|
[Sponsors] |
![]() |
![]() |
#1 |
New Member
Sony
Join Date: May 2010
Posts: 14
Rep Power: 16 ![]() |
Hello All !
I am using System Coupling in Ansys 14.5 for a 2Way- FSI transient Analysis. I am facing an issue in the restart of the simulations. Every time I restart, after finishing the first iteration of Ansys structural and fluent the simulation crashes display an error as below - "pure virtual method called Terminating/opt/ansys_inc/v145/aisol/.workbench: line 141: 9637 Aborted" Has anyone come across this. Thanks !! |
|
![]() |
![]() |
![]() |
![]() |
#2 |
Senior Member
Join Date: Apr 2009
Posts: 531
Rep Power: 21 ![]() |
What OS are you running on and are Fluent and Ansys structural running in serial, shared parallel or distributed parallel?
|
|
![]() |
![]() |
![]() |
![]() |
#3 |
New Member
Sony
Join Date: May 2010
Posts: 14
Rep Power: 16 ![]() |
Hi Stumpy !
I am using Linux (RHEL) I have tried to restart on two separate Linux machines (one with RHEL 5.7 other with RHEL 6.5) , but in both cases the restart simulation fails after the first coupling iteration and gives the same error (Pure Virtual method called...) Fluent is running parallel across 4nodes(8cpus) each. I am not sure how fluent uses memory (shared or distributed) Ansys is running on 1 node (8cpus) in parallel using the 8 cpus using the memory of the same node. Thanks Last edited by SonyCFD; March 26, 2014 at 19:42. Reason: typo |
|
![]() |
![]() |
![]() |
![]() |
#4 |
Senior Member
Join Date: Apr 2009
Posts: 531
Rep Power: 21 ![]() |
Is the restart consistent with the way the first part of the run was completed? I think Ansys (MAPDL) would be very sensitive to any machine or OS change on a restart. If Ansys was run using the distributed option (even though everything's running on the same machine) then you cannot change the number of CPU's on a restart either. Fluent itself should be robust to changes in OS or number of partitions on a restart, but when combined with System Coupling it may not be. If the number of partitions change then it's possible the node numbering on the FSI interface changes, which would mess up System Coupling. Version 15 is more robust for restarts, but you would not be able to restart a 14.5 case in 15.0; you'd have to run from scratch.
|
|
![]() |
![]() |
![]() |
|
|
![]() |
||||
Thread | Thread Starter | Forum | Replies | Last Post |
Problem in using system coupling in ANSYS 14.5 | yhy20081016 | FLUENT | 1 | April 16, 2016 07:39 |
Problem with Joule-Heating simulation in a high aspect-ratio system | farshadsm | Main CFD Forum | 4 | December 4, 2012 03:09 |
Restarting batch job of system coupling in ANSYS14 | bbhv | ANSYS | 0 | June 25, 2012 08:30 |
CFX11 + Fortran compiler ? | Mohan | CFX | 20 | March 30, 2011 18:56 |
Cool system ,Convection + Radiation problem. | Leo | FLUENT | 1 | February 21, 2001 03:39 |