dyno lies....

[SR71]

There I was thinking that you could make the figures suggest anything you wanted...

How can a cat-back exhaust generate an extra 34 ftlbs of torque?

Thats over 10% extra torque on the stock car from a bit of pipe.

I contend that if this is the case, carbon deposition can never again be disputed to cause a loss of power!

[ArthurPE]

There I was thinking that you could make the figures suggest anything you wanted...

How can a cat-back exhaust generate an extra 34 ftlbs of torque?

Thats over 10% extra torque on the stock car from a bit of pipe.

I contend that if this is the case, carbon deposition can never again be disputed to cause a loss of power!

how are deposits and decated related again?

let's say it is decat'ed and chipped...
ie, not stock, and makes 420 HP
I could see cats making more low end torque...

then obviously stock is less, say 380? 400?
275-295 with losses (easily 105+ HP)
so the % difference goes up, approaches 50%

and we come full circle...since we are talking about stock cars...
a car with a 45+% power/wt disadvantage is as fast as the more powerful car...

how????

[SR71]

how are deposits and decated related again?

Wasn't it you (or the Bosch Handbook?) that mentioned:

T = V x Cr x Vol eff/4Pi

By affecting the airflow you're affecting Veff correct? I don't believe a catback system can affect the Veff by >10%.

So the absolute values on that graph are bogus....you're always keen to tell us dynos deliver bogus figures....unless, perhaps, the car is not stock.

You can also tell it is not stock just from the shape of the curve...

Where is the ramp at 5500rpm?

So now, like you say, how do we square the circle?

All we're doing is squaring the "circle" by asserting those premises that support our prejudice.

The truth is no doubt somewhere in the middle.

Out of curiousity, how do cars which dyno in 2WD mode, cater for the loses of all 4 tyre wheel interfaces which a car when dynoed in 4WD mode suffers? Surely in this case the drag associated with two wheels vanishes which is why 2WD dynos deliver artificially low loss figures...

[ArthurPE]

I do...at low engine speeds where it is low eff to begin with...
but 10% is pushing it...I agree, hence no trust in dynos

but going from say 70% at 3000 rpm to 77% is a huge difference than 90 to 99 at 7800...

rolling resistance is low, especially on a non-driven wheel...
but this is all a distraction...we are not measuring (guessing) those losses anyways, anymore than air drag...only driveline loss, a factor is plugged in for slipp, and the wheel/drum speed is measured...should negate any concern

let's assume an M3 weighs 3550 (from numerous sources this seems resonable)
RS4, euro stripper, cf brakes, no sunroof, no nav, etc. 3850 fair?

power
M3 let's compromise, say 360 HP

now it get's tricky
most dynos on here show 300 + 20% losses ~ 375 HP give or take
but measured losses are closer to 25% or 100+ HP
let's use 300 HP (it goes much lower if you use the 380 - 105+ in losses as most claim)

M3 101 HP/kip
RS4 78 HP/kip
a 30% difference
cube 0.93, base 13, adjusted 11.9
square 0.88, base 13, adjust 11.4
no where near real world results
btw, it is square, not cubed, 1/2 a t^2 and the second derivative...
but that is moot...

now if we look at Ring times, +/- <1% or avg 0, it makes no sense
with a 30% advantage? the relationship there is ~ linear (but not 1:1, slope varies)

the 'truth' or reality is the RS4 is faster than it should be, and the M3 slower, or both...

imho you have a very strong mind, you need to do the same as I do at times (apparently not often enough, lol), open it...I TRY to let the numbers/science dictate my position...but I am frail, like all, and can get 'wrapped up' in the debate...

how are deposits and decated related again?

Wasn't it you (or the Bosch Handbook?) that mentioned:

T = V x Cr x Vol eff/4Pi

By affecting the airflow you're affecting Veff correct? I don't believe a catback system can affect the Veff by >10%.

So the absolute values on that graph are bogus....you're always keen to tell us dynos deliver bogus figures....unless, perhaps, the car is not stock.

So now, like you say, how do we square the circle?

All we're doing is squaring the "circle" by asserting those premises that support our prejudice.

The truth is no doubt somewhere in the middle.

Out of curiousity, how do cars which dyno in 2WD mode, cater for the loses of all 4 tyre wheel interfaces which a car when dynoed in 4WD mode suffers? Surely in this case the drag associated with two wheels vanishes which is why 2WD dynos deliver artificially low loss figures...

