Hierarchical Solution of Large-Scale Three-Dimensional Topology Optimization Problems

Abstract

Abstract A new solution strategy for topology optimization in 3D elasticity is discussed. This solution strategy uses principles from hierarchical data structures and image analysis to reduce the computational resources necessary to solve large-scale topology optimization problems. The savings in computational resources result from successive use of increasingly detailed hierarchical models starting from a coarse approximation. These models, stored using octree data structures, are used to determine the finite element discretization at a given hierarchy. Through the use of the hierarchical models, large-scale topology optimization problems in 3D elasticity may be solved on desktop workstations.