Other substances for wcing.

[l33t p1mp]

How about mercury? Other than the safety problems, would it do a better job than water? I will keep looking to see if there are any other possibilities, should make for a fun experiment. :D

Water (k=.58) is 3x better than alcohol (k=.17) and mercury (k=8) is 14x better than water at removing heat. (Formula used was k = W/mK).

[coolmiester]

Its finding a pump that could cope with the weight of mercury though

[HESmelaugh]

That would be one bold experiment!

I guess the reason we don't see alternatives to water used very often is that while there are substances that could theoretically be used and would remove more heat than water, using them is usually a lot more of a hassle and a lot more expensive than water.

[T_Flight]

Mercury?!? :eek: No way! That stuff is toxic as all getout. Totally impractical for watercooling. In some places it is regulated, and in other it is forbidden...meaning you can't get it. It would be a PITA to handle, and if it got on anything would be an instant short. Trying to keep from making a mess with the stuff or avoiduing spills wouldn't be impossible, but it wouldn't be easy either. The stuff is dangerous. Stay away. Stay far away from Mercury.

Here is the MSDS Sheet For Mercury. Please I beg anyone that is thinking of trying this or who reads this. Please stay away from Mercury. It is bad news.

http://fscimage.fishersci.com/msds/96252.htm

[l33t p1mp]

Its finding a pump that could cope with the weight of mercury though

Any pump off a car, I have seen it done, a guy had a loop that was over 60 feet long and had 2 inch wide pvc.


Mercury?!? :eek: No way! That stuff is toxic as all getout. Totally impractical for watercooling. In some places it is regulated, and in other it is forbidden...meaning you can't get it. It would be a PITA to handle, and if it got on anything would be an instant short. Trying to keep from making a mess with the stuff or avoiduing spills wouldn't be impossible, but it wouldn't be easy either. The stuff is dangerous. Stay away. Stay far away from Mercury.

Here is the MSDS Sheet For Mercury. Please I beg anyone that is thinking of trying this or who reads this. Please stay away from Mercury. It is bad news.

http://fscimage.fishersci.com/msds/96252.htm

Lol, I wasnt going to try mercury, thats just asking to die from toxic fumes.

[coolmiester]

Any pump off a car, I have seen it done, a guy had a loop that was over 60 feet long and had 2 inch wide pvc.


Well i hold my hands up on that one i must admit.

While looking for other substances for cooling, how about a material to replace copper??

[Polizei]

Mercury isn't as deadly as you make it out to be. It's not one sniff and you have cancer or you're dead. It takes years and years before there is even a possibility of a birth defect or genetic mutation (cancer). I'm not saying it's safe to take a bath in, but you are making it out to be worse than building a playground at Chernobyl.

[l33t p1mp]

Mercury isn't as deadly as you make it out to be. It's not one sniff and you have cancer or you're dead. It takes years and years before there is even a possibility of a birth defect or genetic mutation (cancer). I'm not saying it's safe to take a bath in, but you are making it out to be worse than building a playground at Chernobyl.

Since its liquid it gets all over the place, and it most likely eats away at the plastic and cause a mercury leak in a running pc...not fun, but nice fireworks. Im not taking risks though, my pc is in my bedroom and I dont want fumes in my room lol. :rofl:


Well i hold my hands up on that one i must admit.

While looking for other substances for cooling, how about a material to replace copper??

Aluminum. :D

[coolmiester]

Aluminum. :D

I think that one's been tried already!!!!

........how about carbon??

http://www.coolercases.co.uk/carbon/carbon_1.jpg

[fart_plume]

Silver is a better conductor of heat than copper . just darn expensive.

[l33t p1mp]

Silver is a better conductor of heat than copper . just darn expensive.

And gold is better than silver. Blingy computer lol. :rofl::ROTF::clap:

[Polizei]

Since its liquid it gets all over the place, and it most likely eats away at the plastic and cause a mercury leak in a running pc...not fun, but nice fireworks. Im not taking risks though, my pc is in my bedroom and I dont want fumes in my room lol. :rofl:
:D

"most likely" != fact


And gold is better than silver. Blingy computer lol. :rofl::ROTF::clap:

Gold is a heat reflector. Gold is better than silver when it comes to conducting electricity, not heat.

[Waterlogged]

Gold is a heat reflector. Gold is better than silver when it comes to conducting electricity, not heat.

:yepp:

Silver=429
Copper=401
Gold=310
Aluminum=250
Mercury=8

http://www.engineeringtoolbox.com/thermal-conductivity-d_429.html

[coolmiester]

But again, we have had many silver blocks including the DD RBX, DD TDX, Storm and G5 but it was found that the added expense did not justify the less than 1degree temp drop.

Personally i think we should be looking at a cheaper substitute that could be easier to machine while still remaining on par or better than copper and light in weight to boot ;)

[atomic ferret]

Mercury is a dangerous substance, but keep in mind that years ago it was played with as a toy with bare hands. The vast majority of the people exposed directly will not suffer any adverse side effects, but prolonged exposure will cause serious health issues. Clearly, using it to cool a computer is impractical and dangerous, but the stuff won't kill you immediately on contact.

[l33t p1mp]

Lol, reminds me of the lead paint on the baby toys...

[rocket733]

But again, we have had many silver blocks including the DD RBX, DD TDX, Storm and G5 but it was found that the added expense did not justify the less than 1degree temp drop.

Personally i think we should be looking at a cheaper substitute that could be easier to machine while still remaining on par or better than copper and light in weight to boot ;)

Sounds like you have something in mind...

[cegras]

thing is passes the skin a takes residency in organs....want to have normal children?

I see a lot of 'may', as well as no indication of time-dependence for the effects.

[CyberDruid]

How about mercury? Other than the safety problems, would it do a better job than water? I will keep looking to see if there are any other possibilities, should make for a fun experiment. :D

Water (k=.58) is 3x better than alcohol (k=.17) and mercury (k=8) is 14x better than water at removing heat. (Formula used was k = W/mK).

the Blood of a freshly pwned n00b:up:

[Sparky]

Gold is a heat reflector. Gold is better than silver when it comes to conducting electricity, not heat.

No it isn't. Actually it is worse than copper (not a huge amount, but still copper is better). However, what it has going for it is it doesn't tarnish or corrode. That's how a gold-plated connector is better than just copper, the gold keeps it from tarnishing (which reduces electrical conductance) but the gold itself is thin enough that it doesn't really impact overall electrical conductivity. Drops it a little but much less than a dirty tarnished copper connection.

[zeroibis]

You could use that liquid metal thermal past...

[exe163]

At the price of replacing water and copper, you might as well move 1 step up from liquid cool altogether.

[l33t p1mp]

the Blood of a freshly pwned n00b:up:

Blood is better than water believe it or not, but I wont start gouging myself or go vampire on my local blood bank just to get some nice temps. Would be horrible to dispose of without contaminating anything though.

[Aberration]

No it isn't. Actually it is worse than copper (not a huge amount, but still copper is better). However, what it has going for it is it doesn't tarnish or corrode. That's how a gold-plated connector is better than just copper, the gold keeps it from tarnishing (which reduces electrical conductance) but the gold itself is thin enough that it doesn't really impact overall electrical conductivity. Drops it a little but much less than a dirty tarnished copper connection.

+1 :up:

[Kayin]

Some of us are working on alternative materials.

We're just being quiet about it...

[XS_Rich]

If the processor was running hot enough, liquid sodium is quite a handy coolant... :D

http://en.wikipedia.org/wiki/Liquid_metal_cooled_reactor#Sodium_and_NaK

Might be a problem if there're any bubbles of air in the loop....

[[XC] riptide]

NOBODY TRY MERCURY - XS Warning. :D

[gurusan]

Gold is a heat reflector. Gold is better than silver when it comes to conducting electricity, not heat.

I'm pretty sure silver is a far better electrical conductor than gold. Gold is only used because it's very stable, so contacts do not corrode.

[just a noob]

what about barrilium(or however its spelled) liquid at room temperature, is it still up there with mercury in terms of thermal transfer?

[Kayin]

Galinstan, if you want a liquid metal.

Honestly, I think the focus is all wrong.

I'll return from my lab when I have my answer to all this.

