jsloan

Well, there are certain countries whose technicians I wouldn't trust. Neither the US nor France is on that list; I agree that either would have done fine. My problem was the insinuation that the FAA was somehow tainted by this conflict of interest -- the very same conflict of interest that the DGAC has.

VoiceActor

I would trust a trained professional human over a computer any day. When my GPS wants to take me into a ditch because it's not up to date with the latest road construction routings, it's my knowledge of the roads and the human capability to exercise clear and sound judgement based on the criteria, that keeps me from driving into the ditch. While self-driving cars continually get rear-ended because they stop at yellow lights, which is completely counter to human nature, it's my awareness that the vehicle behind me is going too fast to stop that makes me speed up to get through the yellow. It's unconscionable - if indeed this is the case - that a flight system would prevent a manual override because it's been "designed to know better". Really? Our over reliance on tech will doom us. Who's supposed to be in control here?

FlyngSvyr

Looks like the examination of the black boxes is not getting off to a good start Saw this in a news article this morning.

If the investigation is not run correctly following international protocols, it may make it hard to come to definitive conclusions, or at worst might have/imply biased conclusions. Everything needs to be above board with agencies on both sides of the pond for international public confidence.

bman1002

Nearly every country in Europe has their finger in the Airbus pie. I think it would be hard to find a country a country on the continent completely free of any conflicts of interest.

mduell

I can't think of a single airliner that doesn't have a runaway trim procedure. I also can't think of a single jetliner without multiple systems for activating the trim.

The content of these questions betrays a fundamental misunderstanding of aircraft design, certification, and human factors. Given the tone and loaded language presented here, I don't get the impression that even detailed answers will provide any clarity.

I have very little faith in the ability of pilots to interpret voltages from the hardware. Computers running software are providing large amounts of useful augmentation to help the pilots do their job.

1970s called, they want their human factors mentality back. Sometimes we're going to let the computer override, sometimes we're going to let the pilot override. In this case, runaway stab, there's a well designed/trained/practiced procedure for pilot override; in the worst case, you can grab the control wheel.

Naked political grandstanding; NTSB has far more experience with 737 FDR data analysis than BEA. And there's already reports of Annex 13 issues.

EWR764

I wouldn't exactly call it "fraught", but Airbus products have similar procedures. For instance, the 320 series has automated stabilizer trim built into the flight control architecture, but there's still a manual trim wheel that will override automation if a pilot intervenes. There's no cutoff switch like on the 737.

Airbus FBW aircraft have a different flight control philosophy which is probably beyond the scope of our discussion here, but a) there is a way to manually trim the 320 series, and b) both Airbus and Boeing products have protections built in to​​​​​ address runaway stabilizers, they just have different ways of doing it.

It's a single point of failure insofar as a failed instrument will likely lead to a series of successive errors before the cascade can be stopped, and in this case, manual intervention is probably required. If true, that is, by definition, less redundant than other control augmentation systems.

A stab runaway isn't limited to phases of flight where the MCAS protection could theoretically apply, and there's no evidence that the Lion Air crash was actually a stab runaway. It presents the same way to pilots, so the response should be the same, but the Lion Air crash appears to be related to the faulty sensor continuing to feed data to the FCC that the airplane was being flown into a dangerous situation, and the MCAS kept doing its job (activating to prevent that low-energy state). The way it is designed, if instrumentation is working properly, any manual intervention by the pilot will disengage the MCAS because it suggests the crew is actively managing the situation. MCAS will reassess its criteria five seconds later to see if the situation has resolved. Simplistically, the theoretical conversation between MCAS and the crew would go something like this:

MCAS: Hey guys, looks like you're flying into a possible stall, let me start to pitch the nose down to help you out.
Pilot: Thanks, MCAS, we've got it (trims nose down).
MCAS: Ok, I am disengaging.
Pilot: Thanks, MCAS.
MCAS: (five seconds later) Hey guys, still looks like there's a problem, starting to pitch down.
Pilot: (trims nose down) Nope, still good.
MCAS: (five seconds later) Just checking... looks good now.

If the situation doesn't resolve, either because the airplane is still being flown into the stall, or the faulty instrumentation fools the computer into thinking it is, it will keep trying to kick in. At that point, pilots should be able to recognize either that the airplane is approaching a stall, or the response (uncommanded nose down input) is incongruent with other apparent factors, or is not resolving, and steps should be taken to correct the discrepancy. 737 pilots have weighed in on the issue in the course of this thread, but to reiterate, if the controls keep getting heavier (requiring more back pressure on the yoke) while opposite inputs are commanded, the next step is to run a non-normal runaway stab checklist. This will 100% of the time correct an unwanted activation of MCAS protection, because it completely cuts out automated pitch trim.

You emphasize "flaw". What makes it a "flaw"? Is there some objective criteria you apply to arrive at that conclusion? If that's a flaw, then yes, all aircraft have lots of flaws which are mitigated in many ways... mechanically, aerodynamically, by automated flight controls, etc. Again, no different than systems like yaw damper or Mach trim that prevent seriously dangerous consequences that are incidental to the physics of hurling a gigantic aluminum tube through the air at 500mph+.

