CFD Online Discussion Forums

CFD Online Discussion Forums (http://www.cfd-online.com/Forums/)
-   FLUENT (http://www.cfd-online.com/Forums/fluent/)
-   -   Fluent Tutorial on Heat transfer (http://www.cfd-online.com/Forums/fluent/122681-fluent-tutorial-heat-transfer.html)

Far August 25, 2013 16:18

Fluent Tutorial on Heat transfer
 
Here I have prepared tutorial on Heat transfer and calculation of local heat transfer coefficent and local Nusselt no in Fluent.

It is well known test case "Baughn Pipe Expansion at ReD = 40750

http://img94.imageshack.us/img94/1462/da8j.png

ICEM CFD, Mesh and Fluent Files are attached along with excel sheet and PPT slides.

https://dl.dropboxusercontent.com/u/...n_ReD40750.zip

1. Units = meters

2. Boundary condtions Velocity and temprature are given in excel sheet

3. Materials properties : K calcualted from prandlt number and Cp as 1005

4. De in the formula for Nu (see image) is calcualted as 2*D*3.14

5. Data is normalized with Nu from Dittus-Boelter correlation for a straight pipe. Value comes out to be 97.79

6. Two models are also compared V2F and SST

7. Yplus is below 0.06

8. Local heat transfer coeffcient is found from Fluent. Go to Wall Fluxs > Surface heat transfer coef. Please note that I have used local heat transfer coefficent in formula for local Nu nuber instead of q/(Tw-Tb). Heat transfer coefficients for both turbulence models are also saved in the zip folder attached herewith.

Far August 28, 2013 20:44

with different turbulence models
 
Results with different turbulence models are shown in following Fig.


http://imageshack.us/a/img20/5203/qn64.png

Far August 29, 2013 03:27

Above results were obtained for mesh size of 17000 nodes. Mesh is also refined and fine mesh has size 34000 nodes approx. There is no difference in results for V2F and SST model with baseline and fine mesh Results are attached...

Y+ for both meshes (baseline and fine) is below 1. It is approx 0.3-0.5

To match the Reynolds number, one must run the simulation and get the mass flow rate and put in the formula of Reynolds number. keep on iterating until you get the required Reynolds number

Moreover to provide the fully developed flow condition at inlet, one must run the simulation for pipe of dia d and import the velocity and turbulence profiles to the simulation. Or one must extend the length of the smallar cylinder by 35-40 dia upstream.

http://img6.imageshack.us/img6/2996/83gw.png

ousegui September 1, 2013 00:47

hi
Great efforts
thx

tiandabeiyang December 30, 2013 09:59

great!A question about heat transfer coefficient
 
Thank you for your effort.Here I have a question:the local surface heat transfer coefficient in dependent on the reference temperature.Different temperature get different heat transfer coefficient .So if we need a UDF to get the Tbulk at different position,and then from the fomular:h=q/tw-tbulk?
this is my thesis,so i do really appreciate if you can give me some advices.thank you very much.

Far December 30, 2013 13:13

yes we need local Tb and we have to use udf.

tiandabeiyang December 30, 2013 23:02

Udf
 
thank you for your reply.Rencently i spent a lot of time on the UDF,but it is too much for me to read,because i am not familiar with UDF.So i want to kow if it is convinient for you to send me a copy of your UDF?thanks very very very much.
this is my email:xusai1206@gmail.com
thanks again.

tiandabeiyang January 1, 2014 03:43

Hi,Far
 
I want to know how to get a specific region of Tbluk?I mean how can we get local region Tbluk which is normal to the local wall face.If we need a position judge to realise it?Thank you very much.

Far January 2, 2014 12:53

Just few ideas !!!
 
i have few ideas and i would to have comments on my opinion from experienced members in this field:

1. Draw a vertical line where you want to have the heat transfer. Take the vertex average fluid temperature on that line.

2. take a point just outside the boundary layer (thermal boundary layer)


I have UDF for this purpose sent by some one for the case i have mentioned above.

siefdi February 16, 2014 22:35

Hi, great effort, thanks.

Sorry for a noob question, but could you please explain in more detail about this step:

Quote:

Originally Posted by Far (Post 448666)

To match the Reynolds number, one must run the simulation and get the mass flow rate and put in the formula of Reynolds number. keep on iterating until you get the required Reynolds number


thanks and regards,

stalon67 February 21, 2014 15:39

I need you udf!!please
 
Quote:

Originally Posted by Far (Post 468390)
i have few ideas and i would to have comments on my opinion from experienced members in this field:

1. Draw a vertical line where you want to have the heat transfer. Take the vertex average fluid temperature on that line.

2. take a point just outside the boundary layer (thermal boundary layer)


I have UDF for this purpose sent by some one for the case i have mentioned above.

hi far.may you send me a copy of your udf? really i need it.thank you so much.(jamalim3@gmail.com)

Ismael April 28, 2014 11:47

Udf
 
Hi. I have seen the discussion and a have the same problem with the Tbulk temperature calculation. I will be very grateful with you help me to develop this UDF code or send me the UDF file from this calculation?(ismael.marchi1@gmail.com)
Thanks for your attention.

vga67 December 18, 2014 05:24

hello

excuse me if here is not the right place for asking such question.

i am simulating a two phase (water and air bubble) model which is the interaction of a shock in water with an air bubble. my problem is that when the shock hits the bubble the temperature of the bubble should rise but the fluent does not show this(i have selected the energy equation to be solved). i do not know what the problem is. should a UDF be written for heat transfer inside the bubble?
can anyone help me on this?
thank you in advance
Hamid


All times are GMT -4. The time now is 16:34.