A question about volume fraction of solid

Hello everyone!
Now I’m working on a 2-D modeling of liquid-solid fluidized bed with CFD-DEM.
And recently, I made a numerical simulation. Comparing volume fraction of solid with experiment, I found there’s a deviation between simulation and experiment. The simulation value can’t reach the experimental value. I mainly compared experimental values by changing friction coefficient and restitution coefficient. And now friction coefficient is 0.1,restitution coefficient is 0.95. And then I tried to increase the number of particles about 10400. But nothing much has changed. Anybody knows what should I do with this?(As shown in the figure)

By the way, I wanna try to change the packing way of particles to increase initial volume fraction of solid.
The maximum initial volume fraction of solid which I did can reach 0.5.I intend to write a “particle_input” file about the information of particles to increase initial volume fraction of solid. But I don’t know how to write this .Hope someone can help solve it!!
112
Thanks,
Zihan.

In Mfix, there is a folder labeled “tools” where there is a “DES_Particle_Generator” folder where there is code to create particle input files outputted as “particle_input.dat”. You can fairly easily modify the particle generating code to incorporate whatever packing you desire. Before running your simulation, be sure that your input file denotes GENER_PART_CONFIG = .False. Hope this helps!

The simulated solid volume fraction in the radial direction is always lower than the experimental data, is that because you used a 2D simulation to model a 3D fluidized bed? Can you correctly predict the expansion bed height?
I guess when using a 2D model, the solid maximum packing volume fraction may not able to reach the real maximum packing in a 3D system (say 0.63). Try to increase the cell size to allow more particle in one cell or increase the cell number in the third direction (perpendicular to you 2D bed) to simulate more particles.

Thanks to Andrew, it’s helpful for me. But I still have a problem. In the “DES_Particle_Generator” folder, there are two f files named"DESParticleGen" and"DESParticleGen_MP", can you tell me the difference of both. And how can I use these files to create “particle_input.dat”,thanks again to Andrew.

Firstly, thanks for your reply! Then I wanna illustrate my case simply. The experiment I wanna simulate only told me some parameter of solid and liquid, inlet velocity of liquid and the mass of solid. Through calculating, there would be about 400 thousand particles if I used 3D modelling. It’s beyond the capability of my computer. So I decided to use 2D modelling. The experiment’s initial bed height is about 0.39m, so I simulated with a 0.5 meter bed height and a 0.7 meter bed height. But I found there is no obvious difference between them. As you say,the maximum packing in my simulation is only 0.5. So I think the way(increase the width of z direction ) that you told me may be a good method. However, I have a question. After increasing the width of z direction, it belongs to 2D modelling or 3D modelling?
Thanks again to gaoxi, and can you tell me whether my thought is feasible or not. Appreciate it!

In your 2D simulation (it is actually 3D simulation with one cell in the Z direction, sometimes is called pesudo 2D ), you can increase the cell length in the Z direction and allows more particles in one cell.
Also, you can do a test by setting the inlet gas velocity to ZERO and let the particles pack, then you can determine the effect of cell length (Z direction) on the maximum solid packing volume fraction.