We evaluate a parallel Schur preconditioner for large systems of equations arising from a finite element discretization of the Navier-Stokes equations with streamline diffusion. The performance of the method is assessed on a biomedical problem involving oscillatory flow in a human abdominal bifurcation. Fast access to flow conditions in this location might support physicians in quicker decision making concerning potential interventions. We demonstrate scaling to 8 processors with more than 50% efficiency as well as a significant relaxation of memory requirements. We found an acceleration by up to a factor 9.5 compared to a direct sparse parallel solver at stopping criteria ensuring results similar to a validated reference solution. © 2009 Springer.
CITATION STYLE
Burckhardt, K., Szczerba, D., Brown, J., Muralidhar, K., & Székely, G. (2009). Fast implicit simulation of oscillatory flow in human abdominal bifurcation using a schur complement preconditioner. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 5704 LNCS, pp. 747–759). https://doi.org/10.1007/978-3-642-03869-3_70
Mendeley helps you to discover research relevant for your work.