Silent power, a designed from scratch computer cooled with cooper foam

[senso]

So what are your toughs about the newest crazy crowd funding campaign?
http://hackaday.com/2014/08/03/behold-the-most-insane-crowdfunding-campaign-ever/#comments

[mikeselectricstuff]

Copper foam smells like bullshit to me.
High surface area, yes, but poor conductivity to all but the bottom layer, and probably poor natural airflow 

And it will look shit once it starts to oxidise

[AndyC_772]

Copper foam? Really? Take a material with good thermal conductivity, but then replace 90% of the volume with trapped, still air in order to turn it into a good insulator...?

[rob77]

that's bullshit - any foam is an excellent heat insulator because of the trapped air.

if i would design a silent heat-sink, then i would rather investigate a possibility of using a honeycomb structure with vertical hexagonal pipes with open ends to allow convection take the heat away (in fact a shitload of small diameter hexagonal copper chimneys packed together )

[Marco]

Copper isn't really that good a thermal conductor to begin with ... if you need transport more than a few cm you use heatpipes.

[rob77]

Copper isn't really that good a thermal conductor to begin with ... if you need transport more than a few cm you use heatpipes.

yes, for sure - heat-pipes made of copper

copper is a good heat conductor, but it's not as good as black anodized aluminium when talking about heat radiation/dissipation.

[rs20]

that's bullshit - any foam is an excellent heat insulator because of the trapped air.

Only closed-cell foams trap air -- this is very much an open-cell foam, a coarse one at that, and it looks like the airflow (and therefore convection and potential heat transport) through that particular foam would be pretty free-flowing.

The remaining question is whether the thin strands of copper leading from the bottom to the top will conduct well enough. Mental exercise -- imagine placing the heat sink in a super table vice and crushing it into solid copper. What would you be left with, a block of copper maybe a centimeter thick?  The heat conductivity through that solid block is clearly going to be very good (A 2 inch x 1 cm cross-section block with 1 inch height works out to 0.13 K/W). But here's the kicker -- it seems like the average copper path length from bottom to top is maybe 50%-100% longer in the foam than the direct line possible in a solid block, so the thermal conductivity is only going to be 33%-50% worse. Let's say 0.26 K/W. Except it's radiating the whole way. Hardly a massive penalty for the foam factor. 100W works out to 26 degrees C over ambient in the copper alone (so yes, neglecting the actual radiation/conduction into air), which is pretty lax, but vaguely workable.

Will this foam radiate heat as well as an traditional heatsink with the same volume of copper? Probably not. But might it be an acceptable compromise of interestingness/aesthetics vs performance? Pretty shaky, probably not. But it's certainly not as intrinsically absurd or off by orders of magnitude as many here are making it out to be. If it helps, think numerically about how electrically conductive the foam would be (to clear your mind of the "omg it's foam and there's air everywhere run away" preconception), and then refer those results back to the thermal world.

Edit: Just to be clear that I basically agree that this isn't a good idea (with reservations), I offer another disadvantage: this foam would be a massive pain to clean dust from...

[wraper]

Copper isn't really that good a thermal conductor to begin with ... if you need transport more than a few cm you use heatpipes.

yes, for sure - heat-pipes made of copper

copper is a good heat conductor, but it's not as good as black anodized aluminium when talking about heat radiation/dissipation.

What funny about copper? Heatpipes are filled with water which circulates inside. Therefore heat is transferred through copper to very short distance. Black surface do not give any advantage when cooling is active. Strange how you compare heat conductivity vs radiation. Apples vs oranges.

[rob77]

Copper isn't really that good a thermal conductor to begin with ... if you need transport more than a few cm you use heatpipes.

yes, for sure - heat-pipes made of copper

copper is a good heat conductor, but it's not as good as black anodized aluminium when talking about heat radiation/dissipation.

