Temporal Thermal Behavior and Damage Simulations of ... - CiteSeerX

Temporal Thermal Behavior and Damage Simulations of FRP Deck Wael I. Alnahhal1; Methee Chiewanichakorn2; Amjad J. Aref3; and Sreenivas Alampalli4 Abstract: The finite-element method 共FEM兲 has been employed to study the structural behavior of the fiber-reinforced polymer 共FRP兲 bridge deck. The numerical results were verified with the field-test results provided by New York State Department of Transportation. Fully coupled thermal-stress analyses were conducted using the FEM to predict the failure mechanisms and the “fire resistance limit” of the superstructure under extreme thermal loading conditions. Furthermore, damage simulations of the FRP deck as a result of snow and ice plowing process were performed to investigate any possibility of bridge failure after damage occurs. Thermal simulations showed that FRP bridge decks are highly sensitive to the effect of elevated temperatures. The FRP deck approached the fire resistance limit at early stages of the fire incident under all cases of fire scenarios. The damage simulations due to the snow plowing showed minimal possibility of bridge failure to take place under the worst-case damage scenario when the top 5 mm of the FRP deck surface was removed. The results of both phases of simulations provide an insight into the safety and the reliability of the FRP systems after the stipulated damage scenarios were considered. Moreover, this paper provides discussions concerning the recommended immediate actions necessary to repair the damaged region of FRP deck panels and possible use of the bridge after the damage incident. DOI: 10.1061/共ASCE兲1084-0702共2006兲11:4共452兲 CE Database subject headings: Fiber reinforced polymers; Thermal factors; Damage; Finite element method; Simulation.

Introduction The United States has one of the largest highway networks in the world. The efficiency and safety of this network plays an essential role for the continued economic health of the country. Fiber reinforced polymer 共FRP兲 composite materials are becoming more attractive to engineers due to several advantages they offer, including high strength-to-weight ratio and corrosion resistance. Because this type of advanced material has a light weight, dead load of the deck is significantly reduced after rehabilitation, which in turn increases the load rating. Recently, the first FRP deck has been installed to a state highway to improve the load rating of a 60-year-old truss bridge over Bentley Creek in Wellsburg, New York. Aref and Chiewanichakorn 共2002兲 conducted failure analysis of Bentley Creek Bridge by performing static–stress analysis up to failure. In addition, thermal-stress analysis was also performed under two scenarios, high ambient temperature gradient and burning truck 1 Research Assistant, Dept. of Civil Eng., SUNY at Buffalo, Buffalo, NY 14260. E-mail: [email protected] 2 Post-Doctoral Research Associate, Dept. of Civil Eng., SUNY at Buffalo, Buffalo, NY 14260. E-mail: [email protected] 3 Associate Professor, Dept. of Civil Eng., SUNY at Buffalo, Buffalo, NY 14260. E-mail: [email protected] 4 Director, Bridge Program and Evaluation Services Bureau, New York State Dept. of Transportation, 50 Wolf Rd., Albany, NY 12232. E-mail: [email protected] Note. Discussion open until December 1, 2006. Separate discussions must be submitted for individual papers. To extend the closing date by one month, a written request must be filed with the ASCE Managing Editor. The manuscript for this paper was submitted for review and possible publication on May 2, 2005; approved on July 26, 2005. This paper is part of the Journal of Bridge Engineering, Vol. 11, No. 4, July 1, 2006. ©ASCE, ISSN 1084-0702/2006/4-452–464/$25.00.

