Configuration Optimization of Trusses

A minimum weight design of trusses is considered by including sizing and configuration variables simultaneously in the design space. Second-order Taylor series expansions of displacement quantities are used along a search direction in order to avoid large numbers of complete structural analyses. Disjoint feasible regions are recognized. Side constraints on nodal coordinates are used to avoid analyses for structures with statically unstable configurations. A primal-dual method, called the multiplier method, is employed as an optimization tool.