Learning Forum

[DrAcAryS]

Hi,

I am simulating a cyclone separator. I wanted to know the efficiency with respect to particle size.

I have applied the suitable conditions:

k-epsilon RNG swirl dominated model, I have used a DPM model with ash-solid as the inert particle. I changed its density as I was concerned with starch.

Reflect condition at inlet, trap at the bottom outlet, escape at the top outlet and reflect conditions at the wall

the problem I am facing is I am getting a higher efficiency each time the DPM model iterates. The cyclone model which I am using has an efficiency of 50 percent at 5 microns. Could you tell me why the efficiency is coming higher each time the model iterates and the simulation is varying from the experimental data. Is it due to the high DPM iterations?

and also as I am concerned with starch is it ok to use ash-solid and change its density or could you suggest other material.

Thanks

[DrAcAryS]

Hi again,
When I have used Calcium carbonate as the inert material and then changed its density to starch i.e 1500 kg/m3 the efficiency is way less i.e 30 percent than that of using ash solid material as the injection and then changing its density to 1500 kg/m3 since starch isn't available. (85%)
Could you tell me which inert material injection will be the most appropriate one?

[Rob]

That doesn't sound right. Assuming you got the model and injection set up correctly you can alter any material to a set density and should get the same answer. Note, for strongly swirling flows I'd use RSM, and k-w with curvature correction is also suitable. Cyclones tend to be fairly sensitive to both mesh and injection setting so please post some images.

[DrAcAryS]

Can you please check if I am doing it in the right way?
I used the k omega standard model with curvature correction. The factor was 1.
The velocity magnitude is 8 m/s and I enabled the Interaction with continuous face and update source flow with every iteration
Boundary conditions:
reflect at inlet
escape at pressure outlet overflow
wall reflect
pressure outlet at underflow trap

most of the particles are incomplete, efficiency is high

[Rob]

Cell quality is OK but I suspect you don't have a good refinement in the cyclone body: you need to make sure you resolve the velocity profile and tets aren't good for that.

[DrAcAryS]

Could you recommend me some refinement techniques or videos

[Rob]

Only what's in the Help. I generally set a fairly small max size and use Fluent Meshing to get a poly mesh.

[DrAcAryS]

Thanks, I will give it a try and see what happens.
So the boundary conditions are correct and I am unable to mesh properly?
So if I get the refinement right I would get the required results?

[Rob]

You'll get the results for the conditions you have set assuming the model is converged. Currently the mesh isn't helping achieve this.

[DrAcAryS]

Hi, I have improved the mesh and model is converged, yet I am getting a higher efficiency than usual, could you please tell me why I am facing these problems