Simulations M-DCT M3 vs. R8

[swamp2]

OK perhaps I am getting too excited with this little software toy but it is damn good and damn fun. Please before you bash this simulation have a look at how well it blindly predicted the numbers in this test. By the way the simulation also completely NAILED the 0-100, 500 and 900 ft numbers as well. I keep getting more and more impressed with the CarTest software. Of course the next criticism will be M-DCT is not out. No biggie, I added a very close estimate of the weight gain, 40 lbs and (rougly) optimized the gear ratios for acceleration (hmmm I guess BMW might do the same...).

Why is the M-DCT M3 almost universally faster (if only by a hair...again insert appropriate comments about "a drivers race")? Look at the curves attached for the power losses (R8 is first). M3 looses about 20% less power to its drivetrain. If you claim the R8 drivetrain is significantly less lossy than this perhaps you can explain why when you tweak these parameters you get a car much faster in the simulation than the actual tests.

Now all that is left is for the M3 to best the R8s N'ring time, all this for half the price

[Epacy]

Thanks Swamp.
Interesting to see the R8 begin to pull in the M3 after 70mph.

[level002]

here are some numbers to plug into your new toy. they're from an r8 with a manual transmission, not the manually shifted auto.

0-60: 4.0
1/4 mile: 12.5 @ 113.2

and, incidentally, 1.01g on the skidpad

the m3 won't be beating the r8 at the ring at any price - sorry, mate. the m-dct m3 versus the auto r8 will be the closest it gets. does audi have a dual clutch option for the r8? if they do, i'd hate to see how ugly that gets.

it's all irrelevant, anyway. the m3 was never meant to take on the r8, it's just too much to ask, really.

[swamp2]

Originally Posted by level002

here are some numbers to plug into your new toy. they're from an r8 with a manual transmission, not the manually shifted auto.

0-60: 4.0
1/4 mile: 12.5 @ 113.2

and, incidentally, 1.01g on the skidpad

the m3 won't be beating the r8 at the ring at any price - sorry, mate. the m-dct m3 versus the auto r8 will be the closest it gets. does audi have a dual clutch option for the r8? if they do, i'd hate to see how ugly that gets.

it's all irrelevant, anyway. the m3 was never meant to take on the r8, it's just too much to ask, really.

Those are darn good numbers, what is the source? As you know all the mags tend to get different numbers. Is it best to take the absolute best ones of all, maybe, maybe not. Since the M3 6MT has already been clocked to 60at 4.4 with bad traction it might get as good at 4.3. The difference between the two cars in 0-60 is about .2 according to the sim, that puts the M3 at a potential 0-60 of 4.1. This along with the rumors of the M3 N'ring lap time of 8:02 as well as besting the R8 0-200 km/hr put this at a VERY close contest IMO. Again it will be a drivers race in all real circumstances.

Heck for half the price I know which one I would choose. The Audi is certainly a great car but I not very fond of the looks personally.

[esquire]

Originally Posted by level002

here are some numbers to plug into your new toy. they're from an r8 with a manual transmission, not the manually shifted auto.

0-60: 4.0
1/4 mile: 12.5 @ 113.2

and, incidentally, 1.01g on the skidpad

the m3 won't be beating the r8 at the ring at any price - sorry, mate. the m-dct m3 versus the auto r8 will be the closest it gets. does audi have a dual clutch option for the r8? if they do, i'd hate to see how ugly that gets.

it's all irrelevant, anyway. the m3 was never meant to take on the r8, it's just too much to ask, really.

the real issue here isn't that the r8 is a better performance vehicle, but the vast amount more your paying for the r8 and its marginal performance gains. for all that money it should be blasting the M3 out of the water. you have to ask yourself... is this another nsx?

[level002]

for swamp:

http://www.roadandtrack.com/article....rticle_id=5532

for esquire:

i'd have to say that the vast amount of money you're paying goes to far more than just marginal performance gains. and, though i find your comparison of the r8 to the nsx laughable, it really doesn't matter to me one way or the other. the way i see it, if cars needed to blast each other out of the water in order to justify a higher price tag - whoa, nellie - what an unattractive scene the auto industry would be...

[Epacy]

Whoa, $9k for the auto.

