Estimate of THE 7:29 GT-Rs power based on Nurburgring straight sections

[swamp2]

Originally Posted by bruce.augenstein@comcast.

The SAE says that, at a given boost level, forced induction engines gain and lose more power due to ambient temperature and humidity than normally aspirated engines do, but they lose less due to altitude (which is obvious if you think about it*). However, turbo engines now do even better at altitude because they typically compensate for altitude (up to a point) by simply dialing up the boost accordingly.

Isn't even the first part of this claim a bit irrelevant since all modern FI (and NA) systems adjust for temperature? I don't know if they adjust for humidity but I think that is a 2nd order effect compared to both temp and altitude.

[swamp2]

Originally Posted by footie

And this comment from the man who has posted here almost 12000 times.

How it cooking hairy palms. (j/k)

Thanks foot! I think that he has reset his post count more than once so 24000 may be more accurate!

Mantis: What is the real number?

As well I would rather post meaningful and thought provoking posts, no matter how long rather than post all nonsense, opinion, things on cosmetic mods and "socializing" posts.

Just FYI I am recently and happily married and don't have any problems there as you allude to thank you very much.

[bruce.augenstein@comcast.]

Originally Posted by swamp2

Isn't even the first part of this claim a bit irrelevant since all modern FI (and NA) systems adjust for temperature? I don't know if they adjust for humidity but I think that is a 2nd order effect compared to both temp and altitude.

As per the norm for us, we're talking apples and oranges here.

Of course they all adjust for temp, pressure and humidity, but the SAE says (in standard J1349) that a forced induction engine gains and loses proportionately more when it does so in response to those conditions. Increased temperature, for instance, may lead to more aggressive spark retard (and possible reduction of boost in turbo engines) than would be the case with a normally aspirated engine. So, when a manufacturer puts a test engine on a stand to test for SAE Certified power, they are allowed a greater "adjustment" with forced induction to get to SAE Net standards when the air inlet temp is, say, 100 degrees F (SAE standard is 70something degrees). So, 100 HP observed may become, say, 110 adjusted in a normally aspirated engine under whatever hot/humid conditions, and 111 or 112 in a forced induction motor under those same conditions.

In other words, footie is correct.

Bruce

[T Bone]

Originally Posted by swamp2

Isn't even the first part of this claim a bit irrelevant since all modern FI (and NA) systems adjust for temperature? I don't know if they adjust for humidity but I think that is a 2nd order effect compared to both temp and altitude.

Humidity and ambient temp affect intercooler efficiency which has a big impact on FI engine performance....hence the first part.... FI engines lose / gain more power depending on ambient conditions. (Why Gustav wants to rerun the RS6 vs. the M5 ..... The RS6 beat the M5 in very cold conditions but Swedish summers kill FI performance - listen to all the AMG sl / e / s 55 people)

Altitude, as Bruce mentions, is easily adaptible by boost control (through the wastegate).

If I lived in the Rockies, I would never own anything but a turbo car.

[swamp2]

Originally Posted by bruce.augenstein@comcast.

In other words, footie is correct.

Apples oranges, whatever. The point seems totally irrelevant to discuss something you can not actually buy anywhere, i.e. a modern FI engine without such controls/systems/software in place. Talk about nitpciking. To make sure we are crystal clear I will refine my previous statement to be.

"Any modern turbo engine system is much better in dealing with issues related to input air "quality" such as temperature and density variations as opposed to modern NA engine systems."

I think the meaning of the above is truly an incredibly tiny leap of faith/assumption away from my original wording. However, I do appreciate your added clarification.

[footie]

Sorry for turning this debate on it's head. But then maybe the official discussion had run it's course anyway.

I like TB and Bruce, thought that turbos show wide variation in climatic conditions but as I had no hard data as proof I didn't want to push the argument further.

[footie]

None of this really answers the question of why the humble S3 with it's modest awd system by the GTR's standards, not mind blowing acceleration by similar PTW 335i's standards, poor weight balance, manual gearbox, normal suspension and tyres can pull off the kind of result only the GTR can better.

In fact this puts more credibility to the stock GTR claims of a 7:29 and the car not being a ringer after all.

[bruce.augenstein@comcast.]

Originally Posted by swamp2

Apples oranges, whatever. The point seems totally irrelevant to discuss something you can not actually buy anywhere, i.e. a modern FI engine without such controls/systems/software in place. Talk about nitpciking. To make sure we are crystal clear I will refine my previous statement to be.

"Any modern turbo engine system is much better in dealing with issues related to input air "quality" such as temperature and density variations as opposed to modern NA engine systems."

I think the meaning of the above is truly an incredibly tiny leap of faith/assumption away from my original wording. However, I do appreciate your added clarification.

I'll try to get the point across another way.

At any given altitude or barometer reading, a forced induction engine will gain or lose proportionally more power than a normally aspirated engine will as temperatures change. Therefore, according to SAE Standard J1349, your statement in quotes (above) is partially incorrect. Turbos are comparatively lousy at dealing with high temperatures, but go like striped apes when it's cold.

We have all noticed this anyway, but the SAE made it official with the revised net power standard published in 1970, I believe.

Bruce

[swamp2]

Originally Posted by footie

None of this really answers the question of why the humble S3 with it's modest awd system by the GTR's standards, not mind blowing acceleration by similar PTW 335i's standards, poor weight balance, manual gearbox, normal suspension and tyres can pull off the kind of result only the GTR can better.

