8

I was wondering why specific coolants or distilled water is used as a heat transfer liquid for engine cooling instead of just using a separate oil circuit with some type of oil as a heat transfer liquid.

Cost issues aside I guess there is (could be created) an oil type available with similar properties (viscosity, heat transfer, heat capacity) to distilled water or coolant mixtures. Even if it would be necessary to re-design some engine elements (radiator tubes, coolant pumps etc) to cope for the differences in liquids properties wouldn't it be an advantage to have oils mixed in case of gasket failure and also if these two had the same properties (and since coolant liquid doesn't get extremely hot with use as to deteriorate) you could use the present "cooling" oil in engine oil changes.

What might be some preventing issues with this idea?

kokobill
  • 1,005
  • 1
  • 9
  • 28
  • 2
    Possible duplicate of Oil Cooled vs Water Cooled. Motorcycles do have oil coolers. But with noise emission regulations, a water jacket around the engine helps to muffle the sound, and more recent bikes have water cooling. – Weather Vane Nov 14 '19 at 10:11
  • @WeatherVane many cars have engine oil coolers and gearbox oil coolers, (there are even axle oil coolers), as a support to the standard cooling system - don't think the OP meant that. – Solar Mike Nov 14 '19 at 10:46
  • If you think an oil change is expensive now, can you imagine if you had to change the 5+ qts required to lubricate your engine and the 6-8+ qts required to cool it?!??!? – FreeMan Nov 14 '19 at 21:00
  • Most folks just use water or water with antifreeze in their motors. Mineral water often has too many ... minerals... in it to be really useful. Besides which, most places you can only buy mineral water in bottles for stupidly high prices (compared to the water from the taps in your house.] – JRE Nov 15 '19 at 08:26
  • Add in a factor for everything with "heat" .. water won't ignite or burn ... – eagle275 Nov 15 '19 at 14:42
  • 1
    While the world has, for the most part, avoided oil cooling in favor of water cooling, the one notable exception I'm aware of is the SACS line of engines produced by Suzuki . For many years, I owned a Katana 600 with one of these in it, and it was a very capable and reliable motorcycle. – dgnuff Nov 15 '19 at 20:40

3 Answers3

23

There's a very straightforward answer to this question: Water has a much, much higher capability to transfer heat than oil.

Even if you were able to overcome the safety and design issues brought up in other answers, you'd still have to overcome the fact that water has a huge performance advantage at transferring heat compared to other common fluids (including oil). Water is essentially the best possible fluid in terms of thermal conductivity; the only fluids that beat it are metals in liquid state (which isn't a practical choice for cooling a car engine made from metal itself). At 75 C, water's thermal conductivity is around .65 W/mK. Typical heat transfer oils are usually around .14 W/mK.

Of course, coolant in a typical automobile isn't water, it's usually water mixed with alcohol. Typical mixes cause the conductivity to drop, because water also beats alcohol at heat transfer, but even pure ethylene glycol is .27 W/mK. A typical 50/50 mix puts you in the neighborhood of .45 W/mK, which means that typical water-based coolant has a thermal conductivity three times better than purpose-designed heat transfer oil.

Let that sink in. If you wanted to cool your car's engine with heat transfer oil instead of typical water-based coolant, you're operating at a three to one disadvantage in terms of thermal conductivity. This basically means that your coolant system needs to be three times bigger in order to have the same capacity. Radiator size is already somewhat of a limiting factor in terms of how the front end of a vehicle is designed. I don't think anyone wants to drive a car where the radiator needs to be three times bigger than a typical car radiator is today.

Oil is really only used for thermal transfer when water is impractical - due to corrosion, conductivity, evaporation, pressure or temperature requirements above what water can handle, or other limiting factors. As pointed out in other questions on mechanics.SE related to air vs water cooling, the transition to water cooled engines was basically because of this fact. An "air cooled" engine - which typically relies not only on air cooling, but also oil cooling - just can't compete. For a given application, if relying on oil or air cooling, you have to really limit your thermal load. In other words, for a given package size, it's easier to get more performance and efficiency with water cooling than with oil and/or air cooling.

