]>
https://gitweb.dealii.org/ - dealii-svn.git/log
bangerth [Thu, 21 Aug 2008 00:58:17 +0000 (00:58 +0000)]
Regenerate.
git-svn-id: https://svn.dealii.org/trunk@16616
0785d39b -7218-0410-832d-
ea1e28bc413d
bangerth [Wed, 20 Aug 2008 20:37:13 +0000 (20:37 +0000)]
add
git-svn-id: https://svn.dealii.org/trunk@16615
0785d39b -7218-0410-832d-
ea1e28bc413d
bangerth [Wed, 20 Aug 2008 17:12:09 +0000 (17:12 +0000)]
Use a better cutoff for the velocity in computing the time step.
git-svn-id: https://svn.dealii.org/trunk@16614
0785d39b -7218-0410-832d-
ea1e28bc413d
bangerth [Wed, 20 Aug 2008 16:54:44 +0000 (16:54 +0000)]
Comment on timing.
git-svn-id: https://svn.dealii.org/trunk@16613
0785d39b -7218-0410-832d-
ea1e28bc413d
bangerth [Wed, 20 Aug 2008 14:33:49 +0000 (14:33 +0000)]
Instead of copying into the global matrix/vector by hand and later calling condense on the resulting object, use ConstraintMatrix::distribute_local_to_global.
git-svn-id: https://svn.dealii.org/trunk@16612
0785d39b -7218-0410-832d-
ea1e28bc413d
bangerth [Wed, 20 Aug 2008 13:13:36 +0000 (13:13 +0000)]
A few more minor cleanups.
git-svn-id: https://svn.dealii.org/trunk@16611
0785d39b -7218-0410-832d-
ea1e28bc413d
bangerth [Wed, 20 Aug 2008 13:08:59 +0000 (13:08 +0000)]
Make the drop tolerance a parameter.
git-svn-id: https://svn.dealii.org/trunk@16610
0785d39b -7218-0410-832d-
ea1e28bc413d
bangerth [Wed, 20 Aug 2008 13:06:50 +0000 (13:06 +0000)]
Simplify copying elements into Trilinos in the same way.
git-svn-id: https://svn.dealii.org/trunk@16609
0785d39b -7218-0410-832d-
ea1e28bc413d
bangerth [Wed, 20 Aug 2008 13:02:45 +0000 (13:02 +0000)]
Simplify the code that computes row lengths in the AMG.
git-svn-id: https://svn.dealii.org/trunk@16608
0785d39b -7218-0410-832d-
ea1e28bc413d
bangerth [Wed, 20 Aug 2008 12:54:24 +0000 (12:54 +0000)]
Use a shared_ptr and thereby also remove the memory leak that we currently have.
git-svn-id: https://svn.dealii.org/trunk@16607
0785d39b -7218-0410-832d-
ea1e28bc413d
bangerth [Wed, 20 Aug 2008 12:52:50 +0000 (12:52 +0000)]
Use the new scalar_product function in base/.
git-svn-id: https://svn.dealii.org/trunk@16606
0785d39b -7218-0410-832d-
ea1e28bc413d
bangerth [Wed, 20 Aug 2008 12:41:11 +0000 (12:41 +0000)]
New function scalar_product.
git-svn-id: https://svn.dealii.org/trunk@16605
0785d39b -7218-0410-832d-
ea1e28bc413d
bangerth [Wed, 20 Aug 2008 12:06:54 +0000 (12:06 +0000)]
Sequester everything that has to do with equation data into its own namespace.
git-svn-id: https://svn.dealii.org/trunk@16604
0785d39b -7218-0410-832d-
ea1e28bc413d
bangerth [Wed, 20 Aug 2008 11:48:58 +0000 (11:48 +0000)]
Use dynamic range of temperature to define viscosity, not maximum norm.
git-svn-id: https://svn.dealii.org/trunk@16603
0785d39b -7218-0410-832d-
ea1e28bc413d
kanschat [Wed, 20 Aug 2008 05:01:08 +0000 (05:01 +0000)]
ignore dll
git-svn-id: https://svn.dealii.org/trunk@16602
0785d39b -7218-0410-832d-
ea1e28bc413d
bangerth [Wed, 20 Aug 2008 03:54:01 +0000 (03:54 +0000)]
Write more.
