First driving reviews of M3 M-DCT / M-DKG transmission!

[Hammer]

Originally Posted by southlight

Added the last part to my post. Hope it's somewhat comprehensible.


Best regards, south

ach mach dir nix draus, gut gemacht!

well done the translations. now i ve no work left

[ersin]

Originally Posted by footie

If BMW offers equally as quick acceleration from the lesser modes than no harm done, but if not then it will be a major F--k up in my opinion.

I'm confused now. Why would having less acceleration in the lesser modes be a bad thing? If you want to accelerate fast then choose S5 or S6. You want a leisurely drive then choose S1 or S2 or A1, A2, etc.

It also sounds like the actual shift speed is dependent upon throttle position too just like the SMG II was. But, I'm not 100% convinced of this yet. This would make sense: you want fast, you push the throttle pedal farther down so the system shifts faster for you too.

[southlight]

Originally Posted by Hammer

ach mach dir nix draus, gut gemacht!

well done the translations. now i ve no work left

Danke, sehr nett.


Best regards, south

[swamp2]

Originally Posted by T Bone

The Cartest2000 simulations don't support significantly faster acceleration for the MDCT.....the reason the MDCT is faster is quicker shift times and not gear.

Depends on what you mean by significant and depends on the particular contest as well.

BMW has released official acceleration differences, but yes only to fairly low speeds and indeed a couple tenths here or there. However, my CarTest runs are quite consistent with BMW results. We are only talking about a couple tenths advantage up to 100, past 100 at 120, 140 and 150 the time to speed advantages grow to about 0.5 to 1.0 seconds. However, time to speed contests to Vmax are won by the 6MT. It is all put in much better perspective looking at simulation results for distance vs. time, M-DCT will have about 100' or 7 car lengths at 150 mph. It should get the jump right from the start and continue to pull away most noticably at each shift.

On to what matters more, shift times or transmission ratios. You need to be really clear to make this comparison. Are we talking about time to a given speed, in gear roll ons with time for speeds 1 to speed 2, distance ahead in a race, etc.? What I can say is that simulation indeed supports your claim that MORE benefit comes from the shift time advantage as opposed to gear ratio differences for timed acceleration runs to a given speed.

For in gear time, from speed 1 to speed 2, without allowing gear shifts M-DCT really exhibits a significant advantage over 6MT. All of this is obviosuly from gear ratios. Lastly giving both cars a chance to use the most appropriate gears this advantage grows even stronger for times from speed 1 to speed 2 ALLOWING shifts. Here both the shift times and gear ratios really contribute.

As you know when in the 400-500 hp range it takes a huge addition in power to make even minor improvements in performance. Simply the law of decreasing marginal return. 20 hp is absolutely significant and I stick by my conclusions that M-DCT is good for about an equivalent +20 hp, +20 ft lb compared to 6MT.

Cheers.

[swamp2]

Originally Posted by T Bone

I will tell all of you guys this.... With SMG2, the E46 M3 didn't have enough power to really showcase the upshift tire chirp phenoma.... The M5 / M6's V10 with SMG3 does and it is a GLORIOUS sensation and not an annoyance.

With the E9x M3, it will have enough power to make this chirp....

Don't fear it, it is like a good thing.

ABSOLUTELY. Don't forget one of my favorite videos of the M6 getting a bit of "scratch" all the way into 5th. Play this mutha loud!

This is simply heavenly! I wonder if the M3 will "chirp" into 4th, I'd say 3rd for sure.

[footie]

Originally Posted by ersin

I'm confused now. Why would having less acceleration in the lesser modes be a bad thing? If you want to accelerate fast then choose S5 or S6. You want a leisurely drive then choose S1 or S2 or A1, A2, etc.

It also sounds like the actual shift speed is dependent upon throttle position too just like the SMG II was. But, I'm not 100% convinced of this yet. This would make sense: you want fast, you push the throttle pedal farther down so the system shifts faster for you too.

I don't want to go into the old debate of the amount of modes on offered, been there, done that and it's a point that some of us will not see eye to eye on.

My only concern is that I hope BMW offer equally as good an accelerate with modes which don't disconnect the DSC and demand the jerk/surge. My guess is this was included to appeal to the current SMG customers. My understand on the technology is that each of the modes should be capable of shifting equally as quick as each other, the only difference I see between S1 and S6 would be this build-up of free energy. What difference this will make to the acceleration I don't know but until I sample it I don't know how it will feel.

Who knows, I might even like it.

