Boeing 737 MAX family NEWS ONLY

Status
Not open for further replies.
It's frightening how many things can break in a modern aeroplane and it still be deemed flightworthy. Laudable that there are enough fail-safes, fallbacks for the failsafes and so on and on that it is indeed pretty much safe enough - well, unless it all gets so complicated that bugs in the procedures get buried. One wonders whether a little more spares stock and time spent fixing things between flights might cut the need for such labyrinthine safety layers and actually cut parts counts, failure modes, periodic maintenance tasks and pilot training to such levels that the thing actually gets cheaper and safer to operate. The MAX's Master Minumim Equipment List (see previous link) being a case in point: deemed safe to fly with the autopilot broke, yet at least one emergency procedure requires that you engage autopilot.
 
steelpillow said:
The Verge being cynical about the MAX refresh a couple of weeks ago. Not sure whether I agree with their conclusion, but it makes for an informed read:
https://www.theverge.com/2020/4/9/21197162/boeing-737-max-software-hardware-computer-fcc-crash
Click to expand...

I once spent a few days working out how common a single bit flip on the 777 might be. Turned out to be significantly under 1x10^-9/flight hour, but that one was only safety critical on approach. It's possible to design around these, read values twice to make sure one wasn't flipped in transit and the like, but for bit flips due to cosmic rays_in_ the processor at a vulnerable point, you're probably never going to be able to fix every possible vulnerability. (Technically you could, by storing every bit multiple times, but that would be an entirely new computer architecture). And contrary to what the Verge article tries to say, there's a positive safety bonus to using elderly processors - we know exactly how reliable they are. The level of safety mandated for aerospace cannot be done using the latest hyper-speed processors off the line. It takes someone taking a specific design iteration of a processor and taking it off into the lab for a couple of years to generate a version that's hardened against cosmic rays and the like, and then taking the time to demonstrate that to the satisfaction of the safety authorities. By which time the state of the art has moved on several generations.

As for hardware versus software, the answer is neither is the issue. I've worked on systems in which parts of the control algorithms are built into the hardware, and others where it's entirely software. If you want to change an existing system, it's easier and safer to do it in software. What Verge should have been taking issue with is the systems design that meant hardware and software were singly redundant, even with two channels in the aircraft.
 
I've been using downtime to sit in the back garden and read through a bunch of old AvWeeks I never got around to reading for various reasons, and I found a Max-relevant article in the Feb 2-15 2015 issue.

"Rule out Rote" describes NASA Ames/San Jose University research in which they put 18 experienced 747 captains and first officers (with 4,000-20,000 hours) in the simulator and hit them with unexpected failures, but ones like stall, wind shear, and engine out on takeoff that they'll have repeatedly experienced in their normal simulator training (and which they'd replicated beforehand as a baseline). The sole difference here is they didn't know they were going to happen during the session. In all cases they took longer to react than normally anticipated, two of the pilots didn't go to full power to escape a low altitude stall, most of them made configuration changes to the windshear, making the loss of altitude worse, and two aborted the takeoff above the speed where they should have continued. Level of experience didn't seem to be significant.

So we've got research, reported in the leading trade journal, that if you surprise pilots they will not react immediately and follow the standard drills, yet at the same time Boeing was writing a safety justification for MCAS that pilots would be on top of it in 6s. And there've been unsavoury allegations that the only reason the Lion Air and Ethiopian pilots didn't save the aircraft is that they weren't American. Yet here's research from years earlier saying that even experienced American pilots don't react that quickly when surprised (and that's going to go double for situations with multiple possible explanations, as with MCAS).
 
DWG said:
And there've been unsavoury allegations ...
Click to expand...
Yes. Such allegations always have an unsavoury motive behind them. I don't think we should go there in this thread, suffice to remember that "every aircraft has four dimensions; length, breadth, height and politics." -- Sir Sydney Camm.

Bringing the Max market out of lockdown is going to be interesting.
 
As ever, those of us without WSJ subscriptions are left in the dark as to its relevance here. Does it even mention the Max?
 
Apologies, forgot to quote the relevant bit:
Boeing is halving output of its 787 twin-aisle jet and said Wednesday that it had resumed limited production of its 737 MAX jetliner. It halted MAX assembly in January as it awaited regulatory clearance for the plane to fly again following two fatal crashes, which it hopes to secure in the third quarter.


Industry and government officials have projected that following the restart of MAX production, it is likely to take the Federal Aviation Administration at least two months to issue detailed safety directives ending the global grounding. After that, according to these officials, it could take an additional month or to two to hammer out revised pilot-training procedures. The FAA’s effort to schedule simulator sessions with foreign pilots and regulators is still under way.


A number of suppliers to Boeing’s airplane factories near Seattle have already announced job cuts, according to state employment records.
Click to expand...
 
TomcatViP said:
https://theaircurrent.com/aviation-...gulators-plot-more-changes-after-jets-return/
Click to expand...

https://www.seattletimes.com/busine...-max-safety-upgrades-were-rejected-over-cost/
linked from the first article is well worth reading for a more in-depth discussion of both the synthetic airspeed concept and the internal politics. It's particularly interesting that Boeing are confident they can calculate airspeed from weight, coefficient of lift, angle of attack and aircraft configuration to a degree of reliability sufficient to overrule physical airspeed sensors.
 
