M4 dyno test by Sportauto 465hp/577 nm

[grimlock]

Originally Posted by 325rider

seriously. cliffs. not trolling

Crank HP is 465PS
'Scientifically' adjusted for drivetrain losses, it is 370PS
(note: drivetrain losses increase with RPM, as friction)

So its appears BMW is spot on with their official number.

Alot of the assumption that BMW must ALWAYS understate their turbo engines HP can perhaps be attributed to the early N54 (?) or certain maps that gave a 'bonus' - but it would be wishful thinking to assume it automatically implies on all subsequent turbo engines, especially as they squeeze more output /liter.

This is exacerbated by less scientific "dyno runs" in which neither the equipment nor method is well understood.

[aus]

Originally Posted by grimlock

Crank HP is 465PS
'Scientifically' adjusted for drivetrain losses, it is 370PS
(note: drivetrain losses increase with RPM, as friction)

So its appears BMW is spot on with their official number.

Alot of the assumption that BMW must ALWAYS understate their turbo engines HP can perhaps be attributed to the early N54 (?) or certain maps that gave a 'bonus' - but it would be wishful thinking to assume it automatically implies on all subsequent turbo engines, especially as they squeeze more output /liter.

This is exacerbated by less scientific "dyno runs" in which neither the equipment nor method is well understood.

How do you explain how damn fast the turbo cars are relative to their HP rating??

.

[grimlock]

Originally Posted by aus

How do you explain how damn fast the turbo cars are relative to their HP rating??

.

Because the HP comes on much higher at lower RPM.. like S65 gets 30% of max power at 3000rpm versus 60% for s55 (random numbers, I do not know the actual numbers) so in a given period of time, the S65 only works at 60% or max power versus 80% for S55 (again, made up numbers).

Also, turbo 'feels' faster because the change in HP/time (steepness of the HP curve) is more abrupt, so the kick in the pants feeling is much greater.
In NA, the pull is steady and longer so it doesn't feel as fast, but it's stealthy fast..

Also, for two turbo engines with the same HP rating the turbo car will still be faster because the HP curve is nearer to the max for longer, like in the S55 you shift at 7200 and it goes down to 5500 or whatever but it's still producing max HP, wheras in the s65 you shift at 8300 or whatever and it drops to 6700 but it only makes 75% of max HP at that number - so time spent at the stated maximum HP is longer for turbo cars, as a product of the power curves, versus NA, which is not plateau but still increasing into max HP.
So this makes turbo faster, but NA feels better.

[swamp2]

Originally Posted by grimlock

Alot of the assumption that BMW must ALWAYS understate their turbo engines HP can perhaps be attributed to the early N54 (?) or certain maps that gave a 'bonus' - but it would be wishful thinking to assume it automatically implies on all subsequent turbo engines, especially as they squeeze more output /liter.

I've never heard of the common acknowledgement that many BMW turbo engines (and many other turbo engines) are underrated as being due to this bit of N54 trivia. It is about how fast the cars are. The E9X M3 is just about right on the money for it's performance given it's stated power level.

Does BMW also not significantly underrate many of their performance figures? Most notorious is their specification of extremely conservative 0-60 numbers.

[swamp2]

Originally Posted by grimlock

Because the HP comes on much higher at lower RPM.. like S65 gets 30% of max power at 3000rpm versus 60% for s55 (random numbers, I do not know the actual numbers) so in a given period of time, the S65 only works at 60% or max power versus 80% for S55 (again, made up numbers).

This is not the answer.

Although the shapes of the power/torque curves have some relevance to observed performance (and feel) it is much more governed by peak power. In addition one can accurately account for these effects in simulation. Again the E92 M3 performs in a way such that the best/fastest observed results are very consistent with simulations that take into account its exact power/torque curves and peak values. The same is true for the F8X M3/4, but ONLY with a significant under rating, to the tune of at least 20-30 hp. Simulation allows us to make this very careful controlled "apples to apples" comparison. The same is true for the 335i, 1M, F10 M5 and other turbo vehicles from other manufacturers as well. The first year GT-R was another classic case of under rating. The car was simply faster than it's stated hp could deliver.

It is still a very interesting question that no one has been able to fully answer. Why does BMW (and other) under rate and why does it appear to be much more common on turbo vehicles.

[grimlock]

Originally Posted by swamp2

I've never heard of the common acknowledgement that many BMW turbo engines (and many other turbo engines) are underrated as being due to this bit of N54 trivia. It is about how fast the cars are. The E9X M3 is just about right on the money for it's performance given it's stated power level.

Does BMW also not significantly underrate many of their performance figures, most notorious is their specification of extremely conservative 0-60 numbers.