temperature gradient. A sequential thermal-stress analysis method was employed. A major disadvantage of this approach was that degradation of material was not taken into account. Hence, the results were too conservative, and the fire resistance limit was not anticipated in that study. A significant concern in any application of organic matrixbased composites is the possibility of an accidental fire. As compared with conventional construction materials, FRP composites show lower heat resistance. It is important to investigate the behavior of FRP decks under accidental fire, which may degrade the composite material. This may release heat and generate potentially toxic smoke. Fire threat can come from many sources at various levels of severity. On a bridge, it may be an oil truck involved in an accident resulting in an oil spill. The postfire mechanical properties of FRP composite materials decrease rapidly with increasing heat exposure time and heat flux mostly due to combustion of the vinyl-ester matrix 共Mouritz and Mathys 2001兲. Mouritz 共2003兲 proposed simple-to-use analytical models that were based on rule-of-mixtures analysis for predicting the postfire tension, compression, and flexural properties of a wide variety of burnt composite materials. Dao and Asaro 共1999兲 performed experimental and theoretical studies of compressive failure of single-skin composites under fire degradation. The experimental results indicated that most structural properties of E-glass/vinyl-ester composites used in their study were lost as temperatures approach 130°C. The effect of damage to the FRP deck due to accidental events such as fire must be investigated to take appropriate remedial measures to predict the damage and possibility of failure. There is no information available in this regard in the literature and it needs further investigation for their proper use. It is widely accepted that FRP materials cannot be used efficiently in the

452 / JOURNAL OF BRIDGE ENGINEERING © ASCE / JULY/AUGUST 2006

Fig. 1. Bentley Creek Bridge plan view

civil engineering market until rational fire safety guidelines are established 共Sorathia et al. 2001兲. This paper presents the results of a study that focused on the damage simulation of an FRP deck system. The study is divided into two phases: Phase I includes the temporal-thermal simulation of the FRP deck on a truss bridge when subjected to extreme thermal incident induced by a truck on fire; Phase II focuses on the damage simulations of the FRP deck when parts of the composite laminate are accidentally removed by snow plowing. The fire resistance limit is defined as the period of time from the initiation of fire until the moment that failure occurs at any part of the FRP deck. Furthermore, damage simulations of the FRP deck were conducted to evaluate the behavior of the damaged FRP deck panel when subjected to a snow-plowing process during the winter period.

Fiber-Reinforced Polymer Bridge Superstructure Bentley Creek Bridge is a simply supported, single-span, steeltruss bridge with an FRP deck. The bridge has a length of 42.7 m, a width of 7.3 m and a skew angle of 27°. The flooring system is comprised of transverse steel floor beams at 4.27 m spacing with longitudinal steel stringers 共see Fig. 1兲. The FRP composite deck was designed to span between the steel floor beams, which were in the transverse direction. Composite action was entirely neglected in the design of the FRP deck. However, an unintended composite action was expected due to the presence of bolt connections between the FRP deck panels and top flange of the floor beams. The panels were joined to one another using epoxy both in longitudinal and transverse directions.

Finite-Element Modeling and Verifications Finite-Element Model In this study, a pre- and postprocessing software package MSC/Patran 共MSC 1997兲 was used in modeling of Bentley Creek bridge along with ABAQUS 共Hibbitt et al. 2002兲. An entire model was composed of 370,544 elements and 462,050 nodes. The FRP “top faceskin” component was modeled by three layers for damage simulations 共Phase II兲. Fig. 2 illustrates a portion of the bridge that was modeled and analyzed in this study. The FRP deck panels can be categorized into three main parts: top faceskin, bottom faceskin, and core 共see Fig. 3兲. The

core elements were oriented in both longitudinal and transverse directions. All FRP deck elements were modeled using a solidcomposite element. Lamina thickness and fiber orientation were accurately modeled according to the information provided by the panel manufacturer. Two different types of FRP materials were used to fabricate the panels, which were generically designated as QM6408 and Q9100, as summarized in Table 1. Detailed stacking sequences of FRP deck components were taken from the report 共Aref and Chiewanichakorn 2002兲. Finite-Element Model Verifications Finite-element model 共FEM兲 verification was conducted by comparing the FEM strain values with the measured strain values from static load tests performed by Alampalli and Kunin 共2001兲. Fig. 4 shows a layout of strain gages on the FRP deck and steel girder. The FRP strain gages were mounted on the bottom faceskin underneath the FRP bridge deck, while the floor beam strain gauges were attached on the third-floor beam only. The FEM of Bentley Creek Bridge was subjected to the same axle loads from Trucks A and B as shown in Fig. 4. Static–stress analysis was performed using ABAQUS 共Hibbitt et al. 2000兲. Strain results obtained from the FEM were taken as an average of the strain values computed at the integration point of the elements. The FEM strain values were compared with the measured strain values in Fig. 5. The results confirmed that composite action between FRP deck panel and floor beam did not exist. Compressive strains at the top flange were almost identical to tensile strains at the bottom flange. Measured strains at midheight of the web were almost zero. Fig. 5 also shows that both FEM and experimental strain values agreed reasonably well, especially those that were located away from the rear axle of Truck A. As a result of this comparison, the model used in the verification process was adopted in the fully coupled thermal-stress analysis 共Phase I兲 and the damage simulations 共Phase II兲, which are independently discussed in subsequent sections of this paper.

