The dune macro used to add the target only adds it if doxygen was found.
Then the target was not know in add_custom_command which is an error
in CMake > v3.2.

Feature/doxyref



See merge request !39

Cleanup/multidomain

One try to cleanup, unify, and purge the multidomain models by:
 - removing duplicate code
 - avoiding unnecessary subfunctions
 - consistent procedure in isothermal and non-isothermal models
 - adding warnings for undesired use (e.g. useMoles, not defined BC)
 - speed up the tests

See merge request !38

[multidomain] Various small cleanups, remove unnecessary/potentially buggy lines of code, add useMoles.

Note: this leads to a change how to set boundary conditions for the momentum balance. Inflow now becomes Dirichlet and Outflow becomes Neumann. This is consistent with the new notation.

Feature/preconditionerblocklevel

Add a new property LinearSolverPreconditionerBlockLevel that sets the block level (in case of nesting of blocks) to be inverted by the preconditioner. Defaults to 1. Needed for matrices for coupled matrices.

See merge request !37

Timo Koch authored 9 years ago and Christoph Grüninger committed 9 years ago

The blocklevel parameter to preconditioners defaults to 1.
For higher nested block matrices e.g. matrices for coupled
problems the blocklevel needs to be higher. This can now be
set via the property LinearSolverPreconditionerBlockLevel.

[el2p] Fix the test if pdelab is not present

The test has to have a main function. The test should
be skipped if pdelab is not present.

See merge request !35

The test has to have a main function. The test should
be skipped if pdelab is not present.

In the section where the assembly of the linear system is explained,
the RHS of the linear system should be the residual r(u) and not the
primary variable u itself. It should read J(u)*x=r(u).

Feature/drop constexpr check



See merge request !33

We require GCC 4.8, the feature is supported since GCC 4.6

[problem] Move isCornerPoint() to more upstream position, because it may be used in other circumstances, too.

[multidomain,BeaversJoseph] Use BJCoeff() from the Darcy spatial parameters which is more meaningful.

This includes proper warnings and removal of the SpatialParams objects.

- removed TODOs in deprecated functions
- add use of shared pointer (in multidomainmodel.hh and multidomainconvergencewriter.hh)
- adapt to nonlinear, common classes
- adjust indentation (which changed due to removal of tabs)
- adapt doxygen docu

This time:
  - deprecated old coupling12() and coupling21() and replaced them by one coupling()
  - try to group boundary condition types
  - unified naming and steps how fluxes are calculated (nothing new was added)
  - further replacing _s, _n by 1, 2

boundaryLayerModel: spurious changes in (~4% for 1e-11 and 1% for 1e-9 in velocity)

This has the following advantages:
 - reuse the BL model in the non-isothermal case
 - possible move of the functions to more problem specific place (due to the use of runtime parameters)
 - better comparison of what is actually done in the code

Update reference solution, because for the BL Model the gas viscosity from Darcy was used.