2014-2020 Formula One 1.6l V6 turbo engine formula

[ringo]

You actually did propose it in a form or fashion. We were all discussing the probable power output and boost pressure values that these engines run. You consistently denied it when other people had a a figure higher than yours and then referred them to calculations you made. The problem is you made the calculations based off a number, cannot recall right now, lower than lambda .98. How can you deny somebody else's claim and say it is wrong then show them calculations based off of a figure you just admitted would be at least slightly to low?

That sounds like claiming that they run a rich mixture to me.

Another question. Modern road car engines run under different conditions than F1 engines do they not? Those modern road car engines only produce power through the crankshaft and whether you want to admit it or not having a compounded engine does change up how things are worked out.

They are what I would consider medium stressed as they are designed for years of running, low duty cycle engines (not designed to be run at full power for long amounts of time). F1 engines are the exact opposite of that, highly stressed and high duty cycle.

Let's look at this scenario. A road car engine runs at full load. It will run slightly rich. Why would they do this? They know that they already have as much air as possible in the cylinders (hence airflow restricted) and know that by pumping in extra fuel that it ensures all the air will be consumed making the maximum power. In our F1 engine series we have a situation where we are limited by fuel. Therefore designers will want to at least ensure all the fuel is burnt releasing the maximum amount of energy. For this to happen there has to be at least a stoichiometric ratio. Many designers say that you add 2% to the airflow stoichiometric figure to ensure complete fuel combustion. Therefore wouldn't it be fair to say that these engines run at least lambda 1.02?

So you are saying that the writght turbo compound possesses these similarities to the F1 engine over a modern sports car?
I beg to differ.
A turbo compound does not produce power through the compound setup. Power production is still only from combustion.
It simply harnesses that power through two mechanical devices, the crank and the hot wheel, or whatever you want to call it. This still has nothing to do with efficient combustion. It changes nothing.
Look I am not saying the engines do not run lean, all i am saying is that you do not have evidence to support if it runs lean or rich, as both are not the most thermodynamically efficient.

Have a look at this:
I use this to find my in cylinder conditions. The higher the combustion temperature the higher the combustion efficiency.
I have mentioned this calculation earlier, but didn't want to over detail the chat. But this is what we really use to know exactly what happens with combustion mixtures. I use a program to do this of course, where other factors like moisture and fuel state is a consideration, but this is the foundation of your discussions.
http://web.mit.edu/16.unified/www/FALL/ ... de111.html

a graph of combustion temps for different lambda would look like this:

As you can see lambda =1 is ideal.
However if we look either side of 1, the temperatures are much higher when you go to the lean side, and there is less sensitivity to lambda above 1, and i think this is why what you suggest that more than 1 was found to be better in practice. The truth is that it's really just because you may not have to be dead precise and given the limitations in the precision of the current technology, lean running somewhat guarantees a close enough peak temperature. (and even then it depends on which fuel you use, as some of them are quite symetrical on either side of lambda =1)

So again, knowing this, i can tell you that i did not promote the use of a rich mixture. I only did it to compare the past turbo engines to the current, like for like. Stoichometry is ideal. Simple as that.
There is no existing evidence of what the F1 engines use currently, in fact i think it changes through the rev range and load conditions. What we need to do is look on a modern engine, which is still the most similar thing to an F1 engine despite limitations for emissions and longevity.

[gruntguru]

As you can see lambda =1 is ideal.
However if we look either side of 1, the temperatures are much higher when you go to the lean side, and there is less sensitivity to lambda above 1

Your graph shows "equivalence ratio" i.e. 1/lambda. High numbers are rich not lean.

There is nothing "ideal" about the highest temperature.

The temperatures peak at 1.0 because there is no dilution by excess fuel or air to absorb some of the heat i.e. the heat of combustion is heating the minimum mass of products so temperature is maximised.

The rich side drops off more gradually because there is much less fuel mass than air at 1.0 (typically 14.7 kg air per 1 kg fuel), so moving to say 10% rich involves adding less extra mass than moving to 10% lean.

[J.A.W.]

A turbo compound does not produce power through the compound setup. Power production is still only from combustion.
It simply harnesses that power through two mechanical devices, the crank and the hot wheel, or whatever you want to call it. This still has nothing to do with efficient combustion. It changes nothing.

