Altair RISE

A program to recognize and reward our most engaged community members

Nominate Yourself Now!

Convergence issues with Acusolve and EDEM coupling

Hi all,

Recently, I have been trying to use Acusolve-EDEM coupling to simulate a case of granular column collapse in water. Unfortunately, despite trying various approaches, the granular column always explodes (or bursts apart) in the water during the initial stages of the coupling. Could anyone please help me with this issue? Attached are some of the simulation files and screenshots.

Many thanks!

Yang

IMGC.slb

Flow.1.Log

Granular column explodes in water.png

Find more posts tagged with

AcuSolve

EDEM

Accepted answers

All comments

Stephen Cole

Hi,

It looks like the time-step in EDEM is 5e-6 s and AcuSolve 5e-5 s? These do seem reasonable values, the time-step ratio of 10:1 also is OK but you could try lowering this ratio (reducing AcuSolve time-step) to help stability.

Also if possible try introducing the EDEM particles more gradually to the system, if you introduce them all at the same time it may be harder to converge.

Regards

Stephen

ydigit

I am yet to open SLB, but based on Log file I am guessing that the elements size for AcuSolve is pretty large.

There are 15309 EDEM particles, for 20k CFD Elements. This does not seem like a good combination.

Please try with smaller mesh size.

Have you been through at least these 3 tutorials before getting started? If not, I would highly recommend solving them, to get hands on experience.

Ken7713

Thank you for your previous advice. I attempted to use a finer mesh resolution, but the simulation still exhibits pressure non-convergence issues. I'm trying to model the free collapse of a granular column in a water-filled tank. My current hypothesis is that the problem may be related to boundary conditions - whenever I apply a free-slip boundary condition at the top of the tank, the pressure fails to converge. However, when I change the top boundary to a pressure outlet condition, the simulation converges, despite knowing this configuration is not physically realistic for my scenario. Is it possible to improve the convergence characteristics by modifying pressure-related parameters within the software? Thank you again for your assistance.

Ken7713

Thank you, Stephen.

I will attempt the operations you suggested. However, as shown in the model, the granular column in my simulation should already be positioned at the bottom of the tank at t=0, so it exists from the very beginning of the coupling. At the moment, I suspect that imposing a no-slip boundary at the top of the tank is causing the pressure to fail to converge, but I cannot use a pressure-outlet boundary. Do you have any alternative recommendations to address this issue? Thank you for your assistance.