[gmx-users] Strange pressure coupling behaviour, rapid expanding box

[Justin Lemkul]

On 12/23/16 6:40 AM, Kamps, M. wrote:
> Dear Justin,
>
> Thanks for your reply! I have tried the simulation again but this time
> with opls_140, which is the same Hydrogen atom as is being used in the
> polymer. This does use small values for sigma and epsilon. However
> this made no difference, the box behaved the same.
>
> Since I'm using semiisotropic coupling, and use different values for
> the xy and z compressibility, I tried altering these values. In the
> current situation the xy compressibility is derived from the
> isothermal compressibility of gold, which is 5.77e-7. For the z
> compressibility, I used 4.5e-5, which is the compressibility of water.
> My thoughts were that maybe the surface caused a lot of pressure on
> the system, therefore I greatly increased the compressibility in the
> xy direction, to the point is higher than that of water (5.77e-4), and
> I've decreased the compressibility in the z direction to 4.5e-6. This
> only caused the system to expand less rapidly, but still only expand
> in the z direction. The surface atoms (placed in their optimal
> position), perfectly maintain the xy box.
>
> I'm really unsure on what is causing this error, are there any other ideas?
>

Sounds buggy, but this is a combination of options I have never used, so I'm out 
of ideas.  If you don't want the z-dimension to vary, set its compressibility to 
zero.

> Also, what happens when you set nwall=1? If there is no wall there,
> the atoms can pass through the upper z plane, with just as the other
> periodic planes, however, where do these atoms come back to? Since the
> bottom plane is defined as a wall? How does this work? I can't find
> this answer in the manual.
>

 From http://manual.gromacs.org/documentation/2016.1/user-guide/mdp-options.html 
in the entry for pbc:

"Without walls or with only one wall the system size is infinite in the z 
direction. Therefore pressure coupling or Ewald summation methods can not be 
used. These disadvantages do not apply when two walls are used."

-Justin

-- 
==================================================

Justin A. Lemkul, Ph.D.
Ruth L. Kirschstein NRSA Postdoctoral Fellow

Department of Pharmaceutical Sciences
School of Pharmacy
Health Sciences Facility II, Room 629
University of Maryland, Baltimore
20 Penn St.
Baltimore, MD 21201

jalemkul at outerbanks.umaryland.edu | (410) 706-7441
http://mackerell.umaryland.edu/~jalemkul

==================================================