To quote the famous Professor Farnsworth.. "Whaaaa?"

It is surely, the purpose of a high efficiency ICE to derive maximum work from the fuel mix combusted..
Clearly, the turbo-compound has shown how this is best done - as you write "harnesses" this - for shaft output..

Your seemingly purblind reiteration that "modern" passenger car mills - which are designed to meet all manner of road rule, drivability, cost & longevity/warranty/production compromises.. ..are somehow more closely related to F1 engines..
..as maximum efficiency exemplars.. ..than the historical examples of max-efficiency ICEs as cited..
..really serves no technically informative purpose..

[trinidefender]

You actually did propose it in a form or fashion. We were all discussing the probable power output and boost pressure values that these engines run. You consistently denied it when other people had a a figure higher than yours and then referred them to calculations you made. The problem is you made the calculations based off a number, cannot recall right now, lower than lambda .98. How can you deny somebody else's claim and say it is wrong then show them calculations based off of a figure you just admitted would be at least slightly to low?

That sounds like claiming that they run a rich mixture to me.

Another question. Modern road car engines run under different conditions than F1 engines do they not? Those modern road car engines only produce power through the crankshaft and whether you want to admit it or not having a compounded engine does change up how things are worked out.

They are what I would consider medium stressed as they are designed for years of running, low duty cycle engines (not designed to be run at full power for long amounts of time). F1 engines are the exact opposite of that, highly stressed and high duty cycle.

Let's look at this scenario. A road car engine runs at full load. It will run slightly rich. Why would they do this? They know that they already have as much air as possible in the cylinders (hence airflow restricted) and know that by pumping in extra fuel that it ensures all the air will be consumed making the maximum power. In our F1 engine series we have a situation where we are limited by fuel. Therefore designers will want to at least ensure all the fuel is burnt releasing the maximum amount of energy. For this to happen there has to be at least a stoichiometric ratio. Many designers say that you add 2% to the airflow stoichiometric figure to ensure complete fuel combustion. Therefore wouldn't it be fair to say that these engines run at least lambda 1.02?

So you are saying that the writght turbo compound possesses these similarities to the F1 engine over a modern sports car?
I beg to differ.
A turbo compound does not produce power through the compound setup. Power production is still only from combustion.
It simply harnesses that power through two mechanical devices, the crank and the hot wheel, or whatever you want to call it. This still has nothing to do with efficient combustion. It changes nothing.
Look I am not saying the engines do not run lean, all i am saying is that you do not have evidence to support if it runs lean or rich, as both are not the most thermodynamically efficient.

Have a look at this:
I use this to find my in cylinder conditions. The higher the combustion temperature the higher the combustion efficiency.
I have mentioned this calculation earlier, but didn't want to over detail the chat. But this is what we really use to know exactly what happens with combustion mixtures. I use a program to do this of course, where other factors like moisture and fuel state is a consideration, but this is the foundation of your discussions.
http://web.mit.edu/16.unified/www/FALL/ ... de111.html

a graph of combustion temps for different lambda would look like this:
http://upload.wikimedia.org/wikipedia/en/5/59/Cpft.jpg
As you can see lambda =1 is ideal.
However if we look either side of 1, the temperatures are much higher when you go to the lean side, and there is less sensitivity to lambda above 1, and i think this is why what you suggest that more than 1 was found to be better in practice. The truth is that it's really just because you may not have to be dead precise and given the limitations in the precision of the current technology, lean running somewhat guarantees a close enough peak temperature. (and even then it depends on which fuel you use, as some of them are quite symetrical on either side of lambda =1)

So again, knowing this, i can tell you that i did not promote the use of a rich mixture. I only did it to compare the past turbo engines to the current, like for like. Stoichometry is ideal. Simple as that.
There is no existing evidence of what the F1 engines use currently, in fact i think it changes through the rev range and load conditions. What we need to do is look on a modern engine, which is still the most similar thing to an F1 engine despite limitations for emissions and longevity.

Ok what on the conditions of stratified charge running. The idea being that you have the flame concentrated in the centre of the piston with a 'wall' of air essentially lining the combustion chamber. In those conditions the resultant combustion can be at a stoichiometric mixture and yield its results with those conditions while at the same time if you take the combustion chamber as a whole it is not considered lean in the traditional sense.

