Significant Recant - The New M3/4s are not Under Rated (long, technical)

[swamp2]

Introduction:

There has been quite some discussion here recently with the S55 in the M3/4 hitting the dynos. Many will recall one of my favorite expressions about dynos – I trust them about as far as I can throw them... Member CanAutM3 has written about reasons why (traditional/inertial) dynos are both inaccurate and imprecise (link). Furthermore, I did some really basic “statistical” analysis of the M3 dyno database backing up these claims showing how most dynos predict a range of results with a 13-17% spread (depending on brand). Further interesting discussion happened in the thread with the Maha dyno results, which by all accounts is hardware, quite a few notches technically (i.e. accuracy and precision) and of course cost above those mentioned in the link above. However, that dyno showed (calculated!) 465 crank hp, which I believe is way too high for the crank hp. Lastly EAS finding the S55 with M-DCT to basically have its quoted crank hp available at the wheels certainly stirred things up. 425 crank hp with a standard 15% loss is ~500 hp at the crank which is just not realistic.

With these observations I decided to head back to the “drawing board” to attempt to gain a better understanding of this situation. I am using a combination of magazine reported performance values, forum member reported test results and two physics based vehicle performance simulation tools, CarTest and Quarter Pro, along with basic hand calculations. For those that have followed some of my prior efforts here, I have used simulation to help make/bolster the claim that BMW typically significantly under rates their turbo engines. I made this claim about both the 335i as well as the F10 M5 and until now had found the same with the S55 engine in the F8X M3/4. This under rating idea has rested on 3 types of evidence all of which lead to this conclusion; dynos (evidence is fairly weak whether the “prediction” is an under rating or over rating), observed performance results and physics based vehicle performance simulation.

Here is the big news and a preview of my surely controversial conclusion: I am significantly, if not entirely adjusting my claim of under rating (at least immediately for both the M4 and F10 M5).

It now appears perfectly reasonable that the prior generation M3 and the current one can maintain their significant observed differences in performance (appropriately a stomping in anyone’s book…) without the need of an under rating of the S55. As many of us have discussed here in the forum prior, the average power (not rpm averaged but time averaged) in the new M4 and M5 are significantly higher than that of the outgoing models. In such cases of highly plateau-ish power curves, an engines peak power is somewhat less informative of the vehicles ultimate performance and certainly less informative in a comparison vs. a car with a linear power curve like that of the S65. Of course 100 lb or so less weight helps here as well along with things like a lower mass and inertia CFRP drive shaft. Although I have embarked on some significant and substantial changes in my approach for modeling losses based on this effort, this does not invalidate all previous wonderful agreements I’ve demonstrated between simulation and test, it does though surely invalidate some of them, with the simulation error (or estimated amount under rating) growing with the engines peak torque. Ultimately this appears to be getting the right answers (performance results via simulation vs. real tests) for the wrong reasons (too high crank power and too high drivetrain losses). Live and learn…

E9X vs. F8X Power Losses in Simulation:

Let’s start with my first key observation that something was awry with CarTest (and of course that means my application of it – after all garbage in, garbage out…). Look at these power loss curves from CarTest for the E9X with S65 vs. F8X with S55, specifically the green traces.



These are the power lost, in hp, during a WOT acceleration run, across gears for the entire driveline from flywheel to hubs. This number, as you can see from the other traces, does not include the aero nor tire losses. The first graph is for the E9X with non-default input losses (for reference my general custom inputs to CarTest for losses are 1% for auxiliaries, 3% for the 6MT, 3% for the differential and 5% for the axles and shafts (12% total). 414 crank hp, as specified from BMW was used for this simulation. Performance for this particular model compares quite well with the upper end of the spectrum of results achieved for the E92 M3 M-DCT from a wide variety of vehicles and magazines. Of course I have accounted for the lightning fast shifts of the M-DCT as well as the US standard 1 foot roll out (timer is frozen until vehicle moves 1 foot) as per US magazine and drag racing standards.



Please compare away vs. test, a wide range of test results are here on this forum. As has been discussed many times here on the forum, just as there is really no one single “right” simulation answer, there also is no one single most correct test result! This was some reasonably strong justification that the S65 was neither under nor over rated (something generally also agreed upon by many of the more knowledgeable enthusiasts here on the forum and elsewhere). CarTest also captured the reduced performance of the heavier E90 M3 sedan with a 6MT as well (significantly slower shift times, results not shown here).

Back to the first graph… Drive train power losses for the E9X M3, let’s just say somewhat arbitrarily, at redline in 3rd gear, are about 63 hp, or right about 15% of peak crank hp. But why does the sum of the input losses, 12% not equal the observed 15%? Well this does indeed have an impact in simulation but certainly would not be the end of the world in itself. There is quite a bit more on this later… Now look at the loss curve for the F80 M3 on the lower graph! For these losses I assumed an under rating such that the engine is putting out 440 crank hp. Peak drivetrain losses are about 99 hp again at redline in 3rd gear. That is a whopping 23% loss and 57% more absolute lost power. This is unrealistic by just about anyone’s standards (way too high) and again my “requested” loss inputs did not change.

What gives, same inputs, radically different outputs? Also, as to the actual drivetrain hardware, we should expect quite similar losses from generation to generation. Gear engagement losses are the primary contributor to drivetrain loss and are already small for each set of two engaging gears, providing perhaps 98-99% efficiency and these shouldn’t change significantly if at all over a 5-10 year period (much more detail on gear loss background and calculations here). There just aren’t any massive improvement margins left in gear efficiency engineering as gears have been highly designed, engineered, refined, tested and simulated for literally thousands of years (well OK only simulated for about 40 years…). Along with ultimate torque capacity and durability, gear efficiency is a core area studied by the engineering community. And of course we are not talking about the use of more efficient straight cut gears as often used in race transmissions, as they are clearly unreasonable from an NVH perspective (strong gear whine). Is there any way the F80 drivetrain “burns” 36 hp/57% more power than the E90 drivetrain? Again, no, just not possible.

Some additional investigation indicates that calculated peak loss percentages in CarTest scale strongly with peak torque but very weakly on peak hp. This has some theoretical justification in Gillespie, Fundamentals of Vehicle Dynamics, but only if “mass factors” are also used which basically convert inertial effects of the drivetrain into effective non-rotational mass. According to the author of CarTest it does not use these mass factor approximations. That is actually quite good as mass factors are also pretty well a gigantic guess and they vary too much depending on the combined transmission and final drive gear ratios (i.e. in particular which gear is selected). Also according to CarTest software’s author there are both inertial “losses” as well as parasitic (true) losses in the graphed loss curves. This appears to be mixing loss “apples with oranges” to some extent. Unfortunately, the author of CarTest, although friendly and knowledgable through email, is not entirely willing or able to explain this situation and observations. However, he was clear that whatever the loss graphs show are what are directly used by the software’s core physics simulation solver (specifically power to the ground = crank power – graphical values for the sum of (drivetrain loss, aero loss and tire loss). Some quick checks demonstrated this is in fact the case just using a form of P = Fv and F=ma,

a = (P(crank)- ΣP(lost, graph))/(v*m) , all metric/SI units, of course

at various single speed/rpm/gear combinations (and of course when there is no wheelspin) the acceleration from this expression matches the predicted CarTest instantaneous acceleration simulation results. In short, from the author, more or less ignore the input dialog boxes for power losses, just look at the graphs… Not an unworkable practical approach but it does require some iteration to achieve the desired real/graphed losses.

New M3/4 Performance Simulations:

The accompanying simulation results below for a hypothetical 440 crank hp F8X car has correlated reasonably well with a few key/early empirical test metrics for the car , however, the cars weight wasn’t provided. I have achieved a much better match between these particular test results and simulation, however, since these results are only Vbox ones we shouldn’t focus too much on matching this particular cars observed results exactly. Metric units are shown since the post with the measured results are also metric. Also these results are simulated with no roll out as per the Vbox set up, thus comparisons with US standard numbers aren’t appropriate even if converting the speed and distance units.



Very good correlation exists for 0-100 km/hr time, 0-400m time (which is very close to ¼ mi) and the standing 1 km time. Many other metrics are not comparing too well. But again some of that was discussed just above and further in that thread. We are at least in the ballpark here.

This is obviously a big problem. If the loss percentages input to the tool do not yield the same loss percentage of peak hp then the software isn’t providing adequate consistency. Furthermore, if a car appears to be performing similarly vs. test but is both under rated and has losses that are too high, those effects can in part or in whole cancel each other. A very practical next step was also just to reverse engineer the losses that appear to be giving very good simulation result for the E9X M3. Yes, to some extent this is just curve fitting, but given prior great agreement, across a wide range of vehicles and performance metrics there is no reason to throw out the baby with the bathwater.

Here is a prior F80 M3 simulation result, along with an updated one, manually adjusting the loss inputs to mimic this same 15% loss of peak hp in 3rd gear.



I was quite surprised, to find a few results quite close (1/4 mi time in particular). We indeed have too much crank hp (a supposed under rating) along with loss estimates too high and the effects counter each other and in some places nearly cancel entirely! Again, these results include a 1 foot rollout.

Here is the kicker. This new “matched loss” F80 M3 (left column) has the factory quoted 425 hp (crank hp), the shape of it’s hp curve is also input precisely to match the factory claimed curve. For the total weight I’ve used 3792 lb, which includes 90% fuel and 165 lb (75 kg) for the driver (source of this weight is post #20 in this thread, a car actually on a scale). This also requires just a hair of torque under rating to as closely match the shape of the digitized factory engine dyno, from 407-415 ft lb across its torque plateau. Really probably not worth mentioning as under rating since the factory claimed value is 406 ft lb, but for completeness, and since we are talking quite a bit here about under rating, I mention it.

For you EU folks concerned with the standing 1 km (with no rollout of course), and since BMW has quoted us a figure at 21.9 s (almost certain this is with M-DCT but not sure if M3 or M4), the number resulting from this updated “matched loss” simulation is 22.1 s, but do note this is a heavier M-DCT F80 vehicle and I suspect the 21.9 s number is for the M4 M-DCT.

What about the M5?

A similar approach for the F10 M5 provides the same conclusion. A consistent 15% peak (redline, 3rd gear) drivetrain loss loosely equates in actual performance results to a 20 hp under rating. Both I and others have indicated that 20 hp seems to be a reasonable estimate for how much the M5 is under rated.



Again comparing to actual test data this comparison looks quite favorable. Note, I did not take nearly the effort here with weights and insuring everything was perfectly “apples to apples” nor examining a broad range of test results (I think fastestlaps.com cherry picks all of the best results for straight line performance). I just wanted to convince myself that I can also recant my prior claim about the F10 M5 being under rated. I don’t think I need to go back and also revisit the 335i, but heck I probably should for the 2008 Nissan GT-R given some of those epic debates here on the forum about under rating in that car. The AWD and the behavior of it’s front-rear and even left-right torque splitting ATESSA-ETS system will surely complicate a reasonable loss estimate for the car under WOT acceleration (most power just after cessation of wheelspin is understood to shift to the rear wheels likely requiring a hybrid efficiency/loss).

In short, I’m pretty well taking a complete reversal here on BMW turbo engine under rating. Of course this is accompanied by an apology to those who I may have too intensely debated this topic with in the past. Knowledge evolves and so do those with the capacity to accept change and even error.

Focus On Wheel Power:

OK, OK enough with this “nonsense” of losses and crank hp you might say. Just use the bloody wheel power curve from a dyno with the corresponding wheel power option in CarTest, completely bypassing losses. How convenient, I thought you would never ask... In particular forum member regular guy has advocated this methodology and has claimed some particularly strong comparisons of test data with simulation. I have carefully digitized the EAS dyno here using an average of the "red" and "green" runs from the first set of dyno runs (i.e. not the highest power run). Here are those simulation results.



Wow, a 425 whp M3 is darn fast. If you don’t believe that a nicely optioned F80 M3 with M-DCT can cover the ¼ mi in 11.6 s trapping at a whopping 121.4 mph then this car is simply NOT making its factory claimed crank hp as whp, period. I would be more than happy to perform a similar analysis using the Maha dyno results, but I can already guarantee that with either 370 whp or 465 crank hp, or however we are supposed to interpret Maha results, such a car will be significantly over performing reality (read that ENTIRE thread) for a bit of a wild ride - some said this dyno “measures” both whp AND drive train losses, until I pointed out that this can’t be done in neutral and also does not account for tire losses).

