My Community

Hi,
I am interesting in doing ab-initio molecular dynamics simulations using VDW DFT-D2 corrections. However, as I understand vasp uses predictor corrector method which uses both the forces and the wave function. The problem is that DFT-D2 applies only for the energy and forces but not for the wave function. So my question is if it correct to do so.
Alternatively, I think that using IWAVPR=0, solves this problem altough it is writen in the manual not to do that.
Thanks

Hi, from what I know if you are using the Verlet algorithm you should use only positions and velocities. So you may force Verlet instead that predictor-corrector by working on the source code (there must be something to comment out, unfortunately I never looked at that). About IWAVPR=0, I think this would not help you anyhow, and it would make your MD much slower worsening the convergence of each electronic SCF cycle.