Hi, I'm trying to model two phase flow of gas and dense solid particles. Solid particles are stiff and big comparing to gas molecules so they can tolerate pressure separately from gas, therefore there will be two different pressure domains. I've used Eulerian multiphase model but it doesn't work for this problem, because of assuming single pressure domain. Do you have any idea for modeling two phase flow with predefined volume fractions and two unknown pressure domains?

Thanks

What do you mean by "Two different pressure domains"? If you mean packing of the particles is an issue, like in a fluidised bed combustor then CFX has particle collision models to handle this sort of thing. Look in the documentation under "solid particle collision model".

Glenn Horrocks

Thank you for your response,

For clearing the problem, consider a domain of sand with large grain size (about 5 cm) and Volume Fraction = 0.5 which flow like a fluid and has fluid properties like pressure (becuase of buoyancy-hydrostatic pressure) and velocity domain. Now we have another gas phase which flow throw this domain and has complete different and uncoupled velocity and pressure domain. I just want to model the two flows to calculate heat transfer between them. Eulerian multiphase flow by solid particle model seems to be a good solution but as you probably guessed I couldn't make solver to maintain volume fraction 0.5 and because of changing in flow area, particle volume fractions goes very high. I've tried to solve this matter with particle collision model but Gidaspow model (as mentioned in CFX doc) is very stiff (convergence problem) and above that could not restrain volume fraction. I've tried many choices for elastic module and compaction parameter but none of them work as they should. I wonder there is any way to model two flows with known volume fractions, no momentum interaction (separate continuity and N.S. equation for each) and just thermal energy coupling. or could I run two seperate models on CFX that exchnage data during solution so I could transfer one domain thermal conditions to another by external variables?

Regards

Hi,

This sounds like you need a discrete element model for the grains, such as EDEM (http://www.dem-solutions.com/) and link this to CFX for the fluid flow. CFX has only basic capabilities for things like maximum packing fractions so I don't think CFX alone can do what you are trying to do.

Glenn Horrocks

Thanks for your advice,‎ I've take a look at EDEM website, it was magnificent, but unfortunately I don't have it.‎ If I could only do it by writing additional code and maintain the solid volume fraction it ‎would be nice!‎

Regards