What funny about copper? Heatpipes are filled with water which circulates inside. Therefore heat is transferred through copper to very short distance. Black surface do not give any advantage when cooling is active. Strange how you compare heat conductivity vs radiation. Apples vs oranges.

i'm not comparing conductivity vs radiaton - i was describing the copper - good heat conduction but not so good at radiation (because of it's color - and unfortunately copper can't be anodized to make it black)
black surface will give you a relatively big advantage when talking about radiation without the aid of forced air circulation.

so in fact i could say that you're comparing apples with oranges 

Black surface do not give any advantage when cooling is active.

you're comparing radiation with active cooling (forced airflow)

[wraper]

The heat conductivity through that solid block is clearly going to be very good (A 2 inch x 1 cm cross-section block with 1 inch height works out to 0.13 K/W). But here's the kicker -- it seems like the average copper path length from bottom to top is maybe 50%-100% longer in the foam than the direct line possible in a solid block, so the thermal conductivity is only going to be 33%-50% worse. Let's say 0.26 K/W. Except it's radiating the whole way.

Funny maths. Path length maybe, but you forgot that the cross section of the copper became close to zero. Therefore calculate thermal conductivity for remaining copper cross section, but anything left for air  .

[wraper]

you're comparing radiation with active cooling (forced airflow)

Heatpipes = active cooling in 99% of cases. PC won't have passive cooling unless it is very low power.

[rs20]

The heat conductivity through that solid block is clearly going to be very good (A 2 inch x 1 cm cross-section block with 1 inch height works out to 0.13 K/W). But here's the kicker -- it seems like the average copper path length from bottom to top is maybe 50%-100% longer in the foam than the direct line possible in a solid block, so the thermal conductivity is only going to be 33%-50% worse. Let's say 0.26 K/W. Except it's radiating the whole way.

Funny maths. Path length maybe, but you forgot that the cross section of the copper became close to zero. Therefore calculate thermal conductivity for remaining copper cross section, but anything left for air  .

I didn't forget; the crushing the block in a vice part was accounting for that. Again, think about the electrical analogy. All those paths are in thermal parallel, so all that matters is the total cross section of the entire piece. If there's a decent density of copper in there, such that crushing it would bring it to 1 cm thick solid block, then that hypothetical solid block has the same thermal conductivity as the original foam, modulo the factor of 1.5 or 2 from the increased path length.

What funny about copper? Heatpipes are filled with water which circulates inside. Therefore heat is transferred through copper to very short distance. Black surface do not give any advantage when cooling is active. Strange how you compare heat conductivity vs radiation. Apples vs oranges.

Well put, worth just putting it in a table for clarity:

                      Thermal Conductivity        Emissivity (radiation)
Aluminium              Good (230 W/m/K)            Fair
Anodized Aluminium     Good (230 W/m/K)            Excellent
Copper                 Excellent (380 W/m/K)       Good
Heat pipe              Hyper-excellent*            N/A


* Heat pipes don't really get much worse as they get longer (at least not as a simple reciprocal relationship), so a normal bulk thermal conductivity doesn't really make sense.

One thing that's not clear to me -- radiation and conduction to air/convection are both relevant to heatsinks, I wonder what the breakdown is for the typical heatsink? Radiation must be some part of the story if manufacturers bother to anodize the heatsinks black, but if it was all about radiation, why does the fan need to move the air around? Is the heat radiated at wavelengths easily absorbed by air? I think thinking this through would be relevant to this discussion as well.

[rob77]

you're comparing radiation with active cooling (forced airflow)

Heatpipes = active cooling in 99% of cases. PC won't have passive cooling unless it is very low power.

are you even reading the posts you're replying to ?

[wraper]

you're comparing radiation with active cooling (forced airflow)

Heatpipes = active cooling in 99% of cases. PC won't have passive cooling unless it is very low power.

are you even reading the posts you're replying to ?

Reread 5 times your first post I quoted, It's my English got that wrong  . So yeah, you did not compare conductivity/dissipation, talked as about separate things, sorry.

Again, think about the electrical analogy. All those paths are in thermal parallel, so all that matters is the total cross section of the entire piece. If there's a decent density of copper in there, such that crushing it would bring it to 1 cm thick solid block, then that hypothetical solid block has the same thermal conductivity as the original foam, modulo the factor of 1.5 or 2 from the increased path length.

Exactly, the same as electrical conductivity. Now remove 95% of the copper volume from the wire and replace it with the air. Then measure electrical conductivity. Strange how all people on the earth did not find this way how to save copper in the wires by replacing it with foam.