git-svn-id: https://svn.dealii.org/trunk@16601
0785d39b -7218-0410-832d-
ea1e28bc413d
bangerth [Wed, 20 Aug 2008 01:05:30 +0000 (01:05 +0000)]
Add reference to step-31
git-svn-id: https://svn.dealii.org/trunk@16600
0785d39b -7218-0410-832d-
ea1e28bc413d
bangerth [Wed, 20 Aug 2008 00:22:02 +0000 (00:22 +0000)]
A few more comments.
git-svn-id: https://svn.dealii.org/trunk@16599
0785d39b -7218-0410-832d-
ea1e28bc413d
kronbichler [Tue, 19 Aug 2008 22:26:20 +0000 (22:26 +0000)]
At the start of each Stokes solve, initialize solution vector to zero. This drives down the number of iterations by a full one third :-)
git-svn-id: https://svn.dealii.org/trunk@16598
0785d39b -7218-0410-832d-
ea1e28bc413d
bangerth [Tue, 19 Aug 2008 20:18:47 +0000 (20:18 +0000)]
Rebuild temperature matrices only when needed. Do not forget to condense the temperature system with hanging nodes.
git-svn-id: https://svn.dealii.org/trunk@16597
0785d39b -7218-0410-832d-
ea1e28bc413d
bangerth [Tue, 19 Aug 2008 19:41:14 +0000 (19:41 +0000)]
Split the 'degree' variable into two, one for the Stokes and one for the temperature system, to make experimenting with different choices simpler.
git-svn-id: https://svn.dealii.org/trunk@16596
0785d39b -7218-0410-832d-
ea1e28bc413d
bangerth [Tue, 19 Aug 2008 05:42:14 +0000 (05:42 +0000)]
Fix a bug in transfering the solution.
git-svn-id: https://svn.dealii.org/trunk@16595
0785d39b -7218-0410-832d-
ea1e28bc413d
bangerth [Tue, 19 Aug 2008 05:35:21 +0000 (05:35 +0000)]
Re-assemble the solution vectors into a combined DoFHandler so that we can output everything into the same output file.
git-svn-id: https://svn.dealii.org/trunk@16594
0785d39b -7218-0410-832d-
ea1e28bc413d
bangerth [Tue, 19 Aug 2008 03:48:14 +0000 (03:48 +0000)]
Use two DoFHandlers, matrices, sets of vectors, etc, so that we don't always have to rebuild all FEValues data for all components, independent of what we're currently doing.
git-svn-id: https://svn.dealii.org/trunk@16593
0785d39b -7218-0410-832d-
ea1e28bc413d
bangerth [Mon, 18 Aug 2008 23:05:06 +0000 (23:05 +0000)]
Remove the currently problematic lines removing coarsening flags.
git-svn-id: https://svn.dealii.org/trunk@16592
0785d39b -7218-0410-832d-
ea1e28bc413d
bangerth [Mon, 18 Aug 2008 21:58:07 +0000 (21:58 +0000)]
Fix the mesh_smoothing_0[12] testcases. A description of the problem is found in the mesh_smoothing_02.cc testcase.
git-svn-id: https://svn.dealii.org/trunk@16591
0785d39b -7218-0410-832d-
ea1e28bc413d
bangerth [Mon, 18 Aug 2008 21:56:16 +0000 (21:56 +0000)]
Check in correct output.
git-svn-id: https://svn.dealii.org/trunk@16590
0785d39b -7218-0410-832d-
ea1e28bc413d
bangerth [Mon, 18 Aug 2008 04:52:52 +0000 (04:52 +0000)]
Split existing test into two: the original one, and a redux.
git-svn-id: https://svn.dealii.org/trunk@16589
0785d39b -7218-0410-832d-
ea1e28bc413d
bangerth [Mon, 18 Aug 2008 04:41:16 +0000 (04:41 +0000)]
Use proper header files.
git-svn-id: https://svn.dealii.org/trunk@16588
0785d39b -7218-0410-832d-
ea1e28bc413d
bangerth [Sun, 17 Aug 2008 23:43:15 +0000 (23:43 +0000)]
Narrow down a bit where the problem appears to happen.