Tsai–Hill Failure Criteria for Fiber-Reinforced Polymer Deck Hill 共1950兲 proposed a yield criterion for orthotropic materials. For plane stress in the 1–2 plane of a unidirectional lamina with fibers in the 1-direction, the governing failure criterion in terms of

JOURNAL OF BRIDGE ENGINEERING © ASCE / JULY/AUGUST 2006 / 453

Fig. 2. FRP deck geometry of Bentley Creek Bridge

Fig. 3. FRP deck compositions 454 / JOURNAL OF BRIDGE ENGINEERING © ASCE / JULY/AUGUST 2006

Fig. 4. Strain gauges and truck positions

Fig. 5. FEM verification results 共Trucks A and B兲 JOURNAL OF BRIDGE ENGINEERING © ASCE / JULY/AUGUST 2006 / 455

Table 1. Mechanical Properties of Composite Materials

Temporal Thermal Simulations „Phase I…

Properties

Value 共MPa兲

QM6408

Modulus of elasticity, E Shear modulus, G Ultimate tensile strength, Xt Ultimate compressive strength, Xc Ultimate shear strength, Xs

18,479 5,861 310 221 114

Q9100

Modulus of elasticity, E Shear modulus, G Ultimate tensile strength, Xt Ultimate compressive strength, Xc Ultimate shear strength, Xs

29,724 6,206 621 476 121

Name

the lamina principal strengths X, Y, and S can be seen in Eq. 共1兲. The appropriate values of Xt or Xc and Y t or Y c would be used depending on the signs of ␴1 and ␴2 ␴21 ␴1␴2 ␴22 ␶212 − 2 + 2 + 2 =1 X2 X Y S

共1兲

where ␴1⫽normal stress in a local 1-direction; ␴2⫽normal stress in a local 2-direction; ␶12⫽shear stress in a local 1–2 plane; X⫽lamina principal normal strength in 1-direction; Y⫽lamina principal normal strength in 2-direction; and S⫽lamina principal shear strength in 1–2 plane. A Tsai–Hill failure criterion was applied to the FRP deck panels at different ply levels. As soon as the Tsai–Hill failure index reached 1.0, a specific part and layer of the FRP deck panels was designated as the “first-ply failure.” The first-ply failure criterion provided a very conservative estimate of the laminate failure condition since first failure in the form of matrix cracks may not lead to laminate failure 共Aref and He 2001兲.

Material Modeling In order to define the fire resistance limit of the FRP deck, the material must have both thermal properties, such as conductivity, specific heat, coefficient of thermal expansion 共CTE兲, and mechanical properties, such as elasticity and density. All properties are temperature dependent. As the temperature increases, the material would become more flexible and weaker. Data available in the literature covered mechanical and thermal properties of a variety of fiber and matrix types of FRP composites at elevated temperature. Data showed that deterioration of mechanical properties at elevated temperature depends on fiber and matrix type of FRP composites. Thus, it was difficult to provide a generalization about the deterioration of mechanical properties for various types of FRP composites. To predict the mechanical properties of FRP deck during the fire, it required a simple degradation law to describe the reduction of mechanical properties with temperature for both FRP and steel materials. To achieve this degradation law, a polynomial function was fitted to a database available from the literature using the least-squares regressions analysis 共Callister 2003; Seifert et al. 2003兲. Fig. 6 illustrates a simple degradation law to describe the reduction in properties with temperature for FRP 共Seifert et al. 2003兲, and steel 共Callister 2003兲. As illustrated in Fig. 6, most structural properties of FRP composites were lost as temperature increases. In comparison to steel, FRP composites showed lower heat resistance. All values of FRP and steel thermal properties used in this study were obtained from literature because of the paucity of experimental data. Table 2 summarizes the thermal properties of steel and FRP that were utilized in the FEM. Steel has a relatively high coefficient of thermal conductivity and a relatively low specific heat. Although both specific heat and coefficient of thermal conductivity of steel and FRP change with temperature, they were assumed to be constant with temperature in this study.

