Discussion Groups

Simulation CFD

Reply
Member
dpounds9999
Posts: 4
Registered: ‎03-31-2014
Accepted Solution

Heat Flux Profile

151 Views, 4 Replies
03-31-2014 11:36 AM

Hello All,

I am working on project for a concentrated solar power application.  I am looking for a method of specifying a heat flux distribution, or profile, across a single surface.  Meaning, the magnitude of the heat flux varies (non-linearly) with position at the heat input surface.  Is this possible to do with Autodesk Simulation CFD?

 

Thanks,
DP

Dan Pounds
Please use plain text.
Product Support
apolo.vanderberg
Posts: 299
Registered: ‎08-31-2011

Re: Heat Flux Profile

04-02-2014 08:10 AM in reply to: dpounds9999

No method of specifying a flux gradient

The only method would be to manually split the surface in to smaller surfaces and then to assign respective Fluxes to each subset to create the gradient

Please use plain text.
Member
dpounds9999
Posts: 4
Registered: ‎03-31-2014

Re: Heat Flux Profile

04-03-2014 06:33 AM in reply to: apolo.vanderberg

This is what I feared.  For a Gaussian distribution like function, the slope can be very steep near the peak and therefore can be quite painful to discretize.  I have a put in a suggestion to the idea forum to include a way to define a vertex and a function that would define the heat flux as a function of position, possibly with definite bounds.

Dan Pounds
Please use plain text.
Valued Contributor
nhahn
Posts: 92
Registered: ‎01-18-2006

Re: Heat Flux Profile

04-03-2014 06:45 AM in reply to: dpounds9999

An interesting application.  Off the cuff suggestion - If your problem is not transient, perhaps modeling a thin layer of highly conductive material between your parts and where you apply the heat flux will smooth out the distribution - this will allow you to use fewer regions and still achive a more continuous near-gaussian profile, instead of a stepwise one.

Please use plain text.
Member
dpounds9999
Posts: 4
Registered: ‎03-31-2014

Re: Heat Flux Profile

04-08-2014 05:22 AM in reply to: dpounds9999

While the approach of breaking up the surface into discrete regions is a valid approach, it can be a painful approach depending on the situation.  Please note that I have accepted this as a solution but I have also put in a request in the idea forum to incorporate an option to define a spatial function.  Thanks.

Dan Pounds
Please use plain text.