GROMACS version: 2018
GROMACS modification: No
Dear all,
For a system which contains slab, I have to use semiisotropic pcoupltype with reference pressure of 1 bar, however, the pressure in the system is some order of magnitudes higher that the target reference pressure even after running several ns of equilibration!
I wonder if that is what you would expect to see for such system?
And what should one could do to reduce the pressure closed to 1 bar in such conditions?
Thank you
Alex
Check your pressure in desired direction (e.g. along x, y, or z)
Thanks Masrul.
The reference pressure for all direction is 1.0 bar.
Do you mean, in the case of semiisotropic, no problem to have high pressure in x/y directions and only pressure in z direction is important to reach to the reference pressure? If so, then I guess I reach to that in z direction in different scenarios I have checked, please see the attached figures.
Regards,
Alex
In the first plot, if the x-y compressibility is zero, you’re getting high pressure because the system can’t relax. In the latter two plots, you just haven’t reached equilibrium with respect to the pressure. Heterogeneous systems may require quite a long time to reach a stable state.
Thanks Justin.
That mean everything is fine with the first plot and I can proceed with the production simulation by ignoring the high pressure in the x-y directions? If so, is that a correct approach when solid slab is involved in a system?
I have no experience with solid slabs but a compressibility of exactly zero seems artificial to me. Functionally speaking, a compressibility of zero deactivates the barostat in that direction, so you’re not ever going to be able to achieve the target pressure, therefore the NPT ensemble you simulate is not the same as if the pressure is actually 1 bar.
