Isn't aluminum better than copper for heatsinks and waterblocks?

[imported_Lucent]

I don't understand why copper is used instead of aluminum. Theoretically, the thermal conductivity of both are that there is less than 1 degree Celcius difference between using the two materials in typical surface area and thickness present in heatsinks and waterblocks. Is there some other property that makes copper superior? Maybe it is the reason that some aluminum heatsinks have copper inserts even if the size of the copper insert theoretically can't make even .5 degree Celcius difference compared to all aluminum heatsink.

 

[Henny]

Copper is a better conductor and Aluminum is a better at disapating heat. That's why you see copper bases and aluminum fins. It's also done this way to reduce weight/cost.

 

[Appledrop]

surely better heat transfer with one solid metal block (ie, 100% alu/cu) than a bit of copper here and a bit of alu here, stuck together? But then again, I know nothing about this .

 

[Zepper]

Copper can soak up more heat per gram, but doesn't gain it or lose it I as quickly as aluminum. I like the combo Al/Cu type too. A good ratio of the masses and plenty of surface area contact between the Cu and the Al makes for a good design - I also consider a radial fin design to be optimal. The Thermal Integration V77L (available from Newegg et al.) is an example of an extremely simple but very effective design for Socket A.

.bh.

 

[Howard]

Actually, aluminum has a higher specific heat capacity than copper. However, per volume, it holds less heat since it's not as dense.

I don't follow the "aluminum dissipates heat better" theory.

 

[SparkyJJO]

I had an aluminum heatsink on my XP Tbred 1700+ then switched to a solid copper heatsink which was physically smaller. Used the same fan, same thermal goop. Got several degrees cooler. nuff said

Actually I think part of it is copper is stronger than aluminum so therefore you can have thinner fins and a thinner base which lets it dissipate heat better. Aluminum is softer than copper so you can't get it as thin and retain strength and stability. I have yet to see aluminum fins as thin as some of the copper ones I've seen.

 

[Navid]

http://www.thermshield.com/ThermshieldPages/Copper_vs_Aluminum.pdf

 

[GalvanizedYankee]

SILVER I TELL YOU!! We need water blocks & heat piped sinks made out of pure silver.

...Galvanized

 

[jonnyGURU]

Originally posted by: GalvanizedYankee
SILVER I TELL YOU!! We need water blocks & heat piped sinks made out of pure silver.

...Galvanized

Check that. :thumbsup:

 

[Howard]

Yeah, I'd wear my waterblock around my neck.

 

[TrevorRC]

Originally posted by: jonnyGURU

Originally posted by: GalvanizedYankee
SILVER I TELL YOU!! We need water blocks & heat piped sinks made out of pure silver.

...Galvanized

Check that. :thumbsup:

They do have them. They're just ridiculously expensive, considering the .1C you'll gain
G5 Silver Block $200 dollars. Nice "entry" into watercooling GY.

Edit: .5C at high flow rates*

 

[Griswold]

Originally posted by: Howard
Actually, aluminum has a higher specific heat capacity than copper. However, per volume, it holds less heat since it's not as dense.

I don't follow the "aluminum dissipates heat better" theory.

QFT. Its really not that hard to understand why an aluminum block of equal volume cools down faster than a copper block.

Not like there are some magic molecules on the surface of aluminum that "mate" better with air molecules.

 

[imported_Lucent]

Found some sites:

http://www.tufts.edu/as/tampl/en43/lecture_notes/ch2.html

http://www.electronics-cooling.com/html/2001_feb_a1.html

 

[JEDIYoda]

Originally posted by: Howard
Actually, aluminum has a higher specific heat capacity than copper. However, per volume, it holds less heat since it's not as dense.

I don't follow the "aluminum dissipates heat better" theory.

I totally agree with that consensus!
Thats why the XP90copper is still right up there with the best heatsinks!

Also Scythe attempted to mass produce there Ninja in a copper model and it was way too heavy so they let that beast die!

The problem with what sounds logical and looks logical especially when dealing with cooling devices that looks and sounds and appears to be logical in real world usage is just the opposite!

Then we jump into the whole buisness of heatpipes and the disipation of heatpipes made with copper or those made with aliminum or some other metal.
Just a note....I could be wrong but aren`t most heatpipes made out of copper and the covered with aluminum or some of there metal?

This is a whole subject that is just fascinating becuase there literally are 2 distinct seperate groups with differing opinions on which isbetter,

In my case I use the XP120 on the rigg that I do not game on yet this rigg at present is OC`d from 3.2EE to 3.8.
My gaming rigg is OC`d to apprx 4.0 and I use the Zalman 9500 on that rigg.

