Hi,
Please see the attached script that uses a pair_style hybrid/overlay combining hertz/stiffness with lj/cut.
The simulation crashes as soon as the first particle touches one of the walls (at timestep ~9400). If I remove the walls, everything runs fine.
Am I doing something wrong in the script or is there a bug in LIGGGHTS that is preventing one to use (granular) walls when applying the pair style hybrid/overlay?
NOTE: This problem was NOT occurring in LIGGGHTS_2.3.8!
Thanks in advance!
Best
Eric
Attachment | Size |
---|---|
![]() | 2.24 KB |
ckloss | Fri, 04/04/2014 - 09:55
Hi eric,
Hi eric,
thanks - I will put it on my list of things to look at for release of 3.0.1 which will happen next week!
rberger | Thu, 04/10/2014 - 12:41
Your bug
Hi Eric,
I just tried to reproduce your bug, both on LIGGGHTS-PUBLIC 3.0 as well as our current development branch 3.0.x. Unfortunately in neither cases I got a crash. We did fix a bug introduced by our current patches concerning hybrid pair styles, so maybe that will also solve your problem. If not, we would need further information from you. E.g. operating system, gcc version, mpi implementation/verison, etc.
Cheers,
Richard