Learning Forum

[Mahsa Asadi]

Hello everyone, I am simulating a cyclone separator with the RSM viscous method and DPM. particles with 2mm diameter are injected, 200 particles are injected per time step and my time step size is 0.001s. I used hex dominated mesh method and the size of mesh is 0.005m. The inner vortex does not form and I do not have escape particles. 261151 particles are injected and 160 particles are trapped, does anyone know what is the problem? And is the problem related to the damping factor? Is it similar to the relaxation factor in the "solution method" section?

[Rob]

It could be linked to many things. Please can you post a contour plot of the tangential velocity on the centre x-plane? Then look very closely around the bottom of the cyclone to see where the particles are relative to the viscous sub layer. 

[Mahsa Asadi]

Hello, thank you for your response
I posted three velocity contours and pressure as well, I am simulating a two-way couple and I considered particle-fluid interaction. Is this problem related to the damping factor?
Thank you so much

 

[Rob]

OK, that looks sensible. Have a look at angular velocity too - we're trying to identify the vortex core. 

From the above the cyclone is doing what it's supposed to. So, next step is to see where the particles are. Ie are they getting stuck in the vicous sublayer or in a bulk flow feature? Zoom in on the flow around the base of the cyclone: does the reverse flow in the core trap particles at the base of the cone? 

[Mahsa Asadi]

Could you please tell me how can I see particles' behaviors in the core section? according to the picture below,  particles have not come out in the trap or the escape outlet, and I considered the restitution of coefficient = 0.6, so they should not stuck in cyclone separators I was wondering how particles behave in the core section.

[Rob]

Zoom in at the bottom, and also look down the axis. You may want to switch the view to "orthographic" in Preferences. I'm assuming you set the bottom face as trap or escape? 

[Mahsa Asadi]

I set the bottom face as the trap and the vortex finder outlet as the escape. Is it possible to see particle behavior inside the cyclone near the trap section? because when I zoomed in, I could not see how they turn in the mentioned section (core section)  
I have two more questions: I considered restitution of coefficient = 0.6, It means that particles could not stick to the walls? according to the picture below, particles are stuck near the wall, why is the reason for that?

how can I get a solid volume fraction contour for this cyclone? because in the contour section, there is no solid fraction variable. 

[Rob]

If there's no outlet at the bottom for the fluid, how is the vortex core forming, and where is the fluid going in the bottom of the cyclone? 

0.6 coefficient of resitituion means you'll eventually lose particle momentum and they will flow along the wall boundary rather than bouncing. 

[Mahsa Asadi]

fluid flow should go out from the outlet of the top ( vortex finder outlet) and particles are gone from the bottom outlet. Am I right?

Is it okay to increase the restitution of the coefficient to 0.8 or should I consider polynominal for the mentioned parameter?

[Rob]

Mostly, so you fill the bottom of the cyclone with the powder? 

The coefficient of restitution is a function of the materials, so it's a parameter that can need some tuning. Equally, it's often left as default (1.0) and tends to give good results.