Home  |  About  | Last |  Submit  |  Contact
AllQuests.com



Previous Question:  What is the release date for FF13 and WKC ? - GameTrailers.com Forums  PS3Next Question:  Jaffe wouldn
Question Carbon-Carbon tiles + Peltier Question ( bit-tech.net Forums Hardware Software )
Updated: 2008-02-29 17:07:10 (24)
Carbon-Carbon tiles + Peltier Question

Good morning all,

I've been watching a great deal about the discovery space shuttle in the past few weeks, and have noticed that the underbelly of the shuttle is made from a material named carbon-carbon - which acts as a thermal sheild - soaking up the vast temperatures generaged by re-entry and discipating over a long period afterwards...

...now being on this forum we all know that a TEC (Peltier) will cool components or whatever the hell down rapidly with one side and will heat the hell out of the other side - thus requiring a suitible method of heat transfer away from the TEC to prevent it killing itself and whatever it is supposed to be cooling: Namely a CPU.

Now call me crazy...well actually don't please - but does anyone have experience with Carbon-Carbon/know where the hell could buy a tile from.
Because would possibly mean could attach a TEC to it on one side, and a 120mm fan on the other to move the heat away (but given the thermal properties of carbon-carbon wouldn't need to be a high throughput of air) but it would require a circuit to keep the fan on for a calculated amount of time after shutdown as to keep the heat regulated again given due to the fundamental properties of carbon-carbon.

If in theory this system would work - it would be possible to acheive peltier level cooling at a reasonable cost - as current systems also require concurrent watercooling just for the hot side of the TEC. Not only that, but how cool would it be to say part of your shuttle pc has some of the same components as the space shuttle.... ... hmmm maybe not that cool but in my opinion - definately worth a try?

Anyone agree?

Answers: Carbon-Carbon tiles + Peltier Question ( bit-tech.net Forums Hardware Software )
Carbon-Carbon tiles + Peltier Question

Quote:
Originally Posted by clocker
While this is certainly true in many ( possibly most) cases, there are good ideas that lanquish unused for other, primarily economic, reasons.
A great heatsink design might remain unavailable if it only excels on one platform or if it requires expensive hand finishing to complete.

The carbon tile/peltier concept fails basic science theory (it just wouldn't work), but I wouldn't be surprised if ideas previously considered unworkable/too esoteric become more common as requirements demand ( like the liquid metal/pulsed magnetic "pump" sealed system).

If you want to see examples of interesting cooling, look to laptops and SFF cases.
Both suffer from increasing heat loads in shrinking packages so their solutions are (sometimes) ingenious and worth using.
For instance, I've been studying laptop heatpipe HSs for use on the northbridge of my DFI-SLI board.
The ultra low profile necessitated by the overhanging vid cards makes regular fixes difficult, but this is a constraint faced by laptop designers from the onset, so it seemed logical to see what they came up with.
No luck so far (they are seemingly in love with cheap, cast zinc frames and cheesy heatpipes), but hope springs eternal.
Yes, but you're missing my point. I'm saying that whatever the reasons for rejecting (or at least not implementing) novel ideas, other brighter minds have often already thought of them (before rejecting/not following through on them). It is very unlikely that we hobbyists will come up with a fundamentally new, groundbreaking idea in cooling before the specialist engineers did.

Nexxo

Carbon-Carbon tiles + Peltier Question

I think the idea of not having to watercool a pelt is a tough one. I've been playing with the idea of aircooling a mere 80W pelt and I was told by a lotta people that there was no chance. Getting a material to absorb the pelt's heat and release it slowly wont work either, as the heat will build up until the pelt goes bye.

Also, just to clear up a few things about carbon-carbon. It's not the stuff they use under the shuttles in tiles, that stuff is just ceramic. Carbon-carbon is used only in the wing leading edges. AFAIK it was this stuff that caused the last shuttle crash, as it was hit with foam and they got a nice big hole in the wing leading edge. It's very physically strong (but not in all ways, as NASA found), almost certainly VERY expensive (it's not naturally occuring or anything) and yea, it's a great insulator.

Hazza

Carbon-Carbon tiles + Peltier Question

definatly worth tryin, u will have to show us teh results if you can find some carbon-carbon , have u tried googleing?

EDIT:OMG OMFG look at teh price , hope u have deep pockets mate.
Linkage to Carbon-Carbon to buy

r4tch3t

Carbon-Carbon tiles + Peltier Question

Its not just the carbon you need, its the rest of the tile as well... Its a good idea tbh, the tiles are amazing, blow torch one side and the other is cool... You'd be very hard pressed to get a shuttle tile though tbh....