[Extigy]

carbon nanotubes ftw. (A few years off tho :P)

[b@llz0r]

not that any of us use Alu in our loops, but mercury will disolve aluminium http://xmb.stuffucanuse.com/xmb/viewthread.php?tid=315

I worked at a gas refinery in central australia, come of the wells were sucking up some mercury so we installed a mercury trap.
One of the boffins took a jar of the stuff with him on the flight out... the jar broke and melted a hole in the fuselage... expensive :banana::banana::banana::banana:-up


I guess from a liquid cooling POV the main properties are
- specific heat capacity, how much heat can the liquid hold for each 1C increase in temperatur and,
- thermal conductivity. how readily will heat be transfered to the liquid

for common liquids water has the best specific heat capacity, but other elements have better thermal conductivity... id be very interested in experiments with different mixtures and solutions... I guess we are almost talking nanofluids now


I was thinking of a poormans nanofluid... just mixing some thermal paste with water

[csjsi]

I think that one's been tried already!!!!

........how about carbon??

http://www.coolercases.co.uk/carbon/carbon_1.jpg

Has anyone tried it ??

[Kayin]

Nanotubes only transfer heat along their axes, and they're hydrophobic. They basically fail at cooling in their stock config. It takes more money than I have sense to make it work right.

[demonkevy666]

I think that one's been tried already!!!!

........how about carbon??

http://www.coolercases.co.uk/carbon/carbon_1.jpg

is that ceramic coated ?

because I have a model engine that has some aluminum Oxide ceramic on it and it's smooth and Extremely powerful also very cool running.

[Kayin]

That's a path worth following...

[Salad Fingers]

Isn’t this all a bit irrelevant??? I mean the single most important thing here is always going to be the radiator/s. Want lower temps get bigger ones and more of them!

And anyway for fluids wouldn’t it be better to have something that’s got the highest thermal conductivity with the lowest heat capacity. Sure it wouldn’t be best to run blocks in series but you would get awesome drops in temp over the radiator.

Edit: Do processors use copper in their internals??
If so why not swap to silver, higher electrical conductivity (http://www.engineeringtoolbox.com/conductors-d_1381.html) = less heat yeah?

[Sparky]

processors are silicon, a semiconductor. Can't use a metal for a CPU ;)

And btw most normal water loops have the blocks in series.

[zeroibis]

Here is the efficiency of some possible blocks or plating that can be used for heat sinks: (calculated by atomic weight / specific heat)

(Higher is better)
Aluminum: 1.114939612
Zinc: 2.568080094
Copper: 2.600081833
Silver: 4.255155819
Platinum: 7.543851508
Gold: 8.066449709
Nickel: 2.251377062
Thallium: 7.765322948

The order is in specific heat, so you can see the relation that is has. I expect that people already know which ones are heavier than others.http://www.ibisgaming.com/forums/showthread.php?t=860

[b@llz0r]

isnt thermal conductivity the property you want in a block?

I would have also thought that a lower specific heat would be preferable so that it can easily transfer the heat to the liquid

by the looks of these web sites silver is the winner with pure copper a close second

http://www.engineeringtoolbox.com/thermal-conductivity-metals-d_858.html
http://www.reade.com/Particle_Briefings/thermal_con_ceramics.html

[l33t p1mp]

I guess thallium might be worth a try.

[Polizei]

I guess thallium might be worth a try.

And you were worried about Mercury poisoning. Thallium is an awful death.

[Serpentarius]

Has anyone tried it ??

it's very "uncommon topic" .. there are ppl using "graphite" (aka pencil lead) as a form of thermal transfer since graphite is also another form of carbon ... just it's very highly electric conductive (might cause discharges)

in fact i'm reading some article if it's possible to use DIY pencil leads as TIMs ..

[upsidedown]

thermal conductivity of Diamond: 900 - 2320 :D

[Serpentarius]

why dont you use gallium, germanium, bismuth, antimony? atleast their safer?

i've read bout gallium is being used as TIMs .. in some company .. (Galiums are liquid in room temperature) ..

antimony & bismuth is normally used in soft soldering (lead, tin etc) .. i havent seen virgin antimony & bismuth before though

hope this helps

[b@llz0r]

thermal conductivity of Diamond: 900 - 2320 :D


You could have something there

Synthetic diamonds are used in many industrial aplications and nowhere near as expensive as the real thing.

impossible to work with tho so they would need to be manufactured as a block.

[b@llz0r]

Industrial diamond industry
The market for industrial-grade diamonds operates much differently from its gem-grade counterpart. Industrial diamonds are valued mostly for their hardness and heat conductivity, making many of the gemological characteristics of diamond, including clarity and color, mostly irrelevant. This helps explain why 80% of mined diamonds (equal to about 100 million carats or 20,000 kg annually), unsuitable for use as gemstones and known as bort, are destined for industrial use. In addition to mined diamonds, synthetic diamonds found industrial applications almost immediately after their invention in the 1950s; another 3 billion carats (600 metric tons) of synthetic diamond is produced annually for industrial use.

The dominant industrial use of diamond is in cutting, drilling, grinding, and polishing. Most uses of diamonds in these technologies do not require large diamonds; in fact, most diamonds that are gem-quality except for their small size, can find an industrial use. Diamonds are embedded in drill tips or saw blades, or ground into a powder for use in grinding and polishing applications. Specialized applications include use in laboratories as containment for high pressure experiments (see diamond anvil cell), high-performance bearings, and limited use in specialized windows.

With the continuing advances being made in the production of synthetic diamonds, future applications are beginning to become feasible. Garnering much excitement is the possible use of diamond as a semiconductor suitable to build microchips from, or the use of diamond as a heat sink in electronics.


:up:

[Salad Fingers]

isnt thermal conductivity the property you want in a block?

I would have also thought that a lower specific heat would be preferable so that it can easily transfer the heat to the liquid

Exactly as I said above, well at least tried too.

I've been thinking some more and I am convinced that to obtain the best temperatures for your components your loop should be constructed from materials that have the highest thermal conductivity with the lowest heat capacity possible. Now that two don't go hand in hand so you'd have to ratio the figures for all the possible materials and select the best choice from there.

I might have a quick look after work today to see how hard this would be to do.

[b@llz0r]

Exactly as I said above, well at least tried too.

I've been thinking some more and I am convinced that to obtain the best temperatures for your components your loop should be constructed from materials that have the highest thermal conductivity with the lowest heat capacity possible. Now that two don't go hand in hand so you'd have to ratio the figures for all the possible materials and select the best choice from there.

I might have a quick look after work today to see how hard this would be to do.

Go steal a diamond tip off a drill rig and strap it to your CPU :)

DO IT NOW!!!

[Salad Fingers]

Go steal a diamond tip off a drill rig and strap it to your CPU :)

DO IT NOW!!!

hahahaha. Got the dudes from sandvik and major drilling outside my office right now, I should make some enquiries.....

EDIT: we got bought out today (http://www.thewest.com.au/default.aspx?MenuID=3&ContentID=112292) so I think a few missing drill bits wont impact the bottom line like it could have.. hmmmmm

[basserdan]

Liquid salt! Just watching a show about alternative energy. A solar power station that focuses reflected sunlight onto tubes containing oil which heats water to steam blah, blah... Anyway they also heat "liquid salt", flowing in tubes, to heat the water at night. Supposedly the liquid salt has a very high heat capacity. Does high heat capacity = good for removing heat? You only asked for ideas, you didn't say it couldn't be stupid!

[b@llz0r]

ideas that seem stupid to start with usually end up being the best ones

[Polizei]

Liquid salt! Just watching a show about alternative energy. A solar power station that focuses reflected sunlight onto tubes containing oil which heats water to steam blah, blah... Anyway they also heat "liquid salt", flowing in tubes, to heat the water at night. Supposedly the liquid salt has a very high heat capacity. Does high heat capacity = good for removing heat? You only asked for ideas, you didn't say it couldn't be stupid!

I think salt needs to be over 1000c before it becomes a liquid... it's got a huge boiling point.

[relttem]

I guess from a liquid cooling POV the main properties are
- specific heat capacity, how much heat can the liquid hold for each 1C increase in temperatur and,
- thermal conductivity. how readily will heat be transfered to the liquid

for common liquids water has the best specific heat capacity, but other elements have better thermal conductivity... id be very interested in experiments with different mixtures and solutions... I guess we are almost talking nanofluids now


I was thinking of a poormans nanofluid... just mixing some thermal paste with water

bunch of that on here:

http://www.xtremesystems.org/forums/showthread.php?t=209373&highlight=nanofluid

http://www.xtremesystems.org/forums/showthread.php?t=202442&highlight=nanofluid

http://www.xtremesystems.org/forums/showthread.php?t=205442&highlight=nanofluid

[crazy1323]

Mercury is a dangerous substance, but keep in mind that years ago it was played with as a toy with bare hands. The vast majority of the people exposed directly will not suffer any adverse side effects, but prolonged exposure will cause serious health issues. Clearly, using it to cool a computer is impractical and dangerous, but the stuff won't kill you immediately on contact.

we used to put it on lab tables in the engineering labs and put electromagnets on the other side. The mercury would "slide" across the table.

