Oil filled water cooling loop?

[cadaveca]

I'm sure most AIO coolers have distilled water and a antimicrobial. That's it. Its why they only last and couple of years is my guess. Why would you spend more money just to have it last longer

Glycol/water mixes in every single one. Glycol is the one additive that help water's thermal conductivity at the termps we use, and also prevents the AIO's from popping under expansion while shipping (or you'd never get one in the mail from Newegg).

 

[RejZoR]

I remember people using water and antifreeze for cars when they made custom loops using heating radiator from Lada and Yugo cars as a heatsink rad Not sure if anyone still bothers with those since dedicated water cooling accessories are so widely accessible now.

 

[Brusfantomet]

Have been on the lookout for the optimal water cooling cooling medium for some time now, as i see it there is three chemical properties that are important:
Heat capacity
Viscosity
Electrical conductivity

Heat capacity is important because the better it is the better cooling one unit (L/Gal/mL wathever) of the fluid passing through the block will grab more heat from the block.
Viscosity is important because it denotes how difficult it is to force the fluid though the loop, so lower Viscosity better flow.
Electrical conductivity is mostly a nice to have, as it only makes the fluid NOT kill your computer should you get a spill.

In addition, a fluid that is not reactive to the other metals and plastics of the loop, and no thigns can live in (algae)

Something with low viscosity and high heat capacity that wont kill you computer should you spill and is stable would be ideal

Now for the fluids i have found:

Water:
Very high Heat capacity for a fluid: 4.1813 [J/(g*K)]
Very Low Viscosity: 0.894 [CP]
But it is conductive and algae likes to live in it

Propylene glycol: its the stuff in engine coolant additives
Heat capacity: 2.521 [J/(g*K)]
Viscosity: 42 [CP]
Still conductive but nothing lives in it, and it has a lower freezing point so its possible to use radiators outside in winter (if you live in a part of the world that gets that kind of temperatures. Also, Here is some interesting stats about water/glycol solutions.

Mineral oil: there is a lot of different oils, but they are high Viscosity.
Heat capacity: 2.521 [J/(g*K)]
Viscosity: 35 to 400 [CP]
Not conductive, i dont know if stuff grows in it tho.

There are probably other fluids but the Heat capacity and Viscosity will be the most important ones comparing it to water.

 

[silentbogo]

In the regular LC loop there will be no reason to use oil (probably less effective than aftermarket air cooling), but a few months ago I saw a video on Youtube of 2 kids testing mineral oil vs. vegetable oil in a tank. Maybe not the best test, but the proof of concept definitely works.

EDIT: here's the video

 

[MxPhenom 216]

I watched this yesterday:

And it made me wonder. Milk would turn into yogurt because of proteins in it and acidity from it would eat the metal parts, orange juice would also corrode metal stuff, Powerade would eventually grow algae etc.

But then it struck me. What about motor oil? A very thin (very liquidy one)?

- it is made to be used in hot conditions
- is not corrosive
- won't allow algae to grow
- won't go bad like ever

Seems like a perfect candidate. There might be some concerns regarding pump and maybe rubber gaskets here and there, but in theory, it should work. I'd love to see how it performs thermally as I don't know how oil performs as heat transportation medium.

.there are systems that are fully submerged with oil, not just flowing through tubes.

 

[peche]

the difference between using just distilled water and silver coils in tubing vs pre-mixed coolant is not even noticeable and using that distilled water and the solver coil will make the loop almost and AIO that will just need to have some cleanliness on the rad...
so why getting another fluids on the loop? and a ultra pain in the ass with the ood results you might get... distilled water and silver dice are cheap enough to get... also coolants are pretty cheap...

.there are systems that are fully submerged with oil, not just flowing through tubes.

correct... oil chilled systems... well the biggest disadvantage there is that everything you'll submerge there is almost lost... [cleaning all back for a future sale ...nahh not worth at all ]

 

[erocker]

Why not the most simple form of liquids... Water? Automotive coolants are designed to run through various metals and operating temperatures outside that of a computer. Other oils just put unneeded stress on the pump. Computer water pumps are designed to operate best using water.

