Dear all
In my model, a set of TIP4P waters are enclosed by organic molecules. In order to utilize GPU capability, I made TIP4P by my own by using fix rigid. The input file is attached.
rigid waters and organics are integrated by fix rigid/nvt/small and fix NVT, respectively.
Everything (pressure, temperature and energy) took typical values, however, the aggregate of water molecules gradually moves in a direction, -Z. I used molecule keyword in fix rigid, but I cannot figure out this result?
Sincerely
in.Water_Phenol (1.56 KB)
I believe it is the flying ice cube effect, you can use fix momentum to counter it.
Michal
I believe it is the flying ice cube effect, you can use fix momentum to counter it.
fix momentum may not work for rigid bodies as the rigid fix styles maintain the dynamic properties for the rigid bodies internally and then derives and resets per-atom properties like coordinates and velocities from them.
it may be sufficient to use the velocity command to remove any residual total momentum before and after equilibration. in a well balanced/tuned simulation setting no drift should form after that.
axel.
Dear Axel
I’ve already used “velocity all create $T 123456 rot yes mom yes” in equilibration state, but drift occurs at the same time.
Please re-read my suggestion. You are only implementing half of it.
Axel
Thanks, Axel
But, I can’t get around the problem.
Would you illustrate the workaround, please?
What should I exactly do to avoid molecules drifting?
Thanks, Axel
But, I can’t get around the problem.
Would you illustrate the workaround, please?
What should I exactly do to avoid molecules drifting?
do exactly as i had suggested: use the velocity command to remove the drift before and after equilibration (of course the second time, you cannot use the “create” variant, as that would collide with the equilibration you just did).
axel.
Dear Axel
I exactly pursued your advice in my work, input file attached, however, the drift even occurs in the equilibration stage.
Is it due to the incompatibility of ensembles or any other reasons?
in.Water_Phenol (1.54 KB)
Dear Axel
I exactly pursued your advice in my work, input file attached, however, the drift even occurs in the equilibration stage.
no, you didn’t!
and it doesn’t matter whether there is a (small) drift during equilibration.
Is it due to the incompatibility of ensembles or any other reasons?
right now, the main problem is, that you are not paying attention. i cannot help with that, and i have no more patience left.
good luck,
axel.
Dear Steve
Would you please help me with the following issue?
In my model, a set of TIP4P waters are enclosed by organic molecules. In order to utilize GPU capability, I made TIP4P by my own by using fix rigid. The input file is attached.
rigid waters and organics are integrated by fix rigid/nvt/small and fix NVT, respectively.
Everything (pressure, temperature and energy) took typical values, however, the aggregate of water molecules gradually moves in a direction, -Z. I used molecule keyword in fix rigid, but I cannot figure out this result?
Sincerely