forcing the h-FEM adaptivity

[Tomas Svaton]

Hello everybody,

I am blocked on the simple problem. I have the displacement-pressure formulation of the elasticity theory and for the reasons of stability I need to keep the elements order unchanged, otherwise does not converge. How could I force to keep the order of the elements for proper variables. I tried to set the candidate list to H2D_H_ISO , but anyway the polynomial degree grows for some elements.

I will be glad for any hint.

Thank you in advance,

Tomas.

[Lukas Korous]

Hi Tomas,

could you send some output (preferably both screenshots and console output)?

Thanks,
Lukas

> --
> You received this message because you are subscribed to the Google Groups
> "hermes2d" group.
> To unsubscribe from this group and stop receiving emails from it, send an
> email to hermes2d+u...@googlegroups.com.
> To post to this group, send email to herm...@googlegroups.com.
> Visit this group at http://groups.google.com/group/hermes2d.
> For more options, visit https://groups.google.com/d/optout.



--
Lukáš Korous

[Tomas Svaton]

Hi Lukas,

Thank you for your fast response. I have taken the screenshot of the solutions and corresponding meshes after the one iteration of adaptivity. The corresponding output is:

tomas@tomas:~/projects/nl-poissons/04-ln-elast-dist-mixed-h$ ./04-ln-elast-dist-mixed-h 

Reading mesh in XML format.

---- Adaptivity step 1:

Solving on reference mesh.

NonlinearSolver: initial residual norm: 16.6816

NonlinearSolver: Damping factor handling:

convergence test

NonlinearSolver: solution duration: 1.372735 s.

NonlinearSolver: done.

Projecting reference solution on coarse mesh.

OpenGL Warning: Failed to connect to host. Make sure 3D acceleration is enabled for this VM.

ndof_coarse[0]: 255, ndof_fine[0]: 2303

ndof_coarse[1]: 255, ndof_fine[1]: 2303

ndof_coarse[2]: 81, ndof_fine[2]: 1089

ndof_coarse_total: 591, ndof_fine_total: 5695

err_est_rel_total: 64.8168%

Adapting coarse mesh.

Total running time: 3.62201 s

  << close all views to continue >>

Then the solutions, meshes and output after three iterations:

tomas@tomas:~/projects/nl-poissons/04-ln-elast-dist-mixed-h$ ./04-ln-elast-dist-mixed-h 

Reading mesh in XML format.

---- Adaptivity step 1:

Solving on reference mesh.

NonlinearSolver: initial residual norm: 16.6816

NonlinearSolver: Damping factor handling:

convergence test

NonlinearSolver: solution duration: 1.052958 s.

NonlinearSolver: done.

Projecting reference solution on coarse mesh.

OpenGL Warning: Failed to connect to host. Make sure 3D acceleration is enabled for this VM.

ndof_coarse[0]: 255, ndof_fine[0]: 2303

ndof_coarse[1]: 255, ndof_fine[1]: 2303

ndof_coarse[2]: 81, ndof_fine[2]: 1089

ndof_coarse_total: 591, ndof_fine_total: 5695

err_est_rel_total: 64.8168%

Adapting coarse mesh.

---- Adaptivity step 2:

Solving on reference mesh.

NonlinearSolver: initial residual norm: 16.6816

NonlinearSolver: Damping factor handling:

convergence test

NonlinearSolver: solution duration: 1.434071 s.

NonlinearSolver: done.

Projecting reference solution on coarse mesh.

ndof_coarse[0]: 255, ndof_fine[0]: 2407

ndof_coarse[1]: 255, ndof_fine[1]: 2407

ndof_coarse[2]: 289, ndof_fine[2]: 2513

ndof_coarse_total: 799, ndof_fine_total: 7327

err_est_rel_total: 87.7447%

Adapting coarse mesh.

---- Adaptivity step 3:

Solving on reference mesh.

NonlinearSolver: initial residual norm: 16.6816

NonlinearSolver: Damping factor handling:

convergence test

NonlinearSolver: solution duration: 1.418305 s.