git-svn-id: https://svn.dealii.org/trunk@16587
0785d39b -7218-0410-832d-
ea1e28bc413d
kronbichler [Sun, 17 Aug 2008 23:23:39 +0000 (23:23 +0000)]
Since the merge from the anisotropic branch, this program did not work any more because it used GeometryInfo::subfaces_per_cell. This is now replaced by face->n_children().
git-svn-id: https://svn.dealii.org/trunk@16586
0785d39b -7218-0410-832d-
ea1e28bc413d
bangerth [Sun, 17 Aug 2008 23:21:57 +0000 (23:21 +0000)]
Improve clarity of code even further, by putting things into a predicate.
git-svn-id: https://svn.dealii.org/trunk@16585
0785d39b -7218-0410-832d-
ea1e28bc413d
bangerth [Sun, 17 Aug 2008 22:43:22 +0000 (22:43 +0000)]
Simplify code a slight bit.
git-svn-id: https://svn.dealii.org/trunk@16584
0785d39b -7218-0410-832d-
ea1e28bc413d
bangerth [Sun, 17 Aug 2008 22:32:56 +0000 (22:32 +0000)]
Fix typo.
git-svn-id: https://svn.dealii.org/trunk@16583
0785d39b -7218-0410-832d-
ea1e28bc413d
bangerth [Sun, 17 Aug 2008 22:13:34 +0000 (22:13 +0000)]
Copied content has no meaning for this new test.
git-svn-id: https://svn.dealii.org/trunk@16582
0785d39b -7218-0410-832d-
ea1e28bc413d
bangerth [Sun, 17 Aug 2008 22:12:36 +0000 (22:12 +0000)]
Add a test that currently fails inside Triangulation<dim>::prepare_coarsening_and_refinement :-(
git-svn-id: https://svn.dealii.org/trunk@16581
0785d39b -7218-0410-832d-
ea1e28bc413d
bangerth [Sun, 17 Aug 2008 20:59:50 +0000 (20:59 +0000)]
The subversion executable is named 'svn', but 'subversion'.
git-svn-id: https://svn.dealii.org/trunk@16580
0785d39b -7218-0410-832d-
ea1e28bc413d
bangerth [Sun, 17 Aug 2008 20:22:55 +0000 (20:22 +0000)]
Not only test whether we can include mpi.h, but also whether we can link. On charm, for example, the former succeeds, but the latter fails with plaen g++.
git-svn-id: https://svn.dealii.org/trunk@16579
0785d39b -7218-0410-832d-
ea1e28bc413d
bangerth [Sat, 16 Aug 2008 03:07:50 +0000 (03:07 +0000)]
Add some reasoning to the documentation of a mesh smoothing flag.
git-svn-id: https://svn.dealii.org/trunk@16573
0785d39b -7218-0410-832d-
ea1e28bc413d
bangerth [Fri, 15 Aug 2008 23:54:40 +0000 (23:54 +0000)]
Implement adaptivity.
git-svn-id: https://svn.dealii.org/trunk@16571
0785d39b -7218-0410-832d-
ea1e28bc413d
bangerth [Fri, 15 Aug 2008 23:48:08 +0000 (23:48 +0000)]
Fix a few cases that I accidentally got wrong.
git-svn-id: https://svn.dealii.org/trunk@16570
0785d39b -7218-0410-832d-
ea1e28bc413d
bangerth [Fri, 15 Aug 2008 23:39:08 +0000 (23:39 +0000)]
Detect whether the compiler can include <mpi.h> and use it in utilities.h
git-svn-id: https://svn.dealii.org/trunk@16569
0785d39b -7218-0410-832d-
ea1e28bc413d
bangerth [Fri, 15 Aug 2008 23:24:43 +0000 (23:24 +0000)]
Undo accidental checkin.
git-svn-id: https://svn.dealii.org/trunk@16568
0785d39b -7218-0410-832d-
ea1e28bc413d
bangerth [Fri, 15 Aug 2008 23:22:29 +0000 (23:22 +0000)]
Get a lot of cases right where we compare RefinementCase objects that belong to different dimensions.
git-svn-id: https://svn.dealii.org/trunk@16567
0785d39b -7218-0410-832d-
ea1e28bc413d
bangerth [Fri, 15 Aug 2008 21:05:44 +0000 (21:05 +0000)]
New testcase.