An Additional Simulation Software Tool – Quarter Pro:

A further step is investigations is an additional physics based vehicle performance simulator that comes highly recommended from forum member bruce.augenstein@comcast, called Quarter Pro (Bruce has usually mentioned “Quarter Jr.” but this is basically the same tool). As engineering professionals who rely on physics based simulation also do in the regular course of their product design work, having a “second opinon” with an additional simulation tool is very valuable and often quite insightful. I have some additional results from Quarter Pro below and have been able to get very close to CarTest results. A big limitation of this tool though is that shift times are unknown and fast DCT type shifts are simply not possible. Thus we will be comparing MT only (which, unfortunately, I have not seen test data for yet). Actual drivetrain inertias must be somehow calculated but the tool provides some nice built in rough calculators which provide estimates based on component and system masses and sizes, which I obtained from official BMW parts catalog (bmwfans.info) and other estimation via google. So, Bruce, if you are reading, do note that Quarter Pro does not include axle losses since most drag cars typically have a solid rear axle and as noted in the Maha thread above, axles are absolutely lossy, power sapping devices. I’m confident this along with the overall content of this post explains some of our prior disagreements about simulation results. However, if not already accounted for, your efficiency for the differential in Quarter Pro/Jr. should very likely be bumped down a couple/few percent to account for axles typical in a road car.

Here are my Quarter Pro (left side obviously) vs. CarTest results for the same 3792 lb F80 6MT M3 (again this includes 90% fuel and 165 lb driver). I have used a 0.3 s shift time in CarTest.



The ¼ mi times are within 0.2 s, the traps within 0.4 mph and the 0-60 times within 0.3 s. As discussed elsewhere the 60 foot time of 1.84 s from Quarter Pro vs. 1.76 s in CarTest is almost for sure going to be a better match with tests on street tires. Both simulations are using the full factory power curve and the stated peak 425 hp (crank) number. Unfortunately, although the input losses appear to be very close, 8% Quarter Pro vs. 7.5% CarTest, we’ve already been through the large variation between input loss and observed loss in CarTest. This is only coincidence as the same strong correlation of results for the E9X M3 in both tools does not have this same correspondence in loss. Either way good correspondence in the results indicates that ultimately, computed wheel powers are close and thus the losses (parasitic or inertial) must also be. This inability to a-priori determine the requisite inputs is certainly a strike against simulation. Additional work should be performed spanning a wider range of vehicles, i.e. heavy low torque vehicles and light high torque vehicles to be able to better determine these inputs a-priori. Also, once I get the full “theory” manual for Quarter Pro I plan to investigate further drivetrain loss and drivetrain interia “loss” as implemented in that software.

Concluding Remarks:

One certainly may critique these efforts in CarTest as a bit of “curve fitting” as I am taking a combination of what I now see was some skill, some luck and some curve fitting for the E9X M3 and other successful simulations, along with a new key observation of unrealistically high losses in CarTest for these “new” high torque BMW engines, to obtain a good match between test and simulation. Either way, I’ve clearly made some mistakes in prior simulations I have posted here by not examining losses in greater detail (in the “Maha” thread in specific see my post #342, in which I clearly dove into this topic in significantly greater detail). Perhaps it doesn’t rise to the level of "the scientific process", but I willing to admit errors, identified by me or by others and incorporate the findings into improved work. This is at the very least a solid and reasonable engineering approach. Despite still lacking a truly firm/fundamental understanding of the loss implementation in CarTest, I believe this and other such software tools are still invaluable and substantially better than any dyno (less the massively spendy, high tech, carefully controlled factory engine dynos, of course). Simulation is better than dynos in terms of offering a broad and detailed prediction of vehicle performance given correct inputs. Given a set of inputs simulation is so powerful because it will always give only one answer, repeatedly, no matter who or how many times the “experiment” is repeated. All variation that occurs in the real world is completely removed. Heck, dynos can’t even predict a hair of real world performance at all without another known nearly identical car to compare to. All this being said, one should not extrapolate to the general conclusion that no vehicles are under rated. This link from the fairly prestigious Rototest Research Institute pretty definitively shows under rating, by as much as 18% does indeed occur (scroll down to the 3rd from last chart on that page). This page also discusses potential errors (variations) from inertial effects and acceleration rates when dyno testing and shows actual test results (in some cases showing as much as 45% variation in power results due only to acceleration rates). Someone really needs to take a similar but more "deep dive" look into inertial dynos to understand why they are so significantly over rating the S55 (and thus other vehicles as well, but then again seem much closer, at least in an average sense, for the S65). “Dynos suck” just isn’t a good enough answer here in my opinion.

