can CFX simulate centrifugal separator by different density

myleader · February 4, 2017, 04:00

Hello friends, now I meet a problem, hope somebody professional can help me.

as we know, by the gravity, the big density gas would go to lower place, and smaller density gas would go to higher place.

If we replace gravity by centrifugal force, that means the big density gas would go to out side, and the small density gas would go to inner side.

now I want to take a simulation of this phenomenal because the current engineering project is to separate methanol CH3OH and alcohol CH3CH2OH.

I tried to use multi material in the domain, but they don't separate, even I give separated material at inlet, they mix in the flow field and I got mixture at the outlet.

also I tried mix-material but also failed, they don't separate.

and I tried very strong centrifugal force which is equal to 10000 gravity and used hydrogen and alcohol which have much more density difference, but also failed.

so I feel confused, if CFX contains some functions which can evaluate the conservative force and diffusion equation, the result should be good. maybe my operation was not right, can any body give me some guidance?

ghorrocks · February 4, 2017, 05:33

CFX cannot model buoyancy in multicomponent mixtures, that is (for instance) the separation of light and heavy gas fractions which start mixed together.

I suspect this is because the buoyancy force is applied onto the entire control volume, so moves the entire gas content of the control volume together (both the heavy and light fractions) and there is no separation.

What you are looking for is different buoyancy forces per component which can lead to separation. But you cannot have different velocity fields for different components as the multicomponent model in CFX assumes a single velocity field for all components.

The only work-around I can see for this is to try some form of multiphase model. This will allow relative slip between "phases" and therefore allow buoyant separation. But you will have to do some thinking about how you handle things like molecular diffusion which is not modelled in multiphase models. It will not be straight forward - that is why it is not in CFX by default, and I give no guarantees it can be done at all.

February 4, 2017, 04:00	can CFX simulate centrifugal separator by different density	#1
myleader New Member Join Date: Dec 2009 Posts: 9 Rep Power: 16	Hello friends, now I meet a problem, hope somebody professional can help me. as we know, by the gravity, the big density gas would go to lower place, and smaller density gas would go to higher place. If we replace gravity by centrifugal force, that means the big density gas would go to out side, and the small density gas would go to inner side. now I want to take a simulation of this phenomenal because the current engineering project is to separate methanol CH3OH and alcohol CH3CH2OH. I tried to use multi material in the domain, but they don't separate, even I give separated material at inlet, they mix in the flow field and I got mixture at the outlet. also I tried mix-material but also failed, they don't separate. and I tried very strong centrifugal force which is equal to 10000 gravity and used hydrogen and alcohol which have much more density difference, but also failed. so I feel confused, if CFX contains some functions which can evaluate the conservative force and diffusion equation, the result should be good. maybe my operation was not right, can any body give me some guidance?

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Efficiency prediction of centrifugal pumps in CFX	peiji1984	CFX	2	November 19, 2022 10:24
Possible to simulate a working Centrifugal Compressor for jet engine on OpenFOAM?	fazlur	OpenFOAM	1	August 21, 2013 05:30
How to simulate steam cavitation event CFX	JRL4444	Main CFD Forum	2	February 13, 2009 14:33
CFX buoyancy ref density	Luis Angel	Main CFD Forum	0	February 28, 2006 15:06
how to simulate centrifugal pump?	Alex	FLUENT	0	February 21, 2006 17:35

February 4, 2017, 05:33		#2
ghorrocks Super Moderator Glenn Horrocks Join Date: Mar 2009 Location: Sydney, Australia Posts: 17,852 Rep Power: 144	CFX cannot model buoyancy in multicomponent mixtures, that is (for instance) the separation of light and heavy gas fractions which start mixed together. I suspect this is because the buoyancy force is applied onto the entire control volume, so moves the entire gas content of the control volume together (both the heavy and light fractions) and there is no separation. What you are looking for is different buoyancy forces per component which can lead to separation. But you cannot have different velocity fields for different components as the multicomponent model in CFX assumes a single velocity field for all components. The only work-around I can see for this is to try some form of multiphase model. This will allow relative slip between "phases" and therefore allow buoyant separation. But you will have to do some thinking about how you handle things like molecular diffusion which is not modelled in multiphase models. It will not be straight forward - that is why it is not in CFX by default, and I give no guarantees it can be done at all.