[gmx-users] System expansion

[Oliver Schillinger]

Dear Gromacs users,

I experience a very strange problem.
I have a well equilibrated system consisting of a protein, a ligand, 
ions and water. On most machines I run my simulations on, everything is 
just fine. Except for one compute cluster, where the system undergoes an 
extreme expansion when I use the md-vv integrator for an NPT-simulation. 
To give you an impression I uploaded a short video showing the first 100 
picoseconds of the run to my Dropbox:
https://www.dropbox.com/s/d1ef94bbns0hasr/explode.mpg

The system expands so heavily that it consists mainly of vacuum.
If I switch the integrator from md-vv to md, everything is fine again.
The same simulation (same .mdp file, same starting positions and 
velocities from previous simulation) runs fine with the md-vv integrator 
on our GPU-cluster.

The .mdp file can be downloaded here:
https://www.dropbox.com/s/k9it6eiqb259o59/explode.mdp

I have no idea if it is me who is doing something wrond, a bug in 
Gromacs, a problem with the cluster or something else.
Any suggestions and ideas are appreciated.

Regards,
Oliver

-- 
Oliver Schillinger
Master's student

Forschungszentrum Juelich GmbH
52425 Juelich | Germany

Building 5.8v, Room 3010
Phone: 02461-61-9532

Eingetragen im Handelsregister des Amtsgerichts Dueren Nr. HR B 3498
Vorsitzender des Aufsichtsrats: MinDir Dr. Karl Eugen Huthmacher
Geschaeftsfuehrung: Prof. Dr. Achim Bachem (Vorsitzender),
Karsten Beneke (stellv. Vorsitzender), Prof. Dr.-Ing. Harald Bolt,
Prof. Dr. Sebastian M. Schmidt