[dsmith]

Hi guys,

I'm an undergraduate fairly new to the cfd world but am currently undergoing a cfd research project in my final year after enjoying and doing well in an introductory cfd course last semester.

the project involves using lagrangian particle tracking to model the dispersion of dust in a workshop from the sanding of wood during wood turning on a lathe. then find a solution to reduce the dust spread as it can cause health problems.

I know it's possible to set the wood on the lathe up as a rotating boundary etc. and I have done both the axial turbine and multiphasemixer tutorials.

but what would be the best way to model the production of dust in cfx from my rotating boundary?

also is it possible to enter a distribution of particulate sizes produced from the sanding as I had this information?

Thankyou!

[Josh]

Hi Dane -

The Blunt Body tutorial uses particles of varying size flowing over a body.

J

[ghorrocks]

Hi,

Your first decision is whether to go lagrangian particle tracking or if a eularian approach will work.

Firstly consider the simplest approach. If the sawdust is small enough that it has no significant relative slip compared to the velocity field (that is the dust is small enough and light enough to follow the velocity field) then forget particle tracking and use a passive scalar. In CFX talk this is an additional variable. If this assumption is good then this will be much faster and more accurate.

If the particles are big enough that there is some relative slip then a eularian approach may still be valid. You just need to define a drag equation which is appropriate.

You should only really go lagrangian particle tracking if you need a history for the particles. Sounds like you don't care about this so a eularian approach sounds more appropriate.

So at first glance I would recommend you look at eularian approaches rather than lagrangian approaches.

As for the sander creating the dust: If you only care about the flow in the room then don't bother modelling the details of the sander. Just put an inlet or a source term where the dust is created, squirting air and dust in the direction and velocity of the exhaust from the sander. No need for rotating boundaries or multiple frames of reference.

And yes, you can do particle size distributions in both eularian and lagrangian approaches.

Glenn Horrocks