A nodal‐based evolutionary optimization algorithm for frame structures

This work proposes a nodal-based evolutionary design optimization algorithm to frame structures whose edges are the Delaunay triangulation of homogeneously distributed nodes in domain. The remaining can freely sway domain except for loading and boundary nodes. As result, it extend space admissible solutions this problem reduce number variables. Then, sensitivity objective function, namely, sum compliance its volume, is input into method moving asymptotes update nodal coordinates member thickness. most inefficient node deleted each iteration based on average strain energy until reaches prescribed limit. 2D numerical examples Michell arch, L-shaped bracket, Messerschmidt–Bölkow–Blohm beam show that proposed get optimal structure within few iterations. Compared with initial configurations 3D, crane arm, transmission tower, aquatic dome have less fewer materials, showing great potential actual frames.