Jump to content


Photo
- - - - -

Race Engine Tuning


  • Please log in to reply
19 replies to this topic

#1 blueprint2002

blueprint2002
  • Member

  • 113 posts
  • Joined: May 19

Posted 30 September 2020 - 10:54

In the marine world, the term “tuning” has a somewhat different meaning to what is understood by the same term in the automotive world.

Apart from the small engines used in the common sort of pleasure craft, most marine diesels are fairly large and correspondingly powerful, and represent proportionate financial investment. Perhaps for this reason, the engines are so equipped as to enable the owner to establish, when thought fit, that they are delivering the rated power, as they did when new. Two factors make this possible: each cylinder has its own high-pressure fuel injection pump, and also has a fitting to which may be attached a “peak pressure” gauge, or an “indicator” which traces out a graph of cylinder pressure vs crank angle. When the latter indicates that a certain cylinder is below par, its fuel pump may be adjusted as needed to correct the deficiency. In any case, every cylinder has its own exhaust temperature gauge permanently in place, which can give early warning if all is not well, and that “tuning” may be needed.

(Recent engines with "common-rail" injection systems are somewhat different, but the same principle applies).

I have been wondering what the builders of F1 engines do to ensure that each cylinder is performing with equal efficiency. Compression pressure measurements tell only part of the story, because it is the actual combustion process that determines cylinder output, this in turn depending on flow through the valves, turbulence, timing of both injection and spark, spray pattern and droplet size, and so on. And of course, the dynamometer can only give the sum total power developed, not how that is shared by the cylinders.

On the other hand, when seeking the ultimate power output, it makes sense to ensure that every cylinder performs as intended, instead of varying randomly in response to the dozens of factors that may have an effect. Apart from the few mentioned in the previous paragraph, the airbox flow pattern, tolerances in every part that figures in the gas flow path, surface finishes, lube oil flow quantity and pattern…….

Anyone have any idea of the state of the art in this field?

 



Advertisement

#2 Greg Locock

Greg Locock
  • Member

  • 5,848 posts
  • Joined: March 03

Posted 30 September 2020 - 21:12

You fit a cylinder pressure sensor. 



#3 gruntguru

gruntguru
  • Member

  • 7,051 posts
  • Joined: January 09

Posted 01 October 2020 - 00:29

AFAIK all current F1 engines run a pressure sensor on all cylinders.



#4 blueprint2002

blueprint2002
  • Member

  • 113 posts
  • Joined: May 19

Posted 02 October 2020 - 09:58

You fit a cylinder pressure sensor. 

 

 

AFAIK all current F1 engines run a pressure sensor on all cylinders.

Thanks for the quick responses. Would you know

1. Are the sensor outputs continuously monitored when the engine is running, on the test bed or out on the track?

2. At say 15000 RPM, each cylinder performs 125 cycles per second and combustion occupies about a quarter of each cycle, so these sensors would need practically instantaneous response to give a worthwhile output. Any idea what sort of sensors they are? Nothing mechanical could possibly do that, so what could? Piezoelectric?

3. What are typical peak pressures in an F1 engine cylinder today? 



#5 desmo

desmo
  • Tech Forum Host

  • 21,475 posts
  • Joined: January 00

Posted 02 October 2020 - 13:43

I think you'll find even dimensionless PV loops are unobtanium in F1.



#6 Canuck

Canuck
  • Member

  • 2,103 posts
  • Joined: March 05

Posted 02 October 2020 - 19:09

1) Yes, definitely yes, possibly yes

2) 15,000 RPM = 250 rotations / second which is 90,000 crank degrees per second.You need to sample beyond the resolution of the data points you want, so say 5x oversampling. That's 450 kHz which is absolutely trivial on the processor end and within the realm of a piezo-based sensor with adequate circuit design.

3) ?



#7 Greg Locock

Greg Locock
  • Member

  • 5,848 posts
  • Joined: March 03

Posted 02 October 2020 - 23:24

They are basically microphones. https://www.kistler....callee=frontend



#8 gruntguru

gruntguru
  • Member

  • 7,051 posts
  • Joined: January 09

Posted 04 October 2020 - 01:55

1) Yes, definitely yes, possibly yes

2) 15,000 RPM = 250 rotations / second which is 90,000 crank degrees per second.You need to sample beyond the resolution of the data points you want, so say 5x oversampling. That's 450 kHz which is absolutely trivial on the processor end and within the realm of a piezo-based sensor with adequate circuit design.

3) ?

3) I think it was 4 elephants on the piston according to Cowell. Assuming they are adult elephants it works out to be a very high pressure (400 bar +) - regardless of whether they are African or Indian.



#9 Canuck

Canuck
  • Member

  • 2,103 posts
  • Joined: March 05

Posted 04 October 2020 - 04:19

...how??  What's the math required to develop 400 bar cylinder pressure? We can make educated guesses about the combustion chamber volume, so what quantity of fuel burning is required to develop that? That's enormous.



