|
Rob Palin has now written more than 5000 words of feedback on my book, chapter by chapter. That's a stunning commitment to an amateur like me, and one for which I am truly grateful. (No mistakes identified, but a lot of amplification.)
Unlike some some people on this discussion group (ie Aerohead) I don't pretend that I know more than the experts.
So for people who want to actually learn, here are some of his other recent comments on lift. (I have de-identified some car companies.)
The Audi TT [mentioned in the book for its scary rear lift results - they crashed] had a CLR that wasn't completely ridiculous - circa 0.11-0.13, I think. It did meet Audi's vehicle stability criteria at the time, but they got stricter afterward. XXXXXX applied the same standard, which was CLF>CLR up to 15 degrees of yaw, and the actual forces at Vmax need to be less than 10% of the axle weight. The TT became kind of a scare story within the aero community, and the name had incredible power within the VAG. Certainly, cars like the Mercedes CLS of that era were much worse (0.18-0.19 or so), but didn't gain any kind of notoriety.
My own view is that the reality is buried in the little-regarded fact that mean lift values leave out a hugely important part of the story, which is the actual frequency spectrum of the lift forces. In research I did back at MIRA, we saw easily 100% fluctuations in lift forces, on all manner of vehicles, with the associated spectrums varying wildly between vehicle shapes, and with individual modifications. A rear spoiler, for example, may have zero net effect on the mean lift, yet reduce the standard deviation of the forces away from that mean quite dramatically - and vice versa.
In the case of the TT, with its curved rump, I would imagine the fairly moderate mean value of CLR didn't reveal that the peak values could be >0.2, nor perhaps that such the underlying flow behaviours could be coherent, or prolonged, or at least of frequencies that perhaps interacted with some resonant modes of the suspension. Ordinarily, lift forces fluctuate at around 30Hz, which should be much higher than the suspension's natural frequencies, but maybe there was a higher order mode which overlapped. Frankly, I doubt anyone knows.
It's very difficult to examine such behaviour in the wind tunnel, due to conflicting demands of consistency of measurement and clarity of signal. Now that most CFD is done transiently, however, the awareness of the true nature of the forces is right in our faces every day, and there's more awareness of just how variable these forces are. Sadly, that also brings a growing realisation that the desire to run transient CFD for just 1-2 seconds of real time (to be practically feasible) is in terrible conflict with the reality of huge force variations as low as fractions of a Hertz. Even sampling for 10 seconds of real time is insufficient on a road car, yet we have to make do with ~3 seconds in CFD or we risk waiting many days for each result.
Stuff from a real, practicing expert.
Dr Wolf of Porsche was equalling fascinating. Note the lack of definiteness - something Dr Wolf was also happy to state on topics like lift. The real experts don't pretend everything was decided in the 1930s...
|