I would like to use the Eulerian model to model the separation of water and oil in a vessel. When setting the size of the oil droplet I can only set a constant droplet size. Is it possible to put in a size distribution for the secondary phase?

You should be able to define particle diameter as a function using UDF.

See: http://www.engr.subr.edu/Support/flu...153.htm#190982

Hi

ap

From what I gather, you have to specify additional secondary phases for each additional particle size you want to model. This is one of the drawbacks of the Eulerian model. Each additional secondary phase means yet another momentum equation that has to be solved.

Hi, According to what i know distribution size of particle occurs because of the break-up/coalescence. To set a diameter you can use the macro C_PHASE_DIAMETER(c,t). Acording to me this would be helpful olny if you have a simple correlation giving you your droplet diameter depending on the height, volume fraction or something else. Also by setting a secondary phase you consider that the number of droplet is the same in each phase. In fact you consider there is no transfer inbetween due to the break up coaslence phenomena. The correct theory of that is the 'population balance theory' (Not sure if it's the right translation in English thought). I do know this can be done by UDF. Some other CFD code release paper about that too. I dunno think it is yet available as a default option in Fluent. You should ask to your nearest Fluent distributor for this.

Hope this help. thomas

I agree with Thomas. A good approach to model your system (oil in water) should be the use of population balances to keep into account coalescence and break-up. However this is not implemented in FLUENT yet. I know of some work on bubble columns and boiling beds which uses this tecnique (See MFIX Open Citations - Chemistry at www.mfix.org),

The approach proposed by Anton (many secondary phases to reproduce a distribution) has been applied to simulate particle size distribution in gas-solid fluidized systems, obtaining good results using 3 or more phases to represent the solid powder. Anyway, considering your system, where coalescence is an essential phenomenon to be taken into account, this approch doens't seem a good idea.

Hi

ap

Hi,

Thanks for all your suggestions. After speaking with Fluent it seems the only way to model this is to use Antons suggestion of having muliple phases of oil with differnt droplet sizes. This clearly has limitations on computer resources, but is currently the only way at the moment.

Thanks

Maybe FLUENT peple mean it's the only way if you don't want to implement population balances, which should be possible using UDF.

Hi

ap