question of DPM model
 

Hi, everyone I have two questions on the DPM model. 1. If my volume friction is larger than 10%, let say about 40%, can I still use DPM model to get rough results. If not, which kind of model I should use. ( I need to know the partical distribution)

2. How can I know the grid resolution is collect for the DPM model. I have a simple case, with different size of grids. One has 20,000 grids, for this case, I got 2500 particles. The other has 50,000 grids, I got more than 5000 particles.

Many thanks

Re: question of DPM model
 

If your volume fraction is greater than 10% you must use the granular model as DPM won't account for particle-particle interactions that are important.

For the other question, the grid resolution is immaterial.

regards.

CFD Newbie

Re: question of DPM model
 

Thanks for your reply.

Could you tell me that why the grid resolution is not important? how can I tell my result is correct if I got different results by using differnt number of grids? And why I got differnt particle number? Sorry for so many question.

Many Thanks

Cindy

Re: question of DPM model
 

For the sake of the particles, resolution is not important but you must do the grid analysis until you obtain independence for the resolution of the flow equations.

Fluent releases a particle for every cell you have in the surface, so the more cells you have the more particles.

If you analyse impact of particles, look at the percentages instead of the absolute number.

The only way to know if your simulations are OK is by comparing them with quality experimental results.

Cheers

Newbie

Re: question of DPM model
 

Thanks so much for your help.

Have a nice weekend

Cindy

Re: question of DPM model
 

some comments:

--with volume fraction of 40 % you should indeed not use DPM. Granular can be an option, but maybe the particles are not granular. What kind of particles do you have, Cindy? Droplets, powder, ...?

--CFD Newbie tells Fluent releases one particle per cell. That's indeed the case with a surface-injection (and more correctly: it's one particle per face), but there are numerous other possibilities. You can also create an injection-file and release as much particles as you want.

--The grid resolution needed for good results have indeed nothing to do with the dpm-model. If you don't have experience with modelling for your application, you should test several grids and compare the results.

good luck,

Laika, still orbiting

Re: question of DPM model
 

where is Granular model?

Re: question of DPM model
 

it's one of the options for Eulerian multiphase simulation.

Laika, still orbiting

Re: question of DPM model
 

Hi Newbie, you said that fluent releases particles at every cell? Does not fluent release particles on the surface that you define (for instance, inlet surface?)