Learning Forum

[GridMan]

[Rob]

Check the cell count, you may find as the bodies separate that the mesh is very refined in the gap.

[GridMan]

HiRob,thanks for your answers!As you said, the cell count did increase, but only from three million to four million, does that cause the computation time to increase rapidly from five minutes to two hours?

[Rob]

Not normally, although it can affect the time step size. How much RAM have you got?

[GridMan]

Thanks for your answer Rob, I have solved this problem. The smoothing method I used before was diffusion which would slower and slower. After I changed it to spring method, the calculation speed was greatly improved and the remained stable. It seems that for a large mesh , the diffusion method will significantly slow down the calculation speed.

[Rob]

I suspect as the cell count goes up the number of checks/cell alterations becomes excessive. Depending on how much RAM you have you may then struggle with the amount of compute effort. Thanks for posting a solution.