Discrete Optimum Design of Steel Frames by Genetic Algorithm

Abstract

This paper presents genetic based algorithm for the optimum design of multistorey steel frames with sidesway subjected to multiple loading cases. The design method obtains a frame with the least weight by selecting appropriate sections for beams and columns from the British standard for universal beam and column sections. The member grouping is allowed so that the same section can be adopted for each group. The serviceability design constraints include the drift limit of overall height of the frame / 300 and interstorey drift limit of storey height / 300 as specified by BS 5950. The combined strength constraints are considered for beam-columns which are subjected to axial force and bending moments which take into consideration the lateral torsional buckling of frame members. The effective lengths for columns are obtained from the solution of nonlinear equation, which is used to produce Jackson and Moreland nomographs. The use of BS 5950 also necessitates to find out the classification of each universal beam or column section selected for frame members. The design algorithm carries out this process automatically following the steps given in BS 5950 before it computes the design capacities of each frame member. The design examples considered have shown that genetic algorithm provides an efficient tool for the practicing designers in designing tall steel frames. Keywords: Genetic algorithm, Steel frame, BS 5950. Optimum design, Minimum weight