Can't read the Seattle Times. It blocks my access unless I let its adverts through. Well, Seattle Times, I don't need you as much as you need readers.:p

But as the 787 appears to work its instrument indicators backwards by synthesizing what it wants them to display, and has not yet started falling out of the sky for that reason, I hesitate to dismiss such a horrific-sounding kludge out of hand. Only, I wonder what will happen when the synthetic and real AoA sensors decide to disagree?
 
Yes and it can be summed here:
But installing it in the MAX would likely have meant 737 pilots needed extra training in flight simulators. Running thousands of pilots through simulator sessions would have delayed the jet’s entry into service and added substantial costs for Boeing’s airline customers, damaging the MAX’s competitive edge against the rival Airbus A320neo.
Click to expand...

General aviation owners did not wait any approval to install GPS and moving map in their cockpit.

This industry is sick (over-regulated). If you are not convinced, transpose that to your mobile phone to understand how you'll probably hold a Chinese Abacus if that was the case.

Boulier1.JPG


Hullo?
 
Last edited:
Isn’t the whole MAX fiasco/ tragedy a product of internal corporate culture and overly soft-touch regulation?
Not exactly a poster-child for claims of the aviation industry being over-regulated....
 
steelpillow said:
Can't read the Seattle Times. It blocks my access unless I let its adverts through. Well, Seattle Times, I don't need you as much as you need readers.:p
Click to expand...

I'm reading it with an ad-blocker, and it's definitely not whitelisted, ads show as grey spaces with 'advertisement' top centre.

But as the 787 appears to work its instrument indicators backwards by synthesizing what it wants them to display, and has not yet started falling out of the sky for that reason, I hesitate to dismiss such a horrific-sounding kludge out of hand.
Click to expand...

It's more sophisticated than that. Airspeed, weight, coefficient of lift, angle of attack and aircraft configuration are linked, essentially by a control law, and because control laws are simply mathematical equations, you can reconfigure them to output whichever value is missing. Which means you can look at the way the aircraft is flying with specific control surface settings and generate the expected airspeed. It's a sanity check on the ASI using a mix of other sensors and known values. If they disagree, something is definitely wrong.

We actually did something similar, though not so sophisticated, on the C-17 HUD, which was considered safety critical during LAPES. The air data was processed to generate the symbology on the HUD, and a second processor path then reversed the maths to verify the back-calculated air data was within tolerance of the input air data. If they weren't, something had gone very wrong. Of course USAF then never used LAPES on the C-17 in service.
 
Citing a record of "delay and non-responsiveness" to requests for information, Chairman Roger Wicker said. "It is hard not to conclude your team at the FAA has deliberately attempted to keep us in the dark."
The Mississippi Republican said it appears the FAA is trying to hide damaging information from his committee.
"I can only assume that the agency's stonewalling of my investigation only suggests discomfort with what might ultimately be revealed," he said.
Click to expand...

https://amp.usatoday.com/amp/3207403001
 
TomcatViP said:
Boeing cut 737 production by ~50% pushing Spirit aero on the verge of collapse.

https://finance.yahoo.com/amphtml/news/spirit-aero-falls-hours-financial-064711044.html
Click to expand...

It's pretty far from "on the verge of collapse" as it won't hit the issue until Q4, so 3-6 months from now depending when the payments are due - generally end of the quarter in the UK, not sure about the US. Renegotiating debt repayments probably covers 80% of companies right now (including Boeing which has used up a $14Bn loan facility and then raised $25Bn more by a bond offer it'll need to pay off at some point). I'd fully expect Spirit to successfully renegotiate terms, though it may have to make further redundancies. It's the smaller suppliers who are probably more at existential risk, little companies with niche products that no one else makes and who are entirely dependent on aerospace for income, but don't have the financial weight of a company the size of Spirit when it comes to renegotiating their debts.

Few more details in the Reuters report: Boeing's order to Spirit for MAX/P-8 fuselages is down from 216 at start of the year to 125 in May to 72 now, and it's already delivered about half of those.

https://uk.reuters.com/article/us-s...after-deeper-737-production-cut-idUKKBN23T368
 
Grey Havoc said:
That was a bit rushed, to say the least.
Click to expand...

Not really, the certification flights are just a formality that was only attempted after numerous test flights and hundreds of hours worth of ground testing by Boeing.
 
Among other problems with that is the fact that not many people trust Boeing these days.
 
In my years of travel, as long as the aircraft had jet engines and swept wings, they really did not care much about the builder. With the MAX news now ancient history I doubt people will concern themselves with the aircraft except for how it keeps "the Corona" out.
 
The biggest risk facing all the aircraft manufacturers is how quickly the airlines can bounce back from the effects of Covid 19 to be able to afford all the aircraft they currently have on order.
 
Grey Havoc said:
That was a bit rushed, to say the least.
Click to expand...

There'll have been a pre-agreed set of test-points, and unlike a development test-flight, where you're doing it to find if the aircraft does something unpredicted and doing analysis between each flight and possibly between each test point, here the aircraft will have already flown each point, probably repeatedly, with full analysis after each flight. And the FAA analysts will have already been through all of that. So what they'll have been doing is simply a demonstration flight so the FAA flight crew can state they've seen it done and didn't see anything amiss.

It's like a Critical Design Review, if you do it right it's over in a morning. But that's because you've done all the actual work in the fortnight beforehand, with a team of people working hand-in-hand with the customer to verify every question has been answered before it has been formally asked. The meeting itself is just a tick-box exercise to sign everything off.
 
Status
Not open for further replies.
Back
Top Bottom