Computational efficiency and validation of bi-directional evolutionary structural optimisation

摘要: Abstract The evolutionary structural optimisation (ESO) method has been under continuous development since 1992. Originally the was conceived from engineering perspective that topology and shape of structures were naturally conservative for safety reasons therefore contained an excess material. To move design to a more optimum would involve removal ESO algorithm caters by allowing material all parts space. With appropriate chequer-board controls on number cavities formed, can reproduce traditional fully stressed topologies. If restricted surface-only material, then problem (along lines Min–Max type problem) is solved. Recent research authors presented benchmarked additive (AESO) that, with decision making, starts procedure minimal kernel structure connects loading points mechanical constraints. Naturally this unevenly overly stressed, subsequently added surface reduce localised high stress regions. AESO only adds surface, present work describes combining basic produce bi-directional (BESO) whereby be removed. This paper shows provides same results as ESO. two benefits, it validates concept, examples demonstrate, BESO arrive at faster than especially true 3D structures, grows small initial one rather contracting a, sometimes, huge where around 90% gets removed over many hundreds finite element analysis (FEA) cycles. Both 2D are examined multiple load cases applied.