Why water for coolant?

[Bibio]

why do we fill radiators with water and antifreeze when we could fill it with oil?

ok apart from the system springing a leek i cant see a problem. but i know some smart person will tell me why.

[Drumming_Animal]

expense?
waste on the worlds resources of oil?

to be honest, no idea..

[embee]

Well, in very simple terms, the coolant is for cooling hot bits of the engine. In order to do that it needs to absorb heat well, and flow round the engine easily, and give up the heat via the radiator.

Water has a very high specific heat capacity (the amount of energy required to heat a given mass up by a given temperature difference), pure water is around 4.2kJ/kg.K , typical oils are only around half this, give or take, so you'd need to flow something like twice as much in order to remove the heat with the same temperature rise.

Water also has a quite useful property in that it absorbs a lot of heat when it boils, so if you get a local hotspot it can cool it very effectively by boiling at that spot but as it circulates the steam condenses again in the rest of the liquid.

Water is pretty runny (until it freezes), so pumps easily and flows into all the nooks and crannies well, plus the radiator can have small bore pipes to improve the heat transfer. The viscosity of water doesn't tend to change as much as oily liquids either when it gets cold. OK so some oily type stuff is runny too (paraffin etc), but generally speaking it's not as runny as water, and if it wasn't the engine oil what would be the advantage over water based coolant?

Water does have a habit of freezing and also tends to corrode stuff, but fortunately cheap and cheerful (relatively) additives can be mixed in to address both these issues.

Some engines have used oil cooling as a strategy ( GSXR (http://en.wikipedia.org/wiki/Suzuki_Advanced_Cooling_System) ) in conjunction with air cooling, so it can be done but has it's limitations.

[squirrel_hunter]

Excellent informative answer, thanks.

[MattCollins]

Okay, not oil, but it is waterless. I hope a change in direction is not going to upset anyone. I figure Embee's response was fairly comprehensive.

http://www.evanscooling.com/
http://motorcyclecoolant.com.au/Home_Page.php

I am using the latter in the TE630 and it is doing the job despite a lower specific heat value. Claims are made that this is offset by the lower vapour build up in the system. I don't know.
The advantage for me is that it has a very high boiling point and operates at near zero pressure, which makes it very hard to blow the coolant out of the system or to blow a core.

I'd be interested hear what others have to say about these type of coolants.

[embee]

I haven't gone through those links thoroughly, but a glance says they are talking about propylene glycol, as opposed to ethylene glycol which is a main constituent of conventional coolant/antifreeze.

Many moons ago I did do a trial exercise with a supplier based in Connecticut who were trying to sell this idea. They had a car from us and we did some work on it.

A big selling point for propylene glycol is that it isn't toxic (it's used in the food industry I think). I don't know how big a problem ethylene glycol toxicity is in real life?

It does work as a coolant. Whether some of the arguments about boiling point and nucleate boiling are valid is open to debate, as I suggested in the earlier post nucleate boiling or "strategic boiling" can actually be pretty useful sometimes, the latent heat of vapourisation of water means that a lot of heat can be absorbed without increasing the metal temperature much whereas using conductive cooling the temp increases significantly more for the same heat transfer.

The metals used in traditional engine construction have temperature limitations, and the effects on combustion/performance/lubrication need to be borne in mind too. Water/EG coolant matches the requirements quite well. You normally aim to keep the delta-T (coolant temp rise across the engine) in the range of 7-10degC and design the system accordingly, so pumps and rads will be sized for this. If you do use a coolant with a significantly different specific heat capacity it might call for some mods.

We never really arrived at any good reason to use propylene glycol instead of water/ethylene glycol in production vehicles, but I wouldn't say don't use it if it suits your application. It does allow an effectively un-pressurised system if that's important. As far as I recall there aren't any significant material compatibility issues (hoses/seals etc).

[MattCollins]

Thanks for that.