The objective of dual clutch systems is to change gear without any noticeable interruption in power and forward motion, noticeable interruption to me means smooth transition between changes. That might be Audi's interpretation of this, BMW might have an altogether different opinion of that. I wish I knew someone who has driven the new GTR or EVO to get an opinion on their gear changes to see it each manufacturer has a different take on what Dual Clutch should behave like.

Steved or anyone?

[enigma]

Originally Posted by footie

T-Bone,

If BMW offers equally as quick acceleration from the lesser modes than no harm done, but if not then it will be a major F--k up in my opinion.

You clearly have no understanding of phyisics. It would be great if I could stop from 100mph in 0.as without a jerk also, but thats equally problematic.

Where do you think the energy of fast spinning engine goes? Options are

1: Quickly to the rear tires producing a jolt of extra acceleration
2: Gradually to the rear tires while reducing engine output to allow the engine to slow without extra acceleration.

#1 = S6
#2 = S1

And you have 4 other choices between them to suite the drivers taste. This is why people buy BMW and not the "one size fits all" audi solution. There will be a performance difference. Its unavoidable.

[footie]

enigma,

Can you explain then why does the Veyron has no jerk in it's acceleration, after all it is the quickest accelerating car in the world (production spec). Surely by your reckoning if VAG wanted the Veyron it be even quicker they would have offered the equivalent to BMW's mode S6.

Also Audi quote an identical 0.2s difference between the manual and DSG system that BMW quote for theirs, so this proves to me that this horse shit of Jerk/surge is manufactured.

There seems to be an arrogance amount some of you when discussing things, it seems like the only way to try and win a debate is to talk down to someone. Bad form.

[enigma]

Originally Posted by footie

enigma,

Can you explain then why does the Veyron has no jerk in it's acceleration, after all it is the quickest accelerating car in the world (production spec). Surely by your reckoning if VAG wanted the Veyron it be even quicker they would have offered the equivalent to BMW's mode S6.

Also Audi quote an identical 0.2s difference between the manual and DSG system that BMW quote for theirs, so this proves to me that this horse shit of Jerk/surge is manufactured.

There seems to be an arrogance amount some of you when discussing things, it seems like the only way to try and win a debate is to talk down to someone. Bad form.

Car companies quote many things. I seriously don't know how to have a debate with you when you flat out deny what most everyone else on this thread realizes. Seriously, its not just me thats tired of it.

The other way to dissapate energy is to overlap the clutch engagements. That is to start engaging the next gear WHILE the previous is still engaged.

The next gear clutch slips
Then you increase tension on the next gear and reduce it on the previous gear clutch. both clutchs are slipping but the car is still accelerating.

Eventually the next gear clutch if fully engaged and the previously one disengaged. Basically you are slipping the clutches to dissipate the energy. Lots of wear and tear on the clutchs but smooth aceleration. From what I have read this is how the veryon does it. Not something I would want in a car that gets driven a lot. You are also wasting the stored energy in the flywheel and turning it into heat in the clutchs instead of acceleration.

One thing to remember. Audi is 4wd which would put extra stress on the clutches if they did it the BMW way since there would be less give or slip in the tires in case of overload.

[ersin]

Originally Posted by footie

I don't want to go into the old debate of the amount of modes on offered, been there, done that and it's a point that some of us will not see eye to eye on.

My only concern is that I hope BMW offer equally as good an accelerate with modes which don't disconnect the DSC and demand the jerk/surge. My guess is this was included to appeal to the current SMG customers. My understand on the technology is that each of the modes should be capable of shifting equally as quick as each other, the only difference I see between S1 and S6 would be this build-up of free energy. What difference this will make to the acceleration I don't know but until I sample it I don't know how it will feel.

Who knows, I might even like it.

The objective of dual clutch systems is to change gear without any noticeable interruption in power and forward motion, noticeable interruption to me means smooth transition between changes. That might be Audi's interpretation of this, BMW might have an altogether different opinion of that. I wish I knew someone who has driven the new GTR or EVO to get an opinion on their gear changes to see it each manufacturer has a different take on what Dual Clutch should behave like.

Steved or anyone?

I'm sorry, but I still don't get it. I thought the whole point of offering the lesser modes is for gentler driving -- less acceleration, more smooth. Who drives their car at maximum acceleration all the time? Lower settings give the engine time to slow its rpm's down to match the new selected gear so that when the clutch engages everything is synced up. Higher, more aggressive settings engage the clutch while the engine is still spinning faster. If the lower settings gave the same shift speed then it would make no sense to offer them.

