Problem Nusselt Number

[WallaWalla]

Hello,

I have a question about the calculation of the Nusselt number. In particular, I have to determine this value in a problem inherent a countercurrent heat exchanger and I need to find the value of the heat exchange convection coefficient h. The problem is that I have a concentric annular duct and I don't know what formula I must to use.
The convection is forced and then the Nusselt number should depend essentially on the Reynolds number and the Prandtl number, right? However, now I have many different empirical formulas for the calculation of Nu and I don't know which one I must to use. Also in the book "Handbook of Heat Transfer" of Rohsenow and Hartnett they use a table that provides the values ​​of this number as a function of the problem boundary conditions; however, in this case, the Nusselt number, is independent of Re and Pr. What relationship should I use for the calculation?

Thank you, best regards

Emanuel Bernardi

 

[25362]

Visit:

http://en.wikipedia.org/wiki/Nusselt_number

http://web2.clarkson.edu/projects/subramanian/ch302/notes/Convective Heat Transfer 1.pdf

 

[WallaWalla]

Thanks, but I had already read these things as I wrote above. In fact, in these document there are some relationship about Forced convection in turbulent pipe flow (Sieder and Tate, Dittus-Boelter ,etc.. ). My question is: I can use these relationships for the concentric annular duct?

 

[MortenA]

WallaWalla,

Im risking sounding "condescending" but im not (really!): This is where book keeping stops and engineering starts! Generally speaking for T&S HX (liquid/liquid) i would use dittus-bolter with the classical parametres.

Best regards

Morten

 

[25362]

I don't see why not.

See also:

http://en.wikipedia.org/wiki/Concentric_tube_heat_exchanger

As ballpark comparative figures for Re>10,000 you can use the following nuumbers which I took, with my old eyes, from a log-log graph in Fluid Flow and Heat Trnasfer by professor Lydersen (Wiley); usually (μ/μ[sub]w[/sub])[sup]−0.14[/sup] ~ 1.0. Basis: Sieder and Tate equation.

Re --------------------- Nu[×]Pr[sup]-1/3[/sup][×](μ/μ[sub]w[/sub])[sup]−0.14[/sup]​

10,000---------------------- 40​

20,000---------------------- 70​

30,000---------------------- 95​

40,000---------------------- 130​

100,000---------------------- 250​

200,000---------------------- 450​

Don't forget all these formulas are empirical; and errors of +/-15%, or more, are common.

 

[25362]

See also:

http://c.ymcdn.com/sites/www.saimec...yer-2003_07__600_dpi_-_2003__19_2___17-21.pdf

to get an idea of the possible estimated errors.

 

[ione]

For forced convection you can use the Petukhov and Roizen correlation (please check it at this link

http://faculty.ksu.edu.sa/Assiry/Documents/Correlations for Convective Heat Transfer.pdf

), where Nu tube is calculated with the Sieder-Tate correlation. What pointed out by 25362 about possible errors stands still.

 

[WallaWalla]

Thank to all. In the nest days I will compare the theoretical data with experimental data so as to estimate errors.