CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   CFX (http://www.cfd-online.com/Forums/cfx/)
-   -   dye simulation (http://www.cfd-online.com/Forums/cfx/22413-dye-simulation.html)

Mike April 7, 2006 10:19

dye simulation
 
Hi everybody,

I am trying to do a simulation of dye in water channel released from a designated area, does anyone know how to do it in CFX 10. I could do particle tracing in transient run. But what I want is the exact modeling which will look like movement of "volume" not "particles" in steady flow. Any help would be very much appreciated.

Thanks alot,

Mike

James April 7, 2006 13:40

Re: dye simulation
 
Hi, Mike

I guesss you want to do a transient tracer dye simulation. If that's you want, just go to the ANSYS CFX forum, and search 'tracer'. You will find some posts descibing every details of tracer simulation.

By the way, it seems that you did some particle tracer simulation in steady flow. Are there special reasons to do that?

Regards!

James

mike April 7, 2006 14:22

Re: dye simulation
 
Hi James,

Thanks alot for your help. Actually what I want is the following. I have some experimental visualization of dye patterns for some period of time in steady state flow condition . And I want to compare my simulation directly with that experiments. What I know with particle tracer, you can get particle motions in flow. However I want volume movement like dye markers. That is the reason, I am looking for that kind of simulation.

Thanks again,

Mike


Glenn Horrocks April 9, 2006 18:36

Re: dye simulation
 
Hi,

If the dye just moves with the fluid I think you will find the additional variable approach better. This uses a Eularian approach rather than Lagrangian and has vastly less computational load than particle tracking as there is coupling of the marker dye to the mass/momentum/heat equations.

It's been a while since I've done one, but I think you declare an additional variable and define it to convect with the flow and you define a diffusivity constant (or just give it 1e-20 if you want it just to convect around with the flow).

Regards, Glenn Horrocks

James April 9, 2006 18:59

Re: dye simulation
 
Hi,

Personally I think 'computational load' may be just one small issue for choosing Eulerian approach or lagrangian approach.

The two approaches have fundamental differences. Lagrangian approach may be good for modelling particle behaviour. For dye simulation, I doubt if good results can be obtained by using lagrangian approach.

Regards!

James


James Date April 10, 2006 05:03

Re: dye simulation
 
The additional variable approach is probably the best way to do this. I think the CFX manual gives an example of this or at least talks about it under the Additional Variable section of the manual.

mike April 10, 2006 10:57

Re: dye simulation
 
Hi Guys, Thanks alot for your help. I will try to use additinal varible as you suggested instead of particle tracer approach. I think that is what I want.

Thanks again,

Mike


All times are GMT -4. The time now is 10:46.