[rs20]

Exactly, the same as electrical conductivity. Now remove 95 of the copper volume from the wire and replace it with the air. Then measure electrical conductivity.

I've taken the volume drop into account in my calculations by reducing the width of the heatsink to 1cm (that's a foam density of around 10%, not orders of magnitude off) from its initial width of many inches; I got a result for the foam of 0.26 W/K.

2.5 cm high, 5 cm deep, 10 cm wide, 10% density, 2.0 path length factor google link.

Strange how all people on the earth did not find this way how to save copper in the wires by replacing it with foam.

If you took the copper in a wire and fluffed it up so you had the SAME AMOUNT OF COPPER spread over a great cross section, then the conductivity would only be moderately worse. But that would obviously be an expensive and pointless operation since solid wire is rather easy to make and, say, twice as conductive as the foam. If the air around all those wires is actually needed for heat dissipation though, then it might just be a useful idea after all.

[senso]

Even more striking for me is how 3 guys can get all the datasheets to implement a intel i7 processor, plus north-bridge, plus the graphics chip, don't you need to have access to the intel yellow and red books to create/design a motherboard?

[rob77]

Even more striking for me is how 3 guys can get all the datasheets to implement a intel i7 processor, plus north-bridge, plus the graphics chip, don't you need to have access to the intel yellow and red books to create/design a motherboard?

probably they're planning to buy an Intel NUC computer with i7 CPU and reverse engineer the schematics and then do the same with the graphics card 

[wraper]

If you took the copper in a wire and fluffed it up so you had the SAME AMOUNT OF COPPER spread over a great cross section, then the conductivity would only be moderately worse. But that would obviously be an expensive and pointless operation since solid wire is rather easy to make and, say, twice as conductive as the foam. If the air around all those wires is actually needed for heat dissipation though, then it might just be a useful idea after all.

So lets say that foam contains 10% of copper. Then you crush it to 10x smaller size and compare to solid copper block of the same size. Calculate 2X worse thermal conductivity. So you basically compare a foam with cross section 10x of the (very small) solid copper block, yet have 2x worse performance than it. When you expand that foam to previous size, you have 10x of the cross section with 2X worse thermal conductivity than that tiny solid copper block. Therefore 20 times worse per same cross section. Then we for example compare just some random regular copper heatsink with the same size as of that foam. 10% metal of the volume copper plates will have 2x better conductivity than foam. I don't get it. And yeah, regular copper heatsink plates are pretty thin, so using 10% copper foam actually might require more metal than copper plates of the same volume. Moreover so dense foam will be barely able to pass any air through it, so you will be forced to use much thinner foam for that reason, which again will make thermal conductivity much worse.

[wraper]

BTW all those calculations are not precise anyway. Here I found some special foam with 8% copper of the volume http://www.ergaerospace.com/Copper-properties.htm  Thermal conductivity: 10.1 W/m-C = W/m*k
solid copper: 401 W/m*k.

[Hole]

One thing to remember: if there is heat radiation there is absorption, too.

Inside the foam all the radiation (good or bad due to color) is absorbed by its neighbors to a certain degree. So effectively only the outside surfaces of the foam do transport heat away by radiation.

[rob77]

BTW all those calculations are not precise anyway. Here I found some special foam with 8% copper of the volume http://www.ergaerospace.com/Copper-properties.htm  Thermal conductivity: 10.1 W/m-C = W/m*k
solid copper: 401 W/m*k.

thanks for the link !  i was wondering what the exact figure looks like... so it's more than clear that the copper foam is the worst heatsink ever considered in a computer design

[wraper]

Looked at this stuff closely. It's not anything near the proper copper foam with a solid structure as in the link I gave. It more like freaking metal sponge.

[dave_k]

I'm still waiting to find out what cooper foam is....

[rs20]

If you took the copper in a wire and fluffed it up so you had the SAME AMOUNT OF COPPER spread over a great cross section, then the conductivity would only be moderately worse. But that would obviously be an expensive and pointless operation since solid wire is rather easy to make and, say, twice as conductive as the foam. If the air around all those wires is actually needed for heat dissipation though, then it might just be a useful idea after all.