 

[peche]

Other oils just put unneeded stress on the pump.

f*cking waterbumps....

Computer water pumps are designed to operate best using water.

correct... that's why they are named "WaterPumps"

thanks

 

[Loosenut]

Distilled water and some Redline Water Wetter. I use a 5:1 ratio

Quote:
"Red Line Water Wetter provides corrosion and rust protection for modern aluminum and cast iron cooling systems. This unique additive reduces coolant temperatures by as much as 20 degrees F. It can be used in plain water to provide much better heat transfer properties and protection than glycol-based antifreeze, or added to new or used antifreeze to fortify inhibitors and reduce foaming."

http://www.summitracing.com/int/parts/red-80204/overview/

 

[RejZoR]

People apparently think only oil is a thick oil. Apparently some don't know that oils can be almost as thin as water. Not identical, but so close it's almost hard to tell.

 

[kenkickr]

Back in the day when I had all the time in the world and was building my water cooling systems from car heat exchangers I used to use Water Wetter.

 

[thebluebumblebee]

People apparently think only oil is a thick oil. Apparently some don't know that oils can be almost as thin as water. Not identical, but so close it's almost hard to tell.

I don't think you understand how truly special water is.
While @Brusfantomet mentioned water's ability to carry heat, what was not mentioned was water's high heat transfer rate. Those 2 facts, and that water is cheap and nonreactive, is the reason that water is what is normally used for cooling.
Also, water is easy to pump, normally with (easy to implement) impellers. Oil normally needs a positive displacement pump, which is MUCH more difficult to implement in a closed system. (this barely scratches the surface on this point)

 

[xorbe]

for what is worth super high voltage 240K underground power cables are oil filled.. using oil as a coolant aint a new idea.. he he

I think this is primarily because (1) water evaporates and sneaks out much more easily, (2) water boils at 100C, and (3) rust.

Car coolant is for the anti-freeze properties -- it doesn't cool as well, actually.

Regarding that Redline Water Wetter (a surfactant) a very tiny dash of baby shampoo should give a similar effect in water. Or HE detergent (anti-suds). I don't know how the water pumps will react to "finer" water though.

 

[lilhasselhoffer]

Have been on the lookout for the optimal water cooling cooling medium for some time now, as i see it there is three chemical properties that are important:
Heat capacity
Viscosity
Electrical conductivity

Heat capacity is important because the better it is the better cooling one unit (L/Gal/mL wathever) of the fluid passing through the block will grab more heat from the block.
Viscosity is important because it denotes how difficult it is to force the fluid though the loop, so lower Viscosity better flow.
Electrical conductivity is mostly a nice to have, as it only makes the fluid NOT kill your computer should you get a spill.

In addition, a fluid that is not reactive to the other metals and plastics of the loop, and no thigns can live in (algae)

Something with low viscosity and high heat capacity that wont kill you computer should you spill and is stable would be ideal

Now for the fluids i have found:

Water:
Very high Heat capacity for a fluid: 4.1813 [J/(g*K)]
Very Low Viscosity: 0.894 [CP]
But it is conductive and algae likes to live in it

Propylene glycol: its the stuff in engine coolant additives
Heat capacity: 2.521 [J/(g*K)]
Viscosity: 42 [CP]
Still conductive but nothing lives in it, and it has a lower freezing point so its possible to use radiators outside in winter (if you live in a part of the world that gets that kind of temperatures. Also, Here is some interesting stats about water/glycol solutions.

Mineral oil: there is a lot of different oils, but they are high Viscosity.
Heat capacity: 2.521 [J/(g*K)]
Viscosity: 35 to 400 [CP]
Not conductive, i dont know if stuff grows in it tho.

There are probably other fluids but the Heat capacity and Viscosity will be the most important ones comparing it to water.

