Optimal design of double layer scallop domes using genetic algorithm

An efficient methodology is proposed for optimal design of large-scale domes with various topologies and dimensions in plan. The major concern with the optimal design of large domes is the difficulties arising from plurality of design variables, i.e., size and shape variables. This complexity has propounded the optimal design problem of large scale domes as a great challenge over the years. Thus, in current study, extending the novel idea of using parametric mathematical functions, design variables are correlated to the geometrical properties of domes through a new point of view. Additionally, a modified sizing approach is taken up while treating with element sections. In this way, the number of design variables is decreased. Consequently, fewer numbers of these variables provides an impressive condition that considerably takes down the computational efforts needed to explore the design space for finding the solution of optimization problem. Optimization task is performed by the robust technique of genetic algorithm. The presented approach is applicable to a wide variety of enormous domes with outsized number of nodes and members. However, to show applicability as well as computational advantages of the presented algorithm, a numerical example of scallop domes is investigated. (C) 2012 Elsevier Inc. All rights reserved.