2D counter flow heat exchanger analysis using chtMultiRegionSimpleFoam
 

Dear Foamers,

I am doing Maters project in CFD using openFOAM. As a part of a project, I am working on chtMultiRegionSimpleFoam solver using the test case "2D counter flow heat exchanger" from the literature "A note on the solution of conjugate heat transfer problems using SIMPLE-like algorithms. Xi Chen *, Peng Han"

I am getting the fluid part temperatures almost correct but a difference of 50K is observed in solid part. The difference is constant thought the solid part. I have checked BCs and input values and they are ok. I am clueless where the problem is.

I have attached the case and the comparison graph. can anyone please go through the case and suggest?

Thanks in advance,
Prasad

Greetings Prasad and welcome to the forum!

It looks like:

1. You're using OpenFOAM 2.1.x.
2. The Allrun script seems to have been edited on Windows.
3. You are using blockMeshDG for generating the mesh: http://openfoamwiki.net/index.php/Contrib_blockMeshDG

This kind of information is crucial for anyone to help you! ;)

About your case:

• It looks similar to this one: http://openfoamwiki.net/index.php/Ge..._-_planeWall2D - this is an incomplete example. But you might want to have a look into the chapter "4 References" on that page.
• I have ran your case with openFOAM 2.1.s, but I don't know where you are measuring the Temperature.
• Therefore, all I can do is guess:
  1. The settings in "system/*/fvSchemes" for the air regions might not be well done, because most of the divergence schemes are set to 1st order, namely "upwind". Switching to second order might help.
  2. You might not have enough mesh resolution near the walls of the solid.

Best regards,
Bruno

Hi Bruno,

Thanks for your time. I am using blueCFD2.1.1. I have taken the values across the HE at L = 0.5.

I have tried couple of variations in numerical schemes and got mixed results. First I tried with SFCD, linearUpwind and both had no improvement in the results. With QUICK scheme, the results were improved and the deviation this time was around 10K. :-) The mesh refinement in the fluid and solid regions did not help much.

Revised comparison graph is attached.

I have a quick question. Interpolation scheme 'linearUpwind' is described as 'first/second order, bounded' and how do we ensure to enforce first or second order specifically? Or on what basis program will consider the specific order?

Thanks,
Prasad.

Dear Prasad,

I don't have access to the paper, so I don't know the specifics of how the paper generated the data, in order to estimate what might be the best approach. It could also be a problem with the turbulence model or fluid properties.

As for "linearUpwind", a quick search lead me to this thread, although I did not fully read it: http://www.cfd-online.com/Forums/ope...tedlinear.html

Best regards,
Bruno

Dear Bruno,

Sorry for the late reply. I got tied up with some other work.

The case is based on laminar flow and fluid properties are given as stated in the literature. I could not attach the literature document because of large file size. Instead please download it from the below dropbox link.

http://db.tt/6qrj6vRH

I presume chtMultiRegionSimpleFoam can be used for incompressible fluids such as water. I searched in forum and found no concluded answer or did miss the right post?

I am also trying to work out on incompressible version of cht...foam chtIcoMultiRegionSimpleFoam which is available in forum. I am finding it difficult to set up a test case. Is there any test case already run for this solver? I could not locate one on forum.

Thanks,
Prasad

Hi Prasad,

Sorry, I don't have much patience to look at this with much more attention (it's been a very long week), but the following properties look suspicious to me, at least on the solid side:
The comments about the units are based on this: http://www.openfoam.org/docs/user/ba...18-1000004.2.6

The paper stipulates that the solid part has the following properties:
I'm too tired to double-check the units, but my suspicion is that the unit conversion is not direct for all 3 properties.


As for the other solver you mentioned: I'm not familiar with it.

Best regards,
Bruno