Both rigg`s temps never are separated by more that 2c`s. Most of the time 1c.

Good subject!!

 

[acegazda]

Originally posted by: JEDIYoda
Also Scythe attempted to mass produce there Ninja in a copper model and it was way too heavy so they let that beast die!
Good subject!!

That would have been a force. The ninka is already one of the best on the market, copper would have made it the sole leader for air probably.

 

[aigomorla]

Originally posted by: acegazda

Originally posted by: JEDIYoda
Also Scythe attempted to mass produce there Ninja in a copper model and it was way too heavy so they let that beast die!
Good subject!!

That would have been a force. The ninka is already one of the best on the market, copper would have made it the sole leader for air probably.

Till the review shows 7/10 motherboards were either snapped or warped due to torque on that hugh monster.

 

[MadScientist]

Good Explanation

 

[imported_Lucent]

I made a mistake in my calculations. In my physics textbook, I thought the values were in kilo because of the k. But it is not in kilo. So my calculations were off by 1000. There is significant difference in temperature between using copper and aluminum. Using a value of 170W/mK for common aluminum alloy and 385W/mK for copper and conducting 100 watts across the thickness of a square box of length=3cm, width=3cm, and thickness=1cm, the temperature difference across the thickness of the square box is 2.886 degrees Celcius for the copper and 6.536 degress Celcius for the aluminum alloy. I used length=3cm and width=3cm because those are the dimensions of a Pentium 4 500 series IHS. So the copper insert does make a large difference theoretically.

Picture of the physics table(easy to mess up if assuming k is kilo):
http://img261.imageshack.us/img261/3796/picture47ol.gif

 

[Soubriquet]

I remember doing a science experiment where we heated rods of Al and Cu over bunsen and the Al wnet black melted and burned before the other end got hot, whereas Cu just got hotter at the other end quickly and the whole length became red hot, showing that the Al was a crap conductor and a crap radiator.

Copper is used as a heat spreader on some HS models, just on the base, and this works well because the small chip area is converted to a larger heatsink base by the copper base and then Aluminium takes over the cooling. The key is to have the copper base lapped mirror flat and good contact assisted by a thermal compound like Arctic Silver. With a cheap £10 HS you can get cooling just as good as a £40 solid copper model, from experience.

Really the best HS design should have a copper base plate contacting the chip connected directly to copper fin elements to take the heat along the fins, but use aluminium for mechanical support in both the fins and base but that would be a lot harder to manufacture.

 

[Howard]

The aluminum turned black and burned? Over a bunsen burner? Interesting.

 

[TrevorRC]

Howard--
Aluminum Oxide, perhaps?

--Trevor

 

[Howard]

Aluminum oxide is definitely not black.

EDIT: Maybe the black was carbon buildup, although from what I remember the bunsen burners I used in high school burned pretty cleanly. It would take a hell of a lot of "flash" heat for a non-powdered piece of aluminum to burn.

 

[wwswimming]

Originally posted by: Henny
Copper is a better conductor and Aluminum is a better at disapating heat. That's why you see copper bases and aluminum fins. It's also done this way to reduce weight/cost.

are you saying air convection off of aluminum is more effective than air convection off of copper ?

 

[wwswimming]

copper is 9.65 and alum is 5.3 - that's the ratio for thermal conductivity by volume.

most of the time with a thermal circuit, you look for the weak links in the chain. if you're having
a 10 degree C rise from the fins to the ambient air, and 3 degree C gradient in the heat sink
with alum., switching to copper, very roughly without a finite element model, reduces the
gradient by 1.5 degrees. plus you have a "delta" or a gradient from the CPU to the bottom
of the heat sink.

so, yes, aluminum is only about half as effective as copper by volume in terms of (volume)
thermal conductivity. but, a lot of times, you will barely notice the difference.

working in industry ...
* for avionics, airborne electronics, we usually used aluminum heat sinks for largely conductive
heat transfer, and copper vias on circuit boards. in some case, LOTS of copper vias.
* for ground based equipment - usually aluminum heat sinks.

if you take apart a PC power supply, that's what you'll see on the major heat producers -
aluminum heat sinks.

copper is getting WAY expensive. $3 a pound.

 

[herm0016]

a bunson burner will easy melt aluminum. solid will not burn though. and alum. oxide is what makes your alum. not corrode like it will without the coating. try putting some murcury on aluminum, it will prevent the oxide layer from froming and the alum will be a pile of powder in a day or so depending on the size of the peice. copper is much better at conducting and being corision resistant.