git-svn-id: svn://svn.iws.uni-stuttgart.de/DUMUX/dumux/trunk@4873 2fb0f335-1f38-0410-981e-8018bf24f1b0

git-svn-id: svn://svn.iws.uni-stuttgart.de/DUMUX/dumux/trunk@4872 2fb0f335-1f38-0410-981e-8018bf24f1b0

IMHO it still has some issues with consistency and the level of detail
in some places...

git-svn-id: svn://svn.iws.uni-stuttgart.de/DUMUX/dumux/trunk@4871 2fb0f335-1f38-0410-981e-8018bf24f1b0

git-svn-id: svn://svn.iws.uni-stuttgart.de/DUMUX/dumux/trunk@4869 2fb0f335-1f38-0410-981e-8018bf24f1b0

git-svn-id: svn://svn.iws.uni-stuttgart.de/DUMUX/dumux/trunk@4866 2fb0f335-1f38-0410-981e-8018bf24f1b0

git-svn-id: svn://svn.iws.uni-stuttgart.de/DUMUX/dumux/trunk@4865 2fb0f335-1f38-0410-981e-8018bf24f1b0

git-svn-id: svn://svn.iws.uni-stuttgart.de/DUMUX/dumux/trunk@4864 2fb0f335-1f38-0410-981e-8018bf24f1b0

and move the BoxAssembler class out of the PDELab namespace

git-svn-id: svn://svn.iws.uni-stuttgart.de/DUMUX/dumux/trunk@4863 2fb0f335-1f38-0410-981e-8018bf24f1b0

git-svn-id: svn://svn.iws.uni-stuttgart.de/DUMUX/dumux/trunk@4862 2fb0f335-1f38-0410-981e-8018bf24f1b0

newton: rename: u->uLastIter, uOld->uCurrentIter, use separate memory for the next iterative solution and the delta vector

git-svn-id: svn://svn.iws.uni-stuttgart.de/DUMUX/dumux/trunk@4861 2fb0f335-1f38-0410-981e-8018bf24f1b0

git-svn-id: svn://svn.iws.uni-stuttgart.de/DUMUX/dumux/trunk@4860 2fb0f335-1f38-0410-981e-8018bf24f1b0

git-svn-id: svn://svn.iws.uni-stuttgart.de/DUMUX/dumux/trunk@4858 2fb0f335-1f38-0410-981e-8018bf24f1b0

git-svn-id: svn://svn.iws.uni-stuttgart.de/DUMUX/dumux/trunk@4857 2fb0f335-1f38-0410-981e-8018bf24f1b0

git-svn-id: svn://svn.iws.uni-stuttgart.de/DUMUX/dumux/trunk@4856 2fb0f335-1f38-0410-981e-8018bf24f1b0

git-svn-id: svn://svn.iws.uni-stuttgart.de/DUMUX/dumux/trunk@4855 2fb0f335-1f38-0410-981e-8018bf24f1b0

git-svn-id: svn://svn.iws.uni-stuttgart.de/DUMUX/dumux/trunk@4854 2fb0f335-1f38-0410-981e-8018bf24f1b0

git-svn-id: svn://svn.iws.uni-stuttgart.de/DUMUX/dumux/trunk@4853 2fb0f335-1f38-0410-981e-8018bf24f1b0

git-svn-id: svn://svn.iws.uni-stuttgart.de/DUMUX/dumux/trunk@4852 2fb0f335-1f38-0410-981e-8018bf24f1b0

git-svn-id: svn://svn.iws.uni-stuttgart.de/DUMUX/dumux/trunk@4851 2fb0f335-1f38-0410-981e-8018bf24f1b0

