Copper and Aluminum

[JigPu]

Hey! I have an idea! I don't know if it would work, if it would make temps better/worse, or anything at all, but it is an idea...

What would happen if you electoplated (or plated in any way) a copper heat sink with aluminum?? I read somewhere that copper has a higher heat transfer rate than aluminum, and aluminum conducts heat into the air better than copper... So what if we took the best of both worlds and coated a low resistance copper heat sink with a high air transfer rate aluminum coating?

Just wondering if this would work. Might there be any chemical reactions between the copper and aluminum that would be bad for the HS, or would the small amount of aluminum not make much of a difference? If so, are there any HS manufactures that make copper heatsinks with aluminum fins? I've been thinking of such a heatsink for a while now, and was wondering...

JigPu

 

[Hoot]

I think you want to go the other way. IE Copper clad Aluminum. We all know that pure copper conducts heat almost 100% better than aluminum, but copper is much more dense than aluminum and consequently does not want to relinquish the heat to air or water as readily as aluminum. I would think that the last thing you want in the core of your thermal radiator is copper, but I could be wrong. I am working on a long term experiment with a Swiftech MC-462A. I have extracted the aluminum pins and found a source for copper clad aluminum pins that I intend to try instead. Staking the pins into the copper baseplate has been quite a challenge since I don't have access to more sophisticated metal working equipment, but I am progressing nevertheless. The summer up here is short and I have been consumed with many seasonal maintenance chores on my house, not to mention recovering from a severe hail storm that levied in excess of $15,000 worth of damage to it and the cars. I still hope to complete it before the end of summer, but one must have their priorities in order. It bugs me not being able to devote the time to this project, but then I'm not sure how many people it would benefit. I mean, who else wants to take the chance of trashing an $80.00 HSF? ;D Stay tuned...

Hoot

 

[It_The_Cow]

Best wishes Hoot. Your article will be most interesting

 

[JigPu]

Hoot (Jul 11, 2001 11:48 p.m.):
IE Copper clad Aluminum. We all know that pure copper conducts heat almost 100% better than aluminum, but copper is much more dense than aluminum and consequently does not want to relinquish the heat to air or water as readily as aluminum.

Huh? Sorry, but thats just what I wanted... The copper to quckly transfer the heat to the aluminum which will quickly transfer it to the air. The other way around and the heat is more slowly conducted ("pure copper conducts heat almost 100% better than aluminum") and then is more slowly transferred to the air ("but copper is much more dense than aluminum and consequently does not want to relinquish the heat to air or water as readily as aluminum.")

So use the aluminum for the metal -> air transfer and copper for the die -> aluminum transfer. Aluminum coated copper....

Thats how I understand it...
JigPu

 

[JigPu]

Anyone else have anything to say? I am really interested in wether this would work or not, and I would like suggestions, comments, ect.

JigPu

 

[inertia]

Hoot I know you're a respected member of this board, but I gotta' call ya' out on the copper coated aluminum and say that I think you've got it backwards. But, let us know how it turns out.

Now if ya wanna electroplate a copper heatsink with aluminum, first ya' gotta look at your handy dandy periodic table of the elements. Let's see here, it looks like copper who's symbol is Cu and for which the island of Cypress was named is # 29. Aluminum is # 13. This means something, but I don't remember what. Get an 8 foot piece of wire. Cut it in half. Strip all of the ends. Tie one end of one wire to a copper heatsink. Get some Tums and crush 'em up into a bucket of water. Swish it around. Throw one end of the other wire into the bucket. Hook up the unused ends of the two wires to your DC power supply (could be a battery.) Now, drink a beer. Ok now it should be electroplated with aluminum. Unless you got the polarity wrong. Oh, we determine the polarity with the periodic table. That's what that's for. Um, the copper should be on the hot side. I think. Hook it up like that, if it doesn't work throw away the heatsink and start over, but hook it up the other way this time.

It might be Rolaids. One of 'em is a water-soluble aluminum salt. Use that one. Oh, and hold everything together with Duct Tape. And strip the wires with a Swiss army knife. That's it.

 

[JigPu]

Thanx for the electroplating procedure! I would do it myself, except I don't have a heatsink to mess around with. Both are being used at the moment, and I dont got no backups... Mom would kill me if I messed up her pooter cuz the HS didn't work, and I don't want my pooter conking out...

Would do it if I could...
Anyone else??
JigPu

 

[Hoot]

A few microns of plating will not be enough to realize enhancement. You would probably need something like a mil (.001) or more.