[esquire]

Originally Posted by level002

for esquire:

i'd have to say that the vast amount of money you're paying goes to far more than just marginal performance gains. and, though i find your comparison of the r8 to the nsx laughable, it really doesn't matter to me one way or the other. the way i see it, if cars needed to blast each other out of the water in order to justify a higher price tag - whoa, nellie - what an unattractive scene the auto industry would be...

translation: you're paying more money for looks. sound about right? your nsx argument just fell apart.

i don't have any doubt the r8 will be a great car, but from everything we've read so far its performance doesn't justify its pricetag. if i'm sitting in my 100k+ sports "supercar", the last thing i want to see is a substantially cheaper car keeping pace with me. but from the sound of it, it seems that performance doesn't really rank that high on your list of priorities for a performance car, so i doubt you'd have much of a problem with a 70k M3 riding right alongside you instead of in your rear view mirror.

[level002]

oh, i see how this goes now. an r8 buyer, who's likely a bloke who's got the means to afford damn near any car out there, needs to know that nothing cheaper will keep pace with him - nothing but nothing has more gravity than bang for the buck, his singular goal is to be The Value King.

it doesn't speak too well about you, mate, if you're saying that should something cheaper hang with you while you're 'sitting in your 100k+ sports "supercar"', your day is ruined. and if you want to continue to make assumptions about what ranks how on my list of priorities for a performance car... all day long, my man, have at it.

but, please... for your sake, please don't think that you're going to be upgrading to an m3 and riding right alongside any r8 whose driver is even trying - all you'll be doing is giving him a bit of a chuckle because, really, you'll be in his rearview mirror for as long as he feels like keeping you there.

but... whatever, right? i'll play along - THE 70K M3 WILL HANG WITH THE R8! THE 70K M3 WILL HANG WITH THE 4-DEAD, 12.5, 1.01G 110K R8!!!

cheers

[Keto]

*cough* Corvette *cough*

[RI_RS4]

Swamp,

I posted some graphs of 3rd and 4th gear pulls for my RS4 when it was stock that you can use to calibrate your models. To make it a bit easier for you, I can send you the data tables if you like. But I do have the following times that can be used for comparison to a simulation.

3rd gear 50 to 70 mph - 2.76s
3rd gear 70 to 90 mph - 2.91s

4th gear 30 to 50 mph - 3.83s
4th gear 50 to 70 mph - 3.56s
4th gear 70 to 90 mph - 3.77s


As an additional point of reference, a stock RS4 posted the following 1/4 mile times in Australia.

12.883s 108.02 mph

http://img.photobucket.com/albums/v7...kotoyRS4-1.jpg

We normally calculate drive line and tire losses for Quattro to be about 28%. I'm not quite sure what the losses are in the R8's system. But, the R8 should be faster with a better power-to-weight ratio. I suspect that the R8 has not been pushed too hard on the track by the magazines. And, of course, the V10 version should be quite a monster. However, the M3 should post better 1/4 mile times than the RS4 by virtue of a higher power-to-weight ratio, and lower drive line losses.

[Garrett]

Originally Posted by level002

oh, i see how this goes now. an r8 buyer, who's likely a bloke who's got the means to afford damn near any car out there, needs to know that nothing cheaper will keep pace with him - nothing but nothing has more gravity than bang for the buck, his singular goal is to be The Value King.

it doesn't speak too well about you, mate, if you're saying that should something cheaper hang with you while you're 'sitting in your 100k+ sports "supercar"', your day is ruined. and if you want to continue to make assumptions about what ranks how on my list of priorities for a performance car... all day long, my man, have at it.

but, please... for your sake, please don't think that you're going to be upgrading to an m3 and riding right alongside any r8 whose driver is even trying - all you'll be doing is giving him a bit of a chuckle because, really, you'll be in his rearview mirror for as long as he feels like keeping you there.

but... whatever, right? i'll play along - THE 70K M3 WILL HANG WITH THE R8! THE 70K M3 WILL HANG WITH THE 4-DEAD, 12.5, 1.01G 110K R8!!!

cheers

Everything comes down to price. If not we would all be driving Ferrari's on the weekends and M5's to work!

Nobody is trying to take accolades away from Audi with their R8, but it's already been tested and the performance numbers are out and more detailed #'s are emerging. It's an outstanding car and one to compare other car too.

Thats what we are doing.

BMW (as said before) could build a mid-engine V10 that would be about the same price..etc. But that isn't their design philosophy with their current line up of 1, 3 ,5 ,6 & 7 series cars. So there is no point in being stuck up about the R8, Audi's reaching out into other markets (ie Porsche) with their new vehicle, but as a stand-alone performance coupe it's performance CAN be debated/critiqed or even compared to other vehicles.