It is not as bad as people make it out to be. Just wear gloves and make sure it is disposed of properly.

[rocket733]

I'm interested in hearing some more about the carbon ocz baseplate, Coolermeister :poke:

[Serpentarius]

we used to put it on lab tables in the engineering labs and put electromagnets on the other side. The mercury would "slide" across the table.

It is not as bad as people make it out to be. Just wear gloves and make sure it is disposed of properly.

wont mercury melt copper??

i recall smelting gold requires mercury to melt it .. and rephased to extract the gold from its ore ..

i do know know if copper is the same case ... since mercury could melt aluminium (as per read above post) which has higher tenstile strength than the copper

surely u guys want to encourage him to try on mercury? might cause more harm than crunching

[Polizei]

surely u guys want to encourage him to try on mercury? might cause more harm than crunching

More harm than crunching?

[MaddHawk]

I was gonna quote a few posts I saw with some serious misconceptions and general ignorance in them but, seeing the large volume of them and that others were correcting them I decided not to. I will weigh in on this topic as I have been wondering just how much materials study there has been done for use in water block construction, cooling fluid, and the radiator.

First a little about my own background and experience in this. Now I am certianly not an expert of any nature but I do have alot of experience in a job where proper cooling of equipment was of be-all-end-all of our existance.
Specifically, I used to be a nuclear trained electrician's mate for the US Navy. As you might imagine keeping a nuclear reactor properly cooled is extremely important.

Now I would first like to take my hat off to b@llz0r. You sir have a solid grasp on common sense and knowledge of the preferable characteristics needed for cooling systems.

I assume we all know the basic components of any cooling loop that are involved with the thermal performance of said loop. But to properly list what characteristics we need I'll list them anyway. One note, it has been over a year since I was in the navy and dealt on the issue of thermal and fluid dynamics so I am on the rusty side, especially where the mathematics is concerned but, I do have my "Applied Engineering Principles" prepared by Bettis Atomic Power Lab and Knolls Atomic Power Lab wich is a compressed version of the basics I used while in the Navy.

The thermal components of a cooling loop are the heat source, the first heat exchanger, the primary coolant, the heat sink aka second heat exchanger, the secondary coolant, and the pump. The for the purpose of these forums our heat source is our CPU, the first heat exchanger is our CPU heatsink or for liquid cooling our water block, the primary coolant is most usually DI water, the second heat exchanger is our radiator, the secondary coolant is the air being blown through the radiator, and the pump is included not because its vital (indeed it is possible to use natural convection to provide the motive force of our primary coolant when the loop is setup properly) but because it also adds a small measurable amount of heat to our loop. Basic equation for heat flow in the system is:

Qsource + Qpump = Qradiator

Where Q - amount of thermal energy.

Now for the characteristics we want in each component of our cooling loops:

Please note Wikipedia is an encyclopedia and done in the wiki format. As such there is a chance that the information contained may be inacurate and/or incomplete. Please use the information in the following links with a grain of salt.
Thermal Conductivity (http://en.wikipedia.org/wiki/Thermal_conductivity)
Specific Heat Capacity (http://en.wikipedia.org/wiki/Heat_capacity#Heat_capacity)
Area (http://en.wikipedia.org/wiki/Area)
Viscosity (http://en.wikipedia.org/wiki/Viscosity)
Flow Rate (http://en.wikipedia.org/wiki/Flow_rate)
Pump Head aka Discharge Pressure (http://en.wikipedia.org/wiki/Pressure)
Pressure Drop (http://en.wikipedia.org/wiki/Pressure_drop)
I will try and list the properties we want in each component in order of most important.

Heat source - there is nothing we can do about this component since its characteristics are determined by the materials selected by AMD and Intel, which is in turn governed by what they need in order to make efficient and powerful processors, best we can do is hope that they give us large surface areas to make contact with our water blocks

First heat exchanger - this is what removes the bulk of the thermal energy from our heat source and transfers it to our primary cooling medium, we want:
Area - the larger the area in contact with our primary coolant the better
Thermal conductivity - the larger the better
Pressure drop - the lower the better, a lower pressure drop allows our primary coolant a higher mass flow rate, the more coolant mass flowing over the heat exchanger the more thermal energy is removed

Primary coolant - this is our liquid flowing through our pump, cooling blocks, radiator, and associated piping, usually water, we want:
Specific heat capacity - larger is better, as the heat capacity goes up our coolant can absorb more heat from our water block and carry it away to our heat exchanger, do not confuse this with thermal conductance which is good for a heat exchanger walls but very bad for the conveying medium linking heat source to heat sink
Viscosity - lower is better as the fluid will have smaller pressure drops as it passes through various components in the system, though too low and our pump won't be able to move it

Heat sink aka second heat exchanger - this is where the thermal energy is removed from our primary coolant to the secondary coolant which ultimatly takes this energy away from our PCs, we want:
Area - again the larger the better
Thermal conductivity - again the larger the better
Pressure drop - again the lower the better, due note the secondary heat exchanger is usually has the largest pressure drop in the system do to being typically much larger than the heat sources as it is being required to dissipate the heat generated by multiple sources and possibly to a secondary coolant with worse heat capacity than the primary coolant

Secondary coolant - this is the coolant that carries heat away from our second heat exchanger and is our ultimate heat sink, we want:
Specific heat capacity - the larger the better
Viscosity - again lower is better but, not so low our pump can't move sufficient mass

Pump - final component in our thermal loop, this is what provides our primary and secondary coolants with their forced flow, I am including both liquid pumps and fans for air in this catagory since all liquids and gasses are considered fluids, we want:
Pump head aka discharge pressure - the larger the better, all the piping, heat exchangers, and various fittings to link them together have various pressure drops across them as the fluid flows by, our pump must be able to overcome this cumulative pressure drop and do it while still maintaining sufficient mass flow rate through the system
Flow rate - this one is dependant upon the size of the system piping, lower flow provides for lower pressure drops in the system thus less pumping power needed but also removes less heat from the heat source, a higher flow will create a larger pressure drop in the system but will also move more mass across through the heat exchangers and therefore move more heat from our heat source

I hope this helps to explain a bit more of what people should be looking for in their components in their cooling loops with a tiny bit of scientific reasoning. The Wikipedia links above will help explain a bit more in detail and as I bone up on my own rusty knowledge of thermodynamics and fluid dynamics, I'll try and help explain it to anyone not full grasping it.

Also, hopefully, the above will help explain slightly what properties they need in various materials when building the cooling system components.

[cegras]

If the primary and secondary coolants have poor thermal conductivity, that heat isn't going anywhere.

[b@llz0r]

(deleted for readability)

probably the best overview on the science of watercooling ive ever read.
this should be made into a sticky
:clap:

[T_Flight]

:DI think leet pimp has an itching to play with deadly substances. Maybe we should put him in a containment suit and let him play with some of these things to get the full effect. I suggest Hypergolic Fuels. That requires Class 5 Containment, and some of these fuels are extremely hazzardous. Not only will they burn you if they come into contact with each other, the chemicals themselves are extremely hazzardous all by themselves. Wanna play with chemicals? Those are the big daddy's of them all. :yepp: :D

[l33t p1mp]

:DI think leet pimp has an itching to play with deadly substances. Maybe we should put him in a containment suit and let him play with some of these things to get the full effect. I suggest Hypergolic Fuels. That requires Class 5 Containment, and some of these fuels are extremely hazzardous. Not only will they burn you if they come into contact with each other, the chemicals themselves are extremely hazzardous all by themselves. Wanna play with chemicals? Those are the big daddy's of them all. :yepp: :D


STOP GRAMMAR TIME! Lets make a list, shall we?
1. My name is l33t p1mp.
2. Hazardous has only 1 z.
3. Other than that, good job.
4. I am cutting you some slack on "wanna".
5. Containment does not need a capital in the middle of the damn sentence.
6. You do not say has an itching, the preferred method is "has an itch" or "is itching to".

:D

Now for a reply.

Damn how many times do I need to say that I wasn't even going to try freaking mercury???

