An algorithm for incremental elastoplastic analysis using equality constrained sequential quadratic programming

A new method for the incremental elastoplastic analysis of structures is presented. Retaining all the equations and variables of the problem at the same level, a particular implementation of sequential quadratic programming with equality constraints is applied to solve the elastoplastic step. The proposed method and standard strain driven formulations based on return mapping by closest point projection schemes are compared by viewing them as dual decomposition techniques suitable for solving the mathematical programming problem associated to the finite elastoplastic step.The numerical tests performed for 2D problems show the good performance and the great robustness of the proposed algorithm.

► Finite elastoplastic step equations are obtained by an optimization problem. ► Strain driven method are seen as decomposition technique to solve this problem. ► A new method based on sequential quadratic programming is proposed. ► Several numerical results are reported to show its robustness.