[ArthurPE]

as much as I have enjoyed this, and believe it or not, I have lol
I have to get ready to go to a freakin' work dinner (Pennsylvania Municipal Authorities Assoc.)
suit, grown-up shoes, the whole 9 yards...
good people, but I'll miss my hockey teams playoff game!
at least it's a nice drive through the mountains, it's at a ski resort...

don't figure the 30-50% conundrum out until I get back!!!

[SR71]

Lets assume losses comprise the following:

x=loss at wheel tyre interface
y=other gearbox transmission losses

Lets assume Audi and BMW design mean that losses at each wheel on each car are similar, but that a "driven wheel", even when coasting is 20% worse than a "non-driven" wheel, and that the gearbox losses are, being generous, 50% worse on the RS4?

So, for B7, we have

4(1.2)x + 1.5y = 80.7

So, for M3, we have

2(1.2)x + y = 43.5

Solving for x, x (the loss at each "non-driven" tyre wheel interface)= 12.9hp & y (gearbox losses) = 12.6hp

So I reckon losses for the M3 should actually be = 4.4x + y = ~16.7% versus 19.5% for RS4.

So the true losses between cars should only be approx 2-3% different.

But like you say, the observed trend in wheel powers suggest a 7-9% (30-40whp) difference which coupled with a weight difference of ~8% (3550 versus 3850) means a good performance delta.

Obviously to some extent negated by better traction off the line in the 4WD car and better torque in the RS4 when not at max power.

I suggest most owners experience regards the relative performance of the two cars is not instructive as it will only be "flat out" that it manifests itself...

[ArthurPE]

interesting analysis, what is the 80.7 and 43.5?

you assume the losses at each RS4 wheel to be the same as those of the M3, BUT, the power is only 1/2 as much at each of the RS4 wheels, so the losses ~1/2 as much...reducing your equations to:
1/2 x 4(1.2)x + 1.5y = 80.7 = 2(1.2)x + 1.5y

the first terms cancel, resulting in the RS4 having 1.5 times the loss of the M3...I would argue the 1.5 should be closer to 2 (2x's the axles, power reversal, f to r diff and f diff extra, maybe 1.75, but you get the point...
this is good stuff....I'll get back into when I get back from the shindig (and I'm sure I'll be turning it over in the brain box during, lol...wife: what's that blank look on your face? me: damn SR71!!!!! lol)

Lets assume losses comprise the following:

x=loss at wheel tyre interface
y=other gearbox transmission losses

Lets assume Audi and BMW design mean that losses at each wheel on each car are similar, but that a "driven wheel", even when coasting is 20% worse than a "non-driven" wheel, and that the gearbox losses are, being generous, 50% worse on the RS4?

So, for B7, we have
4(1.2)x + 1.5y = 80.7

So, for M3, we have
2(1.2)x + y = 43.5

Solving for x, x (the loss at each "non-driven" tyre wheel interface)= 12.9hp & y (gearbox losses) = 12.6hp

So I reckon losses for the M3 should actually be = 4.4x + y = ~16.7% versus 19.5% for RS4.

So the true losses between cars should only be approx 2-3% different.

But like you say, the observed trend in wheel powers suggest a 7-9% (30-40whp) difference which coupled with a weight difference of ~8% (3550 versus 3850) means a good performance delta.

Obviously to some extent negated by better traction off the line in the 4WD car and better torque in the RS4 when not at max power.

I suggest most owners experience regards the relative performance of the two cars is not instructive as it will only be "flat out" that it manifests itself...

[SR71]

80.7 = 19.5% of 414.
43.5 = 10.5% of 414.

Your figures from RRI Sweden?

Note, those RRI figures come with quoted weights of 1664kg (3660) for M3 and 1728kg (3800) for B7.

Both cars rated @ 414.

On the coastdown, power is not being delivered to the wheels, they are freewheeling, so the 1.5 x gearbox losses should take into account the effect of all the extra spinning gear. That said, I wonder if that is still conservative? But even if you said losses in B7 were = 2y, the effect is small.

But I suggest my point(s) is....are:

1) Wheel power is the best thing to measure I agree, but it can vary hugely depending on tyres, pressures, profiles etc etc so these variables need to be controlled strictly

2) Comparing 2WD and 4WD cars on dynos is fraught with difficulty.

3) Judging by the figures (chemical_ali - Put your E92 on a weighbridge would ya!) the B7 does a great job keeping up with a E92 including at the track.

4) I'm sure there is another one but I'm off to bed...