I silver plated my copper baseplate on my MC-462A to approximately 8 microns and it did not result in a measurable improvement in heat transfer from the core to the baseplate.

Hoot

 

[William]

actually I am with hoot on this as you want the heat to spread, but it has to be dissapated better. Put the metal that gives away heat more easily on the inside. Hmm, how to explain it. Essentially aluminum would be an insulator as it has a lower conductance. By insulating the inside you are forcing the heat outwards and into the air.

 

[Hoot]

The son I never had...sigh ;D

Hoot

 

[JigPu]

I have seen a few aluminum heat sinks with "copper cores" but never the other way around. Maby I just don't understand what they mean by "core", but I've never seen a copper heatsink with aluminum core, and so wouldn't the HS manufactures agree with me on aluminum plated copper?

As for the thickness, I was thinking about putting on more than a few microns. I don't know how much, but definatly more than that.

And as for the explination for copper plated aluminum, are you trying to say that since aluminum transfers heat to other substances better, it should be on the inside transferring to the copper? If so, then why even bother with the copper plating? It has lower thermal resistance, but won't transfer the heat to the air as good (the heat would sort of "build up" waiting to get off the copper). My idea was to get the heat through the heat sink as fast as possible (using a low resistance metal) and then get it off the heat sink as fast as possible (using the aluminum which is better at that).

JigPu

 

[3DMike]

Does anyone remember the name of theat heatsink from which the thermoengine is a cheap copy? Its basically the same but the round core piece is solid copper and the fins are Aluminium soldered on. This would seem like a real good idea in the copper/alu tests.

Now where did I see that review??

 

[cjtune]

JigPu (Jul 14, 2001 11:46 a.m.):
I have seen a few aluminum heat sinks with "copper cores" but never the other way around. Maby I just don't understand what they mean by "core", but I've never seen a copper heatsink with aluminum core, and so wouldn't the HS manufactures agree with me on aluminum plated copper?

I think they meant 'base' as the 'core'. You'll need a higher conductivity base for the fins or pins to sit on so that more heat gets distributed to the outer fins rather than just concentrating on the fins just right above the CPU core. A cold heatsink fin doesn't lend much hand in throwing heat to the ambient. Note also that axial fans cannot blow air right smack above the center of the HS, above the CPU core by virtue of their build (anybody wanna try to fix a centrifugal blower to replace the 'normal' fans above a heatsink? The airflow from them blowers are more distributed -ie. uniform mass flow flux). There you have it, two reasons why you need more conductive metals at the base of the fins/heatsink. But of course, these hybrid copper/aluminium heatsinks are either just a gimmick (to sell the aluminium better!) or a economical measure compared to all-copper (or silver!!) heatsinks. Of course, if the CPU core is large enough (old Pentiums, K6s, etc.) so that the entire bottom of the HS is in contact with it, there'll be very little reason to have a copper-cored/based aluminium heatsink.

 

[cjtune]

Oh, BTW, plating an extremely thin layer of highly conductive metal onto a lower conductivity metal would do little to 'spread' heat as it would be quite hard for heat to travel width-wise as the cross-sectional area in the horizontal plane is very small. As all of you know, thermal resistance is inversely proportional to cross-sectional area. What is the optimum thickness? Beats me. 3-dimensional heat transfer is a really tough subject. As for the fins/surface in contact with the air/water, I don't think it really matters whether it's aluminium, copper, gold, silver, or unobtanium, as long as CUMULATIVELY (considering space and cost constraints as well) they appear to be a super-large surface area to your CPU core -but of course, whenever possible, use a highly conductive metal. Take it like this: you can imagine a fin of metal X which is much more a conductive material as metal Y to be of thinner and taller build than a fin made out of metal Y in proportion. Similarly, a heatsink base of copper would appear to the CPU core as a thinner base of aluminium. Of course, the base cannot be infinitesimally thin as the heat transfer in the horizontal plane would be obstructed, as mentioned earlier. Like most things in life, there's an 'optimum curve graph' for it or something like that.

 

[Badger]

The idea that copper does not transfer heat to the atmosphere as well as aluminium is a complete myth.
Given the same surface area, same surface finish (rougher the better) , same temp difference and air flow, both materials will transfer heat to air the same.
The only possible difference would be due to colour, dark matt colours radiate reat better than light shinny ones, but there's not much in it between copper and aluminium (unless the aluminium is anodised black). Anyway most of the heat is tranfered by convection not radiation so colour is pretty insignificant.