Highland3r

Carbon-Carbon tiles + Peltier Question

you will kill your TEC in about 20 seconds. the space shuttle tiles work by having attrotious thermal conductivity and so dont absorb or emitt heat very well. So you put that bettween a TEC and a cooling source its like you have no cooling at all. bye bye computer.

Lynx

Carbon-Carbon tiles + Peltier Question

Plus it would probably fall off if the Shuttles tiles were anything to go by

sorry.......couldn't resist!

coolmiester

Carbon-Carbon tiles + Peltier Question

err those tiles reflect the heat they dont absrob it.

so you would be sending the heat straight back to the pelt.

dom_

Carbon-Carbon tiles + Peltier Question

i agree with dom, isnt that what would happen?

simosaurus

Carbon-Carbon tiles + Peltier Question

Apparently that isn't what would happen, from what I have been able to gather, the core temperature of the carbon-carbon tile gets to phenomenal heats, whilst the outside of it remains 'cool' - the last time I checked you can't reflect heat caused by friction...it could be my physics thats wrong...but I don't think so - thus the tile must absorb it.

If it did reflect it then it would be bye bye PC, but I remain confident that it wouldn't do so. I think I'll draft an email to my friends at MIT and see if can get a 'sample' from nasa.

And the jokes about the shuttle and the tiles falling off - bad taste IMO.

So the question is out there - anyone familiar enough with this technology to give and EDUCATED opinion on the feasability.

RFID

Carbon-Carbon tiles + Peltier Question

Quote:
Originally Posted by RFID
So the question is out there - anyone familiar enough with this technology to give and EDUCATED opinion on the feasability.
Yes, me. They're a very good high-temperature INSULATOR. The surface layer radiates heat well (high emissivity) and the body doesn't conduct it much (low thermal conductivity).

Quote:
the core temperature of the carbon-carbon tile gets to phenomenal heats, whilst the outside of it remains 'cool'
You can demonstrate the same principle by sticking a lump of polystyrene foam in the freezer for a few hours. Pick it out, it will feel normal temperature to the touch.

cpemma

Carbon-Carbon tiles + Peltier Question

Quote:
Originally Posted by RFID
Apparently that isn't what would happen, from what I have been able to gather, the core temperature of the carbon-carbon tile gets to phenomenal heats, whilst the outside of it remains 'cool' - the last time I checked you can't reflect heat caused by friction...it could be my physics thats wrong...but I don't think so - thus the tile must absorb it.
Nope. The tile acts as an insulator, as cpemma says. Therefore it follows that it hardly conducts any heat at all (otherwise it wouldn't be able to act as an insulator). Therefore it would, in effect block the TEC from passing on its heat anywhere.

The pretty pictures you see of someone holding a glowing tile is really just to demonstrate its insulating properties (and surface heat dissipating abilities) again --except that this time the heat is trapped in the body of the tile, unable to burn the fingers of the hand holding it. Of course it spent a good half hour in a 1300 C oven to get to that state, and of course while it was there its surface got just as hot as its environment --it wasn't somehow cooler while the tile absobed all that heat in its core.

You indeed cannot reflect heat caused by friction (at least not at the point of friction while it is in progress), but the Shuttle does not get hot because of atmospheric friction but because it compresses the air ahead of it to a superheated plasma as it slams into it at a speed of about 27360km/h.

Nexxo

Carbon-Carbon tiles + Peltier Question

Diamond is a great conductor of heat, if I remember correctly.

Stormtrooper

Carbon-Carbon tiles + Peltier Question

Quote:
Originally Posted by Stormtrooper
Diamond is a great conductor of heat, if I remember correctly.

My Pockets are deep, but not THAT deep...ok then, so we have nullified carbon-carbon as a possible intermediary to soak up heat, and replace it with another carbon structure

Anyway I'm not too bothered - was just curious about the feasibility of the idea - which has now been shot down in flames .... ....so other than diamond, is there any material people can think of that would soak up the heat and release slowly from - just so can use a peltier without the hassle of watercooling it too.

Any materials chemists around?

Anyway - let me know.

RFID

Carbon-Carbon tiles + Peltier Question

You say "just so can use a peltier without the hassle of watercooling it too"

but surely your trying to eliminate a small hassle (watercooling) to replace it with a massive hassle (peltier)

I can't see the end goal in what your trying to achieve TBH

coolmiester

Carbon-Carbon tiles + Peltier Question

I'm with coolmiester here. You want to keep things simple and you propose to use a pelt?

Look, no offence, but every now and then a relative newcomer to extreme cooling posts a new idea and thinks he brought us the holy grail of cooling (last time was mineral oil as coolant...). But you have to realise (and this is important, people, so listen up) that engineering minds far brighter and knowledgable than all of us have been looking at ways to efficiently cool increasingly hot processors for years now.

