Sorry to disappoint you guys, but it is a fact that aluminum sheds heat faster than copper, information backing up this statement abounds everywhere. Does the words"Thermal Resistance" rings a bell?

Geesh a poor n00b like me managed to get three accomplished geeks in a row to bite dust today. Quite an accomplishment if you ask me
:evilaugh:

iggiebee,

Please read this article. I've linked to one of the more relevant pages, but the whole article is good.

You really don't understand what you are talking about.

Gloating just makes you look like even more of a fool.

I'll never make a silver block, but I know how to plate bullets in silver now, if it's any help ;)

Please, no need for name calling to prove a point; totally uncalled for, it leaves a lot to say about the person making such comments, and does nothing to further validate your arguments. It would have been better and mature in the first place, to post relevant links and arguments instead of going into a trigger happy, newbie bashing, witch burning, and vampire staking expedition.

Having said that, the article in question is well written and quite extensive, BUT you have to admit that does not seems to have been written by any expert authority in such matters, and has to be taken with a grain of salt, as such most of what is said are simple "personal" hypothesis IMO.

I found some contradictions: "Copper's higher conductivity means is that a thinner copper fin can transmit as much heat as a thicker aluminum fin.".. This affirmation is very hard to swallow, what are the basis for making such statement. Are thin fins impossible to make if using aluminum?

The the author contradicts himself when saying that: "However, on a weight-basis, aluminum can conduct more heat than copper. "

I am including a link to relevant information to support my statements written by a Ph.D. at Advanced Thermal Solutions, Inc.


http://www.coolingzone.com/Guest/New...S_Feb2003.html

Well, that's the problem isn't it? If there's the possibility of contradiction, then it's in error?

Read it again: it's correct. I believe that you're confusing the material density, with the thermal resistance.

The machinability of Aluminium is not in question.

You have to consider the article in context of what me/we are trying to prove or disprove, which is the argument that "Aluminum sheds heat faster than Copper". So in this context, yes the author is in error when makes a claim to the contrary IMO.

I am not questioning the machinability of Aluminum, but the author of the article most certainly is, more specifically by omitting such an important fact, in order to advance his conclusions.

What is important is that in general you will not able to find one single shred of experimental data supporting any of the arguments exposed by the author at amdmb.com's article. Please consider that I am not denying the (very valid) use of copper plate on a very concentrated heat source, such as is the case of an Athlon CPU core, but as pointed out in the article I linked in the prior message, you should leave open the possibility that same or higher efficiency can be achieved by optimizing for Aluminum. There is no shame in admitting these simple facts.

Actually, there are very few involved in watercooling forums who have anywhere near the expertise of Dave Smith.

The basis for the statement is contained in the statement.

Copper has a higher thermal conductivity than aluminum. Do you dispute this?

Granted that copper has a higher thermal conductivity than aluminum, then by definition a thinner section of copper will move the same amount of heat for a given temperature differential than a section of aluminum.

Manufacturing feasibility has nothing to do with it.

There is no contradiction, the earlier statement was discussing heat conduction as a function of the volume of the material.

This statement is discussing heat conduction as a function of the mass of the material.

What contradiction do you think you see?

Here's a quote from your link (with some added commentary pointing out some flaws in the author's statements):

"This model demonstrates that an Aluminum heat sink is as good as a Copper heat sink for a uniform heat source. [Previous sentence should have said 'nearly as good'. The author's own graphs show that aluminum is not "as good" - Since87] When the heat source becomes smaller a Copper heat sink might be advantageous depending on the source size. [The fact of the matter is that when cooling small CPU dies, the lower spreading resistance of copper is a very significant advantage.] Also at higher velocities, percentage difference between Copper and Aluminum becomes more pronounced. [Translate this to 'At higher convection coefficients such as when watercooling rather than aircooling, the percentage difference between copper and aluminum becomes significantly more pronounced.']"

Furthermore, the author's statement:

"Optimally designed Aluminum heat sinks meets or exceeds thermal performance of Copper or Copper based heat sinks."

Only holds true when weight is considered as part of "optimally designed". When weight is not a consideration, copper always beats aluminum for steady state heat transfer. The author of that article has not stated, much less proved, that aluminum exceeds the thermal performance of copper when the heatsinks compared are of equal volume and geometry.

