C&D comparison: S6 vs M5 vs AMG E63

[adsgreen]

Just to clarify - I wasn't arguing against Arthur about torque as such.
Torque is good but you can't make any conclusion from a peak engine torque figure.
You can't do a huge amount more with peak power but it's more rpresentitive when comparing like with like.

The best picture is always painted with the full facts and in car cases this is the area under engine curve. A FI ESP a turbo will generally always score better when comparing with an NA . That's all.

Otherwise why the hell aren't we all drag racing diesels

Don't change at 8k - it works well up to 8250

[sakimano]

Just to clarify - I wasn't arguing against Arthur about torque as such.
Torque is good but you can't make any conclusion from a peak engine torque figure.
You can't do a huge amount more with peak power but it's more rpresentitive when comparing like with like.

The best picture is always painted with the full facts and in car cases this is the area under engine curve. A FI ESP a turbo will generally always score better when comparing with an NA . That's all.

Otherwise why the hell aren't we all drag racing diesels

Don't change at 8k - it works well up to 8250

Yeah, was just messing with you.

p.s. changing at 8k is slower. Changing at 8250 is even slower stil. Car starts to fall flat on its face at about 7500 so that's where I initiate my gear shifts (2-3, 3-4, 4-5)...I don't bother worrying about it in 1st as it's so quick/short/pulls nicely to redline.

so...NEW ARGUMENT!

Here's a dyno of a stock RS4, a pbox longitudinal acceleration in g plot, a VCDS RPM/second log in 3rd gear, and a VCDS torque plot that I brought into EXCEL and turned into a dyno of sorts,

All show the car struggling up top. All tell me to shift well before redline. So that's what I do.

Dyno



pbox g plot



VCDS log


VCDS torque log turned into a crude ft-lb and hp plot

[adsgreen]

No new argument.... You're just WRONG!

he he.

The key point here (and kinda relevant to my other argument) is that yes, torque is falling flat on it's face and it's much higher if you change gear and drop rpm. The problem is that the lower torque of the engine at 8000-8250 rpm coupled with the lower gear (say 2nd) still results in a higher torque at the wheels than changing to 3rd and dropping to say 5-6k rpm.
Here's a very rough graph to illustrate.


For gears 1-3 you want to go all the way up to the limiter (or as nears as you can without tripping it).
It's only from 4th upwards you can see that it is better to change into 5th before the redline (about 8100) as the torque at the wheels is better in the higher gear from that point.

[JCviggen]

Car starts to fall flat on its face at about 7500 so that's where I initiate my gear shifts
Dyno

The plot kind of contradicts your statement. At ~7800 its still making the same power as it was at ~6500. How is that falling flat? If you shift early you'll drop down to considerably lower horsepower. HP= the work being done. Less HP= less acceleration, it's pretty straight forward.
It's pretty obvious from that plot that you want to AVOID at all cost dropping down below 6K rpm, that's where the power dissapears.

[sakimano]

Car starts to fall flat on its face at about 7500 so that's where I initiate my gear shifts
Dyno
The plot kind of contradicts your statement. At ~7800 its still making the same power as it was at ~6500. How is that falling flat? If you shift early you'll drop down to considerably lower horsepower. HP= the work being done. Less HP= less acceleration, it's pretty straight forward.
It's pretty obvious from that plot that you want to AVOID at all cost dropping down below 6K rpm, that's where the power dissapears.

are you looking at the RS4 curve, or the R8 curve? The RS4 hp curve (red) clearly starts falling at 7000.

p.s. funny...you didn't mention the g plot, the VCDS torque plot or the VCDS measure of RPM/second. Neither did adsgreen. Odd. I think the VCDS plot is the most compelling. A quick look at RPMs per second (I use a smoothed little grouping of 4 data points so =(B46-B42)/(A46-A42). The car rips through the RPMS at about 600 RPM/second for most of the gear...but starts to slow down quite a bit once you hit the 7000s. If you look at the last reading, which measures RPM/Sec from 7520-8000 RPM you're only picking up 432 RPM/second there.

[sakimano]

No new argument.... You're just WRONG!

he he.

The key point here (and kinda relevant to my other argument) is that yes, torque is falling flat on it's face and it's much higher if you change gear and drop rpm. The problem is that the lower torque of the engine at 8000-8250 rpm coupled with the lower gear (say 2nd) still results in a higher torque at the wheels than changing to 3rd and dropping to say 5-6k rpm.
Here's a very rough graph to illustrate.


For gears 1-3 you want to go all the way up to the limiter (or as nears as you can without tripping it).
It's only from 4th upwards you can see that it is better to change into 5th before the redline (about 8100) as the torque at the wheels is better in the higher gear from that point.

I get the thrust plots...however when you drive the car and it doesn't do what the thrust plot says, you have to re-evaluate. I'd submit that other than maybe Dan @ JHM I have more passes in my car at the dragstrip...i.e. more timed runs measuring acceleration in a gear shifting rip through multiple gears, than anyone on earth. (not many at around 40 passes though lol). I have found that I go faster when I shift in the mid 7000s than I do when I shift at 8000 or 8250 rev limiter.

When I look at my logs, and my pbox times and my dragstrip passes and everything else, it tells me that I don't want to be riding each gear to the limit. Don't forget, I ran 12.75 in my RS4, faster than any other stock RS4 on earth. Don't think I didn't try every possible shift point in finding what works best!

[adsgreen]

But saks - the rpm per sec will be (or should be) considerably lower than 432 / sec in the next higher gear.