[rocket733]

STOP GRAMMAR TIME! Lets make a list, shall we?
1. My name is l33t p1mp.
2. Hazardous has only 1 z.
3. Other than that, good job.
4. I am cutting you some slack on "wanna".
5. Containment does not need a capital in the middle of the damn sentence.
6. You do not say has an itching, the preferred method is "has an itch" or "is itching to".

:D

Now for a reply.

Damn how many times do I need to say that I wasn't even going to try freaking mercury???

:rofl: Thanks for the laugh.

[l33t p1mp]

I plan on doing this to every post I reply to, its amusing, and shows people how bad they are with "ye olde keyboard". :D :rofl: :rotf: :clap:

[cegras]

Halt! Hammerzeit!

[l33t p1mp]

False, it is halt, hammerzeit. Keep trying...my turn, more languages (yes I speak these).
Pare, es el tiempo del martillo!
Arrêtez, c'est l'heure du marteau!
Stop, hammer time!
Arresti, es il tempo del martello!

[Jass]

Did anyone try to use anti freeze?

I would also want to experiment using urine. Seriously, it absorbs a lot of heat from the body, so it should also be able to do the same for the computer :)

[rocket733]

Did anyone try to use anti freeze?

I would also want to experiment using urine. Seriously, it absorbs a lot of heat from the body, so it should also be able to do the same for the computer :)

Anti-freeze has been used previously but did not help the heat transfer. Urine is actually an interesting case because it potentially contains ammonia (if it's not being completely converted into urea) which has a higher specific heat capacity than water (4.86 -at 104 F- vs 4.19). Of course since the boiling point is -27 F it isn't too practical.

[l33t p1mp]

Did anyone try to use anti freeze?

I would also want to experiment using urine. Seriously, it absorbs a lot of heat from the body, so it should also be able to do the same for the computer :)

I aint pissing in my loop, you can try it if you want. :p:

[MaddHawk]

If the primary and secondary coolants have poor thermal conductivity, that heat isn't going anywhere.

Not true. Heat will always go from hot to cold so long as both substance have any thermal conductivity at all and a difference in thermal energy exists between the two.

To know which property is more important you need to understand how the heat is being moved. With out water block on the CPU heat is moved from the CPU by direct contact via a method called conduction. The purpose of the water block though, is not to be the heat sink but merely provide a safe and reliable means of transfering that heat to our coolant. In fact if AMD and Intel were to manufacture the CPU case with a grid pin array directly on the case and provide a means of mounting a water cooling top directly to the case and seal it properly, you wouldn't even need a water block base. This would remove this conduction middleman. Point is for conduction you want a high thermal conductivity, for which the best material so far pointed out would in fact be the diamond form of carbon.

Now your comment was about our coolant and not about the block. So what about the coolant? Well our coolant absorbs heat from the block by convection, which is the transfer of thermal kinetic energy via a moving fluid. Since we use a pump to provide he motive force for moving our fluid, PC liquid cooling systems are called 'forced convection cooling'. Blowing air directly over a heatsink is also forced convection cooling. For convection style cooling the principle thermal property of concern is no longer thermal conductivity but rather specific heat capacity. We are using the movement of the mass of our coolant to remove the heat from our source and convey it to our sink, not the transfer of heat directly through the water.


Specific heat capacity, also known simply as specific heat, is the measure of the heat energy required to increase the temperature of a unit quantity of a substance by a certain temperature interval.

Basicly, specific heat capacity, tells us how much energy it takes to raise the temperature of a substance of a certian mass or volume by 1 degree.


Heat capacity (symbol: Cp) — as distinct from specific heat capacity — is the measure of the heat energy required to increase the temperature of an object by a certain temperature interval.

This is basicaly the same as specific heat capacity but with out the reference to the mass or volume of the substance in question. As such to be of value we must specify the amount of mass or volume, i.e. a bathtub of water has a larger heat capacity than a cup of water.

Lets look at the specific heat capacity for Mercury and water. The values for this is specified per mole of molecules at 25°C.



Water liquid 4.186 J cm−3 K−1
Mercury liquid 1.888 J cm−3 K−1

As you can see water more than twice the amount of thermal energy to raise its temperature by 1 degree. So for a given temperature rise in the coolant, water exracts more thermal energy from our heat source.

"Well since Mercury has almost 14 times the thermal conductivity of liquid water, wouldn't it more readily absorb heat from the block and transfer it to the radiator?" Initially, yes it would but, this would provide a benefit only for a short while. Because Mercury has a greater thermal conductivity it would saturate its heat carrying capacity from our water block much faster than water would. Since our secondary coolant is still air the amount of heat the CPU dumps into the mercury coolant would exceed that of the air's ability to remove it. Since its heat capacity is much less than that of water it would rapidly saturate the cooling systems ability to absorb heat from the CPU. Once our mercury cooling system becomes saturated the total system temperature begins to rise starting at our heat source. Eventually this rise in temperature will reach the point that our CPU fries itself or that the ⌂T across the radiator from the mercury to air is sufficient to bring our system to an equalibruim temperature.

Now lets look a water. Water has 0.60 W·m−1·K−1 thermal conductance. Nearly 14 times less than that of mercury at 8.30  W·m−1·K−1. But it has more than double the heat capacity of mercury (1.888 J cm−3 K−1) at 4.186 Joules per cubic centimer per kelvin (1°K = 1°C, just the basis for 0° is different.) This results in that water takes longer to heat up and maintains a larger ⌂T across the water block and removes more thermal energy as result. Water also is much closer to air as far as thermal conductivity goes and as such the rate of heat dumped into the air is able to follow much more closely to heat absorbed from the CPU. Water is able to store much more thermal energy in itself while the radiator heats up to a point that the ⌂T across it is able to pass sufficient heat to the air and cool the water for return to the CPU. This allows us to go with smaller radiators and cooling systems for a greater amount of heat transfered by the cooling fluid.


Now to go on to potentially better substances, one fluid I saw listed as having a higher heat capacity than water is ammonia. In fact Wikipedia listed it as having the highest specific heat capacity of any known chemical. Beyond this I really haven't read on since their are better ways to improve our cooling system efficiencies besides using hazardous chemicals and elements.

What can be done to improve our liquid cooling systems? For starters improving the surface area of the radiators and water blocks in contact with the water. Second we can change our secondary coolant to provide us with colder temperatures on the secondary side of the second heat exchanger. Instead of using a radiator we could use a water/water heat exchanger with chilled water flowing flowing as our secondary coolant.

Take the Cooler Master 1000 Coolermeister chopped (http://www.xtremesystems.org/forums/showthread.php?t=182967) up and put in a chiller system into it and then run your coolant loop outside your PC case the chiller case and into the water/water heat exchanger. This would provide a much larger degree of cooling on our primary coolant without needing to run hefty insulated tubing into our primary case. You could even set it up so the junction going into your PC is done via quick disconnect and the external tubing is part of the chiller system.

Another way to improve thermal performance is to use all silver water blocks and all silver radiators. Of metals, silver has the best thermal conductance of all. For those with the truly xtreme budget to match their xtreme system they could try and order a custom artificial diamond block and radiator.

Another option is to improve the flowrate of the coolant with more pumps and more powerful pumps. Try putting pumps in parallel, not series to get greater flow rates. Also try using educator-jet pump (http://en.wikipedia.org/wiki/Eductor-jet_pump) in your loop. The discharge of your normal pump goes into motive connection with your main loop going into the suction of the educator with a small side connection going to your pumps suction. This allows for the use of a smaller pump to provide a similar flow through the system or to use a stronger pump to provide an even higher flow rate that it could achieve on its own.

[Polizei]

I aint pissing in my loop, you can try it if you want. :p:

"Aint" isn't a word, even if you put the apostrophe in it. Grammar time comes to bite you in the ass, huh?

Stop spamming, please. You have enough posts to view the For Sale forum.

[l33t p1mp]

Wow take a chill pill...did the local pub run out of ale or something? And yes, aint isn't a word, I got lazy.

[Riggs]

Wow take a chill pill...did the local pub run out of ale or something? And yes, aint isn't a word, I got lazy.

Calm down. Im kinda tired of the spam tbh also.

[Omastar]

STOP GRAMMAR TIME! Lets make a list, shall we?
1. My name is l33t p1mp.
2. Hazardous has only 1 z.
3. Other than that, good job.
4. I am cutting you some slack on "wanna".
5. Containment does not need a capital in the middle of the damn sentence.
6. You do not say has an itching, the preferred method is "has an itch" or "is itching to".

:D

Now for a reply.

Damn how many times do I need to say that I wasn't even going to try freaking mercury???