I don't know anything about these cars/engines compared to you guys, but I was commenting as an outsider from the perspective of science/common sense.

It's quite ridiculous the N55 is only rated at 300PS, but it's probably true because of the way "max HP" can be manipulated in turbo engines. Just cut the top off and make it a long flat plateau, that way it really is no more than 300PS but it stays there from like 4700-6000rpm

"Max HP" is just the highest point in the power graph, even if it only reaches that at 8300rpm and 500rpm either side of it is only 87% of it..

Power in turbo cars is like a steady burn, and NA is more like a brief explosion.

0-60 performance times are so prone to human error.. try pressing 0.5s on a stopwatch and it's probably the time it takes you to fully depress the pedal.

[grimlock]

Originally Posted by swamp2

This is not the answer.

Although the shapes of the power/torque curves have some relevance to observed performance (and feel) it is much more governed by peak power. In addition one can accurately account for these effects in simulation. Again the E92 M3 performs in a way such that the best/fastest observed results are very consistent with simulations that take into account its exact power/torque curves and peak values. The same is true for the F8X M3/4, but ONLY with a significant under rating, to the tune of at least 20-30 hp. Simulation allows us to make this very careful controlled "apples to apples" comparison. The same is true for the 335i, 1M, F10 M5 and other turbo vehicles from other manufacturers as well. The first year GT-R was another classic case of under rating. The car was simply faster than it's stated hp could deliver.

It is still a very interesting question that no one has been able to fully answer. Why does BMW (and other) under rate and why does it appear to be much more common on turbo vehicles.

Ok, I can't refute that turbo cars are underrated as I don't have a dyno.. but all we have to go on are the more scientific dynos, and some less scientific ones.
If you have calculated it's not possible to achieve such times without more power, than you may be correct.. it is so easy to get more power from a turbo, why wouldn't they?
At the end of the day we are trying to explain performance, higher crank HP might be the reason - but I was just trying to debunk the unscientific dynos out there. I don't know how this one fits in the picture, perhaps this one that shows less power is the unscientific/incorrect one?
I am not stating any conclusions, just debating the method.

Here is one example N20 vs N52 in the F10, the N20 is rated at 245 vs 240, but dynos 240vs208.
http://f10.5post.com/forums/showthread.php?t=682146
You are correct, I cannot explain this difference in turbo vs NA dynos.

[grimlock]

Originally Posted by swamp2

It is still a very interesting question that no one has been able to fully answer. Why does BMW (and other) under rate and why does it appear to be much more common on turbo vehicles.

One explaination is that NA power is hard to come by, so you naturally want to make it seem bigger.
But Turbo power is like stealing - its cheap and easy so you don't want to admit to it and so naturally understate it.

[swamp2]

Originally Posted by grimlock

Ok, I can't refute that turbo cars are underrated as I don't have a dyno.. but all we have to go on are the more scientific dynos, and some less scientific ones.
If you have calculated it's not possible to achieve such times without more power, than you may be correct.. it is so easy to get more power from a turbo, why wouldn't they?
At the end of the day we are trying to explain performance, higher crank HP might be the reason - but I was just trying to debunk the unscientific dynos out there. I don't know how this one fits in the picture, perhaps this one that shows less power is the unscientific/incorrect one?
I am not stating any conclusions, just debating the method.

Unfortunately dynos are not the final answer. A "scientific" dyno is almost a non-sequitur...

Although rri.se has some accurate equipment and Maha certainly makes some nicer, more thorough and more accurate equipment than Mustang and Dynojet products, there is simply too much variation in the results from dynos. This is due to a wide variety of both deterministic and random errors present in the equipment, calibration, set up, operation and post processing of dyno data.

Have a look here. Look for stock vehicles and sort by increasing or decreasing hp. Do you think a stock E9X M3 can have anywhere between 312-365 whp? That is about +/-8% or a 16% range. These engines are probably putting out a fixed amount of power within a couple of percent!

Dyno's are a very reasonable tool when used carefully for a before/after estimate of a change in power from a modification(s). Other than this, I trust them about as far as I can throw them...

If you can't trust a manufacturers specification (and about the only time I do is when they quote their power using the SAE Certified Power standard - GM does this a lot) and you can't trust a wheel/hub dyno what is left? We have measured performance results and physics based performance simulation. Often impressive agreement can be obtained between the two.

