An efficient SLP code for design of large scale structures

Design optimisation of complex structures may become computationally very expensive if it entails many structural analyses. Therefore, optimisation techniques are attractive if they are capable to obtain the optimum design in few optimisation cycles. In practical design, Sequential Linear Programming (SLP) is very popular because of its inherent simplicity. This paper presents sizing and configuration problems of large-scale truss structures optimised under multiple loading conditions with up to 200 design variables and 3500 constraints. The relative performance and merits of some SLP algorithms well referenced in literature are compared. Besides, the efficiency of an advanced SLP based algorithm called LESLP is tested. The numerical performance of LESLP is also compared to that of the Sequential Quadratic Programming method (SQP), universally considered a powerful algorithm and theoretically better founded than SLP. The numerical results proved the high efficiency of the LESLP code in solving large-scale problems