A program for inverse analysis of triaxial tests in consolidated drained condition using a genetic algorithm

Abstract

For a model to be used in a numerical procedure it is necessary to find appropriate parameters that reproduce the best response of the model in relation to the available experimental results. To avoid subjectivity and facilitate the search for parameters of constitutive models, this study aims to develop a program (GATriaxial) for inverse analysis of isotropically consolidated drained triaxial tests (CD) results using a nonlinear elastic model and a genetic algorithm (GA). The program was validated by applying different configurations to the GA in the search for a known optimal solution and its efficiency verified by the comparison between the representation of the calibrated model using the traditional calibration and GATriaxial, using CD tests. It was shown that the use of GA in the model calibration employing GATriaxial software produces an automated estimate, less subjective, fast and still provides an indicator of the quality of the fit. In addition, calibration by inverse analysis was found to be more efficient in defining model parameters than traditional calibration.