CFD Online Discussion Forums

CFD Online Discussion Forums (
-   CFX (
-   -   Heat transfer Coefficient (

san.pirate October 21, 2009 19:26

Heat transfer Coefficient
Dear all,

I couldn't find Heat Transfer Coefficient in the Variables list in the CFX-post. How can such an important variable be missing.

There is only Wall heat transfer coefficient. What if a person needs the local or average heat transfer coefficient of the fluid.:eek:

Anticipating help


ghorrocks October 21, 2009 21:42

???? What other coefficients are there? The wall heat transfer coefficient is the convective HTC.

If you want to calculate the average or global HTC you can average/integrate the Wall HTC over surfaces. Be careful about the reference temperature when doing this, you will probably want to specify a constant reference temperature for the HTC when you do this.

san.pirate October 22, 2009 13:01


Originally Posted by ghorrocks (Post 233610)
The wall heat transfer coefficient is the convective HTC.

yes you are right. But I am not interested in the wall heat transfer coefficient, but the convective HTC of the fluid. The fluid being the air.

Normal method is h=Q/(A* LMTD), but how will employ this formula for a 3-D domain.

I was wondering whether CFX had a direct method to find the HTC of the fluid.

Thanks for the initial response

ghorrocks October 22, 2009 17:21


But I am not interested in the wall heat transfer coefficient, but the convective HTC of the fluid. The fluid being the air.
As I said, the wall heat transfer coefficient IS the convective HTC. They are the same thing.

CFX calculates the h locally on the wall as a function of the local flow conditions.

Miky1020 October 22, 2009 20:32

You can't define convective HTC for an isolated fluid. HTC has sence between a fluid and a surface.

ghorrocks October 22, 2009 20:48

I think I see what you are talking about. Do you want the convective component of heat transfer in the bulk of the fluid from the energy equation?

You should be able to write an expression in CFD-Post to evaluate it and plot it from there.

Why do you want to see it? What are you looking for?

san.pirate October 23, 2009 17:06

I want the Heat transfer Coefficient, so that I can find out the Nusselt number and then certain other factors from there. But the value of wall heat transfer coefficient is in the range of 200W/m^2K Where as the Heat transfer coefficient of air is normally in the range of 10-100W/m2K


Miky1020 October 23, 2009 17:33


HTC has sence between a fluid and a wall. Nusselt number is defined by dimensionless fluid temperature gradient evaluated AT the wall. HTC which you are looking for and wall heat transfer coefficent are the same thing, as Glenn has pointed out.

In the other hand, you will find reference values for air HTC, but this coefficient is higly dependent on flow regimen, flow geometry, temperature differences between fluid and wall and fluid properties, so you must calculate it for every single case, even when you have a correlation for Nusselt number. I mean, 200 W/m^2K its a possible value for HTC with air like fluid.



ghorrocks October 24, 2009 05:49

Two issues about wall HTC:
1) It is defined using a reference temperature. By default CFX uses a reference temperature calculated from the local flow conditions but usually when you compare against literature values you need to use a defined reference temperature, usually the far field temperature. So you need to either:
1a) Tell CFX to use a specified reference temperature for the wall HTC value OR
1b) In post-processing use the wall heat flux to calculate a wall HTC based on your reference temperature.

2) Getting accurate wall heat transfer is tricky. Much trickier than getting the boundary layer and wall shear stress right. Don't be surprised if you have significant errors (50% error is not too bad)

3) as Miky says, 200w/m^2K is quite possible. As we have no idea what you are modelling we cannot comment on that.

san.pirate, your posts are very vague. Every posting you are adding a new piece of important information which really should have been posted on the first post. Can you please make sure you post enough information to describe what you are trying to do so we don't have to guess?

All times are GMT -4. The time now is 08:45.