Both I and other forum members have demonstrated this using at least two different simulation software tools across a wide range of vehicles. If a significant discrepancy is observed between measured and simulated results it is a pretty safe bet that the vehicle is under rated (or perhaps sometimes over rated). One can then easily vary the input power (again the full curve) to match observed test results in this way actually quantifying a level of under rating. Now that being said both test results and simulation results can show a wide variation in results. To prove the former point have a look here at this "database" of measured E9X M3 test results. Generally is makes good sense to compare the best achieved test results vs. the simulation. Parasitic losses in simulation, and other subtleties in getting "correct" inputs certainly inject some uncertainty into simulation. Nothing is perfect and absolute here with this many variables. However, it is my evaluation that this combination of the best observed test results (of course less any outlier types of best result) together with simulation are the most sensible way for enthusiasts to judge a manufacturers power claims.

[swamp2]

Originally Posted by grimlock

One explaination is that NA power is hard to come by, so you naturally want to make it seem bigger.
But Turbo power is like stealing - its cheap and easy so you don't want to admit to it and so naturally understate it.

I think that is a gross simplification and is not an adequate explanation.

Now that being said there is probably a shred of truth here. Under rating makes the entire rest of the car look more impressive; great chassis, good traction, great gear box, great tires, top notch differential, amazing traction control software, etc. all "conspired" to give the car 450 hp worth of performance when the manufacturer says it only has 400 hp...

[M4TW]

Originally Posted by swamp2

I think that is a gross simplification and is not an adequate explanation.

Now that being said there is probably a shred of truth here. Under rating makes the entire rest of the car look more impressive; great chassis, good traction, great gear box, great tires, top notch differential, amazing traction control software, etc. all "conspired" to give the car 450 hp worth of performance when the manufacturer says it only has 400 hp...

Normally I can follow along with what you are saying Swamp - even if I disagree with you. But I must say. That doesn't make a lick of sense. There are a myriad of reasons why BMW might want to err on the side of under-rating HP - one of them being to fool their competition into benchmarking the wrong number. To use an Occam's razor approach, however, it may simply be the methodology with which BMW uses to rate HP in it's turbo charged engines.

I'm just glad that they don't err in the wrong direction.

[turbo8765]

Originally Posted by swamp2

It is still a very interesting question that no one has been able to fully answer. Why does BMW (and other) under rate and why does it appear to be much more common on turbo vehicles.

The answer is that they don't. That is the explanation.

Straight line performance vs Power:weight has steadily increased due to gradual, incremental improvements in some or all of the following parameters (depending on the vehicle in question).

Suspension
Powertrain Efficiency
Decreased Shift Times
Greater number of gears and better optimized gear ratios
Tires
TC and LC Systems
AUC Torque
Aerodynamics

The notion that understating power would lead a potential owner to go "wow this is much faster than the rated power would seem to suggest… and I am therefore more likely to buy it" is ludicrous. Mostly because, as we all know, butt dyno's are SUPER inaccurate. If that were the technique the manufacturer were attempting to use they'd be better served by simply installing a louder exhaust. That adds +10% to the butt dyno reading.

Perhaps a more objective owner wonders "wow, my calculations indicate the car should trap 117 based on the rated power but it somehow it traps 119… I was on the fence at 117 but I'm sold at 119" is nearly equally ridiculous.

Most buyers aren't bimmerpost regulars. They're not buying cars based on the delta between anticipated butt dyno output and measured butt dyno output. Nor are they buying cars based on trap speed.

If anything, BMW may be more likely to lose potential buyers by understating power. Most buyers are technically unsophisticated… If the MB salesman can say "But the C63 makes XX more hp than the M3/4…..

All that aside, the bottom line is the BMW are not lying to you and gain nothing by lying to you. Manufacturers, BMW included, aren not pulling HP numbers out of their collective arses.

Can you imagine the meeting at BMW….

"We know that the s55 makes X power, but we must say it makes less. Despite the fact that we built a reputation based on accurate and precise German Engineering, we must in the case be intentionally inaccurate"

"Brilliant Wilhelm… but precisely how inaccurate should we be?"

"Market research suggest that understating hp by 7.2% would optimize test drive butt dyno impressions relative to stated horsepower"

"Wilhelm… you're a genius"

I joke… but you have to wonder, under your supposition of an underrating, how did they determine the extent to which it would be underrated?

In the end, which is more likely to be wrong… your analysis software OR BMW? I hear they have a couple guys at BMW that are actually pretty good at the math.

[CanAutM3]

Originally Posted by swamp2

furthermore the alternator is clutched open (mechanically or electrically...) on a WOT (i.e. dyno type of acceleration). Hence it uses almost nothing (less belt/pulley perhaps)! I believe something similar happens with the AC.

That is the whole point of my speculation. When run on a bench dyno at steady state during the "official" rating by the manufacturer, the accessories need to be engaged.

However, on a chassis dyno operated in transient (under acceleration) in the field, the accessories are disengaged by the ECU, hence a higher power reading.

