[gmx-users] Is pressure coupling required in a simulation?

[Mark Abraham]

Darrell Koskinen wrote:
> Hi,
> I was wondering if you could tell me if pressure coupling is always 
> required in a simulation?

Trivially, no. If you want to compare with experimental data from an NPT 
ensemble, then it might be a good idea.

> The reason I am asking is because when I run 
> my simulation with pressure coupling disabled my simulation completes 
> without any warnings, but when I run my simulation with pressure 
> coupling enabled I experience a segmentation fault and am presented with 
> the following message:
> 
> Warning: 1-4 interaction between 27252 and 28244 at distance 2.183 which 
> is larger than the 1-4 table size 2.000 nm
> These are ignored for the rest of the simulation
> This usually means your system is exploding,
> if not, you should increase table-extension in your mdp file

Most likely, you didn't equilibrate at the right density for your 
desired pressure. So, the change in volume required for NPT was too 
large for the integration step you were using. This also suggests your 
NVT simulation might not correlate with reality either. See 
http://www.gromacs.org/Documentation/How-tos/Steps_to_Perform_a_Simulation 
or probably various tutorials.

> In my simulation, I have an infinite graphene sheet parallel to the 
> xy-plane and thought that maybe the pressure coupling is causing extreme 
> pressure to build up within the graphene sheet. Could this be the case?

Certainly, if the pressure coupling is isotropic (or at least on in the 
xy-plane) and the box wants to resize noticeably, you'd have to stretch 
the bonds and that will lead to instability. If the bonds are 
constrained then the system is intrinsically unstable.

> If this is the case, then can I simply run my simulation with pressure 
> coupling turned off or is pressure coupling required?

That's your job to determine given whatever science you're trying to do.

Mark