Efficient optimization of aircraft structures with a large number ofdesign variables

Abstract

The number of variables within structural optimization procedures must be reduced to achieve acceptable low computer time. A method is presented that reduces the number of variables for a large finite element wing-box structure by representing the thickness distribution of the wing-box elements using mathematical functions. The control points of these mathematical functions are handled within the optimizer, thus greatly reducing the number of variables. The advantage of this procedure is the variational freedom achieved using functions to represent the thickness distribution of the structural elements is virtually unchanged. Element thickness of a wing box consisting of 430 finite elements is optimized using the proposed method. The number of variables is reduced to 57, and computation time is decreased by a factor of 4.2. Results indicate a very good optimization behavior. Furthermore, this method is used with the classical variable slaving or coupling method to yield an additional reduction in computation time.