a two-stage procedure for optimum design of plate girders using a meta-heuristic algorithm

The main aim in the design of welded plate girders is to minimize the weight of the beam while satisfying design requirements which turns the design procedure into a nonlinear and complex optimization problem. in this paper, a two-stage optimization procedure is introduced to find the best design of welded plate girders in terms of both safety and economy. the total weight of the girder is considered as the objective function where some predefined constraints are applied using penalty functions to restrict the solution space. in the first stage, the theoretical optimum values of girder dimensions are obtained using the artificial bee colony (abc) algorithm based on the optimum use of steel material while satisfying restrictions on the flexural and shear strength, permissible deflection, and proportioning limits. since the plate dimension values obtained by the first stage may not be available in the market, the second stage of optimization is carried out to reach a safe and economical design by considering available plate dimension values. the critical ratio of the designing procedure is obtained equal to one, which is the ideal design. to demonstrate the effectiveness of the proposed method, two examples are considered at the end of this article. the results show that the dimension of the plate girder calculated by the proposed approach is more economical and practical than that obtained by the traditional trial and error techniques.