A numerically scalable dual-primal substructuring method for the solution of contact problems––part I: the frictionless case Original Research Article

We present a dual-primal non-linear domain decomposition method with Lagrange multipliers for solving iteratively frictionless contact problems. This method, which is based on the feti-DP substructuring algorithm, features a non-linear Krylov-type acceleration scheme. It addresses both cases of restrained and unrestrained bodies. When the bodies in contact behave linearly, it does not perform any Newton-like iteration to solve the non-linear contact problem. We present performance results for several numerical simulations which suggest that the proposed method is numerically scalable with respect to both the problem size and the number of subdomains. We also illustrate the parallel scalability of this method on a cluster of Silicon Graphics systems for a three-million degree of freedom static contact problem.