Edit: Punctuation

The article that you linked to, has one drawback that you seem to have missed: it used the exact same dimensions, for both the Aluminium and the copper heatsinks that are tested.

If you understand how significant that is, then you understand that because of the different natures of the material, the design needs to be different, and that comparing the same exact dimensions in both materials is flawed.

Let me give you a follow up scenario: take the same heatsink, and lower the base thickness to 2 mm. what you'll find is that the Aluminium heatsink performs worse, because it's unable to dissipate the heat which is now too concentrated, and the copper might perform a little better, because it is more than capable of handling the concentration.

That's theoretical, and not fact, but that's the essence of what I'm trying to relay.

The author probably took an off-the-shelf Aluminium heatsink that was optimized: it was specifically designed to run at the highest level of performance that Aluminium can give, under the expected conditions. If you took a design that was optimized for copper, and reproduced it in Aluminium, the difference would be huge.

Typically, copper HS will have thinner fins, and a thinner baseplate than Aluminium. The Aluminium counterpart of the copper-optimized design would perform dismally.

If you caught any of the above, then I'd like to get into a discussion of the fin-to-gap ratio, for different fluids ;)

Please quote arguments of Dave Smith's, that you don't think experimental support can be found for.

Dave acknowledges that when the heatsource is removed from two heatsinks, that only vary by material, the aluminum heatsink will cool faster than a copper one.

He also points out that this is irrelevant to the steady state heat transfer of the heatsink.

Interesting indeed.

Just ignore him guys. He just wants to get you guys going.

heh, this topic always spawns a bit of a furor...



About the AMDmb article - that particular paragraph you picked out is indeed unclear and possibly misleading. Attempting to quantize heat in this manner is confusing in this instance, when being able to conduct "as much" or "more" heat is easily misunderstood without further definition of the statement.


[Edit: I turned out to be wrong. see my next post below]
In particular, the statement "on a weight-basis, aluminum can conduct more heat than copper" is in fact incorrect.
The conductivity value of these metals is measured in W/m K - Cu being 385 and Al 205. Density and mass/weight is no part of this value - they only come into play regarding specific heat, of which aluminum does have a higher value than copper. However specific heat has no relevance whatsoever to conduction, and is of no consequence in a system in thermal equilibrium.

Another thing to note: any weight/mass distinctions previously made are pretty much inapplicable, since on the surface of the earth they're functionally equivalent.


The rest of the article seems accurate, though, during a quick skim of its contents.

I believe that what Dave was trying to convey, was that given two rods of equal length and mass, one made of copper and one of aluminum, the aluminum rod would conduct more heat for a given temperature differential between the ends of the rods, than the copper rod. Is this not correct?

What weight/mass distinction were you referring to?

Nice to meet you too JayDee :)

Was wondering how I managed to make four posts without getting the "mandatory JayDee Slam". It must be a record :D

I havn't slammed you and don't slam newbs unless they need it. You done a good job slamming yourself though with hypocritical remarks. Go troll another forum eh?

Wha can I say..yehh you're some fine descent person.

Abvious attempt to start shit.

Just called 3 senior members "accomplished geeks" and praises yourself for being ignorant as you are wrong.

No need for name calling? Ummmm what was the previous post about then ehh? Your allowed to name call senior members of the forum names and the senior members are supposed to not do anything about it. Interesting logic.
Your taking it out of context. We know Dave as an engineer and as one who knows what he is talking about. We are not about to bow down to someone like who who posted BS as in the above quotes.

Flame bait.
More flame bait.

Your posts speak for themselfs. You have absolutly nothing to back yourself up about me trying to slam you or you questioning MY character. As you obviously read some of my posts you know I have no problem at all handling the torch and don't care if someone thinks I am an ass for it. I back myself up though, which most do not. So carry on with your BS if you want or we can get to talking about relevant info like water block design and construction and make some progress on that subject which this forum is here for. Your call.

First, jaydee and iggiebee, take it to PMs if you are so inclined.

Yeah, it seemed like he was trying to simplify things, but didn't word it quite well enough.

Well, i'll do the calculations for the two metals.
Let's assume a rod of 1kg mass, 1m long, with a permanent temperature differential of 100°C across it.
For aluminum, its diameter will be 1.085cm, for copper, 0.597 cm.

