A penalty interior point algorithm for a parameter identification problem in elastoplasticity

This paper deals with a special class of parameter identification problems in structural plasticity. Specifically, we address the problem of identifyi ng yield limits and hardening moduli from knowledge of the displacement res ponse of the structure under a given set of proportional loads. Under the a ssumptions of piece-wise linear holonomic (reversible) plasticity and a sui tably discretized structure, the inverse problem can be formulated as a spe cial optimization problem known as a Mathematical Program with Equilibrium Constraints (MPEC), a key feature of which is the presence of complementari ty conditions, involving the orthogonality of two sign-constrained vectors. A recently developed numerical scheme, the Penalty Interior Point Algorith m (PIPA), is proposed for solving the identification problem. Some computat ional results for a hypothetical beam on elastoplastic springs are also giv en.