Commits · 98408c4283909fdbd7e1f84ecaf1df5e2a271688 · Sander, Oliver / dune-gfe

Jan 31, 2011
- use system method mkdtemp to create a temporary directory without name clashes · 98408c42
  Oliver Sander authored 14 years ago
  
  [[Imported from SVN: r6909]]
  98408c42
- allow to have the coupling boundary on either side of the rod · e77275da
  Oliver Sander authored 14 years ago
  
  [[Imported from SVN: r6908]]
  e77275da
- for testing: support having a Dirichlet boundary on either side of the rod · 4c957737
  Oliver Sander authored 14 years ago
  
  [[Imported from SVN: r6907]]
  4c957737
- bugfix: need to multiply with the _transposed_ rotation to get the proper local coordinates · d1beae28
  Oliver Sander authored 14 years ago
  
  [[Imported from SVN: r6906]]
  d1beae28
- use new method RodWriter::writeBinary · c9b5e4c3
  Oliver Sander authored 14 years ago
  
  [[Imported from SVN: r6905]]
  c9b5e4c3
- new method to write rods in simple binary format · 82dc65b3
  Oliver Sander authored 14 years ago
  
  [[Imported from SVN: r6904]]
  82dc65b3
- fix the canonical --> local force transformation · a506a0e7
  Oliver Sander authored 14 years ago
  
  [[Imported from SVN: r6903]]
  a506a0e7
- use the BoundaryPatch iterator to iterate over boundary patches · 3aa5ad53
  Oliver Sander authored 14 years ago
  
  [[Imported from SVN: r6902]]
  3aa5ad53
- bugfix: the linearized rod NtD map gets its Neumann values in local coordinates (wrt the directors) · 15279bd3
  Oliver Sander authored 14 years ago
  
  [[Imported from SVN: r6901]]
  15279bd3
- make the cgsolver not print any output -- at a paranoia check to make sure the... · 940ff625
  Oliver Sander authored 14 years ago
  
  make the cgsolver not print any output -- at a paranoia check to make sure the residual really is zero afterwards [[Imported from SVN: r6900]]
  940ff625
- remove two unused variables · 71dd9d0c
  Oliver Sander authored 14 years ago
  
  [[Imported from SVN: r6899]]
  71dd9d0c
- use the boundary patch iterator to implement a loop over a boundary patch · 7cd85c22
  Oliver Sander authored 14 years ago
  
  [[Imported from SVN: r6898]]
  7cd85c22
Jan 30, 2011
- don't call the preprocess method of the iteration steps right before solving -- it is not necessary · 32543273
  Oliver Sander authored 14 years ago
  
  [[Imported from SVN: r6897]]
  32543273
Jan 28, 2011
- bugfix: initialize the neumannRegularization_ member in the constructors · b3756d0f
  Oliver Sander authored 14 years ago
  
  [[Imported from SVN: r6896]]
  b3756d0f
Jan 26, 2011
- and another parenthesis · 2134056d
  Oliver Sander authored 14 years ago
  
  [[Imported from SVN: r6895]]
  2134056d
- Stupid bug: forgot a few parentheses. Arrgghhh!!! · 52e3f4a6
  Oliver Sander authored 14 years ago
  
  [[Imported from SVN: r6894]]
  52e3f4a6
- transform the weak boundary stresses to the strong ones also for the case with contact · da0cb6ba
  Oliver Sander authored 14 years ago
  
  [[Imported from SVN: r6893]]
  da0cb6ba
- bugfix: forgot to turn the residual (== weak boundary stresses) into strong boundary stresses · 9aa6f282
  Oliver Sander authored 14 years ago
  
  [[Imported from SVN: r6892]]
  9aa6f282
- Add method to turn weak boundary stresses into strong boundary stresses. · ee640ac4
  Oliver Sander authored 14 years ago
  
  This is done by solving a mass matrix system. [[Imported from SVN: r6891]]
  ee640ac4
- compile the contact-supporting code only if dune-contact has been found · 48770fc3
  Oliver Sander authored 14 years ago
  
  [[Imported from SVN: r6890]]
  48770fc3
- bugfix: forgot to set vector size · 4176606f
  Oliver Sander authored 14 years ago
  
