Dear all,
I use LIGGGHTS-PUBLIC 3.8.0 to perform 2D simulations for particle mixing in a rotary drum.
When I use the real coefficient of restitution (0.82) for the mixing material, the real time spent is much higher than I use the default value 0.5 in many other simulation cases.
So I made a test for different coefficients of restitution (CoR) in a 6s 2D simulation. Real time spent for CoR=0.1 is 72 s, for CoR=0.5 is 136 s and for CoR=0.9 is 542 s. Moreover, when I take CoR=0.9 for a longer simulation process like 60 s, the real time increase is not linear (much longer than 10 times of the real time for a 6 s simulation). Attached you can find three log.liggghts files and the mesh I use.
I didn't find such phenomenon in the corresponding 3D simulation.
I wonder if anyone knows the reason.
Regards,
Wu, Wencong
Attachment | Size |
---|---|
![]() | 3.49 KB |