Engine cooling numbers?

[xpensive]

Surely this has been covered within a thread before, but I honestly cannot find the proper one to tag on to.

Does anyone have relevant numbers of an F1 engine cooling system, in terms of flow, delta T and as a consequence, kW conveyed to the ambience?

[marcush.]

I ´m still collecting data for this thred to be launched mate ! I have data from current cars and as far as I know the cores are differing in that area not that much....but maybe someone with more insight could add here -do the top teams go as far as producing their own fins and tubes? or is it just using what is avaialble in the market...My understanding is that a lot can be done in that area but the cost are immense as you will need to order a huge amount of material and attract the supplier somehow to go into it and have no production to recover the cost...

[strad]

They run Xtremely hot.

[xpensive]

Cooling calc's was not sexy a topic to the F1T community as I had imagined, which I guess leaves me and you marcush?

Let's see, try this for starters; Water's capacity for coneying energy can be simplifeied to 70 Watt per Lpm and C.

How to make use of this humble piece of info; If you have 20 liters per minute waterflow, passing through a radiator,
70C going in - 40C going out, the power convection is 20 * (70-40) * 70, equalling 42 000 Watt or 42 kW. Simple, eh?

A wise man at F1T once said something like the energy content of gasoline is equally split between mechanical-, xhaust- and cooling-power. If that's the case, there would be some 550 kW dissipated through those radiators at full song.

According to the 70W/K*Lpm formula above, a delta T of 40 C would call for 200 Lpm, which calls for comments?

[machin]

My final year project at Uni was to determine the ideal heat exchanger design for the Ilmor powered Voodoo target drone...

http://www.aviationweek.com/shownews/02 ... rfrm23.htm

I wrote a program into which you could put required heat rejection figures and it would come up with core size, tube number, diameter, number of rows and fin size and number... I'll have to dig it out....

[747heavy]

is it not more an attempt to maintain a sort of equilibrium between the heat/energy input into the water while passing through the engine, and the need to reject the same energy in the radiators, while maintaining approx. the same temperature?

I can´t speak for F1 but in other automotive/racing applications the delta T from input to output of the radiator is more like ~5K.
I never have seen an automotive radiator/cooler application where you would have a delta T of 40K across the radiator.

We would need to keep in mind, what is the capacity of air to coneying energy.
I know that you have an optimum flow rate through an radiator.
If you pump you water or oil any slower or faster, you won´t have the optimum heat/energy rejection.

This depends on the ability of your other medium to coneying energy and the flow rate of this medium through the radiator.
In simple terms for an F1 application it will depend from your car speed.
Ideally you would have a mechanism to keep the flow rate constant with variying speeds (e.g. having adjustable inlet or outlets, similar as what is used in turbins for fighter jet´s).

For this reason an electrical waterpump, or a variable rate water pump is much more efficient, then the standard water pump, where flow rate changes with engine rpm.

[Belatti]

The approach I would use to try design the cooling system:

1) gather good data of the heat generated in permanent regime
2) calculate the parameters for most efficient cooling of the heat in 1) given a "base" inlet size to start with and its airflow (radiator config, angle and size, water flow rate, etc.)
3) massage the parameters in 2) to see the effects of, for example, having a smaller air intake with increased water flow rate or a larger radiator with more angle.

I guess you can do a super complex model if you have the tools, otherwise some basic hand calculations to start with will tell you much.

[Edis]

According to Toyota regarding it's F1 engine:
Coolant temperature to radiator: 130 degC
Coolant temperature from radiator: 122 degC

There are simulation software than can be used to calculate heat flow into the cooling system, but around 15% of the energy released in the engine should be a good starting point.

[xpensive]

According to Toyota regarding it's F1 engine:
Coolant temperature to radiator: 130 degC
Coolant temperature from radiator: 122 degC

There are simulation software than can be used to calculate heat flow into the cooling system, but around 15% of the energy released in the engine should be a good starting point.

Intriguing, I wonder what that might mean in terms of waterflow, how are we going to figure that one out?

[747heavy]

some things remotely related/maybe useful to this topic.

http://www.ret-monitor.com/articles/113 ... -coatings/

http://www.ret-monitor.com/articles/795 ... ng-system/

[riff_raff]

"According to Toyota regarding it's F1 engine:
Coolant temperature to radiator: 130 degC
Coolant temperature from radiator: 122 degC"

Edis,

If you accept those coolant deltaT through-the-core numbers from Toyota, then that's a start. I believe the rules require plain water as the engine coolant, so you also know the specific heat value of the coolant. If we also assume that the heat flow to the coolant is 15% of the fuel's energy content, then that would be about 13,000 Btu/min for a 700hp engine at 35%BTE.

Based on a 15degF temp drop and a specific heat of 1 Btu/lb-degF, that would require a coolant mass flow rate of 867 lb/min, or about 110 gal/min. Since that number seems high, I'd speculate that the coolant deltaT across the core is much greater than 8degC.

Maybe someone can check my math.

riff_raff

[xpensive]

Precisely my point riff_raff, I have 236 kW and 421 liters per minute. Through a 20 mm ID pipe, that's 22 m/s!

[Belatti]

The assumption of the heat flow to the coolant of 15% of the fuel's energy content seems fine to me. Its possible that it could be less? How would you compose the heat "cake"? (heat taken by the oil, the exhaust, the block radiation, etc.)

I guess that riff_raff calculations make waterflow numbers high when he says "for a 700hp engine". I wonder what would be the mean power over a lap. The same way those 8°c deltaT may vary from top speed to hairpins.


Here is a plot of data acquisition of a low power sport proto (something similar to a radical SR3). The resolution of the temp meter is only 1°c. Still, you can clearly see how the temp goes up in the slow part of the track, even if the engine is not delivering power, because of the reduced air flow.

[747heavy]

Precisely my point riff_raff, I have 236 kW and 421 liters per minute. Through a 20 mm ID pipe, that's 22 m/s!

that´s maybe not too far off.

According to some "rule of thumb" calculations, which specify the flowrate needed with 0.757 lpm*hp @ max. rpm. (this would made ~530 lpm for 700 hp)

Again can´t talk about F1, but in other race cars the water pipe ID is more like 40-43 mm.

From my expirience, I would say Edis delta T number is pretty realistic, that it is a bit higher then the number I quoted, is due to the fact that a "normal race engine" is not designed to operate at 130°C.
The higher delta T from of the water in relation to the ambient temp (let´s say it´s 25°C) makes the radiator more efficient, then one where your inlet temp is ~85-90°C, as in the case of most "normal" engines.

Some water pump flow rate discussions are going on here:

http://www.eng-tips.com/viewthread.cfm? ... 237&page=2
http://www.toyota-supra.info/forums/mki ... -pump.html

Assuming the pump was sized for a 320hp (240kW) engine, rejecting a typical 60% of its brake power to coolant, the water flow rate required to keep the temperature rise across the engine to 8degC would be 292litres per minute.
-------

The difference between the pump outlet and the engine outlet will tell you the temp rise across the engine (which should stay below 8degC). The engine outlet should also stay below yout chosen max temp - 120degC or so

[747heavy]

2001 SAE paper on cooling systems

Conventional cooling needs on smaller
engines with mechanically driven water pumps vary
between 2.0 to 2.6 L/min/kW. Some advanced engines
already run at 1.0 - 1.7 L/min/kW.
It is predicted that with an electric pump & diverter valve with
precision cooling & nucleate boiling sensing & control, flows under 1.0
L/min/kW would be achievable.