[trinidefender]

Oh and ringo, if you didn't suggest the use of a rich mixture then please don't dispute other people's power numbers and say they are wrong when you say you did calculations and are now saying you did your calculations with a purposefully wrong air:fuel ratio that you "used it for comparison"."

I am curious though. The graph you provided doesn't seem to take into account the usual 2% extra air needed to burn the fuel or take into account anything like stratified combustion or anything. All it tells us is the maximum temperature rise fora particular A:F ratio and does not, as gg pointed out, take into account the fact that the high temperature is diluted down by the extra air.

[Tommy Cookers]

you are quoting 60s NASA material on the efficiency benefits of higher PR in the turbo parts of a compounded engine this I have repeatedly said,( 40s NACA descriptions of raising mean exhaust pressure retaining PU power and boosting efficiency) essentially the same thing, an effect that is in principle unrelated to mixture strength

True - however the principle of efficiency increasing with PR and CR when applied to the Brayton cycle or the Otto cycle or the Diesel cycle is independent of mixture strength. The case in hand - where the displacement of the recip' machine is fixed and the rate of heat addition is fixed will unquestionably see an efficiency improvement as the PR of the turbo cycle is increased. Yes the returns will diminish more rapidly due to real-world isentropic efficiencies in the compressor and turbine, and due to fixed friction in the 1.6 litre recip' and rising heat losses compared to the fixed heat input rate as PR is increased.
However, the turbo machinery on the F1 PU does not exist simply to produce sufficient airflow for the 1.6 L recip' to burn the available fuel. It is an integral part of a compound heat engine which has two stages of compression and two stages of expansion. The PR chosen for each stage will have an effect on the efficiency of energy recovery from the entire system.

The primary benefits of leaning are reduced dissociation (effectively more complete combustion) and reduced heat loss to the cylinder wall due to lower combustion temperature. The second of these can be reduced even further with stratified charge by concentrating heat toward the centre of the chamber - away from the walls.

There is general agreement that the MB customer engines have lower power (circa 30hp) due to fuel and lubricant advantages enjoyed by the MB team and their development relationship with Petronas.

for the fine phrases above to be representative we would need an engine like the NASA-study compound CI/turbine
ie turbine power greater than the CI core power (so needing and having an artificially low CR in the CI core)

in such an engine the turbine runs as a pressure turbine
ie the mean exhaust pressure is raised ie there is 'back' pressure and so a cost in crankshaft power
but turbine power increases, power is moved from crankshaft to turbine
by improving pressure conservation backpressure raises efficiency, so PU power benefits

here in F1 we have a rather small turbine power, and a rather large core ie crankshaft power is the main power source
because the '120 kW' rule effectively limits the turbine power (no accident, and the MM mgu-h was quoted as 70 kW)
so the PRs or whatever characteristics of the compressor and turbine will have only a rather small effect
history shows (if seperate exhaust pressure pulses are conserved by design) turbine power recovery equal to 18% of crankshaft power
for a fixed fuel amount and without any 'backpressure' and so without any loss of crankshaft power
(this at sea level and a 'boost' of 2.2 bar abs) - corrected for our F1 CR and boost we would get 16%
this running without raised mean exhaust 'backpressure' is blowdown running of the turbine
as obtains in the turbines of all road car turbos and most race ones
from this 16% we must deduct the supercharger power of course, but without backpressure this would be small unless we ran very lean

turbine power can be further increased by engineering some added backpressure (ie combined blowdown and pressure running)
though with some exhaust arrangements (eg 12 cylinders and 1 turbine) blowdown working is impossible
(the scavenge stroke always 'sees' backpressure)

with their exhaust manifold design Renault and Ferrari seem to intend blowdown working
Mercedes have a different exhaust manifold design, intending combined or even pressure-only running ?
or since Merc use better fuel etc than their customers, maybe they are also ahead of R's and F's fuel ?

all the above is independent of mixture (lean or otherwise)

