Heat Sink Material

[tetwin11]

We are placing three electrical components in close proximity to one another and it has been determined that there must be a heat sink/heat spreader plate to keep the temperature within operating range. What would you suggest as a material to use for this heat spreader plate. I've heard aluminum 6061 and 6063 suggested.

 

[ornerynorsk]

Silver is an outstanding heat sink, if cost will allow. I think the 6000 series aluminums are used chiefly for their extrudability. I don't have my material chart handy to see which aluminum is best for thermal cond.

It is better to have enough ideas for some of them to be wrong, than to be always right by having no ideas at all.

 

[tetwin11]

Thank you for the reply. I would shy away from something like silver for cost concerns. I suppose I would be looking for the highest conductivity/$ material. Does anyone have experience in this field a more economical suggestion?

 

[TheTick]

http://en.wikipedia.org/wiki/Thermal_diffusivity

 

[telecomguy]

Also check out

http://en.wikipedia.org/wiki/Heat_sink

 

[MikeHalloran]

Conductivity is not the sole concern in heat sinking. In fact it's a relatively minor factor. The largest thermal resistance in the system is the boundary layer of air surrounding the heat sink. ... which in turn implies that surface area of the sink is more important than material selection.

However, I suggest that you back up a bit, and drag out the detailed spec sheets for each of the devices you are trying to cool, read the background information provided by the device manufacturer, and attempt to size a commercial heat sink for each device, using thermal resistances and device heat dissipation to limit junction temperatures to safe limits, before deciding that some arbitrary piece of sheet metal is going to be adequate.




Mike Halloran
Pembroke Pines, FL, USA

 

[dhengr]

Aluminum, and then extrude fins on the underside of the plate to disburse heat and keep it standing off the bottom of the cabinet. The attachment of your equip. to the top of the plate will be critical too to assure best transmission of the heat to the sink.

 

[tetwin11]

TheTick,
Good link. It seems I then want the highest diffusivity/$.

telecomguy,
It seems that I am using the wrong term to describe what I am looking for. This article defines a heat sink as " a passive component that cools a device by dissipating heat into the surrounding air". We are really trying to sink the electronics to the sheet metal enclosure. There is no moving air inside, but we can create a path to the air on the outside of the enclosure.

I think the term I am looking for is more "heat spreader".

 

[ajamnia]

I agree with both MikeHalloran and dhenger in that you need to be very mindfull of your heat flow path. As for material zinc is also a good choice.

 

[MiketheEngineer]

How about a computer fan - about $5.00

 

[MikeHalloran]

You really need to do the math on this problem.
If you can handle Ohm's Law, with series resistances, you can do it.


Mike Halloran
Pembroke Pines, FL, USA

 

[IRstuff]

Copper is can used for short thermal paths. For longer ones, you might want to look at heat pipes.

TTFN
faq731-376

 

[1gibson]

X2 on heat pipes, look at some high end CPU coolers. Most of the designs are way over the top, but will get you a good feel for how the heat pipes are set up.

 

[TheTick]

Not saying diffusivity is the only ctiterion. Just that if you are comparing materials in a given design, it says a lot. e.g. if you are comparing grades of Al and Cu.

 

[drawoh]

tetwin11,

Pure aluminium and pure copper are much better conductors than any of their alloys. I agree with the comments above about boundary layers and doing the calculations.

This can be a tough discussion with co-workers. I was analysing an enclosure into which I would put a circulating fan. I worked out that a nylon cover 2mm thick would see a temperature drop of 1.5C[°] while dumping my heat. This meant that nylon, 2mm thick, with a substantial cross sectional area, was a perfectly good heat sink.

You may be working with electronics people who think in terms of electronics analogies. In electronics, you can assume materials either conduct, or they don't. Thermal is not that simple.

JHG

 

[KENAT]

6061 is one of the most readily available Aluminum alloys so for 'general' use is often one of the first considered because it's readily available and hence tends to be fairly cheap.

Posting guidelines faq731-376

http://eng-tips.com/market.cfm?

(probably not aimed specifically at you)
What is Engineering anyway: faq1088-1484​

 

[MintJulep]

Wow! A lot of solutions here.

Hard to see how, as it's not very clear yet exactly what the problem is.

 

[MikeHalloran]

If you are worried about the difference in conductivity between, e.g. copper and its alloys, well, that's analogous to emptying the ashtrays on an airliner to save weight. Boundary layers and mechanical interfaces have orders of magnitude more effect than material conductivity differences.

If the thermal path is quite long relative to its breadth, then, yes, coppermetals may be a better choice than aluminum. (Heat pipes are even better, of course.) However, brass is a better choice than copper because pure copper is so gummy that it gets expensive to fabricate.

The other argument against coppermetals is that they have a high rate of what retailers call 'shrinkage', because of their scrap value, so you have to add security costs, or lack of security costs, to the material costs.


This might be a good time to disassemble some old laptop computers and study the extreme measures used to cool their CPUs.



Mike Halloran
Pembroke Pines, FL, USA

 

[tetwin11]

Thank you all for the advice. This application has a rather short heat path (~5-10 mm) over a large area (200x300mm).

 

[jtowne]

Use 6063-T5. That's what most extruders use. Plus the thermal conductivity is approx 205 versus 180 for 6061-T6. There are trade offs. 6063 extrudes easier that 6061. 6063 has higher thermal conductivity. 6061 machines easier than 6063.