CFD Online Discussion Forums

CFD Online Discussion Forums (
-   FLUENT (
-   -   Discrete Phase Modeling (

revanth June 18, 2002 15:19

Discrete Phase Modeling
Hi All,

I am Revanth and I am currently woking on Discrete Phase Modeling. I modeled a centrifuge which is very close to having 2 cylinders, one inside the other. The outer cylinder rotating at 100rad/sec(low speed). I meshed it in GAMBIT and then exported it into Fluent. The flow problem was run using water as the fluid at a flow rate of 0.3lit/sec. I wanted to see how different particles behave when sent into the centrifuge. So I started injecting particles using the Injections panel and also, I used surface injection. And that surface is velocity inlet. Now the particle tracks make sense for 100 micron particle. In other words a particles of density as high as steel would get trapped on the wall, whereas for particles of less density like that of a cell(say 1010kg/m3), some escaped while some got trapped. Now when i did the trackings for 10 micron particle, it would give me the same numbers(of the particles escaped & trapped) no matter how high the density is. Now is it that Fluent doesnt track particles of less size like 10 microns or I am doing something wrong. I would really appreciate if people can come up with some ideass regarding this.

Thank You,


Jin-Wook LEE June 19, 2002 04:16

Re: Discrete Phase Modeling
I think that you are correct. But Fluent's result may be also correct, considering your particle density is very similar to that of water, and considering that, for small partilce, the trajectory is very similar to the fluid pathline.

Sincerely Jinwook

revanth June 19, 2002 16:24

Re: Discrete Phase Modeling
Thank you. So how can we get particle tracks of these particles, i mean the 10 micron particles in Fluent? Have anyone of you worked with Discrete Phase Modeling? If yes, then please let me know how to do this?

Thank you,


Jin-Wook LEE June 19, 2002 23:58

Re: Discrete Phase Modeling
I think that your concern was that the particle track was nearly same even though the density of particle was changed. Now, your question seems to see the particle tracks.

So, I can not understand your question exactly. But if you want to see the particle track, you can do it by 'DISPLAY --> PARTICLE-TRACK'. Then you can see the trajectories graphically and you can save the trajectory at the text file.

Sincerely, Jinwook

revanth June 20, 2002 12:22

Re: Discrete Phase Modeling
I am sorry for the confusion. I know how to do particle tracking in FLUENT. My question is, if we inject different 10 micron particles of different densities and they all are giving the same trackings(number escaped and trapped), how can we get the correct results. To be more clear, a 10 micron particle with density 1005 kg/m3 and a 10 micron particle of density 5000kg/m3 are giving the same tackings(number escaped& trapped), which doesnt make sense. Is there any cut off in FLUENT which doesnt allow us to track particles as small as 10 microns?

Thank you


Jin-Wook LEE June 21, 2002 05:35

Re: Discrete Phase Modeling
First of all, I would like to know how many number of particles do you use ?

I mean that, maybe, your results does make sense for small particles even though the density is different. Probably the difference in the real world is very small so that CFD can not detect it with small number of particles. To check this, try with many particles(say thousands of particles). Then, I guess that you may get slightly different result.

Sincerely, Jinwook

revanth June 21, 2002 09:51

Re: Discrete Phase Modeling
Thank you for your reply. Well, as I said earlier I am using SURFACE Injection. And the default number of particles is 104. This is the way I get the results:

Number tracked -- 104

Number escaped -- 40

Number trapped -- 60

Number incomplete--4

I am not sure how to increase the number of particles. And if you have any idea why I am getting Incomplete tracks as said above, please let me know.

Thank You,


Jin-Wook LEE June 23, 2002 20:12

Re: Discrete Phase Modeling
As far as my experience is concerned, density difference of about five times for your case may be big, but may not be big. My personal opinion is that the difference is not so big if your geometry does not have sudden change in flow direction.

To increase the number of injection points, you can create new points or groups or ......

Or, you can conclude that 'the difference is very small' even though you can not conclude that 'the result is exactly same', after couple of numerical tests.

For your test, you can increase the density about 1000~10000 times than the refernece value. Say, 10^6 or 10^7, then you may have different result.

Incomplete calculation is caused by, 1) insufficient 'MAX. NUMBER OF STEPS', or 2) your particle is travelling in the recirculation zone.

Sincerely, Jinwook

david July 18, 2002 05:30

Re: Discrete Phase Modeling
A) Jin-Wook LEE is right. Increase the 'MAX. NUMBER OF STEPS B) May I know your ratio of particulate flow rate (kg/s)/fluid flow rate (kg/s) at a x-section of next to injection? If it is close to zero, then what do you expect ? It's like injection of neutrally buoyant particle. In that case you don't need to inject particles to see the motion of particles !!


All times are GMT -4. The time now is 06:07.