Summary:

• Multiple physics based simulation tools, with careful accounting for drivetrain inertia and losses indicate no under rating in the S55 M3/4 nor in the F10 M5.
• Observed performance differences between the prior and current generation M3 may be explained simply through the vehicle weight and shape of the S55 power curve. Average power is considerably higher when an engine produces such a plateau shaped power curve and average power (even more strongly than peak power) dictates performance.
• Prior simulation efforts for many vehicles including the M4 and F10 M5 with crank power levels including under rating had unrealistically high losses and these factors counter each other.
• Crank power estimates from EAS and using a Maha dyno both substantially overestimate wheel and crank power.

[1stm3f80]

Yeah but you forgot to consider the thing a ma gig. And the vertical metal thingy's. In conclusion the Max HP depends on multiple factors such as the position on the do hickey and tangible transmitters. These release powered units of horse torque which crystallize the friction dockets.

[jc05e46m3]

Thanks for the explanation....though, I hope this is part of your job.

I'll leave my testing to real-world conditions with other vehicles and see visual deltas. In Mexico, of course.

[04RC51]

Swamp is a really smart dude, but alot of these threads seem to have overtones of E9x owners in denial. Just me calling it like I see it without any fancy charts to back up my claims so I guess Swamp wins.

[M3 Esq]

Swamp, great as always, but is there an executive summary by any chance?

[GBPackerfan1963]

Originally Posted by M3 Esq

Swamp, great as always, but is there an executive summary by any chance?

It hauls ass.

[swamp2]

Originally Posted by 04RC51

Swamp is a really smart dude, but alot of these threads seem to have overtones of E9x owners in denial. Just me calling it like I see it without any fancy charts to back up my claims so I guess Swamp wins.

Thanks.

Perhaps you missed this part...

Originally Posted by swamp2

It now appears perfectly reasonable that the prior generation M3 and the current one can maintain their significant observed differences in performance (appropriately a stomping in anyone’s book…) without the need of an under rating of the S55.

I've never denied the cold hard facts about the performance numbers themselves since long before we knew power or weight. The performance improvement march WILL continue its steady progression.

Under rating is more of a technical interest and it certainly has relevance to dyno testing. But of course ultimately neither really matter, the real world results do. It is sort of like the specific power argument (hp/liter). High ones are generally more impressive engineering, but ultimately just the peak power is infinitely more important. In this regard we should be pretty much on the same page.

[swamp2]

Originally Posted by M3 Esq

Swamp, great as always, but is there an executive summary by any chance?

Thanks. Good point. I had considered including one given the length and detail. However, the post title itself sums it up about as briefly as possible.

[swamp2]

Originally Posted by jc05e46m3

...I hope this is part of your job.

Just a fun hobby... Car enthusiasts can in general greatly benefit from a bit more science/engineering.

[Boss330]

Impressive as always swamp!!!! Takes a big man to admit being mistaken and also explaining what went wrong. Thumbs up to you!

[jc05e46m3]

Originally Posted by swamp2

Just a fun hobby... Car enthusiasts can in general greatly benefit from a bit more science/engineering.

THANKS!

[consolidated]

So swamp, your continued curiosity about this car is starting to give you away. Are you inching towards jumping in with us?

Thanks for the in depth post as usual, it's great to have one-off members like yourself here.

[swamp2]

Originally Posted by consolidated

So swamp, your continued curiosity about this car is starting to give you away. Are you inching towards jumping in with us?

Thanks for the in depth post as usual, it's great to have one-off members like yourself here.

Despite some negative views about the car and BMW M in general, I remain a fan of the car. I'm just not as big of a fan as I genuinely wanted to be. It has never been entirely crossed off the list as a potential. As I often say, as fast as a 991 C2S in a drag race at 2/3rds the price and potential for 4 useable seats is very compelling.

[swamp2]

Originally Posted by Boss330

Impressive as always swamp!!!! Takes a big man to admit being mistaken and also explaining what went wrong. Thumbs up to you!

Thanks, very kind words.

[gthal]

I understand that a dyno is best used to compare changes to a car versus actual power. Having said that, what I don't understand, then, is why does the F8X dyno so much higher than the E9X at the wheels? There is a SIGNIFICANT increase in WHP/torque variance between the cars on multiple different dynos... much more than the difference in their stated crank ratings. Any explanation for this?

P.S. All I can say is WOW to the time/effort you put into this!