  [[Imported from SVN: r6889]]
  4176606f
- bugfix: the totalDirichletNodes field wasn't set up correctly · 86536dc5
  Oliver Sander authored 14 years ago
  
  [[Imported from SVN: r6888]]
  86536dc5
- add a scaled identity to the rod stiffness matrix if there is no Dirichlet... · 626e141f
  Oliver Sander authored 14 years ago
  
  add a scaled identity to the rod stiffness matrix if there is no Dirichlet boundary: this regularizes the problem [[Imported from SVN: r6886]]
  626e141f
- bugfix: store the continuum subdomain solutions in the approprate place when iterating with contact · fac46458
  Oliver Sander authored 14 years ago
  
  [[Imported from SVN: r6884]]
  fac46458
- bugfix: forgot to set initial value · 0db13f55
  Oliver Sander authored 14 years ago
  
  [[Imported from SVN: r6883]]
  0db13f55
- bugfix: forgot to set vector length · 67ae92dd
  Oliver Sander authored 14 years ago
  
  [[Imported from SVN: r6882]]
  67ae92dd
- bugfix: do not mix boundary patches for different grids · b2990574
  Oliver Sander authored 14 years ago
  
  [[Imported from SVN: r6879]]
  b2990574
Jan 24, 2011

Finish implementation of SteklovPoincareStep with contact problems. I don't... · 9c16108a

Oliver Sander authored 14 years ago

Finish implementation of SteklovPoincareStep with contact problems.  I don't dare to test this today...

[[Imported from SVN: r6871]]

9c16108a

rearrange code that calls the linearized NtD maps · 88fad907
Oliver Sander authored 14 years ago
```
[[Imported from SVN: r6870]]
```
88fad907

Add a new method iterateWithContact · 3dda20ef

Oliver Sander authored 14 years ago

This handles the case that there are contact/mortar relations
between the continua.  Unfortunately, this does not quite fit
into the framework: so far we have treated all continua individually.
However, the contact solvers lumps everything into a single
algebraic system.  Consequently this is now getting a bit
hacky.  I hope to find something prettier in the future.

The linearized NtD map for continua is not implemented yet.

[[Imported from SVN: r6868]]

3dda20ef

minor refactoring: move a common std::map operation into a separate method · db983376
Oliver Sander authored 14 years ago
```
[[Imported from SVN: r6866]]
```
db983376
add a constructor for a general complex · 56d64c20
Oliver Sander authored 14 years ago
```
[[Imported from SVN: r6863]]
```
56d64c20
export iterators over rods and continua · c8229582
Oliver Sander authored 14 years ago
```
[[Imported from SVN: r6862]]
```
c8229582
bugfix: set the correct Dirichlet nodes for the linearized continuum NtD map · a40035bb
Oliver Sander authored 14 years ago
```
[[Imported from SVN: r6858]]
```
a40035bb
bugfix · f85c1df4
Oliver Sander authored 14 years ago
```
[[Imported from SVN: r6857]]
```
f85c1df4

Make the linearized continuum NtD map handle arbitrary number of couplings. · c964d8a6

Oliver Sander authored 14 years ago

This completes the implementation.  In principle, the Steklov-Poincare step
should now work for arbitrary configurations of rods and continua.

Known issues:
- rods must not stick to the same continuum with both ends
- all objects must have a nonempty Dirichlet boundary
  (will be fixed very soon)

[[Imported from SVN: r6856]]

c964d8a6

make the linearized NtD map work for rods with zero, one, or two couplings · 92b8b9c3
Oliver Sander authored 14 years ago
```
[[Imported from SVN: r6855]]
```
92b8b9c3

Jan 23, 2011
- call the linearized NtD maps for all rods/continua · 49c072c2
  Oliver Sander authored 14 years ago
  
  [[Imported from SVN: r6854]]
  49c072c2
- hand over all residual forces to the linearized NtD maps · 845b4819
  Oliver Sander authored 14 years ago
  
  [[Imported from SVN: r6853]]
  845b4819
- hand over the subdomain names to the linearized NtD maps · 3d4f637b
  Oliver Sander authored 14 years ago
  
  [[Imported from SVN: r6852]]
  3d4f637b