Comparative study of optimum design approaches and artificial neural network based optimum design of a singly reinforced concrete beam

Abstract

The primary objective of this paper is to comparatively study the results produced by the two approaches of optimum design viz. cost minimization and weight minimization and to present a parameter (cost ratio) which demarcates the domains of their plausible application in the context of optimum design of reinforced concrete members, by considering the design instance of a singly reinforced concrete beam. The paper comprehensively studies for different combinations of concrete and steel grades and for a range of design moments, the variation of difference in cost and self weight of the beam as produced by the two approaches with respect to cost ratio i.e. ratio of unit cost of concrete to that of steel and presents the critical values of cost ratio to serve as a criteria for selection of a suitable optimum design approach. The paper also presents the application of a backpropagation neural network to model the design expertise generated using the Neural Network Toolbox of MATLAB.