiffness. However, variability in the duplicate tests is very high and it is doubtful whether the FRP reinforcement bars will have sufficient reliability in
140
Y.C. Wang et al. / Composite Structures 80 (2007) 131–140
their tensile mechanical properties at such high temperatures for practical applications.
[3]
Acknowledgements This collaborative research study was conducted when Dr. Kodur was employed at the National Research Council of Canada. The authors would also like to thank Mr. Jim Gee and Mr. Jim Gorst of the Structures and Fire Test Laboratory of the University of Manchester for preparing the test specimens and assistance with the tests.
[4]
[5]
[6]
References [1] American Concrete Institute (ACI 2001), ACI 440.1R-40, Guide for the design and construction of concrete reinforced with FRP bars. American Concrete Institute, Farmington Hills, MI. [2] Blontrock H, Taerwe L, Matthys S. Properties of fiber reinforced plastics at elevated temperatures with regard to fire resistance of reinforced concrete members. In: Dolan CW, Rizkalla SH, Nanni A, editors. Proceedings of fourth international symposium on fiber
[7]
[8] [9]
reinforced polymer reinforcement for reinforced concrete structures, 188. ACI Special Publication; 1999. p. 44–54. Committee of European Standardization (CEN 2001), ENV 1994-1-2: 2001, Eurocode 4 Design of composite steel and concrete structures, Part 1.2: Structural rules – Structural fire design, British Standard Institution. Harade T, Idemitsu T, Watanabe A, Khin M, Soeda K. New FRP tendon anchorage system using highly expansive material for anchoring. Modern Prestressing Techniques and their Applications. Proceedings, FIP’93 Symposium, Vol. II. Japan Prestressed Concrete Engineering Association; 1993. p. 711–8. Kodur VKR, Baingo D. Fire resistance of FRP reinforced concrete slabs. RC Internal Report No. 758. National Research Council of Canada, Ottawa, ON, 37, 1998. Kodur VR, Harmathy TZ. ‘‘Properties of building materials’’ SFPE handbook of fire protection engineering, 3rd ed., P.J. DiNenno, Quincy, MA: National Fire Protection Association; 2002. pp. 1.155– 81. Lees JM, Gruffydd-Jones B, Burgoyne CJ. Expansive cement couplers, A means of pre-tensioning fibre-reinforced plastic tendons. Construction and Building Materials. 1995;9(6):413–23. Pultrall Inc. Isorod guide specification. A ADS Company; 2001. www.pultrall.com. Saafi M. Effect of fire on FRP reinforced concrete members. Journal of Composite Structures. 2002;58:11–20.