[Sina R]

Hello everyone
I have a simulation problem and I'm not sure quite what to do, It's a rotating shaft with blades on it used to extrude a mixture of nylon (the one's used in plastic bags) and water.
Should I use CFX or Fluent? Can CFX be coupled with ANSYS ROCKY so I can simulate the nylons? Can fluent handle the rotation of this shaft?

Thanks for your help

[ghorrocks]

It is likely CFX and Fluent can both do this - but we would need more details to be sure. Use the software you prefer to use. They both can couple with Rocky.

[Sina R]

Thank you ghorrocks for your prompt response.

Here are more details about the problem
It's a shaft which acts as both a centrifuge and an extruder, and rotates with an angular velocity of 750 RPM, the geometry is quite complex, it has a main inlet and some other inlets which are nozzles, we need it to deliver the mixture of nylon and water at a rate of 500 kg/s (50% nylon, 50% water), what we need to do is to find the best (optimized) length and angle for the blades on the shaft so we can meet this flow rate (or even more).
Also some of these inlets have steam entering them (110 degrees steam).
So basically we have three phases, water, vapor and the nylon.

If could help me with the coupling of CFX and ROCKY I'd be very grateful.

[ghorrocks]

Where does the vapour come from? Is the water boiling or condensing? Or is it just air as there is a free surface?

Why do you need Rocky for this? Why is the particle tracking in CFX unacceptable?

[Sina R]

The vapor is coming from nozzles, the nozzles are on the top side of this system, they basically inject the vapor downward, by the water if you mean the water near the blades after being mixed with the vapor, it's not boiling, the water is just warm (assume 50 degrees).

The nylon pieces have some qualities like being flexible (basically it's chopped up plastic bags for recycling), which I think is not possible to simulate in CFX or fluent (defining non-spherical particles for FLUENT is quite the hastle), and we need an animation of the motion of these pieces too.

The steam (and overall the energy equation and temperatures) can be neglected for now, the coupling is what matters now.

[ghorrocks]

So it sounds like we have rotating domain with liquid water and water vapour - it sounds like that means this is a free surface multiphase model with phase change. So already this is a complex model. Make sure that you are being realistic about attempting this model - already this is not a topic for beginners, and we have not considered the particles yet.

CFX has a simple model for accounting for non-spherical particles. You just choose a drag model which covers the drag regime you want.

But are you trying to account for the anisotropy caused by the chopped up plastic bags - where they bend and flex with the flow, and this is coupled with the flow? I cannot see any easy way of directly modelling this with any approach. CFX does not have a flexible particle model, and if Rocky has a flexible particle model it will cpuple back to CFX with just a point so the anisotropy of the particles will be lost.

You could model this as an FSI simulation, but modelling a single piece of plastic in the flow is about all you are going to be able to do here. You will not be able to model lots of particles.

Are you being realistic in modelling these flexible particles? Can you see a way of doing it because I cannot. If I was attempting a model like this I would consider developing an empirical model for the particle drag.

Also - how much interphase slip do you get between chopped up plastic bags and water anyway? The nylon is very light and high drag, which means you might not get much slip anyway, which means the details of the drag are not important, so accounting for the flexibility is not going to make any difference to the results so you can ignore it.