Separation efficiency in an straight through cyclone with water

Cloudseeker · February 9, 2016, 08:00

Hi all,

I am investigating an axial, straight through cyclone. The fluid is water and the particles are sand with a diameter of around 100 micrometer. The swirl is generated by vanes. The sand should accumulate at the outside wall and being removed with a portion of the water stream. We made experiments with a labor model and a simulation using cfx.
The calculated separation efficiency is around 3 times higher than the measured one. The trajectories match pretty well with matlab calculations. I used RSM and turbulent dissipation for the particles in cfx.
I tried to find some reference papers for the problem. But there is not much information on straight through cyclones, either CFD or experiments. The few actual studies are done with air.
Studies show, that particles tend to accumulate at the edges of eddys. Since these studies were done with air, the effect here with water should be even greater. So it is even possible to get accurate results with the current models ?

Thanks for your comments on my problem.

Best Regards
Jürgen

Ps: Sorry for the bad English.

ardismo · February 11, 2016, 04:15

Hi,

For 4 months I only work on cyclone seperators. As I see, results are very dependent to URFs ,spatial discretization order and so on. If you can give a little bit more detail maybe I can help you .

Cloudseeker · February 12, 2016, 04:17

hi,

The vanes generate the swirl, then the tube tube diameter is reduced to accelerate the flow. then comes a straight tube with an inner tube. The outer diameter is around 0.25 m and the length around 0.7 m. after the tube the outer 4 mm are separated from the rest of the flow.
The average axial velocity in the tube is around 3 m/s and the circumferential velocity is 4.5 m/s. At the inlet the average volume fraction of the particles is between 1*10^-5 and 1*10-3 (different cases). Since the VF increases in some areas i tried it with 1-way, 2-way coupling and 2-way + collisions (4-way). The difference are a few percent points in separation-efficiency, but all are around 3 to 4 times the experimental data.

thank you

February 9, 2016, 08:00	Separation efficiency in an straight through cyclone with water	#1
Cloudseeker New Member Join Date: Aug 2014 Posts: 12 Rep Power: 11	Hi all, I am investigating an axial, straight through cyclone. The fluid is water and the particles are sand with a diameter of around 100 micrometer. The swirl is generated by vanes. The sand should accumulate at the outside wall and being removed with a portion of the water stream. We made experiments with a labor model and a simulation using cfx. The calculated separation efficiency is around 3 times higher than the measured one. The trajectories match pretty well with matlab calculations. I used RSM and turbulent dissipation for the particles in cfx. I tried to find some reference papers for the problem. But there is not much information on straight through cyclones, either CFD or experiments. The few actual studies are done with air. Studies show, that particles tend to accumulate at the edges of eddys. Since these studies were done with air, the effect here with water should be even greater. So it is even possible to get accurate results with the current models ? Thanks for your comments on my problem. Best Regards Jürgen Ps: Sorry for the bad English.

Similar Threads
Thread	Thread Starter	Forum	Replies	Last Post
Cyclone separation efficiency using CFX	Benfa	CFX	20	April 9, 2019 13:10
Simulation of rotating undershot water wheel	Miss Anna	CFX	25	June 7, 2016 23:24
Oil water separation	Eric1	CFX	1	September 1, 2010 15:07
Leading edge separation and efficiency	KK	Main CFD Forum	2	December 3, 2008 06:19
Separation phenomenon of water, droplet and air	Raymond	CFX	0	July 9, 2004 22:59

February 11, 2016, 04:15		#2
ardismo New Member uk Join Date: Jul 2012 Posts: 16 Rep Power: 13	Hi, For 4 months I only work on cyclone seperators. As I see, results are very dependent to URFs ,spatial discretization order and so on. If you can give a little bit more detail maybe I can help you .

February 12, 2016, 04:17		#3
Cloudseeker New Member Join Date: Aug 2014 Posts: 12 Rep Power: 11	hi, The vanes generate the swirl, then the tube tube diameter is reduced to accelerate the flow. then comes a straight tube with an inner tube. The outer diameter is around 0.25 m and the length around 0.7 m. after the tube the outer 4 mm are separated from the rest of the flow. The average axial velocity in the tube is around 3 m/s and the circumferential velocity is 4.5 m/s. At the inlet the average volume fraction of the particles is between 110^-5 and 110-3 (different cases). Since the VF increases in some areas i tried it with 1-way, 2-way coupling and 2-way + collisions (4-way). The difference are a few percent points in separation-efficiency, but all are around 3 to 4 times the experimental data. thank you