git-svn-id: https://svn.dealii.org/trunk@16566
0785d39b -7218-0410-832d-
ea1e28bc413d
bangerth [Fri, 15 Aug 2008 20:41:32 +0000 (20:41 +0000)]
Fix the template argument of the refinement case: We are comparing with face refinement flags, so it needs to be 2, not 3.
git-svn-id: https://svn.dealii.org/trunk@16565
0785d39b -7218-0410-832d-
ea1e28bc413d
bangerth [Fri, 15 Aug 2008 16:09:42 +0000 (16:09 +0000)]
Fix a grammar error.
git-svn-id: https://svn.dealii.org/trunk@16564
0785d39b -7218-0410-832d-
ea1e28bc413d
bangerth [Fri, 15 Aug 2008 16:08:28 +0000 (16:08 +0000)]
Fix a grammar error.
git-svn-id: https://svn.dealii.org/trunk@16563
0785d39b -7218-0410-832d-
ea1e28bc413d
bangerth [Fri, 15 Aug 2008 15:53:03 +0000 (15:53 +0000)]
Fix a problem in the ConstraintMatrix.
git-svn-id: https://svn.dealii.org/trunk@16562
0785d39b -7218-0410-832d-
ea1e28bc413d
kronbichler [Fri, 15 Aug 2008 15:12:23 +0000 (15:12 +0000)]
Add an entry to the news list about the new function I implemented recently.
git-svn-id: https://svn.dealii.org/trunk@16561
0785d39b -7218-0410-832d-
ea1e28bc413d
bangerth [Fri, 15 Aug 2008 04:58:01 +0000 (04:58 +0000)]
Make output patches adaptive on polynomial degree. Remove a number of variables that are no longer needed.
git-svn-id: https://svn.dealii.org/trunk@16560
0785d39b -7218-0410-832d-
ea1e28bc413d
kronbichler [Thu, 14 Aug 2008 23:21:57 +0000 (23:21 +0000)]
Removed some left-over from times when we also had FEFaceValues.
git-svn-id: https://svn.dealii.org/trunk@16559
0785d39b -7218-0410-832d-
ea1e28bc413d
bangerth [Thu, 14 Aug 2008 23:04:21 +0000 (23:04 +0000)]
Remove the function that outputs the artificial viscosity parameter.
git-svn-id: https://svn.dealii.org/trunk@16558
0785d39b -7218-0410-832d-
ea1e28bc413d
bangerth [Thu, 14 Aug 2008 23:03:11 +0000 (23:03 +0000)]
Commit a version in which artificial diffusion appears to work. This version also contains a function that writes the artificial diffusion parameter into the output file.
git-svn-id: https://svn.dealii.org/trunk@16557
0785d39b -7218-0410-832d-
ea1e28bc413d
kronbichler [Thu, 14 Aug 2008 21:06:01 +0000 (21:06 +0000)]
Made the AMG preconditioner class even more generic. Deleted the ILU/UMFPACK typedef in the beginning since we won't need that part any more.
git-svn-id: https://svn.dealii.org/trunk@16556
0785d39b -7218-0410-832d-
ea1e28bc413d
kronbichler [Thu, 14 Aug 2008 20:27:39 +0000 (20:27 +0000)]
Changed interface to Trilinos to make it generic. Now the initialize call provides some options to be able to adapt the preconditioner to the problem.
git-svn-id: https://svn.dealii.org/trunk@16555
0785d39b -7218-0410-832d-
ea1e28bc413d
bangerth [Thu, 14 Aug 2008 18:59:51 +0000 (18:59 +0000)]
Default to cube geometry in 3d. Use proper direction of gravity.
git-svn-id: https://svn.dealii.org/trunk@16554
0785d39b -7218-0410-832d-
ea1e28bc413d
kronbichler [Thu, 14 Aug 2008 17:58:01 +0000 (17:58 +0000)]
Added calculation of artificial viscosity.
git-svn-id: https://svn.dealii.org/trunk@16553
0785d39b -7218-0410-832d-
ea1e28bc413d
bangerth [Thu, 14 Aug 2008 17:22:07 +0000 (17:22 +0000)]
Fix a typo.
git-svn-id: https://svn.dealii.org/trunk@16552
0785d39b -7218-0410-832d-
ea1e28bc413d
bangerth [Thu, 14 Aug 2008 17:12:38 +0000 (17:12 +0000)]
Add the artificial diffusion term to the rhs.
