|
[Sponsors] |
February 19, 2009, 06:35 |
Heat transfer coeffcient
|
#1 |
Guest
Posts: n/a
|
Hi all,
I want to find out the average heat transfer coefficent associated with a wall. The problem is Water flows inside a pipe with the surface temperature of pipe is 60 deg C and the water enters at 27 deg C. Flow is turbulent. Whether surface heat flux and mean temperature are to be taken for finding h? If yes, how to find the mean temperature of fluid? Else can we take the surface heat transfer coefficient given directly by the software? which is correct? Expecting a reply Raj. |
|
February 19, 2009, 17:14 |
Re: Heat transfer coeffcient
|
#2 |
Guest
Posts: n/a
|
FLUENT does exactly what you what. see fluent user's guide
|
|
February 20, 2009, 03:40 |
Re: Heat transfer coeffcient
|
#3 |
Guest
Posts: n/a
|
Thanks for your reply Mr. Pablo
From Fluent Users Guide, Surface heat transfer coeffcient is defined by h = q /(Twall-Tref) where q is the total surface heat flux Twall is the temperature of wall Tref is the reference temeperature given in Reference values table which is a constant. The result of the above equation is that 'h' will keep on decreasing over the length of the pipe. But, when the flow is thermally developed, 'h' should become constant (pipe flow) which does not happen here. Hence suggest me how to get the constant 'h' for the above stated problem. Regards Raj. |
|
February 20, 2009, 10:42 |
Re: Heat transfer coeffcient
|
#4 |
Guest
Posts: n/a
|
maybe you can take an area average of h over the pipe area? alternatively calculate h yourself by exchanging Twall and Tref to average wall temperature and fluid temperature respectively
/M |
|
February 23, 2009, 00:05 |
Re: Heat transfer coeffcient
|
#5 |
Guest
Posts: n/a
|
Thanks for your reply Mr. Mange
My aim is to simulate the Dittus-Boelter's equation for a turbulent pipe flow with heat transfer. Through this equation,we will get a heat transfer coefficient which I want to get from the fluent analysis. I cannot take a average 'h' for the whole pipe since 'h' will be very high at the starting of pipe. As you said, I already tried with taking the average fluid temperature all over the pipe length and then I calculated 'h' by myself. Still the 'h' values does not match with Dittus-Boelters equation. This may be simple problem of solving a benchmark problem of Dittus-Boelters equation, but I could not get the results correctly. Kindly suggest me how to go about it? Regards Raj. |
|
February 23, 2009, 04:21 |
Re: Heat transfer coeffcient
|
#6 |
Guest
Posts: n/a
|
If you feel confident that you have an sufficiently accurate solution from fluent, maybe you need to double check how the parameter you are comparing with is defined in your source material.
/M |
|
February 23, 2009, 04:47 |
Re: Heat transfer coeffcient
|
#7 |
Guest
Posts: n/a
|
Hi!
some points: 1. the heat transfer coefficient is normally calculated by the reference inlet temperature. But I agree, when fluid is heating up while passing the geometry, reference temperature has to be calculated different! 2. to calculate the reference temp. for your problem, you might do one of the following: - calculate the temperature on the axis of your pipe, or - create a number of axis perpendicular planes through your pipe and calculate the area averaged temperature on each plane. 3. The accuracy of heat transfer in fluent is not the best!!! Especially for turbulent flow, you have so many parameters that can affect our results: The turbulence model, the wall function, the grid close to the wall (y-plus), the turbulence intensity at the inlet, the turbulence scale length,.... So, varying fluid temperature is the less problem you have! Best wishes Ralf |
|
April 9, 2009, 21:30 |
|
#8 |
Member
amar
Join Date: Mar 2009
Posts: 31
Rep Power: 17 |
Hello dears Raj and Ralf
Sorry for inserting my question in your discussion, but it's very urgent for my simulation. My question is as follows: My geometry is on a section from flat plate fin and tube heat exchanger to simplify my geometry for you to be able to imagine., you can imagine it like the air flows over a non-zero thickness flat plate inside the computational domain and there is a tube perpendicular on the plate and through it. I'm using the energy and turbulence models. my problem is how to calculate the heat transfer coefficient to compare with experimental result. Also when I use the report ----> surface integral ----> Area-weighted average ----> wall fluxes ----> surface heat transfer coeff. I choose all the faces of the computational domain, I found that there is heat transfer coefficient for symmetry faces and also for the sold thickness of the flat plate. Is that correct or may be I have something wrong? Please this very urgent if you can help me. Thanks a lot |
|
|
|
Similar Threads | ||||
Thread | Thread Starter | Forum | Replies | Last Post |
Forced convection heat transfer in heat sink | Sidy | FLUENT | 1 | October 18, 2008 04:27 |
flow/heat transfer through heat sink (with fins)? | Pei-Ying Hsieh | Main CFD Forum | 1 | January 26, 2008 07:33 |
HEAT-TRANSFER & EXCHANGE HEAT MODEL Problems | Jordi | FLUENT | 1 | March 31, 2006 11:30 |
pressure coeffcient | shami | FLUENT | 0 | October 8, 2004 07:59 |
Heat Transfer Coeff. at Heat Flux Boundary | Rushyen | CFX | 6 | January 18, 2001 06:09 |