Carbon Nano-Tube technology. Heatsinks and thermal-plates

[ZytheEKS]

I am adding a potted peltier block in between my CPU and Waterblock-Heatsink. I am also trying to increase the space between my waterblock and motherboard to allow more vertical clearance as I have a Swiftech Apogge Waterblock/pump on a Crosshair V motherboard. There is absolutely no space in-between the R.A.M. chips and the angled Barb inserts on the waterblock. I would like to add a liquid heatsink to my R.A.M. as I will be soon installing 2400MHz R.A.M.

It matters not whether you know the specifics on all this, because all that is needed to know is that I need a thermal-plate to add spacing between my CPU and Heatsink. I'm adding a potted peltier to increase thermal transfer, and am wondering if anyone knows where I can acquire a Carbon Nano-Tube thermal-plate.
Like this, only made out of carbon nano-tube
http://www.frozencpu.com/products/2410/exp-03/40x40mm_Copper_Cold_Plate.html?tl=g30c105

 

[mokrunka]

I'm not sure you'll find carbon nanotube based thermal standoffs yet. I'm sure you know this, but theoretically, your nanotubes have relatively high thermal conductivity, so they would be a good choice for your standoff. They also do not have very good strength in compression. From my brief exposure, and limited knowledge (from a graduate materials science course) there has not been a lot of work done in creating composite structures from CNTs; please enlighten me if you're aware of applications that have done this.

There are places that you can buy Carbon nanotubes, but they are very expensive (like $30/gram expensive).

 

[merz.38]

You could see if someone would write you a research grant?

 

[ZytheEKS]

I'm not sure you'll find carbon nanotube based thermal standoffs yet. I'm sure you know this, but theoretically, your nanotubes have relatively high thermal conductivity, so they would be a good choice for your standoff. They also do not have very good strength in compression. From my brief exposure, and limited knowledge (from a graduate materials science course) there has not been a lot of work done in creating composite structures from CNTs; please enlighten me if you're aware of applications that have done this.

There are places that you can buy Carbon nanotubes, but they are very expensive (like $30/gram expensive).

I know of two current applications in the modern industry.
Some british researchers developed a carbon nano-tube CPU heatsink the U.S. Airforce uses. Extremely expensive, not suitable for civilian or professional use.

I also have heard of a LOT of people who mix carbon nano-tube powder in with their thermal-paste they use on their CPU/heatsink. Supposedly it makes a significant difference. People seem to mix a large amount of it in with their thermal paste so that it almost creates a thin dense wall between the CPU and heatsink after compressed and heated. Apparently since Carbon Nano-tube's thermal conductivity is significantly higher than copper it transfers heat from the CPU to the heatsink at a notably higher rate than a thermal paste would alone.


Edit: If you don't believe me about the heat-sink here's this
http://www.sbir.gov/sbirsearch/detail/81968

 

[mokrunka]

Yes, the thermal conductivity of CNT is about 10 times that of Cu. I'm not sure how well it would work to add it to your thermal paste, although it couldn't hurt to try. You can buy a small amount in "powder form" in a film canister type container and give that a try.

You could also see if you could somehow mill out a standard heat sink of copper, and try to fill it with the material. Outside of something like that, you may actually need a research grant to try and figure out how to manufacture something in the form of a heatsink.

I don't doubt that the SBIR found someone to make one for the airforce, but to me, that only reaffirms the idea that this type of part will not be found on the open market!

 

[Conumdrum]

Maybe the op should look into Unobtainium.

Read up on it on Wiki, AMAZING STUFF.

 

[ZytheEKS]

Hmmmm, this all gives me an idea. Within the year I'm going to try something, and I'll bump this forum topic once I try it with the results.

Also, my father works for Boeing, and I know for a fact that with some of the insane products they are very likely to use nano-tube based heatsinks on not only computer parts but engine cooling, and for core-fuselage temperature maintenance.

Here is a declassified prototype aircraft that might use if for core-fuselage cooling. (Heat-shielding isn't nearly enough as the 2011 crash into international waters clearly showed. The heat weakened the integrity of the hull and it ripped itself apart in mid flight)
http://en.wikipedia.org/wiki/DARPA_Falcon_Project
Direct your attention to the HTV-3X


Again, it asserts your notion for open market it is not available, but I am confidant that if they know how to make it they would definitely use it if it had significantly higher thermal transfer.

 

[Conumdrum]

Good luck. Big dreams does much for the future.

 

[bluezero5]

your concern is actually cost.

if you want the best thermal conductivity, look no further, DIAMOND is what you are looking for, at around twice the conductivity of any carbon tech I have seen. However, you will run into a huge bottleneck, which is the CPU IHS, making such good thermal conductivity somewhat pointless.

 

[Seebs]

your concern is actually cost.

if you want the best thermal conductivity, look no further, DIAMOND is what you are looking for, at around twice the conductivity of any carbon tech I have seen. However, you will run into a huge bottleneck, which is the CPU IHS, making such good thermal conductivity somewhat pointless.

Simple solution... De-Lid CPU and slap $10K worth of diamond dust on the actual die... then clamp on your $10M solid diamond heatsink straight to the CPU die... After that; it's just a matter of explaining to the wife why it is that the Feds are knocking on your door wanting to know how you got into Fort Knox.

 

[mokrunka]

LOL-this has gone south quickly

 

[snowbrd]

Simple solution... De-Lid CPU and slap $10K worth of diamond dust on the actual die... then clamp on your $10M solid diamond heatsink straight to the CPU die... After that; it's just a matter of explaining to the wife why it is that the Feds are knocking on your door wanting to know how you got into Fort Knox.

I think thats perfectly viable. But I heard that unobtanium can transfer and dissipate all the heat of any cpu out there.

 

[NamesLucky]

Carbon nanotubes do have a very high thermal conductivity in the axial direction, but radial thermal conductivity is somewhere along the lines of soil, which is very bad. Thus, in order to achieve improved thermal conductance over other thermal gels/pastes, you would have to have the CNTs aligned from the heatsink case to the processor. At Georgia Tech people are researching using aligned CNT forests as chip packaging, and I believe they are making headway, however use in a thermal paste seems unlikely since alignment is hard to accomplish in a paste. Possibly, in the future, a hard, CNT forest thin film could be adopted as a replacement to thermal pastes, or maybe there's something like that out there now that I'm not aware of.

 

[bobolot]

Sorry for bumping this up again,
oh, and hello forum,

how much extra height do you need? and what are the dimensions of the CPU heat spreader?


As for why i'm interested:
I'm working on high conductivity (but not diamond high) carbon based materials, and trying to find marketable uses for them is for them is surprisingly hard. Any chances for people wanting to look at the materials i will dive at!

 

[Leegit]

Carbon nanotubes do have a very high thermal conductivity in the axial direction, but radial thermal conductivity is somewhere along the lines of soil, which is very bad. Thus, in order to achieve improved thermal conductance over other thermal gels/pastes, you would have to have the CNTs aligned from the heatsink case to the processor. At Georgia Tech people are researching using aligned CNT forests as chip packaging, and I believe they are making headway, however use in a thermal paste seems unlikely since alignment is hard to accomplish in a paste. Possibly, in the future, a hard, CNT forest thin film could be adopted as a replacement to thermal pastes, or maybe there's something like that out there now that I'm not aware of.

Correct. Since they must be axially align, you would have to build a scaffolding with an array of chambers out of nanomaterial which would allow you to insert CNTs and maintain there vertical position. The scaffolding would also have to be out of a material high thermal conductivity or it would defend the purpose of even using CNTs.

I begin my PhD studies on graphene nanocomposites in 2 months