Originally Posted by
positiverate20
No idea how to retrieve a 20 minute post I'd written- where does the "Auto-save" go??
Anyway, quickly, and quite ironically, I was explaining that I've little to no "software coding" literacy nor any in-depth knowledge as to how it works.
My point was basically (acknowledging this is off-topic at this stage for flight in question, but in same tangent as previous MCAS discussion):
We know, with Lion Air anyway, MCAS caused problems.
We know MCAS failed- in that it wasn't designed to have the effects that it had.
With this in mind, how can we be so confident that MCAS adhered to:
- the pitch limitations
- the 10 seconds of operation
- the 5 seconds 'wait and see' period
- not functioning if flaps extended
etc. etc.
My point was- if the system failed- how, at this stage anyway, can anyone be certain that one component of the system failed entirely and that every other single component performed perfectly? Despite whatever failure mode it's in, does it still adhere to the pitch input rates, same operating window etc.?
I'm more comfortable with hardware, because I understand it, and know how it can fail- which is why it shocks me that the MCAS system is dependent on one single AoA vane's data- ludicrous. However, we can't at this stage determine the hardware was the only fault for Lion Air, are there other flaws or weaknesses hidden in the software, circuitry, functioning, logic of the MCAS elsewhere? Maybe MCAS was never an issue? Maybe errant data from a faulty ASI corrupted the MCAS or made it behave erratically?
I'm uncomfortable with the encroachment and infringement of automation by way of control inputs during what was considered to be manual flight. It frightens me that even with all AP functions off, computer software can still manipulate controls in this way.
Actually MCAS functioned as it was supposed to in the Lion Air crash, it received a high AOA signal and kept trimmed accordingly. The broken part was the AOA sensor. There's no need to be afraid of computers adding control inputs with the auto pilot off, the 320 does it all the time and has half the hill loss rate of the 737.