LTT cooling idea

[Supersonicwolfe]

Liquid metal is good at heat conducting right? I was thinking why not just use that as medium in a loop.

 

Or use long heat pipes sticking directly into open air with fins and fans Mad MAX style?

 

This sounds like a LTT idea to me

[Neo-revo]

the idea has been discussed

 the possible benefits far outweighed by the risks/ problems associated,  even with Gallistan being a relatively inert non-poisonous liquid metal

[Neo-revo]

relatively inert compared to say mercury, or NaK

[Supersonicwolfe]

9 minutes ago, Neo-revo said:

the idea has been discussed

 the possible benefits far outweighed by the risks/ problems associated,  even with Gallistan being a relatively inert non-poisonous liquid metal

LTT isn't known for thinking safety first   . But I know it is a radical idea, the filled system won't be cheap, nor it will be portable, probably going to weight good 100-200KG

I did consider toxicity but in a sealed system it shouldn't be a big issue, gallium is kind of the best choice of the 3 you mentioned, however gallium will also destroy some metal it contacts, especially aluminum. Mercury is chemically safe, but biologically dangerous, and NaK will just explode when in contact with water...

 

Also I think Linus did feature a fan-less passive system with massive heat pipe and radiators at a show, ECS or something.

[Neo-revo]

its the same chemically active as galistan it can almagam certain metals. and you mention the obv risk/problems. a cubic cm of gallistan weighs 6.24 grams.

 

my current loop used about 1500cc so  if i could just replace it and do this,  (need a good pumps, redo all seals, remove certain fittings, upgrade to epdm or neoprene orings.... )

 the fluid would weigh around 9.5 kilo's

[For Science!]

19 minutes ago, Supersonicwolfe said:

-

and not to mention that it probably won't improve the temperatures that much anyway since the waterblock - fluid - radiator transfer is not limiting the system. It is the radiator - air interface that is most limiting for a typical watercooled loop (which is still air cooled).

 

This is why changing from propylene glycol, ethylene glycol, to water, makes no impact on performance.

[Neo-revo]

11 minutes ago, For Science! said:

propylene glycol, ethylene glycol,

only really necessary if you are getting extreme hot water temps, putting too many parts on a tiny loop/rad.     Or happen to be messing with subzero

[For Science!]

1 minute ago, Neo-revo said:

only really necessary if you are getting extreme hot water temps, putting too many parts on a tiny loop/rad.     Or happen to be messing with subzero

But this is what I'm getting to. Take for example a typical loop using EK-CryoFueld (water based) or just distilled water. You then change the fluid to something like Mayhems X1 (EG based) and you will see virtually no difference in temperatures.

 

This is despite going from water to a lower thermally conductive coolant (Ethylene Glycol), this shows clearly that in a typical water loop, the thermal conductivity of the coolant is not rate limiting. Therefore increasing the thermal conductivity of the fluid will not have a significant impact either, since you're just improving a non-bottleneck.

 

If on the otherhand you were dunking the radiator in water (watercooled - watercooled PC), then since the radiator-air bottleneck is alleviated, it would make more sense to look into this. Many people have been seen to be dunking their radiators in ice-cold water, if this were something like liquid metal instead, the transfer would indeed be much more efficient, however since the specific heat capactiy is not very high, it wouldn't last very long.

[Neo-revo]

3 minutes ago, For Science! said:

heat capactiy is not very high, it wouldn't last very long.

you could be Ozzy O this last summer and order what was it 2 tons of ice to fill the out door pool one day on whim. and get at least a few hours  gaming in