Am I crazy? Silicone amorphous waterblock

[MadHacker]

Originally Posted by Butcher

Been done with oil (though you can;t submerge HDDs), water would be somewhat more risky as one little leak and zap.

how much more is water over oil in thermal conductiveness?
what i have read in the past about system submersion not including HD and stuff...
they use a liquid that is bad for the enviroment. and you don't realy want to breath it in either...

[Althornin]

Originally Posted by MadHacker

how much more is water over oil in thermal conductiveness?
what i have read in the past about system submersion not including HD and stuff...
they use a liquid that is bad for the enviroment. and you don't realy want to breath it in either...

you are thinking flourinert, maybe?
but you can use simple mineral oil....

[Butcher]

Water has roughly twice the specific heat capacity of transformer oil. Other oils will vary but I doubt you'll see much better.

[MadHacker]

Originally Posted by Butcher

Water has roughly twice the specific heat capacity of transformer oil. Other oils will vary but I doubt you'll see much better.

I imagine once the issues of shorting out the computer is fixed by sealing everything... why not use mercury... other then if it gets into your blood stream in any way... it is pretty safe... would it have a higher heat capacity?

Now higher heat capacity means it can absorb more heat or does it mean it can transfer heat better?

[Cptn. Foo Foo]

Originally Posted by MadHacker

I imagine once the issues of shorting out the computer is fixed by sealing everything... why not use mercury... other then if it gets into your blood stream in any way... it is pretty safe... would it have a higher heat capacity?

Now higher heat capacity means it can absorb more heat or does it mean it can transfer heat better?

Thats just scary.

[|kbn|]

Mercury isnt that good though. You are all getting confused by two different groups of numbers.
Conductivity = how fast heat can move through a object. Its best to thing of the object as having no flow, so not liquid, although liquids do diffinatly conduct heat.
Specific Heat = how much heat energy it takes to make the temperature of an object rise.

If you intend to pump a coolant or have a moving coolant in any way it needs to have a high specific heat (expeption is phasechange, i dont know enough about that) so that there is the most differnce in temp between the source and the coolant.
Conduction is most important in the block/heatsink becuase it will spread the heat to a larger surface area so it is easyer to remove using coolant.

relying comleatly on a coolant (direct die, submersed system or not) is not going to get good results without jets (to maximise the use of the coolant) and high flow rates. because there would be much less spreading of heat the jets would need to be aimied well at hotspots or some areas will not get cooled aswell and be hotter.

Moving the heat away is much more effective that waiting for it to be disipated by air the normal way becuase it has a large amount of coolant which can absorb a lot of heat, with just a heatsink (although air is still a coolant) you miss that layer.

Therefore there is no point in submerging systems except to cool components which do not have a high heat output, which are pointless in cooling anyway as the stock coolers shoudl manage...

Also with a submerged system there is a lot of coolant that is not moving (= wasted) with a high heat capacity. If a waterchiller/other chiller was used it would not be able to cool it becuase of the amount of heat that the coolant holds.

Mercury is not a relistic idea. Its poisnous, expensive, not availble in quantity (and i think yo need a licence or something to buy it) and iirc has rubbish conductivity. Also it would be almost imposible to pump around. This makes it bad becuase conductivity through it would not be faster than just pumping water at moving the heat. It would basicly act like one BIG insulated metal heatsink with one of the worst thermally conductive metals you can get...
Transfer of heat needs a thermal gradient too... it wouldnt have much thermal gradient as it would reach eqilibrium.

I expect Ive made lots of mistakes, Im not an engineer and I typed this very quickly.

[Butcher]

A table of specific heat capacities I made in another post:

Material   Heat Capacity (J/(cm³˚C))
Aluminum   2.43
Copper     3.4788
Iron       3.5433
Water      4.186
Mercury    0.7602
Hydrogen   0.001273056

They are per unit volume - as you can see water trashes pretty much everything.

Basically higher specific heat is good as it allows your coolant to carry more energy at a lower temperature - which reduces temps overall. For a given volume if you pump 100W in you're going to raise them temp of mercury a lot more than water. Additionally it's a lot harder to handle than water and probably harder to pump. Water really is about the best coolant known to man, if it were scarce people would probably buy it to put in cooling systems.

[nicozeg]

Even if this was working OK, I needed a final test to see if everything was like planned: The easy removal.

After about 10 days of use I decided it was enough, and proceeded to dismantle it in the name of progress.

First test was trying to take out the hoses, fortunately silicone has a strong adhesion over silicone hoses and was like one piece. The plastic top and sides were also very firm, so it could tolerate a lot of abuse without coming loose.

It was needed a sharp knife to take out the plastic parts, and then just pulling one corner was enough to take out the silicone. only a small bit of it remained inside the coils, that required a precision screwdriver to take out.

Now the mobo looks like new it's just a little oiled, but i don't care, another day I may clean it with some degreaser.

[bigben2k]

Still a valiant effort. If you ever come up with a new way of doing it, let us know, ok?

The main issue, as far as I can see, is using just enough "stuff" to prevent the part from overheating, since it's an isolator, and balancing that with the cooling solution.


BTW, in the military, they actually spray fluorinert onto electronic components, then recycle and cool the liquid...

[redleader]

Originally Posted by Incoherent

The implication being that any conductor of heat will also conduct electricity?
Not the case as I am sure you are aware.
Heat transfer in non-metallic solids can get pretty quantum mechanical, but suffice to say that movement of electrons is not the only way to shift heat. Although it is the most efficient way to do it.

Cheers

Incoherent

Heat conduction from electrons is quantum mechanical. The heat is transfered when the electrons jump between energy bands in atoms, all of which are governed by that nasty wave equation no one can seem to solve. Jump up a level and you decrease the thermal energy of that atom (or at least you decrease the thermal energy somewhere nearby). Jump down and you increase it.

Electrical insulators that conduct heat do exist, however the mechanism is quite a bit easier to understand, basically tightly bound atoms vibrate in sync with each other. Thus one you heat one you effectively heat the other atoms connected to it and the energy passes very quickly as a phonon vibration or something like that.

At least those are the two types of thermal conduction mechanisms I'm familar with. There may be others.

[dima y]

that is Abit NF7-S right? can you measure the hole distance for the northbridge center to center? approximatly? please pretty please?

[satanicoo]

Originally Posted by dima y

that is Abit NF7-S right? can you measure the hole distance for the northbridge center to center? approximatly? please pretty please?

58.5 milimeters.

[dima y]

thank you very much