CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   CFX (http://www.cfd-online.com/Forums/cfx/)
-   -   Anisotropic thermal conductivity in an insulated stranded coil (http://www.cfd-online.com/Forums/cfx/86844-anisotropic-thermal-conductivity-insulated-stranded-coil.html)

 Attesz April 4, 2011 05:17

Anisotropic thermal conductivity in an insulated stranded coil

Hi CFX users,

I'm modeling an electromotor inner fluid domain including the heat source solid domains. The main problem is that there are little air gaps between the copper conductors (little circles nearby) and they are insulated by varnish. So towards the conductors the heat transfer coefficient is about the value of the copper, but in the orthogonal directions it is reduced because of the air gaps and the insulation.

My first idea is to write to the support:p to help me how to model the anisotropic thermal conductivity via CCL commands as I read in other topics here.

But I'm interested in your other suggestions about the modelling of this problem.

Attila

 singer1812 April 4, 2011 09:46

Contact support.

This can be done by modifying the material properties in CEL. The limitation is that the orthotropic material properties must be aligned with either the Global Cartisian Coord System, or Global Cylindrical System, so your body has to be aligned right and be of a geometry allows this limitation.

 Attesz April 4, 2011 10:27

Thanks Edmund, fortunately my geometry is aligned with the GCS, I thought that this is necessary. I'm waiting for the support.

On the other hand, am I right with the approach of the modelling?

 singer1812 April 4, 2011 10:36

No idea.

I am not clearly picturing what you are trying to model. If you are just determining an HTC to use as a BC, then I wouldnt think you need to use ortho props at all.

On the other hand, if you are trying to model thermal gradients within the solid domain, then perhaps yes.

 Attesz April 4, 2011 11:08

Quote:
 On the other hand, if you are trying to model thermal gradients within the solid domain, then perhaps yes.
Yes, I'm trying to model this. Thanks!

 All times are GMT -4. The time now is 12:28.