Aluminum vs Copper

[Nanook]

For the longest time copper has been king; however, now there is research stating that while aluminum can't transfer heat as well as copper, but aluminum does dissipate it into the air better. 

 

Basically, Copper is better at heat transfer and aluminum is better at heat dissipation.

 

So what cools better now? Is copper still king, or were we wrong the entire time? Maybe a new copper/aluminum allow would be better than single elements?

 

 

https://www.researchgate.net/post/For_better_heat_transfer_which_material_should_prefer_first_aluminium_or_copper#:~:text=It is clear that thermal,heat transfer application over copper.

 

[jaslion]

Almost all coolers nowadays are a copper heatpipe aluminum fin array hybrid. Taking advantage of the metals strongpoints whilst avoid their weaknesses.

[Nanook]

5 minutes ago, jaslion said:

Almost all coolers nowadays are a copper heatpipe aluminum fin array hybrid. Taking advantage of the metals strongpoints whilst avoid their weaknesses.

I didn't think of that. Good observation.

[For Science!]

No idea where this idea that dissipation and conductivity are two different things (apart from that one dude on ResearchGate, which means nothing imo). The reason why aluminium is used is from a weight/cost perspective and nothing else.

[Nanook]

41 minutes ago, For Science! said:

No idea where this idea that dissipation and conductivity are two different things (apart from that one dude on ResearchGate, which means nothing imo). The reason why aluminium is used is from a weight/cost perspective and nothing else.

Yes, I will gladly take the hit for posting research gate lol. I was too lazy to look up peer review at the moment. However, Conductivity and dissipation are not really the same thing and I can give an example I suppose. At the core it has to do with density and surface area.

 

Imagine a bird feather that you got pretty hot, but without catching on fire. After words you introduced it into the surrounding ambient air maybe with a slight breeze. At the same time you had a small iron cube of the same amount of mass as the feather. The iron cube was also heated to the same temperature as the feather. Which one do you think is is going to dissipate its heat to the surrounding air faster? Clearly the feather will dissipate its heat faster. It is less dense therefore has a much larger surface area in contact with the ambient air In contrast, due to its density the small iron cube has much less surface area in contact with the air.

 

However, this does not conflict with conductivity. Clearly because the iron cube is denser it has the ability to transfer heat around its own mass much better. This makes it a better conductor of heat. Again due to its density the iron cube will spread the heat around its own mass much quicker because adjacent atoms are much closer. Also the conduction of electrons play a role in quicker heat distribution within its own mass. The feather however, cannot spread heat around its own mass quickly. At whatever point heat is added to the feather, it localizes at that point and spreads much slower because it lacks density of the iron and electron conductivity. As a result the feather must take all the heat added at a small point, which results in failure or burning of the feather.

 

The ideal cooling unit for feathers and iron would be a piece of iron to soak up the heat better, and a stack of many feathers to take the heat from the iron and dissipate it into the air. This is starting to sound similar to the standard aluminum and copper air heatsink. Copper heat pipes with a stack of aluminum fins.

 

So the larger surface area of the feather dissipates heat much better. While the iron cube is a much better conductor and pool for heat absorption. This is an extreme example of opposites, but it is still analogous to the copper and aluminum. 

 

So yes, dissipation and conductivity are different things.

[Guest]

1 hour ago, Nanook said:

Yes, I will gladly take the hit for posting research gate lol. I was too lazy to look up peer review at the moment. However, Conductivity and dissipation are not really the same thing and I can give an example I suppose. At the core it has to do with density and surface area.

 

Imagine a bird feather that you got pretty hot, but without catching on fire. After words you introduced it into the surrounding ambient air maybe with a slight breeze. At the same time you had a small iron cube of the same amount of mass as the feather. The iron cube was also heated to the same temperature as the feather. Which one do you think is is going to dissipate its heat to the surrounding air faster? Clearly the feather will dissipate its heat faster. It is less dense therefore has a much larger surface area in contact with the ambient air In contrast, due to its density the small iron cube has much less surface area in contact with the air.

 

However, this does not conflict with conductivity. Clearly because the iron cube is denser it has the ability to transfer heat around its own mass much better. This makes it a better conductor of heat. Again due to its density the iron cube will spread the heat around its own mass much quicker because adjacent atoms are much closer. Also the conduction of electrons play a role in quicker heat distribution within its own mass. The feather however, cannot spread heat around its own mass quickly. At whatever point heat is added to the feather, it localizes at that point and spreads much slower because it lacks density of the iron and electron conductivity. As a result the feather must take all the heat added at a small point, which results in failure or burning of the feather.