As such, none of us are likely to come up with some groundbreakingly new idea that no-one had ever thought of before. If it hasn't been done before, it's because there was a darn good reason not to.

Nexxo

Carbon-Carbon tiles + Peltier Question

The primary problem with TECs is that the hot side needs to be cooled, or else it fails to keep the cold side cool. If there was some hot object on the hot side, slowly dissapating heat, it wouldnt work for very long.

Stormtrooper

Carbon-Carbon tiles + Peltier Question

Double-carbon composite is made by layering raw carbon fibers with some carbon dust in-between, and then compressing and heating in an autoclave. I very seriously doubt that one could make this at home (in any normal home workshop at least), and I don't know anything else it is used for.
------
I don't know anything about the thermal properties of the shuttle tiles, other than they can withstand some extreme temperatures. One material that sounds familiar in that regard is monolithic graphite, which is used for making cruciables that are used for melting glass and metal in.
By the looks of prices at this site: http://www.graphitestore.com/
it doesn't appear to be very expensive. A sheet .25 thick and 4 x 4 inches = $12.38, not bad at all. But their spec sheet says the metric thermal conductivity is 83 W/(m^2/K/m) where this page: http://www.engineeringtoolbox.com/th...ty-24_429.html
says copper is 401, about 5X better. They list aluminum as 250 but that is probably for pure aluminum; silicon-Al is somewhat higher than that but still not higher than copper. (carbon and graphite are close but not quite the same thing) The only material listed better than copper seems to be silver, a bit better at 429 but not much.
------
Synthetic diamonds seem to be the best solid available material now, price notwithstanding: http://www.aip.org/pnu/1993/split/pnu131-2.htm
-however-
carbon nanotubes do lots of odd things too and large-scale manufacturing is only now becoming possible.
~

Lubb

Carbon-Carbon tiles + Peltier Question

Quote:
Originally Posted by ConKbot of Doom
I'd like to see the diamond for one of the A64s That with phase change cooling on the back side... sounds good to me, who wants to sponsor me?
How about no?

Dude its only going to make a marginal diference anyway. At some point its down to the phase units capacity to handle heat more than it is on the jackets capacity to transfer it. Copper is just fine IMO. If you want better temps the increase offloading capacity insted of trying to increase load.

jaguarking11

Carbon-Carbon tiles + Peltier Question

Quote:
Originally Posted by clocker
If you want to see examples of interesting cooling, look to laptops and SFF cases.
Both suffer from increasing heat loads in shrinking packages so their solutions are (sometimes) ingenious and worth using.
For instance, I've been studying laptop heatpipe HSs for use on the northbridge of my DFI-SLI board.
The ultra low profile necessitated by the overhanging vid cards makes regular fixes difficult, but this is a constraint faced by laptop designers from the onset, so it seemed logical to see what they came up with.
No luck so far (they are seemingly in love with cheap, cast zinc frames and cheesy heatpipes), but hope springs eternal.
Well heatpipe tech is still relatively new in the enthusiast market. Im afraid you wont find a high power heat dispurcing unit on a laptop. But the ides on some of them havent changed mutch since the p1 days. The idea is removing the heat from one spot and transfering it to a larger area. For you chipset you could implement a heatpipe sink and move the heat towards the rear of the mobo where the exaust fan is (abit and asus are both doing this) and the heat output is not high enough to warant anymore. If you want better temps sandwitch a 20w or a 30w pelt and copper plate and heatpipes with large fins on the end to disperce heat and it will be more than enough for a chipset.

80w pelt can be cooled with air but you need massive amounts of surface area. I have a 320w unit that I powered at 7v and it frezes on one end after a fiew seconds but the heat reaches peak heatsink capacity after a fiew minutes and then the cold side becomes hot and I just shut it off for fear of pelt suicide.

jaguarking11

Carbon-Carbon tiles + Peltier Question

Quote:
Originally Posted by Nexxo
My point is that these new ideas have often already been considered (and rejected) by people much more knowledgable than us.
While this is certainly true in many ( possibly most) cases, there are good ideas that lanquish unused for other, primarily economic, reasons.
A great heatsink design might remain unavailable if it only excels on one platform or if it requires expensive hand finishing to complete.

The carbon tile/peltier concept fails basic science theory (it just wouldn't work), but I wouldn't be surprised if ideas previously considered unworkable/too esoteric become more common as requirements demand ( like the liquid metal/pulsed magnetic "pump" sealed system).

