ET crash ADD NBO

[ian16th]

It is possible that Boeing could go bankrupt over this.

The US Government wouldn't let it happen, any more than RR in 1971.

The defence of the Realm, or the Republic in the case of Boeing, is paramount.

[Fox3WheresMyBanana]

Well, I don't think it's likely, but if the rest of the world stops ordering them, and cancels their existing orders, I'm not sure there's much the US Government can do about that any more. That's not to say they won't try, of course.

[Boac]

I still think the best solution is a comparator left to right which isolates MCAS (??and stick shaker??) when they differ by a determined amount , plus a big 'We're gonna die, Frank' light to announce.

[boing]

BOAC,

I imagine you flew the VC10 at some time. Remember the stick-pusher system. Two lights, two sensors, two test switches, two separate systems. And on the ground test you must only get a push if both test switches were moved at the same time simulating a genuine high AOA event. For God's sake, I was doing this in 1974.

Probably the AOA vanes have barely changed since then.


.

[Wodrick]

Probably the AOA vanes have barely changed since then.

A little, in the era of which you speak AOA sensors were an analogue function and a very simple sensor. The more modern sensor has an A/D converter so the output is digital. The normal reason for change is physical damage, usually a bent shaft although sometimes internal binding. There was a simple, pretty straightforward go-nogo test for freedom of movement where you balanced a coin on the blade and if it didn't rotate it was rejected. I figure this wouldn't work on the latest blade type sensors there must be something else.
I have no experience of the later 737 series (7, 8, 9) but AOA on an Airbus goes to multiple systems and a failure is flagged by them all at the end of flight, sometimes an ECAM if it was summat serious like Stall, not that Airbusses can stall

[OFSO]

Design philosophy: I've just designed and build a remote switcher for my central heating. Requirement (2) was "default to off", i.e. under all fault modes remote system must never leave the heating running continuously but revert to "off". Requirement (3) was on-the-spot manual override under all conditions. Requirement (4) was LEDs to show what system is doing. Requirement (5) was clear labelling of lights and switches

[Boac]

I imagine you flew the VC10 at some time.

- no, Sir. 737 only in airlines. My only VC10 was two out and backs to Belize for Harrier work.

[barkingmad]

Price of safety??

Not a lot that's free from any manufacturer!

IIRC all the NGs were capable of accepting a 15kt tailwind, but the piece of Boeing vellum certifying your particular registration cost approx $15-$20k.

When I raised the issue of the Bae146 simulator control column roll forces versus the real aircraft, the bill for a minor tweak and certification of the simulator was quoted as £20+k.

It's bad enough when "first world" airlines baulk at such costs, so what happens with lo-cos and "third-world airlines?

Incidentally, the Lionair crew were reported as desperately trying to consult the manual to help them with their problem (despite trim runaway being a QRH memory item) so I wonder was it one of those awful EFB laptop widgets they were trying to use or was it a hard copy manual?

In my last years on the beast when a non-urgent tech query arose on the flight deck I usually had the answer out of the Bulfer book long before the Magenta child had extracted, warmed up and interrogated the company laptop.

End of rant, now to find a darkened room..........

[Fox3WheresMyBanana]

I believe part of the problem was that they didn't have trim runaway, their problems were definitely intermittent. Furthermore, the trim still clearly worked - when the trim is activated the MCAS cuts out - until the trim stops being used.

[Ex-Ascot]

They seem to have found a fix but a survey (do not know the poll number) says 70% of people will not fly on it.

https://simpleflying.com/american-airli ... -fix-test/

[Slasher]

I believe part of the problem was that they didn't have trim runaway, their problems were definitely intermittent. Furthermore, the trim still clearly worked - when the trim is activated the MCAS cuts out - until the trim stops being used.

Thing is Fox whenever the stabiliser is doing dumb and dangerous things you hit the cutout switches first and work out what's going on later. You don't really need a published drill to tell you that, as it's an instinctive survival thing.

[Boac]

I am still very confused by the lack of 'previous' incidents involving MCAS. Ex-A in post#242 linked a Politico article about 'unexplained' malfunctions involving the autopilot, but not one case have I seen where the stick shaker is on permanently with IAS/ALT disagree and when flaps up/autopilot off there is a persistent attempt to trim nose down at intervals. NOT ONE! So, why the 3, two of which were fatal?