Right track, only a few problems.

1) Water is not conductive. The impurities commonly found in water (and those introduced whenever a metal oxidizes and partially dissolves) conduct electricity. Don't believe me? Get a glass of tap water, salt, and a digital multimeter. As more salt is dissolved into the water you get less resistance.

2) Viscosity isn't really a factor in PC water cooling. The flow rates are comparatively low, which means that the energy required to move the fluids is negligible. Viscosity is only of concern when you've got tiny tubes, and high flow rates. While standard (and cheap) centrifugal pumps would have a harder time with oil, it's not a deal breaker.

3) Everyone's missed the really fundamental part. The difference between inlet and outlet temperatures on radiators varies very minimally (1-2 degrees). As such, oils generally can't heat up enough to actually have a significant difference between their temperature and surrounding air. If the delta between ambient and fluid is very small no heat transfer occurs. This is why we don't use oils.

4) Glycol is an alcohol. Glycol has the tendency to bond to the polarized section of water, and make it "slippery." This is why it's used with water (if the water is less polarized it tends to oxidize less), but not on its own. On its own glycol is another hydrocarbon with relatively poor conductivity.



To all of those people who are commenting about water being special, you're about 80% of the way there. It's a unique substance, in that similar chemicals have a vaporization point so much lower that it isn't funny. Water has a huge heat capacity, it's relatively abundant, and the cost is very low. Oils are used in things like power transformers because the delta between operational and ambient temperature would vaporize water. If you ever get a chance run an old school transformer for some time, and just feel how hot it gets. As industrial transformers are several times larger than than, it isn't hard to imagine 200-300 F temperatures. This is why oil is used. Computers have a delta that is comparatively so small as to be non-existent, which is why we use water.



TL;DR:
Water isn't an arbitrary choice. Water is demonstrably the best option available for PC cooling. We add things to water that make it a worse cooling fluid because it's pretty much the only way chemicals reactions occur (read: aqueous solutions), and water is a great oxidizer. The chemicals we add are uniformly less efficient coolers (in this instance), but their improvements (namely biocidal, aesthetic, or stability) outweigh the substantial costs. While liquid metals do technically have better properties, it's difficult to justify the slight improvements while looking at the smoking slag heap that was your computer.

I think this is primarily because (1) water evaporates and sneaks out much more easily, (2) water boils at 100C, and (3) rust.

Car coolant is for the anti-freeze properties -- it doesn't cool as well, actually.

Regarding that Redline Water Wetter (a surfactant) a very tiny dash of baby shampoo should give a similar effect in water. Or HE detergent (anti-suds). I don't know how the water pumps will react to "finer" water though.

Not quite (to the oil).

The reason that oil is used in high energy lines is that even tiny resistance produce huge amounts of heat. Between that, and the fact that oil is completely non-conductive, oil is used for safety reasons.

Glycol and surfactants are one in the same. Molecules with one polar end and a relatively non-polar end allow water to "stick" to impurities but remain in solution.

 

[xorbe]

The reason that oil is used in high energy lines is that even tiny resistance produce huge amounts of heat. Between that, and the fact that oil is completely non-conductive, oil is used for safety reasons.

Ah very good point.

 

[cadaveca]

3) Everyone's missed the really fundamental part. The difference between inlet and outlet temperatures on radiators varies very minimally (1-2 degrees). As such, oils generally can't heat up enough to actually have a significant difference between their temperature and surrounding air. If the delta between ambient and fluid is very small no heat transfer occurs. This is why we don't use oils.

inlet/outlet temps with only 1-2 degrees delta is based on a limited amount of water, a specific heat load, water speed, and specific radiator designs. This is NOT a rule of any sort, merely something skinnee measured and verified within his own loop. I can easily show you differently (I asked my teacher how to do this properly, even). Heating and cooling is my specialty (HVAC).

