2014-2020 Formula One 1.6l V6 turbo engine formula

[OO7]

We keep hearing about the importance of combustion designs in the ICE and it seems to be the most numerous type of development in the PU. Would these new high tech combustion techniques such as TJI, provide a significant benefit in a high revving NA F1 engine (20000 rpm) or is its use largely restricted to lower revving force induction designs?

Well, how I see it, it's been reversed. In the pre 2014 engines it was how to maximise the amount of O2 (air) available in the combustion chamber, fuel wasn't so much the problem. Much of the mix could be done even before it entered the chamber (with all kind of spray nozzles). A bit of extra fuel didn't matter too much.

Now air is not the problem. Because of the turbo charging it's unlimited (in theory) but it's the fuel that needs to be burned as effectively as possible and it's only injected right into the chamber, so all of the magic happens there.

I think I understand what you mean Jolle, but if the NA engine is limited by airflow, why pump in more fuel. Surely you only want to use as much fuel as there is air for a complete burn?

[Tommy Cookers]

We keep hearing about the importance of combustion designs in the ICE and it seems to be the most numerous type of development in the PU. Would these new high tech combustion techniques such as TJI, provide a significant benefit in a high revving NA F1 engine (20000 rpm) or is its use largely restricted to lower revving force induction designs?

Well, how I see it, it's been reversed. In the pre 2014 engines it was how to maximise the amount of O2 (air) available in the combustion chamber, fuel wasn't so much the problem. Much of the mix could be done even before it entered the chamber (with all kind of spray nozzles). A bit of extra fuel didn't matter too much.
Now air is not the problem. Because of the turbo charging it's unlimited (in theory) but it's the fuel that needs to be burned as effectively as possible and it's only injected right into the chamber, so all of the magic happens there.

I think I understand what you mean Jolle, but if the NA engine is limited by airflow, why pump in more fuel. Surely you only want to use as much fuel as there is air for a complete burn?

high rpm combustion efficiency is improved by a rich mixture because combustion becomes more consistent
said tests of Ferrari 2007 F1 engines shown in an instrumentation journal (posted here years ago)
the reason why engines were run within the race to different rpm and corresponding degrees of richness in the rpm limit days ?

and a rich mixture is chemically a deterrent to dissociation, as well as (if we access this) temperaturewise a deterrent

presumably the fall in combustion efficiency that occurs in lower speed boosted engines with 'excessive' leaning is similarly related to inconsistency ?
part of the reason why TJI etc works so well ?

an NA engine in a fuel-limited formula would still be best on a mixture close to stoichiometric
(Mahle TJI gave NA 40% indicated efficiency stoichiometric and 44% at 1.7 equivalence ratio, ie this a much bigger engine for equal power))
despite (remember) a lean mixture giving more benefit (eg than stoi) for a given high CR

[gruntguru]

and a rich mixture is chemically a deterrent to dissociation, as well as (if we access this) temperature-wise a deterrent presumably the fall in combustion efficiency that occurs in lower speed boosted engines with 'excessive' leaning is similarly related to inconsistency ?
part of the reason why TJI etc works so well ?

Yes, definitely consistency - and at higher AFR than otherwise possible. TJI papers show extension of lean limit (in NA test engines) to 2.2 and above.

The other benefit - equally significant - is rapid combustion. This has a direct effect on TE.

an NA engine in a fuel-limited formula would still be best on a mixture close to stoichiometric and the benefits of eg very high CR are less than with a lean mixture

Provided the engine could breathe enough air, best efficiency (and performance) would still occur at a lean mix.

[Tommy Cookers]

TJI papers show extension of lean limit (in NA test engines) to 2.2 and above.
The other benefit - equally significant - is rapid combustion. This has a direct effect on TE.
Provided the engine could breathe enough air, best efficiency (and performance) would still occur at a lean mix.

iirc your lean limit of 2.2 etc is for the prechamber fuelled on gas - nowhere near this is possible on liquid fuelling
anyway regardless of this, F1 boost figures will show the leaning actually used

yes I assume the TJI combustion speed is the reason why Mahle are able to show NA indicated efficiencies thus .......
stoichiometric engine running giving about 40% and an engine lean to 1.7 lamda giving about 44%

(being at the same massflow) the lean engine has 65% of the stoi engine indicated power for 59% of the stoi engine's fuelling rate
relative to output its mechanical losses are disproportionate ie about 50% greater, ie its brake power is about 62% of the stoi engine's
and the bte will be about 35% for the stoi engine and 37% for the lean engine

ie to produce competitive power the lean engine will need 59% more massflow and so need to be 59% 'bigger' in displacement
the disadvantage from this package will surely nullify the higher bte ?
in a 'fair' F1 ie without minimum weight and dimensions that allow the engine package to escape the consequences of its size etc

of course slightly lean of stoich might be the best

[Brian Coat]

This is probably a spot of copyright banditry but this rather old picture illustrates the trade off very clearly (Heywood).



It's not a 2013 F1 engine of course but the fundamentals are the same.

In this example you can clearly see the interplay between (1) the ITE which rises with lambda; and (2) the combustion efficiency falling away at higher lambda, due to cyclic variability, slower burn, misfires ...

Leaner AFR will improve Full load ITE if the combustion system will put up with it.

[gruntguru]

TJI papers show extension of lean limit (in NA test engines) to 2.2 and above.
The other benefit - equally significant - is rapid combustion. This has a direct effect on TE.
Provided the engine could breathe enough air, best efficiency (and performance) would still occur at a lean mix.

iirc your lean limit of 2.2 etc is for the prechamber fuelled on gas - nowhere near this is possible on liquid fuelling

Nearer than you think: 2.1+ in this paper http://www.mahle-powertrain.com/media/m ... 1_1146.pdf

EDIT. Had a re-read and found the pre-chamber was fueled with vaporized gasoline. Nevertheless, with the pre-chamber fueled with well atomized liquid at lambda 0.9 or thereabouts, I can't see why it wouldn't ignite main-chamber mixtures of similar "leanness".

