10-05-2013, 04:30 PM | #199 | |
General
21117
Rep 20,741
Posts
Drives: 2021 911 turbo
Join Date: Jan 2012
Location: Montreal
|
Quote:
I could understand a very slight difference due to parasitic losses in the gearbox, but not this much of a difference. Let me explain. Let's take the 2-3 shift for example. Wheel torque in 2nd gear at 7500RPM is 3004lb-ft (296 x 3.218 x 3.154) and in 3rd at 5600 is 3001lb-ft (395 x 2.402 x 3.154). If you do the math for any RPM shift below 7500, you will find that wheel torque in 3rd will be lower than in 2nd, this means that shifting before redline would not be optimal. The same also applies to all gears. One element that my estimations does not consider, is a portion of the parasitic losses in the gearbox. The torque and power numbers are at the flywheel, so any losses upstream of the gearbox are already accounted for in the torque numbers. For a given road speed (shift point), the drivetrain losses downstream from the gearbox are the same for both gears (drivetrain is spinning at the same speed). The CarTest output suggest a 7200RPM optimal shift point from 2nd to 3rd. This implies that 2nd gear at 7200RPM has 25hp more loss in the gearbox compared to 3rd at 5375RPM. I don't buy it. Inside the gearbox, the losses caused by the higher engine RPM would be in the faster spinning clutches, input shaft, lay shafts and gears. IMO, not enough to make such a significant difference. Last edited by CanAutM3; 10-22-2013 at 04:24 PM.. |
|
Appreciate
0
|
10-05-2013, 04:38 PM | #200 | |
General
21117
Rep 20,741
Posts
Drives: 2021 911 turbo
Join Date: Jan 2012
Location: Montreal
|
Quote:
But BMW will likely stick with a more traditional spacing like they did on the M5. If it is the case, it is a shame, because it would not allow to fully exploit the capabilities of the S55 . I can only hope |
|
Appreciate
0
|
10-05-2013, 04:49 PM | #201 | |
General
21117
Rep 20,741
Posts
Drives: 2021 911 turbo
Join Date: Jan 2012
Location: Montreal
|
Quote:
Last edited by CanAutM3; 10-05-2013 at 04:55 PM.. |
|
Appreciate
0
|
10-05-2013, 08:18 PM | #202 | ||
Lieutenant General
609
Rep 10,407
Posts |
Quote:
Quote:
Transmission Differential Axles Some of these move at road speed (rpm) whereas some move at engine rpm (well technically the tranny moves at both...). The total drivetrain loss I am using for these predictions is 11% which is over 40 hp. In addition to this 40 hp you must also account for the effects of drivetrain intertia. This also acts very similar to an actual loss (even though iti is not a truly dissipative loss which converts to heat, the energy is used up to angularly accelerate all rotating components) and is about the same size in peak value. Thus peak total "losses" can be as much as 80 hp. A typical time average total loss might be about 50 hp (for this M4). I don't think I really need to respond to the rest of your post. Build a spreadsheet, run the numbers. You can also refer to Chapter 2 in "Fundamentals of Vehicle Dynamics" by Gillespie. He covers in detail the calculations required and points out the huge importance of losses and drivetrain intertia terms in the overall equations. I don't only use CartTest, I've also written a complete detailed spreadsheet program to do the same calculation, it is about 5000 or so rows just to give you a feel for the size and complexity. Of course shift points can be calculated in a fraction of that. Not to sound too condescending but it will significantly help your overall understanding to go through the exercise yourself. Or heck purchase CarTest. Cheers.
__________________
E92 M3 | Space Gray on Fox Red | M-DCT | CF Roof | RAC RG63 Wheels | Brembo 380mm BBK | | Vorsteiner Ti Exhaust | Matte Black Grilles/Side Gills/Rear Emblem/Mirrors | | Alekshop Back up Camera | GP Thunders | BMW Aluminum Pedals | Elite Angels | | XPEL Full Front Wrap | Hardwired V1 | Interior Xenon Light Kit | |
||
Appreciate
0
|
10-06-2013, 01:17 AM | #203 | |
General
21117
Rep 20,741
Posts
Drives: 2021 911 turbo
Join Date: Jan 2012
Location: Montreal
|
Quote:
I do have a good enough understanding of physics and dynamics. I have done this type of calculation numerous times both professionally testing aircraft engines and personally developing track/race cars. And yes, I have built spreadsheets to calculate shift points on several occasions. Chapter 2 - "Acceleration Performance" does provide a good overview of the losses and effects of drivetrain inertia. I am not disputing any of the principles exposed there. However, when calculating shift points, as I explained in my previous post, only a small fraction of the total drivetrain friction and inertia losses have an impact. When the car travels at 94km/h (using my previous numbers), be it in 2nd at 7500RPM or in 3rd at 5600RPM, the losses downstream of the transmission (driveshaft, differential, axle shafts, wheels) will be identical. The only differences are the losses in the transmission and in the engine inertia generated by the 1900RPM difference in engine speed. I don't believe that this delta loss (not total loss) can equate to 25hp. I still could be mistaken though and will keep an open mind, if you are able to precisely point where my logic is flawed. Last edited by CanAutM3; 10-18-2013 at 05:58 PM.. |
|
Appreciate
0
|
10-06-2013, 02:01 AM | #204 | |||
Lieutenant General
609
Rep 10,407
Posts |
Quote:
Quote:
Quote:
P.S. If the moment of inertia of the components of the entire drivetrain are known, mass factors are not required but except at the OEM these are rarely if ever known, hence the use of mass factor. P.S.S. I got this whole topic wrong myself here on this forum long ago in a long debate. I was making more or less the same error you are - ignoring the inertial terms and how significant they are.
__________________
E92 M3 | Space Gray on Fox Red | M-DCT | CF Roof | RAC RG63 Wheels | Brembo 380mm BBK | | Vorsteiner Ti Exhaust | Matte Black Grilles/Side Gills/Rear Emblem/Mirrors | | Alekshop Back up Camera | GP Thunders | BMW Aluminum Pedals | Elite Angels | | XPEL Full Front Wrap | Hardwired V1 | Interior Xenon Light Kit | |
|||
Appreciate
0
|
10-18-2013, 04:41 PM | #205 | |
General
21117
Rep 20,741
Posts
Drives: 2021 911 turbo
Join Date: Jan 2012
Location: Montreal
|
Quote:
Using "equivalent mass" for the different drivetrain components would be more precise. I would have thought that Cartest would be using various moments of inertia (engine/flywheel/transmission, transmission/driveshaft/diff and axleshafts/wheels) as input assumptions. I stick with my view that all the inertias (as well as friction losses) downstream of the transmission (drive shaft, diff, axle shafts and wheels) do not have any impact on the shift points. As I stated in my previous post, only engine and transmission inertia have an impact on optimal shift points. What I do realize though is that I may have underestimated their impact on optimal shift points; especially so in the lower gears. I need to go back to the text books and do some calculations . In principle, the engine inertia has much less impact in the higher gears. Yet, the Cartest results still show shift points around 7200RPM in all gears; I would have expected higher shift points in the higher gears. Any thoughts why? Last edited by CanAutM3; 10-19-2013 at 10:12 AM.. |
|
Appreciate
0
|
10-18-2013, 06:18 PM | #206 |
Private First Class
1
Rep 116
Posts |
Really an interesting read, CanAutM3 and swamp2. Thank you! I have no deeper engineering knowledge and struggle to understand why engine inertia should have an impact at all. Isn't that already 'included' in power or torque charts?
Also, like CanAm, I do think that downstream of the transmission, there should be only speed dependent losses, no rpm dependent losses, hence no impact on shift points? |
Appreciate
0
|
10-18-2013, 06:38 PM | #207 | |
General
21117
Rep 20,741
Posts
Drives: 2021 911 turbo
Join Date: Jan 2012
Location: Montreal
|
Quote:
Bench dynos are brake dynos that measure engine output at the flywheel. These are the type of dynos used by car companies to develop engines and publish official engine outputs. Usually the numbers produced by such dynos are obtained in steady state and are much more precise and repeatable. Since the engine is not accelerating when the torque is measured, the engine inertia is not accounted for. However, chassis inertia dynos have to be operated in transient mode. They calculate the engine output based on the capability of the drivetrain to accelerate a flywheel (the dyno drum). Numbers obtained from such dynos will include the losses created by the drivetrain inertia. Any dyno operated in transient mode will include interia losses, what makes it complicated is the fact that the inertia losses vary depending on the acceleration rate. Combining results from steady state and/or different transient runs allows to back calculate the moment of inertia of the drivetrain components. This begs another question regarding Cartest. When inputing the power/torque numbers, how does Cartest know if they were obtained in steady state (excluding inertia losses) or transient (including inertia losses)? Last edited by CanAutM3; 10-18-2013 at 10:43 PM.. |
|
Appreciate
0
|
10-18-2013, 06:53 PM | #208 |
Private First Class
1
Rep 116
Posts |
Ah, I see, thanks again. Now I think I also understand why you said inertia losses have less impact at higher gears. But iirc the shift points swamp2 posted did get lower in higher gears. Isn't that contradictory?
|
Appreciate
0
|
10-18-2013, 07:39 PM | #209 |
General
21117
Rep 20,741
Posts
Drives: 2021 911 turbo
Join Date: Jan 2012
Location: Montreal
|
Yes it is. And this is why I am asking Swamp for his thoughts... there might be other parameters that have an impact.
Last edited by CanAutM3; 10-18-2013 at 10:41 PM.. |
Appreciate
0
|
10-19-2013, 12:27 AM | #210 |
Private First Class
1
Rep 116
Posts |
And here comes another question... Aren't engine inertia losses largely saved as the momentum is entrained (is this correct english?) when you shift up? I.e. if you shift early to avoid engine inertia losses you also have less momentum, thus short shifting didn't help that much? In other words: If you use force-at-the-wheels to road speed charts that include engine inertia losses but don't include engine inertia gains you'll find shift points that actually are too low? Then again, I might be completely wrong.
Last edited by Kadema; 10-19-2013 at 01:36 AM.. |
Appreciate
0
|
10-19-2013, 05:42 AM | #211 | |
General
21117
Rep 20,741
Posts
Drives: 2021 911 turbo
Join Date: Jan 2012
Location: Montreal
|
Quote:
Last edited by CanAutM3; 10-19-2013 at 07:16 AM.. |
|
Appreciate
0
|
10-19-2013, 03:15 PM | #212 | |
Lieutenant General
609
Rep 10,407
Posts |
Quote:
1. Peaks -Peak hp, rpm of peak -Peak torque, rpm of peak torque -redline Then the software does a very good job of curve fitting to come up with a realistic curve. Although some refinement can improve performance prediction this method is generally pretty accurate. 2. Actual curves at the crank -A full power OR torque curve at the engines crank This captures a bit more subtlety of the shape of the curve. It is more important for non linear power curves (typical turbo) and less important for a typically linear power curve (E9X M3). Redline is also needed. 3. Wheel values -Again power or torque by rpm and the other curve is computed. Redline is also needed.
__________________
E92 M3 | Space Gray on Fox Red | M-DCT | CF Roof | RAC RG63 Wheels | Brembo 380mm BBK | | Vorsteiner Ti Exhaust | Matte Black Grilles/Side Gills/Rear Emblem/Mirrors | | Alekshop Back up Camera | GP Thunders | BMW Aluminum Pedals | Elite Angels | | XPEL Full Front Wrap | Hardwired V1 | Interior Xenon Light Kit | |
|
Appreciate
0
|
10-19-2013, 03:31 PM | #213 | |
Lieutenant General
609
Rep 10,407
Posts |
Quote:
Imagine this thought (or real) experiment. Let a block with simple wheel(s) attached slide down an inclined plane. Here the only work done is by gravity not an engine. If you compared the speed of the block/wheel assembly at the end of a certain height/length incline it will be going a different speed if the axle is locked or unlocked (of course assume no complicating friction effect whether or not axle is locked). When the axle is unlocked there is extra rotational energy stored in the wheels angular momentum. In both cases gravity did the exact same amount of work but again the final speeds will be different! In each successive gear the flywheel, transmission shafts, driveshaft, differential axles (and wheels) all must be "spun" up over and over again (regardless of the small conservation of angular momentum effect depending on shifting style). Some of these terms depend on road speed and some depend on engine speed. Again, in a typical sport road car, the car may actually accelerate as if it has as much as 15-25% more total gross mass than in weighs just from these inertial effects. The effects are made larger by high power/high redline engines (compared to low redline ones typically with lower gear ratios) because the terms have factors which are the square of the gear ratios. Again a 3000 lb vehicle accelerates much more like a 3600 lb one in first gear solely due to drive train interial effects. The effect is clearly not small. The thought experiment here is that if you could magically make all of the relevant spinning components have no mass, ones vehicle would accelerate just like it was ~600 lb lighter. The engine (crank, pistons and flywheel) intertias are all important each time these components are accelerated. This is unrelated to the particular method that establishes the crank power of the engine. These terms are only in effect when the entire vehicle is accelerating. Steady state power measurement does basically eliminate these inertial effects from the power rating. Bringing things full circle shift points depend on the actual dynamic force delivered to the wheels in each gear, which in turn depends on both parasitic losses and drivetrain inertia.
__________________
E92 M3 | Space Gray on Fox Red | M-DCT | CF Roof | RAC RG63 Wheels | Brembo 380mm BBK | | Vorsteiner Ti Exhaust | Matte Black Grilles/Side Gills/Rear Emblem/Mirrors | | Alekshop Back up Camera | GP Thunders | BMW Aluminum Pedals | Elite Angels | | XPEL Full Front Wrap | Hardwired V1 | Interior Xenon Light Kit | Last edited by swamp2; 10-19-2013 at 03:38 PM.. |
|
Appreciate
0
|
10-21-2013, 06:27 AM | #214 | |
Banned
17
Rep 394
Posts |
Quote:
Your focus on the redline seems misplaced, since the car's power curve seems to plateau after about 6500 and the demonstration videos I've seen had the driver shifting well before the redline. Once we start considering aftermarket FI solutions, the S65 wins, so let's not go there. Regardless, I'm holding out on making any final judgements until I drive the car myself. I would love to have the new M3 (love the new sedan's wide body) as a fun daily driver that I don't feel guilty putting 20s on and doesn't kill me on the gas, and then gut and supercharge my E92 to be a true monster. Plus, that way I avoid having that heathen 4 after the M. Savages. |
|
Appreciate
0
|
10-21-2013, 10:14 AM | #215 | |
Major General
1715
Rep 5,110
Posts |
Quote:
The E30 M3 LOOKED like a "pure, raw gocart", but wasn't really that fast. Especially compared with the Ford Sierra RS Cosworth that was one of it's direct competitors. The three door Cosworth looked even more aggressive than the E30 with it's big (functional) rear wing, deep front spoiler and wide fenders. In reality the E30 M3 was a E30 with a body kit and a 2,3l engine that developed well under 100hp/liter. Very minor suspension upgrades to the front suspension, rear suspension was carried over from the E30. The street version of the E30 M3 had all the "creature comforts" available at the time, and was just as comfortable to drive as would be expected of a production vehicle. But E30 will always be considered a classic M3. Then the E36 came, which possibly is the least special M3 so far. It has just a minor bodykit, no wide fenders etc. The engine and handling is good, but just like the E34 M5, it doesn't look like a special car. Seems like the M3 that is the least different from the std. production models. It even came as a 4 door and was much more likely to find use as a family carrier than the E30. Most likely to become the Ferrari 348 of the M3 range. Then we have the E46, which returned to the E30 M3 legacy of special looks. But this time also backed up with a strong engine. The E46 was a return back to the original idea, and especially models like the CSL will ensure that the E46 will be remembered as a classic M3. The E9X model is something between the E46 and the E36, but more towards the E46. It's got a wide body and crazy engine (I'm not a fan of a V8 M3, but the S65, like the S85, will probably be remembered as some of the great M engines). But you could get a 4 door and it became HEAVY... With the F8x models it seems we will be getting the best handling M3 ever and probably the M3 that differs the most from the car it's based on. Wide body and mechanical differences in suspension beyond what has been seen before. And the engine is also based around a completely new block with different bore spacing etc from the N55, requiring new tooling and means that it's a M3/M4 specific engine, not just a tuned N55. IMHO the F8x looks like it will have even more M specific development than any M3 before it, and I'm pretty sure BMW hasn't done all of that work with light weight components and suspension development just to create a car that isn't a better track car than it's predecessor... Until we have read independent reviews, we have no way of knowing that the F8x won't be as thrilling to drive... I think many will be in for a surprise on how good the F8x actually will be (or at least i hope it will be that good ) Last edited by Boss330; 10-21-2013 at 10:26 AM.. |
|
Appreciate
0
|
10-21-2013, 10:33 PM | #216 | |
General
21117
Rep 20,741
Posts
Drives: 2021 911 turbo
Join Date: Jan 2012
Location: Montreal
|
Quote:
This is the same reason I think that only the engine and part of the transmission inertia have an effect on optimal shift points. I still need to take some time and do the math . BTW, any thoughts as to why CarTest recommends similar shift point (7200RPM) in every gear in our previous simulation? Last edited by CanAutM3; 10-21-2013 at 10:41 PM.. |
|
Appreciate
0
|
10-21-2013, 10:38 PM | #217 | |
General
21117
Rep 20,741
Posts
Drives: 2021 911 turbo
Join Date: Jan 2012
Location: Montreal
|
Quote:
What are you thoughts on the E36 engines? I thought folks in Europe were pretty pampered with the 286ps and 321ps engines compared to the lowly 240hp we had here in North America. I had the opportunity to drive the 286ps 3.0L and found it pretty awesome. I think the 3.2L variant was the first ///M street engine to achieve 100ps/l... |
|
Appreciate
0
|
10-22-2013, 12:27 AM | #218 | ||
Lieutenant General
609
Rep 10,407
Posts |
Quote:
Quote:
I do not recall any simulations I provided that showed a vehicle shifting always at or around 7200 rpm. We talked about and older 5er here in this thread as well as the M4. Both share the trend of a at or near redline shift in the first gear (or two) followed by an ever decreasing ideal shift point.
__________________
E92 M3 | Space Gray on Fox Red | M-DCT | CF Roof | RAC RG63 Wheels | Brembo 380mm BBK | | Vorsteiner Ti Exhaust | Matte Black Grilles/Side Gills/Rear Emblem/Mirrors | | Alekshop Back up Camera | GP Thunders | BMW Aluminum Pedals | Elite Angels | | XPEL Full Front Wrap | Hardwired V1 | Interior Xenon Light Kit | |
||
Appreciate
0
|
10-22-2013, 10:14 AM | #219 | |
Major General
1715
Rep 5,110
Posts |
Quote:
I drive a F10 with the N53 NA 3l engine. And even though the 3,2l M3 has more top end power, the 3l only has 14hp more... (272 vs 286). And frankly they aren't that different. At the time the engine was great, but today it really doesn't feel that special IMHO. |
|
Appreciate
0
|
10-22-2013, 04:23 PM | #220 | |
General
21117
Rep 20,741
Posts
Drives: 2021 911 turbo
Join Date: Jan 2012
Location: Montreal
|
Quote:
|
|
Appreciate
0
|
Post Reply |
Bookmarks |
|
|