Single face with excessively high Heat Trans Coeff
 

All,

Here I am again with the same problem I faced months ago-

http://www.cfd-online.com/Forum/flue...cgi/read/34467

I am analysing 3D cylinders in cross flow heated from one end - they are a periodic section simulating a pin fin heat sink (air is the fluid).

The problem is that I'm getting single faces with very high (in the 1000s for Re around 400) heat transfer coefficents - there tends to be only one face in the whole domain to exhibit this behaviour.

I've asked Fluent about this and the suggestions they put forward haven't worked - take a look at my post-

http://university.fluent.com/forum/viewtopic.php?t=116

After trying different turbulence models/meshes/double precision I attempted to adapt the grid at regions of high HTC but with no success - the face with high HTC always comes back.

Anyone any ideas on this problem? It's really messing up my PhD!

Re: Single face with excessively high Heat Trans C
 

1- Heat transfer coefficient is a function of the Reynolds number and the Prandtle number. Do you believe that your velocity field is well computed ? How do you calculate your HTC? 2- Can you email me a case and data file?

Re: Single face with excessively high Heat Trans C
 

I am aware of how Fluent calculates the Heat transfer Coefficient - but even if the total surface heat flux is the parameter being monitored, the cell still appears to be giving incorrect results.

The case and dat file are very large (130MB) as they are quite detailed and being solved with double precision, so I'm not sure if I can email them.

Any ideas how I can solve this problem? What is causing this high heat transfer coefficient - what parameter in the wall adjacent fluid cell is causing this?

any help would be much appreciated- thanks, mat w

Re: Single face with excessively high Heat Trans C
 

Mat Are you happy with the Velocity field? Your geometry seems to be producing a complex velocity field. As you are preparing for a Phd, I would ask you to follow the rules of science, if there is a kind of anomolous result, we look at the definition and look for what might have caused it, my starting point is the velocity field, if you agree! I guess you can use the <send mail> button so I can send you an email account that can receive these big files, Cheers and Good Luck