Jump to content


Photo
- - - - -

AFR and the 2014 F1 rules


  • Please log in to reply
10 replies to this topic

#1 gruntguru

gruntguru
  • Member

  • 7,635 posts
  • Joined: January 09

Posted 22 December 2014 - 02:35

The current F1 rules mean the power is no longer "air-flow restricted", it is now "fuel-flow restricted".

 

Under an air-flow formula (restricted capacity, rpm etc), the AFR is set to a value that produces the most power per gram of oxygen. This usually accurs at lambda 0.8 - 1.0 ie 0 - 20% excess fuel, depending on engine design.

 

Under a fuel-flow formula, the AFR is set to a value that produces the most power per gram of fuel. This usually accurs at lambda 1.2 - 2.0 ie 20% - 100% excess fuel, depending on engine design.

 

There are many other factors at play, but experience with PFI engines (current engines are DI which can permit even leaner operation) and a clue from Renault (MAP = 3.5 bar) suggest full power mixtures around 1.3.

 

Oh - another clue. The power levels being mentioned around the traps (around 700 hp?) suggest a thermal efficiency of 40% plus. At lambda 0.8 20% of the fuel cannot be burned meaning the 80% of fuel that is burned, is being converted to work at an efficiency of 50% plus - a figure I would suggest is impossible.


Edited by gruntguru, 22 December 2014 - 05:37.


Advertisement

#2 GdB

GdB
  • New Member

  • 22 posts
  • Joined: December 14

Posted 22 December 2014 - 04:45

The current F1 rules mean the power is no longer "air-flow restricted", it is now "fuel-flow restricted".

 

Under an air-flow formula (restricted capacity, rpm etc), the AFR is set to a value that produces the most power per gram of oxygen. This usually accurs at lambda 0.8 - 1.0 ie 0 - 20% excess fuel, depending on engine design.

 

Under a fuel-flow formula, the AFR is set to a value that produces the most power per gram of fuel. This usually accurs at lambda 1.2 - 2.0 ie 20% - 100% excess fuel, depending on engine design.

 

There are many other factors at play, but experience with PFI engines (current engines are DI which can permit even leaner operation) and a clue from Renault (MAP = 3.5 bar) suggest full power mixtures around 1.4.

 

Oh - another clue. The power levels being mentioned around the traps (around 700 hp?) suggest a thermal efficiency of 40% plus. At lambda 0.8 20% of the fuel cannot be burned meaning the 80% of fuel that is burned, is being converted to work at an efficiency of 50% plus - a figure I would suggest is impossible.

 

Good observation, but a correction (typo?) is needed :

 

"...lambda 1.2 - 2.0 ie 20% - 100% excess air..."

 

MAP does not imply any air mass flow or AFR.  The exhaust back pressure could be very high to power MGU-H, thus requiring higher MAP.

 

The fuel restriction is why I suggested D-EGR is or will eventually be used.



#3 gruntguru

gruntguru
  • Member

  • 7,635 posts
  • Joined: January 09

Posted 22 December 2014 - 05:36

lambda 1.2 = 20% excess air

lambda 2.0 = 100% excess air. Not sure of the typo?

 

1.6 Litres, 10,500 rpm, 110% VE, air density = 1.2 kg/m3, PR = 3.5, CAT = 70*C (343*K)

Air massflow = 1.6/2/1000x10500/60x1.1x1.2x3.5x(273/343)

= 0.515 kg/s

 

Fuel flow = 100/3600 = 0.0278 kg/s

 

AFR = 0.515/0.0278 = 18.5:1

Lambda = 1.26 (whoops did I say 1.4?)

 

If backpressure was so high that VE was really terrible (say 0.9) the AFR would still be 15.1 (lambda = 1.03). (There is no chance the VE would be this low. It would require BP>MAP and zero scavenge. Scavenge is essential for internal cooling and consequent knock reduction.)


Edited by gruntguru, 22 December 2014 - 05:37.


#4 TDIMeister

TDIMeister
  • Member

  • 318 posts
  • Joined: January 06

Posted 22 December 2014 - 12:49

I come up with lambda=1.23 using the same numbers as you.  However, this presumes several things and ignores several other factors that have shown to make lean-burn SI engines not survive at high power outputs.

 

1. You're putting absolute faith in Renault's 3.5 bar MAP number, which I find to be rather high for an SI engine running spec'ed 98 RON gasoline knock free with a reasonable compression ratio for the desired high efficiency.  3.5 bar MAP might be the peak value, but counterintuitively is not necessarily the value operated at rated power.  Furthermore, lean mixtures can also make engines more knock-prone, more on this in a moment.

2. Notwithstanding #1 above, the value of which is plausible given the following, methinks you overestimate the trapped VE.  One of the problems with lean-burn engines is high EGT - how is this possible you may ask?? There is one common thread that connects 1. and 2.  But to discuss only VE for a moment, one way the EGT might be addressed to protect the turbo is to run tons of valve overlap, in effect  to scavenge the cylinders with pure charge air - the fuel is then injected only at a point after the exhaust valves close.  Therefore, the VE one back-calculates out of the MAP includes a not-inconsequential portion that goes straight out the exhaust to protect the exhaust valves and turbine from limiting high temperatures. Edit: I wrote this before reading thoroughly through gruntguru's original post, which covered this subject.  My apologies to you, gg.

