DESIGN OPTIMIZATION OF FIBER-REINFORCED PLASTIC COMPOSITE SHAPES

A global approximation method to optimize material architecture and cr oss-sectional area of new fiber reinforced plastic (FRP) composite bea ms is presented. The sections considered are intended for applications in short-span bridges. The beams are subjected to transverse loading, and the optimization constraints include deflection limit, material f ailure, and elastic buckling. Assuming a laminated structure for the p ultruded FRP shapes, experimentally-verified micro/macromechanics mode ls are used to predict member structural behavior. The design variable s include the cross-sectional geometric dimensions and the material ar chitecture. The constraint functions are defined through a global appr oximation at a number of design points, and the approximate constraint equations are obtained through multiple linear regressions and are de fined as power law functions of the design variables. The proposed met hod can concurrently optimize the dimensions and material architecture of a given shape, and as an illustration, a new winged-box (WE) shape is optimized. The present optimization approach combined with existin g knowledge on FRP shapes can be used to develop various new shapes an d to create a new family of efficient FRP geometries for the civil str uctural market.