I have previously tried to show Ganesan's (uncalibrated) representations of the potential of leaning for efficiency
these seem quite unencouraging - but we should note that they are power-normalised
GG has shown Bosch's (uncalibrated) representations, as usual these 'actual engine' values are not normalised
so they are influenced by the fall in friction and other losses with falling rpm etc as power falls inevitably with leaning
and usually engines that burn lean mixtures well have excessive turbulence at full (non-lean) power
as would the Honda in the greencarcongress abstract
so the BMW and Honda figures are in a way tending to exaggerate the benefits to F1
eg unless Honda's engine uses 99 kg/hr of fuel running at 30 AFR we cannot in F1 reach its apparent efficiency
we must in airflow terms 'resize' its design to use 99 kg/hr (and make it a 1.6 litre V-6 running 10500 rpm etc)
the F1 sizing must come from the right combination of boost and backpressure
resizing eg for Honda 30 AFR means greater losses etc (to be weighed against the benefits of leaning)

FWIW I am impressed by the historic lack of advocacy of seriously-lean-mixture design
eg in aviation or other WOT use 20% lean was as far as they went (though rpms were usefully low wrt combustion)


@gg
how about some figures for your suggested 3.5 bar abs boost ? (supercharging power and turbine power)

btw about dissociation .....
fuel chemistry might be made to manage dissociation
though dissociation might anyway raise HUCR
isn't dissociation time-dependent ? (we never hear about it in high-rpm race engines)

[Brian Coat]

High OPR is needed for high overall efficiency but why does this need high compressor boost?
Can't the piston CR do this without the need to pump lots of excess air into the cylinder? Up to the knock limit, of course.

On another note, I think the choices about the work split between the two shafts is interesting but I have not seen it discussed. Cosworth suggested that you can easily short change the MGU-H (in favour of crankshaft) and deteriorate lap performance (e.g. ES depletion).

[gruntguru]

for the fine phrases above to be representative we would need an engine like the NASA-study compound CI/turbine ie turbine power greater than the CI core power (so needing and having an artificially low CR in the CI core)
here in F1 we have a rather small turbine power, and a rather large core ie crankshaft power is the main power source
because the '120 kW' rule effectively limits the turbine power (no accident, and the MM mgu-h was quoted as 70 kW)
@gg
how about some figures for your suggested 3.5 bar abs boost ? (supercharging power and turbine power)

We do have just such an engine in F1. The PR is much lower - probably 3.5 - 4.0 compared to 10 for the Garrett proposal and 6.5 for the Napier Nomad. Nett turbine power for the Garrett proposal was only 32% of crankshaft power, 24% of total power.

70kW MGUH is an interesting number. (Does MM = Mercedes?)
- A compressor running at 0.58 kg/s (lambda = 1.4), 3.5 PR and 80% eff needs 93 kW to drive it.
- A pressure turbine running on 1173*K exhaust at 0.61 kg/s, 3.0 PR and 85% eff produces 164 kW
- Nett power to MGUH = 164 - 93 = 71 kW
Harnessing some blowdown energy will clearly improve the picture.

GG has shown Bosch's (uncalibrated) representations, as usual these 'actual engine' values are not normalised so they are influenced by the fall in friction and other losses with falling rpm etc as power falls inevitably with leaning

F1 engine on the dyno at 10,500 rpm. Fix the fuel flow at 100 kg/hr and vary the airflow by varying the boost. As you lean the mixture through stoichiometric, the BMEP increases as ITE and BTE increase. FMEP will change slightly but the mixture at which BTE is maximised (also max BMEP) will be almost identical to the mixture at which ITE (and IMEP) is maximised.

Besides, the Bosch Automotive Handbook states "Spark ignition engines with intake manifold fuel injection achieve the lowest fuel consumption at constant engine output dependent on the engine at 20% - 50% surplus air (lambda = 1.2 - 1.5)" (8th edition, Page 559). Do you think by "lowest fuel consumption" they mean BSFC or ISFC?

