CarTest results and comparisons vs. MB AMG

[swamp2]

Originally Posted by Garrett

Yes, an adept human can shift a gear in about 200ms ~ 700ms. The latest version of SMGIII could shift in about 80ms~90ms.

DCT are said to shift in about 10ms..!!





-Garrett

We had some debate here a while back about the real shift times. I think the BMW quoted times for SMG are literally the "gear shift" times, proabably not the sum of clutch in+shift+clutch out.

Link:
The whole thread is a bit long but some good debate and opinions.

However with a DCT the shift has already been pre-selected so there is only a single simultaneous one clutch in and the other out = very fast. All that being said not sure I am confident of .008s (8 ms) time. IN reality maybe a bit slower, but compared to a human or SMG still no contest.

[Garrett]

Originally Posted by swamp2

We had some debate here a while back about the real shift times. I think the BMW quoted times for SMG are literally the "gear shift" times, proabably not the sum of clutch in+shift+clutch out.

Link:
The whole thread is a bit long but some good debate and opinions.

However with a DCT the shift has already been pre-selected so there is only a single simultaneous one clutch in and the other out = very fast. All that being said not sure I am confident of .008s (8 ms) time. IN reality maybe a bit slower, but compared to a human or SMG still no contest.

I can't remember where I've seen it, but BMW put out a graph illustrating power to the wheels shift time and it was (about) 80ms. I'll try and see if I can find it!

[southlight]

Originally Posted by swamp2

We had some debate here a while back about the real shift times. I think the BMW quoted times for SMG are literally the "gear shift" times, proabably not the sum of clutch in+shift+clutch out.

Link:
The whole thread is a bit long but some good debate and opinions.

However with a DCT the shift has already been pre-selected so there is only a single simultaneous one clutch in and the other out = very fast. All that being said not sure I am confident of .008s (8 ms) time. IN reality maybe a bit slower, but compared to a human or SMG still no contest.

As I said in that thread some time ago, that's indeed the clutch in+shift+clutch out time IMO. That's what I wanted to show with the graph. The SMG III needs almost no time for the gear shift itself.

Best regards, south

[swamp2]

Originally Posted by southlight

As I said in that thread some time ago, that's indeed the clutch in+shift+clutch out time IMO. That's what I wanted to show with the graph. The SMG III needs almost no time for the gear shift itself.

Best regards, south

I remeber the graph. The shift times quoted by BMW of 65 or 80 ms or whatever (clearly < 100 ms) are entirely inconsistent with the results obtained in this article. These tests show total shift times of >= 1/4 sec. That is awfully different from < 100ms, 2.5 times as long. Who is telling the truth and who is lying?

BTW, if I remember correctly you never finished this small "debate" back then. I claimed the article seemed to use fairly scientific techniques and even reported a margin of error. You never came back with any solid proof to validate BMWs claims.

[southlight]

Originally Posted by swamp2

I remeber the graph. The shift times quoted by BMW of 65 or 80 ms or whatever (clearly < 100 ms) are entirely inconsistent with the results obtained in this article. These tests show total shift times of >= 1/4 sec. That is awfully different from < 100ms, 2.5 times as long. Who is telling the truth and who is lying?

BTW, if I remember correctly you never finished this small "debate" back then. I claimed the article seemed to use fairly scientific techniques and even reported a margin of error. You never came back with any solid proof to validate BMWs claims.

You're right I was lazy about that. But to be honest, I don't have any proof on this topic. That's why I said IMO: All we have is BMW's statement and the graph I posted on one side and the test results for SMGII on the other side. To me it's not that interesting now where the new M-DCT is knocking on the door.

EDIT: OK, it seems like you're right and I was wrong about that. The 80ms and 65ms respectively are indeed for the gear shift time only! Some did also measure a time of about 250-300ms on the M3 E46 for a S6 shift. The graph that was shown in the german magazine did fool me. Now you know why I never came back to the debate.

Best regards, south

[esquire]

Originally Posted by esquire

well not exactly, your 4.43 figure is for the M-DCT, whereas car and driver pulled 4.4 with the 6mt. in your analysis, the 6mt hits 60 mph in 4.68 seconds - a significant .28 second departure.

swamp, i'm quoting myself here because i'm not sure if you even saw this post (which btw is the last post on the first page). it's safe to say that this could only be the result of a discrepancy between the computer's performance model and the real thing. given that the car as tested by car and driver was a 6mt, it's safe to assume that the flaw in your simulation wasn't a shift time flaw. my intuitive thought on it, is perhaps bmw is even underrating the hp on the M3.

what are your thoughts on this discrepancy?

- esquire

[swamp2]

Originally Posted by esquire

swamp, i'm quoting myself here because i'm not sure if you even saw this post (which btw is the last post on the first page). it's safe to say that this could only be the result of a discrepancy between the computer's performance model and the real thing. given that the car as tested by car and driver was a 6mt, it's safe to assume that the flaw in your simulation wasn't a shift time flaw. my intuitive thought on it, is perhaps bmw is even underrating the hp on the M3.

what are your thoughts on this discrepancy?

- esquire

The 4.68 was for 0-100 km/h, not to 60 mph. I posted the metric figures for south who asked about what gear the M-DCT was in at 200 km/h.

Even if the 6MT comes in at 4.3, 4.47 is within 2%. Did you see south's post on tire size discrepancies? He found a 3% error right there in all CarTest sims. But the existence of that error is probably only accurate at 70 degrees F, with tires at 33 psi and at sea level (kind of joking here, but again, not really). I could tweak some of the parameters further but when almost all the numbers come in within 5% or so I am pleased.

It all depends on the simulation and discipline. Some folks like in some areas of physics are happy with a prediction within a factor of 10. In acoustics 3dB is pretty good (a factor of 2!). For this type of thing I am very pleased with 5%. Sure wish I could get in the ball park on the 335i though... Cheers.