Computerized layout (or “topology”) optimization was pioneered almost four decades back. However, despite dramatic increases in available computer power and the application of increasingly efficient optimization algorithms, even now only relatively modest sized problems can be tackled using the traditional “ground structure” approach. This is because of the need, in general, for the latter to contain every conceivable member connecting together the nodes in a problem. A simple, but effective solution method capable of tackling problems with large numbers of potential members (e.g. >100,000,000) is presented. Though the method draws on the linear programming technique of “column generation”, since layout optimization specific heuristics are employed it is presented as an iterative “member adding” method. The method requires a ground structure with minimal connectivity to be used in the first iteration; members are then added as required in subsequent iterations until the (provably) optimal solution is found.

Layout optimization of large‐scale pin‐jointed frames

An efficient structural optimization methodology is presented for the design of minimum weight space frames subject to multiple natural frequency constraints. A powerful class of generalized hybrid c onstraint approximations which require o nly the first order constraint function d erivatives have been developed to overcome inherent nonlinearity of the frequency constraint. The generalized hybrid constraint functions are shown to be relatively conservative, separable and convex in the region bounded by the move limits based on the formula described in this paper. The optimization methodology is implemented in an automated structural optimization system which has been applied to solve a variety of space frame optimization problems. N umerical results obtained for three example problems indicate that the o ptimization methodology requires fewer than 10 complete normal modes analyses to generate a near optimum solution.

Space frame optimization subject to frequency constraints

A multilevel method for structural synthesis based on finite element modeling at both the system and the substructure level is set forth. At the system level, the design is described using basis vectors generated via substructure design modification. Whenever possible, substructure constraints are satisfied via local material redistribution. At both the system level and the substructure level, design modification is carried out by solving a sequence of approximate problems. The method generates upper bound approximations for the optimum design and it has potential for substantial gains in computational efficiency, because the uncoupled substructure design modification problems can be processed in parallel.

A multilevel method for structural synthesis

A decomposition procedure is proposed in this paper for solving a class of large-scale optimum design problems for perfectly-plastic structures under several alternative loading conditions. The conventional finite element method is used to cast the problem into a finite dimensional constrained nonlinear programming problem. Structures of practically meaningful size and complexity tend to give rise to a large number of variables and constraints in the corresponding mathematical model. The difficulty is that the state-of-the-art mathematical programming theory does not provide reliable and efficient ways of solving large-scale constrained nonlinear programming problems. The natural idea to deal with the large-scale structural problem is somehow to decompose the problem into a collection of small-size problems each of which represents an analysis of the behaviour of each finite element under a single loading condition. This paper proposes one such way of decomposition based on duality theory and a recently developed iterative algorithm called the proximal point algorithm.

L.A. Schmit

Papers

Decomposition in optimal plastic design of structures