Does aluminum dissipate heat better than copper?

I have seen this debated a LOT, and looked for true scientific research on the matter, but via Google I have not been able to reach a definite answer. A search on the boards yielded no results in the Technical Discussion forum, and opinion-slinging in the General Cooling forum (and every other forum i've visited).

Some arguments that are made:
1)Copper has better thermal conductivity than aluminum, and therefore is superior in dissipating this heat to the air.
2)Despite copper's better thermal conductivity, it is denser than aluminum and therefore keeps the heat in more. (A less dense structure allows for better airflow through that structure, at the molecular level? I don't understand the idea behind this argument)
3)Aluminum radiators work better in cars than bronze/copper ones. This argument does not apply here - a quote from http://www.dewitts.com/pages/whyaluminum.asp

The trick to better cooling is wider tubes. This increases the “tube to fin” contact area, which determines the radiator efficiency. A typical copper radiator uses 3/8” wide tubes (Fig. 1) while the aluminum radiators (Fig. 2) use tubes from 1” to 1 1/4” wide. When a radiator is designed with wide tubes, the tubing wall thickness must be increased to prevent the tube from expanding or a term known as “ballooning”.

Click to expand...

So, in the end, does aluminum dissipate heat better than copper, meaning that a well-done (defined as a good connection between the copper and aluminum - this is possible, i believe) hybrid heatsink is better than a copper heatsink of the same mold?

This is an opinion, but I'd say copper is better. There was an article linked to a year or so ago that proved this out.
Aluminum has the ability to absorb heat faster than copper, and when removed from the heat source, will cool faster because it is less dense than copper.
BUT..in a system with steady heat input, like a computer cpu, copper is better at keeping heat going into and out of the metal, much as it is with electricity.

I also like copper over aluminum for it's solderablity, and lower ammount of destruction when faced with corrosion.


I wish I'd saved that link, but alas, that was many formats and two computers ago.

The amount of heat disapated is dependant on the surface area of the material. There are other variables such as the substance the heat is dissipated to and the flow rate of the substance.

Air cooling is the most common form of cooling. The best preforming heatsinks have a large surface area. When it comes to the amount of heat dissipation the values that matter in calculating the actaul amount is the surface area and the differnce in temperatures of the heatsink and the adjacent air. When you apply a constant load onto the heatsink then the temperature of it will rise untill the amount of heat entering the heasink and the amount of heat dissapation are equal. The airflow improves the amount of dissipated heat because it provides cooler air to the surface of the heatsink. With no airflow, the air warms up around the heatsink and has to dissipate heat through the air.

The reason copper heatsinks with the same surface area as aluminum preform better is because copper transfers heat better. The temperature drop from the die to the surface of the heatsink is very important. If the drop is 20c then the heatsink can't cool very well because there isn't as much of a difference between the air and the heatsink surface. A perfect heatsink setup would have it's entire surface area the same temperature as the die. This would provide better cooling than a real world situation.

Water cooling is a different story. It provides a means of transfering heat to a radiator via water. The reason it works much better than air cooling is that water holds much more heat for the weight than air or even metal. Also if I'm not mistaken, it transfers to and from a metal 32 times faster than air. Basicly you transfer heat from a very small waterblock to a big radiator to get much more surface area.

The reason you get better radiator cooling is because aluminum radiators have much more surface area to transfer heat.

The reason copper heatsinks with the same surface area as aluminum preform better is because copper transfers heat better.

Click to expand...

GigaForce310 - you took it for granted that copper heatsinks perform better than hybrids. is this necessarily true?
and i understand that surface area, ambient temperature, the medium of heat transfer (air or water) all make a big difference, but is it true that aluminum dissipates heat better than copper, with a given airflow, equal surface areas, etc?
your post implies that you think copper is better in every way, but it's never directly stated, and there's no empirical data to back it up =(

As I recall. The formulas for determining convection from a solid to a fluid do not include a variable for the composition of the solid. No solid material "gives off heat" better than any other.

