does heatsink requires sunshield?

[ribin]

Hi

I have an electronics enclosure with heatsinks at 360K mounted vertically on outside of box, with heatsink emissivity 0.8, absorptivity 0.3, and solar gain of ~1000W/m^2, dissipating 100W of heat, ambient temp 320K. Heat transfer coefficient for convection estimated as 2 W/Kelvin.

Judgement tells me sunshields (completely shielding 'sinks from solar gain, approx 15mm from 'sink fins) will give lower steady state temps for heatsinks, and hence any components on sink's thermal path.

However, using simple electrical analogy (i.e. summation of heat fluxes, estimation of intermediate temperatures), I get very similar sink temps for both cases, because heat loss from heatsink via thermal radiation is almost eliminated with sunshield, due to high temp and proximity of shield.

Please can somebody point out my error(s).

TIA
Robin

 

[poetix99]

Dear "ribin" (Robin),

I am not sure that you have specified enough of the problem, but I get some general idea, and will give some "stream-of-consciousness" thoughts. (This is part of the fun of these threads; not to mislead, but to "think aloud".):

My sense is the same as yours seems to be - that the sunshield will be beneficial on sunny days.

At the nominal temperatures that you have cited, I think that the majority of heat transfer will be convective. I think that a sunshield will have the effect of maintaining your sink temperature low enough that it does actually cool the enclosure(!!??). There must, of course, be a temperature difference btwn enclosure and sink in order to "drive" that heat exchange in the first place.

- The shield temperature is much less than the sun's temperature as you calculate and compare radiative effects without and with a shield. I do not believe that the radiative gain is the same; shielded and unshielded.
- You can ensure an even lower temperature of the shield by making it two-ply with an air gap. Orient the shield so that it has natural convection flows between the two layers. (I think that this is overkill, by the way.)
- Could the heat sink simply be given a northern exposure (southern exposure in the southern hemisphere), i.e. no direct sunlight? This would negate the need for the shield altogether.

Sincerely.
poetix99

 

[ribin]

Thanks Poetix

I had over looked the considerable temp difference across the sunshield. I am using a composite of a Nomex honeycomb plus laminates, so I have now assumed ambient temp for 'cool' surface of sunshield. This gives more believable results.

Robin