Tube Heat Transfer Coefficient (Forced Convection) 1

[murrylee]

Hi,

I have a steel tube 73.5" ID X 412" long. For cooling purposes I would like to run a large fan (60" dia direct drive, ~60,000 CFM) inside the tube . I would like to calculate the heat transfer coefficient for the inside of the tube for this scenerio. This apparatus will be used in outdoor conditions ranging from 108F to -52F.

I have been looking through my heat transfer books and I don't think that the equations for Turbulent Forced Convection apply for ducts that have a Length/ID ratio of only 5.6 (412"/73.5"). Should I just assume a flat plate?

Any help would be greatly appreciated!

Thank you very much,

Murry

 

[IRstuff]

What exactly are you trying to cool? The outside? or the inside?

TTFN

FAQ731-376

 

[vpl]

What's running through your tube?

What are you trying to cool?

This isn't a homework problem, is it?

Patricia Lougheed

Please see FAQ731-376: Eng-Tips.com Forum Policies for tips on how to make the best use of the Eng-Tips Forums.

 

[murrylee]

Sorry for not being more clear. I have many layers of heat generating cable wrapped around a steel winch drum (the tube). As the cable conducts heat into the winch drum, I am attempting to cool the cable by means of forced convection on the inside of the winch drum.

Inside the winch drum is ambient air.

Basically I would like to find the correct equation for calculating the heat transfer coefficient, h. I can't find an equation that applies to a tube interior cooled with turbulent flow that has a Length/Inner Diameter ratio less than 10.

The best I can find is the Nusselt equation (Ozisik, Heat Transfer pg. 317)

Nu=0.036*Re^0.8*Pr^(1/3)*(D/L)^0.055 for 10 < L/D < 400

All other equations for turbulent flow inside a tube seem to be for L/D > 60

Thanks for the help,

Murry

 

[IRstuff]

You'll probably need fins...

TTFN

FAQ731-376

 

[LOUDOG]

For something like this I wouldn't worry about doing a really scientific calculation, because chances are you wont get the right answer anyways. What I would do is check if the heat transfer is adequate for the maximum allowable temperature you need for a convective coefficient of 100 W/m^2K. This is a fairly typical value for forced convection gases. If this gives you a large safety factor then call it good.

If you have the apparatus built already, it wouldn't hurt to just try it.

 

[vpl]

I have many layers of heat generating cable wrapped around a steel winch drum (the tube). As the cable conducts heat into the winch drum, I am attempting to cool the cable by means of forced convection on the inside of the winch drum.

Let me start off by saying when I read the words "layers of heat generating cable" my mind immediately thinks of energized multi-phase electrical power cable. It then has a tendency to go off and cower at the idea of many layers of this cable being wrapped around a steel winch drum. I realize that this might not be the type of cable murrylee is talking about (and just shows what can happen when people from different backgrounds think they're clearly communicating). Hopefully he will come back and reassure me this isn't what he intends.

The problem is that, if the cables are generating heat, the hottest area may not be in the center of the winch drum. It might very well be part way through the cable layers. Additionally, there might be localized hot spots that differ in distance from the average.

So, no matter what type of cables, if they truly are heat-generating, then the first thing that needs to be done is to determine how much heat the cable are generating. Then murrylee would need to figure out where his hottest area is. If it's not very close to the center of the winch drum, adding forced air cooling probably isn't going to help. This is because air, even forced air, is not very good at cooling (there is a reason IRStuff mentioned using a finned surface) and it might not be getting to the hottest area. So, no matter the flow rate, the cooling may very well be insufficient to prevent localized hot spots -- which could, in turn, create electrical shorts and fires.

This could be a rather complicated heat-transfer problem to solve as it would depend on the amount of heat being generated per layer of cable, whether the cable was insulated, how tightly packed the cables were, whether the layers were uniform, and probably a dozen other factors I haven't thought of. Obviously it can be done, after all, generators exist and work fine. But I don't think it is just a simple equation to calculate the heat transfer coefficent.

My two cents worth...

Patricia Lougheed

Please see FAQ731-376: Eng-Tips.com Forum Policies for tips on how to make the best use of the Eng-Tips Forums.

 

[IRstuff]

Unless he's lucky??? enough to have bare copper cables, there is the equally critical issue of what thermal conductivity he has through the wraps to even get to the winch barrel,

If it's electrical cable, so much the worse, since the insulation, well, insulates, both electrically and thermally. Big electrical cables can be made from steel wrapped around an aluminum core, which is not exactly conducive to thermal conduction.

TTFN

FAQ731-376

 

[chicopee]

Murrylee, when you say that you have "many layers of cable", how many layers do you have on the tube? or are you saying that you have many wraps of cable on the tube but only one layer?