Here's a relevant thread.

http://forum.oc-forums.com/vb/showthread.php?s=&threadid=14199

and

http://forum.oc-forums.com/vb/showthread.php?s=&threadid=14419

(first I want to apologize for not yet reading the corresponding threads that nihili posted, but i will get to them)


But what nihili (and GigaForce) said is correct, there is no variable for heat transfer from a surface for the surface composition, it only depends on the fluid (air or water) flow and other related properties OF THE FLOW AND THE FLUID, but not the metal/surface (as long as the same surface areas are considered, roughness, etc...).

Thus for a typical heatsink, the one with the HIGHEST thermal conductivity should always (in theory) perform the best, because it provides the lowest temperature drop inside the material, which means a higher temp. differance from the surface of the metal to the air, which results in higher rate of heat transfer.

The general consensus seems to indicate that aluminium radiators dissipate heat beter than copper radiators better than copper ones for three main reasons.

1. Copper radiators use soldered joints which are well known to be poor thermal conductors.

2. Aluminium (Al-u-min-e-um) radiators usually have a greater internal surface area than that of traditional copper radiators. This is obviously an importatant factor.

3. Aluminium radiators can have a greater external surface area (fins) than a copper radiator due to the dirrerent ways copper and aluminium are machined.

Now when comparing heatsinks instead of radiators the following factors need to be taken into account:-

A. Is the copper heatsink skived or is it solder jointed?
B. Does the aluminium heatsink share exactly the same design and surface area as the copper heatsink?

The skived copper heatsink would perform better than an all aluminium heatsink. If the copper heatsink has solder joints then it's possible that it could be better or indeed worse than an aluminium heatsink of exactly the same design.

Pressure fitted copper cores inside hybrid heatsinks offer the best of both worlds provided the bond between the two metals is good enough. Firstly the wieght of the heatsink can be significantly reduced. Secondly the shape of the pins/fins of an aluminium heatsink can be can far more dynamic/efficient than copper due to the differing construction techniques.

A hybrid copper/aluminium heatsink could in thery be more efficient than either an all copper (skived or soldered) heasink or an all aluminium heatsink.

thingi

it's aluminium in britain, in the united states it's pronounced aluminum and in fact SPELLED aluminum, just fyi
and i don't know the aussis' stance on the matter

i don't think heatsinks are subject to the same restraints that radiators are, in terms of machining them, so i don't think the extra "machinability" of aluminum could really help it on a heatsink. in any case i haven't seen an aluminum heatsink design that couldn't be done in copper, that i can remember.

in terms of heatsinks then, all copper will always perform the best

Here is a post i ran into at another forum:

Now about the aluminum copper thing..

Take 2 same sized blocks of metal... one aluminum one copper... heat them to the same temperature. Now monitor temperatures as they cool... one probe in contact with the metal, one a half inch above it's surface, note what happens... The copper block will stay hotter, longer with lower free air temps. The aluminum will cool faster with higher free air temps... because the copper, being higher mass, will retain heat longer.

Yes, because of it's higher thermal conductivity copper soaks up more heat more quickly, but because of it's higher mass it's going to STAY hot. Aluminum isn't as good a heat absorber but, because of it's lower mass, it releases the heat more quickly.



what do we make of this whole more mass = holds onto heat better idea?
i think that to get the copper block to the same temperature as the aluminum one, more energy must be put into the copper block, because if it is denser (more molecules to heat up) then the same amount of energy will not heat up each molecule as much as in aluminum. so in this example, the copper block must dissipate more energy than the aluminum, and hence the aluminum gets rid of it faster (because they both give off the energy at the same rate to the air).

make sense?

