Ajuste de modelos de cortante en hormigón estructural mediante algoritmos evolutivos

The behavior of reinforced and prestressed concrete members subjected to shear has been traditionally studied by the so-called Compression-Field Theories. In the last years several authors have modified the mechanical model of such theories in order to include the tension-stiffening effect of the concrete in the steel constitutive model; the resulting system of equations is highly non-linear and involves a great number of unknowns. From an experimental database of RC beams, this work performs an adjustment of the shear model proposed by one of these theories using evolutionary algorithms. Evolutionary algorithm is a branch of Natural Computing. Its computations start from an initial population of individuals (randomly generated) and proceed according to rules of selection and other operators, such as recombination and mutation. The main objective of such algorithms is to find good solutions in a reasonable time-frame. Numerical results show a relative error between numerical results and experimental data less than 10%, what validates genetic algorithms as a good alternative for solving these computationally-intensive adjustment problems.