F1 sizing must come from the right combination of boost and backpressure resizing eg for Honda 30 AFR means greater losses etc (to be weighed against the benefits of leaning) FWIW I am impressed by the historic lack of advocacy of seriously-lean-mixture design eg in aviation or other WOT use 20% lean was as far as they went (though rpms were usefully low wrt combustion)

10% lean was for many years the "optimum" mixture for fuel efficiency and engines like the Wright Turbo Compound running 20% lean were the exception. Direct injection on the WTC was clearly one enabling factor. Improvements in ignition technology (spark energy but particularly also Electronic Spark Timing) and combustion chamber design have enabled effective combustion of much leaner mixtures. My 1968 Bosch Handbook states "minimum fuel consumption with roughly 10% excess air" (Page 360) and now the current Bosch Handbook states 1.2 - 1.5 as optimum for homogeneous mixtures. So clearly - times are changing.

btw about dissociation .....
fuel chemistry might be made to manage dissociation
though dissociation might anyway raise HUCR
isn't dissociation time-dependent ? (we never hear about it in high-rpm race engines)

All chemical reations are to some extent time dependent. Dissociation is mostly temperature dependent. Exothermic reactions tend to be driven backwards as the temperature increases - think of the very high temperature trying to "force" the heat energy back into the molecules driving them back to the higher energy state that existed before combustion. I seem to recall Ganeson's book showing a graph of dissociation v's AFR with dissociation peaking around stoichiometry.

[gruntguru]

High OPR is needed for high overall efficiency but why does this need high compressor boost?
Can't the piston CR do this without the need to pump lots of excess air into the cylinder? Up to the knock limit, of course.

TE for a real SI engine doesn't keep increasing at high CR's as predicted by the Otto cycle efficiency. This is similar to the old single expansion steam engine problem where too much expansion in one cylinder results in cooling of the cylinder by the highly expanded working fluid. The cool cylinder then absorbs a lot of heat from the hot working fluid during the early stages of expansion. The solution was double and triple expansion engines where steam was partly expanded in a very hot cylinder then piped to a lower pressure, lower temperature cylinder for further expansion and so on.

It is more thermodynamically efficient to limit the expansion in the recip' stage and continue the expansion in the turbine.

On another note, I think the choices about the work split between the two shafts is interesting but I have not seen it discussed. Cosworth suggested that you can easily short change the MGU-H (in favour of crankshaft) and deteriorate lap performance (e.g. ES depletion).

This is more related to your first comment than you think. Again, the best overall efficiency and therfore power, will occur at some ratio of turbine power to crankshaft power. For normal full-power running, this would be the setting used. If the ES is full, more power would be temporarily available by "short-changing" the turbine to increase crankshaft power while running the MGUK (and possibly the compressor) entirely from the ES.

[Tommy Cookers]

It is more thermodynamically efficient to limit the expansion in the recip' stage and continue the expansion in the turbine.

as predicted by the arbitrary maths of eg Brayton, who assumed steady flow and steady combustion
the real characteristics of intermittent combustion and flow and ....
the real characteristics of the real turbines (and blowers) used can/will be the dominant factor
here we have only a low turbine PR available without loss of efficiency ?

backpressure improves efficiency by conserving pressure in blowdown (denser exhaust means less supersonic flow/heat)
backpressure means higher turbine PR but the main credit does not belong to the turbine
and it's said CI is better suited to higher backpressure (higher turbine PR), presumably due to higher CR, but we have SI

improving PU efficiency by raising turbine PR becomes counterproductive when the electrical power becomes more than can be used

raising blower ('compressor') PR would seem to demand/benefit from a lot more charge cooling


btw ....
MM is Magneti Marelli, they made some mgu-h publicity last year at Monza time
checking, it seems MM gave 80 kW (presumably a continuous rated value)

Wright TC DI was only 500 psi, for burning lean the fuel was better prepared (by time & heat) in carburetted engines

[gruntguru]

It is more thermodynamically efficient to limit the expansion in the recip' stage and continue the expansion in the turbine.

as predicted by the arbitrary maths of eg Brayton, who assumed steady flow and steady combustion

No. I presented a case in terms of
1. Otto efficiency vs CR compared to real SI engine efficiency with increasing CR.
2. Heat-loss issues associated with excessive expansion in a single stage piston expander.

the real characteristics of intermittent combustion and flow and .... the real characteristics of the real turbines (and blowers) used can/will be the dominant factor here we have only a low turbine PR available without loss of efficiency ?