Fig. 6. Stiffness-temperature curve of FRP and steel at elevated temperature 456 / JOURNAL OF BRIDGE ENGINEERING © ASCE / JULY/AUGUST 2006

Table 2. Thermal Properties of Steel and FRP Thermal property Thermal conductivity 共W/mk兲 Specific heat 共J/kgK兲 Coefficient of thermal expansion 共/°C兲

Steel

FRP

55 450 1.25⫻ 10−5

0.577 1500 14.4⫻ 10−6

Coupled Thermal-Stress Analysis Thermal-stress analyses of the FRP deck were carried out with the FEM. A fully coupled thermal-stress analysis method was required since stresses induced were temperature-distribution dependent, and temperature distributions were also dependent on stress solution. For such a case the thermal and mechanical solutions must be obtained simultaneously rather than sequentially. The FRP deck on truss bridge was exposed to a heat flux approximately equivalent to heat emitted from an oil fuel fire 共100 kW/m2兲 共Mouritz and Mathys 2001兲. Radiation and conduction loss of heat flux was not considered during the thermal analyses and the applied heat flux used in this study was assumed to be 50 kW/m2 for the period of the heating time. The initial temperature of all nodes was assumed to be zero. The FRP deck on truss bridge was subjected to thermal loading from a truck fire incident applied on the top face of the bridge deck, combined with mechanical loading from the truck axles. Fig. 7 summarizes various scenarios of a fire incident combined with mechanical loading from the commercial trucks. Results and Discussion Thermal-stress analyses of the FRP deck shows a temperature increase with time at various locations of the FRP deck under various fire scenarios. The temperature gradient increased rapidly even though the heating time was small. This was due to the high flammability and poor fire resistance of the FRP composites used in this study. The reduction to the mechanical properties with increasing heating time was determined by calculating the maximum Tsai–Hill failure index of the FRP bridge deck versus heating time.

Figs. 8–10 show the maximum Tsai–Hill failure index versus heating time for various parts of the bridge deck for Fire Scenario Case 1. Based on these figures, we can make several observations: • The fire resistance limit 共FRL兲 of the top Faceskin Layer Number 1 共top兲 of the FRP deck was approximately 450 s for Fire Scenario Case 1A, while FRL was approximately 870 s for Fire Scenario Case 1C; • After 900 s of fire burning on deck, the maximum Tsai–Hill failure index for FRP laminates at bottom faceskin is well below 1.0 for all fire scenarios—Cases 1A, B, and C; • Both Fire Scenario Cases 1A and 1B almost had similar fire resistance limit for the various parts of FRP deck, particularly because they had been subjected to identical mechanical loads; • Since the FRP deck was only subjected to one truck at the Fire Scenario Case 1C, the fire resistance limit of FRP deck under the Fire Scenario Case 1C was larger than the fire resistance limit of FRP deck under the Fire Scenario Cases 1A and B; • During the first 300 s of burning on deck, the rate of increase in thermal stresses for the top Faceskin Layer Number 1 共top兲 共TគFគ1兲 was higher than that of top Faceskin Layer Number 3 共bottom兲 共TគFគ3兲 and the longitudinal core 共LគC兲; but the top Faceskin Layer Number 3 共bottom兲 共TគFគ3兲 and longitudinal core 共LគC兲 approached the fire resistance limit before the top Faceskin Layer Number 1 共top兲 共TគFគ1兲; and • Since steel had a good fire resistance and its temperature was not high, thermal stress at floor beams and stringers could be ignored at early stages of incidental fires. Fig. 11 shows the longitudinal stress contour at various parts of the bridge deck under fire at 150 s for Fire Scenario Case 1B. The fraction of the FRP deck subjected to fire deflected upward even if the fire duration time was small 共Aref et al. 2004兲. This behavior is mainly due to two reasons: anisotropy of FRP, and a relatively low thermal conductivity and high thermal expansion coefficient of FRP that induced a significant difference in strains between the bottom face and the top exposed face 共Dutta 2004兲. Results of thermal stress analyses of the bridge deck under Fire Scenario Cases 2 and 3 show that they had the same trend of

