CFD Online Discussion Forums

CFD Online Discussion Forums (
-   CFX (
-   -   Hydrocyclone and its outlets (

KK March 2, 2008 03:52

Hydrocyclone and its outlets

I want to simulate a hydrocyclone whose outlets are open to atmosphere,It is a general phenomenon that these outlets will draw the air in to the cyclone because of low pressure region created and hence an air core will exist.

Having said that, now I want to know how to set up the problem in CFX-11?

1) I need to specify these outlets as 'opening' type boundary conditions. right? (so that air can enter)

2) As the fluid region ends at these 'opening' type boundary conditions, how could I specify that air is available to be sucked in? I guess 'opening' requires to specify volume fractions for multiphase flow, how could I know it before hand ?

Please help, Thanks, KK

Glenn Horrocks March 2, 2008 18:03

Re: Hydrocyclone and its outlets

1) If the flow is entirely into the domain you should use inlets.

2) Set the total pressure of the inlet at atmospheric pressure, or maybe a little below it if there are entry losses. You know the volume fractions at the inlet - air=1, fluid=0.

Glenn Horrocks

Kk March 2, 2008 18:09

Re: Hydrocyclone and its outlets
Thanks Glenn,

1) The flow will not be entirely in to the domain at these boundaries. Because of radial pressure distribution, area near the axis will draw the lighter fluid in (air) and the area near the circumference will make the fluid (water) flowing out of the domain. Hence it should be an 'opening' as per my understanding.

2) Now, how can I decide about volume fractions at these openings?

Many Thanks, KK

Glenn Horrocks March 3, 2008 17:44

Re: Hydrocyclone and its outlets

1) Flow direction is in both direction - then you have to use openings. Be aware that these boundaries are less numerically stable so it may be harder to converge. It may be worthwhile considering moving the boundary further away to a region where the flow is entirely in one direction.

2) Volume fractions - Unless you know the volume fraction at the boundary location you can't use the location! Again, if you move the boundaries further away to a region where the volume fractions are known then it can be correctly specified.

Glenn Horrocks

All times are GMT -4. The time now is 23:40.