EDIT, EDIT. Found it. Liquid gasoline extended lean limit beyond 1.9. (TE same as gas-fueled pre-chamber up to 1.9)

[Brian Coat]

A lot of Mahle's published data is at low speed and lighter loads.

All other things being equal, do we expect the lean limit to improve a bit more at full load?

[gruntguru]

TJI papers show extension of lean limit (in NA test engines) to 2.2 and above.
The other benefit - equally significant - is rapid combustion. This has a direct effect on TE.
Provided the engine could breathe enough air, best efficiency (and performance) would still occur at a lean mix.

yes I assume the TJI combustion speed is the reason why Mahle are able to show NA indicated efficiencies thus .......
stoichiometric engine running giving about 40% and an engine lean to 1.7 lamda giving about 44%

(being at the same massflow) the lean engine has 65% of the stoi engine indicated power for 59% of the stoi engine's fuelling rate
relative to output its mechanical losses are disproportionate ie about 50% greater, ie its brake power is about 62% of the stoi engine's
and the bte will be about 35% for the stoi engine and 37% for the lean engine

ie to produce competitive power the lean engine will need 59% more massflow and so need to be 59% 'bigger' in displacement
the disadvantage from this package will surely nullify the higher bte ?
in a 'fair' F1 ie without minimum weight and dimensions that allow the engine package to escape the consequences of its size etc

of course slightly lean of stoich might be the best

The paper I quoted above shows a 20% increase in ITE (compared to lambda = 1.0) at 2.0, which is also the AFR for peak ITE. Peak BTE will occur at some lower figure depending on FMEP - probably around 1.8.

IMO the engine (for a NA fuel-flow restricted formula) would be sized and operated somewhere near there - with a displacement about 40 - 50% greater than a stoichiometric engine. The weight penalty would probably be in the order of 40% which is not a lot (modern engines being very light) - for a 15% power increase.

[Tommy Cookers]

The paper I quoted above shows a 20% increase in ITE (compared to lambda = 1.0) at 2.0, which is also the AFR for peak ITE. Peak BTE will occur at some lower figure depending on FMEP - probably around 1.8.
IMO the engine (for a NA fuel-flow restricted formula) would be sized and operated somewhere near there - with a displacement about 40 - 50% greater than a stoichiometric engine. The weight penalty would probably be in the order of 40% which is not a lot (modern engines being very light) - for a 15% power increase.

we are not interested in the leaning at which some particularly favourable 'peak' BTE is obtained ie lowish rpm and low friction
though it would be nice to know the manufacturer claims and the associated conditions

we are interested in the leaning at which the high power peak BTE is obtained ie high or very high rpm
this will always be less lean (for consistency/'efficiency' of combustion) and again less lean for transient response
and this lesser leaning approach also has more scope for further leaning at partial powers

before the 3 way catalyst late-designed cars ran about 25% lean in 'freeway' cruise
piston-engined aircraft run about 30% lean in cruise
(but the engine is of course sized for the required full power unleaned)

in NA F1 or MotoGP etc the lean running engine's greater bulk would be the prime aspect detrimental to track performance
I assume a mechanical efficiency of 87.5%, a significantly less favourable value tends to degrade further the NA lean engine case

[Brian Coat]

Some good points there, TC.

I figured high speed (say 12K-18K PPS) nat. asp. racing engine mech. eff. is more like 80-84%?

This is based on 3 Bar FMEP and 12K and 4 Bar at 18K [Both excluding pumping] and BMEP at 16 Bar.

Friction taken/extrapolated from:

http://www.ricardo.com/pagefiles/35412/ ... ngines.pdf

[Tommy Cookers]

Test Post unrelated to the above -
a good read on Ricardo Hyboost (already mentioned by a poster to another thread)
it seems easily googleable as Hyboost, but the link below doesn't seem to work for me

http://www.cpowert.com/assets/IMECHE%20 ... %20(3).pdf

[Brian Coat]

If anyone is interested in HyBoost, you may enjoy this FISITA paper from 2012 which we last discussed on here a couple of years ago.

(The link posted in Sep '14 no longer works because that FISITA URL has lapsed and/or been mildly hacked).

Thanks to web.archive.org, there is a copy here.

http://web.archive.org/web/201212230454 ... 01-041.pdf

Here is a working link to the IMeche paper, which is more 'vehicle-centric'.

http://www.cpowert.com/assets/IMECHE%20 ... %20(3).pdf

[NL_Fer]

In contrary to street engines, these f1 hybrids run at full power most of the time it is very different. I found allot of interesting features in gas engines for powerplants and ships, but they run much slower.

Another viewpoint is the fact that f1 engines don't need to comply with emission rules, so they can produce NOx without any problems.

I also saw a document describing the use of fueltank vacuum vapour to fill a pre chamber and use port of direct injection for e main chamber.

[gruntguru]

Another viewpoint is the fact that f1 engines don't need to comply with emission rules, so they can produce NOx without any problems.

NOx peaks slightly lean of stoichiometry but declines rapidly with further leaning. At 2.0 (using TJI) NOx is about 1/40th the level at 1.0 - low enough to meet many emissions regulations without a catalyst.

[gruntguru]

So I converted the ITE versus lambda chart from the above TJI paper to show BTE at three different values of mechanical efficiency. Best power coincides with peak BTE ie approx lambda = 1.6, 1.5 and 1.3 for the cases of M.E. = 90%, 85% and 80% respectively.