Does aluminum dissipate heat better than copper?

[Finken]

Ok, thanks for the reply..

All of these things that concern overclocking is a big part of my life, and I spend at least some time every day thinking about things like these.
I needed to get some confirmation to know if I was on the right track

About my first post, that was pretty much what I had thought about.

I also hope to hear some more about thermal diffusivity!

EDIT - Btw.. would I then be right in the assumtion that a passive cooler could be Al as well as Cu (that Cu wouldn't perform much better)?

 

[Stedeman]

This tool may help I don't know for sure it all depends on what you are looking for / to do its called QFin Optimized Heat Sink Design and Analysis Tool http://www.icepak.com/prod/qfin/index.htm

 

[Finken]

Oh cool, thanks a lot

 

[NeoMoses]

I think justdoitjerky may know something about heat transfer, but it doesn't sound like he truly understands its applications in cooling PCs. It's true, there is a term called "Thermal Diffusivity" and it is defined as he stated. Unfortunately, he is not looking at steady state cooling. Thermal diffusivity is more a measure of a material's transient properties than its steady state properties. PC cooling is mainly interested in steady state performance, i.e. how will this perform with my PC folding/crunching for 2 weeks?

JustDoItJerky, please prove your statements, either through experimental data or a disclosure of your derivations, including any assumptions made. I would be interested to see this.

 

[AMD FanBoy]

yea considering most high end heatsinks are made with all copper, copper cools better. It conducts heat like 1.5 times more than alluminum

 

[a c i d.f l y]

I would just like to state that the main reason copper was discontinued for use in automobiles is its tarnishing/oxidating factors, and how it corrodes with water and anti-freeze. This, and it weighs a ton more, and is much harder to manufacture (due to its much higher melting point). I am not too sure about the mallability (I know I butchered that word, I'm up late) of the two elements, but from my experience, copper forms less easily than alluminum.

It is all in the manufacturing and weight that limits us.

This is why we all do not use SILVER as heat sinks. It would tarnish. Same goes for GOLD. Both have higher thermal conductivity than copper or alluminum. However, a solid gold heat sink would be disgustingly heavy, and the pressure applied to hold it in place would more than likely deform the metal as well.

.....and that was probably useless information.

...but I am strongly for the belief in the 'heat pipe' theory. Stick a cylindrical copper piece against the core of the cpu, and wrap it with thin alluminum fins. Decreasing the tarnishing of the copper to only the base (which would be sealed by grease/paste) and allows for the alluminum sized heat sink with copper's conductivity of heat.... the biggest problem, and has always been the problem with hybrids, is the conductivity/convection of heat between the copper and alluminum.... who has information on this?

...something I just thought: couldn't there be a molecule that would conduct heat better than gold, yet be strong, light, and not tarnish or be resistant to tarnishing?....

-Frank

 

[Gnufsh]

Gold does not have better thermal conductivity than silver or copper.
# Silver 430
# Copper 390
# Gold 320

 

[a c i d.f l y]

Gold does not have better thermal conductivity than silver or copper.
# Silver 430
# Copper 390
# Gold 320

Maybe I misread my source.... but even still... I want a silver heat sink / radiator?!

-Frank

 

[Since87]

This is why we all do not use SILVER as heat sinks. It would tarnish.

...something I just thought: couldn't there be a molecule that would conduct heat better than gold, yet be strong, light, and not tarnish or be resistant to tarnishing?....

The reason we don't use silver for heatsinks is that it would cost more than its weight in umm.... Silver!

As a matter of fact, I know of several silver waterblocks that have been made.

As far as a, "molecule that would conduct heat better than gold, yet be strong, light, and not tarnish".

It's called diamond. Half the density of copper, five times the thermal conductivity, virtually indestructible when used correctly.

One guess why we don't use it.

 

[a c i d.f l y]

I read in another post that there is form of 'fake-diamond' (true diamond molecules with all the properties of diamonds, except they're made, and not natural), and that were pretty cheap for use in machine that needed that kind of strength to grind through tough ores and such...... wouldn't it be interesting to get that kind of production to have windows made out of it? 2" thin glass on space ships?... heh, reminded me of Star Trek IV, "Transparent Alluminum"... *tangents*

-Frank

 

[Carnil]

I read in another post that there is form of 'fake-diamond' (true diamond molecules with all the properties of diamonds, except they're made, and not natural), and that were pretty cheap for use in machine that needed that kind of strength to grind through tough ores and such......

