problem with using nve and temp/rescale

Dear LAMMPS users,

I am trying to model wetting of water on a solid substrate. I used LJ potential for interaction between water molecules and interaction between water and substrate atoms. Stillinger-Weber potential was used for interaction of atoms in the substrate. I freezed the substrate atoms using setforce command. To integrate the equations of motion, NVE ensemble was used and only temperature of the water was rescaled at 298 K. The problem I faced was that the output temperature of the system was given 150 K on average. Could anyone please explain to me why I get this discrepancy? Am I setting something wrong in my simulation?

Regards,

Farshad

Dear LAMMPS users,

I am trying to model wetting of water on a solid substrate. I used LJ
potential for interaction between water molecules and interaction between
water and substrate atoms. Stillinger-Weber potential was used for
interaction of atoms in the substrate. I freezed the substrate atoms using
setforce command. To integrate the equations of motion, NVE ensemble was
used and only temperature of the water was rescaled at 298 K. The problem I
faced was that the output temperature of the system was given 150 K on
average.

Could anyone please explain to me why I get this discrepancy?

what do you expect, when half of your system is frozen at 0K?

Am I setting something wrong in my simulation?

apart from the temperature accounting thing, using temp/rescale is a
very bad idea.

axel.

You are getting contributions to the temperature from the substrate which is zero. Go through the compute temp command on how to print the temperature of the water.

On a side note, temp/rescale might get you in trouble down the line.