Optimization of Steel Trapezoidal Box-Girders Using Genetic Algorithm

Abstract

The goal of the structural design is to select sizes of member with an optimum proportion of the overall structural geometry in order to achieve the lowest initial cost design. Traditional steel box-girders, generally made of two plate flanges, two webs and a number of internal diaphragms, were used in many different fields, such as the structural, architectonical and bridge engineering industries. This study aims to establish a three-dimensional finite element model to minimize the total volume of a steel trapezoidal box-girders. The finite element ANSYS program package was used to determine the optimum total volume of the steel trapezoidal box- girder. ANSYS program has been coupled to MATLAB software. The genetic algorithm method of optimization is considered. In this analysis, the objective function is the total volume minimization. The design variables are the top flange width, bottom flange width, top flange thickness, bottom flange thickness, web height and web thickness. The constraints in this study include normal stress, shear stress and the displacement. The results of optimization for box-girder show a reduction in the total volume of a steel trapezoidal box-girder of about 38%.