Dear all, I am using LAMMPS for coarse grain simulation of polymer chains. I want to control the temperature of the chains using thermostat. Which one should I chose? fit nvt or fix langevin
These polymer chains are expecting random kick form implicit solvent.
Thank you.
It really depends on what you want to obtain in your simulations. Both Noose-Hoover chains and Langevin are able to reproduce canonical ensemble pretty well. However for Langevin, you can implement higher timesteps and to my knowledge, it better resembles the implicit solvent case for big particles like proteins etc.
These protocols are not the options to comment on, which of them is good. Little knowledge on statistical thermodynamics, particularly on ensemble issue, can be helpful how to implement them for doing fairly realistic simulations.
Both are thermostat that make sure that the average kinetic energy or "instantaneous (kinetic) temperature" of the system remain constant as per equipartition theorem. their selection depends on the thermodnamic ensemble in which you are going to simulate your system.
Nose-Hover includes the heat bath explicitly as an additional degree of freedom whereas Langevin method simulates interactions with a solvent. More importantly, two terms added to the equation of motion (i.e, frictional drag on the system moving through a solvent and a random term adds energy from collisions/interactions with solvent molecules).
for detailed understanding, refer standard textboks as suggested above.
- Badges
- Science topic