#10 Greg Locock

Greg Locock
  • Member

  • 5,848 posts
  • Joined: March 03

Posted 04 October 2020 - 05:13

peak not bmep, maybe?


Edited by Greg Locock, 04 October 2020 - 05:23.


#11 Kelpiecross

Kelpiecross
  • Member

  • 1,662 posts
  • Joined: October 10

Posted 04 October 2020 - 05:18

 I think the peak cylinder pressure is more  like  1000 to 1500psi at most.  400 bar is about 6000psi.  F1 engines have only much the same peak pressures as a road car - only they do a lot more of  them a lot quicker.     



#12 Charlieman

Charlieman
  • Member

  • 2,000 posts
  • Joined: October 09

Posted 04 October 2020 - 11:55

We have to consider repeatability of manufacture. Engines are designed, redesigned and during the development stage their long term performance attributes can be assessed. Some people advise whether the design can be built consistently, and they do an incredible job.

 

F1 has had a decade or more of rules limiting the number of engines per season. Every engine has to deliver power on a consistent life time scale, power being determined using several measures. You can't fit a measuring device to everything which you wish to monitor.

 

Mission control style monitoring is clearly beneficial, but it is better to get things right in the first place.



#13 gruntguru

gruntguru
  • Member

  • 7,051 posts
  • Joined: January 09

Posted 05 October 2020 - 02:05

Andy Cowell stated in 2016 that the PCP (peak cylinder pressure) was equivalent to having 4 elephants standing on the piston. Assuming 5 ton elephants gives about 400 bar. Modern diesel engines operate at over 200 bar. F1 engines are high compression (18:1) high boost (4 bar abs) and very lean (2) all similar to diesel - but with rapid combustion.

 

I think somewhere between 300 and 400 is realistic. 



#14 Fat Boy

Fat Boy
  • Member

  • 2,334 posts
  • Joined: January 04

Posted 06 October 2020 - 17:38

Andy Cowell stated in 2016 that the PCP (peak cylinder pressure) was equivalent to having 4 elephants standing on the piston. Assuming 5 ton elephants gives about 400 bar. Modern diesel engines operate at over 200 bar. F1 engines are high compression (18:1) high boost (4 bar abs) and very lean (2) all similar to diesel - but with rapid combustion.

 

I think somewhere between 300 and 400 is realistic. 

It seems like we should be able to come up with a better unit of measure than 'elephants on a piston.'



#15 Kelpiecross

Kelpiecross
  • Member

  • 1,662 posts
  • Joined: October 10

Posted 07 October 2020 - 03:22

How on earth do they make 18:1 CR work with 60 Psi boost?  



#16 malbear

malbear
  • Member

  • 293 posts
  • Joined: September 02

Posted 07 October 2020 - 08:28

How on earth do they make 18:1 CR work with 60 Psi boost?  

probably an enormous amount of overlap and late intake valve closure so much that the actual compression ratio after valves close probably about 12 to 1  just my 2 cents 



#17 Charlieman

Charlieman
  • Member

  • 2,000 posts
  • Joined: October 09

Posted 07 October 2020 - 12:16

Andy Cowell stated in 2016 that the PCP (peak cylinder pressure) was equivalent to having 4 elephants standing on the piston. 

It's one of those expressions, like the area of a football pitch or of Wales, or the volume of Loch Ness.

 

You have no idea what they are talking about and when converted into numbers using the best methodology, it is no more helpful.



#18 gruntguru

gruntguru
  • Member

  • 7,051 posts
  • Joined: January 09

Posted 07 October 2020 - 21:37

So you think Andy Cowell might have said four elephants when the reality is three? Or perhaps he meant four small elephants? Cowell doesn't strike me as that type.



#19 gruntguru

gruntguru
  • Member

  • 7,051 posts
  • Joined: January 09

Posted 07 October 2020 - 21:40

It seems like we should be able to come up with a better unit of measure than 'elephants on a piston.'

Obviously stated in those terms to convey the order of magnitude while limiting disclosure.



Advertisement

#20 Canuck

Canuck
  • Member

  • 2,103 posts
  • Joined: March 05

Posted 07 October 2020 - 21:58

This borders on argument from authority - Cowell is an authority in this domain, therefore what he said must be true (and not flippant, casual or otherwise misteading).

 

Again though, can't we work this backwards to determine it's accuracy? Or is peak cylinder pressure one of those measureable-but-uncalculable / unpredictable values?

 

If we have a mechanical compression ratio, displacement and speed, we should be able to determine within a margin of some error the bore, stroke and total compression chamber volume and volume delta with crank angle (but might go off the rails on the delta without rod lengths). We can take known engine efficiencies and turbocharging efficiencies to estimate air volume in that space and from that the maximum quantity of fuel that could be burned. That gives us the maximum theoretical energy that could be produced...I'm already in way past my depth here.


Edited by Canuck, 07 October 2020 - 21:59.