Fig. 7. Various scenarios of fire incident on bridge deck surface JOURNAL OF BRIDGE ENGINEERING © ASCE / JULY/AUGUST 2006 / 457

Fig. 8. Maximum Tsai–Hill index versus heating time at various parts of the bridge deck for Fire Scenario Case 1A 共heat flux⫽50 kW/m2兲

Fig. 9. Maximum Tsai–Hill index versus heating time at various parts of the bridge deck for Fire Scenario Case 1B 共heat flux⫽50 kW/m2兲 458 / JOURNAL OF BRIDGE ENGINEERING © ASCE / JULY/AUGUST 2006

Fig. 10. Maximum Tsai–Hill index versus heating time at various parts of the bridge deck for Fire Scenario Case 1C 共heat flux⫽50 kW/m2兲

Fig. 11. Longitudinal stress contour at various parts of the bridge deck for Fire Scenario Case 1B at 150 s 共heat flux⫽50 kW/m2兲 JOURNAL OF BRIDGE ENGINEERING © ASCE / JULY/AUGUST 2006 / 459

Fig. 12. Maximum Tsai–Hill index versus heating time at various parts of the bridge deck for Fire Scenario Case 2A 共heat flux⫽50 kW/m2兲

Fire Scenario Case 1. Also, it was observed that the Fire Scenario Case 2 was the most critical case to the effect of elevated temperatures from fire incident 共See Fig. 12兲. Fig. 13 shows the fire resistance limit for various parts of the bridge deck for Fire Scenario Case 4, where the bottom faceskin of the FRP deck was subjected to fire incident. It was observed that it took more time to approach the fire resistance limit than the previous fire scenarios. That was because the thickness of the bottom faceskin was equal to that of the top faceskin’s three layers; meanwhile, it was subjected to less mechanical stress. From the previous results of FEM of FRP deck under fire, it can be concluded that FRP materials were very sensitive to the effect of elevated temperatures. FRP experienced degradation in strength and stiffness at early stages of the fire incident. This preliminary attempt to model a full-scale FRP deck under fire demonstrated that structural fire modeling is possible, and more importantly, when actual material degradation is available, the simulation will indicate accurately the actual time to failure due to fire.

FRP Deck Damage Simulations „Phase II… Every year, New York state experiences several heavy snow events during the winter period. In highway industry, a vital role in highway maintenance is to achieve highway-user satisfaction in terms of mobility and safety. Snow and ice reduce pavement friction and vehicle maneuverability, causing slower speeds, reduced roadway capacity, and increased crash risk. Salt was first used in the 1930s for snow and ice control to make roads safe and

passable 共Salt Institute 2004兲. It was not until the 1960s that the use of salt in conjunction with snow plowing became widespread after winter maintenance personnel learned of its effectiveness. Bridge-deck corrosion has been a problem in the snowbelt areas. Research on the subject is an ongoing process. Various anticorrosion methods have been considered. The method offering the most promise for old bridge decks is cathodic protection, where a small reverse current halts the rusting process. Another approach is applying a sealant to the concrete. Epoxy-coated reinforcing bars and air-entrained concrete and/or high-density concrete are used in the construction of new deck surfaces. FRP is an advanced material that has a light weight, high strength, high durability, and corrosion resistance. This is one of the reasons the New York State Department of Transportation 共NYSDOT兲 replaced the deteriorated concrete deck by a fully FRP Bridge on Bentley Creek Bridge. For protecting the deck, an additional wearing surface was applied to the FRP deck panels during fabrication. Portions of the wearing surface covering panel joints were applied on-site 共Alampalli and Kunin 2001兲. In extreme winter conditions, the snow-plowing process can be detrimental to a wearing surface after a certain period of time. By plowing snow and ice from a roadway, the wearing surface can be gradually removed and eventually the plow can penetrate into an FRP deck panel surface. The main purpose of damage simulations is to investigate different scenarios of having damaged FRP deck panel surface that could lead to a bridge failure. Damage simulations using FEM analysis results and recommendations are presented in the subsequent sections.

