1: \begin{abstract}

2: Highly heterogeneous, anisotropic coefficients, e.g. in the simulation

3: of carbon-fibre composite components, can lead to extremely

4: challenging finite element systems. Direct solvers for the resulting

5: large and sparse linear systems suffer from severe memory requirements

6: and limited parallel scalability, while iterative solvers in general lack robustness. 

7: Two-level spectral domain decomposition methods can provide such

8: robustness for symmetric positive definite linear systems, by using coarse spaces

9: based on independent generalized eigenproblems in the subdomains. Rigorous

10: condition number bounds are independent of mesh size, number of subdomains, as

11: well as coefficient contrast. However, their parallel scalability is still

12: limited by the fact that (in order to guarantee robustness) the coarse

13: problem is solved via a direct method. In this paper, we introduce a multilevel variant in

14: the context of subspace correction methods and provide a general convergence

15: theory for its robust convergence for abstract, elliptic variational

16: problems. Assumptions of the theory are verified for

17: conforming, as well as for discontinuous Galerkin methods applied to a scalar diffusion problem.

18: Numerical results illustrate the performance of the method for two- and three-dimensional

19: problems and for various discretization schemes, in the context of scalar diffusion and linear elasticity.

20: \end{abstract}

21: