BMW M3 M4 RPM Redline Visual, Finally

[Kadema]

Originally Posted by swamp2

I never said that turbo engines are either inferior or always need to be short shifted. Also the numbers are not entire "my" numbers they have been gleaned from both defaults in the software, the literature, papers and books (which of course is based on testing) and a small bit of curve fitting. That along with tons of validation efforts on predicted vs. simulated performance results. Either way the numbers are the same I used for the majority of my E9X M3 simulations. I did not purposefully make the numbers larger to reflect personal engine biases. That would be massive scientific dishonesty. Cheers.

Yes, I somehow feared you'd say something like that and didn't literally imply you'd purposefully "make numbers larger". Maybe I hope you have an unhealthy dogmatic character that unconciously influences your estimation. Why? Of course I hope that BMW stay true to their claim of having constructed a machine that delivers best of both worlds- turbo torque and high revving. I'd probably not buy a car whose dct in sport mode always revvs up to the red line, and does so inefficiently in terms of fuel efficiency and performance, only to make me believe the above mentioned claim is true...

[pkimM3r]

Originally Posted by ScarecrowBoat

this is hands down the biggest nerd fest of a thread i have ever seen

Talk about an ///NFest

What are the extra variables in these formulas to compute tire compound, size, wheel spin, imperfections in the road, size of drivers balls, wind conditions, etc.

[swamp2]

Originally Posted by pkim1079

What are the extra variables in these formulas to compute tire compound, size, wheel spin, imperfections in the road, size of drivers balls, wind conditions, etc.

I initially had some skepticism about the power of physics based simulation to accurately predict vehicle performance. All such skepticism has been pretty well laid to rest (mostly due to my own efforts to test the tool and concepts and author my own tool).

Part of the issue is that tests conducted in the real world also vary. This sounds obvious, but I'm sure you are not appreciating the subtlety involved. There is no one right (test based) answer for any given performance metric for any given car - there is a range of realized results in the real world. This variation is significant. Simulation provides a benefit because it truly removes most of the uncontrollable variations that occur in the physical world. Driver and environmental conditions remain as the most significant variables that contribute to producing this variation. These conditions are ENTIRELY eliminated through simulation in that they can be absolutely identical each and every "test".

To some extent these can be addressed with simulation. Launch rpm, launch technique and tire friction can all be modified from nominal values. Furthermore tire friction both static and dynamic friction are used in the tool I use regularly and the tool I wrote myself. Of course shift times can be altered from novice to expert drag racer to DCT shift times. Furthermore these effects can be isolated and explored individually!

The power of physics as a truly predictive science should not be underestimated. Similarly it's limitations, uncertainties and applicability must be understood and respected.

I'm probably barking up the wrong tree but I hope this helps.

[paddy335]

Originally Posted by ScarecrowBoat

this is hands down the biggest nerd fest of a thread i have ever seen

damn right.....but you've really got to admire someof the work that's gone into some of those posts.

[Carl L]

Originally Posted by b33g33

If you're ever in the ATL area, PM me beforehand and I'd be happy to give you some seat time.

Thanks man Would love to get out that way some day, heard good things about the area.

[pkimM3r]

Originally Posted by swamp2

I initially had some skepticism about the power of physics based simulation to accurately predict vehicle performance. All such skepticism has been pretty well laid to rest (mostly due to my own efforts to test the tool and concepts and author my own tool).

Part of the issue is that tests conducted in the real world also vary. This sounds obvious, but I'm sure you are not appreciating the subtlety involved. There is no one right (test based) answer for any given performance metric for any given car - there is a range of realized results in the real world. This variation is significant. Simulation provides a benefit because it truly removes most of the uncontrollable variations that occur in the physical world. Driver and environmental conditions remain as the most significant variables that contribute to producing this variation. These conditions are ENTIRELY eliminated through simulation in that they can be absolutely identical each and every "test".

To some extent these can be addressed with simulation. Launch rpm, launch technique and tire friction can all be modified from nominal values. Furthermore tire friction both static and dynamic friction are used in the tool I use regularly and the tool I wrote myself. Of course shift times can be altered from novice to expert drag racer to DCT shift times. Furthermore these effects can be isolated and explored individually!

The power of physics as a truly predictive science should not be underestimated. Similarly it's limitations, uncertainties and applicability must be understood and respected.

I'm probably barking up the wrong tree but I hope this helps.

I was just messing around and fully understand your simulation and math. And to piggyback on your post, yes different reviewers will have real world performance tests with different variables. Often times manufacturers will also provide a "tuned" car for reviews so ymmv.

But this car will not do sub 12 sec 1/4mi times. Thats at best with no friction from tires and air .

Id rather have similar 1/4 mi times to the current car but with significant increase in trap speed so that the car is actually faster (not just in such a short distance) if that makes sense.

[V8FunNaturally]

Swamp could you reverse the plotting order of your curves (plotting the highest gear first, lowest gear last), this would better show the optimal shifting points occuring much below redline for the boosted M4 engine.

Also typo correction, in your original post 2nd to last paragraph you typed M4, instead of M3 (E92) in "whereas the M4 tapers off much less dramatically".

These curves show in a mathematical way what could be expected that a turbo engine is going to be much less pleasurable to drive than naturally aspirated. Simply because turbos cannot spool fast enough to keep a low compression ratio engine (off boost) breathing at high RPM AND have zero lag at low RPM. This will not be a driver's car (as far as the engine is concerned for starters) the way the E92 and E46 are.

Yes, it is going to be silly/idiotic kind of fun in the low to midrange RPM because of the available torque, but not the mature linear high reving fun that is much more exhilarating from a naturally aspirated engine.

Now I understand why all the M4 used for the taxi ride BMW pre-views with BMW DTM drivers were all manuals. Any efficiently designed DCT will not use the upper RPM range above the optimal crossing. So it should sound even more dull than the manuals.

The only way to improve this situation with a DCT is to have as many gears as possible, that means not 7 but 10 or more. Obviously this isn't possible and desirable. So that's it, this engine is essentially wothless above those RPMs. Even for mileage. The only benefit for going to redline will be sound and we already know it is not going to be the forte for that engine.

[swamp2]

Originally Posted by pkim1079

But this car will not do sub 12 sec 1/4mi times.

Although its a bit early to speculate with power and weight still not precisely know but I'd be willing to bet that some journalist will crack into the 11s with the car.

Originally Posted by pkim1079

Id rather have similar 1/4 mi times to the current car but with significant increase in trap speed so that the car is actually faster (not just in such a short distance) if that makes sense.

Times and traps correlate pretty well. Yes you can have two cars with the same time and one trapping slower or faster. Either way power to weight is what drives improvement in both. Thus what you hope for is next to impossible.

[swamp2]

Originally Posted by sunsweet

Swamp could you reverse the plotting order of your curves (plotting the highest gear first, lowest gear last), this would better show the optimal shifting points occuring much below redline for the boosted M4 engine.

Not sure what you are asking for but CarTest can not produce different plots for the force in gear vs. speed.

Originally Posted by sunsweet

Also typo correction, in your original post 2nd to last paragraph you typed M4, instead of M3 (E92) in "whereas the M4 tapers off much less dramatically".

Thanks.

Originally Posted by sunsweet

These curves show in a mathematical way what could be expected that a turbo engine is going to be much less pleasurable to drive than naturally aspirated. Simply because turbos cannot spool fast enough to keep a low compression ratio engine (off boost) breathing at high RPM AND have zero lag at low RPM. This will not be a driver's car (as far as the engine is concerned for starters) the way the E92 and E46 are.

I agree but it seems many prefer to short shift, drive lazy and have the midrange over the top end.

Originally Posted by sunsweet

Yes, it is going to be silly/idiotic kind of fun in the low to midrange RPM because of the available torque, but not the mature linear high reving fun that is much more exhilarating from a naturally aspirated engine.

Correction, because of the available power, not the torque...

Originally Posted by sunsweet

Now I understand why all the M4 used for the taxi ride BMW pre-views with BMW DTM drivers were all manuals. Any efficiently designed DCT will not use the upper RPM range above the optimal crossing. So it should sound even more dull than the manuals.

The only way to improve this situation with a DCT is to have as many gears as possible, that means not 7 but 10 or more. Obviously this isn't possible and desirable. So that's it, this engine is essentially wothless above those RPMs. Even for mileage. The only benefit for going to redline will be sound and we already know it is not going to be the forte for that engine.

DCTs will benefit the performance of any car, NA or FI, turbo or supercharged. In fact they can be of even more benefit to a turbo car by helping reduce lag with very short shifts and little to no time off throttle.

However, it remains a good question as to what the DCT shift points will be. Perhaps the differences between redline and optimal shifts will be very small in favor of the optimal point and BMW will take the "phony" route and choose redline just for more excitement. Doubly (or I guess triply when you consider your point about acoustics...) contradictory since the car is already dropping in redline regardless of shift points.

[Boss330]

Totally different engine, but I drove the 428i yesterday. I'm sure that also would come up with a "shift before redline" characteristic. But what a fun little engine!!! No way did I ever feel the need, or even desire, to short shift! It just pulled to the redline in every gear and really seemed to want to be driven that way as well!

That engine won me over by it's willingness to rev and very unlaggy response!

I'm sure that it also can be beneficial to do short shifting during some types of driving.

The thing is, a simulation cannot recreate how a engine sounds and feels in real life!

Not sure whether the 428i has active sound, but it even sounded great on the inside!!!

My guess is that people who dismiss the S55 might be in for a surprise...

[pkimM3r]

Originally Posted by swamp2

Although its a bit early to speculate with power and weight still not precisely know but I'd be willing to bet that some journalist will crack into the 11s with the car.

Times and traps correlate pretty well. Yes you can have two cars with the same time and one trapping slower or faster. Either way power to weight is what drives improvement in both. Thus what you hope for is next to impossible.

So what if traction is horrible in 1st and 2nd gear enough to slow the 1/4 mile time (whether wheel spin or traction control managing the wheels) yet gain huge distances in the higher gears where the end of the quarter mile is just the beginning of easily walking the car next to you?

There is a +10mph difference in the 458 and mclaren mp??? Vs the gtr. Same 12sec times. 10 mph you are gaining and passing pretty significantly at those speeds. Thats all i was referring to.

Plus I believe trap speed is a good indication of power to weight where as times are an indication of traction and launch eg. Awd.

So using your own idea of power to weight, we should be looking at trap speeds not 1/4mi times. No?

[swamp2]

Originally Posted by pkim1079

So what if traction is horrible in 1st and 2nd gear enough to slow the 1/4 mile time (whether wheel spin or traction control managing the wheels) yet gain huge distances in the higher gears where the end of the quarter mile is just the beginning of easily walking the car next to you?

There is a +10mph difference in the 458 and mclaren mp??? Vs the gtr. Same 12sec times. 10 mph you are gaining and passing pretty significantly at those speeds. Thats all i was referring to.

Plus I believe trap speed is a good indication of power to weight where as times are an indication of traction and launch eg. Awd.

So using your own idea of power to weight, we should be looking at trap speeds not 1/4mi times. No?

Sure if traction (tires, suspension, chassis, diff, etc.) are great or AWD is involved 1/4 mi times will be comparatively better but traps won't improve (at the same power level). Similarly power level and CdxA become the most important factors at high speeds as opposed to power/weight which entirely dominates most all typical performance metrics as the single best indicator. No disagreements there. Just generally speaking across all cars, times and traps are well correlated. Because of the overall similarities in the E92 M3 and F82 M4, it wouldn't really make sense for one number to improve but the other not to.

For some "math" behind this have a look here. The quality of the regression (curve fits) for trap speed vs. power to weight is a bit higher than for time vs. power to weight. In the math language the former simply has a larger R^2 value. The difference though is not too large!

Also on the same page note the truly excellent correlation between 0-60 time and 1/4 mi time. Basic common sense, have a high power to weight AND get a good 0-60 (i.e. good launch) and you will have a similarly good 1/4 mi time.

[CanAutM3]

Here are my thoughts of the ideal ratios for the F8X DCT. These ratios would allow to keep the engine mostly in the power plateau (5700-7300); with redline shifts (7500RPM) in each gear that would drop the RPM to 5600 in the next gear. This would also yield a nice lofty mpg oriented 7th gear for 1850RPM cruising at 100km/h. (Note: I am assuming a power curve for which the power drop off at 7500RPM is equal to the power at 5600RPM)


FD 3.154 RPM km/h
1 4.310 7500 70
2 3.218 7500 94
3 2.403 7500 126
4 1.794 7500 169
5 1.340 7500 226
6 1.000 7500 303
7 0.747 1850 100

[pkimM3r]

Originally Posted by CanAutM3

Here are my thoughts of the ideal ratios for the F8X DCT. These ratios would allow to keep the engine mostly in the power plateau (5700-7300); with redline shifts (7500RPM) in each gear that would drop the RPM to 5600 in the next gear. This would also yield a nice lofty mpg oriented 7th gear for 1850RPM cruising at 100km/h. (Note: I am assuming a power curve for which the power drop off at 7500RPM is equal to the power at 5600RPM)


FD 3.154 RPM km/h
1 4.310 7500 70
2 3.218 7500 94
3 2.403 7500 126
4 1.794 7500 169
5 1.340 7500 226
6 1.000 7500 303
7 0.747 1850 100

Also allows not having to drop to second gear much on the track!

[swamp2]

Originally Posted by CanAutM3

Here are my thoughts of the ideal ratios for the F8X DCT. These ratios would allow to keep the engine mostly in the power plateau (5700-7300); with redline shifts (7500RPM) in each gear that would drop the RPM to 5600 in the next gear. This would also yield a nice lofty mpg oriented 7th gear for 1850RPM cruising at 100km/h. (Note: I am assuming a power curve for which the power drop off at 7500RPM is equal to the power at 5600RPM)


FD 3.154 RPM km/h
1 4.310 7500 70
2 3.218 7500 94
3 2.403 7500 126
4 1.794 7500 169
5 1.340 7500 226
6 1.000 7500 303
7 0.747 1850 100

Cool, can you describe the torque curve you used more precisely, perhaps in tabular form as a function of rpm? I would be surprised if given the shape of the torque curve BMW has show us that one could achieve this many redline shifts. I would like to repeat the calculation using your torque curve and redline but using the CarTest methodology.

[Boss330]

Originally Posted by swamp2

Cool, can you describe the torque curve you used more precisely, perhaps in tabular form as a function of rpm? I would be surprised if given the shape of the torque curve BMW has show us that one could achieve this many redline shifts. I would like to repeat the calculation using your torque curve and redline but using the CarTest methodology.

Question:

As I mentioned the other day, I drove the 428i and was really amazed by that engines characteristics. It was really egging you on to take it to the redline. Not just for 0-100km/h acceleration, but also for overtaking or "just for fun". Unlike the M550d I drove a while ago, which really wanted you to shortshift, this engine felt so much more like a highly tuned NA engine in it's power delivery. Much more so than the N55 (in a 640i), which I frankly found a bit underwhelming... That 428i put a BIG smile on my face. Possibly because I was so pleasantly surprised by it. I had driven a N20 320i and that felt more similar to the N55 in it's power delivery. The 428i however was like a different engine!

My point is, the N20 2,8i version, has a power graph as you can see here. Not sure if you want to do all the calculations/analysis, but shouldn't this engine also, on paper, be a engine that should favour short shifting?

It could possibly be that short shifting gives a few tenths better acceleration, but the engine really wants to go to the redline! Which feels really fun and inspiring, unlike the M550d where you just want to change up way before redline.

If the analysis shows that the N20 2.8i engine also "should" be shortshifted, I would say that one should reserve judgement on how the S55 will feel until it has been driven.

[CanAutM3]

Originally Posted by swamp2

Cool, can you describe the torque curve you used more precisely, perhaps in tabular form as a function of rpm? I would be surprised if given the shape of the torque curve BMW has show us that one could achieve this many redline shifts. I would like to repeat the calculation using your torque curve and redline but using the CarTest methodology.

I did not use very detailed power/torque curves. I simply figured gear ratios that would always keep the engine in the 5600-7500 RPM band (1.3393 ratio between ratios). As specified, I assumed that the power at 5600RPM was equal to the power at 7500RPM. Note that this is very basic and does not factor in drivetrain losses.

If you want to input in CarTest, use the following. I am curious to see the results.

RPM Tq Hp
1500 246 70
1600 296 90
1700 346 112
1800 396 136
1900 396 143
2000 396 151
2100 396 158
2200 396 166
2300 396 173
2400 396 181
2500 396 188
2600 396 196
2700 396 204
2800 396 211
2900 396 219
3000 396 226
3100 396 234
3200 396 241
3300 396 249
3400 396 256
3500 396 264
3600 396 271
3700 396 279
3800 396 287
3900 396 294
4000 396 302
4100 396 309
4200 396 317
4300 396 324
4400 396 332
4500 396 339
4600 396 347
4700 396 354
4800 396 362
4900 396 369
5000 396 377
5100 396 385
5200 396 392
5300 396 400
5400 396 407
5500 396 415
5600 396 422
5700 396 430
5800 389 430
5900 383 430
6000 376 430
6100 370 430
6200 364 430
6300 358 430
6400 353 430
6500 347 430
6600 342 430
6700 337 430
6800 332 430
6900 327 430
7000 323 430
7100 318 430
7200 314 430
7300 309 430
7400 302 426
7500 296 422

[CanAutM3]

Originally Posted by Boss330

Question:

As I mentioned the other day, I drove the 428i and was really amazed by that engines characteristics. It was really egging you on to take it to the redline. Not just for 0-100km/h acceleration, but also for overtaking or "just for fun". Unlike the M550d I drove a while ago, which really wanted you to shortshift, this engine felt so much more like a highly tuned NA engine in it's power delivery. Much more so than the N55 (in a 640i), which I frankly found a bit underwhelming... That 428i put a BIG smile on my face. Possibly because I was so pleasantly surprised by it. I had driven a N20 320i and that felt more similar to the N55 in it's power delivery. The 428i however was like a different engine!

My point is, the N20 2,8i version, has a power graph as you can see here. Not sure if you want to do all the calculations/analysis, but shouldn't this engine also, on paper, be a engine that should favour short shifting?

It could possibly be that short shifting gives a few tenths better acceleration, but the engine really wants to go to the redline! Which feels really fun and inspiring, unlike the M550d where you just want to change up way before redline.

If the analysis shows that the N20 2.8i engine also "should" be shortshifted, I would say that one should reserve judgement on how the S55 will feel until it has been driven.

Looking at the N20 power graph in this application, it still displays a rather peaky power curve. Doing a quick estimation based on the graph you provided, if a redline shift brings the RPM in the next gear at or below 4500RPM (the RPM at which power equals power at the 7000RPM redline), then redline shifts are worth it. If RPM in the next gear is above 4500RPM, shifting at redline is not optimal.

[swamp2]

Originally Posted by CanAutM3

I did not use very detailed power/torque curves. I simply figured gear ratios that would always keep the engine in the 5600-7500 RPM band (1.3393 ratio between ratios). As specified, I assumed that the power at 5600RPM was equal to the power at 7500RPM. Note that this is very basic and does not factor in drivetrain losses.

