Realising CFD-DEM coupling conserdeing irrgular particle shape, which is the way to go?

Submitted by Rachel on Thu, 04/29/2021 - 03:31

Hi all,

As we can see, until now, the public version of CFD-DEM scheme cannot be used to coupling simulation considering particle shape. To realise this function, the easiest way may be using bonded multi-sphere modes to simulate particle shape, while not using superquadric function. Also, the solid-fluid interaction forces, e.g. drag force model, should also be modified. My confusion are:

From the literature we can see, the projected particle area in the flow direction is a very important factor for the calculation of drag force for irregular particle shapes (clumps). We may need two informationto compute the projected area: 1. the orientation of clumped particles; 2. the flow direction of the seeapge flow.

If anyone can give me any clues about how to realize this function when modifying a drag force model? Any clues or suggestion are greatly welcome.

Thanks so much.

Best,
Rachel

The mentioned literature title:
'Investigating effect of particle shape on suffusion by CFD-DEM modeling' Construction and Building Materials 2021

Forums:

mschramm | Thu, 04/29/2021 - 07:08

Hello,
I believe the Gidaspow Drag Model allows a user to input a constant (Phi?) that can help describe Irregular particle shapes.
The public release does not handle superquadratics.
If you are worried about clumps, look into the Immersed Boundary method (but this requires a very fine mesh).

If these do not work for you, you will need to look into compiling your own drag model. I would start by copying an existing model and adding changes where needed.

Rachel | Thu, 05/20/2021 - 06:56

Thanks for your response!

Based on your suggestion, I believe I have to compile my own drag model for this. To modify the drag force model for irregular particle shape, an important factor is the projected particle area in the flow direction. Would you have any idea about how to obtain such projected area?

Maybe two information would be needed, i.e. the flow direction, and the orientation of the clumped particles. Do you know how to get these two message? Thanks.

Best,
Rachel

mschramm | Thu, 05/20/2021 - 20:58

Hello,
You will need to pass quaternion information of your particle(s) to determine how your particle has changed overtime from your initial clump.
As for the projected area, you will have access to the velocity of the fluid at an atom/superquadratic particle. You can use this fluid property to create a normal plane for your calculations.

This becomes more difficult if your clump can deform (flexible fiber model...)

syq20 | Fri, 09/23/2022 - 08:49

Hi Rachel,

Have you solved the problem? I also need to implement a customized drag model where the projected particle area is needed. Any clues for me?

Best,
syq