simtk.org

Hello,

We ran a simulation with 50000 time steps (restart files each 1000 time steps) on 128 processors, but can't seem to post-process it.

It always stops at the same point while saying "results have been converted" and "no mesh face directory exists":
Done reading (vin plane traction) results : C:/Users/Vitaly/Google Drive/PulmFSI/Results/128-procs_case/restart.40000.17

Could this be a problem in 18th geometry?

Thanks,

Vitaly

Below is the output (I shortened it as best as I could):

Will reduce all available data
Opening C:/Users/Vitaly/Google Drive/PulmFSI/Results/128-procs_case/restart.40000.1 to scan for number of variables...
Opening C:/Users/Vitaly/Google Drive/PulmFSI/Results/128-procs_case/geombc.dat.1 to scan for number of processors and global modes...
Reducing : C:/Users/Vitaly/Google Drive/PulmFSI/Results/128-procs_case/geombc.dat.1
.
.
.
Reducing : C:/Users/Vitaly/Google Drive/PulmFSI/Results/128-procs_case/geombc.dat.128
Number of solution variables found (5)
Number of nodes found (1788000)
Number of elements found (8767051)
Number of processors used (128)
Create vtkPoints with 1788000 nodes.
Create tets for 8767051 elements.
Reducing : wall properties from C:/Users/Vitaly/Google Drive/PulmFSI/Results/128-procs_case/geombc.dat.1
No wall properties found from C:/Users/Vitaly/Google Drive/PulmFSI/Results/128-procs_case/geombc.dat.1. Will search from restart files.
Reducing (solution) results : C:/Users/Vitaly/Google Drive/PulmFSI/Results/128-procs_case/restart.40000.1
Done reading (solution) results : C:/Users/Vitaly/Google Drive/PulmFSI/Results/128-procs_case/restart.40000.1
.
.
.
Reducing (solution) results : C:/Users/Vitaly/Google Drive/PulmFSI/Results/128-procs_case/restart.40000.128
Done reading (solution) results : C:/Users/Vitaly/Google Drive/PulmFSI/Results/128-procs_case/restart.40000.128
Reducing (time derivative of solution) results : C:/Users/Vitaly/Google Drive/PulmFSI/Results/128-procs_case/restart.40000.1
Done reading (time derivative of solution) results : C:/Users/Vitaly/Google Drive/PulmFSI/Results/128-procs_case/restart.40000.1
.
.
.
Reducing (time derivative of solution) results : C:/Users/Vitaly/Google Drive/PulmFSI/Results/128-procs_case/restart.40000.128
Done reading (time derivative of solution) results : C:/Users/Vitaly/Google Drive/PulmFSI/Results/128-procs_case/restart.40000.128
Reducing (displacement) results : C:/Users/Vitaly/Google Drive/PulmFSI/Results/128-procs_case/restart.40000.1
NOTE: No (displacement) in C:/Users/Vitaly/Google Drive/PulmFSI/Results/128-procs_case/restart.40000.1
Reducing (varwallprop) results : C:/Users/Vitaly/Google Drive/PulmFSI/Results/128-procs_case/restart.0.1
NOTE: No (varwallprop) in C:/Users/Vitaly/Google Drive/PulmFSI/Results/128-procs_case/restart.0.1
Reducing (vin plane traction) results : C:/Users/Vitaly/Google Drive/PulmFSI/Results/128-procs_case/restart.40000.1
Done reading (vin plane traction) results : C:/Users/Vitaly/Google Drive/PulmFSI/Results/128-procs_case/restart.40000.1
.
.
.
procs_case/restart.40000.17
Done reading (vin plane traction) results : C:/Users/Vitaly/Google Drive/PulmFSI/Results/128-procs_case/restart.40000.17

Hi Vitaly,

Thank you for your question! Can you run the postsolver with the -sol flag added to the end? This might be an error with how the "vinplane traction" field gets written to the restart files during the solver. Appending the -sol flag to the postsolver will tell it to just reduce velocity and pressure information. If that works, then we can work on a way to get the "vinplane traction" fields back since the wall shear stress is a popular quantity that users like to compute.

Hi Justin,

Yes, we will give it a shot. Thanks!

-Vitaly

Hi Justin,

I am running the postsolver using the GUI in Windows. Is there a way to do what you suggested there?

Thanks,

Vitaly

Hi Vitaly,

Hmm unfortunately I do not know how to do this through the Windows GUI. Did you have a Linux cluster that you ran the simulation on 128 processors on? You can run the postsolver through the command line on the cluster to post-process the simulation results.

Hi Justin,

Thanks for your response. Is there documentation on running postsolver through the command line in Linux? The documentation I see on the website has some info for the solver, but nothing for the postsolver.

Thanks,

Vitaly

Hi Vitaly,

If you run the postsolver with the flag "-h" it should provide a detailed output of the different flags you can provide to get certain outputs. Let us know if you have any questions pertaining to what they mean.

Thanks!

SimVascular Development Team

Hi Justin and Adam,

I've encountered the same problem as Vitaly. As Justin suggested I added -sol flag at the end of svpost on the linux cluster and it did combine them without the error (vin plane traction no available). However, because there is no traction forces, I cannot view shear stress. Could you please suggest a way we can add it back in?

Also, do you know why the restart files are not able to store the "vin plane traction" fields?

Thanks,
Parastou