Optimización de forma de modelos paramétricos bidimensionales basada en ponderación geométrica y reglas evolutivas

This paper presents a method for shape optimization of two-dimensional models subjected to simple or multiple load cases. The optimization is performed iteratively using evolutionary rules, based on the stress level. These rules determine regions on the boundary where the material is underused or overused; the objective is that the model evolves to a minimum weight model with a high and homogeneous stress level. The evolution is performed by modifying the boundary of the model slowly. Since the boundary is defined by parametric B-spline curves, generated changes result in a smooth boundary. The main proposal of the method is that boundary modifications are given by a set of displacements that, its magnitude and direction, take into account the geometrical information of the neighborhood, unlike other methods that simply generate displacement perpendicular to the boundary. Finally, the paper presents the design of a frame bike in order to show the good performance of the method.