If you want to see examples of interesting cooling, look to laptops and SFF cases.
Both suffer from increasing heat loads in shrinking packages so their solutions are (sometimes) ingenious and worth using.
For instance, I've been studying laptop heatpipe HSs for use on the northbridge of my DFI-SLI board.
The ultra low profile necessitated by the overhanging vid cards makes regular fixes difficult, but this is a constraint faced by laptop designers from the onset, so it seemed logical to see what they came up with.
No luck so far (they are seemingly in love with cheap, cast zinc frames and cheesy heatpipes), but hope springs eternal.

clocker

Carbon-Carbon tiles + Peltier Question

Quote:
Originally Posted by jaguarking11
But if everyone shut their mind and never experimented this world would not have anything new. its nice to have an idea and its also good to suport it if it makes sense to you.
My point is that these new ideas have often already been considered (and rejected) by people much more knowledgable than us. So if someone thinks they have a new idea it is best to do some basic research first into whether it has been thought of or discussed before, and whether it might actually work (and not get all huffy when people say it won't).

To illustrate: the latest idea is to embed nano-tubes filled with distilled water in a 1mm copper die. It was dreamt up by a NASA engineer and he has 17 million to throw at it. I don't think we will top that easily.

Nexxo

Carbon-Carbon tiles + Peltier Question

Quote:
Originally Posted by Nexxo
I'm with coolmiester here. You want to keep things simple and you propose to use a pelt?

Look, no offence, but every now and then a relative newcomer to extreme cooling posts a new idea and thinks he brought us the holy grail of cooling (last time was mineral oil as coolant...). But you have to realise (and this is important, people, so listen up) that engineering minds far brighter and knowledgable than all of us have been looking at ways to efficiently cool increasingly hot processors for years now.

As such, none of us are likely to come up with some groundbreakingly new idea that no-one had ever thought of before. If it hasn't been done before, it's because there was a darn good reason not to.
But if everyone shut their mind and never experimented this world would not have anything new. its nice to have an idea and its also good to suport it if it makes sense to you. yes the carbon tiles on the shuttle actualy reflect the heat. And also if it absorbed that mutch heat it would have to go somewere and eventualy the tiles would be at a thermal peak and either disintegrate or transfer it to the shuttle boddy.

In the past pelts were used very efectively and relatively hassle free with large heatsinks and fans. Now cpu's are mutch hotter and need more load capacity.

Now with this being said. I know they do make lab dimonds and if one were to obtain a lab made dimond and actualy imbed it in the bottom of a block and used it directly on the die while the other side got cold fluid it would help with temps. But fact remains its not practical and plainly stupid to do. if you want colder temps then insulate your block and tubes and run a chiller. It will be more eficient than a pelt and possibly cost you less on your monthly electrical bill. this is not realy hard to do and readymade chillers can be picked up used from bar outlets for cheep.

jaguarking11

Carbon-Carbon tiles + Peltier Question

To the best of my knowledge, that being a degree in materials science, alloying materials lowers their thermal conductivity. This is the case with aluminium silicon.

Silicon is added to aluminium to improve its castability. However, it does generally have a higher thermal conductivity than typical engineering alloys.

There has always been a lot of discussion regarding novel materials for waterblocks, particularly the base, but lets be honest, they are expensive and with copper being as good as it is, all you are trying to do is improve the one feature of cooling technology that doesn't need that much improving. By this I mean that getting the thermal energy from the base of the block to the surface in contact with the water is becoming the least significant portion of the total thermal resistance from cpu to water.

Now, interface materials is somewhere where we really could do with some novel ideas. If we consider the sequence of thermal resistances from cpu to water -

cpu to base plate
base plate to block internals
block internals to water

changing the material of the block changes the middle one, and by a multiplication factor, so if we get double the thermal conductivity, we halve the resistance.

However, cpu to baseplate is a pretty darn large resistance in the grand scheme of things, and yet we only attribute 2-5% of the cost of the block to this area.

So anyone with some spare creativity kicking about, please direct it to getting the heat from the cpu into the block itself more efficiently.

8-ball

8-BALL

Carbon-Carbon tiles + Peltier Question

Quote:
Originally Posted by jaguarking11
...
Now with this being said. I know they do make lab dimonds and if one were to obtain a lab made dimond and actualy imbed it in the bottom of a block and used it directly on the die while the other side got cold fluid it would help with temps. But fact remains its not practical and plainly stupid to do. ...
I'd like to see the diamond for one of the A64s That with phase change cooling on the back side... sounds good to me, who wants to sponsor me?

ConKbot of Doom

Previous Question:  What is the release date for FF13 and WKC ? - GameTrailers.com Forums  GameTrailers.com Forums  PS3Next Question:  Jaffe wouldn

- Source: Carbon-Carbon tiles + Peltier Question bit-tech.net Forums Hardware Software
- Previous Question: What is the release date for FF13 and WKC ? - GameTrailers.com Forums GameTrailers.com Forums PS3
- Next Question: Jaffe wouldn't mind a little involvement with God of War III - GameTrailers.com Forums GameTrailers.com Forums PS3