topic Internal Metal Powder Flow in CFD Forum

Internal Metal Powder Flow

Anonymous — Wed, 17 Jan 2018 23:10:02 GMT

I was wondering if there is any way to model powder flow in Autodesk CFD 2018? The model I am trying to create involves a metal powder internally flowing through a valve.

A follow up question to the question above is: if the above is possible, would it also be possible to model both the metal powder and a carrier gas within the same volume?

Thank you in advance!

Re: Internal Metal Powder Flow

Jon.Wilde — Thu, 18 Jan 2018 09:34:33 GMT

Are you able to characterise it using the fluid material settings within CFD?

We can model non-newtonian fluids as well, which might be a help although I am not totally sure.

If you have a carrier gas, what percentage of the total is power?

We could consider modelling just a gas and then adding particles as a post process, but then this is really only modelling gas.

What we cannot really do is model two fluids, you would need to assume a homogeneous fluid.

Does that help?

Thanks,

Jon

Re: Internal Metal Powder Flow

Anonymous — Fri, 19 Jan 2018 18:56:01 GMT

Jon,

I do not believe I can model this as a non-Newtonian fluid, but I will look into it and see if anyone else has done this.

The model basically comprises of a carrier gas (argon) carrying metal powder (45 um to 135um particle size) through a valve, so the solution of modeling the gas and adding in the particles as a post process sounds interesting. But how would adding the particles in post only be modeling the gas? Is it because the particles would not be a part of the solving process?

Thank you!

Re: Internal Metal Powder Flow

David.Short. — Mon, 22 Jan 2018 09:11:23 GMT

Hi,

Massed particle traces sounds like a good option. Yes precisely, the particles behavior is not solved for but estimated using the final flow solution and the set particle characteristics.

Here is some more info.

Hope that helps,

David

Re: Internal Metal Powder Flow

Anonymous — Tue, 23 Jan 2018 18:08:08 GMT

Thank you David! I think this will work fine for my purposes. I've been playing around with the parameters it seems good so far!