[footie]

Originally Posted by ersin

I'm sorry, but I still don't get it. I thought the whole point of offering the lesser modes is for gentler driving -- less acceleration, more smooth. Who drives their car at maximum acceleration all the time? Lower settings give the engine time to slow its rpm's down to match the new selected gear so that when the clutch engages everything is synced up. Higher, more aggressive settings engage the clutch while the engine is still spinning faster. If the lower settings gave the same shift speed then it would make no sense to offer them.

That has always been my point all along.

[ersin]

Ok footie. I think we're on the same page. My understanding is that the higher modes do shift faster and in the fastest modes it fastest modes when the clutch engages the engines speed is faster than the gear selected and so therefore the energy that would have been wasted waiting for the engine to spin down is now converted into the jerk/surge. The slower settings then take a bit longer so that you don't feel the jerk/surge.

If that is the case, and we could be wrong on that, but if it is then it does make sense to offer the lower modes if you want a gentler ride. Same philosophy as EDC I guess.

[VictorH]

Wow! Quite a few opinions on the new tranny and none of us have driven it. I am not wild about the option of jerking between gears, by keeping full throttle during shifting (unless you like to drag race).

I assume S6 is going to have the most rapid shift times, which you would conclude would be the mode to use when at the track. However, if a gear change is going to get scratch from the rear tire then I guess I don't want to be changing gears in any corners. Nothing better to upset the balance of a car mid corner than a jerky shift (well oil in the corner works pretty good too). I need to drive and then I'll decide.

[swamp2]

Originally Posted by enigma

You clearly have no understanding of phyisics. It would be great if I could stop from 100mph in 0.as without a jerk also, but thats equally problematic.

Where do you think the energy of fast spinning engine goes? Options are

1: Quickly to the rear tires producing a jolt of extra acceleration
2: Gradually to the rear tires while reducing engine output to allow the engine to slow without extra acceleration.

#1 = S6
#2 = S1

And you have 4 other choices between them to suite the drivers taste. This is why people buy BMW and not the "one size fits all" audi solution. There will be a performance difference. Its unavoidable.

Footie, there is another effect here. This engine under WOT at redline produces about 250 ft lb. Shifting when getting on it drops the rpm to something like 6000 (of course this varies gear by gear and on gas pedal depression as well). Around that 6k rpm the engine is producing very close it its peak torque, 300 ft lb. Aren't you going to feel a jerk when you try to apply 20 more ft lb over a very short time interval? Unless you make a careful effort to hide it you are going to feel it. You can play tricks with DCT and all sort of parameters as I have mentioned many times in the past like spark, timing, fuel flow, automated throttle control, clutch times and clutch phasing, etc. to control the jerkiness. As long as the jerk is not fake, which I highly doubt it is, I suspect that even with DCT more performance requires more jerk. But in all cases the jerk is less than SMG because there is almost no deceleration. My innacuracy in the past may have been believing that BMW could provide the most performance with little to no jerk. I do think I was wrong about the subtlety myself.

[footie]

Originally Posted by enigma

Car companies quote many things. I seriously don't know how to have a debate with you when you flat out deny what most everyone else on this thread realizes. Seriously, its not just me thats tired of it.

I don't think, I know that the jerk has been engineered. I know for a fact that DSG does exactly what is say it does and more. The 0.2s improvement is the minimum and remember it's only a 6 speed box and not a 7 speed like the DCT.

Originally Posted by enigma

The other way to dissapate energy is to overlap the clutch engagements. That is to start engaging the next gear WHILE the previous is still engaged.

The next gear clutch slips
Then you increase tension on the next gear and reduce it on the previous gear clutch. both clutchs are slipping but the car is still accelerating.

Eventually the next gear clutch if fully engaged and the previously one disengaged. Basically you are slipping the clutches to dissipate the energy. Lots of wear and tear on the clutchs but smooth aceleration. From what I have read this is how the veryon does it. Not something I would want in a car that gets driven a lot. You are also wasting the stored energy in the flywheel and turning it into heat in the clutchs instead of acceleration.

With the Veyron you are not talking about a piddly 400hp, the Veyron has close on 3 times that amount. If you reckon that the clutches will be damaged then I think you are misunderstand the technology.

Originally Posted by enigma

One thing to remember. Audi is 4wd which would put extra stress on the clutches if they did it the BMW way since there would be less give or slip in the tires in case of overload.

I won't disagree that awd does put extra strain on clutches, especially on standing starts. But I know SMG has given more problems because of their S6 mode than anything else, I ask you how many launch control starts will an M5/6 be able it perform after another.

[swamp2]

Originally Posted by VictorH

Wow! Quite a few opinions on the new tranny and none of us have driven it. I am not wild about the option of jerking between gears, by keeping full throttle during shifting (unless you like to drag race).