Actually, the conjugated form is "let's", a contraction of "let us". FAIL.

[Polizei]

The period should go inside the quotation marks on '"wanna".' as well.

[MaddHawk]

And here I was hoping for a thread that would include some serious discusion as to the materials used in cooling loops and now it's all :off:.

[Waterlogged]

Ok, let's just say we did find something better than copper and water, would it make any kind of difference in overall temps if we're still using air as the final link in heat removal? Isn't that really the weak link here?

[Lekko]

I think we should devote more research into utilizing cats as a medium to transfer heat.

I have noticed that cats tend to be found absorbing heat off radiators, sun spots, and other heated areas. If you could train cats to circulate and alternately lie on your CPU, then go outside for a bit and return in a loop, they could possibly be a viable source of cooling. Although I have yet to run the numbers.

[Polizei]

Ok, let's just say we did find something better than copper and water, would it make any kind of difference in overall temps if we're still using air as the final link in heat removal? Isn't that really the weak link here?

The weak link is people don't want to move enough air through a radiator to get optimum cooling. If we are all about getting the heat out of the processor - everyone buys the newest blocks out there - shouldn't they all be about getting all that "new" heat out of the water that the blocks sapped from the processor?

You buy a block better at moving heat from processor to water, but why skimp on fans that move heat from water to air? The weak link is the amount of air people are willing to move.

You can move more air at the same noise level only if you get a bigger radiator and people already complain about the size of radiators. You can move more air but at a greater noise level while keeping the same size radiator.

[MaddHawk]

The weak link is people don't want to move enough air through a radiator to get optimum cooling. If we are all about getting the heat out of the processor - everyone buys the newest blocks out there - shouldn't they all be about getting all that "new" heat out of the water that the blocks sapped from the processor?

You buy a block better at moving heat from processor to water, but why skimp on fans that move heat from water to air? The weak link is the amount of air people are willing to move.

You can move more air at the same noise level only if you get a bigger radiator and people already complain about the size of radiators. You can move more air but at a greater noise level while keeping the same size radiator.

I actually answered that in my previous post.

What can be done to improve our liquid cooling systems? For starters improving the surface area of the radiators and water blocks in contact with the water. Second we can change our secondary coolant to provide us with colder temperatures on the secondary side of the second heat exchanger. Instead of using a radiator we could use a water/water heat exchanger with chilled water flowing flowing as our secondary coolant.

Take the Cooler Master 1000 Coolermeister chopped up and put in a chiller system into it and then run your coolant loop outside your PC case the chiller case and into the water/water heat exchanger. This would provide a much larger degree of cooling on our primary coolant without needing to run hefty insulated tubing into our primary case. You could even set it up so the junction going into your PC is done via quick disconnect and the external tubing is part of the chiller system.

Take my previous job background for example. I was a sailor in the US Navy in the submarine force. We used water cooling for all of the ships primary electronic components including the backbone of the ships lan. This water cooling was in turn cooled by another water cooling system that used sea water for its coolant.

Now I doubt there are many, if any at all, that can use some sort of permanent cool water flowing through a secondary loop that has a more or less permanent source and drain. The closest I can think of would be to use city tap water from the household plumbing and then direct another pipe/hose to drain to the sewer or the yard. Since I know water isn't free in most places this can be rather impractical. But using a chilled water system for our secondary loop we can put the 2nd Law of Thermodynamics to use for us.


Heat generally cannot spontaneously flow from a material at lower temperature to a material at higher temperature. Informally, "Heat doesn't flow from cold to hot (without work input.)"

With a type of refrigeration system, heat can be forcibly removed from a cold object and added to a hot one. This can provide us with a better secondary coolant and increase the efficiency of our system to remove and disperse our PCs heat. The down side is the increase in power consumption.

[jasonelmore]

:yepp:

Silver=429
Copper=401
Gold=310
Aluminum=250
Mercury=8

http://www.engineeringtoolbox.com/thermal-conductivity-d_429.html

Diamond is the most conductive material on the planet... Outside of earth?! not so much

http://geekchic.merseyblogs.co.uk/Luthor%2520Kryptonite.jpg

[relttem]

or you can increase the convective coefficient, h, of the fluid to increase the heat transfer.

[Fresh Daemon]

<Off-topic>

Interesting story, at least to me:

Back in the day, the Soviet Union built some nuclear subs with liquid-metal-cooled reactors. These things were great at producing a very small and efficient reactor, and since the USSR didn't have a safety culture to speak of (see Chernobyl, Mayak) they didn't care that it was pretty dangerous. Anyway, if the temperature in these things dropped too much below several hundred degrees, the liquid metal would solidify in the reactor and permanently ruin it. Therefore, when these subs were in port, they would run tubes from the quay with superheated steam to keep the reactor hot while it was shut down.

But this was the USSR, and people were constantly stealing things, stuff would break and never get repaired, and so on. The subs themselves were the pride of the Soviet fleet, so they were fine, but the steam units were a different story, they'd always be broken. Therefore, the sub crews had to run the reactors constantly, even when in port, so these things gave out much sooner than they should have from overuse (not that they were built that well to begin with, again, Soviet Union = poor design and engineering), not to mention irradiating people from these dubiously-shielded reactors sitting in ports for months.

Sorry for the off-topic, but the subject just reminded me of this. But, slightly more on-topic, the reason the US never bothered with these things was not because they lacked the engineering know-how (being far ahead of the USSR in pretty much every technical/engineering field), but because they realized that it was too dangerous and/or just too much of a design/engineering nightmare for the payoff to be worth it.

So, returning to the topic, you could probably find a substance that was better than water, but if it was more expensive, less readily available and required more complex engineering, really, why bother?

[MaddHawk]

For the temperatures found in our PCs these days, you really can't do better than water for your primary coolant. But that doesn't mean there aren't other areas that can't be improved.

Related to you submarine tidbid, the first nuclear powered ship in the world, the USS Nautilus, was also cooled by liquid metal. I would also beg to differ that a liquid metal cooled reactor could be built to be smaller and more efficient than a water cooled one, after all I was a nuclear reactor operator for the US Navy for 5 years. ;) :up:

PS - I really like playing with the smilies here. :D

[Fresh Daemon]

I'm pretty sure the USS Nautilus had a S2W pressurized water reactor. I think you mean USS Seawolf. They didn't bother building more than one sodium-cooled reactor in the US Navy because they thought it posed too many safety hazards.

Anyway, the Soviet Alfa class was originally intended as an experiment to build a small interceptor submarine with a very small crew and a high degree of automation, and the way they originally got such a small and powerful reactor was cooling it with molten lead. As I said, it wasn't such a good idea, but then the USSR was known for fielding notably crappy designs.

[Fresh Daemon]

Now I doubt there are many, if any at all, that can use some sort of permanent cool water flowing through a secondary loop that has a more or less permanent source and drain. The closest I can think of would be to use city tap water from the household plumbing and then direct another pipe/hose to drain to the sewer or the yard. Since I know water isn't free in most places this can be rather impractical. But using a chilled water system for our secondary loop we can put the 2nd Law of Thermodynamics to use for us.

Although if you live near a large body of water, or can sink a well... actually, wasn't there a guy in the UK some years ago who buried a big tank in his back yard below the frost line and just used that for a "radiator"?

[crazy8s846]

If it were me, i would not worry about some super comductive fluid, it would proably be cheeper to just lower the ambient temp in the room where the computer is. That would bring your temps down more than another rad. of some super fluid.

[T_Flight]

How about Moonshine? :D You could chill that down and get some cooling for real! :yepp: I betcha you'd be careful with that too, because if you didn't you'd be putting out a big clear blue flame that's almost invisible.

How do you like THAT one? :ROTF:

[Oli4v]

what about beer cooling ? ^^, would be fun experiment ^^

[demonkevy666]

what about beer cooling ? ^^, would be fun experiment ^^

well it's mixed with carbonated water too, won't that be bad for pressure build up. also the amount of alcohol is pretty low.
also need to think abotu the Thermal conductivity I found one I was looking for

Canola Oil
Thermal Conductivity (W/m°K) 0.179 - 0.188

lol

[Jah]

A note on using ammonia as a coolant, it has for water cooling bad, proper of dissolving copper, and if memmory serves me well also aluminium. Not to mention it's FANTASTIC smell. It really clears out you nostrils :shocked:

Ammonia also happens to have a boiling point way below zero so a mix of ammonia and water would be useful in you silver block ;)

[zeroibis]

A big part is factoring in the ability for something to "hold" heat. Things can heat up fast and transfer it quickly are ideal for a block that can then be cooled off quickly.