OK, specific heat (Cp) is the measure of how much energy it takes to raise the temperature of 1kg of material 1degree (C or K). The Cp for copper is 385 Joules/kg*K and for aluminum is 903 J/kg*K (at 300K = 23 deg. C). density for both is copper: 8933 kg/m^3 and alum: 2702 kg/m^3

So, if we have two same SIZED blocks of material, say 1m x 1m x 1m for simplicity, we have different masses, namely 2700kg of alum. and 8930 kg Cu. if we put 2.438 x 10^6 Joules of energy into each block, we can calculate the temp raise:

Q = m*Cp*deltaT --> deltT = Q/(m*Cp)

deltaT alum: = 2.4381 x 10^6 J/(2700 kg * 903 J/kg*K) =1 degree

deltaT Cu: = 2.4381 x 10^6 J /(8933 kg * 385 J/kg*K) = 0.71 degree

so for the same SIZE blocks, more heat is energy is required to heat the copper by the same delta T than the aluminum. In other words from the moment you turn your computer on it will take a longer time for a copper heatsink to reach steady state conditions.

Now since these blocks have the same size they have the same surface area (that's what we're comparing, right - same size/shape/etc... heatsinks)

Now, that is really transient heat transfer, where there is a buildup of heat in one of the components of the system (the heatsink). At steady state, BOTH Al. and Cu. heatsinks will stabilize at a certain temperature and the heat per unit time (J/s = Q_dot) will be exactly equal to the heat leaving the heatsink per unit time. It will just take a longer time for the Cu one to reach this point.

Now we're talking steady state:
2 processes going on:
(1) is conduction of heat through the heatsink
(2) is convection of heat from surface of heatsink to air

for conduction: Q_dot = (kAc/L)*(Tdie - Tsurface) <-- equation (1)here Q_dot is joules/second or Watts, k is the thermal conductivity of the material in Watts/meter*K, Ac is cross sectional area through the heatsink in meters^2, L is the lenth between Tdie and Tsurface in meters, Tdie is temperature at the die or temperature at the bottom of the heatsink (here we'll consider them equal) and Tsurface is the temp. at the surface of the heatsink, where the air contacts it. (temps in degrees C or K)

for convection: Q_dot = hA(Tsurface - Tair) <-- equation (2)
where A is the area of the surface of the heatsink (giving heat to the air) and h is the convection coefficient and has units Watts/meters^2*K.

from equation (1): Tdie = Q-dot * L/(k*Ac) + Tsurface
from equation (2): Tsurface = Q_dot/(h*A) + Tair

combining: Tdie = Q_dot*L/(k*Ac) + Q_dot/(h*A) + Tair

Taking the Tair, Across section of the heatsinks, A surface of the heatsinks, L length of the heat must travel between the die and surface, and Q_dot to be constant and the same in both cases (Ac, A, L wil be same because we are comparing the same heatsinks just of different materials, but the physical dimensions are the same, Q_dot is constant because the cpu puts out a fixed amount (ideally) say xxxWatts under load, and Tair is constant for it to be a fair comparison)

k for alum = 237 W/m^2*K
k for Cu = 401 W /m^2*K

--->You can see from the resulting equation that to make Tdie lower the only thing you can change is to make k, the thermal conductivity bigger - which is why COPPER WILL ALWAYS YIELD LOWER TEMPS FOR THE STEADY STATE HEAT TRANSFER.

Now, one more thing - after you turn your computer off, there is no more heat being put into the heatsink, and it is a possibility that the aluminum might cool down faster, but I'm not convinced of that yet. But who really cares then - the computer is off. All we really care about is steady state, which has been shown that copper will always yield lower die temps that for aluminum. (in the same situation, same mounting, same design heatsinks, etc....)


Also, anyone in this thread that hasn't looked over at the "Convection" thread by Neomoses also in the Technical Discussion Forum should give it a quick glance over.

