Liquid Cooling effects of liquids on CPU's

quitwwwww

Prominent
Aug 8, 2017
3
0
510
So for my science fair project I plan on testing different coolants to see which has the best impact on the CPU.

My Independent Variable is the liquid for the cooler

My dependent variable is the CPU Temp or if you have another good one please reply.
I will be using an all in one water cooler from corsair on a LGA 1151 mobo

My real question is I need help deciding on coolants, I plan on using distilled water for a control, a store coolant as one test, and a homemade coolant if there are any out there and if you could send me the algorithm.

I plan on entering for the ISEF under chemistry for the study of inorganic chemicals on liquid.

 
Solution


Technically, yes. Purer coolants tend to have a higher boiling point, like distilled water for instance. Here are some coolants you can try- 100% alcohol (methanol, aka Heet gas treatment, or denatured- ethyl alcohol you can't drink), mix that with distilled water in...

sirstinky

Distinguished
Aug 17, 2012
644
0
19,360
Generally one of the best mediums for thermal absorption is going to be water. Water is amazing because it has a pretty large delta for phase transition (solid/ice 32 F and below, liquid from 33-34 F to 212 F, gas/vapor 212 F) and can absorb large amounts of thermal energy (BTU's). Use distilled water to eliminate any contaminants like you are planning. In a closed system under some pressure, water can absorb even more energy before boiling.

You could also try regular antifreeze 50/50 but I don't know how good that will be for the cooler. You might want to try straight mineral oil as well, or that stuff you get for constipation. They use straight/petroleum mineral oil in heaters and transformers as an insulator and coolant. It transfers heat pretty well I think.

On the downside, I would imagine the more viscous your coolant gets, the less effective it will be since the pump can't circulate it efficiently through the heat exchanger and cooling block.

Good luck at ISEF!
 
1. the corsair cooler is factory sealed. it's far from trivial to replace the coolant in it. you'd be better to use Swiftech H220 X2
2. to keep the variables under control, read this : http://www.tomshardware.com/forum/id-1800828/intel-temperature-guide.html
for repetitive testing, disable power states, speed step etc in BIOS.
3. for results, use delta (difference between liquid and room temperatures )

P.S.
you are not going to see significant difference between different liquids. don't forget to do fun things like milk, soda, juices etc.
 

Olle P

Distinguished
Apr 7, 2010
720
61
19,090
That's the property used in heatpipes.
Energy capacitivity is what's used in regular water cooling. The difference in water temperature within the loop (that's before/after passing the CPU block) is typically a single degree or less.
Transformer oil typically has less than half the capacitivity of water and also less than 1/4 of water's heat conductivity.

Water is used simply because it's the very best option taken several factors into account.
The one disadvantage is that any impurities will lead to negative effects over time, and this is what's countered using additives such as glycol ("anti freeze", mostly to prevent galvanic corrosion) and/or biocide (to prevent growth of algae). A "wetter" to reduce the surface tension might also be used.
The "store coolant" is just water with said additives. It's the water that does most of the cooling.
Note also that all of the chemicals mentioned above are "organic" from a chemistry perspective!

Provided you don't get a build-up of particles on the surfaces inside the loop (and you should expect to get such a build-up) there should be no measurable difference (beyond the margin of error) in cooling efficiency between the various water based liquids such as distilled water, lemonade, fruit juice and pre-mixed coolant.
 

quitwwwww

Prominent
Aug 8, 2017
3
0
510
Thank you for answering! I haven't gotten the okay for my idea from my teacher yet but a question I have is that. Is there any homemade coolant out there I should try to make or should I stick with examining what chemicals absorb and expel heat the best from company coolants?

My hypothesis right now is that a more purer coolant will have better thermal energy transfer. I don't know the best way to explain that I'm working on it till my teacher gives me the okay.
 

sirstinky

Distinguished
Aug 17, 2012
644
0
19,360


Technically, yes. Purer coolants tend to have a higher boiling point, like distilled water for instance. Here are some coolants you can try- 100% alcohol (methanol, aka Heet gas treatment, or denatured- ethyl alcohol you can't drink), mix that with distilled water in various concentrations. You could use something like various oils (mineral, vegetable, etc), The other thing you could try is a saline solution, aka salt water. Do some research in 'vapor pressure' and you'll get an idea of how that works, although I don't know how useful it would be in a closed loop.
 
Solution

Olle P

Distinguished
Apr 7, 2010
720
61
19,090
If you go on with this project some advice are:
* Use parts you can take apart and clean inside. You'll need it!
* Go for cheap and easy to get household liquids: Tap water, milk, juice, oil(s), carbonated soda, ... (the most practical difference between them should be how much they clog up the cooling block).