 

The ideal cooling unit for feathers and iron would be a piece of iron to soak up the heat better, and a stack of many feathers to take the heat from the iron and dissipate it into the air. This is starting to sound similar to the standard aluminum and copper air heatsink. Copper heat pipes with a stack of aluminum fins.

 

So the larger surface area of the feather dissipates heat much better. While the iron cube is a much better conductor and pool for heat absorption. This is an extreme example of opposites, but it is still analogous to the copper and aluminum. 

 

So yes, dissipation and conductivity are different things.

I know you are trying to say something here....

 

But Iron blocks and feathers is laughable to use when describing a transfer of BTU which is what you convert your wattage to.

 

Diamond

Silver

Gold

Copper

Some other stuff

Then

Aluminum.

Way way way down the list

Feathers and Iron

Edit

Gold is better than silver

 

Is a rough off the top of my head thermal transfer order starting from best to worst.

 

Aluminum has no benefits over copper other than cost as mentioned above. 

 

Silver obviously is too expensive, however I have a cold plate I lapped out of a Morgan Silver Dollar.

 

Gold I haven't tried. But from Copper to silver it's single digit difference in PC cooling applications. 

 

Dissipation and conduction is the same on the same piece of material. When you break the material and solder more to it, the solder joints slow dissipation rates.

[Masterjuggler]

1 hour ago, Nanook said:

So yes, dissipation and conductivity are different things.

You're comparing not materials as in the original question, but geometries. You are right that dissipation efficiency depends on more than just the material, but if you keep all other things equal like geometry, the copper will transfer heat much more effectively than aluminum.

 

Dissipation is just the transfer of heat from one mass to another. Water and air are both fluids, and the heatsink that is able to transfer heat to it faster is able to dissipate it faster.

 

I have a physics degree btw, so I have a little bit of qualification to be saying this.

[For Science!]

3 hours ago, Nanook said:

Yes, I will gladly take the hit for posting research gate lol. I was too lazy to look up peer review at the moment. However, Conductivity and dissipation are not really the same thing and I can give an example I suppose. At the core it has to do with density and surface area.

 

Imagine a bird feather that you got pretty hot, but without catching on fire. After words you introduced it into the surrounding ambient air maybe with a slight breeze. At the same time you had a small iron cube of the same amount of mass as the feather. The iron cube was also heated to the same temperature as the feather. Which one do you think is is going to dissipate its heat to the surrounding air faster? Clearly the feather will dissipate its heat faster. It is less dense therefore has a much larger surface area in contact with the ambient air In contrast, due to its density the small iron cube has much less surface area in contact with the air.

 

However, this does not conflict with conductivity. Clearly because the iron cube is denser it has the ability to transfer heat around its own mass much better. This makes it a better conductor of heat. Again due to its density the iron cube will spread the heat around its own mass much quicker because adjacent atoms are much closer. Also the conduction of electrons play a role in quicker heat distribution within its own mass. The feather however, cannot spread heat around its own mass quickly. At whatever point heat is added to the feather, it localizes at that point and spreads much slower because it lacks density of the iron and electron conductivity. As a result the feather must take all the heat added at a small point, which results in failure or burning of the feather.

 

The ideal cooling unit for feathers and iron would be a piece of iron to soak up the heat better, and a stack of many feathers to take the heat from the iron and dissipate it into the air. This is starting to sound similar to the standard aluminum and copper air heatsink. Copper heat pipes with a stack of aluminum fins.

 

So the larger surface area of the feather dissipates heat much better. While the iron cube is a much better conductor and pool for heat absorption. This is an extreme example of opposites, but it is still analogous to the copper and aluminum. 

 

So yes, dissipation and conductivity are different things.

No, this is flawed. You are bringing all matter of things including surface area, total mass, etc. Feathers and a cube of iron cannot be compared in any capacity for this discussion. This is on par with the "a 1 kg ball of feathers vs a 1kg ball of steel, which will fall first" level of arugement.3

[Nanook]

On 7/5/2021 at 3:16 PM, For Science! said:

No, this is flawed. You are bringing all matter of things including surface area, total mass, etc. Feathers and a cube of iron cannot be compared in any capacity for this discussion. This is on par with the "a 1 kg ball of feathers vs a 1kg ball of steel, which will fall first" level of arugement.3

No it's not flawed. It's basic larger surface area cools quicker because it has more contact with the air. Also, a 1kg ball of feather and 1kg ball of steel will fall at  the same speed in a vacuum because gravity will act the same on both.

