Optimizations of Composite Structures

Advanced composites such as graphite/epoxy laminates have been widely applied to practical aircraft and spacecraft primary structures. The composites have also been adopted as structural components for automobiles and even for civil structures. The main advantage of the composites is its high specific strength and stiffness. However, the composite materials have strong anisotropy in that elastic modulus. In order to adopt the composites for structures effectively, optimum designs are indispensable. The optimum designs are not just designs of configurations but include designs of stacking sequences of the laminates.

In my laboratory, stacking sequence optimizations using genetic algorithms (evolutional algorithms) are conducted. Practical composite structures require stacking sequence constraints like a balance rule of number of angle plies. The genetic algorithms adopt a recessive gene like repair strategy for implements of those constraints. In order to improve design reliability, we adopted consanguineous initial population for the genetic algorithms. Moreover, to reduce the optimization cost, we adopt a response surface methodology. The response surface methodology is adopted to obtain approximated fitness function. The response surface methodology uses statistical tools to evaluate preciseness of approximations and designs of experiments to obtain low variant response with small number of analysis.

In the recent study, we are trying to apply the method to practical structures of aircraft and spacecraft.