460 / JOURNAL OF BRIDGE ENGINEERING © ASCE / JULY/AUGUST 2006

Fig. 13. Maximum Tsai–Hill index versus heating time at various parts of the bridge deck for Fire Scenario Case 4 共heat flux⫽50 kW/m2兲

Damage Simulation Method In this study, FRP deck panels were assumed to be damaged from snow and ice removal processes only around the top faceskin of the FRP panel. The maximum damage penetration depth of the top faceskin in the analyses was 5 mm, or 33% of the top faceskin total thickness. In the FEM, the top faceskin was modeled using three layers of 5-mm-thick eight-node continuum element. A portion of toplayer elements of the top faceskin was removed from the entire model. Axle loads from Trucks A and B were applied on the FRP deck panel surface, both on damaged and undamaged areas. Four damage scenarios were considered. A combination of different truck configurations and all scenarios resulted in a total of ten damage simulation cases. Fig. 14 shows different cases based on damaged area 共shaded region兲 and truck configurations. Case 1 represents the worst-case situation, when all top layer elements of the top faceskin were removed from the entire bridge. Results and Discussion Static–stress analyses were performed by subjecting a damaged bridge with different truck loadings. Tsai–Hill failure indices were computed using Eq. 共1兲 based on the FEM-extracted element stress components at different ply levels. Four FRP components were considered: 1. Top faceskin; 2. Bottom faceskin; 3. Core 共longitudinal兲; and 4. Core 共transverse兲.

Fig. 15 illustrates Tsai–Hill failure indices of all cases at different ply levels through top faceskin thickness. Cases 1A and 1B exhibited the highest Tsai–Hill index 共0.637兲 at the middle layer of the top faceskin 共between 5 and 10 mm depth兲. Both of these two cases show insignificant difference in the results. Cases 2B, 4B, and 4C also exhibited similar magnitude of maximum Tsai–Hill failure index. It is obvious that the bridge would experience a high Tsai–Hill index when trucks were placed on the damaged region regardless of how many trucks were present on the bridge. On the other hand, bridges would exhibit a low Tsai–Hill failure index when trucks were placed on the undamaged region, and again, regardless of the number of trucks present on the bridge. Bottom faceskin results show an insignificant Tsai–Hill failure index value for all considered cases. Maximum Tsai–Hill failure indices for longitudinal and transverse FRP core components were 0.297 and 0.127, respectively. Fig. 16 compares Tsai–Hill failure index among different FRP deck components for different damaged cases. In the worst-case situation 共Cases 1A and 2B兲, Tsai–Hill failure index would be approximately 0.64 at the top faceskin component, meaning bridge failure would not occur. However, by imposing a 33% impact amplification factor recommended by the AASHTO code 共AASHTO 2004兲, Tsai–Hill failure index of 0.85 would be expected. Based on the results in damage simulations, it is recommended that a damaged portion of the deck should be closed and repaired as soon as the problem is detected.

JOURNAL OF BRIDGE ENGINEERING © ASCE / JULY/AUGUST 2006 / 461

Fig. 14. Damage simulation cases

Fig. 15. Tsai–Hill indices of FRP top faceskin 462 / JOURNAL OF BRIDGE ENGINEERING © ASCE / JULY/AUGUST 2006

Fig. 16. Maximum Tsai–Hill indices of FRP deck

Conclusions Thermal simulations showed that FRP bridge decks are sensitive to the effect of elevated temperatures. As compared with steel or concrete bridges, FRP bridges exhibited lower heat resistance. At the most critical fire scenario 共Case 2A兲, the top faceskin starts failure after 440 s of a burning truck on deck. Therefore, any FRP bridge under fire incident has to be vacant from people and vehicles quickly, and the damaged region of an FRP bridge should be repaired before any further use. Although the most important factors influencing the fire endurance of FRP composites were thermal conductivity of FRP, CTE of composites, type of resin, and type of glass fiber, in this study, the combined effect of these factors was not addressed. Phenolic resin had shown a greater tendency to produce delamination during fire incident according to a study done by Dodds et al. 共2000兲. More research is required to investigate the effect of these factors on fire endurance of FRP composites.