The list that I showed earlier that took atomic weight/specific heat shows an elements ability to hold heat divided by its ability to transfer heat. (assuming my understanding is correct)


Here is the efficiency of some possible blocks or plating that can be used for heat sinks: (calculated by atomic weight / specific heat)

(Higher is better)
Aluminum: 1.114939612
Zinc: 2.568080094
Copper: 2.600081833
Silver: 4.255155819
Platinum: 7.543851508
Gold: 8.066449709
Nickel: 2.251377062
Thallium: 7.765322948

The order is in specific heat, so you can see the relation that is has. I expect that people already know which ones are heavier than others.

This data demonstrates why in practice gold out preforms the other options, because despite some elements offering the ability to rapidly transfer heat they are not as efficient at storing the heat because there atomic mass is less. I came up with this idea after taking with my room mate about using diamond for cooling and his explanation of why that was a bad idea. (He is going to pharmacy school)

[Zehnsucht]

A big part is factoring in the ability for something to "hold" heat. Things can heat up fast and transfer it quickly are ideal for a block that can then be cooled off quickly.

The list that I showed earlier that took atomic weight/specific heat shows an elements ability to hold heat divided by its ability to transfer heat. (assuming my understanding is correct)



This data demonstrates why in practice gold out preforms the other options, because despite some elements offering the ability to rapidly transfer heat they are not as efficient at storing the heat because there atomic mass is less.

Cheeses, not one more, below is from another post where someone thinks that alu, despite having lower thermal thermal conductivity is better suited for blocks since copper somehow stores heat...
Also, Cu > Au in thermal conductivity (401 vs 318).
Secondly, diamonds are used as heat sinks in IC packages.

My response...

The ability of a metal to transfer heat is described by the term thermal conductivity. The thermal conductivity of pure copper can be expressed as 401 W/m-C, compared to pure aluminum's 220 W/m-C. Higher thermal conductivity means faster heat transfer.

The puzzling, persistent myth among computer heatsink geeks is that, somehow, copper absorbs heat faster than aluminum, yet releases it more slowly.
This is simply wrong.

The metal cannot care or know whether the heat is coming or going, its thermal conductivity remains constant. Copper's thermal conductivity is always higher than that of aluminum.

Also, material density has NOTHING to do with thermal conductivity:

Copper
Density: 8.96 g/cm³
Thermal conductivity : 401 W·m^−1·K^−1

Lead
Density: 11.34 g/cm³
Thermal conductivity : 35.3 W·m^−1·K^−1

Diamond
Density: 3.53 g/cm³
Thermal conductivity : 2000-2500 W·m^−1·K^−1

The thermal conductivity is dependent on how the atoms which build the material in question interact with each other.

When talking about blocks for heat transfer:
We are interested in the surface area which dissipates the heat, ergo thermal conductivity is the key since it is a function of surface area.

When we are talking about a medium which transfers heat between two sources (i.e. water in our case):
Then we are interested in the specific heat capacity, since it's a function of the mass of the substance.
Here we want as much as possible, since we want to transfer as much heat as possible from the source.
Water: 4.2 kJ/kgK
Ammonia: 4.7 kJ/kgK

[demonkevy666]

Cheeses, not one more, below is from another post where someone thinks that alu, despite having lower thermal thermal conductivity is better suited for blocks since copper somehow stores heat...
Also, Cu > Au in thermal conductivity (401 vs 318).
Secondly, diamonds are used as heat sinks in IC packages.

My response...

The ability of a metal to transfer heat is described by the term thermal conductivity. The thermal conductivity of pure copper can be expressed as 401 W/m-C, compared to pure aluminum's 220 W/m-C. Higher thermal conductivity means faster heat transfer.

The puzzling, persistent myth among computer heatsink geeks is that, somehow, copper absorbs heat faster than aluminum, yet releases it more slowly.
This is simply wrong.

The metal cannot care or know whether the heat is coming or going, its thermal conductivity remains constant. Copper's thermal conductivity is always higher than that of aluminum.

Also, material density has NOTHING to do with thermal conductivity:

Copper
Density: 8.96 g/cm³
Thermal conductivity : 401 W·m^−1·K^−1

Lead
Density: 11.34 g/cm³
Thermal conductivity : 35.3 W·m^−1·K^−1

Diamond
Density: 3.53 g/cm³
Thermal conductivity : 2000-2500 W·m^−1·K^−1

The thermal conductivity is dependent on how the atoms which build the material in question interact with each other.

When talking about blocks for heat transfer:
We are interested in the surface area which dissipates the heat, ergo thermal conductivity is the key since it is a function of surface area.

When we are talking about a medium which transfers heat between two sources (i.e. water in our case):
Then we are interested in the specific heat capacity, since it's a function of the mass of the substance.
Here we want as much as possible, since we want to transfer as much heat as possible from the source.
Water: 4.2 kJ/kgK
Ammonia: 4.7 kJ/kgK

then some kinda crystal heat sink would do very well. other then diamond, but what...?

maybe even a magnet lol jk

[just a noob]

a thallium heatsink could be interesting, unless someone mis-handled it, resulting in a death similar to arsenic

[Duniek]

anyone mentioned about mercury ? :eek: :D

check this out (old photo)
http://img257.imageshack.us/img257/6647/hg1dz1.jpg http://img261.imageshack.us/img261/561/hg2or2.jpg

[fart_plume]

A note on using ammonia as a coolant, it has for water cooling bad, proper of dissolving copper, and if memmory serves me well also aluminium. Not to mention it's FANTASTIC smell. It really clears out you nostrils :shocked:

Ammonia also happens to have a boiling point way below zero so a mix of ammonia and water would be useful in you silver block ;)

Ammonia is poisonous in any quantity larger than an ampule for waking someone who is passed out. This is why it is no longer used as a refridgerant in household appliances. It is however still used in some commercial manufacturing plants as a refridgerant.

[DonFaygo]

OK, you find someone to build a diamond water block and I'll pimp that!!

[zeroibis]

Yes, that is what I was saying was that copper DOES transfer faster. Aluminum was a bad example sorry, what I was trying to say was if you have two elements that can transfer heat at equal rates (for 1mol this is the amount used to measure specific heat) then the element with the most atomic weight will be the best choice because while they can both transfer heat at the same rate the one with a greater atomic weight can hold more heat.

Also it was my understanding that diamond is used as an insulator ... (except for blue ones)

[MaddHawk]

I'm pretty sure the USS Nautilus had a S2W pressurized water reactor. I think you mean USS Seawolf. They didn't bother building more than one sodium-cooled reactor in the US Navy because they thought it posed too many safety hazards.

I think you missed the part where I said I was a nuclear operator for the US Navy. :rolleyes: To expound on that, I was a nuclear trained electrician in the US Navy. I went through the full Nuclear Navy Training Program and served on two submarines. The first was the USS Salt Lake City (SSN-716) now decomissioned and the second was the USS Virginia (SSN-774). The USS Seawolf is a water reactor. I was trained on her prototype plant when I was still in training. As part of the program we are also taught a good deal of he history of the program. The USS Nautilus was a liquid sodium cooled reactor and so was the first iteration of the D1G reactor at Ballston Spa, New York.


A big part is factoring in the ability for something to "hold" heat. Things can heat up fast and transfer it quickly are ideal for a block that can then be cooled off quickly.

The list that I showed earlier that took atomic weight/specific heat shows an elements ability to hold heat divided by its ability to transfer heat. (assuming my understanding is correct)

This data demonstrates why in practice gold out preforms the other options, because despite some elements offering the ability to rapidly transfer heat they are not as efficient at storing the heat because there atomic mass is less. I came up with this idea after taking with my room mate about using diamond for cooling and his explanation of why that was a bad idea. (He is going to pharmacy school)

Your roomate is wrong, so is thinking that the heat capacity and specific heat capacity play a role in transfering thermal energy. The thermal conductivity of a material is dependant upon the molecular and atomic structure of the material in question. To understand why, you must understand how thermal energy moves about inside matter. Thermal energy is the kinetic energy of the atoms and molecules in a material and the more they bounce about or vibrate the greater their thermal energy and thus the greater the "heat" of the substance. Heat moves about in a material when its atoms and molecules bounce into each other. When they collide thermal energy from the "hotter" atom/molecule moves to he "cooler" one.

Think of it like this. Take a pool table with billard balls on it. When you strike the white ball it becomes our "heat source or hot atom". It continues to move until it strikes another ball. When it does it transfers its motion or "heat" to the other ball.


