1: \begin{abstract}

2: The Stokes system with constant viscosity can be cast into different formulations by exploiting the incompressibility constraint.

3: For instance 

4: the strain in the weak formulation can  be replaced by the gradient 

5: to decouple 

6: the  velocity components in the different coordinate directions. 

7: Thus the discretization of the simplified problem

8: leads to

9: fewer nonzero entries in the stiffness matrix.

10: This is of particular  interest in large scale simulations where

11: a reduced memory bandwidth requirement

12: can help to significantly accelerate the computations.

13: In the case of a piecewise constant viscosity, as it typically arises in multi-phase flows,

14: or when the boundary conditions involve traction, the situation is more complex,

15: and one has to treat the cross derivatives in the original

16: Stokes system with care.

17: A naive application of the standard vectorial Laplacian results in a

18: physically incorrect solution, while formulations based on the strain increase the computational effort everywhere, even when the inconsistencies arise only from an 

19: incorrect treatment in a small fraction of the computational domain.

20: Here we propose a new approach

21: that is consistent with the strain-based formulation and preserves the decoupling advantages of the gradient-based formulation in isoviscous subdomains. 

22: The modification

23: is equivalent to locally changing the discretization stencils, hence the more expensive discretization is restricted to a lower dimensional interface,

24: making the additional computational cost asymptotically negligible.

25: We demonstrate the consistency and convergence properties of the method and show that in a massively parallel setup, the multigrid solution of the resulting discrete systems is faster than for the classical strain-based formulation.

26: Moreover, we give an application example which is inspired by geophysical research.

27: \end{abstract}

28: