[gmx-users] Periodic Boundary condition

Mark Abraham mark.abraham at anu.edu.au
Tue Feb 5 08:19:57 CET 2008

> Thanks for valuable suggestions.
> The simulation cell has dimension of 5.8nmX5.8nmX3*5.8nm.. And a single
> layer consists of 400 atoms and thus has a dimension of 5.46nmX5.46nm in
> XY plane. Do you think system size is small enough for long-range
> non-bonded interactions and/or velocities to traverse the vacuum gap?

I can't tell from this information. You can, though.

> having the information of surface dimension, can I still use distance
> restraints to hold all my molecules of tetrahydrofuran?

You *can* do it. Whether it's reasonable is a different thing.

> And the physical model is comparable with what people have published.
> Moreover, People have done some monte carlo simulation studies, there they
> have used a repulsive (reflecting) wall to apply periodic boundary
> condition only in a plane.

That is also a kludge to get around an incomplete model of physics. They
should be reporting how much they think that affected the simulation and
how they assessed that. If not, walk away quietly :-)

> And I think not having a proper boundary condition in my system is a cause
> of error in my results.

Well you can just keep making the Z-dimension bigger. The cost of MD
scales with the number of particles, not the size of the box.


More information about the gromacs.org_gmx-users mailing list