fix group and NVT

Hi all

I have a pretty basic question (and may seem dumb) too. In LAMMPS say I have two groups, group A with 100 particles and group B with another 100 particles. If I do fix NVT of group A alone, will the interaction with the particles in group B is considered even if group B is near group A?

The reason why I am asking this is the following

If I have grafts in my system, then which one fo the following is correct

velocity graft_grp set 0.0 0.0 0.0
fix 1 nongraft_grp nvt temp 0.5 0.5 5


fix 1 all nvt temp 0.5 0.5 5
fix 2 graft_grp setforce 0.0 0.0 0.0

I know the above is very basic - but suddenly I got this doubt.


Vaidyanathan M S

You are not thinking this through. Whether particles interact or not is completely different from whether particles move or not. Even if you wipe out the forces on one group of particles, this does not affect the other group.


Because fix nvt does thermostatting and time integration, if both group A and B need to move, you should apply either nvt to both OR nvt to one and nve to another. Mind you, whether or not they are in a fix, the forces between group A and B will always be calculated. The fixes will only influence whether or not the particles’ positions are updated.

Given your input, I think both are equivalent but the first might be faster.

Okay, that was my thought too.
Yes. This exactly was my doubt. So whether they are the group or not, the net force acting on the particle of either groups is determined by sum of all the forces (which is the gradient of potential - which only depends on the RELATIVE position of the particles) - Am I correct? Hence, if I perform NVT on group A, the distance moved by any particle of group A will be given by double integral of the Net Force (from both A and B) groups and if do not apply NVT on group B, the particles in group B DO experience a force, but are not time integrated and hence wont move.

Thanks again for the clarification

I think so, yes. It’s easy enough to double check this yourself, which is probably worth doing.