git-svn-id: https://svn.dealii.org/trunk@16551
0785d39b -7218-0410-832d-
ea1e28bc413d
bangerth [Thu, 14 Aug 2008 16:49:41 +0000 (16:49 +0000)]
Switch on BDF2 scheme -- it seems to work.
git-svn-id: https://svn.dealii.org/trunk@16550
0785d39b -7218-0410-832d-
ea1e28bc413d
bangerth [Thu, 14 Aug 2008 16:25:46 +0000 (16:25 +0000)]
Compute BDF-2 matrices.
git-svn-id: https://svn.dealii.org/trunk@16549
0785d39b -7218-0410-832d-
ea1e28bc413d
bangerth [Thu, 14 Aug 2008 16:10:48 +0000 (16:10 +0000)]
Fix computation of lhs matrix: it needs to depend on the time step, but previously did not.
git-svn-id: https://svn.dealii.org/trunk@16548
0785d39b -7218-0410-832d-
ea1e28bc413d
bangerth [Thu, 14 Aug 2008 16:03:33 +0000 (16:03 +0000)]
Fuse two loops.
git-svn-id: https://svn.dealii.org/trunk@16547
0785d39b -7218-0410-832d-
ea1e28bc413d
bangerth [Thu, 14 Aug 2008 15:59:02 +0000 (15:59 +0000)]
Move building the temperature matrix to the function that also builds the right hand side. This ensures that we use the correct time step, and also that we can next build the BDF-2 matrix.
git-svn-id: https://svn.dealii.org/trunk@16546
0785d39b -7218-0410-832d-
ea1e28bc413d
kronbichler [Thu, 14 Aug 2008 15:39:17 +0000 (15:39 +0000)]
One more typo.
git-svn-id: https://svn.dealii.org/trunk@16545
0785d39b -7218-0410-832d-
ea1e28bc413d
kronbichler [Thu, 14 Aug 2008 15:37:54 +0000 (15:37 +0000)]
A typo.
git-svn-id: https://svn.dealii.org/trunk@16544
0785d39b -7218-0410-832d-
ea1e28bc413d
kronbichler [Thu, 14 Aug 2008 15:34:23 +0000 (15:34 +0000)]
Described the evaluation stabilization parameter in more detail.
git-svn-id: https://svn.dealii.org/trunk@16543
0785d39b -7218-0410-832d-
ea1e28bc413d
bangerth [Thu, 14 Aug 2008 13:11:40 +0000 (13:11 +0000)]
Add the variables we need for the BDF-2 time stepping scheme. Remove the code that dealt with inflow boundary conditions for the temperature field -- it didn't yield a contribution so far because the velocity was always tangential to the boundary, but we also can't enforce this sort of boundary condition any more because we now have diffusion and so we need Dirichlet or Neumann b.c. all around.
git-svn-id: https://svn.dealii.org/trunk@16542
0785d39b -7218-0410-832d-
ea1e28bc413d
kronbichler [Thu, 14 Aug 2008 00:38:03 +0000 (00:38 +0000)]
Read most of the comments and added a few things. The section on stabilization will require a few more details on the calculation of the residual...
git-svn-id: https://svn.dealii.org/trunk@16541
0785d39b -7218-0410-832d-
ea1e28bc413d
kronbichler [Wed, 13 Aug 2008 22:44:17 +0000 (22:44 +0000)]
The function memory_consumption now also accounts for the memory consumed by jacobians, jacobian_grads and inverse_jacobians
git-svn-id: https://svn.dealii.org/trunk@16540
0785d39b -7218-0410-832d-
ea1e28bc413d
kronbichler [Wed, 13 Aug 2008 22:05:24 +0000 (22:05 +0000)]
Add the functionality inverse_jacobian to the FEValues class, which gives the inverse Jacobian of the transformation from the real to the unit cell.
git-svn-id: https://svn.dealii.org/trunk@16539
0785d39b -7218-0410-832d-
ea1e28bc413d
bangerth [Wed, 13 Aug 2008 21:27:16 +0000 (21:27 +0000)]
Use a Q(degree) element for the temperature as well. This allows us to get rid of all the face terms in computing the right hand side of the temperature equation. On the other hand, in its current (unstabilized) version, the scheme is not stable for the transport equation, so we need to add a minimal amount of diffusion and ensure that the time step is small enough. The former restriction will be lifted once we get to proper nonlinear stabilization schemes.
