Constitutive models for concrete based on the microplane concept haverepeatedly proven their ability to well-reproduce its non-linear response onmaterial as well as structural scales. The major obstacle to a routineapplication of this class of models is, however, the calibration ofmicroplane-related constants from macroscopic data. The goal of this paper istwo-fold: (i) to introduce the basic ingredients of a robust inverse procedurefor the determination of dominant parameters of the M4 model proposed by Bazantand co-workers based on cascade Artificial Neural Networks trained byEvolutionary Algorithm and (ii) to validate the proposed methodology against arepresentative set of experimental data. The obtained results demonstrate thatthe soft computing-based method is capable of delivering the searched responsewith an accuracy comparable to the values obtained by expert users.
展开▼