(a few calculations later)
the amount of heat transferred is:
Copper= 431.20 Joules
Aluminum= 758.09 Joules

(oops, proved my last post wrong. guess the values i was thinking of weren't for thermal conductivity after all.)

Aluminum transfers just over 75% more energy than copper does, given a rod of matching mass and length - not to be confused with size/volume. Were these rods the same length and diameter, copper would have transferred 89% more energy than aluminum.


Again, oops. In my first read through the thread I thought I remembered someone making that distinction...

That's what I got, too. It's not implicitly obvious, but it can be checked pretty easily.

The density of Al is 2700 kg/m^3 and the density of Cu is 8920 kg/m^3.

The thermal conductivity of Al is 2.37 W/cm-C and Cu is 4.01 W/cm-C

Writing out Fourier's law for two cylindrical bars of equal length, equal mass, and equal heat load, I get:

delta-T2 / delta-T1 = k1/k2 * rho2/rho1

So if "2" is aluminum and "1" is copper, you get:

delta-T (Al) / delta-T (Cu) = 4.01/2.37 * 2700/8920 = 0.512

So, for cylindrical rods of the same mass, you're going to get around twice the transfer of heat (or half the temperature differential) using aluminum.

Wow, this really takes a while without my texts. Maybe I should keep this computer in my office . . .

Alchemy

Edit - adjusted thermal conductivities. Thanks, flyingass!

So, anyway, back to your first post in this thread iggiebee, and the thing we were about to lynch you for ;)
Please tell us what the thermodynamic equivalent of "shedding heat" is. Conduction? Convection? Radiation? Violation of Conservation of Energy?



Wait, Alchemy, why'd we get slightly different results?
Hmm...i used 205 W/mK 385 W/mK for Al and Cu thermal conductivity, respectively. Why do our values differ?

[Edit]: Found tables - Thermal conductivity varying with temperature. At 300°K/27°C
Al at is 237 W/mK, Cu is 401 W/mK.....
Looks like you were using values for 100°K.

I guess it doesn't really matter since the point is proven either way.

Why, this shit doesn't need to be brought up for the millionth time. Use the search option and dig up the threads that all this has already been covered in and add to them. Let this die as it should.

http://forums.procooling.com/vbb/sho...&threadid=4232

http://forums.procooling.com/vbb/sho...&threadid=1457

http://forums.procooling.com/vbb/sho...&threadid=4760

http://forums.procooling.com/vbb/sho...&threadid=6402

http://forums.procooling.com/vbb/sho...&threadid=6404

http://forums.procooling.com/vbb/sho...&threadid=5707

http://forums.procooling.com/vbb/sho...&threadid=5321

http://forums.procooling.com/vbb/sho...&threadid=4177

http://forums.procooling.com/vbb/sho...=&threadid=476

Flyingass and Alchemy,

Many thanks for being so gracious and doing all that math. God knows I am totally math handicapped, having studied only Architecture in a recent incarnation. ;)

Found this point of interest in another forum somwhere:
"the specific heat of aluminum, by mass, is quite high - about twice that of iron and nearly 2 & a half times that of copper. However, aluminum has roughly a third of the density of iron or copper, so it's specific heat by volume is nearly the same."

IGGY, don't take the;

"AAAAAAaaaaaaaaahhhhh!!!"

"Quick somebody get me a wooden stake!!!"

"::strangles someone::"

"I got some wood, gas, and matches"

personaly, it's an old 'in house' joke about this subject(there's a couple others too), that's all. Nothing personal. They could have been more gracious and explained this to you before jumping in for the kill (cause you are wrong about alu shedding heat quicker :shrug: ) but that would spoil the sport ;) (which some DO take a bit to far eh?)...


With the web something is'nt true just because it's said alot in alot of places. My personal theory is that people think alu sheds heat quicker because it does cool down quicker once the heat source is removed, but while heat is being applied constantly this just does'nt apply...

This argument is old but it has a place for the newcomers. It's the personal bickering, accusations of trolling, 'shit starting' and noob slamming that belong in PM's IMO...

What is Mass by definition?. Is it weight x volume?...

What's yoru ambient temp? :)

a material that can absorb heat faster than it can shed it, by mere virtue of it's chemical character....

"thermal resistance of a substance such as Copper or Aluminum has no relevance to the direction in which the heat is moving."

think about it, but not too hard... :D

oh how I wish such a substance was in existence, I would be the richest man in the world today.