Dear all it seems quite painful to run particle flow in Fluent5,(1)postprocessings are not enough for study of microscopic,results are not correct(2)coupled (interaction)particle flow seems never going to convergence,(3)CPU time is too long. my case is large ratio converged-diverged nozzle,how to decide either using coupled or not,particle mass rate(void fraction)is refering to inlet size,but refering converged size, it's big, how to juge the particle mass rate is high or not,during interaction iteration, how big is enough for number of iteration of continuous phase per DPM,does coupled calculation no need to reach convergence? Are there any CFD code runing particle flow mush better than Fluent5,and more easy & friently to do UDF? many thanks in advance!

Hi,

I don't have an answer to all your questions that are really case dependent but at least I can help you regarding mass flow.

The continuous phase is not affected by the discrete if the solid mass flow rate of the discrete phase is less than 20% of the continuous one (gas mass flow rate) . So usually, when SOLID_mass/GAS_mass (ratio of mass flow rates) is over 0.2, two way coupling is applicable.

regards,

AB

Hi,alice, Thanks for your help,the problem I met is particle flow through a converged-diverged pipe, refer to inlet area,particle mass rate is 5%, but refer to minimum section of pipe,it will be 50%, limitation in FLuent5 is 10%,if injection is from inlet, when particles are going through minimum section,mass rate will exceed the limitation of Fluent5,so coupled calculation is expected,however,I am still confusing how to simulate it,does Fluent5 can solve this case?

Hi,

Ok I see your problem now, you're right for choosing coupled. I'm sorry that I can't help you much on that, but the cases I deal with are not with a convergent part at the inlet. If you're interested by the flow itself and not by the particle motion or properties the only way I see there is to simulate your case as a flow through porous media. Then you have to define different porous media in the different volume areas (different porosity) of your convergent-divergent geometry. That's all I can say about it, good luck.

Regards,

AB

If your particle/gaz volume fraction is greater than 20% in any part of the domain particle/particle interaction may become an important part of the problem.

I suggest you touse an Euler/Euler approach instead of a dpm approach. Try the eulerian granular model in FLUENT 4 if your particles diameter is important.

For particle trajectories use streackline.

Best regards

Alain