 

Now, the connecting thread of both the above that has been so far ignored in the discussion is that lean mixtures in SI engines cause the flame velocity to decline.  This is an irrefutable phenomenon observed from the days of Harry Ricardo.  You have surely heard or read about engines overheating, burning exhaust valves or being more prone to detonation due to running lean  - most spectacularly in dragsters and other race cars, often stories abetted by nitrous oxide, that ran "lean" (which were in fact still substantially rich of stoichiometric) that blew the engines to kingdom come.  All this is because of the effects of diminished flame velocity and extending the combustion duration -- itself not an efficiency-promoting thing!

 

That said, the flame velocity thing in lean-burn engines that I just said above could be completely moot if F1 engines employ a stratified charge and all the brains and CFD power go into designing a combustion system that does an amazing job of stratifying a power-optimal fuel-air ratio core in an otherwise overall lean mixture in the cylinders (a problem that has vexed engine designers since Ricardo himself to this day).  This is the subject of my PhD (so I'm not completely uninformed on this subject, albeit not specifically in an F1 context, unfortunately), but I would love to see the combustion chamber / piston top design and placement of the fuel injectors and spark plug of a state-of-the-art F1 engine; I can tell a lot just by looking at it.

 

P.S.: One tell-tale sign of a stratified charge engine, especially a poorly designed one, is smoke coming out the exhaust despite running allegedly lean global mixtures.  This is that vexing problem to which I alluded earlier.


Edited by TDIMeister, 22 December 2014 - 13:20.


#5 gruntguru

gruntguru
  • Member

  • 7,635 posts
  • Joined: January 09

Posted 23 December 2014 - 04:59

TDIM your response is easily the best so far (having previously made this proposition on another forum). To address your points in order.

 

1. Knock. Assuming some degree of charge stratification at full power, the end-gas will be very lean - possibly just air.

Efficiency. Being a compounded engine TE will depend on CR and PR of the "Brayton" section (the turbo-machinery). Trading off some CR to permit higher boost levels will certainly be part of the optimisation process.

 

2. EGT. High EGT does not go hand in hand with high TE - quite the opposite as you know. For a given burn rate EGT peaks at stoichiometry and falls off to either side. The reason high EGT is associated with lean mixtures is twofold. 1. Full power mixtures are typical rich (0.9 or less) so any leaning from there will see EGT increase. 2. Slow burn rate. For a given engine design, leaner mixture will burn slower, less energy goes to work, more energy remains as heat. There is zero possibility that current F1 engines are running lean enough for burn rate to suffer.

VE. Sure, a decent chunk of the high VE of an F1 engine goes straight through as scavenge. Most tuners still count this in their quoted AFR. I have seen a few cases where highly scavenged engines tuned using a WB lambda meter were set up massively over-fuelled as a result. I did include scavenged air in my AFR estimate (probably erroneously).

 

Flame velocity. The Honda RA168E operated at full load (economy mode) at lambda = 0.98, RON = 102, MON = 90, MAP = 2.5 bar, CR = 9.4:1, 1988 technology and port fuel injection. One of the keys to providing adequate flame velocity at lean mixtures is MAT (70*C in the Honda).

 

I have no doubt that current F1 engines with the benefit of DI, stratified charge, a further 26 years of engine tech development, etc are running leaner than 0.98

 

On stratified charge - road going DI engines tend to run homogeneous charge at full load. Many believe that this is due to a limitation of the technology. More likely IMHO this is purely to utilise all the oxygen. Road cars are not tuned for max TE at WOT - it wouldn't make sense to throw away the 10 - 15% additional power available from running 0.9 instead of 1.2. The current F1 rules OTOH are all about TE. They are tuned to max TE.


Edited by gruntguru, 02 January 2015 - 07:45.


#6 GdB

GdB
  • New Member

  • 22 posts
  • Joined: December 14

Posted 31 December 2014 - 22:46

Perhaps HLSI  and or D-EGR will be applied to F1:

 

http://www.greencarc...40428-hlsi.html

 

The key is working with only DI.



#7 gruntguru

gruntguru
  • Member

  • 7,635 posts
  • Joined: January 09

Posted 01 January 2015 - 04:33

. . . only DI.

:D

Not too many people complaining about that although I agree HLSI is difficult with DI. The technology is pretty good though and no doubt the "rich zone" is fairly "homogeneous". Best of both worlds - homogeneous, stratified charge.  :lol:



#8 bigleagueslider

bigleagueslider
  • Member

  • 1,235 posts
  • Joined: March 11

Posted 02 January 2015 - 04:36

If F1 is trying to promote an image of better efficiency, then why be concerned about how much air passes thru the engine? The cars are powered by extracting the chemical energy contained in the fuel via combustion.

 

Regarding gruntguru's comment about the BTE of F1 powerplants being 40%+, if true that is quite impressive.



#9 gruntguru

gruntguru
  • Member

  • 7,635 posts
  • Joined: January 09

Posted 02 January 2015 - 07:47

http://www.bbc.com/s...rmula1/26943423
"The new F1 engines have a thermal efficiency of "40% and above" - better than that of a road-going diesel." - Professor Dr Thomas Weber, the Daimler board member responsible for research and development.


Edited by gruntguru, 02 January 2015 - 07:47.


#10 7MGTEsup

7MGTEsup
  • Member

  • 2,466 posts
  • Joined: March 11

Posted 09 January 2015 - 11:29

Regarding the RA168E you have boost pressure down as MAP 2.5bar I was under the impression that it was 2.5 bar gauge not absolute as in 3.5 bar MAP.



#11 gruntguru

gruntguru
  • Member

  • 7,635 posts
  • Joined: January 09

Posted 10 January 2015 - 09:02

It was 2.5 bar absolute AFAIK.