[For Science!]

Just now, Nanook said:

No it's not flawed. It's basic larger surface area cools quicker because it has more contact with the air. Also, a 1kg ball of feather and 1kg ball of steel will fall at  the same speed in a vacuum because gravity will act the same on both.

so.....1 cm cubed of aluminium vs 1 cm cubed of copper, which will dissipate heat faster?

[Nanook]

On 7/5/2021 at 1:50 PM, Masterjuggler said:

You're comparing not materials as in the original question, but geometries. You are right that dissipation efficiency depends on more than just the material, but if you keep all other things equal like geometry, the copper will transfer heat much more effectively than aluminum.

 

Dissipation is just the transfer of heat from one mass to another. Water and air are both fluids, and the heatsink that is able to transfer heat to it faster is able to dissipate it faster.

 

I have a physics degree btw, so I have a little bit of qualification to be saying this.

Nah, it wont. The aluminum will dissipate heat better than the copper, but the aluminum will transfer more heat and more efficiently to itself. I also have a physics degree.

[Nanook]

3 minutes ago, For Science! said:

so.....1 cm cubed of aluminium vs 1 cm cubed of copper, which will dissipate heat faster?

The aluminum because it will have a larger surface area than the copper. Just look at the design of modern heat sinks for proof. Almost all of them are copper bases with heat pipes soldered to thin aluminum blades. Noctoa, cooler master, be quite, etc. all have the same copper base and pipes to aluminum blades.

[For Science!]

Just now, Nanook said:

The aluminum because it will have a larger surface area than the copper.

......What are you on about. I just said the are both 1 cm cubed, i,e, they have the same surface area.....

[Nanook]

3 minutes ago, For Science! said:

......What are you on about. I just said the are both 1 cm cubed, i,e, they have the same surface area.....

I seen 1kg for some reason. If they are both 1cm^3 each, then the copper will dissipate heat better. But this argument ignores density and surface area. You can get more heat dissipation from aluminum if both masses of both elements are equal.

[Nanook]

On 7/5/2021 at 1:34 PM, ShrimpBrime said:

I know you are trying to say something here....

 

But Iron blocks and feathers is laughable to use when describing a transfer of BTU which is what you convert your wattage to.

 

Diamond

Silver

Gold

Copper

Some other stuff

Then

Aluminum.

Way way way down the list

Feathers and Iron

Edit

Gold is better than silver

 

Is a rough off the top of my head thermal transfer order starting from best to worst.

 

Aluminum has no benefits over copper other than cost as mentioned above. 

 

Silver obviously is too expensive, however I have a cold plate I lapped out of a Morgan Silver Dollar.

 

Gold I haven't tried. But from Copper to silver it's single digit difference in PC cooling applications. 

 

Dissipation and conduction is the same on the same piece of material. When you break the material and solder more to it, the solder joints slow dissipation rates.

This is not correct. Barring cost of materials, if we have two blocks of aluminum and copper that are the same mass, then the aluminum will dissipate heat into the air faster because it will have larger surface area in contact with the air. 

[For Science!]

Just now, Nanook said:

I seen 1kg for some reason. If they are both 1cm^3 each, then the copper will dissipate heat better. But this argument ignores density and surface area. You can get more heat dissipation from aluminum if both masses are equal.

Sure, but that isn't the realistic comparison. Lets take a 120 mm CPU cooler, if you were to make one out of copper and one out of aluminium, you make them the same size not the same weight, and yes, the copper 120 mm CPU cooler is going to be 3x heavier than the aluminium CPU cooler, and thats the primary disadvantage as I said above already (weight and cost).

 

Similarly for a 120 mm Radiator, you aren't going to stop populating the radiator with copper fins just because you reached the same mass as with aluminium fins. So copper heatsinks and fins do dissipate better than aluminium ones per form.

[Nanook]

Quote

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.

https://www.overclockers.com/forums/showthread.php/187162-Does-aluminum-dissipate-heat-better-than-copper#:~:text=Aluminum has the ability to,as it is with electricity.

