I am running a test simulation (see attached input file) to check energy conservation in liggghts for the case where there is no damping or sliding and only the elastic springs in the normal and tangential direction are active. The damping is set to zero for both normal and tangential contact and the coefficient of friction is set very high to prevent energy loss by sliding. Thus I expect the total energy to remain pretty much constant over time. However the total energy at the end of the simulation is much lower than at the beginning. Any hints of what leads to energy dissipation in this simulation ??
| Attachment | Size |
|---|---|
 Input File | 1.58 KB |
Bias | Tue, 04/16/2019 - 13:57
Tangential sliding
Hi,
If you set the friction to a high value you should get more energy dissipation due to friction, not less. A particle that is sliding "free" would mean no energy dissipation due to tangential friction. The tangential history model gives resistance to sliding by setting a force Ft in the opposite direction if you increase the coefficient of friction you just increase the higher limit for this force Ft.
tapsab | Tue, 04/16/2019 - 15:29
Hi...thanks for responding. I
Hi...thanks for responding. I ran my script with the coefficient of friction to be zero, in addition to all damping coefficients being zero. Still, the energy is not conserved. So there seems to be some source or irreversible dissipation which I am not able to figure out.
Regarding the sliding argument, in liggghts sliding occurs when Ft > mu*Fn. Thus a higher value of mu should ensure that sliding is prevented and instead all energy is stored in the tangential elastic spring. In any event, this is a moot point since as I mentioned in my script I get energy dissipation for both mu = 0 or mu = 100!
Bias | Tue, 04/16/2019 - 16:11
Your right about the spring,
Your right about the spring, sorry for that. But how do you get the potential energy in the springs added to your sum of energy? I didn't see that in the input file.
tapsab | Tue, 04/16/2019 - 16:32
In the first and last step of
In the first and last step of the simulation no particles are in contact with each other thus no potential energy is stored in the springs. Thus the total energy in the initial and final step should just be the kinetic energy of the particles. The collisions occur in the middle of the simulation but as I mentioned, in the absence of damping and sliding, these collisions should not lead to energy dissipation.