Metal temperature increases have been the concern. It has been difficult to discern whether there has been any increase in logged CHT which is about as close as I can get to measuring metal temperatures. There has been what appears to be a slight increase in logged peak coolant temperatures (still within limits), but I don't have a control for valid comparison.
I estimated that there was a need for around 15% greater coolant flow, so I fitted a higher volume water pump (by how much I have no idea) from the Husky performance parts catalogue. The radiators seem to be up to the task for all but the worst conditions and could probably benefit from a small increase in volume or area.
Oil temperature spikes atrociously in heavy conditions, is slow to decrease and corresponds with higher peak coolant and head temperatures. With the bash plate fitted oil temperatures increased further so an oil cooler with a thermal bypass and a fan is going to be fitted when the new side cover arrives for modification. I reckon this should help offset any potential cooling deficit introduced by the coolant change.

[andrewsmith]

Cheers embee was thinking about this after a chat with a bandit 600 owner.

For the record race bikes can only use water, as glycol is a slippery substance when spilled

[embee]

...so I fitted a higher volume water pump .... The radiators seem to be up to the task for all but the worst conditions and could probably benefit from a small increase in volume or area.
Oil temperature spikes atrociously in heavy conditions, is slow to decrease and corresponds with higher peak coolant and head temperatures.

Have you considered a water/oil heat exchanger ("cooler")? The advantage of these compared to an oil/air cooler is that you use the coolant rad for the heat rejection, which usually has spare capability and a fan if you end up at low speed with not much air-on speed. The oil temp then tends towards the coolant temp under most cicumstances, so if it's cold it'll warm the oil quicker, and if oil is hot enough to be an issue it'll always be well above coolant temp so will cool it. No need to mess with oil 'stats etc, and no extra finned rad to worry about.

If you have a spin-on oil filter it's pretty simple to fit one. If it isn't a spin-on, it might not be feasible.

Heat to oil is typically 10-15% of the heat to coolant and is highly power (load x rpm) dependant, so plenty of revs usually makes an oil/water exchanger more effective. There's usually enough spare flow capacity in the coolant pump to accommodate the flow through the cooler (size the pipes accordingly, 1/3 diameter = 1/10 area = 1/10 flow in rough terms, reducing coolant flow 10% should only increase mean temp 1degC typically, rarely a problem

I fitted one from a zx6 Kwak to my SV650, and a zx10 version on my 1.0L Yaris. On the car it has improved fuel economy (only a couple of mpg admittedly) on short runs, but I fitted it mainly because I tow a trailer with it and it's only a littl'un so works quite hard.

Ideally the flow/return should be at the in/out of the pump, but on the SV I simply used the "bypass" hose, this is not gated by the 'stat on the SV unlike on cars and some other bikes so flows at all times though flow drops as the 'stat opens so isn't ideal but it's enough.

[MattCollins]

That was one of the first options considered except that I would have turned up a couple of exchangers and installed them in the lower radiator tanks. An independent oil cooler allows oil temperature lower than coolant temperature and adds something to the total cooling capacity.

[embee]

You've obviously thought through the options well. All I'd say is that it's usually ideal for the oil to be about the same as the coolant temp, but when all's said and done there are many ways to skin that cat. ;)

[punyXpress]

Ethylene Glycol:
Apparently VERY toxic for cats & they are attracted to it for some reason.

[Bibio]

so in laymen's terms, basically the bike would take longer to heat up and harder to cool?

[johnnyrod]

Ethyelne glycol is indeed toxic, it also has a sweet taste. There are various campaigns on to make it mandatory to add Bitrex or something else foul-tasting. As said above propylene glycol is non-toxic and is used in cough syrup etc. i read the stuff in th links on page 1, the main thrust of the efficiency increases seems to be based on fan power consumption, but in a car/bike the fan hardly ever runs. And for sure you could increase the running temp a bit, but is it really a good idea when the rest of the bike is made to run at its usual temp? Presumably you'd need to find a new mechanical thermostat to achieve this anyway? And mod the fan switch so it doesn't come on too early?