DURABILITY CHARACTRISTICS OF NEW GFRP DOWELS FOR CONCRETE PAVEMENT

Transcription

1 DURABILITY CHARACTRISTICS OF NEW GFRP DOWELS FOR CONCRETE PAVEMENT Mathieu Montaigu 1, Mathieu Robert 2, and Brahim Benmokrane 3* 1 Master Student, Department of Civil Eng., University of Sherbrooke, Quebec, Canada 2 Postdoctoral Fellow, Department of Civil Eng., University of Sherbrooke, Quebec, Canada 3 NSERC Research Chair Professor, Department of Civil Engineering, University of Sherbrooke, Quebec, Canada * Corresponding author: Brahim.Benmokrane@usherbrooke.ca ABSTRACT Dowel bar commonly used for load transfer in concrete pavement slab is traditionally made from steel. However, once the steel dowel bar corrodes, it may cause faults, such as binding due to lockout of the dowel bar in concrete pavement, level differences resulting from spalling or decreased efficiency of load transfer. To solve this problem, many studies on the performance and feasibility of FRP dowel bars are being undertaken. In particular, glass fibre-reinforced polymer (GFRP) dowel bar is now considered as a potential solution. However, GFRP dowels could be damaged by the capillary water of concrete, which typically has high alkalinity. This study is conducted through a research collaboration project between the University of Sherbrooke and the Ministry of Transportation of Quebec (MTQ, Pavement Division). One of the main objectives of this research project was to evaluate the durability characteristics of new generation of GFRP dowel bars manufactured using vinylester resin, in simulated environmental conditions experienced by concrete pavement. An accelerated aging test method was applied to evaluate the long-term durability of the GFRP dowel bar. The results of this research have contributed to introduce this type of GFRP dowel bars with enhanced durability characteristics for concrete pavement applications in the province of Quebec. 1. INTRODUCTION FRP composites, mainly based on thermoset polymers and glass or carbon fibres, are being used in infrastructures exposed to harsh conditions involving de-icing salts or marine environments. A typical dowel bar is more vulnerable to corrosion when it is used to transfer loads in concrete pavement due to the penetration of water by the joint opening [1]. When corrosion occurs in a dowel bar, freezing could be caused by the expansion, followed by fractures due to the curling of the concrete pavement slab. Ministry of Transportation of Quebec (MTQ) has encountered these types of problem with steel reinforced jointed concrete pavement. More than 350 km of jointed concrete pavement have been rebuilt since 1994 in the province of Quebec. To minimise the corrosion problems, the MTQ is actually using epoxy-coated steel dowels of 25.1 to 1

2 38.1 mm in diameter in the design of new concrete pavement slabs of 225 to 325 mm in thickness. To solve corrosion problems of metallic dowels (conventional steel, epoxycoated, etc.), GFRP dowel bars could be considered now as a feasible solution [2, 3] with high industrial productivity [4, 5]. The MTQ is looking to improve the long-term performance, durability, and safety of jointed concrete pavement using GFRP dowels. However, long-term durability performance of GFRP dowels, such as in high alkalinity environment, needs to be investigated [6]. This study investigates the durability of GFRP dowels using accelerated ageing alkaline tests at different temperatures and period of time. GFRP specimens were immersed at different temperatures (up to 60 C.) in alkaline cement Portland solution for a period of 180 days, then tested under transverse direct shear. Based on prediction modelling, long-term transverse shear strength of the tested GFRP dowels was determined. 2. EXPERIMENTAL PROGRAM 2.1 Material GFRP dowels made of continuous high strength E-glass fibres impregnated in vinylester (VE) resin using the pultrusion process were used in this study. Durability tests were performed on several diameters from 25.4 to 44.5 mm. The initial properties (reference) of the GFRP dowels were determined by preliminary tests. As there was no effect of the diameter on the degradation of dowels (less than 2%), the durability study reported in this paper only presents results for 34.9 and 38.1 mm-diameter dowels. Note that these two diameters are the most employed diameters in roads infrastructure. The reference mechanical and physical properties of the tested 34.9 and 38.1 mm-diameter GFRP dowels are summarized in Table 1. All GFRP dowels were cut into 300 mm lengths so that the direct shear test can be performed using ACI 440.3R-04. The bars were divided into two series; 1) the unconditioned reference samples; 2) conditioned samples immersed in alkaline solution. 2.2 Conditioning of the GFRP Dowels Accelerated ageing tests were conducted according to the method DBT-2 Recommended FRP Dowel Bar Test Protocol, proposed by Market Development Alliance of the Composite Institute [7], in accordance with the ASTM C 581 standard. Cement extract solution, which is a real representative solution of concrete pores water, was prepared by mixing commercial Portland cement type 30 (highest rate of CaO) with tap water to get a ph of 12.60, measured at the beginning of the test and maintained constant throughout the conditioning period with Ca(OH) 2 and cement. Specimens were immersed at three different temperatures (23, 50 and 60 C). Specimens were place in wood containers waterproofed with a high density polyethylene film and hermetically closed to avoid excessive evaporation and changes of the ph solution (Figure 1). For each diameter and resin type of dowel, six specimens were removed from solution and tested under short beam shear tests after 30, 60 and 180 days at 23, 50 and 60 C of conditioning to measure the mechanical properties of aged dowels. 2

