CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   CFX (http://www.cfd-online.com/Forums/cfx/)
-   -   Wall Transfer Coefficien - Heat Transfer Coefficient - CFX (http://www.cfd-online.com/Forums/cfx/94006-wall-transfer-coefficien-heat-transfer-coefficient-cfx.html)

Gargioni November 2, 2011 18:49

Wall Transfer Coefficien - Heat Transfer Coefficient - CFX
 
Dear friends,

I have simulated the flow in a simple geometry (like 2 paralel walls - "2,5D" case).
My BC are (for laminar flow):

In - low velocity and fixed temperature;
Wall - no slip and fixed temperature;
Out - Openning with static pressure.

In CFX-Post:
I have created a line in wall position, through the length.
When I plot the variable (Wall Transfer Coefficien), the value found isn't the value hoped. :confused: 1) Why this happen?

But when I take the Heat Flow "q" (in same line created) and temperature average of section "Tav" (with formulas in own CFX-Post)... I can obtain "h" correctly: q /(Tw-Tav) = h

2) Why "q" is correct, but "h" isn't?

3) How can I obtain "h" directly?

4) I have seen thread about that. Some users has talked about reference temperature.. But I don't understand that...
Someone can explain it for me? and how can I change this reference temperature?


I'm starting to study the CFD world...

Thanks in advance everyone.

ghorrocks November 2, 2011 21:58

I would start by working through this FAQ:

http://www.cfd-online.com/Wiki/Ansys..._inaccurate.3F

Gargioni November 2, 2011 23:25

First of all, thanks for answer.
I liked the link! There are a lot of good things over there!

but... don't worry... I have done all steps, mainly for to learn how the software works.

My results aren't inaccurate. I'm using: RMS = 10^-09 and highresolution for advection and couple velocity-pressure. The simulation doesn't need great adjustments.. So highresolution scheme is good.

I have problem just with Heat Transfer Coefficient, even the "q" seem is right.

The manual doesn't let clear about "h". There said some thing about: "... using an external heat transfer coefficient, hc...".

As am I getting the value of "q" correct, while the value of "h" seem being wrong? How can I obtain the truly value of "h"?

*I did the same simulation on FLUENT software and I obtained the "h" correct there! :(

ghorrocks November 2, 2011 23:49

There is more to accuracy than convergence tolerance and differencing scheme.

h is referenced to a ambient temperature. By default CFX uses a function of the local fluid temperature which is usually quite different to the engineering definition (which is usually inlet temperature or far field temperature).

To get h as engineers understand it have a look in the output file. There is a discussion in there about HTC reference temperatures and how to define your own temperature.

Gargioni November 3, 2011 00:25

"There is more to accuracy than convergence tolerance and differencing scheme." I said just for let clear that I don't have problems with accuracy.

May you help me to find this thread about HTC?

Thanks in advance =]

ghorrocks November 3, 2011 00:28

It is in the output file. Have a look in your simulation. It is not a thread, but there are plenty of threads in the forum which have discussed the issue.

Gargioni November 4, 2011 16:34

I changed the expert parameter tbulk for htc. And I think CFX considers the tbulk value like constant :confused:

How I said:
I have simulated the flow in a simple geometry, like 2 paralel walls - "2,5D" case. I have created a line in wall position, through the length, for captures the "h", however the value of "tbulk" is variable along the length.

What can I do when the tbulk value isn't constant?

ghorrocks November 6, 2011 05:57

If you have set the tbulk parameter then tbulk will be constant and the value you define, otherwise it will depend on local flow conditions.

You can easily extract h for a constant tbulk anyway. The local tbulk temperature used is an available variable, so between that and the wall heat transfer variable you can calculate h based on any tbulk you like.

Gargioni March 7, 2012 01:19

I solved my problem... doing a CEL routine.

TrII4d April 15, 2013 07:46

maybe u can give a short summary of the steps which u made to help me?


All times are GMT -4. The time now is 06:56.