So lets say that foam contains 10% of copper. Then you crush it to 10x smaller size and compare to solid copper block of the same size. Calculate 2X worse thermal conductivity. So you basically compare a foam with cross section 10x of the (very small) solid copper block, yet have 2x worse performance than it. When you expand that foam to previous size, you have 10x of the cross section with 2X worse thermal conductivity than that tiny solid copper block. Therefore 20 times worse per same cross section. Then we for example compare just some random regular copper heatsink with the same size as of that foam. 10% metal of the volume copper plates will have 2x better conductivity than foam. I don't get it. And yeah, regular copper heatsink plates are pretty thin, so using 10% copper foam actually might require more metal than copper plates of the same volume. Moreover so dense foam will be barely able to pass any air through it, so you will be forced to use much thinner foam for that reason, which again will make thermal conductivity much worse.

  To make it easy for you, I provided a calculation with the path length factor (2) and the volume factor (10) taken out, and got my figure of 0.26 K/W. Still you think you're telling me something new. Here it is again, please don't respond to me again without specifically criticizing this calculation:

2.5 cm high, 5 cm deep, 10 cm wide, 10% density, 2.0 path length factor google link.

BTW all those calculations are not precise anyway. Here I found some special foam with 8% copper of the volume http://www.ergaerospace.com/Copper-properties.htm  Thermal conductivity: 10.1 W/m-C = W/m*k
solid copper: 401 W/m*k.

Firstly, from that page you linked to, "Thermal Conductivity Copper has an extremely high thermal conductivity. Duocel® copper foam is used in very high performance heat exchangers where weight and cost are of low priority." So, let's plug that bulk number of 10.1 into my calculation:

2.5 cm high, 5 cm deep, 10 cm wide, Duocel copper foam google link.

Stop the presses! It works out to 0.5 K/W. Furthermore, plug the density you found on that page, and the "tortuosity factor" of 3 from Duocel Foam Thermal Conductivity ( I had guessed 2 ) into my first equation, and my original answer changes to 0.49 K/W. Pretty much bang on. I find this thread in particular to be a strange one for being called "completely wrong" for being off by 50%.

[wraper]

To make it easy for you, I provided a calculation with the path length factor (2) and the volume factor (10) taken out, and got my figure of 0.26 K/W. Still you think you're telling me something new. Here it is again, please don't respond to me again without specifically criticizing this calculation:

Didn't tell you something new. Just wrote how I understand your calculations and compared that with regular heatsink.

Firstly, from that page you linked to, "Thermal Conductivity Copper has an extremely high thermal conductivity. Duocel® copper foam is used in very high performance heat exchangers where weight and cost are of low priority." So, let's plug that bulk number of 10.1 into my calculation:

Picked best possible foam, not some metal sponge as in that computer. What's wrong?

Stop the presses! It works out to 0.5 K/W. Furthermore, plug the density you found on that page, and the "tortuosity factor" of 3 from Duocel Foam Thermal Conductivity ( I had guessed 2 ) into my first equation, and my original answer changes to 0.49 K/W. Pretty much bang on. I find this thread in particular to be a strange one for being called "completely wrong" for being off by 50%.

But again heat exchanger != regular heatsink. There are very high pressure inside which pushes medium through it. They make it up to size: Duocel® copper foam is limited to 1" x 6" x 18" . As heat will come from both sides of the foam in heat exchangers, longest distance heat will need to travel through the foam is 1.25cm. So suppose in the heat exchangers surface area is more important than thermal conductivity. And this foam is much more dense than you would be able to use with simple fan and especially free air flow. If you think that such thermal conductivity is enough, it's not so simple. Even in regular heatsinks with plates which have better thermal conducticity than that foam, heatpipes are used (talking about PC heatsinks where you need to dissipate a lot of heat in restricted volume, not thick ones used in amplifiers, etc.). Just because if heat need to travel through the plate more than few centimeters, it becomes a problem as plate is hot in one end and cool in another. Yet even heatsink with such plate density as PC ones (= high surface area) are not any good for free convection.