That's the kind of response I was looking for! thanks Albigger

As I have a keen interest in the subject matter having built my own water-cooled overclocked system a couple years ago, I need to correct the comments made by Albigger, hopefully with no offense. In graduate school, you learn to derive those equations from mass, moment and energy equations. With an understanding of the assumptions involved in "convection heat transfer", you can derive the generic formulas. Since I've done numerical calculations/finite element analysis on this very problem, the answer is aluminum. Why? There's a standard term in the energy equation called the thermal diffusivity, which is a function of temperature. In fact, the local heat transfer coefficient is based on a lot of things, including the thermal boundary layer and yes, the material. To set the readers straight here, the thermal diffusivity is simply the thermal conductivity divided by the product of the specific heat and density of the material. Using Albigger's numbers, the thermal diffusivity for copper and aluminum is 11.66E-5 and 9.72E-5, respectively. Since a high thermal diffusivity inhibits convection, http://scienceworld.wolfram.com/physics/ThermalDiffusivity.html, aluminum is superior. Thus, my cooling block is indeed aluminum.

I dont know if this will help or hurt anyones point, but here is a test done by www.powerelectronics.com. They used a couple different heatsinks (take into consideration that they were testing heatsinks weighing (3.19 kg for the Al base Al fin) up to (11.4 kg for the Cu base Cu fin heatsink)

http://powerelectronics.com/ar/power_tests_compare_forced/index.htm

A relevant to our application link about copper and aluminum:

http://www.amdmb.com/article-display.php?ArticleID=105&PageID=4

Read the facts myths and truths.

justdoitjerky:

Please explain how thermal diffusivity has any (negative or otherwise) effect on the rate of convective heat transfer. Perhaps there is something I am missing, and please go into detail. I know this stuff, and that assertion sounds questionable at best.

How do you explain the volumes of experimental evidence that clearly show copper heatsinks maintain a lower temperature than aluminum ones of similar design?

kct2 said:

justdoitjerky:

Please explain how thermal diffusivity has any (negative or otherwise) effect on the rate of convective heat transfer. Perhaps there is something I am missing, and please go into detail. I know this stuff, and that assertion sounds questionable at best.

How do you explain the volumes of experimental evidence that clearly show copper heatsinks maintain a lower temperature than aluminum ones of similar design?

Click to expand...

I agree, I've done some thinking about your (justdoitjerky) information and I can see where you are coming from (i think) but I highly doubt that is the actual case or that you could prove it...

But if you can I would REALLY like to see how, I'm interested in this...

Please provide some more info. for us.

I think I can see why aluminum's lower diffusivity would result in better better convenction initially, but once the system reaches equilibrium the effect would seem to be lost.

Or are you implying that the lower diffusivity will result in a greater Delta-T at equilibrium?

I've been thinking of something else when comparing Cu and Al.

What if you took advantage of the lower density of Al to build a bigger HS with a larger surface area?

Cu can conduct heat almost 100% faster then Al, but Al is less then 50% as dense.

This means that you should be able to build a HS that could be about 2 times the size of a Cu of the same weight, and use the larger area to compensate for the lower conductivity.

If you then would have a contactarea made out of Cur to act kind of like a IHS you could conduct heat away faster from the CPU and then "pass it on" to the Al using a bigger contactsurface.

Am I lost, or could there be something to this?

Ok, first of all I'd like to say that I have started writing this several times and always written something totally different
Every time I have reached new conclusions as I was writing down my own thoughts

Anyway.. I would very much like to hear your thoughts on all of this, and hope that you with a better knowlage of the laws of physics can tell me what you think!

The ability to cool has to be limited by the amount of heat that can be passed on to a heatsink or waterblock. A smaller die would make this limit lower while a bigger die would increase that number. (because of the bigger area)

If that would be true, then I should be thinking about phase change and peltiers. They allow for a LOT better cooling power.

That forced me to reach another conclusion. (still asuming that my first statement is true)
The limit should be rather high, and therefore air- and watercooling should have at least some more left to give. And especially water ofcause.
But even with air it should (at least in theory) be possible to come take cooling a bit further.

