Has anyone got any thoughts on what they think is the best CFD package (i.e. ease of use, quality of results, postprocessing, etc.) for multiphase flow modelling? Thanks in advance!

Dear Atholl

I think it is your own code, but, as your question is quite general, I will tell you more.

If you are looking for commercial packages you have a large list at this web page (see sponsors). I have used PHOENICS during my PhD. and it had worked perfectly in my cases. You can add it all the models you want. For instance solid volatilization models (if you are handling solids' combustion) or surface tension if you have two-phase flows.

If you are looking for free codes then try at Universities web pages. I suggest you our web page for two-phase flows free code: www.lmm.jussieu.fr/~zaleski/codes

Regards

Kike

I agree with you kike. currently, i'm developing my own code to solve two-phase flow problem-this is my ph.d.-, as a mathematician i've to do so. but i belive if i were an engineer i'll use commercial packages....

regards dia

The commercial package CFX-4 performs well. Using userfortran-routines you can add any model you like and change almost everything.

Good luck, Astrid

It really depends on what kind of applications you intend. Most vendors offer the standard Eulerian/Eulerian type of multiphase algorithm, but there may be significant differences on how they handle the whole mess of turbulence production or attenuation by particle gathering or dispersal. And if you have reactions that produce a phase that has a significant density difference with the carrier phase convergence can be very difficult to obtain.

Most codes of this type also do not have appropriate viscosity relationships for flows with more than a few volume percent particles. In fact the whole question of constitutive equations for multiphase flow is a black art unless both phases are fluid phases and even that is messy if there is surface tension.

Eulerian/Lagrangian multiphase is very cpu intensive unless the Stokes number is vanishing in which case the particles are fully coupled to the flow.

I don't mean to sound negative, only that it takes a bit of backgroudn reading to even decide which multiphase theory and model is right for your application. See book and papers by C. Crowe.

I was doing two-phase flow calculations with CFX4 and Fluent5.5. If you want to define your own models by user defined functions I think CFX offers much more flexibility to do this.

Tobias