Long post with correct information..

You sir are totaly correct. :up: I have tried several times to get the same information across and explain why this is so by showing how thermal dynamics is the foundation upon which our material selections must rest. I guess most people don't have the patience to properly educate themselves on the basics. :shakes:


Yes, that is what I was saying was that copper DOES transfer faster. Aluminum was a bad example sorry, what I was trying to say was if you have two elements that can transfer heat at equal rates (for 1mol this is the amount used to measure specific heat) then the element with the most atomic weight will be the best choice because while they can both transfer heat at the same rate the one with a greater atomic weight can hold more heat.

Also it was my understanding that diamond is used as an insulator ... (except for blue ones)

Like I said above, atomic weight has no bearing on heat transfer nor on heat capacity. The mistake you are making with diamond is in assuming that diamond as an electrical insulator also = thermal insulator. Now I don't know of what kind of electrical properties a diamond may or may not have, but I do know that the rigidness of its crystaline structure is what makes it a supurb thermal conductor. Its crystal is so rigid, hence its hardness, that it moves heat through it VERY fast. This property is part of all diamonds. This is why NASA build diamond windows over thermal moniters for is space probes. This is also why chip makers would use diamond for heat transfer from the chip to the outside of the casing.

[T_Flight]

How about DU (Depleted Uranium) for a block material? :ROTF:

Now THAT would make a nice conversation piece.

[MaddHawk]

Make sure you got your lead lined gloves. :p:

[relttem]

The specific heat, Cp, DOES play a role in heat transfer. This can be seen in the equation:

Q = mdot*Cp*dT

So, does density. Since the Reynolds number is proportional to the density of the fluid, and Nusselt number is a function of the Reynolds number, which further breaks down to the thermal conductivity being a function of the Nusselt number.

[zeroibis]

In that case, they should try to make a thermal past that has diamond in it as diamond dust is relatively cheep and would offer better performance over other options :D

[Zehnsucht]

In that case, they should try to make a thermal past that has diamond in it as diamond dust is relatively cheep and would offer better performance over other options :D

Already done, ic diamond 7 carat:
http://www.innovationcooling.com/article1.htm

Review
http://xtreview.com/addcomment-id-2884-view-IC-diamond-7-vs-arctic-silver-5-review.html

This thread is getting way out of track.

[gxavier]

I think you missed the part where I said I was a nuclear operator for the US Navy. :rolleyes: The USS Seawolf is a water reactor. I was trained on her prototype plant when I was still in training. As part of the program we are also taught a good deal of he history of the program. The USS Nautilus was a liquid sodium cooled reactor and so was the first iteration of the D1G reactor at Ballston Spa, New York. The D1G reactor was also the prototype reactor for the USS Nautilus.


That doesn't mean you're right...

The original poster is correct. You have it backwards. The USS Nautilus did indeed use a pressurized water reactor, and the prototype reactor was not the D1G, but the S1W located in Idaho.

The Seawolf (SSN-575) used liquid-sodium... the Nautilus used water. The 575 was the only sub in US history to use a liquid metal sodium reactor.

A simple check at globalsecurity will confirm this.

[T_Flight]

Global security may or may not be right. It is not an official Gov't site eventhough it may sound like it. We've had alot of problems with that site before with info on the F16 that was posted before it was even declassed which we knew was wrong, becasue they couldn't possibly know that info if it was classified. Later as things have been declassed, things in our simulator have had to be changed which we actually thought were correct.

Thrust numbers and percentages for engibne and throttle tables was a big one that took almost 6 months of dev work for one individual to fix and get right. Some of their munitions, range, and missile data is way off. They have re-printed alot of what the Military has told them and given them a between " ____ and _____" range, and then will interpret that as a certain range that they "think" is correct, but in reality none of them are correct and are actually classified. I won't get into some of the ranges on rockets and missiles, but many of them are not even close, and even if they were told what they were they probably wouldn't believe it anyway. It's best that people do not know except for the Servicemen and women that use those.

That site does have alot of good info, and is kinda neat, but it can be innaccurate also. In fact, some of the info they were given was purposely inaccurate. This keeps the other side guessing no matter what country it is. Willful disinformation is used frequently.

The thing about info there is...

If it's not older than 25 years, and has not been declassed by the President of the USA it cannot be regarded as accurate in the USA. That is the law. In the case of a dicrepancy between that and Federal law it would even override it, because it is regarded as a National Security Issue. Some things like Nuclear secrets can be extended out indefinitely. Obviously things that could destroy the entire planet will never be declassified.

Don't take everything at Globalsecurity.org as the gospel.

[gxavier]

Don't take everything at Globalsecurity.org as the gospel.


Agreed... But these aren't F-22 specs we're talking about here. This is rather common information that the first nuclear sub reactor in history (Nautilus) was water-cooled. It's also common information that the second nuclear sub in history (Seawolf) was liquid-sodium cooled.

This is much much further back than 25 years ago. The Nautilus was comissioned in 1954 and the Seawolf only a year later.

[m^2]

There were some CuSil (http://en.wikipedia.org/wiki/CuSil) blocks made in Poland. It was said to have 515 W/mK (silver - 429).

[Zehnsucht]

There were some CuSil (http://en.wikipedia.org/wiki/CuSil) blocks made in Poland. It was said to have 515 W/mK (silver - 429).

From your link to the wiki:


I re-measured it, and calculated Cusil thermal conductivity to be 371 W/m/K.


Sounds quite unscientifically to write "I re-measured it". Who remeasured it?

[zeroibis]

Already done, ic diamond 7 carat:
http://www.innovationcooling.com/article1.htm

Review
http://xtreview.com/addcomment-id-2884-view-IC-diamond-7-vs-arctic-silver-5-review.html

This thread is getting way out of track.

Using that in my next build :up:

[Easybeat]

I have always wondered why Metal Foams have never made an appearance in cooling. Its very light and has a massive surface area so as a radiator it would be excellent.

http://en.wikipedia.org/wiki/Aluminium_foam

[MaddHawk]

The specific heat, Cp, DOES play a role in heat transfer. This can be seen in the equation:

Q = mdot*Cp*dT

So, does density. Since the Reynolds number is proportional to the density of the fluid, and Nusselt number is a function of the Reynolds number, which further breaks down to the thermal conductivity being a function of the Nusselt number.

I agree for a fluid. ;) If you read more carefully when I said specific heat isn't applicable you will see I was refering to the water blocks and radiators.


That doesn't mean you're right...

The original poster is correct. You have it backwards. The USS Nautilus did indeed use a pressurized water reactor, and the prototype reactor was not the D1G, but the S1W located in Idaho.

The Seawolf (SSN-575) used liquid-sodium... the Nautilus used water. The 575 was the only sub in US history to use a liquid metal sodium reactor.

A simple check at globalsecurity will confirm this.

You are telling me, the guy who SERVED and WORKED on nuclear powered submarines for FIVE YEARS that I am wrong and a public non-government related site is correct. :mad: I checked those sources and also looked at Wikipedia as well to see what they said. I don't know why they are wrong but, my guess is people got information about the first Seawolf and the Nautilus mixed up. Furthermore when you check the official history sites the US Navy maintains they do not speak of the reactor technology. This is most likely due to the fact that even if declassified they still don't want their technology to be come common knowledge. I am not going to speak any more on this as I can not honestly say weather or not I have overstepped my own authority by potentialy releasing what might actually be still classified information.

Edit: Just remember, trust everything on the internet, and ignore the guy with first hand knowledge and experience.


From your link to the wiki:
Sounds quite unscientifically to write "I re-measured it". Who remeasured it?


Mr. Toshi Oyama, Ph.D. R&D Manager, WESGO Metals
:cool:

[MaddHawk]

I have always wondered why Metal Foams have never made an appearance in cooling. Its very light and has a massive surface area so as a radiator it would be excellent.

http://en.wikipedia.org/wiki/Aluminium_foam

Possible, you just might be onto something there. Question would be what kind of resistance to water and air flow would it provide.

[Zehnsucht]

:cool:

Uhnnn...

:coffee: :coffee:

[MaddHawk]

Haha, I was using the Wikipedia quote to answer the question as to who did the testing.