 

[JigPu]

Me did not know that... I read that aluminum transfers to air better than copper, and that is why they made some heat sinks out of aluminum. If that IS true, then this whole idea dosen't work... Still I'd like to see a heat sink like that.

If you know of any numbers proving this, I'd like to know (just to convince the reasoning part of myself about this).

JigPu

 

[cjtune]

Badger (Jul 15, 2001 07:38 a.m.):
The idea that copper does not transfer heat to the atmosphere as well as aluminium is a complete myth.
Given the same surface area, same surface finish (rougher the better) , same temp difference and air flow, both materials will transfer heat to air the same.
The only possible difference would be due to colour, dark matt colours radiate reat better than light shinny ones, but there's not much in it between copper and aluminium (unless the aluminium is anodised black). Anyway most of the heat is tranfered by convection not radiation so colour is pretty insignificant.

Yeah I tough heat transfer by radiation was negligable too....until I got that lemon of a Sunon 120mm AC fan. The hub of that fan gets so hot (has an induction motor there) that it is able to pass on the heat to the aluminium casing (warm to the touch)!! Since there is nearly no direct contact between the hub and the outer casing, the heat must have gotten there by radiation. Beware of using AC fans!!!

 

[cjtune]

JigPu (Jul 16, 2001 12:03 a.m.):
Me did not know that... I read that aluminum transfers to air better than copper, and that is why they made some heat sinks out of aluminum. If that IS true, then this whole idea dosen't work... Still I'd like to see a heat sink like that.

If you know of any numbers proving this, I'd like to know (just to convince the reasoning part of myself about this).

JigPu

Most theoretical heatsink and radiator analysis is very involved -to extent of using computer simulations (CFD, FEM). The only numbers you'll get easily, of, course, is through experimentation -that's what WE do here. By sifting through all the heatsink and waterblock reviews here you can see that there are aluminium heatsinks/waterblocks that perform better than all-copper heatsinks/waterblocks. But of course, it's usually the other way around -why, copper IS more conductive (one factor out of many), and the geometry, the shape of the copper heatsinks themselves are already refined, no doubt from previous experiences with ALUMINIUM heatsinks. So, it's more a matter of good geometry but without forgetting that materials (and airflow, etc.) also permits heat to be thrown away more easily to the ambient.

 

[Thelemac]

I'm not really sure if this was implicated or not as I am having difficulty understanding words right now (I'm a tad ti*yawn*red) so here goes:

Even given better conduction to air by AL, coating the CU with it isn't really going to help cause AL conducts the heat within itself much more slowly than the CU...meaning that the CU will be getting the heat to the AL, but the AL won't be taking it that quickly. So you'll have a bottleneck. What you could do to improve CU is get a *really* big fan and stick it on there...CU has a much higher cfm threshold before you get start diminishing returns as it gets heat back to the surface more quickly than AL...so more air would be helpful.

 

[cjtune]

Thelemac (Jul 16, 2001 02:24 a.m.):
I'm not really sure if this was implicated or not as I am having difficulty understanding words right now (I'm a tad ti*yawn*red) so here goes:

Even given better conduction to air by AL, coating the CU with it isn't really going to help cause AL conducts the heat within itself much more slowly than the CU...meaning that the CU will be getting the heat to the AL, but the AL won't be taking it that quickly. So you'll have a bottleneck. What you could do to improve CU is get a *really* big fan and stick it on there...CU has a much higher cfm threshold before you get start diminishing returns as it gets heat back to the surface more quickly than AL...so more air would be helpful.

No, no. Having AL doesn't automatically mean better convection heat transfer to air! Material conductivity is only one aspect of thermal resistance (C/W) of a HSF. It's primarily the way the fins are designed -how tall, how many, how well they let the airflow through. The best possible heat transfer (from a fin) will arise when the ENTIRE fin of a heatsink is at the same temperature as its base (in reality, the temperature at the base of a heatsink is always highest and drops when nearer to the tip since heat has been taken away by airflow or radiation). To have this, you'll need a material with an infinitely high thermal conductivity. The whole fin would have to be made out of that hypothetical material -so why bother with different metals?. To put it in a better way, copper-coated AL fins would be like plastic hot water bottle with a metal outer skin and AL coated copper fins would be like a metal hot water bottle with a plastic outer skin. Why not have an all-metal bottle to cool down the hot water?

Er, have I been beating around the bush?