Coupling Properties

Claudio Wolfer's picture
Submitted by Claudio Wolfer on Thu, 11/24/2011 - 12:01

Hi

I'm testing the adjustments in the couplingProperties-File. Some questions arose while working.
1) When I use the voidFractionModel centre the momentum coming from the particles in each cell act as one force on the fluid, right? And the momentum acting on the particles is the same for each particle?
2) What is the divided model for? When I try to use this (with dividedProps) I get an error message "type divided, constructor not in hash table". Do I have to compile something again?
3) I guess that the locateModel specifies the algorithm to allocate the particles to the cells. Whats the difference between standard and engine?
4) Last questions for today:-) When is an explicit momentum coupling model useful? What are the entries in the explicitCoupleProps-dictionary?

Thanks for help
Claudio

skyopener's picture

skyopener | Thu, 11/24/2011 - 13:22

hi Claudio,
as far as I know..

1) When I use the voidFractionModel centre the momentum coming from the particles in each cell act as one force on the fluid, right? And the momentum acting on the particles is the same for each particle?
>>>yes.
2) What is the divided model for? When I try to use this (with dividedProps) I get an error message "type divided, constructor not in hash table". Do I have to compile something again?
>>>it's a mysterious keyword in the couplingfile:). perhpas a more accurate submodel to calculate the void fraction..
3) I guess that the locateModel specifies the algorithm to allocate the particles to the cells. Whats the difference between standard and engine?
>>>The locate model gets the cell number of every particle in. it seems that only standard model can be touched in the current released cfdem..
4) Last questions for today:-) When is an explicit momentum coupling model useful? What are the entries in the explicitCoupleProps-dictionary?
>>>explicit coupling and implicit coupling are two methods when taking the effect of solid phase into the control equations of fluid phase. also the implicit one is more robust.
and the explicit coupling can be found in the former version of cfdem.

hope this help...

S.L

Claudio Wolfer's picture

Claudio Wolfer | Thu, 11/24/2011 - 13:40

Hi Chris, hi skyopener

Thanks for your answers. That helps a lot.
Have a nice day
Claudio

cgoniva's picture

cgoniva | Thu, 11/24/2011 - 13:23

Hi Claudio!

>>1) When I use the voidFractionModel centre the momentum coming from the particles in each cell act as one force on the fluid, right?
yes, but that is true for any voidFractionModel.

>> And the momentum acting on the particles is the same for each particle?
no, each particle has a different drag, depending on its velocity (imagine one perticle at rest and one moving, both in the same cell)

"centre" just says that: the volume of one particle is accounted for in the cell the particle centre is located in. As a consequence, the momentum exchange term of this particle is also accounted in that very cell. Nevertheless each particle in a cell can have a different drag.

>>2) What is the divided model for? When I try to use this (with dividedProps) I get an error message "type divided, constructor not in hash table". Do I have to compile something again?
not yet in the release, it basically can distribute the partciles volume and momentum over several cells (as big Particle also can).

>>3) I guess that the locateModel specifies the algorithm to allocate the particles to the cells. Whats the difference between standard and engine?
just specifies the algorithm of finding the corresponing cell a particle is in. (engine is not yet released)

>> 4) Last questions for today:-) When is an explicit momentum coupling model useful?
that qustion has to be answered by you, depending on the physical problem you want to solve. Basically a term depending on the velocity (such as drag) should be treated implicitely.

>>What are the entries in the explicitCoupleProps-dictionary?
just enter nothing and it will tell you step-by-step what is missing.

Cheers,
Chris