991 GT3 overspeed events
10-01-2014, 01:30 PM
#46
Rennlist Member
I think what he was saying is that when in auto-mode (different scenario than previously discussed), the car runs right to redline, then shifts in under 100 ms, and in that under 100ms time, how much does the motor over-rev from internal engine momentum, while disengaged from a gear, and if so, how much? At least that was my understanding of what he was saying........ So it would be a question of does the mass of the pistons, etc continue to carry rpm higher, or does the high compression immediately slow the internal engine speed down.
Yes, a different scenario than when gravity, acceleration, and the car's momentum continues on, past when the electronic brain shuts off fuel/spark, and the tranny is still engaged (because in manual mode PDK-S holds it engaged).
Sorry for the confusion...but that was how I read his comment.
Yes, a different scenario than when gravity, acceleration, and the car's momentum continues on, past when the electronic brain shuts off fuel/spark, and the tranny is still engaged (because in manual mode PDK-S holds it engaged).
Sorry for the confusion...but that was how I read his comment.
10-01-2014, 01:44 PM
#47
Would clutch would just reengage and not over rev as there is no longer any acceleration to go over 9k when going into paddle natural.
10-01-2014, 01:47 PM
#48
So the motor's internal "inertia" won't carry it over the Redline- instead the question as we approach the RPM cutoff is "how fast can the electronics cut the air, fuel or spark off?". The answer, particularly with the last one, is going to be "pretty damn fast", especially as you can anticipate that you're going to hit redline and begin to take action before you even get there. Thus I'd expect to see zero over-revs due to the engine itself assuming the engine software guy still has his job.
The other option is mechanically pushing the engine faster than it wants to go, and this gets much trickier.
Things like drive shafts and even tires are really springs, and they get "wound up" by the high torque under acceleration, even if only by a tiny amount. Of course they get wound up the wrong way: the moment you let off the throttle they try to slow the engine, not accelerate it. But whenever you get springs and dynamics involved strange things can happen- resonant bouncing, etc. But by itself this should still be relatively easy to measure and predict.
Finally we get to interactions with the car and tires, and the real opportunity for over-revs. Imagine that you're decelerating over a wet crest. The wheels are lightly loaded and the pavement is slick, so you're getting some longitudinal slip, maybe 5%. The wheels are going slow enough that the car lets you grab a lower gear, and just as you do you hit a dry bump in the pavement. The wheels load up, the wheel slip percentage is nearly instantly reduced, and suddenly the car "pushes" the motor some of that 5% faster, causing an over-rev.
Is the car smart enough to know that this might happen and prevent it? Perhaps it is, and they've programmed for this particular situation, but can they programmed for every possible situation? Do you push in the clutch? If so, how quickly? It's mechanical and therefore not instant, so the over-rev will depend on the tires grip level, etc... I don't know, but my bet is that if people see over-revs this type of thing is where they will come from.
10-01-2014, 02:18 PM
#49
Interesting analysis Petevb, very impressive. Just to mention something I'm sure you already know, that this 991 GT3 engine uses rocker arms with a damper as this engine was designed with high RPM in mind. The reason they have this rocker arm set up with a damper is to prevent the "float" from the cam you speak of at high RPM.
Perhaps, though from the sound of it the decision to go PDK was made late in the game so it's not clear how optimized the motor is for it. In theory, however, they should eventually be able to make a slightly better, lighter motor for the PDK they could for the manual, which must appeal... despite which I still want a manual.
10-01-2014, 03:35 PM
#51
Rennlist Member
"An object in motion stays in motion" according to Sir Isaac. This means that our pistons are always going to keep going exactly the same speed unless something acts on them. Friction slows them down rapidly- that's the strong engine braking. Meanwhile the only things we have to speed them up are burning fuel / air or mechanically pushing them faster.
So the motor's internal "inertia" won't carry it over the Redline- instead the question as we approach the RPM cutoff is "how fast can the electronics cut the air, fuel or spark off?". The answer, particularly with the last one, is going to be "pretty damn fast", especially as you can anticipate that you're going to hit redline and begin to take action before you even get there. Thus I'd expect to see zero over-revs due to the engine itself assuming the engine software guy still has his job.
The other option is mechanically pushing the engine faster than it wants to go, and this gets much trickier.
Things like drive shafts and even tires are really springs, and they get "wound up" by the high torque under acceleration, even if only by a tiny amount. Of course they get wound up the wrong way: the moment you let off the throttle they try to slow the engine, not accelerate it. But whenever you get springs and dynamics involved strange things can happen- resonant bouncing, etc. But by itself this should still be relatively easy to measure and predict.
Finally we get to interactions with the car and tires, and the real opportunity for over-revs. Imagine that you're decelerating over a wet crest. The wheels are lightly loaded and the pavement is slick, so you're getting some longitudinal slip, maybe 5%. The wheels are going slow enough that the car lets you grab a lower gear, and just as you do you hit a dry bump in the pavement. The wheels load up, the wheel slip percentage is nearly instantly reduced, and suddenly the car "pushes" the motor some of that 5% faster, causing an over-rev.
Is the car smart enough to know that this might happen and prevent it? Perhaps it is, and they've programmed for this particular situation, but can they programmed for every possible situation? Do you push in the clutch? If so, how quickly? It's mechanical and therefore not instant, so the over-rev will depend on the tires grip level, etc... I don't know, but my bet is that if people see over-revs this type of thing is where they will come from.
So the motor's internal "inertia" won't carry it over the Redline- instead the question as we approach the RPM cutoff is "how fast can the electronics cut the air, fuel or spark off?". The answer, particularly with the last one, is going to be "pretty damn fast", especially as you can anticipate that you're going to hit redline and begin to take action before you even get there. Thus I'd expect to see zero over-revs due to the engine itself assuming the engine software guy still has his job.