git-svn-id: https://svn.dealii.org/trunk@16538
0785d39b -7218-0410-832d-
ea1e28bc413d
bangerth [Wed, 13 Aug 2008 20:07:52 +0000 (20:07 +0000)]
Silence a warning.
git-svn-id: https://svn.dealii.org/trunk@16537
0785d39b -7218-0410-832d-
ea1e28bc413d
bangerth [Wed, 13 Aug 2008 20:03:08 +0000 (20:03 +0000)]
Don't integrate by parts the terms on the temperature rhs.
git-svn-id: https://svn.dealii.org/trunk@16536
0785d39b -7218-0410-832d-
ea1e28bc413d
bangerth [Wed, 13 Aug 2008 19:43:49 +0000 (19:43 +0000)]
Remove one unneeded header file. Remove a check for the size of a matrix element.
git-svn-id: https://svn.dealii.org/trunk@16535
0785d39b -7218-0410-832d-
ea1e28bc413d
kronbichler [Wed, 13 Aug 2008 18:46:35 +0000 (18:46 +0000)]
The Trilinos AMG preconditioner initialization contained an error before that maked it be non-optimal. Moreover, some options are now changed that avoid warnings (and crashes) during the preconditioner build for systems with many entries per row (higher order elements).
git-svn-id: https://svn.dealii.org/trunk@16534
0785d39b -7218-0410-832d-
ea1e28bc413d
bangerth [Wed, 13 Aug 2008 03:20:39 +0000 (03:20 +0000)]
Implement source locations also for 3d.
git-svn-id: https://svn.dealii.org/trunk@16533
0785d39b -7218-0410-832d-
ea1e28bc413d
bangerth [Wed, 13 Aug 2008 03:08:02 +0000 (03:08 +0000)]
Add a comment to the place where we skip copying elements into Trilinos.
git-svn-id: https://svn.dealii.org/trunk@16532
0785d39b -7218-0410-832d-
ea1e28bc413d
bangerth [Wed, 13 Aug 2008 02:46:03 +0000 (02:46 +0000)]
Let the preconditioner have its own sparsity pattern.
git-svn-id: https://svn.dealii.org/trunk@16531
0785d39b -7218-0410-832d-
ea1e28bc413d
bangerth [Wed, 13 Aug 2008 01:19:26 +0000 (01:19 +0000)]
Use a more explicit description of the sparsity pattern.
git-svn-id: https://svn.dealii.org/trunk@16530
0785d39b -7218-0410-832d-
ea1e28bc413d
bangerth [Wed, 13 Aug 2008 01:04:45 +0000 (01:04 +0000)]
No longer build the preconditioner for the pressure mass matrix in the (1,1) block of the system matrix but in the preconditioner_matrix proper. That also means that we can't throw away the preconditioner_matrix at the end of building the preconditioner any more, though.
git-svn-id: https://svn.dealii.org/trunk@16529
0785d39b -7218-0410-832d-
ea1e28bc413d
bangerth [Wed, 13 Aug 2008 00:34:02 +0000 (00:34 +0000)]
Move an assertion to the place where it was probably supposed to be.
git-svn-id: https://svn.dealii.org/trunk@16528
0785d39b -7218-0410-832d-
ea1e28bc413d
bangerth [Tue, 12 Aug 2008 23:56:42 +0000 (23:56 +0000)]
Remove unused variable.
git-svn-id: https://svn.dealii.org/trunk@16527
0785d39b -7218-0410-832d-
ea1e28bc413d
bangerth [Tue, 12 Aug 2008 23:50:23 +0000 (23:50 +0000)]
Only run when UMFPACK is actually installed.
git-svn-id: https://svn.dealii.org/trunk@16526
0785d39b -7218-0410-832d-
ea1e28bc413d
kronbichler [Tue, 12 Aug 2008 23:43:38 +0000 (23:43 +0000)]
Changed the application of the (inverse) preconditioner matrix on the velocity block to an AMG preconditioner based on Trilinos. Made changes in the Makefile to only try to run the program when deal.II is configured with Trilinos.
git-svn-id: https://svn.dealii.org/trunk@16525
0785d39b -7218-0410-832d-
ea1e28bc413d
bangerth [Tue, 12 Aug 2008 20:03:10 +0000 (20:03 +0000)]
Document our failed experiment with block diagonal preconditioners.
git-svn-id: https://svn.dealii.org/trunk@16524
0785d39b -7218-0410-832d-
ea1e28bc413d
bangerth [Tue, 12 Aug 2008 19:15:57 +0000 (19:15 +0000)]
Use a different approach: Build a matrix that contains the vector Laplacian with the no-normal-flux boundary conditions. The boundary conditions do couple the individual components, but only on the boundary.
git-svn-id: https://svn.dealii.org/trunk@16523
0785d39b -7218-0410-832d-
ea1e28bc413d
bangerth [Tue, 12 Aug 2008 18:59:32 +0000 (18:59 +0000)]
Undo last change -- it turns out that this preconditioner only really works when using Dirichlet boundary conditions for the velocity around the entire domain.
git-svn-id: https://svn.dealii.org/trunk@16522
0785d39b -7218-0410-832d-
ea1e28bc413d
bangerth [Tue, 12 Aug 2008 16:17:28 +0000 (16:17 +0000)]
Do the math for variable time steps.
git-svn-id: https://svn.dealii.org/trunk@16521
0785d39b -7218-0410-832d-
ea1e28bc413d
hartmann [Tue, 12 Aug 2008 13:54:10 +0000 (13:54 +0000)]
aniso/hp_constraints_q*08 tests moved to fail.
git-svn-id: https://svn.dealii.org/trunk@16520
0785d39b -7218-0410-832d-
ea1e28bc413d
hartmann [Tue, 12 Aug 2008 13:48:46 +0000 (13:48 +0000)]
aniso/hp_constraints_q*08 tests moved to fail.
git-svn-id: https://svn.dealii.org/trunk@16519
0785d39b -7218-0410-832d-
ea1e28bc413d
bangerth [Tue, 12 Aug 2008 03:00:32 +0000 (03:00 +0000)]
Link to step-31
git-svn-id: https://svn.dealii.org/trunk@16518
0785d39b -7218-0410-832d-
ea1e28bc413d
bangerth [Tue, 12 Aug 2008 02:58:30 +0000 (02:58 +0000)]
Finish section on solvers for the Stokes part.
git-svn-id: https://svn.dealii.org/trunk@16517
0785d39b -7218-0410-832d-
ea1e28bc413d
bangerth [Tue, 12 Aug 2008 01:50:50 +0000 (01:50 +0000)]
Fix a bad link
git-svn-id: https://svn.dealii.org/trunk@16516
0785d39b -7218-0410-832d-
ea1e28bc413d
bangerth [Mon, 11 Aug 2008 23:44:10 +0000 (23:44 +0000)]
Talk about preconditioners a bit.
git-svn-id: https://svn.dealii.org/trunk@16515
0785d39b -7218-0410-832d-
ea1e28bc413d
bangerth [Mon, 11 Aug 2008 23:12:50 +0000 (23:12 +0000)]
Use only the allowed range h3...h6 instead of h2...
git-svn-id: https://svn.dealii.org/trunk@16514
0785d39b -7218-0410-832d-
ea1e28bc413d
bangerth [Mon, 11 Aug 2008 23:04:05 +0000 (23:04 +0000)]
Use A_T instead of A.
git-svn-id: https://svn.dealii.org/trunk@16513
0785d39b -7218-0410-832d-
ea1e28bc413d
bangerth [Mon, 11 Aug 2008 23:03:36 +0000 (23:03 +0000)]
Make it clear what the right hand side of the T equation depends on.
git-svn-id: https://svn.dealii.org/trunk@16512
0785d39b -7218-0410-832d-
ea1e28bc413d
bangerth [Mon, 11 Aug 2008 23:02:04 +0000 (23:02 +0000)]
Document the stabilization approach we are going to take.
git-svn-id: https://svn.dealii.org/trunk@16511
0785d39b -7218-0410-832d-
ea1e28bc413d
In the beginning the Universe was created. This has made a lot of
people very angry and has been widely regarded as a bad move.
Douglas Adams
Typeset in
Trocchi and Trocchi Bold Sans Serif.