RSM-EB simulation : SIGSEGV signal error

Questions and remarks about code_saturne usage
Forum rules
Please read the forum usage recommendations before posting.
Post Reply
kunalck
Posts: 16
Joined: Fri Sep 21, 2018 2:06 pm

RSM-EB simulation : SIGSEGV signal error

Post by kunalck »

Hello All,

I am trying to run a steady state thermal flow simulation with RMS-EB model restarting from the converged solution of the same mesh with k-omega SST model. I am getting following error as I run the simulation.

SIGSEGV signal (forbidden memory area access) intercepted!

Call stack:
1: 0x7f045467920a <cs_restart_read_variables+0x104a> (libsaturne.so.5)
2: 0x7f0454c36c16 <__cs_c_bindings_MOD_restart_read_variables+0x16> (libsaturne.so.5)
3: 0x7f04546d2eba <lecamp_+0xdcd> (libsaturne.so.5)
4: 0x7f04546d2058 <lecamo_+0x69> (libsaturne.so.5)
5: 0x7f04545ac7bf <caltri_+0x12ff> (libsaturne.so.5)
6: 0x7f0454572d3a <cs_run+0x50a> (libsaturne.so.5)
7: 0x7f04545726ef <main+0x12f> (libsaturne.so.5)
8: 0x7f0451583c05 <__libc_start_main+0xf5> (libc.so.6)
9: 0x400b59 <> (cs_solver)
End of stack

Before this , I tried running the simulation with homogeneous condition and it crashed in the first time-step itself. All the leads are appreciated.

Best
Kunal
Yvan Fournier
Posts: 4070
Joined: Mon Feb 20, 2012 3:25 pm

Re: RSM-EB simulation : SIGSEGV signal error

Post by Yvan Fournier »

Hello,

Could you specify the version of the code and system, and ideally provide the other info recommended in this forum's usage recommendations (only the small files at this stage, not the mesh or restart files).

It is a probable that you are encountering a bug which was fixed in April 2018, but providing the exact version number can help me guess. If this is the case, the next bug-fix releases (within a week or 2) will fix it. In the meantime, I could indicate which file to copy from the source tree (on GitHub) to fix this.

Regards,

Yvan
kunalck
Posts: 16
Joined: Fri Sep 21, 2018 2:06 pm

Re: RSM-EB simulation : SIGSEGV signal error

Post by kunalck »

Hello Yvan,

Thank you for your response.

I am using following configuration
Code : Code_Saturne 5.1
Operating System: Red Hat Enterprise Linux
Kernel: Linux 3.10.0-327.18.2.el7.x86_64
Architecture: x86-64

I have attached listing and error to the post
listing.txt
(18.04 KiB) Downloaded 188 times
. Please let me know file to copy from GitHub to fix this.

Best
Kunal
Yvan Fournier
Posts: 4070
Joined: Mon Feb 20, 2012 3:25 pm

Re: RSM-EB simulation : SIGSEGV signal error

Post by Yvan Fournier »

Hello,

You can get the file with the correction from the GitHub repository: https://github.com/code-saturne/code_sa ... _default.c using the "raw" button and then "save as".

Just copy it to your case's SRC directory, and you should be OK (and remove this file when you upgrade to the next release, which should come soon (within 1 or 2 weeks).

Best regards,

Yvan
kunalck
Posts: 16
Joined: Fri Sep 21, 2018 2:06 pm

Re: RSM-EB simulation : SIGSEGV signal error

Post by kunalck »

Hi Yvan ,

Thank you so much for the bug fix. It worked and I am no more facing that error.
Although, I am facing a bit of problem with Rij solver. That error
listing1.txt
(51.04 KiB) Downloaded 187 times
might not have any connection with above bug. But I still wanted to report it, in case it might have some dependency.

Best
Kunal
Yvan Fournier
Posts: 4070
Joined: Mon Feb 20, 2012 3:25 pm

Re: RSM-EB simulation : SIGSEGV signal error

Post by Yvan Fournier »

Hello,

The error is probably different here. You have a very high maximum Courant Number even at the first time step, which is probably what leads to divergence. Which maximum value did you have in the previous (k-omega) computation ? The mean value seems ok, but the max is very high, so you might have a mesh quality issue, or a mesh which is too refined in areas of fast flow. Visualizing that value could help understand which is the case (running just one or two restarted time steps).

Best regards,

Yvan
kunalck
Posts: 16
Joined: Fri Sep 21, 2018 2:06 pm

Re: RSM-EB simulation : SIGSEGV signal error

Post by kunalck »

Hello ,

As mentioned above about Courant number being high, to fix that I am using non-uniform time stepping IDTVAR=2 and also tried improving the mesh. I ran few simulations of which, I am sharing two log files. I have tried using upwind scheme for turbulence which just delayed the crashing of the solution. I would like to get few comments , and tips if possible about meshing criteria for RSM-EB model and its other simulation parameters. Thanks in advance,
Attachments
listing_1.txt
(927.51 KiB) Downloaded 179 times
listing_2.txt
(532.19 KiB) Downloaded 176 times
Post Reply