[gmx-users] Pressure control in Gromacs

[Mark Abraham]

Rongliang Wu wrote:
> Hello gmx-users,
> 
> 　　I've tried the demos in the folder tutor, the pressure also fluctuate rather fiercely and its value is too high i think, does it mean that there is something wrong with the force field or pressure units of GROMACS?

No. As I posted in reply to a misplaced thread on gmx-developers last
night...

"An instantanous pressure (however you define it) will necessarily vary
quite a lot during pressure-coupled simulation. A positive pressure
value just means the system has a tendency to expand, and a negative
value means the system will tend to contract. This is exactly what you
should expect - the coupling is (effectively) a dynamic equilibrium with
an infinite body with the reference pressure.

After a proper equilibration period, a P-coupled MD simulation should
have an average pressure that is close to the reference pressure for the
P-coupling."

Much of the problem is in defining an instantaneous, rather than average
pressure. The apparently ferocious variation is totally normal.

Mark