Convection / cooling in vertical tubes.

[opticsman]

HI , Can anyone tell me if I can get cooling through convection through a small vertical channel ? The channel will be open –ended and about 200 mm long , cross-section 10 X 5 mm. The closed channel surface will face the outside world , the open face will contact a warm mass at constant temperature ( 30 deg. C) . The channel walls are thin ( <1mm) and will be an insulator ( plastic ) . Will I get any cooling if fthe outside temp. is say 40 Degrees or will I make the mass hotter?
Any qualitative tips on key parameters to help convection would help e.g increase cross-section, increase cross-section upwards ( taper) change material type etc. Many thanks . Opticsman.

 

[25362]

I don't full understand, please clarify:

Is the mass static ?
Is it on the concave part of the channel ?
Do you wish to &quot;cool&quot; this mass at 30 deg C by convection with outside air at 40 deg C ?! Remember that by principle, heat flows from the warmer side to the cooler &quot;sink&quot;.
Is the outside air moving (wind?) or is the whole setup indoors ?

Even to give &quot;qualitative&quot; answers one sometimes needs more information.

 

[opticsman]

Thanks Chemical, I tried to simplify things as my query was getting long. The mass is static and the concave part is towards it. assume no wind. I should have asked if relatively , the mass gains less heat than if it were directly covered with the same insulator and no channel. If so , what parameters would assist this?Opticsman

 

[25362]

to opticsman,

Three dimensionless numbers are considered as factors in estimating the HTC (hear transfer coefficient) by convection: the Prandtl, the Reynolds, and the Grashof.
The Pr refers to air proeprties, and here you can do little. The Re can be improved by the use of a ventilator which increases the air velocity. However, if this is out of the question and you are speaking of natural, aka free, convection, i.e. only buoyancy effects move the air around the object, then you are left with the Gr. The Gr depends, among other factors, on the temperature difference, and on the typical dimension of the tube (channel) to the third power. I.e., by increasing the surface you are bound to improve convection.

However, since the HTC depends on the Gr^0.25 this would mean a relation of the HTC to D^0.75. This is in respect of convection.

You mentioned something about thermal insulation. This is bound to affect, by heat conduction, the temperature of the external surface exposed to air. It all depends on what you want to achieve, metals are good heat conductors, plastics aren't.

Therefore, the best approach to cool the mass in the channel would be: enlarge the surface, use better thermal conductors, move the air around. Hopefully, there isn't a hot radiator around that may be acting in the opposite direction.

 

[opticsman]

Dear Chemical, comments much appreciated. Is D a typical dimension of the channel? i.e. could be length as well as cross-section?
If the outside temperature is less then the mass, say 20 deg C. I assume heat will radiate outwards. If I can only change the channel material ( the bits not in contact ) will I assist the cooling process by making that a conductor ? also the key question for me is will I be better off with the open -ended channel as opposed to a flat insulator against the mass or will the column of air just add insulation? Thanks Opticsmman.(PS from your comments, a tapered channel might seem advantageous (assuming the bottom opening must have restricted size) Would a tapered chinmey thus work better)?

 

[25362]

D is a typical dimension, it could be the length or the height. You could use an average of both. Radiation implies there are colder surfaces around. Air is quite transparent to thermal radiation (diatomic molecules) at these temperature levels, i.e., air is considered diathermic. Non moving (enclosed air) is a good insulator, not as good as vacuum, but neverless good enough, that's the reason insulators are made from foamed materials.
So, whatever you modify, pay attention not to occlude air or prevent its circulation, because it is considered a good insulator.

I'm afraid that's all I can add to the issue in hand. Good luck.

 

[cbiber]

If you can post a little sketch, maybe it would be easier to determine what length scale you should use for the dimensionless parameters.

I would recommend keeping things simple with a straight rectangular channel. Tapering is way complicated to make, and given that you're not even sure that this is a solution for you, probably not worth investigating until you have the basic principles down. Are you thinking that air would flow up through the channel to cool the mass?

It's still not clear to me exactly what you are trying to do. &quot;Constant 30 C&quot; temperature -- with or without the channel? Or are you trying to get to a lower temperature? If all you've got is 20 C air to do the cooling, you'll get no lower than 25 C, maybe higher, with the small channel you propose. Or are you trying to keep the mass from gaining heat from a 40 C environment? And is there any solar loading, if we are to assume &quot;no wind&quot; (and it is therefore outdoors?)?

Cathy Biber

Biber Thermal Design

http://www.biberthermal.com

 

[opticsman]

Cbiber,Thanks. I guess key question is this does the channel produce any cooling benefit benefit compared to a plane insulator under various conditions. Opticsman

 

[cbiber]

My gut tells me it's too long and skinny to get any significant flow through it.

If the channel were open to the atmosphere, it would look like the space between two heat sink fins. There is an optimum spacing for the fins, described at

http://www.coolingzone.com/Guest/News/NL_AUG_2001/Cathy/August_cb_2001.html.

If I run the calcs, a fin array 100 mm wide with optimally spaced fins (10 fins at about 10 mm spacing -- just 10X what you have) would give you only 1.5 W of dissipation with the temperatures you specified. And that's assuming that the channel sides act like fins and can actually convect heat from both sides. In other words, the cooling capacity of this channel is low.

 

[opticsman]

cbiber. Thanks. that helps a lot. opticsma.