Learning Forum

[bowen]

Hello, when i run the simulation of cone penetraation test with the airflow from the holes, which is located in the cone surface. and there are 4 holes in the shaft. the timesetp for cfd is 200 times of that of DEM. after runing 1.4s,  the particles are suddenly flying. The log shows 2.7% of the particles exceeded 50% the CFD critical explicit time step.

[bowen]

this is the formar screenshot before collapsing, it runs normally. 

[Oguzhan]

 

The warning that you are receiving means that a significant portion of your particles are operating with time steps that are too large relative to the critical explicit time step required for stable CFD coupling. To overcome this, you must decrease the DEM time step of the simulation, which is usually done via Young’s Modulus. To keep the CFD coupling interval reasonable, you might need to decrease the CFD time step as well. Also, having particles larger than the CFD cell size can cause unphysical drag forces on them.

 

[bowen]

Thanks, Oguzhan. you are right. The mesh size is near the same as the particle size near the tip cone. I use CGM model. I know the mesh size should large than particle. Howevr, it will genetrate million particles. The CGM scaling factor is 3 near the tip cone, resulting 903000 particles.  the timestep is autiomaticlly determined by software. and i set the timestep for cfd is 200 times DEM timestep.

I use cfd-dem to investigate the particle ejection under different flow rate. In my previous model, there are no four holes near the shaft. the model can run normally. no warning 'the log shows 2.7% of the particles exceeded 50% the CFD critical explicit time step.'.    when i use the same geometry add the four holes near the shaft. when i run the program. it shows me this warning. 

I guess if it has some problems about the new mesh. I also follow your suggestion, use a decrease cfd time step with 150 times. it still collapses.  
i use laminar seeting. is there something error in some meshes when calculate fluid? 

[Oguzhan]

 

Hi Bowen, I see very high fluid velocities in the screenshot that you provided (~40 m/s). They might help to find out if there’s a problematic region in the mesh. In general try to keep the mesh size around the particle size (with CG) at maximum. 

 

1. Do you also have any problem when you try running this case just on Fluent?
2. You mentioned that you’re using the laminar setting, what are the conditions for the air flow and penetration speed in the test?
3. Could you also copy the Rocky and Fluent log files under this thread?