Is it necessary to apply external pressure in the deformation direction?

User discussions
1

GROMACS version: 2022
GROMACS modification: No

Hi All,

When performing deformation in the z-direction with semi-isotropic coupling, should the reference pressure in the z-direction be set to 1.0 or 0.0? I have set the compressibility in the z-direction to 0.

pcoupl = Parrinello-Rahman
pcoupltype = semiisotropic
tau_p = 1.0
ref_p = 1.0 0.0
compressibility = 4.5e-5 0

@MagnusL Could you please comment on this?

Thank you.

2

It doesn’t really matter what pressure you set in a dimension with 0 compressibility. I usually keep ref_p = 1.0 1.0, but ref_p = 1.0 0.0 should work the same.

3

Yes, it is. I tried it myself and obtained the same result in both cases.

Thanks.

4

Hi Again,

What would be a good value for compressibility to use in the deformation of a viscoelastic material? Does compressibility impact the deformation simulation?

5

Hi,
The value of the compressibility itself depends on the type of material you are simulating, but in general I think we should remind that tunning the compressibility is essentially the same as tuning the relaxation time of the barostat (tau_p and compressibility always go together, see the definition of the inverse mass parameter matrix for P-R). So, from a purely numerical point of view, I would argue you should choose a value of tau_p/compressibility such that the box relaxes in x/y faster than it is strained in z (i.e. it depends on the strain rate you are using). I am not sure whether coupling in z would create instabilities/artifacts, I usually do not couple in the strain direction just to be safe.

6

Thanks a lot for the clarification.

7

Hi Again,

In my system, when strain is zero (t=0), I get a stress value, which I believe is an artifact. Can subtracting this stress value from all the others solve the issue? Or is there any other way to avoid this?