The R8 is also the epitomy of Audi engineering, that said, it's a shame Audi didn't design a new engine for it and just re-used their 1 and only great engine. FSI 4,2l V8.



All that said, do not underestimate the new 2-door, 4 seat, 18 speaker (800watt) Stereo, w/nav, cupholders, large trunk, power seats, comfort access 420hp V8 we call an M3 ...!!

It's not the same beast as an R8, nor is it ment to compete directly with such a car, but I'm sure the new M3 will "hang" with an R8 on all accounts. To what degree is a matter of debate, but to me that discernable level is moot, because of the price difference and the over all differences in the 2 cars.





-Garrett

[BruceWain]

+1

[OBI_agent]

Originally Posted by swamp2

OK perhaps I am getting too excited with this little software toy but it is damn good and damn fun. Please before you bash this simulation have a look at how well it blindly predicted the numbers in this test. By the way the simulation also completely NAILED the 0-100, 500 and 900 ft numbers as well. I keep getting more and more impressed with the CarTest software. Of course the next criticism will be M-DCT is not out. No biggie, I added a very close estimate of the weight gain, 40 lbs and (rougly) optimized the gear ratios for acceleration (hmmm I guess BMW might do the same...).

Why is the M-DCT M3 almost universally faster (if only by a hair...again insert appropriate comments about "a drivers race")? Look at the curves attached for the power losses (R8 is first). M3 looses about 20% less power to its drivetrain. If you claim the R8 drivetrain is significantly less lossy than this perhaps you can explain why when you tweak these parameters you get a car much faster ijn the simulation than the actual tests.

Now all that is left is for the M3 to best the R8s N'ring time, all this for half the price

Yo awesome post man! M3 is gonna be awesome hopefully!
I want to find out some other cars performance. Where can i find this software?

[T Bone]

Originally Posted by Epacy

Thanks Swamp.
Interesting to see the R8 begin to pull in the M3 after 70mph.

Nice data generation Swamp. What was your shift time variable?

I really like the Cartest software, it predicted the M3 performance well ahead of time.

I wonder how accurate it is predicting how poorly Quattros do at high speed. Even though the R8 looks like it has a smaller frontal area, I bet the M3 will be faster all the way to vMax.

[RI_RS4]

Originally Posted by T Bone

I wonder how accurate it is predicting how poorly Quattros do at high speed. Even though the R8 looks like it has a smaller frontal area, I bet the M3 will be faster all the way to vMax.

T Bone, I provided Swamp actual 3rd and 4th gear acceleration data from when my RS4 was bone stock, and a 1/4 mile time from another RS4. He should be able to use it to calibrate his software to actual performance numbers. As you can see from comparisons to his numbers for the M3, the M3 is a tad bit faster than the RS4, as would be expected given the power-to-weight ratio difference and the drive line loss. If he can use AWD dyno curves, I can provide those, also, in the interest of accurate science.

I have no problem finding that the M3 is the faster of the two cars. I'd expect an M3 to pull on an RS4 after launch.

[T Bone]

Originally Posted by RI_RS4

T Bone, I provided Swamp actual 3rd and 4th gear acceleration data from when my RS4 was bone stock, and a 1/4 mile time from another RS4. He should be able to use it to calibrate his software to actual performance numbers. As you can see from comparisons to his numbers for the M3, the M3 is a tad bit faster than the RS4, as would be expected given the power-to-weight ratio difference and the drive line loss. If he can use AWD dyno curves, I can provide those, also, in the interest of accurate science.

I have no problem finding that the M3 is the faster of the two cars. I'd expect an M3 to pull on an RS4 after launch.

Hello,

See the attached jpeg. Here are the performance parameters that are inputed for the 2005 RS4 for Cartest 2000. I also have a licensed copy and love it.

The issue as I see it is the drivetrain losses are not linear with speed for Quattros. Looking at the numerous high speed videos for Quattros, they seem to suffer from increasing drivetrain losses as speed increases.

So while the software is pretty good for 1/4 mile, I think the accuracy goes a bit off as speed increases.

[swamp2]

Originally Posted by T Bone

Nice data generation Swamp. What was your shift time variable?

I really like the Cartest software, it predicted the M3 performance well ahead of time.

I wonder how accurate it is predicting how poorly Quattros do at high speed. Even though the R8 looks like it has a smaller frontal area, I bet the M3 will be faster all the way to vMax.

I used .001 seconds for both the "engage time" and "shift time". No DCT settings in the software... I chose physically unrealistic times as when looking at the acceleration graphs the total shift time appears to be about .15 s despite these settings. The manual is taking about .35 using .2 and .1 seconds respectively for the parameters above. Must be some other minimum times controlled by a non user controlled parameter. Again, if M-DCT is as good as the rumors we could have another .1 second gain per shift.

Note the power loss curves I posted does show how the R8 power losses stay fairly constant over a wide speed range whereas the MT model used to simulate the M-DCT shows a steadily improving power loss with speed.

That being said the time to distance and time to speed traces from the software for M-DCT M3 vs. R8 are virtually on top of each other, less the small "cut outs" for the R8 while shifting. The small gains if any ina time to speed graph through the gears looks like this:
1 R8
2 Tie
3 M3
4 M3
5 M3
6 Tie
7 Initial lead M3, overtaken by R8

[swamp2]

Originally Posted by RI_RS4

I posted some graphs of 3rd and 4th gear pulls for my RS4 when it was stock that you can use to calibrate your models. To make it a bit easier for you, I can send you the data tables if you like. But I do have the following times that can be used for comparison to a simulation.

We normally calculate drive line and tire losses for Quattro to be about 28%. I'm not quite sure what the losses are in the R8's system. But, the R8 should be faster with a better power-to-weight ratio. I suspect that the R8 has not been pushed too hard on the track by the magazines. And, of course, the V10 version should be quite a monster. However, the M3 should post better 1/4 mile times than the RS4 by virtue of a higher power-to-weight ratio, and lower drive line losses.

Right out of the box with no additional tweaks I find
........................................CarTest
3rd gear 50 to 70 mph - 2.76s 2.61
3rd gear 70 to 90 mph - 2.91s 2.85

4th gear 30 to 50 mph - 3.83s 3.99
4th gear 50 to 70 mph - 3.56s 3.78
4th gear 70 to 90 mph - 3.77s 3.86

Nice! All within about .2 s, some a bit low some a bit high. On average within .05 s.

I also posted a power losses graph for the RS4. Transmission losses are fairly constant acorss speeds (not across rpm in an individual gear through). An average for that looks to be about 78 hp = 19%. The wheel loss is highly speed dependent but if you have to have a single number about 22% looks about right to me so 28% may be a bit too much.

Last but not least CT gives 13.0@105.6 for the RS4.

[jworms]

This software looks pretty cool. I can't wait for the M3 DCT performance tests to start popping up.

Originally Posted by RI_RS4

T Bone, I provided Swamp actual 3rd and 4th gear acceleration data from when my RS4 was bone stock, and a 1/4 mile time from another RS4. He should be able to use it to calibrate his software to actual performance numbers. As you can see from comparisons to his numbers for the M3, the M3 is a tad bit faster than the RS4, as would be expected given the power-to-weight ratio difference and the drive line loss. If he can use AWD dyno curves, I can provide those, also, in the interest of accurate science.

I have no problem finding that the M3 is the faster of the two cars. I'd expect an M3 to pull on an RS4 after launch.

also, i just want to say thank you to RI_RS4 for not turning this thread into another audi vs bmw war. your posts in this thread are way more mature than other audi enthusiasts who have come to the site

[lucid]

Originally Posted by T Bone

Hello,

See the attached jpeg. Here are the performance parameters that are inputed for the 2005 RS4 for Cartest 2000. I also have a licensed copy and love it.

The issue as I see it is the drivetrain losses are not linear with speed for Quattros. Looking at the numerous high speed videos for Quattros, they seem to suffer from increasing drivetrain losses as speed increases.

So while the software is pretty good for 1/4 mile, I think the accuracy goes a bit off as speed increases.

looking at the jpeg, it seems like the software has stored model parameters for several cars. does it say how those were obtained?

[T Bone]

Originally Posted by lucid

looking at the jpeg, it seems like the software has stored model parameters for several cars. does it say how those were obtained?

The software has been around for years so I imagine they have an extensive trial-and-error system couple with some good models.

For example, I emailed the author about the M6 model when I plugged in the numbers and it didn't work out. He suggested playing with the shift times and bang the new numbers were consistent with real world observations. And the parameter adjustments were not unreasonable (for example, in my case, I had my shift times at 0.08 instead of the 0.5 seconds for manual shifting.