How to Calculate Nusselt Number for Air Flow in a Pipe with ANSYS?

[amirghaderi]

Hi
i need to calculate nusselt number for stimulation of air flow in a pipe with 5mm diameter and the pressure is 5atm.
it should be in the form:
Nu = N1 + N2 Re^N3*Pr^N4
and i want to calculate N1,N2,N3,N4.
but i don't know how?
can anyone Help?

 

[minger]

For simple air flow through a circular conduit and given proper boundary conditions, you should be able to find analytic...well empirical relations.

 

[amirghaderi]

hi minger
but all i have is nusselt, prandtl and reynolds numbers...
so i have 3 known factors but 4 unknown: N1,N2,N3 and N4.
it has not a unique solo...

 

[minger]

If you have all of thsoe parameters, why do you care what the constants are?

Also, here is a solution
N1 = Nu
N2 = 0.0
N3 = 0.0

The point I'm trying to make is that there will be an infinite number of solutions given one data point.

 

[CFDFEAGURU]

ANSYS has no input for a Nusselt number. You need to calculate the film coefficient from the Nusselt Number and then use the calculated film coefficient for your boundary condition on the pipe wall.

 

[amirghaderi]

ok
but how i can calculate film coefficient from nusselt?
i am using fluid116...
and has N1...N4 input.
can i change its options to change required inputs?
tanx...

 

[CFDFEAGURU]

Is this 2D flow or 3D? (I haven't used the old FLOTRAN elements in years) Pipe flow is axisymmetric so you should do a simple 2D analysis. (3D will not show you anymore information than a 2D analysis would) You calculate the film coefficient from the relationship Nu=hD/k. Where k is the fluid thermal conductivity, h is the film coefficient, and D is the inside diameter of the pipe.

 

[amirghaderi]

its 3D analize because it has a wavy shape (it's not straight) so i can't do it axis symmetric..
i create 3D pipe geometry in ansys workbench and i import it in ansys.
but i can't mesh it with element "FLUID116" ...
with witch element i can mesh instead??
tanx

 

[CFDFEAGURU]

Not sure what element you can use in workbench. (I hate workbench, it should be outlawed) I use CFX for all of my fluid simulations. I am having a hard time understanding how a cylindrical corrogated pipe is not axisymmetric. Do you have a picture of the geometry?

 

[amirghaderi]

the pipe is like this..
i don't work in work bench i just create its geometry in workbench.
but i do other analizes in ANSYS .

 

[CFDFEAGURU]

I see. What are you trying to calculate? Pressure drop? Velocity? ... Also, the model is symmetric at the midplane so you could save some time and use symmetry. If I know what you are after I can help you out better.

 

[amirghaderi]

the problem is :
the fluid with 5atm pressure and -50C flows through the pipe , the pipe is heated by heaters on its side and has constant temp 200C.
i want to find the required length to have fluid with 100C.

 

[minger]

Did you check your textbooks as I mentioned earlier? Look for something called the Colburn equation. It is written exactly as you need and is written for computing the local Nusselt number for a fully developed (hydrodynamically and thermally) turbulent flow in a smooth circular tube.

In fact, in my textbook the following example problem is very closely related to what you're trying to do. They give a length and find exit temperature, you need vice versa.

To be honest, do this with a pencil and pad of paper, you'll be much closer than doing a numerical simulation. In order to accurately model the heat transfer from the moving fluid, you'll need to accurately resolve the thermal boundary layer. You're much better off doing a hand-calc.

 

[amirghaderi]

hi minger
can you tell me what is your textbook?
tanx

 

[minger]

Just "Introduction to Heat Transfer" by Incropera and DeWitt. Dittus and Boelter slightly improved on the Colburn equation.

http://en.wikipedia.org/wiki/Dittus-Boelter_equation#Dittus-Boelter_equation