Anyway, it is just a theory that could explain some of the underrating. IMO, if underrating there is, I believe it has to do more with the testing methodology/standard than a marketing ploy.

[CanAutM3]

Originally Posted by coloradoe92m3

I'm not sure the fuss
1. It will make 380-400 who
2. It will feel like the current m5 exactly. Perception of "running out of breath" is from a flat power curve. Flat means no increasing rate of power which means once it hits peak HP, it accelerates consistently at that rate. This creates the feeling of less acceleration given there is no linear slope. S65 classically pulls to redline since up until redline its increasing level of acceleration and truly is accelerating the car faster as you near redline. We feel a change in acceleration but once acceleration goes steady velocity, we perceive that as less power.

I think I understand what you mean, but your explanation is not quite correct.

Acceleration rate closely matches the torque curve, not the power curve.

When a turbo car hits the power plateau, the torque curve starts decreasing, hence so does the acceleration rate. The acceleration keeps on decreasing until redline, hence the feeling that it is running out of breath. On the S65, since the torque curve does not decrease much up to redline, the acceleration rate is almost sustained giving this feeling of pulling all the way to the top of the tach.

[CanAutM3]

Originally Posted by grimlock

P-Schlepp (drivetrain losses) is much greater at 8300 vs 5800rpm.

That is the flaw in your theory, right there.

P-Schlepp does not equate to drivetrain losses. It is the overall drag measured to calculate flywheel power. P-Schlepp also includes the power to overcome the dyno inertia and losses.

, it'a all there .

[AreOut]

Originally Posted by aus

How do you explain how damn fast the turbo cars are relative to their HP rating??

.

because it's only maximum HP rating which is irrelevant if two torque curves aren't equal...

for example F80 has maximum of 370 whp but during the acceleration it will put down an average of 330 whp, while E90 with its maximum 360 whp will put down an average of 300 whp(just an example the number will differ a bit)

now it would be another story if the gearbox was CVT in both cars and it would accelerate always on maximum power rpm

[grimlock]

Originally Posted by CanAutM3

That is the flaw in your theory, right there.

P-Schlepp does not equate to drivetrain losses. It is the overall drag measured to calculate flywheel power. P-Schlepp also includes the power to overcome the dyno inertia and losses.

, it'a all there .

you're right, I deleted my original message to avoid confusing anyone.
cheers.

[w3rkn]

Charts that do not cross over @ 5250 are annoying.

There is zero reason for this, there is zero reason not to use the same scaling for TQ as HP..

[MKE_M3]

I believe there is a very obvious reason for under stating turbo HP. External variables have a much greater impact on turbo cars, especially temperature. I don't think BMW is understating HP. They are just providing a HP number that their customers should expect to obtain under adverse conditions. A reasonable floor for HP.

What is the HP on a hot day, with accessories engaged on a car that is in the low end of manufacturing tolerance for power?

Just my 2 cents. It's a more legally defensive/conservative approach. I hold little credence in marketing conspiracy theories.

Also, with regards to SWAMP's comment about "scientific dyno's," I disagree. An unscientific dyno is an oxymoron. Dyno are tools for measuring. Can't get more scientific than that. It's how they are commonly used that is unscientific . I think we'll agree about that.

[Soorena]

Originally Posted by w3rkn

Charts that do not cross over @ 5250 are annoying.

There is zero reason for this, there is zero reason not to use the same scaling for TQ as HP..

HP in PS and TQ in NM don't cross each other at 5252. They do that in ~3900 RPM.

[M4TW]

The marketing conspiracy theory is easy to validate/debunk:

1. Are all German turbo cars apparently under-rated?

Yes - the methodology employed by German car manufacturers to rate crank HP is responsible for the apparent anomaly not a marketing ploy.

No -

2. Are all of BMW's turbo models under-rated by the same degree?

Yes - the methodology employed by BMW to rate crank HP is responsible for the apparent anomaly not a marketing ploy.

No -

3. It's a marketing ploy.

[bradleyland]

Originally Posted by swamp2

Unfortunately dynos are not the final answer. A "scientific" dyno is almost a non-sequitur...

Only, I would scope that statement to chassis dynamometers. A good water brake engine dynamometer can provide accurate steady-state power measurement for extended periods, which is what is needed to collect enough data to arrive at an accurate power rating. I have a feeling you meant chassis dynos though.

The MAHA chassis dyno is one of the better chassis dynos (I hear it's capable of steady-state testing as well), but there are still too many confounding factors with chassis dynos on the whole. If you want to look at measured wheel horsepower, you're constrained to comparing dynamometers of the exact same type and configuration; otherwise confounding factors clobber any meaningful result.