dwizum
  • 1,134
  • 6
  • 11
  • There are some high output air cooled engines still in use, see radial engines... – Solar Mike Nov 14 '19 at 15:37
  • 3
    And as to size of radiator being *3 why not increase the flow rate instead... much easier... – Solar Mike Nov 14 '19 at 15:39
  • 4
    why not increase the flow rate instead... much easier - you answered that in your own answer - pumping losses. Pushing viscous oil through a radiator at three times the rate you're able to pump water through it would require a gigantic and lossy pump. Switching from oil to water is "free" in terms of total energy budget and gets you 3 times the performance. – dwizum Nov 14 '19 at 16:00
  • 1
    And I'm not sure I know of any modern cars with air cooled radial engines in them? I didn't think typical applications for radial engines were on topic for this question. – dwizum Nov 14 '19 at 16:01
  • And 3 times the drag for the larger radiator - a pump with a hugher flow rate is only a few more hp... – Solar Mike Nov 14 '19 at 16:01
  • You referenced mechanics.SE... so those engines are valid... – Solar Mike Nov 14 '19 at 16:02
  • a pump with a hugher flow rate is only a few more hp I'm not sure why you're making comments here that contradict your own answer. – dwizum Nov 14 '19 at 16:02
  • 1
    You referenced mechanics.SE... so those engines are valid. I'm just trying to answer the question which was asked, which was about typical automotive engines. Maybe we should move this to chat, or ask a separate question, if you want to talk about other types of engines. – dwizum Nov 14 '19 at 16:03
  • why not use ammonia for the fluid - much better at heat transfer than water... – Solar Mike Nov 14 '19 at 16:10
  • 2
    At 75 C, ammonia is around .35 W/mK. Worse than pure water and typical water/alcohol mixes. And, ammonia arguably has safety and practicality issues compared to water. But I didn't see ammonia in the question, and don't see why it's relevant to answering the question. – dwizum Nov 14 '19 at 16:16
  • 6
    Note that thermal conductivity is not the only relevant number that's important for a coolant. Density, viscosity, and specific heat capacity also play a role. According to an article I found, what you really care about is the "Mouromtseff number". (Water still wins, as it turns out, but its entirely possible for something with a higher thermal conductivity to be a worse coolant if its other properties are worse). – mbrig Nov 14 '19 at 19:42
  • 1
    It seems a good cooling fluid would be mercury. High working temperature range, excellent thermal conductivity (15 times that of water), low viscosity, etc. The "Mouromtseff number" I calculated for it is 10 times what I got for water. Just design your engine to use iron cooling channels, and non-reactive plumbing. The radiator can't use aluminum or copper tubing, but I am sure that some substitute materials can be found. – Buddha Buck Nov 14 '19 at 21:22
  • Most engines use water and ethylene glycol. Alcohol can be used as an antifreeze, but is far less common than the ethylene glycol based antifreeze. – JRE Nov 15 '19 at 08:28
  • @BuddhaBuck Mercury's about 13 times more dense than water, greatly increasing the energy needed to pump it through the system and the weight of the engine. It also has much higher surface tension. And there's the dramatically higher cost. And the fact that it's quite toxic. – barbecue Nov 15 '19 at 22:13
5

Water has the nice property that it boils at a fairly low temperature, so the vapour pressure rises quickly with temperature and it is easy to design fail-safe over-pressure protection (a simple spring-loaded filling cap). Also, water is not flammable.

Oil does not have those nice properties, and a cooling system failure which released high pressure, high temperature oil which might be above its auto-ignition temperature would not be a good idea.

Also flexible hoses (and their connectors) to contain high pressure and temperature oil would be more problematic than water hoses, so you might have to redesign the car radiator to be rigidly connected to the engine block, and also solve the problem of how to transfer heat to the air conditioning system.

Of course oil is already used for engine cooling where that is combined with the need for lubrication, e.g. bearings, cam followers, etc.

alephzero
  • 2,944
  • 1
  • 10
  • 17
  • 1
    Pressure control systems exist to deal with the issues you describe and there are some systems where oil has been used at higher temperatures than any found in engine cooling systems - solar heating systems - one obvious point was that the oil was designed to not catch light when hot in case of accidental release... but the cost compared to water and other considerations meant it went out of fashion... – Solar Mike Nov 14 '19 at 12:49
3

Oil has been used as a heat transfer fluid in many applications (especially high temperature ones) but in engines the higher viscosity will require more pumping power, which means more driving power from the belt and so more power from the engine.

Then there are the issues of cost... and, of course, issues with replacing fluid if you are far from an auto parts store.

Solar Mike
  • 34,563
  • 2
  • 28
  • 58