Hi everybody,

I had posted the same message before but I could not recieve any response.!!

My problem is this, I would like to see the heat transfer phenomena in the domain( fluid or solid) in terms of vectors.. Any suggestions about this?

Thanx in advance

Leo

Could you be more specific about what you mean by "heat transfer phenomena".

Secondly, velocity can be represented by a vector because it has 2 (or 3) components depending on whether you are in 2D (or 3D).

What you can do is (if you model is 3D): (1) define 3 UDS's (2) create 3 custom field functions (CFF's) for each of your vector components (3) patch your CFF's to each of the 3 UDS's (4) create 3 more CFF's with each CFF equal to one of the UDS's (5) then, patch your 3 new CFF's to x, y and z- velocity (6) plot your vectors

Hi Lanre,

What I mean is that, I want to see/represent the the heat transfer process in terms of vector form. Suppose a 2D model with a high temp wall on one side and a lower temp wall on the other side. Let the region be solid for simplicity( i.e. solve only conduction)and other remaining two walls have zero heat flux. After solving , you will obtain a temp distribution in the region . Can we somewhat represent the heat transfer in the domain ( from one element to neigbouring elements/cells) in terms of vector form.That way we can obtain the path/region where the more/most heat transfer occurs..

I think we can somewhat use the temp gradients as the components of out new vector( dT/dx as x component, dT/dy as y component etc.)? Is that OK? If so how can I calculate the temp gradiens for each cell using UDS or CFF? I am only able to handle FORTRAN or Visual Basic languages!

Any suggestions?

Regards, Leo

(1). The standard approach is to plot the temperature contour. (2). You can easily see the heat flux direction, because it should be normal to the contour curve or surface.

Plotting of heat flux vectors is a useful aid, I know of one package that even animates based on these vectors. The conductive flux is no problem but choice of Tref for the convective flux will determine the magnitude of the convective flux vectors....

Fred.

Leo,

Following the message from Fred you can compute the heat flux gradient in each component direction to evaluate a vector. Utilizing UDS's or UDM's as storage area for your gradients. Currently, the vector plotting utility works only with velocity (will likely be any variable could be used to plot vectors in the future). So you have to use the patching approach I described previously.

You will probably need to use the DEFINE_UDS_FLUX macro and do a bit of C using pointers, etc. to compute the gradients.

Why don't you contact Fluent to work with you on this? It seems like a useful postprocessing feature that the code should have.

Good luck, Lanre

Correction to my last post: FLUENT 5.5 CAN plot vectors based on any solution variable.