NonlinearSolver: done.

Projecting reference solution on coarse mesh.

ndof_coarse[0]: 255, ndof_fine[0]: 2407

ndof_coarse[1]: 255, ndof_fine[1]: 2407

ndof_coarse[2]: 289, ndof_fine[2]: 2513

ndof_coarse_total: 799, ndof_fine_total: 7327

err_est_rel_total: 87.7447%

Adapting coarse mesh.

---- Adaptivity step 4:

Solving on reference mesh.

NonlinearSolver: initial residual norm: 16.6816

NonlinearSolver: Damping factor handling:

convergence test

NonlinearSolver: solution duration: 1.380896 s.

NonlinearSolver: done.

Projecting reference solution on coarse mesh.

ndof_coarse[0]: 255, ndof_fine[0]: 2407

ndof_coarse[1]: 255, ndof_fine[1]: 2407

ndof_coarse[2]: 289, ndof_fine[2]: 2513

ndof_coarse_total: 799, ndof_fine_total: 7327

err_est_rel_total: 87.7447%

Adapting coarse mesh.

---- Adaptivity step 5:

Solving on reference mesh.

NonlinearSolver: initial residual norm: 16.6816

NonlinearSolver: Damping factor handling:

convergence test

NonlinearSolver: solution duration: 1.439120 s.

NonlinearSolver: done.

Projecting reference solution on coarse mesh.

ndof_coarse[0]: 255, ndof_fine[0]: 2407

ndof_coarse[1]: 255, ndof_fine[1]: 2407

ndof_coarse[2]: 289, ndof_fine[2]: 2513

ndof_coarse_total: 799, ndof_fine_total: 7327

err_est_rel_total: 87.7447%

Adapting coarse mesh.

Total running time: 14.4593 s

  << close all views to continue >>

The solution 0 and 1 are correct. The problem is with the solution 2. I need all three meshes the same and the polynomial degree in the certain way. P2 for the solutions 0 and 1 and P1 for the solution 2. Is this possible to achieve with Hermes?

Thank you for your help.

Kind regards,

Tomas.

--
You received this message because you are subscribed to a topic in the Google Groups "hermes2d" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/hermes2d/9N75UujgmzI/unsubscribe.
To unsubscribe from this group and all its topics, send an email to hermes2d+u...@googlegroups.com.

[Lukas Korous]

Hi Tomas,

I believe you can do that - could you send me a link to the source
code? Preferably on github or something.

Thanks,
Lukas

[Tomas Svaton]

Hi Lukas,

Thank you for your response. Could you take a look on the source which is here ? I would really prefer to upload everything to the devel on the git, like I did it the last time. But I met several problems with the upgrading to the new version (git fetch and then git merge merge). 

I switch off all the stuff and when I run cmake . the BSON library is needed. Like in the file "hermesInstOut.txt (4K)" I do not know where to find it. Here is the CMake.vars (2K)

Then I tried just to git in another directory hermes, tutorial and examples, then compiled and installed and everything was all right. Just the problem with launching the test and tutorial examples persists. You can see this message :

tomas@tomas-VirtualBox:~/workspace/hermes3p0-testing/hermes/hermes2d/test_examples/02-poisson-newton$ ./02-poisson-newton 

Using default WeakForm<Scalar>::clone, if you have any dynamically created data in your WeakForm constructor, you need to overload this method!

Newton: initial residual norm: 9230.57

Newton: iteration 1, residual norm: 4.04915e-11

Newton: solution duration: 0.160805 s.

X Error of failed request:  BadRequest (invalid request code or no such operation)

  Major opcode of failed request:  153 (GLX)

  Minor opcode of failed request:  19 (X_GLXQueryServerString)

  Serial number of failed request:  22

  Current serial number in output stream:  22

I am sorry, I am not an advanced user. What you propose to do?

Thank you very much.

Kind regards,

Tomas.

CMake.vars (2K)