A PR of 3+ is not "low". This is a common range for turbochargers and efficiency can be well over 80%.

improving PU efficiency by raising turbine PR becomes counterproductive when the electrical power becomes more than can be used

Agreed. But as I showed above a PR around 3 is unlikely to generate more than 80kW surplus.

raising blower ('compressor') PR would seem to demand/benefit from a lot more charge cooling

Why? It is thermodynamically preferable to have no charge cooling. Charge cooling will only be provided to the extent required to control detonation and thermal stress in the chamber. It is certainly not needed to increase air mass flow.

[Tommy Cookers]

in what engine is it thermodynamically advantageous to not have charge cooling ?
in an engine that has too low a CR
charge cooling increases HUCR

or in an engine that has too high a turbine PR and so gives more electrical power than can be used
and less crankshaft power than the other cars in the race


btw .....
why worry about dissociation ?
not only does (consistent) dissociation raise HUCR
but, in your lean engine, reassociation would occur in the exhaust manifold and so increase turbine power
though there could be a protest over this

[trinidefender]

in what engine is it thermodynamically advantageous to not have charge cooling ?
in an engine that has too low a CR
charge cooling increases HUCR

or in an engine that has too high a turbine PR and so gives more electrical power than can be used
and less crankshaft power than the other cars in the race


btw .....
why worry about dissociation ?
not only does (consistent) dissociation raise HUCR
but, in your lean engine, reassociation would occur in the exhaust manifold and so increase turbine power
though there could be a protest over this

To my knowledge from a thermal point of view concerning BSFC it is more efficient to have a higher induction temperature.ma charge cooler is generally only used to increase the power density of an engine.

[Tommy Cookers]

..... To my knowledge from a thermal point of view concerning BSFC it is more efficient to have a higher induction temperature.ma charge cooler is generally only used to increase the power density of an engine.

surely (in a given engine) the charge temperature before compression in the cylinder is a driver wrt detonation ?
ie the lower the charge temperature before compression the greater is the maximum useable CR
or the lower the charge temperature before compression the greater is the maximum useable mep (for a given CR)

PN testing for Avgas uses boost air at various pressures but fixed temperature (to establish mixture effects on max useable mep)
fixing air temperature (while maintaining cylinder CR) shows that richening allows greatly raised mep without increased detonation
greatly being eg 30% in the standard 100/130 Avgas
ie the usual charge heating with supercharging has been eliminated

[trinidefender]

..... To my knowledge from a thermal point of view concerning BSFC it is more efficient to have a higher induction temperature.ma charge cooler is generally only used to increase the power density of an engine.

surely (in a given engine) the charge temperature before compression in the cylinder is a driver wrt detonation ?
ie the lower the charge temperature before compression the greater is the maximum useable CR
or the lower the charge temperature before compression the greater is the maximum useable mep (for a given CR)

PN testing for Avgas uses boost air at various pressures but fixed temperature (to establish mixture effects on max useable mep)
fixing air temperature (while maintaining cylinder CR) shows that richening allows greatly raised mep without increased detonation
greatly being eg 30% in the standard 100/130 Avgas
ie the usual charge heating with supercharging has been eliminated

Here is a real world example. Look at this: http://www.k20a.org/upload/HondaRA168EEngine.pdf

"As the intake air temperature increases B.S.F.C. Becomes better" see figure 10 in the paper. They go on to say that as air intake temp increases there is "a tendency to generate knocking, and the ignition timing must be retarded to avoid it."

They talk about how higher air temps helps with vaporisation of the fuel. With a direct injection engine running at the current pressures I'm not sure how much difference that will make though.

I could be wrong but I have known it as a rule of thumb that best BSFC occurs at a temperature round about just below the temperature that starts encouraging knocking.

Point #2. What interests me is how the paper also relates fuel temperature to BSFC improvement. They relate it, again, through higher fuel temps encouraging fuel vaporisation. I actually wouldn't be surprised if F1 teams pass the fuel through a heat exchanger to preheat the fuel. TC and GG do you agree with this statement? If not why?

*EDIT*
Point #3. They equate best fuel consumption at an equivalence ratio of 1.15. They go on to talk about how making the engine leaner than 1.02 creates a situation where "unsatisfactory transient response may appear, thus making the engine become insufficient for racing performance."

So they are saying BSFC is best at lambda 1.15 however for racing they had to run the engine at lambda 1.02 for engine response reasons?