I'm trying to set up a temperature gradient to measure thermal conductivity by non-equilibrium method. Ideally, I would like to do this without any additional programming, but I can't find any obvious options to do it in the DL_POLY manual (using Classic version).
I thought about freezing subsets of the system while thermostatting others, which would work for the initial conditions, but it won't do for a steady-state simulation.
You'll have to code it yourself, see
Non-Equilibrium Molecular Dynamics Simulation of the Thermocapillary Effect
H. A. Maier, Ph. A. Bopp, and M. J. Hampe
Can.J.Chem.Eng. 90, 833 (2012)
MD Simulations of the Soret Effect in Simple Partially Miscible Binary
Biphasic Mixtures
H. A. Maier, M. J. Hampe, and Ph. A. Bopp
Chem.Phys.Let. 518, 55-60 (2011)
As far as I know, there is no simple way of applying different thermostats in one simulation box without doing any programming in DL_Poly.
I had the same problem, applying two thermostats. The programming was a bit challenging, even though I only did it for one special type of thermostats and could exploit some implementation in the code, due to the special structure of my system (rigid and flexible molecules which need to be thermostated to the same temperature separately to decrease the equilibration time significantly and speed up simulations). Probably it is worth a thought of switching the code for this problem. As Ming Ma pointed out probably LAMMPS.
is it possible? how do it in lammps, mabye your trick can use in dl_poly?
sorry ,
did anyone use dl_field to generate the inputs for dl_poly ?
- Badges
- Science topic
- Similar topics
- Chemistry
- Theoretical Chemistry
- Computational Chemistry
- Molecular Dynamics