Investigation Of Design Methods For Concrete Bridge Deck Reinforced With GFRP Bars

The present design methods of concrete bridge deck slabs reinforced with GFRP bars aremainly obtained by amending the design methods of concrete bridge deck slabs reinforcedwith steel reinforcement, basing on research results of experimental tests on simply supportedflexural components. However, it was found in practical engineering that structural behaviourof concrete bridge decks was underestimated by design codes, due to the existing archingaction. In order to investigate the working mechanisms of arching action in concrete bridgedeck slabs reinforced with GFRP bars, and to establish a reasonable design methodconsidering the influence of arching action on serviceability and ultimate states, a series ofexperimental study, finite element analysis and design method deduction were carried out.The main achievements were shown as follows:Simulating the behavior of a single span one-way slab strip in a global bridge deckstructure, a test device was designed to provide longitudinal and rotational restraints at eachend of test specimens. Arching action in specimens reinforced with GFRP bars occurred aftercracking, while arching action in specimens reinforced with steel bars became significant aftersteel bars yielded. The elastic modulus of GFRP bars was lower than that of steel bars, thecompatibility between GFRP and concrete was better than those of steel and concrete duringservice working stage in specimens with moderate reinforcement ratio. The low stiffness ofGFRP bars could not influence behaviors of GFRP reinforced concrete slabs compared tothose reinforced with steel bars, when the reinforcement ratio was similar and moderate. Inthe experimental tests, it was found that reinforcement configuration affected slightly onultimate bearing capacity, but determined direct the racking distribution and failure mode ofspecimens. In the test specimens of plane concrete and slabs with small reinforcementpercentages, concrete crushing in normal section at the failure loads. However, in the slabmodels with moderate to high reinforcement ratio, test specimens failed with concretecrushing in oblique section and strain of reinforcement was far less than the ultimate strain,for the depth of concrete compressive zone was increased due to arching action, and bearingcapacity in normal section was larger than that in oblique section.Nonlinear finite element models were set up to discover the law of arching action in slabstrips changed with structural parameters. It was found that the models performed well witharching action when the ratio of longitudinal restraint stiffness to longitudinal compressivestiffness of slab strips was0.2to1. With higher ratio, restraint at the ends was closer to fixedsupported, and the increased scope of influence of arching action decreased. If the ratio was lower, serviceability of the slab strips degraded. When the span or depth of the models werevaried, influence from longitudinal restraints grown with the decrease of ratio of depth to spanof the equivalent arching, otherwise, ratio of depth to span of the equivalent archingincreased with influence from longitudinal restraints decanted. When the span or depthreached moderate value, arching action worked well. Comparing the analysis of models underuniform loading, central symmetrical loading at one-fourth span and central loading atmid-span, the width of equivalent oblique strut in uniform loaded model was largest, and thearching action was most significant.Design method of concrete bridge deck slabs reinforced with GFRP bars was proposedconsidering arching action, on the base of the results of experimental researches and finiteelement analysis. After cracking, the slab strip was simplified to be a three-hinge arch system.Longitudinal restraints were regarded as compressive springs at arch feet, and tensilereinforcement was look as a tensile spring at mid-span. Critical parameters of the equivalentthree-hinge arch system were defined, and influence from arching action and reinforcementon the mechanism of slab strip were quantified. The max deflection was calculated with thevirtual work principle. Based on the code for design of concrete structures, calculation methodof the maximum crack width was proposed,in which strain of steel bars from flexuralcalculation was replaced by that calculated in the three-hinge arch system, and coefficientswere modified which reflecting the influence from bond strengths between reinforcement andconcrete, and from concrete deformation between cracks on crack widths. Further more,ultimate bearing capacity in normal section was deduced with depth of concrete compressivezone calculated in the three-hinge arch system. Development path of the oblique crack wassupposed, and ultimate bearing capacity in oblique section was derived with depth ofcompressive zone of remanent concrete in oblique section. The limit load was the smallerultimate bearing capacity. Design methods of concrete slab strips reinforced with GFRP barswere induced into the design of a global bridge deck slab reinforced with GFRP bars,combining the present research results of design method of concrete bridge deck slab withsteel reinforcement. The predictions by the proposed methods were validated by test resultswith a good collection.