[cegras]

You've done it now MaddHawk. They're going to knocking on your door soon .. : P

[Fresh Daemon]

<off-topic>


I think you missed the part where I said I was a nuclear operator for the US Navy. :rolleyes: To expound on that, I was a nuclear trained electrician in the US Navy. I went through the full Nuclear Navy Training Program and served on two submarines. The first was the USS Salt Lake City (SSN-716) now decomissioned and the second was the USS Virginia (SSN-774). The USS Seawolf is a water reactor. I was trained on her prototype plant when I was still in training. As part of the program we are also taught a good deal of he history of the program. The USS Nautilus was a liquid sodium cooled reactor and so was the first iteration of the D1G reactor at Ballston Spa, New York. The D1G reactor was also the prototype reactor for the USS Nautilus.

Seawolf SSN-575, not SSN-21.

Seawolf SSN-575 at history.navy.mil (http://www.history.navy.mil/danfs/s9/seawolf-ii.htm):

"Seawolf returned to General Dynamics Corp., on 12 December 1958 for refueling and conversion of her power plant from a sodium cooled to a pressurized water cooled reactor and was out of commission until 30 September 1960."

USS Nautilus SSN-571 at Wikipedia (http://en.wikipedia.org/wiki/USS_Nautilus_(SSN-571)):

"Nautilus was powered by the S2W naval reactor, a pressurized water reactor produced for the US navy by Westinghouse Electric Corporation."



Appreciate our position here: we have no proof that you are or have ever been a US Navy employee, but you want us to accept statements that contradict every piece of evidence we can actually find on the subject, and you do this based on authority that you cannot prove.

That's a pretty hard sell, and I'm sure we can agree that regardless of whether or not you are one of them, there are an awful lot of blowhards and Walter Mittys on the internet who claim to be lots of things that they are not and have never been.

Furthermore, you also want us to believe that you casually let slip some information on an online forum, information so classified that it still has not been published anywhere, even a half-century after the fact? If the information was that sensitive, I'm sure you would have been told at the time that blabbing it in a public forum was tantamount to treason, so what's the reason behind this rather bizarre choice of forum for your revelations? You could have written a book, or gone on the History Channel or something, after all.

Also, if it's a matter of national security, what purpose does it serve to confuse the reactor types in the public mind? It certainly doesn't provide denial of having either technology, or the date at which such technology was developed, or any other reasonable purpose I could see either.

I'm not trying to be difficult here, but there's an awful lot of big holes in this story.

[DonFaygo]

Its the internet, everybodies right!!!

[Baleful]

Its the internet, everybodies right!!!

:shakes:

[DonFaygo]

^^^ I thought it was funny

[MaddHawk]

I know its easy for someone to claim to be or have been something they never truly have. I looked up those same sources you did and saw what they said. I also admit I mixed up the first Seawolf with the current one. But, I still stand by my statement about the reactor technology used on the Nautilus. All I can say is in this case the public and civilian websites are wrong. Its that or they navy fed me false information while I was in the service. I also wish I knew of a way to provide solid proof I am who I say I am without exposing myself to undue risk of identity theft. :shrug:

[Extigy]

From silver water blocks to nuclear powered subs ? WTF?

This thread is really starting to stink of bulls s**t :rofl:

[MaddHawk]

Yeah, its gotten too :off:.

To help get it back on topic, I seem to recall hearing someone say that there has been solid silver water blocks made before? Also I have to wonder, just how big a slab of artificial diamond can be manufactured? If the blocks are big enough I know of a material that is harder than diamond that could probably be used to produce tools that could shape the diamond into a water block.

[cegras]

It's really impossible to manufacture a diamond that's very large. The crystal imperfections from the current synthetic methods are too imperfect.

Even vapor deposition yields an extremely polycrystalline sample.

[MaddHawk]

Even so, the poorest quality diamond was listed on Wikipedia to have a thermal transfer rating more than double silver. More like 2.5x that of copper at its worst. At its best its about 6 times that of copper.

Edit: Having participated in this thread, it seems to me we are more or less at the limit of what we can do material wise with liquid PC cooling technology. Unless, of coarse, we are willing to pay large sums for exotic materials, or until new materials are developed with a cheap means of manufacture.

[OggeOJ]

it's very "uncommon topic" .. there are ppl using "graphite" (aka pencil lead) as a form of thermal transfer since graphite is also another form of carbon ... just it's very highly electric conductive (might cause discharges)

in fact i'm reading some article if it's possible to use DIY pencil leads as TIMs ..

Cant you earth it? A cable from the block to the chassi?

[Acekiller]

Platinum is the best for conducting heat and then rapidly getting rid of it.
The is a better substance what is also a little bit cheaper but very hard to find ( they use it on Space shuttles)
Instead of getting mecury in Lc system rather get 10-20 kgs of platinum then a couple dozen fans and your done for the best cooling system ever.
It would only cost you about 362000 pounds or 559944.32 dollars.:)
But rather than the cost think of the performance.

[Zehnsucht]

Platinum is the best for conducting heat and then rapidly getting rid of it.
The is a better substance what is also a little bit cheaper but very hard to find ( they use it on Space shuttles)
Instead of getting mecury in Lc system rather get 10-20 kgs of platinum then a couple dozen fans and your done for the best cooling system ever.
It would only cost you about 362000 pounds or 559944.32 dollars.:)
But rather than the cost think of the performance.

Wow. You haven't paid any attention at all have you? :down:
Thermal Conductivity (higher is better)
Platinum: 71.6  W/mK
Copper: 401 W/mK


http://www.xtremesystems.org/Forums/showpost.php?p=3499474&postcount=99

[Waterlogged]

Platinum is the best for conducting heat and then rapidly getting rid of it.
The is a better substance what is also a little bit cheaper but very hard to find ( they use it on Space shuttles)
Instead of getting mecury in Lc system rather get 10-20 kgs of platinum then a couple dozen fans and your done for the best cooling system ever.
It would only cost you about 362000 pounds or 559944.32 dollars.:)
But rather than the cost think of the performance.

It wouldn't be much better because the system would still be dependent upon the air to remove that heat from the rad. If you put the rad into anything to help the transfer, you just moved into chiller territory. So again, the weak link as plain old simple air cooled water goes, is...and always will be, air through the rad.

[Acekiller]

Platinum can disperse heat into the air very quickly Not conduct the heat my BAD

[Acekiller]

http://en.wikipedia.org/wiki/Space_Shuttle_thermal_protection_system That has all you need to know about heat dispantion

[Zehnsucht]

/*/ Edit /*/

Nevermind. I'm leaving this thread.

[XS_Rich]

http://en.wikipedia.org/wiki/Space_Shuttle_thermal_protection_system That has all you need to know about heat dispantion

Even though it doesn't mention it once...

[m^2]

From your link to the wiki:


Sounds quite unscientifically to write "I re-measured it". Who remeasured it?

As wiki states, it's a quote from "Mr. Toshi Oyama, Ph.D. R&D Manager, WESGO Metals".

IIRC the story was like this:
In 2003 MCS (a company) (right name?) made blocks out of CuSil, claiming 515 W/mK. They weren't great, got beaten by White Water (?) clones. Later sb. discovered that the conductivity measurements were very old and there are newer ones, made by Mr. Toshi Oyama - showing 371 W/mK, less than copper. I guess that the wiki CuSil article is based on discussions about the block as I found the same quotation in the polish net. Not much information left though.

[[XC] riptide]

-snip-

Hey man. I read a brilliant book by a certain man called John Pina Craven. The Silent War. I expect you know who he is.

Anyways, he didn't expand in GREAT detail... becasue he couldn't as lots of stuff is still classified even going back to Polaris

Anyway... I'll invite you to take another look at yer posts maybe and edit if you feel you have to. You never know what may or may not be still classified. :up:

[MaddHawk]

Platinum can disperse heat into the air very quickly Not conduct the heat my BAD


http://en.wikipedia.org/wiki/Space_Shuttle_thermal_protection_system That has all you need to know about heat dispantion

The ability of metal to conduct heat has nothing to do with how well the air can absorb the heat. The article you linked doesn't even cover heat dispersal, thermal transfer, or any other aspect of thermal dynamics which relates to using water cooling loops to cool our PCs.

To get maximum heat absorption from the air we can do the following:
increase air flow, i.e. faster fan
increase the surface area of the radiator the air touches, i.e. greater fin density, larger fins
increase the ⌂T between the air and the radiator, i.e. turning on the house hold air conditioner and turning down the temp

[Locke]

In the future it will be possible to build a diamond large enough to be the matrix portion of the block. You wouldn't need to use a chip removal process, b/c the diamond could be grown with the proper pin matrix built in. It is technically possible now based on the research I have done, but it would be expensive beyond conventional means.