3 Table 1: Mechanical and physical properties of 34.9 and 38.1 mm diameter GFRP dowels Property VE-based GFRP dowels 34.9 mm 38.1 mm Fibre content (%) Cure ratio (%) Tg ( C) Moisture uptake (%) Relative density LTE long. (x10-6 C -1 ) LTE transv. (x10-6 C -1 ) Direct shear strength (MPa) Short beam shear strength (MPa) points flexural strength (MPa) Flexural modulus of elasticity (GPa) Physical Mechani Fig. 1: GFRP dowels ready for conditioning 2.3 Direct Transverse Shear Tests Direct shear tests were conducted to characterize the GFRP dowels according to ACI 440.3R-04 guide Test Method B.4. The setup consists in a 230 x 100 x 110 mm steel base equipped with lower blades spaced of 50 mm face to face, allowing the double direct shear of the specimen by an upper blade, as shown in Figure 2. Six unconditioned specimens of 300 mm long were tested as reference for each type of dowel at the laboratory conditions using MTS 810 testing machine equipped with a load cell of 500 kn. The displacement speed was 1.5 mm/min giving between 30 and 60 MPa/min, until the failure of the specimen. The direct shear strength is given by the equation 1. P s u (1) 2A Where τ u is the direct shear strength (MPa); Ps is the failure load (N); and A is the specimen section (mm²). 3

4 Fig.2: Direct shear test setup 2.4 Prediction of Long-Term Behavior Arrhenius concept [8] was used for the prediction of long-term behaviour of GFRP dowels. The equation 2 expresses the Arrhenius relation, in terms of the degradation rate [9]. Ea RT k Ae (2) Where k is the degradation rate; A is a constant relative to the material and degradation process; E a is the energy of activation of the reaction; R is the universal gas constant; and T is the temperature in C. The primary assumption of this model is that only one dominant degradation mechanism of the material operates during the reaction and that this mechanism will not change with time and temperature during the exposure. 2.5 Differential Scanning Calorimetry (DSC) Twelve-milligram to 15-milligram specimens from both unconditioned and aged samples were sealed in aluminum pans and analyzed in a TA Instruments DSC Q10 calorimeter equipped with a refrigerated cooling system. Analysis was conducted in modulated DSC mode. Specimens were heated from 25 C to 195 C at a rate of 5 o C/min. Glass transition temperature was determined by DSC for each specimen in accordance with ASTM D 1356 standard. Two scans were performed for each specimen. The first scan is useful to determine the difference of T g between reference and conditioned specimens. If a decrease of T g is observed for conditioned samples, this is an indication of plasticizing effect or chemical degradation. The second scan gives information about the mechanism of degradation and if it is irreversible. 2.6 Fourier Transform Infrared Spectroscopy (FTIR) Fourier transform infrared spectroscopy (FTIR) spectra were recorded using a Nicolet Magna-550 spectrometer equipped with an attenuated total reflectance device. Fifty 4

