Material-structure integrated design optimization of GFRP bridge deck on steel girder

Abstract

Design optimization of fiber-reinforced polymeric (FRP) composite products is essential to facilitate their applications in engineering structures. For bridge structures, the main design optimization goals are the reduction of FRP material consumption and the structure weight, which aim to reduce the initial construction cost and achieve a longer bridge span. Compared with conventional steel–concrete composite bridges, FRP-steel composite bridges possess more design variables and more complex design process, which necessitate the simplified optimization models. This paper aims to propose a two-scale design optimation method for FRP bridge deck on the steel girder. The macro behavior of the pultruded FRP composite bridge deck is analyzed. Regarding the micro level, the equivalent properties of pultruded GFRP lamination are calculated by combining micromechanics and classical lamination theory (CLT). The above-mentioned macro pultruded GFRP bridge level and the micro fiber/resin level were bridged based on the assumption that the micro-component effective homogenized strain equals to the corresponding macro strain. The two-scale lamination optimization of pultruded GFRP bridge deck is finally achieved by finding optimized two-scale design variables that can achieve the minimum bridge weight or the lowest initial construction cost with all listed constraint requirements satisfied. A pultruded FRP deck supported on equally-spaced steel girders was selected as a case study to show how to obtain the optimized two-scale parameters by using this proposed optimization method. The optimized results of the top flange thickness, tu, the bottom flange thickness, tl, the web height, hw, and the web thickness per meter, tw, are 46.02 mm, 45.86 mm, 300.0 mm and 37.42 mm, respectively. Results also showed that the optimized ratio of the 0°-lamina, 45°-lamina, and the 90°-lamina are77.9%, 17.1%, 5.0%. The optimized fiber volume fraction is 65.2%.