Change initial condition to block/parabolic profile

See merge request !1645

[subgridmanager] Fix clang compiler errors

Closes #727

See merge request !1641

* remove unnecessary traits
* add test for alu grid

[bboxtree] use single epsilon for point isection query

See merge request !1643

Cleanup/comment and indentation

See merge request !1644

[mpfa][fix] Use correct type for data member

See merge request !1642

[cleanup] Remove superflouous semicolon

See merge request !1640

[staggered]fluxVars] Improve handling of Neumann BCs for the momentum eq

See merge request !1630

* correctly account for the case when there is a Neumann bc for both
  the normal and tangential part of the stress.

[test] Disable regression test for 2p2c adaptive sequential

See merge request !1638

With our current test tool it is very hard to create regression tests
for adaptive grids. The adaption criterion is usually based on a
threshold, so it might be slightly different on different architectures.
This will currently make the test fail, altough physically the solution
might not be very different. For such cases tests are very complicated.
One possibility might be the projection on a known grid on which the
regression error can be computed.

Until such tools are avaiable, we simply compile and run the test but
do not check for regression in the solution.

Feature/l2 norm

See merge request !1636

We assume the grid function to model a Dune::Functions::GridFunction.

Feature/2d 2d intersections

See merge request !1625

Rename grahamConvexHull2d3d to grahamConvexHull<2>,
deprecate old interface.

Dennis Gläser authored 5 years ago and Timo Koch committed 5 years ago

This actually makes use of the 2d in 3d algorithm at the moment. In the
future a more efficient implementation for 2d hulls could be
implemented.