q24.mfx (10.5 KB)
Dear MFiX developers and users,
I am working on a cold bubbling fluidized bed simulation using MFiX-TFM. I would like to ask for advice about geometry generation, mesh resolution, and the problem that the simulated solids are not sufficiently suspended into the upper region of the bed.
My experimental system is a vertical cylindrical bubbling fluidized bed. The main parameters are:
- Bed diameter: 0.06 m
- Bed height: 1.35 m
- Particle material: iron ore
- Mean particle diameter: about 0.8 mm
- Particle density: about 3822 kg/m3
- Gas phase:ideal air
- Model: MFiX-TFM
- Drag model: mainly Gidaspow
- Particle-particle restitution coefficient: 0.9
- Experimental pressure taps are located at approximately 0.21 m, 0.46 m, 0.71 m, and 0.96 m.
In the experiment, the pressure at 0.45 m, 0.7 m, and even 0.95 m still has noticeable values, especially at higher gas flow rates. However, in my simulation, the pressure decreases too rapidly with height. The solids volume fraction above about 0.4 m becomes very low, and the particles mainly remain in the lower part of the bed. As a result, the simulated pressure at the upper pressure taps is much lower than the experimental pressure.
I would like to ask the following questions.
Question 1: Which method is recommended to generate the cylindrical geometry?
Question 2: What mesh size should be used relative to particle diameter?
My current mesh is approximately:
- imax = 14
- jmax = 320
- kmax = 14
Would this be more suitable for this type of bed?
Also, should I refine the mesh in the y direction near the distributor / bed bottom region? The initial packed bed height is only about 0.145 m for one of my cases. The bottom region contains the inlet, distributor effect, and dense solids. I am wondering whether a non-uniform mesh with finer cells near the bottom and coarser cells in the upper freeboard is recommended.
For example, would it be better to use a finer grid in the lower region, such as from y = 0 to y = 0.3 m, and a coarser grid above 0.3 m? Or is a uniform y-direction mesh preferred for numerical stability in TFM?
Question 3: Why are the solids not suspended high enough, and why is the pressure above 0.4 m much lower than experiment?
This is my main problem.
In the experiment, the pressure along the bed height decreases gradually. Even at higher locations such as 0.46 m, 0.71 m, and 0.96 m, there is still a noticeable pressure signal. This suggests that the particles, bubbles, or dilute solids suspension can affect a relatively high region of the bed.
However, in my simulation:
- Solids volume fraction decreases sharply above about 0.4 m.
- Most of the particles remain in the lower part of the bed.
- The pressure at 0.45 m and above is much lower than in the experiment.
- The simulated bed expansion or solids suspension height is clearly lower than expected.
I am trying to understand what may cause this issue.
- Non-spherical iron ore particles
The particles are iron ore, not ideal spherical glass beads or quartz sand. The particles are irregular and may have higher surface roughness. In the current TFM model, I use a spherical particle diameter of about 0.805 mm.
Thank you very much.
