[FLASH-USERS] Non Converging Simulations

[Hansen, Eddie]

You can increase the max number of iterations and/or increase the tolerance to try to get convergence. Starting with a smaller time step might help smooth things out at first.

Sometimes propaceos files have trouble at lower temperatures. You could try initializing the domain at a slightly higher temperature. Also, making sure the table has a temperature and density grid spanning your simulation's typical values is a good thing to check. Increasing the resolution of that table grid can help too.


******************************************
Eddie Hansen | Research Scientist
Flash Center for Computational Science
Dept. of Physics and Astronomy
University of Rochester, Rochester, NY
Cell 607-341-6126 | flash.rochester.edu
(he / him / his)
******************************************
________________________________
From: flash-users <flash-users-bounces at flash.rochester.edu> on behalf of Kaldawi, Thomas <Tkal at lle.rochester.edu>
Sent: Wednesday, June 28, 2023 11:51:45 AM
To: flash-users at flash.rochester.edu <flash-users at flash.rochester.edu>
Subject: [FLASH-USERS] Non Converging Simulations

Hello,

I am running the LaserSlab simulation with a target, two foils, and a support to hold the foils (shown below). I also added the propaceos patch to read my propaceos EOS and Opacity tables; however, when I run the program with these files, I receive the following error after a while:

refined: total blocks =          218
      12 2.1384E-14 3.1384E-15  ( 1.250E-03,  4.955E-02,   0.00    ) |  3.040E-09 3.554E+84 3.747E+87 0.4000000
      13 2.4523E-14 3.4523E-15  ( 1.250E-03,  4.955E-02,   0.00    ) |  2.970E-09 3.704E+84 3.747E+87 0.4000000
      14 2.7975E-14 3.7975E-15  ( 1.250E-03,  4.955E-02,   0.00    ) |  2.904E-09 3.696E+84 3.747E+87 0.4000000
      15 3.1772E-14 4.1772E-15  ( 1.250E-03,  4.955E-02,   0.00    ) |  2.841E-09 3.492E+84 3.747E+87 0.4000000
  iteration, no. not moved =            0         124
  iteration, no. not moved =            1          52
  iteration, no. not moved =            2           0
 refined: total leaf blocks =          177
 refined: total blocks =          226
      16 3.5950E-14 4.5950E-15  ( 1.250E-03,  4.955E-02,   0.00    ) |  2.780E-09 3.485E+84 3.747E+87 0.4000000
[gr_hypreSolve]: Nonconv./failure ierr=256, descr=[Method did not converge], group no=48, converged=F, |initial RHS|^2=1.3256E-30, num_iterations=500, final_res_norm=4.611224628013419E-004
 [RadTrans] ***** ERROR DETECTED IN Diffuse_solveScalar !!! ****             -1             48              3            256              1              0 F     0


>From what I can see the output looks the same on visit, but I would still like to have all parts of the simulation converge if possible. If anyone else had encountered this issue before any help would be appreciated.

Also, in the flash.par file, I am using the same propaceos file for all of my materials in both the EOS and Opacity sections because I am just testing if it works.

 [cid:9f8f067f-7621-4001-8e01-64dfdfbdd888]
 [cid:1dc5d899-4910-442c-9205-9b35e0888ba0]
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://flash.rochester.edu/pipermail/flash-users/attachments/20230629/52f68a0e/attachment-0001.htm>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: Outlook-eihihxzj
Type: image/jpg
Size: 26528 bytes
Desc: Outlook-eihihxzj
URL: <http://flash.rochester.edu/pipermail/flash-users/attachments/20230629/52f68a0e/attachment-0002.jpg>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: Outlook-tievlwnh
Type: image/jpg
Size: 32011 bytes
Desc: Outlook-tievlwnh
URL: <http://flash.rochester.edu/pipermail/flash-users/attachments/20230629/52f68a0e/attachment-0003.jpg>