In this paper we present an algorithm for the numerical simulation of the cavitation in the hydrodynamic lubrication of journal bearings. Despite the fact that this physical process is usually modelled as a free boundary problem, we adopted the equivalent variational inequality formulation. We propose a two-level iterative algorithm, where the outer iteration is associated to the penalty method, used to transform the variational inequality into a variational equation, and the inner iteration is associated to the conjugate gradient method, used to solve the linear system generated by applying the finite element method to the variational equation. This inner part was implemented using the element by element strategy, which is easily parallelized. We analyse the behavior of two physical parameters and discuss some numerical results. Also, we analyse some results related to the performance of a parallel implementation of the algorithm.

Este artigo apresenta um algoritmo de dois passos para a resolução numérica de problemas de complementaridade linear (LCP). O algoritmo que apresentaremos combina a estratégia de conjunto ativo aliada ao método dos gradientes conjugados quadrático para a resolução do sistema linear reduzido. O desenvolvimento deste método de dois passos foi baseada em um outro, também de dois passos, desenvolvido por Kocvara & Zowe (1994), que combina o método de super relaxação sucessiva simétrica com projeção e o método dos gradientes conjugados pré condicionado. Na seção Experiência Numérica, utilizaremos o método para resolver LCPs com matrizes não singulares pertencentes às classes de matrizes semi positivas definidas, P-matrizes e P0-matrizes. Será feita ainda uma comparação entre o desempenho numérico do algoritmo apresentado neste texto e o método desenvolvido por Pardalos, Ye, Han & Kaliski (1993).

In this article we present an iterative two-step algorithm for the numerical solution of linear complementarity problems (LCP). The algorithm presented combine the active set strategy with the square conjugated gradient method for the linear system solution. This iterative two-step algorithm was based in a similar one developed by Kocvara & Zowe (1994), that combine the symmetric successive over relaxation with projection method and the precondicioned conjugated gradient method. In the Numerical Experience section, we have tested the method for solve some LCP with non-singular matrices. Such matrices belong to semi positive definite matrix class, P-matrix class and P0-matrix class. Furthermore we show a comparision between our method and the one developed by Pardalos, Ye, Han & Kaliski (1993).