[gmx-users] Nose-Hoover NPT ensemble problem

[Lianqing Zheng]

Dear all:

When I simulate a system of SiO2 (1536 atoms in a cubic of 2.8664nm) using
Nose-Hoover NPT algorithm, the oscillations of volume, volume velocity and
acceleration get faster and faster, and finally crash the program. At the
following is an example: 
http://bbs.resnet.rochester.edu/~jeckle/md/npt1.jpg

I tried to prepare the system by using Berendsen NPT algorithm and energy
minimiztion. I also tried to shorten time step to 0.1fs and enlarge pistol
mass for volume. But all these only can postpone the time when the program
crashes, but then will not help anymore.

This problem is in my own code, not in Gromacs. Do you have any idea about
what is wrong?

I'd appreciate for you help!

Lianqing