CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   FLUENT (http://www.cfd-online.com/Forums/fluent/)
-   -   cyclone problem (http://www.cfd-online.com/Forums/fluent/29665-cyclone-problem.html)

Jichun April 23, 2002 17:21

cyclone problem
 
Hello, everyone,

Currently, I am using FLUENT to solve a cyclone problem. I know this is a typical problem that FLUENT can handle using ASM model. My question is that for a cyclone, there are two outlets with pressure_outlet boundary condition. When I turn on the multiphase model, I need to set the volumn fractions for the particulate phase at the outlets. The problem is how can I know the volumn fractions for the particulate phase ahead of time? These values should be obtained by solving this problem. Is there anyone could kindly give me some suggestions on this problem? Thanks a lot.

Jin-Wook LEE April 26, 2002 20:58

Re: cyclone problem
 
Of course, the values are obtained by solving your problem if you finish the calculation. You don't need to input them. Instead, initial guess with arbitrary value is enough.

The exact input(volume fraction) is required if there is backflow. In this case(backflow), CFD(or Fluent) require the input because it is actually 'INLET' because of the backflow.

Sincerely, Jinwook

Jichun April 29, 2002 18:08

Re: cyclone problem
 
I really appreciate your response to my post. Just as you said, if there is a backflow occuring in reality I have to treat it's boundary conditions as pressure_inlet type. The thing is: how can I know the volume fraction ahead of time? Can I set whatever values at that boundary or will these values be updated during the calculation? Again, thanks a lot!

Have a nice day!

JiChun

Jin-Wook LEE April 29, 2002 20:43

Re: cyclone problem
 
Physically, it is OUTLET. The best way is to make the computational outlet as 'real outlet'. Then you don't need to input them. Instead, Fluent calculate the value even though you give any value, which is only initial guess.

Otherwise(if there is backflow through the part domain of outlet or full domain of outlet), the computational iutlet is physically partly inlet or fully inlet. Then, unfortunately, there is no way to predict the value. It is nature and it is CFD. As we can simply guess, we should input the all values for inlet, the same situation for the backflow outlet which is physically inlet(partly or fully).

So, if you don't have the boundary condition for the value of outlet where is backflow, you should make the outlet as 'real outlet'. The simple approach to make the outlet as 'real outlet' is to extend the computational domain to the far downstream in order to prevent backflow.

If you live in Korea, you can touch me by, 016-723-8750.

Sincerely, Jinwook


cwflying May 16, 2002 04:43

Re: cyclone problem
 
Could you tell me which boundary conditions you applied to the aircore?


All times are GMT -4. The time now is 17:01.