 

Aluminum has the ability to sink heat faster and dissipate heat faster. Copper is a better "pool" to sink heat into because it can absorb more heat.

 

The winning ticket is a combination of aluminum and copper. Copper block and heat pipes soldered to an aluminum fin stack.

[For Science!]

1 minute ago, Nanook said:

https://www.overclockers.com/forums/showthread.php/187162-Does-aluminum-dissipate-heat-better-than-copper#:~:text=Aluminum has the ability to,as it is with electricity.

This is even less good than the research gate source, and even that person says that copper is better.

[Nanook]

3 minutes ago, For Science! said:

This is even less good than the research gate source, and even that person says that copper is better.

I understand, but I am having difficulty find a peer review on this particular topic. But there seems to be mass consensus online that aluminum dissipates heat better.

[For Science!]

Just now, Nanook said:

I understand, but I am having difficulty find a peer review on this particular topic.

But do you understand that comparing copper and aluminium by weight is not appropriate? 

Lets say it takes 30 grams of aluminium to cover an entire 120 mm radiator core with fins.

If you then said that you are allowing only 30 grams of copper to fill the 120 mm radiator core, 2/3rds of that radiator will be fin-less.

 

So you can't compare the two materials by weight, as then the copper product will always only be a third complete. 

[Nanook]

Just now, For Science! said:

But do you understand that comparing copper and aluminium by weight is not appropriate? 

Lets say it takes 30 grams of aluminium to cover an entire 120 mm radiator core with fins.

If you then said that you are allowing only 30 grams of copper to fill the 120 mm radiator core, 2/3rds of that radiator will be fin-less.

 

So you can't compare the two materials by weight, as then the copper product will always only be a third complete. 

Mass seems like the very realistic way to compare these things. It is a fundamental unit.

[For Science!]

Just now, Nanook said:

Mass seems like the very realistic way to compare these things. It is a fundamental unit.

My 3070 weighs the same as your 3080Ti, so no issue if we swap them right?

 

Would you then say a fully copper 120 mm radiator should be compared against a fully aluminium 360 mm radiator because they weigh the same? Maybe if PCs were weighed before their benchmark like wrestlers then weight may come into play as a metric. Or perhaps if you needed to fit onto carry-on airplane luggage by weight. But no, that's not how we build PCs.

[Nanook]

2 minutes ago, For Science! said:

But do you understand that comparing copper and aluminium by weight is not appropriate? 

Lets say it takes 30 grams of aluminium to cover an entire 120 mm radiator core with fins.

If you then said that you are allowing only 30 grams of copper to fill the 120 mm radiator core, 2/3rds of that radiator will be fin-less.

 

So you can't compare the two materials by weight, as then the copper product will always only be a third complete. 

But why deny that so many top heat sinks use a combination of copper and aluminum? I am sure people would pay a bit extra to have a purely copper fin stack. In fact, if you look at older heat sink designs, they were all copper. Then a transition happened along the timeline somewhere switching to an aluminum fin stack. The fact that high end manufacturers use this copper and aluminum combination give evidence this combination is the best.

[For Science!]

Just now, Nanook said:

But why deny that so many top heat sinks use a combination of copper and aluminum? I am sure people would pay a bit extra to have a purely copper fin stack. In fact, if you look at older heat sink designs, they were all copper. Then a transition happened along the timeline somewhere switching to an aluminum fin stack. The fact that high end manufacturers use this copper and aluminum combination give evidence this combination is the best.

As I said, weight and cost. Fully copper heatsinks were so heavy they would bend boards and themselves in their own weight. Ultimately the rate limiting transfer step is from the metal-to-air and since air is such a poor conductor of heat, the cost/weight difference was simply not worth it. Still doesn't make aluminium better, from a theoretical point of view as a heat dissipating material.

[Nanook]

8 minutes ago, For Science! said:

My 3070 weighs the same as your 3080Ti, so no issue if we swap them right?

 

Would you then say a fully copper 120 mm radiator should be compared against a fully aluminium 360 mm radiator because they weigh the same? Maybe if PCs were weighed before their benchmark like wrestlers then weight may come into play as a metric. Or perhaps if you needed to fit onto carry-on airplane luggage by weight. But no, that's not how we build PCs.

Why would you not compare an aluminum 360 to a copper 120 if the masses are the same? You are attempting to deny aluminum its natural advantage in density and surface aarea.