Oil Cooling 2012

Domi

Domi

Domi

Associate
Joined
6 Sep 2008
Posts
790
Ok who has tried and how well does it work ?

It seems to me that the current generation of top end hardware is not as power hungry as before and that this just might be an uber silent option.
Most of the cost of this option frankly is in the pump arrangement could it be done without.

Like water cooling Oil cooled still seems to need a flow and so a pump and probably fans, is anybody expert enough to know if this is actually true ?

Could for example a mineral Oil bath cool 350-400w WITHOUT a flow/pump/fans ?
 
Comparing water and oil, not only will heat spread through the oil more slowly (which is important in a water block, as it means the oil further away from the fins will not absorb as much energy), but also, the oil will absorb less energy (and thus cool the block down less). In other words, water wins. For same reasons oil is slower releasing heat energy in the radiator. Again, water wins.
 
Just to clarify Oil cooling would be a submerged system not simply a watercooled system with water replaced with oil.

Without a pump we are talking a sealed system for example in a glass case, my theory is using glass we get good heat convection from the oil to the surrounding air ?

The question is how hot the oil gets , does it convect enough heat away from the components and does it need circulation.

Would even a large fan rotating within the liquid work to cool the oil.
 
Glass has rather unimpressive heat convection and the system really needs decent circulation. Normal PC fans do not perform great under Oil, but I reckon a powerful industrial fan like a Delta or San Ace could be used. The Oil would also greatly muffle the sound.

However removing the heat from the oil is not easy. Some people have used a pump inside the oil to push oil into an external radiator that removes the heat via standard fans. However the pump needs to be strong to push oil through the tubes.
 
So maybe rather than the fish tank type method peeps use, some sort of enclosed metal case.
Ideally a sealed except from the top metal case where the case itself is the radiator with a large fan in the oil giving circulation.

I am of course no expert and no engineer but it just seems like this would work.

Obviously the bit I am not understanding must be just how quickly the oil heats up and its lack of heat dissapation but surely its possible to passively cool 300w over a quite large surface area.

It clearly can work well this is becoming quite common in industrial use I was just hoping we were nearer for more mainstream enthusiasts.
 
You need to move the oil through a radiator to get surface area to cool it. Oil is much harder to work with then water. You will need a radiator designed for oil, and a powerful enough pump so as oil cools it doesn't hang onto the surface thereby insulating surface from hot oil needing cooling. I wouldn't even consider trying oil cooling.
 
Last edited:
I have always liked the idea of this, but I wouldn't like to have to deal with the mess! There is also the problem of cooling the oil, as the temperature will increase steadily it is is not cooled appropriately.
 
Just for a little info... I tried mine to see how it would do passive before i added my gfx cards in.

(It was water on a 2600k at the time with n XSPC raystorm block)

A 10 gallon ish res with pond pump and no radiator
The water started at 19.8c, after 1hr of world of tanks (not demanding) the water was upto 28c, after 4hrs it was 45c and held at that, however at this time the cpu was running around 55c @ 4.8ghz.

One simple 360 rad held the water in the same rig with 2x GTX480 added in! at 25c.
The bmw 318is rad i use now hold the water at ambient pretty much ~21c it says now (that is with a 3930k and the 480's and no fan!)

To summarise, water struggles to passively radiate enough heat (2sq foot of open surface) for a 2600k, im not sure how oil would handle it.
 
You must log in or register to reply here.
Back
Top Bottom