Number of particles in .LOG and paraview less than setting

MFiX
#1

Hi, here is my case, a simple bubbling fluidized bed.

I set 384 particles in GUI
image
However, when I check the .LOG file
image
And in paraview, only 120 particles can be observed(20*6)
image
How should I correct it?
dem_3d_tjh_2021_2_28.zip (36.1 MB)
What’s more, the diameter of particle(dp) is 0.0012 m, while the length of bed in y direction is 0.0048 m(4 dp). But in the following picture, obviously the length of bed in y direction is much longer than 4dp. Where is the wrong?
image

Any suggestions would be greatly appreciated.

#2

You can try to reduce the solid volume fraction in that IC. To maintain the same particle mass, you may need to increase the volume of that IC.

#3

Are you sure those images are not in the x direction? 0.0288/0.0012 = 24. And if this is a paraview screenshot, did you scale the spheres correctly?

#4

The spheres are not scaled properly (You should store the particle diameter in the vtp file). The main issue if the domain size in the y-direction is 4 particle diameters high. The seeding routine will leave a buffer near the wall to guarantee there is no overlap. So here you can have at most 3 particles in the y-direction. If you can spread particles in the z-directions, you should be able to put more particles. You can also manually specify the particles positions in particle_input.dat file and have them touch the wall.

#6

Thanks, f117. 384 particles can be generated by your method. But the particle spacing is too large. I try to use particle_input.dat.

#7

Thanks, onlyjus. I think the spheres are scaled correctly. The y direction is divided into three columns because there needs to be a buffer near the wall.

#8

Thanks, jeff. Since I used the function of automatically generating particles before, I am less likely to manually determine the position of the particles. After writing particle_input.dat file, what should I do in GUI? Or is it enough to put the file in the folder?

#9

Sorry, I think the spheres haven’t been scaled correctly yet. In paraview, I read the pvd file of particles, and set point gaussian, then got the screenshot. What should I do to scale correctly?

#10

I recommend you use the new format for particle_input.dat. See the 2D DEM fluidized bed tutorial for an example. In the Solids>DEM tab, make sure the “Enable automatic particle generation” checkbox is unchecked.

image

#11

You should either manually enter the size of the particles (if uniform) or use the particle diameter data to scale the points (you need to save the diameter in the vtp file). In the “Property” tab, scale the shader using the diameter array, see below. Here the Gaussian radius is a multiplier of the Gaussian Scale array (radius=0.5 * diameter).

image
image594×628 88.3 KB

#12

Jeff, with your help, I understand how to use particle_input.dat and scale particles in paraview correctly now. Thanks a lot!