Simultaneous outline shape and size optimization for stiffeners in practical engineering structures

Abstract

This article presents an engineering method to optimize the stiffener outline shape and cross-sectional sizes simultaneously for stiffened plate structures. A new optimization model for stiffened plate is established, and a hierarchical algorithm including external and internal optimization module is proposed to tackle the shape and size variables respectively. In external optimization, genetic algorithm (GA) is utilized to optimize binary-coded shape variables, and when calculating individual fitness values in GA, an internal optimization only involving size variables is carried out. For each individual, the stiffener outline shape is determined by a given shape variable vector and a reasonable finite element model could be automatically built with a parametric modeling method. The solution of internal optimization problem can be obtained by size optimization method, and an offset vector which updates during each cycle is created to modify the relative position of plate and stiffeners, ensuring the practicality of the optimal results in engineering. The presented approach is firstly verified through a typical example and then it is applied to the structural design of a micro-satellite. The results show that the method is effective in dealing with the simultaneous outline shape and size optimization problem of stiffeners, and the optimal results could be directly applied in practical engineering structures.