CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   CFX (http://www.cfd-online.com/Forums/cfx/)
-   -   Simulating soot sticking to a wall in a boiler flue. (http://www.cfd-online.com/Forums/cfx/105645-simulating-soot-sticking-wall-boiler-flue.html)

Zmur August 5, 2012 08:16

Simulating soot sticking to a wall in a boiler flue.
 
Hello!
I need to simulate how the boiler flue is being contaminated by soot particles.

I have a model of a boiler, where the hot gases (air and soot) travel through a pipe, which is inside a tank of water (this water then is used for heating). The length of the tube is 1m, diameter is 80mm. It has some obstackles inside it (made of sheet metal, size about 50mm), which are used to increase turbulence. The soot is sticking around them. We try to optimize the flow so that less cleaning is required.
Right now I can't make paticles to stick to wall. They only slide along it.

Some settings:
Particles:
Erosion Model Finnie (power factor = 2.0; Reference Velocity 1 ms)
No Particle rough wall model
One way coupling
Drag Force Schiller Naumann
Particle control settings:
max. part. integr. time step is 1e10[s]
Max tracking time is 20[s], Distance is 10[m], steps 100000

Flow:
Flow velocity is 5m/s (a pump is mounted at the air intake of the device), temperature is 300C.

Heat transfer is enabled. I use Heat transfer coefficient setting on the outer side of the tube. The water outside of the tube is about 70C.

Any information source on the problem would be helpful. Also, I think that I don't need tpo simulate particles precisely. It would be sufficient for me to see where they are gathering and\or sticking to a wall. If there is a way to do this without particle simulation, this may also help.

ghorrocks August 5, 2012 09:32

Why are you using an erosion model? I did not think soot could cause erosion.

I think you are looking for the coefficient of restitution parameters, which you set on the wall boundary. You set this to zero and the particles stick.

Zmur August 5, 2012 16:34

Quote:

Why are you using an erosion model? I did not think soot could cause erosion.
My mistake. Turned that off.

Quote:

I think you are looking for the coefficient of restitution parameters, which you set on the wall boundary. You set this to zero and the particles stick.
Looks like they now just disappear after hitting a wall. I made parallel restitution coef. 0.9 and perpendicular 0.

ghorrocks August 5, 2012 20:21

Yes, the particle track stops (ie disappears) after hitting the wall.

Zmur August 6, 2012 02:40

Can I see how many particles have disappeared in some place?

ghorrocks August 6, 2012 06:53

Not sure off the top of my head. It might be in the wall volume fraction but I am not sure.

Also, are you sure your particle tracking technique captures the correct physics of soot in close proximity to a wall? Soot is such a fine particle that things like electrostatic charge and possibly other forces will become important which are not usually considered in particle tracking.

Zmur August 8, 2012 16:28

About that electrostatic charge, thank you, this may be an interesting idea. :)

Right now I want to find a way to see where particles are concentrating at low speeds (they usually stick to wall in these places). Maybe I can model them as a fluid or some sort?

ghorrocks August 8, 2012 19:43

My point is a bit deeper than than - soot is such a fine particle it will have essentially no slip relative to the air. This means that it will just follow the air streamlines and not deposit anywhere. But clearly it does, and so something else is happening to the particles to make them stick onto the walls. It could be electrostatics, it could be something else. But something else is playing an important role in deposition and until you include it in your model your model will not be valid.

Following on from this point, if soot has no significant slip relative to air then a particle tracking approach is not a good choice. Modelling the soot as an additional variable with convection and diffusion is more appropriate - it is also much easier.

Zmur August 9, 2012 07:38

Quote:

Modelling the soot as an additional variable with convection and diffusion is more appropriate - it is also much easier.
I can try. How can I set this up?

ghorrocks August 9, 2012 08:25

Have a look at the flow from a circular vent example.


All times are GMT -4. The time now is 03:52.