AND
updated the "Mother of all Flowcharts" (here to stay)


git-svn-id: svn://svn.iws.uni-stuttgart.de/DUMUX/dumux/trunk@4847 2fb0f335-1f38-0410-981e-8018bf24f1b0

tutorial_coupled


git-svn-id: svn://svn.iws.uni-stuttgart.de/DUMUX/dumux/trunk@4844 2fb0f335-1f38-0410-981e-8018bf24f1b0

git-svn-id: svn://svn.iws.uni-stuttgart.de/DUMUX/dumux/trunk@4843 2fb0f335-1f38-0410-981e-8018bf24f1b0

git-svn-id: svn://svn.iws.uni-stuttgart.de/DUMUX/dumux/trunk@4842 2fb0f335-1f38-0410-981e-8018bf24f1b0

git-svn-id: svn://svn.iws.uni-stuttgart.de/DUMUX/dumux/trunk@4841 2fb0f335-1f38-0410-981e-8018bf24f1b0

git-svn-id: svn://svn.iws.uni-stuttgart.de/DUMUX/dumux/trunk@4840 2fb0f335-1f38-0410-981e-8018bf24f1b0

git-svn-id: svn://svn.iws.uni-stuttgart.de/DUMUX/dumux/trunk@4839 2fb0f335-1f38-0410-981e-8018bf24f1b0

git-svn-id: svn://svn.iws.uni-stuttgart.de/DUMUX/dumux/trunk@4838 2fb0f335-1f38-0410-981e-8018bf24f1b0

git-svn-id: svn://svn.iws.uni-stuttgart.de/DUMUX/dumux/trunk@4837 2fb0f335-1f38-0410-981e-8018bf24f1b0

git-svn-id: svn://svn.iws.uni-stuttgart.de/DUMUX/dumux/trunk@4836 2fb0f335-1f38-0410-981e-8018bf24f1b0

git-svn-id: svn://svn.iws.uni-stuttgart.de/DUMUX/dumux/trunk@4835 2fb0f335-1f38-0410-981e-8018bf24f1b0

(phaseState -> fluidState, material context -> material parameters, etc)

git-svn-id: svn://svn.iws.uni-stuttgart.de/DUMUX/dumux/trunk@4834 2fb0f335-1f38-0410-981e-8018bf24f1b0

now the box problems can be compiled without having any PDELab files around...

git-svn-id: svn://svn.iws.uni-stuttgart.de/DUMUX/dumux/trunk@4829 2fb0f335-1f38-0410-981e-8018bf24f1b0

git-svn-id: svn://svn.iws.uni-stuttgart.de/DUMUX/dumux/trunk@4827 2fb0f335-1f38-0410-981e-8018bf24f1b0

since parallel the linear solver is hard-coded into PDElab, this only
works for sequential runs. to try it out, change the value of
HAVE_DUNE_PDELAB to 0 in the config.h file and add

...

around the definition of the LocalFEMSpace property.

git-svn-id: svn://svn.iws.uni-stuttgart.de/DUMUX/dumux/trunk@4826 2fb0f335-1f38-0410-981e-8018bf24f1b0

divide the relative tolerance by 10 to get the minimum reassemble
error. for the next newton iteration, reassemble all vertices which
exhibit a relative error which is larger than the geometric mean of
the minimum reassemble tolerance and the relative error of the last
iteration.

git-svn-id: svn://svn.iws.uni-stuttgart.de/DUMUX/dumux/trunk@4825 2fb0f335-1f38-0410-981e-8018bf24f1b0

git-svn-id: svn://svn.iws.uni-stuttgart.de/DUMUX/dumux/trunk@4824 2fb0f335-1f38-0410-981e-8018bf24f1b0

this dramatically speeds up things...

git-svn-id: svn://svn.iws.uni-stuttgart.de/DUMUX/dumux/trunk@4821 2fb0f335-1f38-0410-981e-8018bf24f1b0

git-svn-id: svn://svn.iws.uni-stuttgart.de/DUMUX/dumux/trunk@4820 2fb0f335-1f38-0410-981e-8018bf24f1b0

This also has the advantage, that hybrid grids should work again from
the FVElementGeometry's POV. basically the only dependence on PDELab
which is left is the linear solver. This dependcy is easy to get rid
of for the sequential case, but much harder for parallel computations
(at least for clueless persons such as me).

git-svn-id: svn://svn.iws.uni-stuttgart.de/DUMUX/dumux/trunk@4819 2fb0f335-1f38-0410-981e-8018bf24f1b0

as far as I could see it was (1) unused (2) would not have worked.

git-svn-id: svn://svn.iws.uni-stuttgart.de/DUMUX/dumux/trunk@4818 2fb0f335-1f38-0410-981e-8018bf24f1b0