The problem with artificial diamonds is that they're quite small. If someone were to make a heatsink out of them, the thermal resistance of the junctions between crystals would more than counter any gain from the high thermal conductivity of the diamond. A good heatsink or waterblock would have to be a single crystal, and they're nowhere near making diamonds that large.

 

[Finken]

Yeah, we can make diamonds using a huge amount of preassure.

 

[zexmarquies01]

So, would taking a Copper HS, and say........using Aluminum Fin's do any good?

Or like as someone else said, having a coper HS, and then say.....plating it with Aluminum. it could work, except for the fact of the problem of heat transfering between the two metals.

If say, we could get pas the problem of heat transfering between the metals, could we take Aluminum, pit it over a Copper HS ( like plat it ) but on the bottom of the HS, where it touches the die, leave that the original copper base. The copper base would absorb the heat quickly, and move to the aluminum, then because aluminum can dissipate the heat quicker, the air would push the heat from the Aluminum pretty fast.

and you'd want the aluminum to be thin. because as aslready stated, it can't hold as much heat as copper. but it can dissipate it faster.

So, could using aluminum fins, or plating a copper HS in aluminum but leaving the bottom un-plated of aluminum.......would that do any good?

That is if we could somehow fix, or at least make the problem with heat going from one metal to the other a bit better, where we could actually gain cooling abilites with the hybrid?

i bet thie Q has came up before. But i thought i'd ask it in here. get some real technical Answers.

Nothing wrong with idea's, thats how computers came to be in the first place......with an idea.

 

[Albigger]

So, would taking a Copper HS, and say........using Aluminum Fin's do any good?

I doubt it.

Or like as someone else said, having a coper HS, and then say.....plating it with Aluminum. it could work, except for the fact of the problem of heat transfering between the two metals.

If say, we could get pas the problem of heat transfering between the metals, could we take Aluminum, pit it over a Copper HS ( like plat it ) but on the bottom of the HS, where it touches the die, leave that the original copper base. The copper base would absorb the heat quickly, and move to the aluminum, then because aluminum can dissipate the heat quicker, the air would push the heat from the Aluminum pretty fast.

and you'd want the aluminum to be thin. because as aslready stated, it can't hold as much heat as copper. but it can dissipate it faster.

No. It has been established that at STEADY STATE heat transfer, which is ideally (and practically pretty close) what we are dealing with in CPU cooling, copper is ALWAYS better.

The ONLY time aluminum would "dissipate" heat faster is under transient (non-steady) conditions.

So, could using aluminum fins, or plating a copper HS in aluminum but leaving the bottom un-plated of aluminum.......would that do any good?

No, I don't think so, based on what has been discussed here.

That is if we could somehow fix, or at least make the problem with heat going from one metal to the other a bit better, where we could actually gain cooling abilites with the hybrid?

i bet thie Q has came up before. But i thought i'd ask it in here. get some real technical Answers.

Nothing wrong with idea's, thats how computers came to be in the first place......with an idea.

The heat transfer between the metals is still a real issue, one that is being researched by many I'm sure. But even without this barrier the copper would still perform better.

Aluminum doe NOT dissipate heat better than copper at steady state conditions (except for maybe some very obscure geometries, but not for 'normal' heatsinks/waterblocks).

 

[bluce ree]

Here's the deal guys.

Copper absorbs heat better.
ALuminum dissipates heat better.

Copper costs more
Aluminum is very cheap.

Copper is more difficult to machine and to work with
Aluminum is easy to machine and work with.

Both are easy to solder to--as long as you heat them up A LOT--remember copper absorbs heat easier.

Copper weighs more
Aluminum is very light

Electric conductivity

silver conducts at 105%
copper conducts at 100%
gold conducts at 70%
Aluminum conducts at 35%
steel conducts around 12%
magnesium around 6%
cadmium around 5%
vanadium around 1%

What does this tell us? Use Copper tubing and aluminum fins--many radiators are made this way, and have been made this way for 50+ years. Unfortunately, when 2 unlike metals are put next to each other, a "battery effect" occurs, that is, the 2 metals have different amounts of valence electrons in their last shell. this means that over time, the metal with more electrons in its valence shell will cause the other metal with less electrons to rust.

or maybe i got that backwards.... oops

i forget chemistry.

 

[Albigger]

Here's the deal guys.

