RE: heat sink cooling

[Roger Stockton <rstockton@delta-q.com>]

Seeing as I made the comment the touched off this debate, I'll offer some
information from the handbook provided by a manufacturer of extruded aluminum
heatsinks...

"The use of a heatsink exploits all three modes of heat transfer: by
conduction through the aluminum baseplate and fins; then by convection
across the heatsink-air interface; and by radiation from the outer surfaces
exposed to the surroundings.  The contribution of radiation is rather small
for forced convection but can be substantial for natural convection - up to
30% in some cases.  However, care must be exercised in the placement of the
heatsink since it can have the detrimental effect of absorption rather than
dissipation if it is located near a warmer object than itself."

"...For a natural convection system, radiation can account for a significant
portion of the overall cooling.  In some applications heat transfer by radiation
can represent as much as 30-40% of the total heat transferred.  The true
amount is dependent on the actual surface and surounding temperatures, as well
as the exposed heatsink surface area and surface emissivity.  ... exposed
heatsink surface area is that which sees its surroundings, and therefore does
not include the internal surface area of the fins that see and radiate to each
other (i.e. a net radiation exchange of zero)."

There are two important things to note here:

- the significance of the contribution of radiation to the performance of a
heatsink is greatly affected by the design of the heatsink regardless of
emissivity (colour & surface finish). E.g. a flat plate offers the same active
surface area for all three transfer mechanisms, while different finned designs
will have greater (and different from one to the other) amounts of surface area
available for conduction and convection than for radiation, so radiation will
contribute more to the performance of one design than another.

- the significance of the contribution of radiation decreases dramatically in the
presence of moving air due to the order of magnitude increase in the convective
contribution (convection heat transfer coefficient for gasses is typically
2-25 for natural convection vs 25-250 for forced [W/(m^2)K])

Regarding the emissivity of various surface finishes, here are some numbers:

Aluminum
 polished			0.06
 oxidized			0.10-0.33
 irridited or chromated	0.10
 anodized			0.70-0.90
Paints
 Varnish			0.89
 Lacquer, white flat	0.80-0.90
 Lacquer, black shiny	0.87
 Lacquer, black flat	0.95
 Enamel, most colours	0.90
 Oil, most colours	0.90

So, while it is true that a polished aluminum heatsink should probably
be avoided under most cases, bear in mind that Ron's test conditions
(still air, inside an environmental chamber) increases the significance
of radiation by decreasing the contribution of convection, and that the
shape/design of heatsink he tested will be more or less sensitive to the
contribution of radiation than your own.

Certainly, the quoted emissivity values do not indicate a significant
dependance on surface colour for this application, as suggested in the
text:

"For temperatures normally encountered in electronic equipment (25C to
100C), the radiation wavelength is relatively long (infrared region)
and colour is relatively unimportant as well.  Therefore, the most
significant factor affecting the emissivity of a surface in an electronic
equipment application is the texture or nature of the surface emitting
the radiation."

To bring this back on topic, as long as Nathaniel mounts his controller's
heatsink where there is going to be moving air (forced convection), he
need not worry greatly about the effect of colour on its dissipative
performance but should choose a colour that will minimize its absorption
of heat from other warm bodies (e.g. the sun) or shelter it from direct
sun by placing it inside a hood scoop, etc. (But the easiest and cheapest
route is simply to obtain a heatsink anodised an appropriate colour...)

Cheers,

Roger.

-----Original Message-----
From:	Colin Dedman [SMTP:cjd121@rsphy1.anu.edu.au]
Sent:	Wednesday, May 31, 2000 3:19 PM
To:	ev@listproc.sjsu.edu
Subject:	Re: heat sink cooling

[snip]

> I have done the above many times.  In still air black anodizing is
> noticably better.  Polished aluminum is the worst.

Thanks Ron. Looks like we won't need to rip up our Physics text books, just yet.

Regards, Colin