If you want to input in CarTest, use the following. I am curious to see the results.

Results are not at all as you predicted. Shift points are all well below redline. This is with the exact power curve, gear ratios and redline you have used.

Torque is simply falling too fast above ~5500 rpm to "have room" for the force curves not to cross!

[swamp2]

Originally Posted by Boss330

Question:

As I mentioned the other day, I drove the 428i and was really amazed by that engines characteristics. It was really egging you on to take it to the redline. Not just for 0-100km/h acceleration, but also for overtaking or "just for fun". Unlike the M550d I drove a while ago, which really wanted you to shortshift, this engine felt so much more like a highly tuned NA engine in it's power delivery. Much more so than the N55 (in a 640i), which I frankly found a bit underwhelming... That 428i put a BIG smile on my face. Possibly because I was so pleasantly surprised by it. I had driven a N20 320i and that felt more similar to the N55 in it's power delivery. The 428i however was like a different engine!

My point is, the N20 2,8i version, has a power graph as you can see here. Not sure if you want to do all the calculations/analysis, but shouldn't this engine also, on paper, be a engine that should favour short shifting?

It could possibly be that short shifting gives a few tenths better acceleration, but the engine really wants to go to the redline! Which feels really fun and inspiring, unlike the M550d where you just want to change up way before redline.

If the analysis shows that the N20 2.8i engine also "should" be shortshifted, I would say that one should reserve judgement on how the S55 will feel until it has been driven.

If one's "feel" is good it will coincide with the ideal shift points that are based on the engine and transmission characteristics. Now that being said, the performance lost from a non ideal shift point may not be large. The acoustics benefits may be large.

I have no problem to run an analysis for the 320i, just give me hp from 1000-redline in 500 rpm increments, confirm redline and each gear ratio and the final drive. It is quick with this information. That being said, it appears pretty obvious from inspection that this car will need to be short shifted for optimum performance.

[swamp2]

Originally Posted by CanAutM3

I simply figured gear ratios that would always keep the engine in the 5600-7500 RPM band (1.3393 ratio between ratios)

By the way I don't think a (fixed) ratio between ratio is at all a standard transmission gear selection technique. It produces very odd shaped speed in gear curves compared to most similar BMWs. More typical is an even spacing on the rpm vs. vehicle speed plots. Below I've shown yours and the M3 M-DCT. The 1M 6MT is not too dissimilar from the M3.

[Boss330]

Originally Posted by swamp2

If one's "feel" is good it will coincide with the ideal shift points that are based on the engine and transmission characteristics. Now that being said, the performance lost from a non ideal shift point may not be large. The acoustics benefits may be large.

I have no problem to run an analysis for the 320i, just give me hp from 1000-redline in 500 rpm increments, confirm redline and each gear ratio and the final drive. It is quick with this information. That being said, it appears pretty obvious from inspection that this car will need to be short shifted for optimum performance.

Thanks once again. No need to do the 320i

I am really looking forward to the first test drive reviews of the F8x What will the S55 feel like? Does it want to rev etc?

The day I drove the 428i I also drove a E36 M3 3,2. The intersting thing is that the 428i felt more spectacular and entertaining during acceleration. The 428i even had more/better engine noise inside the cabin during hard acceleration and definitely felt quicker. I'm a devoted fan of the BMW I6 engine, but the E36 M3 engine actually felt, dare I say it, uninspiring compared with the N20 in the 428i...

The E36 M3 just has a steady and "slow" rise in revs to redline. The 428i just felt meatier at every stage of the rev scale. I'm sure I still would choose the M3 engine for exhaust sound and smoothness, but that 428i engine just put a big smile on my face. Even though it's slower from 0-100km/h it just feels more spectacular!!

I NEVER thought I'd write what I just did. But that was before I drove the 428i. It felt fast, and it sounded like a sports car engine from the drivers seat, which probably also "helped" my perception of it

Probably active sound?