[ArthurPE]

the 1664 kg is a sedan, the coupe is 1611 (3550 lbs)
since all the Ring times, and accel numbers I've used are for the coupe, it would be the number to use... http://www.rri.se/popup/performancegrap ... artsID=768
M3 power for sedan & coupe 373 & 371, call it 370

it has both RS4's at 3800+, but call it 3800
although the associated accel times are for a wt closer to 4000
3965/3920/3980 R&T/C&D/R&T
sport auto ~ 3800

power is 328/345, call it 337, the avg
if we assume 20% loss 337/0.8 ~ 420 HP
but this assumes 414 HP, otherwise if we assume 380 or whatever, losses are way too small at 11%

RS4 88.7 HP/kip
M3 106 HP/kip
20% difference...
sq rt, base 13, adjusted 11.85...still, way off...

RRI uses a hub dyno, so no slip/tire losses, in either number...
it's also very accurate, it's an actual electric dynamo-meter...

I pulled some numbers from a R&T accel test graph RS4 vs M3 sedan, looked at 40-90, very linear, and the meat of 2nd & 3rd
the shift points were almost identical
plus the awd traction advantage is negated by 40 mph
I first calced the accel for each (times RS/M 6.9/6.6)
RS4 10.6 ft/sec^2
M3 11.1
~5% difference

then I calculated the mass in slugs (RS/M, 123.14/116.92)
from which the F (motive thrust) in lbs was found (RS/M, 1305/1298 lbs)
the RS was actually putting down more power (marginally) and only was slower due to weight, if it were the same weight, it would be the same speed as the M...so this tells me the RS4 is making as much power, and has good gearing...but gearing alone doesn't explain it, since the M's is very similar...so this leaves one conclusion: same power, ~410+ HP
they ran 1/4 mile in this test:
RS4 13/108
M3 13/110
M3 0.7 faster 16 vs 15.3 to 120 mph, again, wt being the difference, not lack of power...

that's why the Ring is 'the test' imho, long enough to amplify differences, fast enough to be power based, long enough to average out errors, as long as it it the same level of driver...
using the same 20% power/wt delta, the fact that the difference is only <1% +/-, call it dang close to even, is very impressive...

80.7 = 19.5% of 414.
43.5 = 10.5% of 414.

Your figures from RRI Sweden?

Note, those RRI figures come with quoted weights of 1664kg (3660) for M3 and 1728kg (3800) for B7.

Both cars rated @ 414.

On the coastdown, power is not being delivered to the wheels, they are freewheeling, so the 1.5 x gearbox losses should take into account the effect of all the extra spinning gear. That said, I wonder if that is still conservative? But even if you said losses in B7 were = 2y, the effect is small.

But I suggest my point(s) is....are:

1) Wheel power is the best thing to measure I agree, but it can vary hugely depending on tyres, pressures, profiles etc etc so these variables need to be controlled strictly

2) Comparing 2WD and 4WD cars on dynos is fraught with difficulty.

3) Judging by the figures (chemical_ali - Put your E92 on a weighbridge would ya!) the B7 does a great job keeping up with a E92 including at the track.

4) I'm sure there is another one but I'm off to bed...

[SR71]

FWIW...

Yeah I made a big error in my analysis as RRI tests are hub dyno.

I knew my gut feeling was right.



So in that case, RRI's figures are what I called "y", and low and behold, what I thought was probably the case is true, namely "transmission" losses for the M are ~x2 of the B7 - 19.5% versus 10.5% and y = 40hp.

So redo calcs using a Cartec figures for "total losses" for B7 ~ 110hp means that losses at a "driven" wheel are ~7.5hp and an undriven wheel ~6.3hp.

The relative size of the figures now makes a bit more sense and thus the "error" in a 2WD braked dyno result on the M is only ~12hp, and thus, the difference in losses between the cars is ~10%, not the earlier suggested 2-3%.

So I reckon M losses ~16% and B7 losses ~26%.

Your figures give ~10% versus 19% so the delta is the same...it'll just ratchet the figures up/down a little. This gives M power at wheels based on quoted 414 as ~350whp and B7 ~305whp, oddly, pretty much what mine did, and what the Cartec figures suggest.

I'd agree with the hp/lb delta of ~20%.

Now times, or rather trap speeds because ET's can vary hugely with rollout, with a typical ET differing by ~0.5 secs depending on a rollout variance of 16"...

Using LRT

mph = 215.39 x (hp/(curb weight+200))^0.3018

Using the above, based on quoted power, which implies the 20% variance in hp/lb at wheels

M3 mph = 110.8
B7 mph = 108.6

so worth ~2mph over the 1/4.

If I plug a 106mph into the LRT curve fit, for a 4000lbs car, the power is 382bhp, and @ 107mph, the power is 394bhp.

I don't know about you but I think most B7's run closer to 105/106 than 109.

So we are now down to a situation where assessing trap speeds to within 3-4% will determine whether or not a car is producing between 380-414bhp.

You can then see why it is quite plausible to believe a 4000lb car running 105mph at the drag strip, developing 310whp, which therefore means a 20% hp/lb deficit over the M3, could be suggested to only deliver 380hp at the crank.

Then when the dyno backs it up....

Whatever...I agree the B7 appears to make a much better job of using its power when you look at the figures.

Its just my contention is it could do even better.

[Sims]

I have put my car on 8 rolling road sessions in a 3 year period and it was orginally dyno'ed at 358bhp with 16,500 miles on a Dyno Dynamics Road all the way through to this year when it was dyno'ed at 406bhp on a Dastek Road. This is a stock car with an x-pipe so as best the modification gives a 3bhp gain so for arguments sake say 403bhp (and at present it has an EPC warning on it so who knows what it could produce once fixed?).

So in three years and 48,000 miles my car has gained 45bhp? Torque has always been fairly consistent at 305-311.

That's a lot on dyno tests, and I was always under the impression you did not believe in dynos.

If you can find a clip on a standard, OEM non modified in any way M3 vs a similar RS4 and they have the same fuel levels and weight i.e. passenger weight in them and the M3 wins then fair enough

Agree. We have the 'Ring times of 8:05 & 8:09. As I have said, hardly a cigarette paper between them.

Until then, please stop trying to compare apples and pears to justify your and perhaps others theory there is something universally wrong with RS4's...

.. That's not denying that there aren't cars out there that do have issues because they have been documented but to suggest that they all have issues is just ridiculous to the point.

We are agreed on that. Not all cars have a problem, but there are cars out there that do, and it's in our interest to find out why that is so. It could be your car tomorrow - debate on this subject should be encouraged for a lot of good is coming out of it..

... The car is much more than just power and apart from the R8, if it were damp or worse still, snow most others would struggle to match the RS4.

There has been 100% agreement on this from everyone.

[dlextreme1977]

[quote="SR71"]

I don't know about you but I think most B7's run closer to 105/106 than 109.

quote]

+1

Think somebofy mentioned it on here earlier but is there any reason the rs246 club have not organised a 1/4mile day or a day at bruntingthorpe or the like. Reckon it would be far more fun and interesting than a dyno day and would certainly put this argument to rest. You could invite rs6's along m3 etc and would be a good laugh

[SR71]

Out of curiousity, on a constant force braked dyno, how does the dyno deal with the 40:60 torque bias?

Does it force the car into 50:50 mode because not all drums are individually braked?

That said, if you don't input the torque bias, how does it know how to brake the individual drums?

If the car is running in 50:50 mode versus 40:60 mode, are the losses now artifically high?

Does a GTR on the dyno show a 50:50 torque split on the dash as opposed to a 95% split to the rear on the road?

The Shootout Day would be great....although it'd no doubt create more questions than answers...

[2manytoys]

Wow, 9 pages of Artur Buildup, and I thought Carbon Buildup was a problem.

Note to self: untick box "Notify me when a reply is posted" so I don't have to return

[larshs]

I think it's only 0.45 bar. But MTM told me that usually all B7's output ~380 bhp, but after a remap ~435 bhp. If thats the case the calculation will probably be a little different or? But with the new exhaust I expected to see 535 bhp on the dyno (Corrrected power).

Never cleaned the valves, and wasn't there when they installed the SC

0.45 bar ~ 6.5 psi...

I don't buy a remap regaining 55 HP, that is 15%, huge...

I'm guessing your baseline is close to stock, 410 or so...low as I said
your SC gain: 1 + ( 0.6 x 6.5/14.7) x 425 ~ 520 HP
I would say you are about right, but remember, the dyno numbers may vary (accuracy) by 5% or more, 20 to 25 HP...or more...

I would love to see some 3rd gear pulls, timed from 3000 to 8000 rpm using the OBC timer...2000 rpm, stomp it, lol

I have to assume whom ever installed the SC would have cleaned the valves if they considered them an issue...they would definitely want to maximize power to make sure you were satisifed, especially if they cost 50HP as is claimed

Did the test you asked for. 3rd gear timed from 3000 - 8000 rpm. Not impressed by the time though. Either people at this forum tell "stories" or my car lack some power. Average timing was 6.7 sec. Used the built in lap timer so it might be inaccurate to +- 0.2sec. Tried one drag 0-100 km/h as well, a little wheel slip in the start, result was 4.1 sec( using the racelogiq performance box).

Weight of the car is 1720kg + me at 80 kg + some extra due to mods, maybe 20-30 kg or so, a total of 1820-1830.

Any comments ?