CFD Online Discussion Forums

CFD Online Discussion Forums (
-   CFX (
-   -   Concentric tubes heat exchanger (

FTM April 9, 2006 07:07

Concentric tubes heat exchanger
I have a two concentric tubes heat exchanger: air in the external tube and water in the internal tube. Air and water are separated by a thin surface of 1mm (the tickness of the internal tube).

Should I mesh the thin surface or there is a way to simulate heat transfer between air and water without meshing the surface?

Can anyone show me a model for this kind of problem?


kirby April 9, 2006 09:32

Re: Concentric tubes heat exchanger
If you simulate the tube,I think you should combine the Fluid calculation and the Solid calculation.

FTM April 9, 2006 16:39

Re: Concentric tubes heat exchanger
Thank you very much for your answer, but the problem is that is very difficult to obtain a resonable mesh for a tickness of 1mm.

Moreover the material of the internal tube is copper and due to its small thickness the tube has a negligible termal resistance. I thought that treating the problem without considering the solid tubes but only a total transfer of heat (no mass) between air and water was a good solution. I sincerely have no idea of how to do that and I am wondering if it is possible.

Otherwise I'll find a way to mesh the internal copper tube and consider the problem as a double conjugate heat transfer (air-solid and solid-water). Anybody thinks that this is the only way?

As usual, many thanks to all.


Glenn Horrocks April 9, 2006 18:31

Re: Concentric tubes heat exchanger

I am not sure, but I don't think using a thin surface will transfer the heat from one side to the other (not easily anyway - please correct me if I am wrong). Therefore the only way to do this is the model the tube thickness and run a CHT simulation.

You will probably find it easier to mesh your tube as an extruded prism or hex mesh and join it to the fluid meshes with GGI interfaces. This will give you a much nicer mesh than using tets, especially if the wall thickness is small compared to the length.

Also, in a CHT simulation do not forget define your convergence critereon on balances AND residuals. If you leave it as the default convegence on residuals only you can have large mass/momentum/heat imbalances across the CHT interfaces.

Regards, Glenn

ste April 10, 2006 09:28

Re: Concentric tubes heat exchanger
Just set a positive Heat flux on the wall of the cold domain and the same heat flux (negative) on the wall of the hot domain...I think it works.


PS: do you have periodic boundaries?

FTM April 10, 2006 13:04

Re: Concentric tubes heat exchanger
Hi, many thanks to all for your interest.

I don't have periodic boundaris and I have just thought to set opposit heat flux on the wall. But to do that I should specify two conditions:

- Heat transfer coefficient: i think that is defined as (k/s), where "K" is the termal resistance of the wall material and "s" its tickness.

- A reference temperature: this is the real problem for me.

I am Italian, excuse me for not having an appropriate technical language.

Good job.

Glenn Horrocks April 10, 2006 18:14

Re: Concentric tubes heat exchanger

That's fine if you know the heat flux, but in general you don't. The heat flux is a function of the fluid and solid temperature field, and the heat transfer is a function of the wall boundary conditions.

Glenn Horrocks

All times are GMT -4. The time now is 16:54.