New plasticity model using artificial neural networks

Concrete is one of the most widely used materials in building construction. Under static loads, the concrete is subjected to various stress states associated with significant deformation. In this paper, we study the feasibility of using artificial neural networks for modelling the mechanical behaviour of plain concrete in compression under static loading using the theory of plasticity. The database used for the development is obtained from a selection of previously published tests results and includes a series of uniaxial, biaxial and triaxial compression tests. This database is used for making and testing predictive models. The results of the ANN model can accurately predict the load resistance and deformation capacity in various compression stress states. Expansion and plastic contraction of concrete under different confining pressures and the nonlinear behaviour of concrete are simulated. The results show that the accuracy of the proposed ANN-based models is satisfactory compared with experimental results. It is also shown that the RBF neural network model may accurately represent the load resistance and deformation capacity for three types of compression tests.