1M Diff and Transmission Specs

[swamp2]

Originally Posted by Pete_vB

Tarheel91 makes a good case for why this is far too over-simplified above, and I think we've shown that well elsewhere in the thread as well, but for another illustration of why this is too simplistic and leads to errors I did the following simulation to illustrate the point. I used the 911 turbo because huge area under the power curve and gearing makes for a dramatic comparison:

Car 1- I took the "Porsche Turbo 1988" in carsim, which you have, and change hp to 400 at 7000 rpm, tq to 400 at 5000 rpm- typical modified 911 turbo numbers.

Car 2- Same car, I left 400 hp at 7000, but I put torque at 300 at 3000 rpm. The flat torque curve looks like an M3 motor with the top 1000 rpm cut off, or a high revving toyota/ lotus motor that makes peak power at redline. For car 2 I also removed 500 lbs.

Car 1 has 3025 lbs / 400 hp = 7.6 lbs per hp.
Car 2 has 2525 / 400 hp = 6.3 lbs per hp.

This contest shouldn't even be close according to your simplistic evaluation, but car 1 wins every acceleration test. That's the problem with ignoring area under the curve. Obviously this is an extreme example, but it's the same error you made in evaluating the 1M and it seems the message still hasn't gotten through. Again, garbage in, garbage out- if you're not evaluating the the important numbers the results are worthless, and the torque/ power curve is important.

You have left out a KEY part of this analysis.

Consider car 3, identical to car 2 but with a 9000 rpm redline, peak power actually at 9000 rpm and a final drive of 4.8. The power and torque specs are all identical to your cases and the power to weight is the best (being still identical to car 2). Car 3 will substantially best both car 1 and car 2. Why is your case unrealistic? Because if an OEM had an engine anything like the hypothetical one you "designed" for car 2, they would raise the redline, the rpm of the peak hp would could rise as well (since the torque curve has plenty of steam left past 7000 as does the hp curve, they are both totally flat at 7000). Another way to think about this is that your car 2 is a bit like an M3 more torque but with an artificial rev limiter set at 6000 rpm!

So as I've stated earlier, assuming the manufacturer has done their gearing selection properly and this is genrally quite true, hp/weight is the most predictive parameter.

Thanks for the example, I think it may help lots of people get why hp sort of naturally includes gearing.

[Pete_vB]

Originally Posted by swamp2

If an OEM had an engine anything like the hypothetical one you "designed" for car 2, they would raise the redline, the rpm of the peak hp would could rise as well...

One might think so. But have you looked at a supercharged Exige dyno? Or a 2.2L S2000? Or more than one ferrari? A couple stock dynos attached- first is a 2.2L S2000, second is an '06 F430. I made the example to be extreme, but cars can and do come close to that...

What you're effectively arguing, though, is that if these cars extended their rev limits they would be faster even if they didn't make more power, because there would be more area under the power curve. Which is true.

For the other extreme look at a Nissan GT-R (dyno also attached). It basically sits on its peak hp for over the top 35% of the rev range. If you look at the gearing that goes with this it never falls off of this power band- the GT-R basically puts down all of its power, all of the time- massive area under the curve.

Between these extremes fall everything else, but it's a continuum, with some cars much better than others.

You made the point that HP to weight is the single best predictor of performance if you need to choose one number. I agree with that. What you seem reluctant to admit, though, is that on its own it's not a very accurate predictor- as we see above there is quite a range where a lower power to weight car can be faster because it's putting down more area under the curve.

[tarheel91]

Originally Posted by swamp2

You have left out a KEY part of this analysis.

Consider car 3, identical to car 2 but with a 9000 rpm redline, peak power actually at 9000 rpm and a final drive of 4.8. The power and torque specs are all identical to your cases and the power to weight is the best (being still identical to car 2). Car 3 will substantially best both car 1 and car 2. Why is your case unrealistic? Because if an OEM had an engine anything like the hypothetical one you "designed" for car 2, they would raise the redline, the rpm of the peak hp would could rise as well (since the torque curve has plenty of steam left past 7000 as does the hp curve, they are both totally flat at 7000). Another way to think about this is that your car 2 is a bit like an M3 more torque but with an artificial rev limiter set at 6000 rpm!

So as I've stated earlier, assuming the manufacturer has done their gearing selection properly and this is genrally quite true, hp/weight is the most predictive parameter.

Thanks for the example, I think it may help lots of people get why hp sort of naturally includes gearing.

A couple things. Extending the rev range isn't always that easy. You've got plenty of other variables that inhibit bringing that up. Ring speeds are essential to engine longevity, and unless the engine is oversquare, you may not be able to bring the red line any higher without bringing engine life down to an unacceptable amount of miles. Unless, of course, you're willing to make the internals stronger (see:F20C from the S2000, S54 from M3/Z4M), but that increases cost, and it still doesn't save your piston rings. From an all out performance perspective, yes it makes sense to extend the redline, but that's not all manufacturers think about.

The same goes for gearing. Manufacturers aren't only thinking about performance when they gear a car. Fuel economy is a major concern as well. That's why my Z4 was originally geared to around 185 despite, maxing out somewhere around 150 most likely. That's why Corvettes have a 6th gear that barely turns 2K crusing down the highway. That's why you can't just look at a power number. Power is an ideal value, but it isn't always COMPLETELY representitive of reality.

[swamp2]

Originally Posted by Pete_vB

You made the point that HP to weight is the single best predictor of performance if you need to choose one number. I agree with that. What you seem reluctant to admit, though, is that on its own it's not a very accurate predictor- as we see above there is quite a range where a lower power to weight car can be faster because it's putting down more area under the curve.

Great, we are finally on the same page... almost.

However, I do contend that this single number I like, is actually on its own a reasonably accurate predictor. Sure as you mentioned there are all types and shapes of torque curves and thus hp curves. That is one key thing that gives engines their own "character". Another thing that makes hp/weight even better is if you simply separate FI vs. non FI cases. But even without doing so it's the best you will get.

There is nothing wrong with knowing the full torque curve and even in addition knowing the gearing. Both of those put you a 1 minute away from a full transient prediction with a simulator. Area under the curve is undoubtedly important. I may have overstated my case about the importance of that factor in making my overall argument.

I continue to learn things here on this forum myself and the sharp folks that hang out here is a reason I like it. Cheers.

P.S. If you get really bored have a look at the Nurburgring lap time regression vs. power to weight discussion in this thread. I mentioned it before. Likely to bring up all sorts of new controversy... The idea is very similar to the theme of the idea I've argued here in this thread. Link.

[swamp2]

Originally Posted by tarheel91

A couple things.
...

Absolutely agreed!

[swamp2]

Originally Posted by Pete_vB

Looks fairly reasonable so far, but where are you getting your M3 weights from? BMW USA is listing unladen weight at 3704 lbs on the website? And DCT is supposed to add 45 lbs... Or is that wrong? You have 3581...

By the way, the unrealistically low loss required to get good speeds is one reason I'm not so sure about cartest. I needed to do something similar, but it doesn't look right vs lapsim.

This is the actual weight of my car with 1/4 tank without me (my car is M-DCT). I weighed the car on a nice 4 wheel system. Post here. Might be more fair to only use only quoted weights but in this case it seems way too much weight. Since CarTest has a separate entry for driver weight it is not clear if the "Curb weight" should or should not include the driver weight.

I think the only unrealistic thing about the standard losses in CarTest are that they are too high. They are unjustified based both on dyno results, specifically rototest.com and on various loss figures quoted in the literature. Like you said, garbage in garbage out.

[Pete_vB]

Originally Posted by swamp2

This is the actual weight of my car with 1/4 tank without me (my car is M-DCT). I weighed the car on a nice 4 wheel system. Post here. Might be more fair to only use only quoted weights but in this case it seems way too much weight. Since CarTest has a separate entry for driver weight it is not clear if the "Curb weight" should or should not include the driver weight.

The cartest input should be without driver: curb weight in the US is total vehicle weight, full fluids and fuel, no driver or cargo, and it looks like that's what cartest is looking for. For your car I'd use 3581 + fuel weight, or 16.6 x 3/4 x 6.3 = 78 lbs = 3659 lbs. I used the listed 3704 for my sim, might be slightly too high in reality.

Originally Posted by swamp2

I think the only unrealistic thing about the standard losses in CarTest are that they are too high. They are unjustified based both on dyno results, specifically rototest.com and on various loss figures quoted in the literature. Like you said, garbage in garbage out.

Yes the losses are too high, but something else appears to be going on that you noticed yourself. In order to get speed correct in the higher gears you need to reduce the losses, but this gives unrealistically fast acceleration at lower speeds (0-60, etc). This suggests tire friction or air resistance is also high.

My preferred method is generally to input a wheels dyno and switch all of the driveline losses off. This has issues of it's own, but it accounts for the very different driveline losses from car to car more accurately.

[Pete_vB]

Originally Posted by swamp2

Great, we are finally on the same page... almost.

However, I do contend that this single number I like, is actually on its own a reasonably accurate predictor. Sure as you mentioned there are all types and shapes of torque curves and thus hp curves. That is one key thing that gives engines their own "character". Another thing that makes hp/weight even better is if you simply separate FI vs. non FI cases. But even without doing so it's the best you will get.

You've suggested earlier in the thread that power to weight is accurate enough to trump the thrust to weight graphs I first posted. However I think you'll now agree that's not the case, as those graphs fill in the rest of the story. Relying on power to weight alone will lead to errors, even when comparing similarly powered, well geared normally aspirated cars (ie Ferrari 430 vs viper GTS).

As you also point out, power to weight will tend tend to overestimate high RPM, low torque cars vs high torque FI cars (as in the case of the 1M vs M3).

I think we're close to on the same page now, though. The data we have predicts that the 1M will do very well at lower speeds vs the M3, but will be overpowered at higher speeds. While we're still reading the tea leaves here, as we don't have good dynos or weights for the 1M yet, the 'Ring times (a nice high speed track) also tend to support this conclusion. However I'll put my money on the 1M at an autocross- as one of the lucky guys that got an early drive said, it looks like it will make a great autocross car, as a comparison vs the already good M3's 2nd gear suggests. With a tune it could be pretty spectacular...

[Pete_vB]

And just for fun...

[HondaGoneRogue]

Where did I put that Tylenol? I can't believe I just read every post here. Strangely, I know I learned something, but still I feel less smart than when I entered the post.

Great discussion guys.

[Pete_vB]

Originally Posted by swamp2

P.S. If you get really bored have a look at the Nurburgring lap time regression vs. power to weight discussion in this thread. I mentioned it before. Likely to bring up all sorts of new controversy... The idea is very similar to the theme of the idea I've argued here in this thread. Link.

I took a quick look at that link, and again the HP under the curve explains partly why the GT-R is well off the projected time. Combined with the shift speed it essentially always puts down its peak 480 hp (and if you believe that power figure I've got a bridge to sell you).

Nissan is also clearly lying through their teeth regarding the true power figures of that car. Side by side analysis of the videos vs the ACR/ ZR1 can be made on the main straight. Simulation shows that the speed the GTR achieves is only possible with either a good size tail-wind or more likely a hotter than average (even for the GT-R) car.

[swamp2]

Originally Posted by Pete_vB

In order to get speed correct in the higher gears you need to reduce the losses, but this gives unrealistically fast acceleration at lower speeds (0-60, etc). This suggests tire friction or air resistance is also high.

CarTest isn't perfect and I have noticed some cases where the low and high end can not seem to be made to match tests well.

Originally Posted by Pete_vB

My preferred method is generally to input a wheels dyno and switch all of the driveline losses off. This has issues of it's own, but it accounts for the very different driveline losses from car to car more accurately.

Yes, a good approach. As I mentioned I like the data from rototest.com for this. Seems to be about the best.

Originally Posted by Pete_vB

However I'll put my money on the 1M at an autocross- as one of the lucky guys that got an early drive said, it looks like it will make a great autocross car, as a comparison vs the already good M3's 2nd gear suggests. With a tune it could be pretty spectacular...

It will indeed be a great autocross car. No doubts. I also agree with your prediction here on besting the M3 in such a venue.

Originally Posted by Pete_vB

I took a quick look at that link, and again the HP under the curve explains partly why the GT-R is well off the projected time. Combined with the shift speed it essentially always puts down its peak 480 hp (and if you believe that power figure I've got a bridge to sell you).

Nissan is also clearly lying through their teeth regarding the true power figures of that car. Side by side analysis of the videos vs the ACR/ ZR1 can be made on the main straight. Simulation shows that the speed the GTR achieves is only possible with either a good size tail-wind or more likely a hotter than average (even for the GT-R) car.

Shape of the engine curve will contribute but the 50 hp under rating, which is the fairly well agreed upon figure, is the larger factor by far . That along with tires that are awfully close to R comps.

Coincidentally I also made an extensive post (turning in to a huge debate...) long ago about the differences among these three cars on the Dottinger Hohe as evidence for the real power the GT-R has. The ZR1 video though is a bit apples to oranges though right at the slight bend as Mero does not show nearly as much balls as Suzuki san did.

[Pete_vB]

Originally Posted by HondaGoneRogue

Where did I put that Tylenol? I can't believe I just read every post here. Strangely, I know I learned something, but still I feel less smart than when I entered the post.

Great discussion guys.

Thank you. Glad you enjoyed it and learned something.

[M3 Adjuster]

Just looked up car and driver specs from their road test digest

http://www.caranddriver.com/features...digest-feature


BMW M3 7/08 $63,600 4.4/12.9 161 158 0.94 14 20
BMW 135 5/08 $42,895 4.7/13.3 143 157 0.89 17 25

With there as tested speed of 0-60 just three tenths lower on the 135... and adding the M3 tires, wider wheels, and a little more horsepower than a 135... and it looks like the 1M could certainly equal the E9x from 0-60 if not be faster... As a matter of fact... it might even be really close up to the 1/4 mile. (love the fact that the M3 brakes are now on the 1M ... it really hauls down well from triple digits with little drama. I expect the skid pad grip will also jump up from that .89 figure

[swamp2]

Originally Posted by tarheel91

A couple things. Extending the rev range isn't always that easy. You've got plenty of other variables that inhibit bringing that up. Ring speeds are essential to engine longevity, and unless the engine is oversquare, you may not be able to bring the red line any higher without bringing engine life down to an unacceptable amount of miles. Unless, of course, you're willing to make the internals stronger (see:F20C from the S2000, S54 from M3/Z4M), but that increases cost, and it still doesn't save your piston rings. From an all out performance perspective, yes it makes sense to extend the redline, but that's not all manufacturers think about.

The same goes for gearing. Manufacturers aren't only thinking about performance when they gear a car. Fuel economy is a major concern as well. That's why my Z4 was originally geared to around 185 despite, maxing out somewhere around 150 most likely. That's why Corvettes have a 6th gear that barely turns 2K crusing down the highway. That's why you can't just look at a power number. Power is an ideal value, but it isn't always COMPLETELY representitive of reality.

Thinking about this a bit more I felt I should clarify something. Not for you, but for others trying to follow the discussion.

My choice of an (unrealistic) 9000 rpm redline above, along with the 4.8 FD ratio were absolutely off the top of my head. They were only numbers used to make a point, not to provide totally achievable/realistic numbers. From how much this hypothetical car outperformed the other (cars #1 and #2) lower and more real world achievable numbers would still won the performance contest and would have also made my case. Perhaps I should have posted the performance results from simulation, it was a sound beating across the board.

[swamp2]

Originally Posted by M3 Adjuster

Just looked up car and driver specs from their road test digest

http://www.caranddriver.com/features...digest-feature


BMW M3 7/08 $63,600 4.4/12.9 161 158 0.94 14 20
BMW 135 5/08 $42,895 4.7/13.3 143 157 0.89 17 25

...
I expect the skid pad grip will also jump up from that .89 figure

Do note the M3 has achieved a much better result than .89. The best I have seen is .98 and .89 is on the very low end of the spead in results that have been published. Link. I'm sure the 1M will achieve the same or better.

[tarheel91]

Originally Posted by swamp2

Do note the M3 has achieved a much better result than .89. The best I have seen is .98 and .89 is on the very low end of the spead in results that have been published. Link. I'm sure the 1M will achieve the same or better.

Skid pad numbers are 99% tires, so I wouldn't really worry that much about it. I'd like to see a standardized mini-autocross used in testing. Skid pad is all steady-state cornering, slalom is all transient, neither are really representative of a car's handling around a normal corner. A mini autocross wouldn't change the fact that tires are a major contributor, but a car with a better suspension set up would be able to outperform a lesser car even if they had equal tires. Well, a miniautocross, or you could just make the slalom use turns of varius radii with very short straights (100 feet) in between.

[swamp2]

Originally Posted by tarheel91

Skid pad numbers are 99% tires, so I wouldn't really worry that much about it. I'd like to see a standardized mini-autocross used in testing. Skid pad is all steady-state cornering, slalom is all transient, neither are really representative of a car's handling around a normal corner. A mini autocross wouldn't change the fact that tires are a major contributor, but a car with a better suspension set up would be able to outperform a lesser car even if they had equal tires. Well, a miniautocross, or you could just make the slalom use turns of varius radii with very short straights (100 feet) in between.

The limited track tests that the American car rags have is indeed disappointing. A few of them have been pretty decent though. When they have section times they are fairly insightful, probably about as much as an autocross course would be. However, road coarses and autocross do require different strengths from a car.

0-60 is still something a very large percentage of folks in America care about, and of course care too much about (IMHO).

[M3 Adjuster]

Originally Posted by formula M

Next year (2012), American's will have a superb track to test on, in Austin..!

Well we already have several, VIR, Watkins Glen, Road America, Road Atlanta, and perhaps Miller Motorsports park. I'm looking forward to the new Austin track though, it is only three hours away.

[Evice]

Originally Posted by M3 Adjuster

Well we already have several, VIR, Watkins Glen, Road America, Road Atlanta, and perhaps Miller Motorsports park. I'm looking forward to the new Austin track though, it is only three hours away.

and Mid-Ohio!!

[swamp2]

Since no one added this here is some information about the differential itself, common with the E9X M3. This is from a very early press release on the E9X M3.

This is a super-sized differential unit. Significant modification in the rear-suspension area, including an all-new subframe, has allowed equipping the M3 with the same heavy-duty differential dimensions as in the 394-hp M5, whose production recently ended. A special high-strength steel alloy, called 18CrNiMo7, is used for the differential gears to achieve superior quietness and durability. A relatively “short” final drive ratio, 3.64:1, exploits the engine’s generous torque and rpm range; the 6th gear keeps it humming moderately at cruising speeds. Here too, targeted airflow under the vehicle helps keep the oil cool, along with a ribbed differential case.

M Variable Differential Lock. Together with the German division of GKN Viscodrive, BMW M engineers developed a special mechanical limited-slip differential for the M3s.

The principal difference between a traditional limited-slip “diff” and this M Variable Differential Lock is that the former senses torque, the latter senses wheel speed (rpm). Under dry to not-quite-dry road conditions, the traditional limited-slip has long enhanced the handling of sporty rear-wheel-drive BMWs; however, under slippery conditions, this differential type has limited ability to improve traction. On all current BMW models, electronic traction control addresses this issue.

The M Variable Differential Lock specifically addresses low- and split-traction situations in a way that reinforces sporty handling, imparting to the M3 a slippery-road ability no previous high-performance, rear-wheel-drive sports car ever had.

Any time a speed difference develops between the two rear (driven) wheels, a shear pump, driven solely by this difference, develops pressure in the silicon viscous fluid in which the lock operates. In turn, this pressure is directed to a multi-disc clutch that transfers driving torque to the wheel with the better road grip (“select high”). The greater the speed difference between the two wheels, the more aggressively the clutch engages. As soon as the difference between the two wheels’ speeds begins to diminish, the clutch starts to ease off.

This mechanism is “elegant,” in that it achieves sophisticated action by entirely natural means. There is no external pump, no external source of lubrication or operating fluid. The very motion to be controlled – differences in speed between left and right wheels – generates its locking action. Viscous fluid is so-called because it develops internal force (via an increase in viscosity) whenever it is sheared; this is why the relatively small difference between one wheel speed and the other can generate the necessary action.

[Dackelone]

I remember touring Tom Milner's shop in WVA many years ago. Milner was the guy who raced BMW's for BMW AG in the US. His old shop was Group 44. This was when he just started to campaign an e46 M3 V8. Anyway... I did a tour with the DC region BMW club... and I remember him saying how the e46 M3 diffs would not make it thruout the races. They were using the e39 M5 diffs and those would last/work.

I will also remember this really sweet McClearn lemans GTR racer.

Dackel