Aluminum vs Copper cooling

[freaksavior]

So I have noticed over the past year, that there are many questions regarding what's a better material for cpu cooling.

Aluminum has a specific heat of 0.9000 J over G x C

Copper has a specific heat of 0.3851 J over G x C

So copper will heat up more quickly than aluminum. I would go into the process of the equation but i don't quite know how to write it out. so hopefully that helped someone. if someone knows the equation for the specific heat and all that stuff, please share (btw it's chemistry)

 

[freaksavior]

sorry for double post, but does anybody know a cpu cooler that is made in aluminum and then exactly the same in copper?

 

[pt]

sorry for double post, but does anybody know a cpu cooler that is made in aluminum and then exactly the same in copper?

the thermalright xp-90 has 2 variants

 

[DaMulta]

a radiator will work better as Aluminum, because Aluminum will disperse heat better. Copper will absorb the heat better than Aluminum, but is not as good as dispersing it.

 

[gR3iF]

The point for this is air:
Air is the medium that will get the heat after getting it away from the cpu.


So to use a metall mix: Use a cooper plate on the cpu and then alu because alu or cooper give the same amount of heat to the air. It just cant take anymore.

To bring in something new: Heatpipes they can transport more heat than cooper. That is the reason for there usage nowadays and something like this:

All in all a mixup of heatpipes and cooper is the effective way for aircooling.
The only thing to use alu is the price and the thing that air absorbs the heat and this isnt affected by alu or cooper usage.

 

[freaksavior]

hmm, so basically what your saying is, it is best to have a copper plate/base, then aluminum fins.

 

[DaMulta]

hmm, so basically what your saying is, it is best to have a copper plate/base, then aluminum fins.

Yes

 

[gR3iF]

Yep that is right

 

[freaksavior]

this is interesting to me.....i didn't know this lol

 

[gR3iF]

Then use the thank button ;D

Its fact so far that Heatpipes are best for transporting heat. The negative point about them ist that they are hardly to work into a base plate or something normally u get then a profile like a rock so the contact area looks really ugly and then the advantage of using heatpipes is gone.

So normally Manufactures uses a flat cooper base plate and then heatpipes soldered to it and then fins out of alu or cooper (this makes the price)

Look at the Zalmann 9500-9700 series they are really good cooler and they work with this simple thing.

 

[largon]

a radiator will work better as Aluminum, because Aluminum will disperse heat better. Copper will absorb the heat better than Aluminum, but is not as good as dispersing it.

Aluminium is inferior for it's thermal specs compared to copper. Alu does not "disperse" heat better than copper. Infact, no such specific physical feature in different substances exists which dictates how well the material gives away heat. Matter can release heat by three methods: conduction, convection and radiation. Convection applies only to liquids and gases.

The only things that matter for conduction and radiability are:
- thermal resistance
- surface area
- color

Aluminium is popular only because it's cheaper than copper.

Real world example:
Thermalright XP-90 vs. XP-90c @ SystemCooling.com
->

XP-90 has a copper base and copper heatpipes with alu fins whereas XP-90c is all copper.
Still, XP-90c wins hands down.

 

[jtleon]

How about a bit of Engineering Data.....

You can see the thermal conductivity (the rate at which heat flows within a material) of copper and other materials here:

http://hypertextbook.com/physics/thermal/conduction/

Aluminum has conductivity value of 237 W/m.K (Watts/Meter x degrees Kelvin)

Copper has conductivity value of 401 W/m.K.

Air has conductivity of 0.025 W/m.K

The biggest engineering challenge is to force the energy to jump from the Aluminum or Copper to the Air. As you can see, air has a very low energy conduction rate. Gigantic amounts of exposed surface area to air are required to facilitate efficient transfer of energy from the metal to the air. Air flow velocity also plays a critical role - more flow=more transfer.

If the HS geometry is the same, then the Copper design should perform almost twice as well as the Aluminum design.

Note that:
Diamond has conductivity of 895 W/m.K

Carbon Graphite is the very best at 1950 W/m.K

Soon we will see HSF's with diamond or graphite components. By the way, carbon graphite is already used in most brushes for electric motors, and pencil lead. OCZ did have a carbon graphite demo heatsink at one time in the recent past.

Regards,
jtleon

 

[theonetruewill]

Aluminium is inferior for it's thermal specs compared to copper. Alu does not "disperse" heat better than copper. Infact, no such specific physical feature in different substances exists which dictates how well the material gives away heat. Matter can release heat by three methods: conduction, convection and radiation. Convection applies only to liquids and gases.

The only things that matter for conduction and radiability are:
- thermal resistance
- surface area
- color

Aluminium is popular only because it's cheaper than copper.

Real world example:
Thermalright XP-90 vs. XP-90c @ SystemCooling.com
->

XP-90 has a copper base and copper heatpipes with alu fins whereas XP-90c is all copper.
Still, XP-90c wins hands down.

I agree, this info is correct. jtleon is also correct.

 

[Oliver_FF]

Aluminium is popular only because it's cheaper than copper.

It's also popular because copper is way heavier than aluminium and doesn't corrode as easily. Sweaty hands on a pure copper heatsink will leave you with dark finger prints all over it, axeing it's usefulness...

 

[a111087]

the thermalright xp-90 has 2 variants

just wanted to tell you guys that i saw a review of both of these and copper of course won, but by ~3C or something like that

 

[DaMulta]

http://www.chevellecooling.com/43.html

 

[POGE]

Copper has all around better thermal properties, aluminum is just easier to work with sometimes, and is also a lot cheaper.

