[lammps-users] KE calculation w/ and w/o SHAKE

Dear all,

I have a question regarding my KE calculation results. In one simulation I ran a job without SHAKE algorithm(time step of 0.5fs) and my another job was performed with SHAKE algorithm(time step of 2.0fs). By accident, I found that KE values of the two are about 20-30% difference(the latter has lower value), even though the temperature was maintained the same(323K). I calculated KE using <KE>=3NkT/2 and found that it is equivalent to that of the run without SHAKE.

What would be the reason(s) for the discrepancy? Is the result with SHAKE fine? Or if there is any mistake, how can I resolve the problem?

Any comments/suggestions are highly appreciated.

Thanks in advance,

Dept. of Chem. Eng.
TAMU 3122 Texas A&M University
College Station, TX 77840


When shake is invoked, degrees of freedom are removed
from the system. So the N in your formula,
<KE>=3NkT/2, decreases. The kinetic energy of the
system must, therefore, be lower at a given
temperature than it would be if shake were not