Copper absorbs heat better.
ALuminum dissipates heat better.

This is true, in the strict sense of "absorb" and "dissipates." HOWEVER, stating this (without further clarification) leads MANY people to believe that aluminum will be better at getting rid of the cpu's heat, when in fact that is NOT THE CASE most of the time.

except for startup and shutdown situations, the amount of heat entering the heatsink is the same as the amount leaving the heatsink, therefore, it doesn't matter if you have copper or aluminum at steady state, the "absorption rate" (if you want to call it that) is equal to the "dissipation rate."

the only thing that matters at steady state is the thermal resistance and the convective coefficient to the air - both of which are covered more thoroughly in the beginning page of this thread.

What does this tell us? Use Copper tubing and aluminum fins--many radiators are made this way, and have been made this way for 50+ years. Unfortunately, when 2 unlike metals are put next to each other, a "battery effect" occurs, that is, the 2 metals have different amounts of valence electrons in their last shell. this means that over time, the metal with more electrons in its valence shell will cause the other metal with less electrons to rust.

or maybe i got that backwards.... oops

i forget chemistry.

I could be wrong but I think many radiators are made that way because, as you stated, copper is a pain to machine, and making a radiator with all copper fins would be too much hassle for what its worth.

in the CPU category, most heatsinks/waterblocks CAN be made in the SAME DESIGN out of copper as out of aluminum, so this is not a factor, and copper should be used (in most cases).


I'm not 100% on the battery effect thing, so I'll leave that to others (or out of this thread, as it might not even pertain to the topic).

 

[Ghoztie]

Just wanted to state some of my own experiences and knowings. I have researched this too in the past.

Copper absorbs the heat faster than any other material (except Silver. Thats the best? ). But it has tendensies to wanna keep the heat more than aluminum does.

Result:
Copper is actually the best. It has a whole lot more thermal absorbation than Aluminum but it will not dissipate the heat as fast as Aluminum.

Haven't read the whole thread. Just wanted to share experiences and info. over and out.

 

[kct2]

People are saying aluminum dissipates heat faster than copper, but not providing any proof. Can someone state the physical property that determines this trait? What determines copper's “tendensies to wanna keep the heat more than aluminum does?”

It seems people are using the word dissipate to represent convective heat transfer. The convective heat coefficient is not a material property; it is a property of the cooling fluid, not the material being cooled.

 

[Thingi]

Except for startup and shutdown situations, the amount of heat entering the heatsink is the same as the amount leaving the heatsink, therefore, it doesn't matter if you have copper or aluminum at steady state, the "absorption rate" (if you want to call it that) is equal to the "dissipation rate."

the only thing that matters at steady state is the thermal resistance and the convective coefficient to the air - both of which are covered more thoroughly in the beginning page of this thread.

I'm not entirely sure that what you've just stated is accurate. I can understand the reasoning behind it, but have you taken into account the increased temp of a CPU at full load compared to when it is at idle?

A CPU is rarely at a uniform temp when a computer is on so I'm not convinced that the "dissipation rate" will equal the "absorbption rate" under most circumstances.

This arguement is of course completely irrerevelant if the cpu is being 'primed - that's the only time I can see the heatsink reaching it's steady state - and that's exactly what I use to make sure my computer is stable.

thingi

 

[bluce ree]

Let's remove the "guesswork"

your link on "specific heat" for aluminum & copper --this is a material's ability to absorb heat with reguards to the temperature you applied to it, the amount of material there is and the actual temperature the metal reaches.
http://www.haverford.edu/educ/knight-booklet/heattransfer.htm


LOTS OF HEAT INFO:
http://tshutt98.princeton.edu/~p101/Lectures/l8_02.pdf

From the looks of it, we're all wrong--there is no 1 answer--it all depends on the shape and amount of material being used.

Copper appears to absorb the most heat per gram & it appears to conduct the most heat, from the hot side of the material to the cold side, ie from the CPU die to the copper fins.

the main thing to consider is that it's a pain to machine, it's very heavy compared to aluminum and A LOT MORE EXPENSIVE.

the problem is, we're talking about PURE COPPER and PURE ALUMINUM in both of these sites. Most of us will only have access to some copper alloy--not pure 100% copper or even close to it. Scratch a penny and you'll see it's made of zinc, only coated in copper--just like it says on the pennies--Caveat Emptor--Let the buyer beware.