Dear list,
I am working with CO2 + CH4 mixtures using TraPPE force fields. Since the CO2 is a 3-particle rigid model and the CH4 is a single particle model, I need to simulate a mixture of rigid and non rigid bodies. I’ve tried to carry out my NPT simulations following the three options reported in fix rigid documentation (1 - rigid/npt with dilate all for CO2 and nvt for CH4; 2 - npt with dilate all for CH4 and rigid/nvt for CO2 and; 3 - fix press berendsen for all, rigid/nvt for CO2 and nvt for CH4). All three options gives similar results, with issues in the temperature control (high temperature peaks appears even after long equilibration runs - 10 000 000 fs). I’ve tried to identify the causes by:
- Using hybrid pair style (lj/cut/coul/long for CO2 and lj/cut for CH4);
- Applying smaller time steps (0,1 fs);
- Removing the thermostat and barostat (nve for CH4 and rigid/nve for CO2);
- Adjusting the Tdamp and Pdamp parameters (in a range from 1 to 10 000).
None of those trials succeed and the temperature still reached very high peaks, moving the average to values higher than desired. It is worth mentioning that the pure compounds simulations (with npt for CH4 and rigid/npt CO2) gives reasonable results, comparable to the NIST standard values. Moreover, the mixture simulation goes ok by adding small flexibility to the CO2 molecule and using a single fix npt for all particles. This works even with very large energy constants for bonds and angles, which means that the CO2 molecule stays almost rigid.
Any help would be very appreciated. My input files are attached.
Best regards,
Cassiano Aimoli
in.lammps (1.74 KB)
lammps.data (92.8 KB)