14 replies to this topic

[bobdar]

Been wondering about engine cooling systems today, and hopefully will find some answers here.

Most engines tuners have an "ideal" coolant temperature they prefer to run (water-based coolants), and usually claim to lose HP above that temperature. However, with a sizable percentage of the input energy going out the radiator, would it make sense to elevate the coolant temperature (using non-H2O-based coolant) and decrease the amount of energy rejected by the radiator?? Assuming that detonation could be avoided, what sort of metal (iron, aluminum) temperature limits are practical? Any merit to this approach?

[Wuzak]

The amount of energy that needs to be rejected by the radiators is, surely, dependent on the amount of energy the engine is putting into the cooling medium. If the energy rejected by the radiators is less than the energy put into the cooling medium the temperature of the cooling medium would rise.

Coolant systems in most cars already are under pressure to prevent the water solution from boiling.

The coolant systems in F1 cars have even higher pressures to allow higher coolant temperatures. This allows smaller radiators, as the heat rejected by a radiator depends, amongst other things, on the temperature difference between the cooling fluid and the air.

[Greg Locock]

Check the NACA library, they've probably got a report on engine out put vs coolant temp. I know it is possible to overcool engines, as I've worked on a couple where we had to raise the ECT to get emissions under control, and that also boosted the max power. It's a trade-off between heat lost through the combustion chamber walls vs some sort of reduction in volumetric efficiency (at a guess). Then, finally the coolant boils, its heat transfer characteristics change (I'm tempted to say it reduces the conductivity, but there is aregime where steam/water mix is more efficient as a conductor), and you get local hot spots, and then material failure.

I suppose in the longer term we'll see higher pressure cooling systems on road cars, for emissions and fuel consumption.

[McGuire]

Originally posted by Greg Locock
Then, finally the coolant boils, its heat transfer characteristics change (I'm tempted to say it reduces the conductivity, but there is aregime where steam/water mix is more efficient as a conductor), and you get local hot spots, and then material failure.

Nucleate boiling -- the exchange rate is a genuine order of magnitude higher than ordinary single phase convection but it's really tricky, needless to say. Locally the coolant is "super-cooling" the surface it is adhered to but is right on the ragged edge of film boiling, and then cavitation. Nasty stuff -- cavitation can blow a pinhole straight through the cylinder liner of a diesel engine.

[Greg Locock]

http://naca.larc.nas...50/naca-tn-2069

[McGuire]

Originally posted by bobdar
Been wondering about engine cooling systems today, and hopefully will find some answers here.

Most engines tuners have an "ideal" coolant temperature they prefer to run (water-based coolants), and usually claim to lose HP above that temperature. However, with a sizable percentage of the input energy going out the radiator, would it make sense to elevate the coolant temperature (using non-H2O-based coolant) and decrease the amount of energy rejected by the radiator?? Assuming that detonation could be avoided, what sort of metal (iron, aluminum) temperature limits are practical? Any merit to this approach?

It's hard to find anything that can beat plain old water. For example, nonaqueous propylene glycol has a boiling point of 370 F, but only about 60% the latent heat of H20. With the larger radiator, pump etc. required you are pretty much back where you started...especially where weight and package size are an issue.

What we would like is a coolant with the heat transfer properties of H2O, but with a higher boiling point and lower surface tension so it doesn't bead. Like water, only wetter. Also, since there are something like 750 million vehicles on the road all pumping at least a gallon of the stuff (except for the Beetles har) non-toxic and biodegradable might be nice too.

For the very reasons you nicely stated, coolant temps in production cars have been slowly creeping up over the years in the effort to gain efficiency...mainly due to better materials, closer engineering and electronic controls. I can see how improving materials technology might some day allow engines to carry small, almost vestigal liquid-cooling systems. The hot spots are the exhaust valve seat and the piston, which demand a considerable percentage of the cooling system's capacity.

[McGuire]

Originally posted by Greg Locock
http://naca.larc.nas...50/naca-tn-2069

Great stuff -- a Packard Merlin being used as a mule for basic cooling system testing. What a cool job they had.

[bobdar]

Cool job indeed. Lots of good info at that site, might take the rest of the year to absorb it all. So, the limits are actually the piston material (which could be cooled from below??) and the head temps between the exhaust valves. How effective are the ceramic coatings for thermal blocking?
F1 engines have anything special in this regard?? Thank you for your replies.

[desmo]

None of the photos of F1 piston crowns or combustion chambers I've seen show any evidence of coatings having been applied. Looks like bare Al to me.

[-RM-]

One could use liquid metal cooling perhaps?

http://www.coolingzo...ke_June_04.html

[J. Edlund]

Originally posted by desmo
None of the photos of F1 piston crowns or combustion chambers I've seen show any evidence of coatings having been applied. Looks like bare Al to me.

Cosworth has also claimed that for example use of carbon-carbon pistons decrease engine output due to detonations. I assume that coatings would have the same effect.

[McGuire]

Originally posted by desmo
None of the photos of F1 piston crowns or combustion chambers I've seen show any evidence of coatings having been applied. Looks like bare Al to me.

This raises an interesting question...if we are attempting to maximize cooling efficiency in the hope of reducing cooling system capacity or even someday eliminating it altogether, do we want a thermal barrier or a thermal conductor on the piston? If it's the latter, aluminum will be pretty hard to beat.

[CFD Dude]

The coatings on pistons are necessarily there for strictly thermal reasons. The sides of pistons can be coated to reduce friction when they come in contact with the cylinder bores, and on the top of the pistons the coatings can be to harden the piston crown. I've seen this done on forced induction engines where burnthrough was a problem. You see it less on naturally asperaited engines, the NASCAR pistons I've seen are just bare aluminum.

The 'ideal' engine temp often has to due with the thermal material properties of the engine block and cylinder head. The coolant temp may run around 200F due to thermodynamic considerations of the coolant, but that's not the temperature in the block and the head. If the temperature gets too hot you can crack the head and when that coolant pours into the combustion chamber it doesn't matter what type of coolant you're running; the engine is toast.

Working with new engines designs in the past we've taken one of the first assembled engines and thermocoupled everything we could inside it and then ran it on the dyno to see how hot the metal was getting. From there we'd adjust the coolant flow and pressure with the waterpump impeller and pulley to find how much cooling the engine needed and from there determine the radiator size required. The limiting factor usually ends up being the valve bridge temps, if you have a cooling issue in the block you can use the oil to take care of that (ie piston squirters), but you're much more limited in the head. Unless of course we dredge up the 'monoblock' again, but maybe we'd best not go there again...;)

[CFD Dude]

In my opinion McGuire is right, you want the piston to transmit as much of the heat as possible, not to block it. THe combustion temperatures are very hot and if you can't transfer that heat away from the combustion chamber as fast as you can you're going to have problems. Remember, not all of the heat is elimiated throught the coolant, a significant amount of heat is also lost through convection to the air on the outer surface of the block, carried out with the exhaust gases and picked up by the oil.

[J. Edlund]

AlBe alloys has a slightly better thermal conductivity than conventional aluminum alloys used for pistons. So with pistons and liners/cylinders of ALBe the cooling should be increased, especially if the AlBe liners are replacing cast iron liners. A machined surface in the coolant passages instead of a cast surface also increase cooling somewhat.

If I remember correctly, NACA had some models of heat transfer from the piston (which mostly goes through the piston rings). But I don't remember the name/number on the report.

There was also a SAE paper (983024) where the testing and optimisation of the cooling system was described, they hooked up several sensors to the engine and modified the coolant flow until the temperatures dropped below a desired value.