gums, Quote:
"Having flown the first "operational" full FBW system, I was a non-volunteer test pilot, as were all of us in those days ( 1979 and a few years thereafter)."
The first year of ops on the A320 (1988) was a bit like that, in a low-key way.
Quote:
"We clever pilots can find ways to "beat the system". We can do things that the engineers and others never envision ( see Perpignan, for example, or the first crash with a high-time pilot making a low pass for public relations)."
Yes, flew with one or two guys who aspired to be test pilots. (Why don't we switch off an ELAC and one/two SECs, and get some useful practice in Direct Law? Not today, thanks.) We managed to avoid a Habsheim.
There was a subtle "trap", though, involving the A/THR and FD. In descent in OPDES, the A/THR is in idle mode, and the selected IAS is controlled by elevator (i.e., conventional). If you come into the circuit and elect to go visual, you may elect to dispense with the AP & FD, but retain A/THR. If you level off a bit at (say) 3000ft when the FMGC thinks you still want descent in idle thrust to (say) 2000ft, the speed will naturally fall below target. In those days, the first intervention from the A/THR would be Alpha-Floor (TOGA thrust) as the stall protections kicked-in. Good way to mess up your visual approach... We soon learned to switch both FDs off: to tell the A/THR we wanted it to control the speed for us (and avoid conflicting FD pitch-commands). That had not been originally established as an SOP.
Word did not filter through to one Asian operator, however, and it lost an aeroplane on the approach at Bangalore. The crew had wound the ALT selector down to zero, I think, retaining OPDES, so the thrust needed at about 500ft never materialised. (Bit like the B777 at LHR; reason totally different.)
Post-Bangalore, one of gums's software changes ensured that the A/THR automatically comes out of idle mode if the speed falls below Vls.
Airbus FBW architecture is now a mature system, with tens of millions of flight hours behind it, so simple surprises like that are highly improbable in the case of AF447. But airline pilots do not generally explore the boundaries of the flight envelope, so the learning process is not as brutally quick as on the F-16.
Quote:
"So my personal observation is that this accident will be found to be in the category of "unplanned flight conditions", possibly complicated by one or two sensor failures.
I also see a few modifications coming for the 'bus control laws. It's digital, right? So not a lotta hardware needed, just some good flight tests with modified control laws/reversion sequences."
gums, you could be proved right.