1: \begin{abstract}

2: Large linear systems with sparse, non-symmetric matrices arise in the modeling of

3: Markov chains or in the discretization of convection-diffusion problems. Due to their

4: potential to solve sparse linear systems with an effort that is linear in the number

5: of unknowns, algebraic multigrid (AMG) methods are of fundamental interest for such

6: systems. For symmetric positive definite matrices, fundamental theoretical convergence

7: results are established, and efficient AMG solvers have been developed.

8: In contrast, for non-symmetric matrices, theoretical convergence results have been

9: provided only recently. A property that is sufficient for convergence is that the

10: matrix be an M-matrix.

11: %

12: In this paper, we present how the simulation of incompressible fluid flows with

13: particle methods leads to large linear systems with sparse, non-symmetric matrices.

14: In each time step, the Poisson equation is approximated by meshfree finite differences.

15: While traditional least squares approaches do not guarantee an M-matrix structure,

16: an approach based on linear optimization yields optimally sparse M-matrices.

17: For both types of discretization approaches, we investigate the performance of a

18: classical AMG method, as well as an AMLI type method.

19: While in the considered test problems, the M-matrix structure turns out not to be

20: necessary for the convergence of AMG, problems can occur when it is violated. In

21: addition, the matrices obtained by the linear optimization approach result in fast

22: solution times due to their optimal sparsity.

23: \end{abstract}

24: