Hi Steve
I normally use 1 fs for CNT simulations and all is fine. I have tried a range of timesteps for AIREBO for deposition simulations from 0.1 fs to 1 fs but all not successful. What was intriguing was that if I perform deposition on a graphite plane, the simulation could be stable if I deposit say 1 atom per 100 timesteps (whatever the timestep size I choose). If I reduce the deposition rate to 1 atom per 1000 timesteps, the computation becomes unstable. For diamond structure, I haven't managed to successfully perform deposition without the cell exploding (except for 1 case with a very small unit cell.
Appreciate if you have any advice.
Thanks
Huck Beng