Optimum Design of Constant-Depth Plate Girders

The iterative flange smoothing method is a computational method for determining the minimum-cost design of girders with constant web height on the basis of a given mathematical model for costs and specification for design, analysis, loading, and material. In this method, the problem is decentralized into an iterative cycle of a few simple, less dimensional operations. Each cycle consists of four major operations where each operation modifies the results of the previous one. These operations are (1) determination of the optimal arrangement of the section modulus along the girder, (2) determination of the cross section along the girder such that material cost is minimized, (3) modification of the design by balancing the cost of material versus fabrication using the smoothing method of operations research, and (4) optimization of the design variables such that the total cost is minimized. The method is found to be efficient, practical, and applicable to different cost models and specification requirements.