[gmx-users] RE : long-term drift and crash

[Mark Abraham]

BON Michael wrote:
> Hi and thanks,
> 
> The pressure is stable, the mean value is -280 bar, which sounds normal, and it is isotropic (Pres-XX = Pres-YY = Pres-ZZ = -280 bar)

That value doesn't sound too normal to me, depending on the size of the 
fluctuations and the length of time over which you've measured that 
average, and/or whether the density is reasonable. Perhaps you should 
try an NPT equilibration stage as suggested here 
http://wiki.gromacs.org/index.php/Steps_to_Perform_a_Simulation, in 
order to fix the density.

> I also tried with no constraint (I forgot to mention it in my previous mail) but it didn't solve anything.
> 
> I did a proper energy minimization, there is no warning, error message, wierd energy at any moment. Moreover, I tried to simulate many different systems, doing each time a good minimization, and I always get this strange behaviour.

Mark