5 scans were routinely acquired with an optical retardation of 0.25 cm to yield a resolution of 4 cm mm diameter dowels were investigated after 180 days conditioning at 60 C. 2.7 Scanning Electronic Microscopy (SEM) Scanning electron microscopy (SEM) observations and image analysis were performed to observe the microstructure of vinylester-based dowels specimens before and after aging in alkaline solution during 180 days at 60 o C. All specimens observed in the SEM were first cut, polished, and coated with a thin layer of gold-palladium by a vapordeposit process. After coating the surfaces, microstructural observations were performed on a JEOL JSM-840A SEM. These observations were conducted to see the potential degradation of polymer matrix, glass fibers, or interfaces, if any. 3. EXPERIMENTAL RESULTS 3.1 Direct Shear Strength Table 2 shows the direct shear strength of vinylester-based dowels after different conditionings in alkaline solution. Figure 3 shows the retention of the direct shear strength after 180 days of ageing of the GFRP dowels at various temperatures. As shown in the Figure 3, vinylester-based dowels were slightly affected by ageing in alkaline solution with a retention of more than 90% of the initial shear performance after 180 days at 60 C in alkaline solution (ph=12.60). All specimens have kept elastic behaviour until the shear failure of the fibres (Figure 4). It can be seen that the degradation rate between 23 C and 50 C is nearly the same that between 50 C and 60 C, characterizing the exponential effect of the temperature on mechanisms of degradation. Table 2: Direct shear properties of aged GFRP dowels Time of immersion (days) Temperature ( o C) Mean Shear Strength (MPa) Standard deviation (MPa) 35 mm 38 mm 35 mm 38 mm

6 Fig. 3: Direct shear strength retention of GFRP dowels after conditioning in alkaline solution at 60 o C a) Fig. 4: Typical mode of failure of 34.9 mm GFRP dowels tested under shear: a) longitudinal view, b) cross-sectional view 3.2 Effect on Polymer Matrix Table 3 gives the values of T g before and after aging in alkaline solution during 180 days at 60 o C. No significant effect of aging was observed on vinylester-based dowels. It can also be seen that no shift of T g was measured during the second run, leading to the conclusion that the initial cure ratio of the vinylester resin was very high. A FTIR analysis of unconditioned dowel and specimens aged in alkaline solution during 180 days at 60 o C was conducted (Figure 5). The most interesting region of the FTIR spectra is located between 3300 cm -1 and 3600 cm -1, which corresponds to the stretching mode of the hydroxyl groups of the vinylester resin. When hydrolysis reaction occurs, new hydroxyl groups are formed and the corresponding infrared band increases. Changes in the peak intensity are quantified by determining the ratio of the OH - peak to the carbon-hydrogen stretching peak of the resin, which is not affected by the conditioning. The experimental ratios of the OH peak to the carbon-hydrogen stretching peak of the core and the surface of vinylester-based dowel immersed in alkaline solution for 180 days at 60 o C were 0.44 and 0.54, respectively, compared to 0.49 for unconditioned samples. The hydroxyl peak did not show any significant changes. This observation lead to the conclusion that no chemical degradation of the b) 6

7 vinylester resin occurred during the immersion of the dowels in alkaline solution at 60 C for 180 days. Table 3: T g of reference and aged dowels Conditioning T g run 1 ( o C) T g run 2 ( o C) 35 mm 38 mm 35 mm 38 mm Reference days in alkaline solution at 60 o C Fig. 5: FTIR spectra for unconditioned and aged samples 3.3 Microstructural Observations External Surface SEM observations of external surface of dowels were performed to investigate the surface deterioration of the polymer matrix after conditioning in alkaline solution. Figure 6 presents micrographs of the surface of reference and aged dowels. a) b) Fig.6: Micrographs of the dowels surface for: a) 34.9 mm VE reference dowel; b) 34.9 mm VE dowel aged in alkaline solution during 180 days at 60 o C 7