[consolidated]

Originally Posted by swamp2

Despite some negative views about the car and BMW M in general, I remain a fan of the car. I'm just not as big of a fan as I genuinely wanted to be. It has never been entirely crossed off the list as a potential. As I often say, as fast as a 991 C2S in a drag race at 2/3rds the price and potential for 4 useable seats is very compelling.

Very much agree...with the latter half. As a luxury and value proposition it's still the car to beat. Even though the engine has been a disappointment for some of the faithful I believe the chassis (and remainder of the car including amenities, interior etc) more than make up for it. It truly is that much better dynamically in my opinion. And we get brakes to be proud of finally.

[CanAutM3]

Some observations:

Looking at the input data in Quarter Pro, have you considered using a better efficiency in 5th gear relative to all the other gears? Since 5th is a direct drive, the power transmissions does not go through the two gear sets that all the others gears need to use, hence the better efficiency.

Also looking at the comparison between Quarter Pro and Car Test, the traction model in Quarter Pro seems to better estimate the 0-60ft where CarTest seems a bit too optimistic (>1g acceleration to 41.5mph).

[GrussGott]

Originally Posted by swamp2

Just a fun hobby... Car enthusiasts can in general greatly benefit from a bit more science/engineering.

You might want to start class with something simpler like:

suck, squeeze, bang, blow that's what makes your engine go. Questions?

[swamp2]

Originally Posted by gthal

I understand that a dyno is best used to compare changes to a car versus actual power. Having said that, what I don't understand, then, is why does the F8X dyno so much higher than the E9X at the wheels? There is a SIGNIFICANT increase in WHP/torque variance between the cars on multiple different dynos... much more than the difference in their stated crank ratings. Any explanation for this?

P.S. All I can say is WOW to the time/effort you put into this!

Correct.

Inertial effects in chassis/inertial dynos are clearly involved. Have a look at the link in the main post to Rototest Research Institute. Engine/crank dynos at the OEMs are very carefully done in steady state WOT, maximum load conditions. Inertial dynos are done in a dynamic fashion. It seems that dynos are not corrected for the rate of acceleration. Although this certainly is an understood factor, on the surface the effect seems in the opposite direction from observation. If the new car has much more power/torque it will accelerate a given roller with a fixed inertia at a higher increase in rpm vs. time that would lose more power to drivetrain inertia registering a lower power level. Unfortunately, as I stated in my post, this needs significantly more investigation/insight. I clearly don't have the answer to this very good question. Almost 100% guaranteed none of the dyno operators do either.

Thanks.

[swamp2]

Originally Posted by consolidated

Very much agree...with the latter half. As a luxury and value proposition it's still the car to beat. Even though the engine has been a disappointment for some of the faithful I believe the chassis (and remainder of the car including amenities, interior etc) more than make up for it. It truly is that much better dynamically in my opinion. And we get brakes to be proud of finally.

No disagreement there!

[swamp2]

Originally Posted by CanAutM3

Some observations:

Looking at the input data in Quarter Pro, have you considered using a better efficiency in 5th gear relative to all the other gears? Since 5th is a direct drive, the power transmissions does not go through the two gear sets that all the others gears need to use, hence the better efficiency.

Also looking at the comparison between Quarter Pro and Car Test, the traction model in Quarter Pro seems to better estimate the 0-60ft where CarTest seems a bit too optimistic (>1g acceleration to 41.5mph).

Despite some of my criticisms of Quarter Pro (in general massively fewer inputs and outputs) the loss by gear is a very nice feature and your suggestion here is probably perfectly valid. I've followed some of the other discussions here on less gear engagements required for certain gears. However, are you certain that a 1:1 ratio absolutely means there should be 1/2 the losses in that gear (for this or in general for all MT transmissions exhibiting a 1:1 ratio). Is the same effect going to be present also in all DCTs with a 1:1? Now lastly the results for all of the metrics output from Quarter Pro don't depend on 5th gear, so in that respect this detail isn't at all relevant.

Fully agree on the latter comment as well. The traction/wheelspin models in Quarter Pro seem to much better capture acceleration right of the line during wheelspin.

[swamp2]

Originally Posted by GrussGott

You might want to start class with something simpler like:

suck, squeeze, bang, blow that's what makes your engine go. Questions?

There are quite a few surprisingly astute folks here on this forum and we have been having (mostly) friendly debates on this nerd stuff for years. Many will understand every bit of what I've written (well at least 95%). I've also learned a lot from many of them. I do appreciate your humor though!