[lammps-users] Granular Tempeature for non dissipative system

Dear LAMMPS user,
I run a simulation bounded with periodic domain, granular particles having no friction, coefficient of restitution=1.0, I gave a initial velocity which generates the 100 units temperatures and then monitor the granular temperature of the system.

I see about 4 unit loss in the temperature value, afterwards it almost maintain a constant value of 96 units throughout the simulation. I did not understand for a completely elastic particles with zero friction why their is decrease in the temperature value (however it is very small).

I am attaching the input file also, which can help someone identifying any error from my side. Also if somebody can explain the reason behind the decrease in the temperature value that would be very helpful.


in.leshear (791 Bytes)

Could you be converting initial kinetic energy to potential (in the contact
springs). Is there any rotational energy you are creating?


Hi Steve,
Yes i can see a little drop in kinetic energy (at very initial time) but the rotational energy remains zero at all time. So you mean to say the some of the kinetic energy is converting to potential energy in contact springs and thats why i am seeing a little drop in the granular temperature?


If you start fully non-overalapped (all KE),
then yes, it is normal that the overlaps will store
some potential energy and reduce the KE.