"fix rigid" NPT stationary hack

I ran your system for 140,000 time steps and did not notice that it
was exploding. (I hope you sent me the correct file.) On the
contrary it was contracting slowly. Because there is no solvent
present in the region where the graphite sheets are located, the
system begins effectively with a large bubble. This bubble grows and
shrinks, causing volume fluctuations. You probably want to prepare
your system more carefully to avoid this problem. I think VMD has a
decent "solvate" extension that should avoid this bubble problem. I
think it works with non-water solvents.

However I did see some differences between our input scripts. I don't
know how significant they are, and I don't know if my way is the right
way but here are the differences between your input script, and the
input scripts Trung and I are using:

Andrew,

Thanks you for taking time to look at my files.

I ran your system for 140,000 time steps and did not notice that it> was exploding. (I hope you sent me the correct file.)

Yup, the one I sent you works. I was asking why if I apply my
thermostat in z alone, it works and why it does not work with the iso
keyword.

I had no problem with these simulations and I get the correct density
for the solvent. I am just baffled why iso does not work even if the
system is periodic in x y and z.

neigh_modify exclude pbcgra pbcgra
neigh_modify exclude pbcgra gra
neigh_modify exclude gra gra

With regards to the neigh_modify command, It is a good idea to do
this but I did not bother to use them as I will free the sheets once
the density of the simulation box has reached equilibrium .Furthermore
the number of solvent atoms is much greater than the number of atoms
in the sheet so most of the calculations will be on the solvent
interactions. Anyway it is still a good idea to do so.

I am not planning to use fix rigid as I will be releasing the sheets.

~ Jan-Michael

Andrew,

Thanks you for taking time to look at my files.

I ran your system for 140,000 time steps and did not notice that it> was exploding. (I hope you sent me the correct file.)

Yup, the one I sent you works. I was asking why if I apply my
thermostat in z alone, it works and why it does not work with the iso
keyword.

I had no problem with these simulations and I get the correct density
for the solvent. I am just baffled why iso does not work even if the
system is periodic in x y and z.

neigh_modify exclude pbcgra pbcgra
neigh_modify exclude pbcgra gra
neigh_modify exclude gra gra

With regards to the neigh_modify command, It is a good idea to do
this but I did not bother to use them as I will free the sheets once
the density of the simulation box has reached equilibrium .Furthermore
the number of solvent atoms is much greater than the number of atoms
in the sheet so most of the calculations will be on the solvent
interactions. Anyway it is still a good idea to do so and was too lazy
to put them.

I am not planning to use fix rigid as I will be releasing the sheets.

Because there is no solvent
present in the region where the graphite sheets are located

There shouldn't be any. The distance between the two sheets is just
around 3.5 Angstrom.

I ran your system for 140,000 time steps and did not notice that it> was exploding. (I hope you sent me the correct file.)

Yup, the one I sent you works. I was asking why if I apply my
thermostat in z alone, it works and why it does not work with the iso
keyword.

I had no problem with these simulations and I get the correct density
for the solvent. I am just baffled why iso does not work even if the
system is periodic in x y and z.

I don't know either, but no that I've taken a look at your system, I
think you were on the right track earlier when you said:

In my system, the particle is not spanning the xy plane. If you look
at the attached jpg file, it almost encompass the xy plan, and maybe
that is why the thermostat doesn't work for the "iso" keyword. This is
my hunch and I don't know how to confirm this.

It's true that you do have a large and very non-spherical solute (a
sheet of graphene) floating around in a narrow box which is initially
too large (solvent too sparse). If during equilibration/compaction,
the system ever gets unhappy due to stress or collision, then (with
"iso") it must expand in all 3 directions (x,y,z) to relieve that
stress which is probably occurring in the xy plane. To get around
that, try the "aniso" keyword (as opposed to "z" or "iso"). If that
fixes the problem, then it's a good sign

Also, I'm assuming you don't want the graphene sheets to interact with
any of their own images in the neighboring cells. If so, you might
want to change the shape of the box to make it a little wider so that
there is more solvent padding around all sides. It sounds like you
realize this too.

Cheers
Andrew