The whole thing is, to me, beginning to point to a software problem rather than hardware (AoA vanes). The consistent 22 deg AoA error shown on the FDR with both LionAir was still present AFTER the vane change, as, I believe, was the stick shaker and 'disagrees'. The other strange thing is the 'side' of the failure which, again I believe, has always been P1.

It now appears that the simulator for the 737Max cannot reproduce MCAS activity - WHAT?!!

[barkingmad]

https://www.nytimes.com/2019/03/23/busi ... crash.html

Apologies if this has been included already.

[Boac]

BM - it does help to explain the 'philosophy' in Boeing but it does not address the disparity in 'AoA' events.

It itself is incorrect also with
"To offset that possibility, Boeing added the new software in the Max, known as MCAS, which would automatically push the nose down if it sensed the plane pointing up at a dangerous angle. The goal was to avoid a stall. "

I do wish 'influential' journos and media would stop peddling this incorrect information, which is confusing many who read them. That is not the function of MCAS. It is not designed to 'push the nose down' - that would be a stick pusher - it is 'designed' to modify the stick force-alpha relationship at high AoA to satisfy regulatory requirements, as it was declared to be too easy to pull the nose-higher near the stall in a Max than previous models. MCAS would not prevent a pilot pulling through to a stall as very air-test pilot can demonstrate. A different issue.

[Fox3WheresMyBanana]

as it's an instinctive survival thing.

..or not

[G-CPTN]

How do you test a fix on an aircraft that hasn't experienced the fault?

[Boac]

You need to simulate a high AoA signal from the controlling vane.

[Mrs Ex-Ascot]

A question for the experts; do simulators have programmes that simulate sensor / instrument failures that could be potential gotchas?

[Fox3WheresMyBanana]

Simulators simulate anything that was written into their specs. For example, all flight instruments can usually be failed both individually and as part of a larger failure (e.g. generator failure).
Given that Boeing was, to my mind, clearly trying to hide the function of MCAS from both its customers and the FAA, it's highly unlikely that the 737 sims were programmed/rewired to simulate any MCAS failure, if indeed they were fitted with MCAS simulation in the first place.
Simulators are rarely able to directly simulate anything which is outside a 'failure of large component x'. So, realistic scenarios such as 'fire in forward luggage compartment' normally means the sim instructors manually working out which systems are adjacent to that compartment, then failing those using individual button presses. Similarly, with small components that in the real aircraft have been found to fail, such as an AoA gauge, the sim instructors have to work out the affected systems and fail those manually. It is rarely possible to replicate the exact knock-on consequences, so the failure may not be given to the crews because the closest simulation possible may not be realistic, or lead crews to the wrong conclusions. In a few cases, it may be possible for the sim technicians to pull a plug or two 'round the back' to give a failure that the operator panel cannot, but usually one has to ask the sim designers for a mod, and this is prohibitively expensive.
That is, of course, assuming the sim instructors have both the systems knowledge AND ARE PERMITTED to deliver a 'realistic' sim experience. I suspect many places deliver the absolute minimum sim experience required on cost grounds.
I was a standards officer on the Tornado F3 sim, and did some work on attempting to simulate battle damage.

[Boac]

Given that Boeing was, to my mind, clearly trying to hide the function of MCAS from both its customers and the FAA, it's highly unlikely that the 737 sims were programmed/rewired to simulate any MCAS failure, if indeed they were fitted with MCAS simulation in the first place.

It now appears that the simulator for the 737Max cannot reproduce MCAS activity - WHAT?!!

As Fox said, it would need a major update of the Simulator software to 'include' the MCAS function AND
A separate means to inject a specific 'reading' - high or low - into either simulated AoA input. That would cost money and.....................

Fox's tales of the Tornado sim remind me of the fixed base Hunter 'sim' at Chivenor in which the final 'wartime' exercise was flown, supposedly leading to an ejection, and the way some instructors got you 'out' was to light oily rags beneath the sim to 'simulate' a fire.