 

[t_ski]

I read a recommendation a long time ago, so I can't remember the source offhand, that using different metals were better in different circumstances. Specifically, using aluminum was better when the unit was to be cooled passively, but copper was better when cooled actively. IIRC the difference was based on how the metal absorbed and dispersed the heat, and how the difference was when the airflow was changed around them.

By the way, there are other metals out there that are more conductive than copper, but obviously cost becomes a limiting factor. Silver is one of those metals, and has been used to make a couple Danger Den blocks (among others): the TDX and the RBX.

 

[panchoman]

You can see the thermal conductivity (the rate at which heat flows within a material) of copper and other materials here:

http://hypertextbook.com/physics/thermal/conduction/

Aluminum has conductivity value of 237 W/m.K (Watts/Meter x degrees Kelvin)

Copper has conductivity value of 401 W/m.K.

Air has conductivity of 0.025 W/m.K

The biggest engineering challenge is to force the energy to jump from the Aluminum or Copper to the Air. As you can see, air has a very low energy conduction rate. Gigantic amounts of exposed surface area to air are required to facilitate efficient transfer of energy from the metal to the air. Air flow velocity also plays a critical role - more flow=more transfer.

If the HS geometry is the same, then the Copper design should perform almost twice as well as the Aluminum design.

Note that:
Diamond has conductivity of 895 W/m.K

Carbon Graphite is the very best at 1950 W/m.K

Soon we will see HSF's with diamond or graphite components. By the way, carbon graphite is already used in most brushes for electric motors, and pencil lead. OCZ did have a carbon graphite demo heatsink at one time in the recent past.

Regards,
jtleon

why didn't they make carbon graphite from the start? seems mad cheap considering its pencil lead.... and i agree with him. the most conductivity the material has, the faster it can move the heat away fromthe cpu.

 

[jtleon]

Some Clarification

If the HS geometry is the same, then the Copper design should perform almost twice as well as the Aluminum design.

My statement above does not mean that the Copper model will be twice as cold as the Aluminum model - it will only dissipate about twice the energy, which in many cases results in only a few degrees cooler operation - considering all three transfer phenomena (conduction, convection, and radiation occurring simultaneously)

In other words, if both the copper and aluminum heatsinks were the same temperature, the energy input would be almost twice as great in the copper model.
Regards,
jtleon

 

[jtleon]

Carbon Graphite manufacturing is big money....

why didn't they make carbon graphite from the start? seems mad cheap considering its pencil lead.... and i agree with him. the most conductivity the material has, the faster it can move the heat away fromthe cpu.

Note that carbon graphite in pencil lead is only a small percentage carbon graphite, the remainder is a polymer that binds the particles together. In this form, the conductivity is very poor, given the low particle population.

After reviewing this question, it seems the greatest obstacle to carbon graphite heatsinks on a large scale is the manufacturing difficulty (read high costs) to create similar heatsinks to those that today are of aluminum or copper. Also recognize that the equivalent carbon graphite heat sink would be very fragile, and could not withstand even those small forces caused by the common clamping mechanisms.

Regards,
jtleon

 

[panchoman]

Note that carbon graphite in pencil lead is only a small percentage carbon graphite, the remainder is a polymer that binds the particles together. In this form, the conductivity is very poor, given the low particle population.

After reviewing this question, it seems the greatest obstacle to carbon graphite heatsinks on a large scale is the manufacturing difficulty (read high costs) to create similar heatsinks to those that today are of aluminum or copper. Also recognize that the equivalent carbon graphite heat sink would be very fragile, and could not withstand even those small forces caused by the common clamping mechanisms.

Regards,
jtleon

well if we add diamond supports to the carbon graphite, the we get a 100k heatsink

 

[KennyT772]

More importantly than the metal of the heatsink is the design. That cannot be forgotten. Properly placed heat pipes are the hardest to find in todays heatsinks.

heat pipes should not be arranged in a inline fashion, or near the outer edge of the heatsink. Many of todays heatsinks group the heat pipes so that the fin area becomes saturated with heat, while other areas never even get warm. The more even the distribution of heat pipes the better the performance end of story.

I believe the best design would be a dual tower style like the Thermalright fx14 with 6 heat pipes. If the heat pipes were pressed square then laid side by side to form a base you would have direct contact with the IHS and the best thermal transfer. You then put a block on top of them (for mounting and rigidity). Each of the 6 heat pipes then should be staggered in the tower for the best distribution. Add a fan with a shroud on each side (to normalize pressure and turbulence) and you have the near perfect heatsink. Maybe I should design this in CAD and send it into Thermalright... This could be easily done with Thermalright's designs and would would great. Anyone think I should send it in?

 

[panchoman]

More importantly than the metal of the heatsink is the design. That cannot be forgotten. Properly placed heat pipes are the hardest to find in todays heatsinks.

heat pipes should not be arranged in a inline fashion, or near the outer edge of the heatsink. Many of todays heatsinks group the heat pipes so that the fin area becomes saturated with heat, while other areas never even get warm. The more even the distribution of heat pipes the better the performance end of story.

I believe the best design would be a dual tower style like the Thermalright fx14 with 6 heat pipes. If the heat pipes were pressed square then laid side by side to form a base you would have direct contact with the IHS and the best thermal transfer. You then put a block on top of them (for mounting and rigidity). Each of the 6 heat pipes then should be staggered in the tower for the best distribution. Add a fan with a shroud on each side (to normalize pressure and turbulence) and you have the near perfect heatsink. Maybe I should design this in CAD and send it into Thermalright... This could be easily done with Thermalright's designs and would would great. Anyone think I should send it in?

sure, sounds great. its never hurts to try (well most of the time)

 

[Chewy]

woot go Kenny GO!