[hcs129]

Hi everybody
Happy new year
I have a question about the ANSYS tutorial, "Modeling bubbling fluidized bed using DDPM+DEM".
In the tutorial 92,000 parcel are initialized in the bed, with 127 particles in each parcel. As I try to decrease the particle number in parcel, injection duration is decreased from 10e-8 to 10e-10. As a result, each parcel has only 1.27 particle. It cause a totally different gas-solid behavior compared to 127 particles per parcel.
With 127 particles per parcel, bubbling fluidized regime are formed as shown in 1.jpg.
While for 1.27 particle per parcel, particles are blown up to top region and distribute uniformly in the whole box, as shown in 2.png.
To my understanding, they have same particle properties such as particle diameter and density. The particle distributions of them should not differ with each other so obviously.
I have also tried 12.7 particle per parcel. It is similar to the 127 particles percel.
So how to give a proper particle number per parcel and why 1.27 particle per parcel will lead to a unrealistic result.
Any one can help?
C.Hu

[ronithstanly]

I too had the same issue (but not with the same problem statement, but a very close one), you should change the parcel diameter when changing the number of particles contained within the parcel-because in DEM the volume should be conserved (See Page 3: https://mfix.netl.doe.gov/download/m...ix_pic_doc.pdf ). Here, in this case, since you are decreasing the number of particles, the diameter should be decreased by the cube root of the initial diameter (to conserve volume of spheres).

In my case, I also had to change the spring constant of the parcels to get a volume fraction distribution similar to my TFM simulation (Because since the experiment of Kuipers 1992 found in the literature did not mention the spring constant of the glass beads they used, so I initially made a wild guess of k=1000 and that sent the parcels flying to space , then I changed to k=100 and it gave realistic flowfields).

This reply might be a bit late, but I will be glad if this helps you or anyone else who encounters the same problem.

[hossein76]

Hi, I want to do a 4-way coupling in Fluent. Can you help me? Thank you (dem-ddpm)
h.mrad13766@gmail.com

[hossein76]

Originally Posted by ronithstanly

I too had the same issue (but not with the same problem statement, but a very close one), you should change the parcel diameter when changing the number of particles contained within the parcel-because in DEM the volume should be conserved (See Page 3: https://mfix.netl.doe.gov/download/m...ix_pic_doc.pdf ). Here, in this case, since you are decreasing the number of particles, the diameter should be decreased by the cube root of the initial diameter (to conserve volume of spheres).

In my case, I also had to change the spring constant of the parcels to get a volume fraction distribution similar to my TFM simulation (Because since the experiment of Kuipers 1992 found in the literature did not mention the spring constant of the glass beads they used, so I initially made a wild guess of k=1000 and that sent the parcels flying to space , then I changed to k=100 and it gave realistic flowfields).

This reply might be a bit late, but I will be glad if this helps you or anyone else who encounters the same problem.

Hi, I want to do a 4-way coupling in Fluent. Can you help me? Thank you (dem-ddpm)
h.mrad13766@gmail.com