Hi Selesta,
yah ..Ray i have realized your argument,it's seems correct,but this things
is happening within few time (.20 ps) so if i want to track few ps,how can i
do, when my system expanded so rapidly(Up = 1.2 Km/s). can i fixed one end
i.e where my wll/piston is placed ? if possible then how?
You need *much* longer sample along the shock direction for that
purpose; that is the nature of NEMD shocks. Yes, you can you "p p fs"
boundary.
another question is to measure shock velocity,please some comments about
that. I have read through the paper you had suggested to Oscar.Is there
other way?
I suggested a paper to Oscar? I don't think so.
You need to read more relevant papers - there is a ton in the
Literature. To measure shock velocity is to measure the velocity of
perturbation in the system, which can be defined by energy, stress,
atom velocity, density and change in neighbors, etc.
BTW your input file is really a help,i go through and already tried to
implement this by my system. But i think it is not clear to me which you are
trying with this five region, and your approach(formulation ) to calculate
run steps.
Five regions are just for monitoring velocities, so that I will know
where the shock wave is. Run steps is just estimated from simple
algebra.
Ray