Simulation randomly stopping without error

Submitted by Misterblobby on Tue, 10/17/2017 - 03:07

Hi,

Has anyone experienced problems with simulations stopping in what appears to be random points, without returning any errors? To make things more confusing, the point it stops at is not repeatable. All I am using is a modified version of the conveyor tutorial fine included with LIGGGHTS. I am currently using 3.7.0, but had a similar problem with 3.3.0. The issue with 3.3.0 seemed to appear just after upgrading to Ubuntu 16.04, but the current 3.7.0 has been recompiled after upgrading.

It just runs along happily, returning the usual data at each timestep, then randomly stops and returns to an empty command line. Restarting returns an error that one or more particles is integrated more than once, which is also new, but will then run happily until randomly stopping later on.

I am guessing it might be a compilation error, but I have no idea where to start looking.

Help!

Forums:

JoG | Wed, 10/18/2017 - 09:50

Did you dump your particle positions and did you have a look at them? Does the motion of the particles look reasonable before the simulation stops?

Misterblobby | Fri, 10/20/2017 - 02:45

Visually, I can't see anything wrong although there is about 10,000 particles. The strange thing is, two separate runs of the same input script will not stop at the same point.

Misterblobby | Fri, 10/27/2017 - 03:23

Same problem, but this time an error message appeared:

SIGINT/SIGTERM caught - Writing restart on next occasion and quitting after that

Does anyone know what it means?

arnom's picture

arnom | Mon, 11/13/2017 - 17:16

Hi Misterblobby,

can you please provide a full case file (input script, mesh files etc.) so that I can have a look at this.

Best,
Arno

DCS team member & LIGGGHTS(R) core developer