jz11 wrote: would be nice to see torque-after-gearbox measurement graph instead of the acceleration, I suspect there should be a small drop and a short spike during the shift, and there is no indication of that on the acceleration graph
Agreed; in effect the mass of the vehicle is acting like a damper (good old F=m.A!), and so the acceleration of the vehicle doesn't really "see" the torque spike in the transmission. As long as the torque spike isn't big enough to damage the gearbox or upset the car (apparently it isn't), who cares? As racers all we care about is accelerating quickly!
Remember that quote I posted earlier: the torque spike from the speed differential in the seamless box is actually smaller than the difference in torque that you get from having a break in the transmitted torque that you get from a conventional dog box gear shift, which is why it is smoother. If it wasn't 5 minutes past midnight I'd do a quick calculation to check that claim. But it's late, so, maybe tomorrow!
EDIT: Some rough numbers based on a 2.4 litre V8 (i.e. when the seamless shift was introduced, before KERS could be used):-
Non-Seamless gearbox, Torque step = from 0 (no torque delivered), to maximum Torque output of engine: Approx 360Nm, therefore max torque step applied to drivetrain =
360Nm.
Seamless gearbox. Rev drop required = ~3000rpm (315rads/sec), Engine inertia: ~0.05kgm^2, time for rev drop: ~100 milliseconds. Applying T=I.α; Torque required = 160Nm. Arresting Torque due to Friction/pumping, approx. 15% of max engine output= 54Nm. Max Torque step applied to drivetrain: 160-54 =
106Nm
Only very rough numbers, but certainly backs up the claims made by the manufacturers, drivers and independent magazine testers that a seamless shift is smoother than a conventional dog box "racing change".