You can't decide gear change point without having competitive stats for both gears.
The accel and VCDS plots are interesting but you need to have the same data for the next gear to draw any conclusions

[ArthurPE]

I'd like to know what the difference is between the S6 and S8 engine (other than 100 HP and 75 lb ft)
I've read:
dual vs single air intakes
turbos
software/boost (obviously)
and compression ratio, lower in the S8, so boost can be raised safely

this engine is going to be a pretty slick device
for those brave enough to mess with warranty I'm guessing the S6 with a chip ought to get 1/2 way to S8, maybe 460-470. 430-440 lb ft

[JCviggen]

are you looking at the RS4 curve, or the R8 curve? The RS4 hp curve (red) clearly starts falling at 7000.

I'm looking at the RS4 curve, and it seems that you are only looking in 1 direction (left to right) when for determining shift points the left part of the curve is just as important.
On the RS4 curve I see max power around 7000 and then it starts to drop off. But it drops off to the left (lower rpm) just the same. It stops at 7800rpm but at that point it is STILL making about the same as it was at 6500rpm. With a curve like this you will always need to shift gear well past peak power if you want to make the best of it. If you shift at peak power it drops down to an rpm that is too low. There is a very steep and sudden dropoff to the left of 6000rpm. You want to avoid that area, so that means shifting well beyond 7500. Unfortunately the curve isn't plotted far enough to determine the actual ideal shift point.

I found an RS4 plot from a MAHA dyno (a whole different league from dynojets) and it's pretty obvious that you'd want to shift that particular car as close to the limiter as possible...

Again, left of 6/6.5K things drop off dramatically. You need to shift well past 8K rpm to drop between 6000 and 6500 where the meat is.
Or if you don't like that one, someone on the dyno at AmD (before and after de-coke)

I don't see that "cliff" that is supposed to be somewhere after 7K that you keep referencing...well running cars don't seem to have it.

p.s. funny...you didn't mention the g plot, the VCDS torque plot or the VCDS measure of RPM/second. Neither did adsgreen. Odd. I think the VCDS plot is the most compelling. A quick look at RPMs per second (I use a smoothed little grouping of 4 data points so =(B46-B42)/(A46-A42). The car rips through the RPMS at about 600 RPM/second for most of the gear...but starts to slow down quite a bit once you hit the 7000s. If you look at the last reading, which measures RPM/Sec from 7520-8000 RPM you're only picking up 432 RPM/second there.

There's insufficient data to go on, we can't see the gear or speed. Speed increases with RPM as well, so naturally more aerodynamic resistance and tire friction will make acceleration slow down the higher you go anyway (assuming power levels off or drops slightly)
That said, I can't make many statements about your particular car. Maybe it doesn't breathe as well at the very top as some others. I've seen some that definitely do not "slow down" dramatically above 7500... http://www.youtube.com/watch?v=mzbBRoIhODY

I ran 12.75 in my RS4, faster than any other stock RS4 on earth

Do you know every RS4 on earth?
Your ET doesn't prove much, except that you got a stellar launch. What was the trap speed?
Have any vids (speedo preferably, not too interested in 1/4 runs)

[sakimano]

I'd like to know what the difference is between the S6 and S8 engine (other than 100 HP and 75 lb ft)
I've read:
dual vs single air intakes
turbos
software/boost (obviously)
and compression ratio, lower in the S8, so boost can be raised safely

this engine is going to be a pretty slick device
for those brave enough to mess with warranty I'm guessing the S6 with a chip ought to get 1/2 way to S8, maybe 460-470. 430-440 lb ft

I think there's much more left on the table Arthur. They cut boost at 5500 RPM. Have a look (at the most simplified power plot ever lol)...

I think you'll pick up 50 hp with just boost control letting the car hold boost to redline. More with better AFR and timing. Add in better dynamics with the almost certainly neutered stock throttle map being opened up and you're looking at good gains.

Incidentally the new M5 plot looks exactly like this....very neutered early on.

[sakimano]

But saks - the rpm per sec will be (or should be) considerably lower than 432 / sec in the next higher gear.

You can't decide gear change point without having competitive stats for both gears.
The accel and vcds plots are interesting but you need to have the same data for the next gear to draw any conclusions

very true...however how bad could it be when I'm losing power so quickly at the very top of each gear? As we've seen from dynos in 3rd, 4th and various logs in multiple gears, the car is strong once the throttle is 100% open on the stock map (around 5250 RPM and up).

While I could play with charts all day, I already did that. I looked at the charts, got some cues, then did some real wrold testing. Then I went 12.81, 12.79, 12.77, 12.75 doing what I'm talking about here. If you want to go hit Santa Pod and go 12.80 or better shifting at 8000 or 8250 rpm, have at it! (no other RS4 owner has ever gone 12.84 or better doing that).

On many of the forums we tend to play with the numbers and bench race/textbook race a bit much. Can't compare to driving these cars!

[ArthurPE]

I plotted the Automobile RS4 curve last night
differently
y axis = thrust = T x gearing / 1.08
x axis = 3000 to 8000 rpm
so the curves are stacked
so you look at the next gear, draw a line over to the lower gear and they intercept THEORETICALLY (more on that later) that is the optimized shift point
for 1, 2, 3 ,4 redline is the point (no overlap)
5th is 7600 or so
6th is 7000 or so
all this proves is Audi picked the gearing correctly, to maximize torque band, and as speed increases, and rate of accel decreases, and gearing becomes closer, you can shift earlier

having said that, momentum factors in, and if the rate of decline of torque is much greater than the the rate in increase in the higher gear, it may be better to shift then
this point is usually right past the power peak ~7500 or so when torque is drops off, you can see it in the thrust plots
but honestly, much of this is derived by 'feel' and timing

[P_G]

[HYFR]

seriously, are you guys still going at it ?!

[ArthurPE]

seriously, are you guys still going at it ?!

it's actually gotten civil and interesting
so the going at it has morphed into an interesting discussion