johnh wrote:
Boeing troll. I’m surprised they keep bothering.
What? How?
but ask yourself why didn’t Ryanair crash the 747 MAX too
Undoubtedly because it hasn’t been built. Probably the reason they haven’t lost any 737 MAX either is that they never had an AOA sensor failure. Also iirc they didn’t have many (any?) in service prior to the crashes and grounding.
There was a lot of w***y-waving back at the time by (mainly American) pilots saying essentially “well I would have caught it and flown out of it”. I’m surprised that’s still going on after for example Sully said it was an unlikely thing to survive.
The MCAS software was MUCH more pernicious than simple trim runaway, because of its persistence and intermittent nature.
It was crap software indeed (and with my 45 years of writing code I would say that every time) but it came into operation only if you brought the plane near to stall, and Western pilots tend to not do that. It’s a dumb thing to do – regardless of any protections.
Apparently some 3rd World Airbus pilots do hold the stick all the way back because the autopilot is very good at holding the plane right on the stall margin, all the way from the runway to the operating ceiling. So an Airbus will outperform a Boeing in a flat-out climb (all else being equal).
The principle I am getting at is that MCAS was a cockup but a cockup which was easily enough exposed once you deviated from any half decent SOPs. We just don’t know how many other software cockups there still are in “FBW” which remain undiscovered because pilots do not exercise that flight regime.
As you know from your software experience John (at least as long as mine) you can have loads of bugs which nobody finds purely because they don’t exercise that part of the code. Which, ahem, is just as well, since testing 100% of any code is impossible.
The Airbus stall warner suppression < 60kt was known to the programmers but was not known to pilots until AF447, and then everybody wondered why do something so stupid. But also AF447 showed – to anyone who didn’t already know it – that airline pilots are not particularly bright (a phrase from a friend with 40 years of long haul big jet flying, btw) and certainly are generally unfamiliar with aircraft systems in detail.
I have no love of Boeing but I think they got too much mud, because a) they handled it badly PR-wise and b) they are American and it is fashionable (especially in the 3rd World) to hate America.
only if you brought the plane near to stall
Well no, not at all. That’s the point. It’s “only if the MCAS THINKS you have brought the plane near to stall”, for example because the one and only AoA sensor it relies on has gone TU. And then (in the Mk 1 version) it would fight you all the way to the ground. You’re at low altitude, with an aircraft that KEEPS pitching itself down into the ground. And nobody has ever told you WHY it is doing this. If you have the presence of mind you can disable the electric trim, but then (owing to a different and much older design c***-up) you can’t move the trim wheel anyway.
If you KNEW that this would stop happening with flaps 10, you could do that. But you don’t know, because Boeing decided not to tell anyone so they could make more money out of SWA.
Of course testing software is impossible. But that’s why (so us ordinary types are told) aerospace software is written separately by three teams who don’t communicate, with majority voting on what to do. Except that in the case of MCAS, it wasn’t.
Boeing is the bad guy here, no question (imo obviously).
As for AF447 – the PF was utterly, totally, beyond all redemption incompetent. He wasn’t safe to fly a Cessna 152, never mind an A330. That’s all there is to it. Sure, the ‘bus is a complex beast, with dozens of different “laws” that all do something different. But if you don’t know not to hold the stick right back for any reason at all (except acro or practising stalls), you should not be flying any aircraft at all.
The AF447 book spends a lot of time explaining the ’bus, but it never ever suggests that the real cause was anything other than the total incompetence of the pilot.
Yes you are right. Pilots are not particularly bright outside of their chosen profession.
Yes the single AoA sensor was dumb, but they can fail too, both / all together. What I find curious is how Boeing thought this was ok. They are pretty competent technologically, with all sorts of military/space projects.
I bet you the Airbus “three teams” thing is vulnerable to all being done to the same spec. Anyway, programmers tend to approach a problem the same way (especially experienced ones, and especially if working to standard robustness guidelines e.g. all storage static, no locals, no heap, etc) so will often produce the same code. I mean, look, nobody wondered about the <60kt stall warner suppression 
If anybody at Airbus had flown a C150, that would not have been done.
Which AF447 book?
In GA, you could have a runaway via various single point failures, too. The saviour is the limited speed range of the plane, and having a Disconnect switch which most pilots know about. Certification calls for a TRIM IN MOTION annunciation after IIRC 9 seconds and a trim servo disconnect after IIRC 15 seconds of continuous trim servo motion. This can be tested on the ground by engaging with AP button and pushing or pulling the yoke. But most pilots don’t know that.
Antonio wrote:
but have you ever tried to move manually or electrically the pitch trim in a Mooney while flying severely out of trim?
Yes. Initially this happened to me a lot. Primarily because I did not have electric trim then. In any go around in the initial training days trimming by hand was quite an exercise. But perfectly feasible. The trim in the Mooney moves the whole empennage.
The scariest one I’ve ever had was many many years ago, again in a go around, with my Trusty C150. That thing had flaps 40. If you put on full power with flaps 40 and landing trim, it will take the nose up like heck but lacks the power to climb at this configuration…
Which AF447 book?
https://www.amazon.com/Understanding-Air-France-Bill-Palmer/dp/0989785726
After reading that and the Concorde book no sensible person would ever fly AF again. I do (just did a round trip to Singapore with them), which tells you as much about me as about AF. But at least I know the risks I’m taking.
[ link fixed ]
But if you don’t know not to hold the stick right back for any reason at all (except acro or practising stalls), you should not be flying any aircraft at all.
Palmer’s book is an excellent, sober analysis which also provides a balanced view on the crew. As the book explains in normal law the memory item for stall recovery is stick right back, in addition the FD had a pitch up. I do recommend reading the book before jumping to conclusions on the crew.
Famously Airbus picked up the ‘phone to Boeing to agree there were serious shortcomings in flight deck crew training, in particular in analysing and parsing the symptoms of the stall. Hence the regulatory requirement now for A-UPRT before a turbine type rating.
There are still some serious shortcomings in training due to some jurisdictions practical test standards on straight wing propeller training aircraft, but hopefully these will eventually be phased out.