Damage simulations showed that the FRP deck would have sufficient reserve capacity at the worst-case scenario when the 33% 共5 mm兲 of the top faceskin was removed by the snow plowing process. However, by imposing the 33% impact amplification factor recommended by the code, maximum Tsai–Hill failure index of 0.64 would increase to 0.85. Therefore, it is recommended that a damaged portion of the deck should be closed and repaired as soon as the problem is detected.

Acknowledgment The work in this paper was conducted in collaboration with New York State Department of Transportation. The views presented in this document represent those of the authors and not necessarily of the NYSDOT.

JOURNAL OF BRIDGE ENGINEERING © ASCE / JULY/AUGUST 2006 / 463

References AASHTO. 共2004兲. LRFD bridge design splcifications, 3rd Ed., AASHTO, Washington, D.C. Alampalli, S., and Kunin, J. 共2001兲. “Load testing of an FRP bridge deck on a truss bridge.” Special Rep. 137, Transportation Research and Development Bureau, New York State Dept. of Transportation, Albany, New York. Aref, A. J., and Chiewanichakorn, M. 共2002兲. “The analytical study of fiber reinforced polymer deck on an old truss bridge.” Rep., Transportation Research and Development Bureau, and Transportation Infrastructure Research Consortium, New York State Dept. of Transportation, Albany, New York. Aref, A. J., Chiewanichakorn, M., and Alnahhal, W. I. 共2004兲. “Temporal thermal behavior and damage simulations of FRP deck.” Rep., Transportation Research and Development Bureau, and Transportation Infrastructure Research Consortium, New York State Dept. of Transportation, Albany, New York. Aref, A. J., and He, Y. 共2001兲. “Finite-element analysis of a fiber reinforced polymer bridge superstructure.” Rep., Transportation Research and Development Bureau, and Transportation Infrastructure Research Consortium, New York State Department of Transportation, Albany, New York. Callister, W. D. 共2003兲. Material science and engineering introduction, Wiley, New York. Dao, M., and Asaro, R. J. 共1999兲. “A study on failure prediction and design criteria for fiber composites under fire degradation.” Composites, Part A, 30共2兲, 123–131.

Dodds, N., Gibson, A. G., Dewhurst, D. J., and Davies, M. 共2000兲. “Fire behavior of composite laminates.” Composites, Part A, 31共7兲, 689–702. Dutta, P. K. 共2004兲. “Thermal measurements of the FRP composite bridge decks.” Proc., 4th Int. Conf. on Advances in Composite Materials in Bridges and Structures, Calgary, Alta., Canada, Canadian Society of Civil Engineers, Montréal. Hibbitt, Karlsson, & Sorensen. 共2002兲. ABAQUS/Standard user’s manual version 6.3, Hibbitt, Karlsson, & Sorensen, Inc., Pawtucket, R.I. Hill, R. 共1950兲. The mathematical theory of plasticity, Oxford University Press, London. MSC Software Corporation 共MSC兲. 共1997兲. MSC/Patran user’s guide version 9, MSC Software Corporation, Los Angeles. Mouritz, A. P. 共2003兲. “Simple models for determining the mechanical properties of burnt FRP composites.” Materials science and engineering, Vol. A359, Elsevier, Oxford, U.K., 237–246. Mouritz, A. P., and Mathys, Z. 共2001兲. “Post-fire mechanical properties of glass-reinforced polyester composites.” Compos. Sci. Technol., 61, 475–490. Salt Institute. 共2004兲. “Highway deicing and anti-icing for safety and mobility.” Salt Institute, 具http://www.saltinstitute.org/30.html典 共Nov. 22, 2004兲. Seifert, O. E., Schumacher, S. C., and Hansen, A. C. 共2003兲. “Viscoelastic properties of a glass fabric composite at elevated temperature: Experimental and numerical results.” Composites, Part B, 34共7兲, 571–586. Sorathia, U., Ohlemiller, T., Lyon, R., Riffle, J., and Schltz, N. 共2001兲. “Effects of fire.” Gap analysis for durability of fiber reinforced polymer composites in civil infrastructure, American Concrete Institute, Detroit, 100–120.

464 / JOURNAL OF BRIDGE ENGINEERING © ASCE / JULY/AUGUST 2006