Hello,
I was thinking of decreasing the strain rate for an uniaxial tensile test which demands an increase of timestep considering same simulation time and adjusting the same intermediate equilibration time periods too. Now for ReaxFF probably the recommended timestep is 0.1 femtoseconds. Will it hamper the simulation if I increase it may be four or eight times by using 0.4 or 0.8 femtoseconds as the timestep.?
Yours Truly
Anubhav Roy
University of Alabama
Department of Aerospace Engineering and Mechanics
Tuscaloosa, U.S.
Hello,
I was thinking of decreasing the strain rate for an uniaxial tensile
test which demands an increase of timestep considering same simulation time
and adjusting the same intermediate equilibration time periods too. Now for
ReaxFF probably the recommended timestep is 0.1 femtoseconds. Will it
hamper the simulation if I increase it may be four or eight times by using
0.4 or 0.8 femtoseconds as the timestep.?
yes. very much so. you have to increase the number of time steps instead.
axel.
Hello,
I am using Periodic BC only. I am using ReaxFF for fracture simulations. Now bigger timestep wont allow me to get high-resolution outputs but if I want to look at the bigger picture, mechanisms should not be affected right?
Hello,
I am using Periodic BC only. I am using ReaxFF for fracture simulations.
Now bigger timestep wont allow me to get high-resolution outputs but if I
want to look at the bigger picture, mechanisms should not be affected
right?
the maximum allowed time step is determined by the physics of your system.
if you increase the time step, beyond that, your simulations will have a
large error or the integration of the equations of motions will fail
completely. for more details on this, please consult a suitable text book
on MD simulations.
axel.
Sure. Thanks a lot. I probably get what you’re saying. Thanks again for your valuable suggestions.
Yours Truly