8 It can be observed that no degradation of the vinylester matrix has occurred after 180 days at 60 C aging in alkaline solution. No increase of the number or dimensions of the pores was observed and the surface remains intact without any cracking or microcracking. No dimension or weight changes of the dowels have been noticed after the accelerated ageing in alkaline solution Microstructural Effects Figure 7 presents micrographs for both reference specimens and specimens aged in alkaline solution during 180 days at 60 o C. The visual and microstructural observations showed no significant damage on vinylester-based dowel after 180 days of immersion in the alkaline solution at the highest temperature (60 o C). Observations of the fibre/matrix interface and of the microstructure, in general, demonstrate that the conditionings of vinylester-based dowel in alkaline solution do not affect the microstructural properties of the GFRP dowel (Figure 7b). a) b) Fig.7: Micrographs at the fibre/matrix interface before mechanical tests for: a)ve reference dowel, b) 34.9 mm VE dowel aged in alkaline solution during 180 days at 60 o C 3.4 Service Life Prediction Long-term predictions of mechanical properties can be made according to the method DBT-2, based on shear strengths obtained after 30, 60 and 180 days at three temperatures of conditioning. Simulations are proposed for the vinylester-based dowels of 34.9 mm diameter. Following the procedure proposed by Bank et al.[10], the natural logarithm of time to reach a set of levels of normalized performances versus 1/T, expressed as the inverse of absolute temperature (1000/K), was used to predict the service life at the Mean Annual Temperature (6.2 o C) in Montréal, Québec, Canada. A coefficient of determination (R 2 ) value close to 1 is desired. However, the ASTM procedures recommend a minimum value of 0.80 for acceptability and the obtained R 2 values are between 0.96 and Predictions are made for direct shear strength retention as a function of time for an immersion at 6.2 C and the general relation between the PR and the predicted service life at the average temperature of 6.2 o C are drawn (Figure 8). It can be seen from Figure 8 that vinylester-based dowels present a very high durability in concrete pavement environment. In fact, the predicted service life of vinylester-based dowels immersed in alkaline solution at an isotherm temperature of 6.2 o C to reach a PR less than 90% can be estimated to be infinite. 8

9 Fig.8: Long-term predictions for 34.9 mm dowels 4. CONCLUSIONS Based on the results of this study, the following conclusions may be: - The tested GFRP vinylester-based dowels present a very high durability after 180 days of accelerated ageing in alkaline solution at 60 C. Shear strength retention of 90 % was measured after 180 days of immersion in alkaline solution at 60 o C. - The microstructural analysis has shown that no damage has occurred to the internal microstructure or at the fibre/matrix interface for vinylester-based GFRP dowel. - According the to long-term predictions, the shear strength retention of vinylester-based GFRP dowels decreased by less than 10% after 200 years. This result clearly demonstrates the high stability of vinylester-based GFRP dowels in concrete pavement environment. For steel reinforced concrete pavement, the ACPA estimates that steel dowel bars can fail in as little as 7 to 15 years depending on design and location due to corrosion problems. 5. ACKNOWLEDGEMENTS The authors would like to express their special thanks and gratitude to the Ministry of Transportation of Quebec, the Natural Science and Engineering Research Council of Canada (NSERC), the Fonds quebecois de la recherche sur la nature et les technologies (FQRNT), Pultrall Inc. (Thetford Mines, Québec) for donating the GFRP dowels, and the technical staff of the structural lab of the Department of Civil Engineering at the University of Sherbrooke. 6. REFERENCES 1. Mauricio, M., Cruz, C. J., Jieying, Z.; Harvey, J. T., Monteiro, P. J. M.; Abdikarim, A. Laboratory evaluation of corrosion resistance of steel dowels in concrete pavements. Pavement Research Center, Institute of Transportation Studies, University of California: Davis, Berkeley,

10 2. Eddie, D., Shalabi, A., and Rizkalla, S. Glass-Fiber-Reinforced Polymer Dowels for Concrete Pavements. ACI Structural Journal, vol. 98(2), 2001, Porter, M. Assessment of dowel bar research. Final Report, CTRE, Department of civil and construction engineering, Iowa State University, Ames, Etats-Unis, 2002, 81 p. 4. Highway Innovative Technology Evaluation Center (HITEC). HITEC evaluation plan for Fiber reinforced polymer composite dowel bars and stainless dowel bars. Ohio Department of Transportation, 1998, Ohio. 5. Saad, A., Abo-Qudais., Imad, L, Al-Qadi. Dowel bars corrosion in concrete pavement. Canadian Journal of Civil Engineering, 27, 2000, Won, J., Cho, Y. et Jang, C. The Durability of Glass Fibre-Reinforced Polymer Dowel after Accelerated Environmental Exposure. Polymers & Polymer Composites, vol. 14(7), 2006, Litherland, K. L., Okley, D. R., Proctor, B. A. The use of accelerated aging procedures to predict the long term strength of GRC composites. Cement and Concrete Research, 11, 1981, Market Development Alliance (MDA). Recommended FRP Dowel Bar Durability Test Protocol. Dowel Bar Team 2 of the SPI Composites Institute, Harrison, Etats Unis, p. 9. Nelson, W., Accelerated testing Statistical models, test plans, and data analyses, Wiley, New York, Bank, L.C, Gentry, T.R, Thompson, B.P, and Russel, J.S., A Model Specification for Composites for Civil Engineering Structures, Construction and Building Materials, vol. 17(6-7), 2003,