In automotive, most production components performance are varied probably from 10-15%. The company I work for now makes shock absorbers, and manufacturers' performance tolerance is about 10%(and imagine that on a car, where you can have one at the bottom end and one at the top end of the spectrum on the SAME car). The tighter you control that, the more expensive it'll be. Since there are no such thing as identical parts, they can be controlled to be as close as possible, from a statistical POV, but nothing will be the same. If you tighten the tolerance and control everything down to blueprint spec, then you should be able to achieve the expected designed performance. But then you are looking at tremandously escaladed cost.
Whilst RacingManiac is correct for the automotive industry as a whole, I think that with the tolerances involved in F1, the margin will be much reduced. Possibly to just a couple of percent.
Since the engines can all easily top the 19,000rpm limit, it will be all down to the torque curve up to that point. I should think they are all within 2-3% of each other.
Scotracer wrote:Whilst RacingManiac is correct for the automotive industry as a whole, I think that with the tolerances involved in F1, the margin will be much reduced. Possibly to just a couple of percent.
Since the engines can all easily top the 19,000rpm limit, it will be all down to the torque curve up to that point. I should think they are all within 2-3% of each other.
Scotracer, even 3% would still be 22hp! With the money and manufacturing processes that are in place, I think that even a 2-3% variance to be huge.
Just 5-8% does make quite a difference, but remember the excessively cold fuel issue that was heading for a punch-up last year? That was expected to have about 1% improvement. In tweaky road cars blue-printing an engine is building to close tolerances making the difference of a pig of an engine being rebuilt and running very nicely - reduced slop, backlash, no tight bearings. Some of these small changes make for big differences - lightened conrods and pistons raising practical rev limit, but lighten them by a further 1% and you might not have much engine left halfway through the race. 1% tweak could make 10% difference.
Dynos do not replicate results exactly and readouts are variable due to ambient temps and pressures.
Engines are different but I also doubt a 50bhp variance in the builds and specs. If 2 motors were built at the same time and dyno'd at the same time then you would expect that the lower output motor would be stripped, checked and fettled to ensure it achieved the projected and expected output.
I think this variance you have is a dyno issue mainly..
alexbarwell wrote:I am an engineer, not a conceptualist
I´m a monkey who only serve the servants
Then I´ll be an engineer but will keep serving the servants
Back on topic: Chris, in Argentinian F.Renault for example, the company that builds all the engines delivers them ramdomly, raffle between, just to make sure that a client gets an "extra powered" unit because of mere luck. These cars have about 200HP +/- 5HP
"You need great passion, because everything you do with great pleasure, you do well." -Juan Manuel Fangio
"I have no idols. I admire work, dedication and competence." -Ayrton Senna
alexbarwell wrote:Just 5-8% does make quite a difference, but remember the excessively cold fuel issue that was heading for a punch-up last year? That was expected to have about 1% improvement. In tweaky road cars blue-printing an engine is building to close tolerances making the difference of a pig of an engine being rebuilt and running very nicely - reduced slop, backlash, no tight bearings. Some of these small changes make for big differences - lightened conrods and pistons raising practical rev limit, but lighten them by a further 1% and you might not have much engine left halfway through the race. 1% tweak could make 10% difference.
I love your signature!
It is nice to inspire others to find their true selves...
Cosworth USA used to claim that their turbo V8 champ car engines (>700hp) all dyno'd out within 1% (7hp?)when new and after every rebuild. And over the years (or decades!), they probably manufactured and rebuilt thousands of those customer race engines. They had a spotless reputation, so I would not doubt that it is true.
Typically with recip. piston engines, an engine will have slightly more power after an initial run-in, due to the fact that the main contributors to friction losses (piston rings, piston skirts, crank bearing journals and cam lobes/followers) have all been polished-up and fit together more closely. Thus producing less friction loss and compression leakage.
F1 engines likely don't lose too much power during their short lives. The typical failure mode is usually a fatigue failure of one of the more highly stressed, high-number-of-load-cycle components, such as a con-rod or crank shaft. These components are designed as light as possible, thus they have absolutely no margin for fatigue life.
The F1 engine designers are very sophisticated in their design and analysis of these components now. In fact, with the telemetry and data logging they now have available, they can probably predict within a minute or two when a particular engine component is going to fail.
"Q: How do you make a small fortune in racing?
A: Start with a large one!"
A division of the company I work at makes >500-hp racing engines for an endurance series and I believe that they all are within maybe 5-hp on the dyno. I wouldn't be surprised if the 5-8% figure that was mentioned for the F1 engines wasn't 5-8 hp.
Each engine will be 'run-in' before ever leaving the factory, probably to the tune of around 100km (bath-tub failure curve). This will be both on static and full dynos. If there is more than one dyno it will be regularly calibrated. I would be amazed to find 5% difference across engines.
