CFD Online Discussion Forums

CFD Online Discussion Forums (
-   CFX (
-   -   Prob in Particle Tracking (

sakurabogoda March 1, 2013 09:56

Prob in Particle Tracking
Dear All,

I have run a particle tracking simulation in a cyclone separator and got the result file. But in CFX post the particle track file shows that it reached up to small distance (in inlet area/ a 600mm long tube) and haven't entered to cyclone body.

The total simulation time was 10s and 5e-5 time steps.

What is the reason with not showing particle tracks? Is this means my particles haven't tracked?

Can anybody pls help me?

ghorrocks March 3, 2013 06:34

Check the particle fate termination criteria. It might be terminating the tracks too early. Also make sure you are visualing the full particle track.

sakurabogoda March 4, 2013 01:21

3 Attachment(s)
Dear Glenn,

Thank you very much for reply. there was a prob with file transfer from cluster to PC. Now I can view the full particle track.

I have another problem now. in out put file it shows particles do not exit from domain.

| Time Integrated Particle Total Mass Flow Summary |

Particle 8um
inlet 6.1375E-10

Particle 5um
inlet 9.9894E-11
oultet -7.2466E-14

Particle 2um
inlet 3.1966E-12
oultet -1.6812E-14

Particle 1um
inlet 3.9958E-13
oultet -4.4929E-15

Particle 10um
inlet 1.1987E-09

By theory 1-5um particles should be exit from domain, but it can be seen nothing is exiting.
As I am using 15m/s inlet velocity, 10S simulation time and 5e-4S time steps, I think there is no matter with simulation status. Also the air flow is developed, but an extremely circulating flow.

I have attached 1um particle track and air flow images from CFX Pre.

Can anybody guess the matter with particles transport inside the domain? I have used 2700kg/m^3 density for particles, is this the problem?

PT1um: Particle track file for 1um particles
V_F_1SL: Velocity of flow (1 stream line)
V_F_25SL: Velocity of flow (25 stream line)

Thank you.

ghorrocks March 4, 2013 05:44

How are you modelling this? Steady state? What turbulence model?

sakurabogoda March 4, 2013 06:02

It is turbulent, LES.

ghorrocks March 4, 2013 17:31

Do you know the expected residence time for the 1um particles? Your particle tracks might not be long enough, or the simulation run long enough.

I would check the general validity of your simulation by using much bigger and smaller particles (say 0.1um and 100um) and checking that they behave as expected. This will tell you whether you have a fundamental problem in the simulation, or if it is just an accuracy issue.

sakurabogoda March 4, 2013 22:59

3 Attachment(s)
Dear Glenn,

Thanks a lot.

But actually I dont know about particle residence time and I want to measure it from CFX. Also I am using 15m/s inlet velocity and so I don't think particles residence time is big.

I have checked flow velocity profile, and there also I can see more re-circulations other than exiting.

Shall I post my output file to have a look?

I am really grateful for your kind advises. It is very kind of you.

ghorrocks March 5, 2013 20:01

Your images are not readable.

Obviously the cyclone will not work unless you set the right flows/pressures at all inlets and outlets. Are you sure your outlet pressure/flows are correct?

All times are GMT -4. The time now is 13:29.