Now if this is still true, then how about this..

The maximum ability to cool a CPU in todays CPUs is not limited by the die and bottom of the HS but by the air and the fins of the HS.
Die -> HS->fins -> air

If the maximum amout of heat that can be dissipated to the air is limited by the thermal resistance of the air then it should not make a difference what material the HS is made out of.
But we know that it does matter.

So to answer that:
A metal of high thermal resistance would benifit less from higher airflow. And a metal of low thermal resistance would benifit more from higher aiflow.
Let's say that you have two identical HSs except the metal they are mad out of.
HS-Pb is made out of lead, and HS-Cu is made out of copper.

The Pb would max out at X cfm, while the Cu would max out at X+Y cfm.

What that would mean for Au and Cu is that you could benifit from a Au when you are using a fan with lower output (lower noise) and a Cu when you are using a fan with higher output (more noise).
It would also mean that to reach maximum cooling power a metal of lower thermal resistance would be better (asuming that we would go with more cfm) such as copper.

I'm done

I'd really want you to tell me what you think about it, even if you think it's all crap

You are pretty much right Finken (I only say pretty much because I am not clear of your wording in some parts). But basically from what I understand this is what you said:

If we increase the amount of air blowing of the heatsink, it will lower the (convective) thermal resistance, that is it will lower the resistance of the heat to go from the metal into the air. Now suppose we blow infinite CFM's over the heatsinks, this would be equivalent to saying that the convective resistance is zero.

Now the limiting factor is the resistance of the heatsinks themselves, and of course copper has a lower conductive resistance (for those into equations, conduction resistance R = L/kA where L is the one dimensional length the heat is traveling through, A is the constant cross sectional area available, and k is the thermal conductivity; also convective resistance is R = 1/hA where h is the convection coefficient and A is the area exposed to the convection medium).

So, as Finken said, and it can be shown with the equations again in my above post, that Aluminum will yeild a higher die temperature for a given CFM (or 'h' value) than copper will (the question is HOW MUCH more)??

Using the rough equation from my previous post, I graphed a sample, just to show the SHAPE of the curve. WARNING! pay not attention to the numbers, as they can not be quantified in this way and I just more or less stuck numbers in the equation that wouldn't be necessarily typical, but here is an idea of what increasing the CFM's (the convection coefficient) will produce:



Attached is a graph of

die temp on vertical axis and convection coefficient on horizontal axis.

equation: Tdie = Q_dot*L/(k*Ac) + Q_dot/(h*A) + Tair

You can see as h approaches infinity that there will only be two terms left determining the die temp, the air temp (or convenctive medium, be it water or air or whatever...) and the term with the equivalent resistance of the heatsink. (note this ignores any contact resistance between the die and cpu, as this is not the focus of the current discussion).

So - as convection reaches its limit with the pumps and fans provided, decreasing the resistance of the heatsink/waterblock has become another alternative to lower temps. This is a reason that sooo many thin based waterblocks (WW, others) are coming out lately. ANY extra base thickness provides SOME resistance to heat flow, and is therefore another obstacle (however some of this is offset by being able to transfer heat to a larger surface area)

Anyhow, now that I've rambled on, I still think copper will always outperform aluminum in an exact comparison, and I will continue to do so unless justdoitjerky offers more explanation about the thermal diffusivity argument...

The practical limit of waterblocks and heatsinks are reaching limit quickly, and people are always trying to buy more powerful fans and pumps, but there is a limit....

EDIT: I wanted to also say about your first post finken that it is in theory fine but not well in practice, as the space around the CPU is often limited anyway by other things, so a bigger aluminum heatsink wouldn't be possible anyhow. Besides, most copper heatsinks are more than big enough and have enough surface area now, but in situations where space isn't limited and weight is a concern, then your Cu-Al hybrid theory might work well.