Hi Ruiyi,
It's better to try things out on a simple model, just to see if there is some problem related to your SimVascular installation. You should try to upgrade to Windows 10 if you can.
I've created a GitHub Issue for the errors produced by svpre (https://github.com/SimVascular/svSolver/issues/35).
Cheers,
Dave
Deformable wall divergence
-
Ruiyi Yang
- Posts: 20
- Joined: Thu Aug 30, 2018 6:11 am
Re: Deformable wall divergence
Hi Dave,
Thank you for your answer and for the post on GitHub.
I tried to run Example 3 and (unfortunately) it worked fine. It tried what was suggested on GitHub (change nwa value in stannspcg.f). The fatal error however there is a warning error (please see enclosure) :
************************************************************
**************** w a r n i n g e r r o r ****************
************************************************************
routine itcg(1x,'c
failure to converge in itmax iterations
************************************************************
As I have an convergence problem in my simulation I don't think this warning is a good news. Do you have an idea of what may causes this warning and how to correct it ? (I tried to increase itmax but it wasn't a good idea).
Thank you in advance,
Have a nice weekend,
Ruiyi
Thank you for your answer and for the post on GitHub.
I tried to run Example 3 and (unfortunately) it worked fine. It tried what was suggested on GitHub (change nwa value in stannspcg.f). The fatal error however there is a warning error (please see enclosure) :
************************************************************
**************** w a r n i n g e r r o r ****************
************************************************************
routine itcg(1x,'c
failure to converge in itmax iterations
************************************************************
As I have an convergence problem in my simulation I don't think this warning is a good news. Do you have an idea of what may causes this warning and how to correct it ? (I tried to increase itmax but it wasn't a good idea).
Thank you in advance,
Have a nice weekend,
Ruiyi
- Attachments
-
- converge.txt
- (589.25 KiB) Downloaded 19 times
-
David Parker
- Posts: 1632
- Joined: Tue Aug 23, 2005 2:43 pm
Re: Deformable wall divergence
Hi Ruiyi,
The converge.txt file you attached in your last post contained the follow lines
warning: illegal maxChild (964656)
warning: illegal maxChild (964656)
root= 482328 bottom=964656 array_size=964656
I think this means that the restart file from the rigid simulation you are using as ICs in your deformable simulation doesn't have the same number of mesh nodes. Make sure that you are using the same mesh for the rigid and deformable simulations. Also make sure that you convert the results of the rigid simulation to get a restart<NumTimeSteps>.0 file that you can use for your ICs file.
Cheers,
Dave
The converge.txt file you attached in your last post contained the follow lines
warning: illegal maxChild (964656)
warning: illegal maxChild (964656)
root= 482328 bottom=964656 array_size=964656
I think this means that the restart file from the rigid simulation you are using as ICs in your deformable simulation doesn't have the same number of mesh nodes. Make sure that you are using the same mesh for the rigid and deformable simulations. Also make sure that you convert the results of the rigid simulation to get a restart<NumTimeSteps>.0 file that you can use for your ICs file.
Cheers,
Dave
-
Ruiyi Yang
- Posts: 20
- Joined: Thu Aug 30, 2018 6:11 am
Re: Deformable wall divergence
Hi Dave,
Thank you for your answer. I redo the rigid simulation to be sure to have the right mesh and to export the right restart X.0.
The good news is the warning error disappeared, the bad news is the "illegal maxChild" warning is still here when I do the presolve part on deformable wall and even worst the divergence is still here when I do the simulation.
Do you have an idea of what may cause the "illegal maxChild" warning ? Do I have to redo the mesh ?
Thank you in advance,
Have a nice day,
Ruiyi
Thank you for your answer. I redo the rigid simulation to be sure to have the right mesh and to export the right restart X.0.
The good news is the warning error disappeared, the bad news is the "illegal maxChild" warning is still here when I do the presolve part on deformable wall and even worst the divergence is still here when I do the simulation.
Do you have an idea of what may cause the "illegal maxChild" warning ? Do I have to redo the mesh ?
Thank you in advance,
Have a nice day,
Ruiyi
-
Ruiyi Yang
- Posts: 20
- Joined: Thu Aug 30, 2018 6:11 am
Re: Deformable wall divergence
Hi Dave,
Here is a dropbox link with my whole projet : https://www.dropbox.com/s/8oun7roy49kgk ... ar.gz?dl=0
Perhaps, it'll help to have a better view of my problem (my deformable wall simulation doesn't converge whereas my rigid wall simulation does).
If you can help me, I would be very grateful.
Thank you in advance,
Have a nice day,
Ruiyi
Here is a dropbox link with my whole projet : https://www.dropbox.com/s/8oun7roy49kgk ... ar.gz?dl=0
Perhaps, it'll help to have a better view of my problem (my deformable wall simulation doesn't converge whereas my rigid wall simulation does).
If you can help me, I would be very grateful.
Thank you in advance,
Have a nice day,
Ruiyi