Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Windows - Record of all DEMO_ runs #12

Closed
siboles opened this issue Dec 19, 2017 · 3 comments
Closed

Windows - Record of all DEMO_ runs #12

siboles opened this issue Dec 19, 2017 · 3 comments

Comments

@siboles
Copy link

siboles commented Dec 19, 2017

@Kevin-Mattheus-Moerman openjournals/joss-reviews#506

My colleague, @azimj, ran all the demos on a Windows machine. He has logged the command window output in the attached file - run_tests_201712191554.txt.

Seems all passed but two.
DEMO_FEBio_beam_L_force.m (this passed for me)

and

DEMO_FEBio_skin_sphere_indentation_tet4_vs_tet10.m (FEBio Error Termination)

Kevin-Mattheus-Moerman added a commit that referenced this issue Dec 20, 2017
DEMO_FEBio_beam_L_force and DEMO_FEBio_skin_sphere_indentation_tet4_vs_tet10 now have control parameters at the top so users can adjust them to improve model stability. Specifically using full-newton iterations may help.
@Kevin-Mattheus-Moerman
Copy link
Member

Kevin-Mattheus-Moerman commented Dec 20, 2017

Thanks for catching these. I think these are simply FEBio issues. The first model is force controlled and the second involves contact and non-linear elements. Both could benefit from altered FEA control parameters and it may be that they solve in one FEBio version and not in another. For instance I think using full-Newton iterations will help with these. To use that set max_ups=0. See below. I've been running these on the latest FEBio version (2.6.4).

For both demos I've now added a set of FEA control parameters at the top so the user can alter them more easily. There may be other e.g. contact related demos that could be "fragile" in this sense so I'll likely update those with similar settings if this seems to solve it.

%FEA control settings
nSteps=20; %Number of time steps desired
max_refs=25; %Max reforms
max_ups=0; %Set to zero to use full-Newton iterations
opt_iter=10; %Optimum number of iterations
max_retries=5; %Maximum number of retires
dtmin=(1/nSteps)/100; %Minimum time step size
dtmax=1/nSteps; %Maximum time step size

@siboles
Copy link
Author

siboles commented Dec 20, 2017

I would recommend running all the demos in full-Newton. The consistency would be worth the trade-off in efficiency. For contact, I always found facet-to-facet sliding, two pass, using penalty rather than Langrange multipliers to be more robust. Of course the penalty method will exhibit some penetration depending on how high relative to the material properties one sets the value. Likewise, for the biphasic contact sliding2 constraint. I think it is fair to close this issue now though.

@Kevin-Mattheus-Moerman
Copy link
Member

All demos now run in full-Newton mode

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants