Laminated plate design using genetic algorithms and parallel processing

Abstract

The use of a parallel computer in the stacking sequence optimization of a composite laminate via a genetic algorithm is studied. Results are obtained for a graphite-epoxy plate under buckling, strain, and ply contiguity constraints. Two parallelization schemes for the optimization problem are proposed and compared. Results are presented showing the computation time as a function of the number of processors, illustrating the benefits and drawbacks of both schemes. A new technique for design population visualization is also developed as a tool for monitoring the performance of the genetic algorithm.