The other option is mechanically pushing the engine faster than it wants to go, and this gets much trickier.
Things like drive shafts and even tires are really springs, and they get "wound up" by the high torque under acceleration, even if only by a tiny amount. Of course they get wound up the wrong way: the moment you let off the throttle they try to slow the engine, not accelerate it. But whenever you get springs and dynamics involved strange things can happen- resonant bouncing, etc. But by itself this should still be relatively easy to measure and predict.
Finally we get to interactions with the car and tires, and the real opportunity for over-revs. Imagine that you're decelerating over a wet crest. The wheels are lightly loaded and the pavement is slick, so you're getting some longitudinal slip, maybe 5%. The wheels are going slow enough that the car lets you grab a lower gear, and just as you do you hit a dry bump in the pavement. The wheels load up, the wheel slip percentage is nearly instantly reduced, and suddenly the car "pushes" the motor some of that 5% faster, causing an over-rev.
Is the car smart enough to know that this might happen and prevent it? Perhaps it is, and they've programmed for this particular situation, but can they programmed for every possible situation? Do you push in the clutch? If so, how quickly? It's mechanical and therefore not instant, so the over-rev will depend on the tires grip level, etc... I don't know, but my bet is that if people see over-revs this type of thing is where they will come from.
I think 2 feet in is now foot and 2 flappies in, but yet I suspect Hall is going to know what's going wrong and neutral out the box for you.
10-01-2014, 05:34 PM
#52
Admin
Lifetime Rennlist
Member
Lifetime Rennlist
Member
Thread Starter
Im my opinion the OP document does not smell at all like an official PAG type recommendation. This must be something locally produce for the NA market (I stand to be corrected)?
The language used is alarmist to say the least the the rev ranges remarkably low. There is no way they built the engine to within 4% over rev failure tolerances. My old 993 has no top end work and has over revved beyond spec on a few occasions (my fault) without damage (at least to springs and valves when inspected).
Apart from the fact you can not over rev this engine it was tested to 10,000 rpm by the factory and im 99% sure that's recorded in an interview with AP or documentation direct from the factory - I will go away and have a look.
If the OP wasn't of such resolute standing on this board Id say this was "trolling" ;-P
The language used is alarmist to say the least the the rev ranges remarkably low. There is no way they built the engine to within 4% over rev failure tolerances. My old 993 has no top end work and has over revved beyond spec on a few occasions (my fault) without damage (at least to springs and valves when inspected).
Apart from the fact you can not over rev this engine it was tested to 10,000 rpm by the factory and im 99% sure that's recorded in an interview with AP or documentation direct from the factory - I will go away and have a look.
If the OP wasn't of such resolute standing on this board Id say this was "trolling" ;-P
I guess that German chauvinism prevents them from translating into proper English.
All kidding aside, I'm located in Europe and that's an official "RoW" document without geographic limitations and exemptions.
Technical Information Service 0
12/13 EN 0335
Vehicle Analysis Log: Evaluating overspeed events (12/13)
Anyone with access to the official PIWIS TSI can verify it.
Best
John
10-01-2014, 08:07 PM
#53
Race Director
10-01-2014, 08:16 PM
#54
RL Community Team
Rennlist Member
Rennlist Member
Doesn't the OEM software for the PDK automatically trigger an up shift when the engine redlines?
10-01-2014, 08:23 PM
#55
Rennlist Member
I've yet to spin the GT3 while cone bashing, but I spun my PDK Carrera S a few times. PDK monitors a whole bunch of vehicle functions including wheel speed, road resistance, and axial and lateral acceleration. I never stalled with PDK in my Carrera and there was no paddle neutral feature in that car; the gearbox disengaged the clutch on it's own. I'm certain PDK-S works the same way.
10-01-2014, 08:51 PM
#57
Race Director
I can (sheepishly) confirm that PDK-S disengaged the clutch and wont stall at least in Sport Auto mode as I put the car into a 180 degree spin in teh pouring rain at the track last week and was impressed with the way it handled it (disengaged, throttle back and when can to a halt changed me down into first gear). I spun my 993 too (twice) same day before quiting (appalling conditions teaming rain, flooded track, greasy rubber from prior event - we lost two cars that day total loss!). Doing all that manually while spinning (almost 360) isnt so much fun LOL!
10-01-2014, 08:55 PM
#58
Race Director
In both default auto and sport auto mode, yes. But in manual mode, with PDK-S in the GT3, there is no automatic upshift under any circumstance, so the engine will bounce off the rev limiter if you don't shift for yourself.
10-01-2014, 08:57 PM
#59
Rennlist Member
I guess that German chauvinism prevents them from translating into proper English.
All kidding aside, I'm located in Europe and that's an official "RoW" document without geographic limitations and exemptions.
Technical Information Service 0
12/13 EN 0335
Vehicle Analysis Log: Evaluating overspeed events (12/13)
Anyone with access to the official PIWIS TSI can verify it.
Best
John
All kidding aside, I'm located in Europe and that's an official "RoW" document without geographic limitations and exemptions.
Technical Information Service 0
12/13 EN 0335
Vehicle Analysis Log: Evaluating overspeed events (12/13)
Anyone with access to the official PIWIS TSI can verify it.
Best
John
Ive asked my guy what could possible cause an over rev in this car other than faulty ECU or aftermarket ECU flash...
10-01-2014, 10:40 PM
#60
Rennlist Member
Join Date: May 2012
Location: Mid-Atlantic (on land, not in the middle of the ocean)
Posts: 12,431
Received 3,782 Likes
on
2,190 Posts
So if you pull the paddles to put it in neutral, and coast in neutral for a while, will the PDK simply select a reasonable gear when you let the paddles go?