Genetic Algorithms in Topologic Design of Grillage Structures

Abstract

The paper describes the use of genetic algorithms in determining the optimal layout and sizing of two‐dimensional (2D) and three‐dimensional (3D) grillage structures for stress, displacement, and element buckling constraints. The design space for this problem is highly nonconvex and not readily amenable to traditional methods of nonlinear programming. The approach develops an optimal topology from a set of predefined structural universe so as to satisfy kinematic stability requirements and other constraints on structural response. A two‐level genetic algorithm–based search is used, wherein the kinematic stability constraints are imposed at one level, followed by the treatment of stress and displacement constraints at a second level of optimization. Since genetic algorithms search for an optimal design from a discrete set of alternatives in the design space, their adaptation in the topologic design problem is natural and is governed only by issues related to computational efficiency. Strategies designed to alleviate the computational requirements of a genetic algorithm–based search are discussed in the paper.