 

[OneMoar]

oil would likely make the rubber seals swell up and fail ...
distilled water/silver coil and some glycol are the only things that should be in a loop if you care about performance

 

[cadaveca]

oil would likely make the rubber seals swell up and fail ...
distilled water/silver coil and some glycol are the only things that should be in a loop if you care about performance

flourinert begs to differ. cost is prohibitive, but oh well. No different than spending $1300 on a CPU, or buying 4 VGAs for one system, or even water-cooling! ROFL.

 

[OneMoar]

flourinert begs to differ. cost is prohibitive, but oh well. No different than spending $1300 on a CPU, or buying 4 VGAs for one system, or even water-cooling! ROFL.

lol iv always wanted to swap out water for a 3m fluid in a loop and see what happens
I think 3M Novec 7100 has a boiling point of like 60C

 

[thebluebumblebee]

The reason that oil is used in high energy lines is that even tiny resistance produce huge amounts of heat.

And when things don't go quite right: (about 100 yards from my house)

 

[xvi]

Glycol/water mixes in every single one. Glycol is the one additive that help water's thermal conductivity at the termps we use, and also prevents the AIO's from popping under expansion while shipping (or you'd never get one in the mail from Newegg).

Distilled water and some Redline Water Wetter. I use a 5:1 ratio

I picked up some water wetter from Walmart for pretty cheap (basically just glycol in a bottle). I ran that in my loop for a while. When I first added it, it managed to find and remove a lot of small bubbles trapped in my loop almost immediately. Worked better than dish soap and gave my water a nice green tint.

I think @Norton has some experience running a bit of radiator fluid in his loops. Care to weigh in, sir?

 

[Norton]

I think @Norton has some experience running a bit of radiator fluid in his loops. Care to weigh in, sir?

Wasn't me? I use straight distilled water in mine and have had no issues thus far.

I'm not against using a bit (<5%) of radiator fluid or methanol in a loop since just that will do a good job inhibiting bacterial growth w/o significantly affecting the performance of just water.

 

[lilhasselhoffer]

inlet/outlet temps with only 1-2 degrees delta is based on a limited amount of water, a specific heat load, water speed, and specific radiator designs. This is NOT a rule of any sort, merely something skinnee measured and verified within his own loop. I can easily show you differently (I asked my teacher how to do this properly, even). Heating and cooling is my specialty (HVAC).

Are you really asking to rehash a discussion in which I already proved my point?: http://www.techpowerup.com/forums/threads/water-cooling-questions.216240/

Jesus, you were the one questioning my units, even after I proved everything out using the only sensible measuring system (and I'm saying this to a Canadian, as someone from the US; if I didn't know better I'd say I was drunk).

I'll ask out of courtesy again, but if you really want to start the math there I'm more than happy to copy and paste.


Respectfully, I trust measurements and a year and a half worth of college classes more than rules of thumb from HVAC. That little blue book is awesome for getting the ballpark figures for gasses (read: stuff people measure frequently but are compressible and very dependent upon atmosphere), but I've yet to see that little blue book produce consistently accurate results (within 5%) for complex situations. If you'd like to change conditions you can prove anything, but it's stupid. Nobody has a lake to cool their system, nobody expects a thimble of water to cool their system (except in phase change cooling, but that's another story), and changing the system specifications is stupid. I proved that a single degree of temperature across a radiator, dissipates 250 Watts of thermal energy. Even a socket 2011 processor (140 Watts) and a trio of 390x GPUs (275 Watts each) would only produce a change in temperature of 4 degrees C (965/250) at most.


Can we please not rehash this?

 

[xvi]

Wasn't me?

Huh. Must have been a different cruncher.

 

[cadaveca]

Can we please not rehash this?

When you stop posting bullshit, I'll stop refuting it. It takes how many watts to raise 1 lb of water one degree? This is science!

EDIT:

to clarify, a watt is a measurement of the rate of energy, not the amount. This is why your power bill reflects kwh, not kw. The rate of energy (kw), over a specified time(h).