Natural convection for radial fins

[cymeryss]

I was wondering if there is a known way of predicting the natural convection coefficient for radial fins. Basically, I have 960 thin copper fins (t=0.13mm, fin pitch = 1.3mm), louvered, and are attached to a cylinder of diameter 210mm. From various sources, empirical data gives me the h value for vertical and horizontal surfaces, but I can't seem to locate any info that would talk about fins at angles, and can't seem to find any data on the problem as a whole. Obviously the angled fins, and the fins pointing downwards since the cylinder is horizontal will have different convective properties. Any help or good source information would be appreciated. Thank you.

Cymeryss

 

[ko99]

It's a simple model if you have CFD software but I'm not aware of anything else that could handle your shape.

ko (

http://www.ecooling.biz)

 

[Montemayor]

cymeryss:

You are right. Angled radial fins on a horizontal tube (why do you call it "cylinder"?) will result in a different natural covection current than for straight, vertical radial ones. I've never seen or heard of the application and I'm interested in knowing what the advantages of such a configuration are. I already know that the resistance to flow would be greater in the angled fins - so the induced current would be less---and so would the resulting heat transfer. What's the advantage?

I know of no source of heat transfer coefficients for such a configuration of radial fins. You don't state that heat transfer coefficients is what you are after - but I assume so. It would help if you were specific in the identification of your goal.

Art Montemayor
Spring, TX

 

[rmw]

Art,

210mm is bumping 9 inches. Why wouldn't you call that a cylinder??

Cymeryss,

This seems like a delicate fin for such a sized cylinder or tube, whatever Art and I settle on calling it.

For our thinking, the fin is roughly .005" thick, (doing the math in my head,) and the pitch is about .050". That to my way of thinking is a fine fin on such a large diameter tube.

What is the height of the fin??? And, how are the fins bonded to the cylinder. That information bears on the answer as well.

rmw

 

[rmw]

cymeryss,

I found the height in your other thread on the tape, which answers my other question. So I have no more questions.

rmw

 

[Montemayor]

rmw:

I would call it a cylinder as well - if it wasn't being used in a heat exchanger. My point is that it is such a large, unusual diameter that I suspect natural convection inside the tube (cylinder) as well - even though we haven't been told that. I've come to expect to be fed piece-meal basic data in queries such as these. With such a large diameter, the flow inside would have to be huge in order to attain turbulence and a reasonable heat transfer film coefficient. It's bad enough that we are looking at a notoriously bad apparatus for efficient heat transfer: natural convection with slanted fins.

There is something about this configuration that raises objections if one is trying to reduce the size of the heat transfer device - which is what I would expect; not make it worse. This is what motivates my question about the application. Perhaps this is a research tool or a lab set up of some kind. I await further data.

Art Montemayor
Spring, TX

 

[rmw]

Art,

I was devoting all my mental juices to the outside, and not even considering the inside.

No matter what the fluid was inside, think of the mass needing to be moved to get the velocity required to get a reynolds number suitable for decent heat transfer.

It is a strange animal, and I, too, would like to get more details. (taped on fins???)

rmw

 

[ko99]

You're right, the dimensions are very strange. In fact, the fin spacing is much too dense for natural convection. Louvers don't make sense either. Sounds like a very elaborate (=expensive) fin arrangement but also very non-optimal.

ko (

http://www.ecooling.biz)

 

[cymeryss]

Art,rmw,
Thanks for the posts and I apologize for not giving any more input on this. Sometimes I think its clear for other what I am trying to do, but don't realize that I have spent time looking at it, and others did not. Anyways, here is the exact scenario that should clear up couple of things.
The fins area attached to the aluminum cylinder (motor case) since this is an Induction motor and the fins are used to cool it. The motor uses natural convection for cooling below 100 degC and any higher endwinding temperatures cause an electric blower (20m/s) to flow air through the louvered fins.
Louvered copper fins are very good for forced convective cooling but I am interested in predicting the thermal resistance value for natural convection of the structure. I have created a thermal model to depict the steady state solution of an Induction motor, and this is the one parameter that I am having difficulty to calculate. Also, I am not sure how you would clasify these fins, maybe not radial, but think of a long strip array of fins being attached to a round structure.
I hope this clears up some of the questions that were raised earlier.

 

[amirhirshfeld]

The only source of information I know that has anything useful on this type of problem:

Extended Surface Heat Transfer by Allan D. Kraus, James Welty, Abdul Aziz.

This is a real good book. However, for anything outside the really simply problems (and the authors do warn you about it) the math becomes extremely complicated.

There are other ways that a problem like this can be simplified and modeled. However they would never be real accurate, a sketch would help.

AH

 

[ko99]

Cymeryss,

I really think the best way to answer your question is with a simple CFD model.

There are consultants who do this (I am one of them, but since we are using this forum I will disqualify myself). If you define the scope of your problem carefully, you should get an accurate answer quickly and inexpensively.

ko (

http://www.ecooling.biz)

 

[25362]

Books on heat transfer tend to link the HTC for free convection to the temperature difference (surface to air) as included in the Grashof number.

Considering the thermophysical properties of air in a given room about constant, the HTC for natural convection would be proportional to (delta T)[sup]0.25[/sup].

Although there are formulas to calculate the HTC for radiation and natural convection, for air moving by buoyancy effects over all kind of surfaces, looking up, down or inclined, these estimates can not be accurate and would just be good approximations.

 

[Tmoose]

If your intended natural convection cooling will be cool air in at the bottom, warm air out at the top, I think even angled fins spacing should be about 1/4 inch.

Somewhere I have a 60s/70s tech manual written by a Massachusetts based electronics company company about air cooling enclosures and components. Sitting here over coffee I think I recall they said (vertical?) fin spacing of about 1/4 inch was best/minimum for natural convection (1 directional flow). I'm thinking a cavity (storm window) is R1 per surface for any spacing from 1 to 4 inches, so that spacing has to be too wide to start to limit a 2 way convection loop. The spacing between panes on our 10 year old Marvin Low E windows appears to about 1/2 inch, which would seem to be a good match if 1/4 inch is optimum for 1 way, 1/2 inch for 2 way natural convective air flow.