Multi objective Optimization of Reinforced Concrete Buildings

Abstract

The application of parametric and topological optimization in the conception of buildings is a problem of high complexity due mainly to the large number of variables of interest to be optimized and to its nature intrinsically multiobjective. Due to the computational development occurred in the last decades, it has arisen the opportunity for a broader study and development of numeric models in this field. For the conception of structural designs, it counts on vary computational software that automate great part of the structural desgin conception process. However, in the stage of definition of the structural elements position, such as columns and beams, there is still a high level of dependency of the designer because it is long the time spent in the conception and not always the solution found is the most viable in economic and executive terms. With that, the current work is the initial development of a computational model of structural optimization of reinforced concrete buildings to decrease the designer dependency with the objective of minimizing the costs – such as concrete volume and steel weight – through the search of columns positions and its dimensions, restricted to an imposed architecture. It must be employed the evolutionary computation philosophy using the heuristic method of genetic algorithms, in the generation of the various feasible solutions, which are obtained by the results of the model of analysis by spatial framed structures, based on the finite element method. For the generation of the cost function, it will be considered the determination of the section area of the column and the steel needed that attends the equilibrium of each reinforced concrete section subjected to biaxial (oblique) bending with axial force state. Lastly, it will be performed a qualitative and quantitative comparative analysis between structural conceptions with and without the optimization technique in order to verify the consequences of its use.