I assume S6 is going to have the most rapid shift times, which you would conclude would be the mode to use when at the track. However, if a gear change is going to get scratch from the rear tire then I guess I don't want to be changing gears in any corners. Nothing better to upset the balance of a car mid corner than a jerky shift (well oil in the corner works pretty good too). I need to drive and then I'll decide.

Analysis based on science/engineering as well as a close following of what other similar units do allows a great deal of (typically accurate) speculation. Driving is only one part of the experience. Understanding is another.

DCT "jerk" (where again jerk specifically means the time derivative of acceleration or the rate of change of acceleration) will almost surely depend on transmission mode and level of aggressiveness (primarily rpm/throttle position). Either way the jerk of DCT will be quite a bit less than the jerk of SMG or MT. There may be further enhancements to DCT to limit drivetrain shock during shifting while cornering, but that is pure speculation.

[hakaida442]

I watched the video of the DCT again and it touts uninterrupted lightning quick shifts and smoothness for more performance. Maybe it was wrong of me to assume that it would not be jerky like SMG.

[ersin]

Originally Posted by hakaida442

I watched the video of the DCT again and it touts uninterrupted lightning quick shifts and smoothness for more performance. Maybe it was wrong of me to assume that it would not be jerky like SMG.

I don't think you're wrong. DCT should be less jerky than SMG due to the fact that there is no/less let up of throttle between shifts.

SMG shifted fast but there was still a moment of time where the clutch was disengaged during new gear selection and acceleration practically stopped, then when the new gear was selected and the clutch re-engaged acceleration resumed. Hence the jerk.

DCT has the next gear pre-selected (not all but most of time I would guess) and therefore no waiting for the gear to change. Just change clutches.

The jerk we are talking about here comes not from the re-engagement of acceleration but the extra energy stored up as engine rotational kinetic energy (angular momentum). My thought is that you will feel less of a jerk but more like a surge. I guess that make us SMG drivers jerks compared to DCT drivers.

[swamp2]

Originally Posted by ersin

I don't think you're wrong. DCT should be less jerky than SMG due to the fact that there is no/less let up of throttle between shifts.

SMG shifted fast but there was still a moment of time where the clutch was disengaged during new gear selection and acceleration practically stopped, then when the new gear was selected and the clutch re-engaged acceleration resumed. Hence the jerk.

DCT has the next gear pre-selected (not all but most of time I would guess) and therefore no waiting for the gear to change. Just change clutches.

The jerk we are talking about here comes not from the re-engagement of acceleration but the extra energy stored up as engine rotational kinetic energy (angular momentum). My thought is that you will feel less of a jerk but more like a surge. I guess that make us SMG drivers jerks compared to DCT drivers.

Mostly correct here but when SMG shifts, just like a MT the car does not "practically stop accelerating", in fact it noticeably and measurably decelerates!

[ersin]

Originally Posted by swamp2

Mostly correct here but when SMG shifts, just like a MT the car does not "practically stop accelerating", in fact it noticeably and measurably decelerates!

Thank you, I stand corrected.

[rai]

when are the DCTs due out in the US??????

[enigma]

Originally Posted by footie

I don't think, I know that the jerk has been engineered.

Someone feel free to check the following math.

Assuming a flywheel of ~10kg, 0.15m radius

I = 0.5 * 10 * (0.15) ^ 2 = .1125

E = 0.5 * I * w^2 = 0.05625 * w^2

At 8400 RPM: E = 0.05625 * (8400/60*2*3.14)^2 = 43.48 KJ
At 6000 RPM: E = 0.05625 * (6000/60*2*3.14)^2 = 22.18 KJ

So after shifting we have ~21KJ of energy stored by the cars flywheel (not counting the other moving engine bits) that has to go somewhere.

If the car (1650kg) is going ~44mph in 1st (20m/s) the car currently has:
1650kg * 20^2 = 660KJ of energy.

If the DCT simply disengages the previous gear and engages the next and dumps that stored energy to the rear wheels you now have 660 + 21 = 681Kj

sqrt(681Kj / 1650) = 20.32 or a sudden velocity increase of 0.32m/s or almost 1 mph. Thats the jerk, its not simulated or made up, they just don't try to hide it under the cover of "smooth" like Audi.

You could apply the energy over time, burn it in the clutches, or reduce the engine output to cover it up. None of those help you go faster.

I suspect by the time you add in the other engine components the stored energy is considerably larger. Enough to put it into the 1-2mph range.

BTW: one "G" is 9.8m/s So if they accelerate the car 0.32m/s in 0.1s. that generates a spike of ~0.3g for that brief instant over the normal acceleration.