Application of virtual ant algorithms in the optimization of CFRP shear strengthened precracked structures

Abstract

Many engineering applications often involve the minimization of objective functions. The optimization becomes very difficult when the objective functions are either unknown or do not have an explicit form. This is certainly the case in the strengthening of existing pre-cracked reinforced concrete structures using external carbon fibre reinforced polymer (CFRP) reinforcement. For a given concrete structure, the identification of the optimum strengthening system is very important and difficult, and depends on many parameters including the extent and distribution of existing cracks, loading capacity, materials and environment. The choice of these parameters essentially forms a coupled problem of finite element analysis and parameter optimization with the aim of increasing the serviceability of the structure concerned. In this paper, virtual ant algorithms combined with nonlinear FE analysis are used in the optimization of the strengthening parameters. Simulations show that the location and orientation of the CFRP reinforcement has a significant influence on the behaviour of the strengthened structure. The orientation of the reinforcement with a fixed location becomes optimal if the reinforcing material is placed perpendicular to the existing crack direction. The implication for strengthening will also be presented. © Springer-Verlag Berlin Heidelberg 2006.