Genetic algorithm based optimum design of prestressed concrete beam

Abstract

Design of prestressed concrete (PSC) beam involves many design variables. Simultaneously handling many design variables computationally expensive and in most cases finds relative optima. An efficient optimization technique, Genetic Algorithm (GA) is well suited for solving such problems and they can find global optimum solution. In the present paper six design variables are considered for optimizing the cost of simply supported prestressed concrete beam with all practical constraints. Since absolute cost of materials changes from time to time, total cost of the beam is expressed in terms of cost ratio. Parametric study has been done to observe the effect of size of population, number of generation, intensity of live load, cable profile and various practical beam lengths. Finally data base is prepared for various practical span lengths. Keyword: Genetic algorithm, cost ratio, prestressed concrete beam. Nomenclature P = Prestrssed force e= eccentricity B = beam width D = depth of beam Dc = diameter of the cable Aco = Area of the concrete �co = density of concrete Cr co = cost ratio of the concrete ft = stress in top fiber