In fact this puts more credibility to the stock GTR claims of a 7:29 and the car not being a ringer after all.

I would call this an argument ad absurdum. You can not prove a model invalid nor justify the over performance of particular vehicle with the existence of another single outlier. There are many aspects of the GT-R that contribute to it being an over performer and indeed there are not many features that contribute to the S3 being an over performer. If you think this in any way affects the validity of the regression model or of the GT-Rs N'Ring time occuring with 100% accurate specs you are way off base.

Outliers point to many possibilities: mistakes, low probability events, immense alignment of many improbable events, false specifications, a ringer, etc.

I would bet if one bothered to dig closely and deeply enough the S3 over performance could either be explained or shown to be not repeatable.

[footie]

Originally Posted by swamp2

I would call this an argument ad absurdum. You can not prove a model invalid nor justify the over performance of particular vehicle with the existence of another single outlier. There are many aspects of the GT-R that contribute to it being an over performer and indeed there are not many features that contribute to the S3 being an over performer. If you think this in any way affects the validity of the regression model or of the GT-Rs N'Ring time occuring with 100% accurate specs you are way off base.

Outliers point to many possibilities: mistakes, low probability events, immense alignment of many improbable events, false specifications, a ringer, etc.

I would bet if one bothered to dig closely and deeply enough the S3 over performance could either be explained or shown to be not repeatable.

Actually the S3 does perform similarly to it competitors in the same way as the GTR does. Meaning it almost always win and by quite a margin.

I know the S3 is a bit of an unknown Stateside but over here it's a car with respected abilities.

[swamp2]

Originally Posted by footie

Actually the S3 does perform similarly to it competitors in the same way as the GTR does. Meaning it almost always win and by quite a margin.

Sure, agreed. But like you said (when you listed how the car is a bit "bland") you can not point to a list of specs or features that would tell you the car will over perform. With the GT-R you easily can.

[footie]

Originally Posted by swamp2

I would call this an argument ad absurdum. You can not prove a model invalid nor justify the over performance of particular vehicle with the existence of another single outlier. There are many aspects of the GT-R that contribute to it being an over performer and indeed there are not many features that contribute to the S3 being an over performer. If you think this in any way affects the validity of the regression model or of the GT-Rs N'Ring time occuring with 100% accurate specs you are way off base.

Can you explain this one to me?

I see you using examples all through these thread highlighting other cars as proof of the GTR's under rating. Though I agree that the GTR is on the outer limits of the regression simulations, the same is too for the S3 to a lesser degree but it isn't equipped with anything like the trick gear that the Nissan is abound with.

Compare the two cars.
Cars :GTR vs S3
Suspension : Race developed active Bilsteins vs stock springs and dampers
Gearbox : Dual Clutch transaxle vs 6speed manual
Bodyshell : Ultra stiff shell with CF cross members vs stock A3
AWD system : full active with LSDs vs Haldex fwd biased, no LSD and only 50% shift to the rear
Tyres : nitrogen-filled Dunlop r-compound vs normal summer tyres similar to PS2

If anything the S3 looks to be the car which is the more surprising when most of the regression data uses PTW as it's main factor.

Originally Posted by swamp2

Outliers point to many possibilities: mistakes, low probability events, immense alignment of many improbable events, false specifications, a ringer, etc.

I would bet if one bothered to dig closely and deeply enough the S3 over performance could either be explained or shown to be not repeatable.

Remember the S3's data was supplied by Sport Auto. To call the S3's test data as a mistake or a ringer is really called SA's credibility into question. They are no fools, they would know a ringer when they see one. And on any test I have seen the S3 does out perform what should be expected, but no more so than what the GTR seems to do.

[swamp2]

Originally Posted by footie

Can you explain this one to me?
...

You keep repeating the same thing over and over.

We both fully agree that the S3 is an over performer given its very modest specifications. I provided a long (but definitely incomplete) list of how/why the S3 may have obtained such a good N'Ring lap time. Although mistakes are possible, I did not state nor directly imply that SA made a mistake in this test. As well, the mistake may be something way much more subtle than simply reading a timer wrong. But of course this or any other variety of mistakes are possible. You know the expression, "shit happens" and it does. Look at the sample size (number of data points), when this gets large the chances of at least one value being an error in some way increases. That is plain, simple, indisputable. Outliers can arise from a variety of statistical or deterministic effects or errors (and of course from a combination of these) and regression is a great way to spot these and then they can be examined on a case by case basis. Interpreting/understanding/explaining the outliers is part art, part science.

Again, I do not have the answer as to how the S3 seems to over perform so greatly on such modest specs. I do believe the sigma number points to something pretty significant though that we are not aware of or something missing. Regression analysis is just a tool, it tells us the S3 is "interesting" but can not reveal the answer. The validity of this particular regression model is more in its high R^2 value rather than its determination of particular +1, +2 or -1 or -2, etc. sigma outliers. Unless you have R^2 very very close to 1 (a perfect mathematical fit of the data to a line or curve) and/or a very small number of samples you will regularly get sigmas greater than and less than 2. Greater than 3 is the massive outlier we talked about before and I explained what the chances are of such events.

I really hope this is clear I am trying to be crystal clear.

[footie]

The best explanation I think was the saying 'shit happens'.