I'm trying to validate a methodology for simulating Raman and IR spectra from classical MD simulations by using dipole and velocity autocorrelation functions. I get some good qualitative results for some cases, but I'm having some difficulties in determining optimal parameters such as simulation time, time step, sampling etc. Can anyone give me suggestions?
Hi Lucas,
Are you using VASP software?
Actually, still I’m a beginner for MD but if you are using VASP software, in the INCAR input file you have NSW and POTIM tags. POTIM is always in femtosecond (fs) and that is the time step. For instance, if you set POTIM = 1, then the time step 1 fs and if you set NSW = 20,000 then
simulation time = NSW * POTIM
= 20,000 * 1 fs
= 20,000 fs
= 20 ns (nanosecond)
- Badges
- Science topic