Moreover, if the 737MAX were to require a new type certification, it probably doesn't sell nearly as well as it has, and costs substantially more to develop as well as for airlines to acquire. Economics absolutely come into play, and the public accepts a reasonable compromise between cost and safety, regardless of whether we like to admit it.

Pilots aren't manually cutting off MCAS; they are cutting out automatic trim, and the 737 can be trimmed manually. MCAS works to prevent a stall in a low-energy, hazardous situation that pilots shouldn't find themselves in during the course of normal operations (high AOA, high bank angle, manual and clean configuration). This isn't something that's actuating on every flight... MCAS is very much a protection toward the edge of the envelope.

An aerodynamic stall should never, ever happen in normal commercial flight operations, but strictly speaking, stall does not = fatal. I would wager that the average traveler would be truly terrified to know why many automated systems on commercial aircraft actually exist. For the most part, they work flawlessly day-in and day-out to prevent major catastrophes, and when they don't, that's one reason we still have at least two well-trained pilots sitting at the pointy end.

The same reason a brand new BMW or Honda can require maintenance at low mileage. Things break, parts fail, and need to be replaced.

skidooman

No, it doesn't. Three different hardwares, three different versions of the program that are contributing to the poll. Otherwise, you are right, it would be pretty dumb.

Got my hands a bit more intimately on one of these projects. The chips, compilers, boards and programs, all different. Takes lots of time, we are talking months and years, but hey, that plane is worth millions and will fly decades, so the investment is worth it.

Of course not ALL the systems are level A. Things like the flight management system, brakes, etc, the stuff that must absolutely not fail, that is level A. The rest have lesser requirements and not the triple redundancy.

Ah, one piece of data, really told to me orally, so cannot vouch for accuracy (and you can bet Boeing and Airbus are NOT publicizing this). On average, how frequently does one of these computers say something different the two others and gets rebooted during TATL? 2-3 times. Not sure if it is for all systems or per system. I am sure a pilot may offer more juicy details here.

I do agree with EWR and stephen on this one: there are two people paid in front of the cabin that are trained to keep the plane safe. When a system goes rogue, they should be the ultimate fail safe. Here, maybe this is what went wrong, or so it seems. The report will tell.

EWR764

You're referencing the Habsheim crash, which, remarkably, was the very first passenger-carrying flight of the A320 (the first production FBW airliner) and occurred at an airshow. A few dozen people lost their lives while thousands of spectators watched the airplane crash into a forest off the end of the runway during a flight demonstration. Could you imagine the social media-driven panic that would ensue if a similar occurrence happened today? Dare I say, it might have killed the A320 and even Airbus. These are very different times...

It's somewhat ironic that you reference that crash, too, because that investigation has always been marked with doubt stemming from the high political stakes involved with the massive investment of the French government in the program and Airbus as a whole. A credible counter-theory exists which posits that the incipient flight control law of the A320 simply overrode the pilots' attempts to execute what was essentially a missed approach, instead behaving as though the airplane was attempting to land. Substantial changes were made to the flight control software in the wake of that crash.

I also have a philosophical problem with investigations carried out by an agency in a country where virtually every commercial air crash results in criminal prosecution... including the Habsheim accident.

I certainly would not hold up the French BEA as any sort of global standard-bearer in air crash investigation and impartiality... but what do I know?

surram

Many thanks EWR764 for your detailed explanations. Appreciate it. The "flaw" reference that I made is the 'inherent aerodynamic instability in the MAX' - based on the opinion of Captain Girish Kaushik - the excerpt from the link below:

Captain (Retd.) Girish Kaushik, who had served in Jet Airways and Air India among other airlines, says:
“Boeing 737-200, then 300, 400, 500, 700, 800, 900 and MAX 8, they seem to be doing a quick fix. They had mounted larger engines far stretching the wings to prevent them from touching the ground. This, I believe caused aerodynamic instability.”

Link to full article:
https://www.thehindu.com/business/ma...?homepage=true

bocastephen

Not looking good so far - early news indicates the jackscrew was found in a position that indicated a full nose-down attitude, similar to the Lion Air crash. This is a contributing factor to the urgent grounding that took place in the USA.

Something shoved that airplane nose down with such veracity the pilot was unable to bring it up.

So now my question is thus - what pilot reports exist that indicate similar behavior by automated flight control systems that the pilot was able to override and restore normal attitude? Do such reports exist that might indicate a trend showing a serious, dangerous defect, and at the same time, pilot error in mishandling the defect with these two crashes?

MSPeconomist

surram

Respectfully jsloan, I believe there is a lot of unease about the very cozy relationship between FAA, Boeing and the Trump administration. It causes me a lot of grief. I read a lot on Vox and is trustworthy and produce excellent articles. So, I can naturally understand why the US has lost its leadership and trust in the eyes of the world. I have lost the trust too.

Please read this article:
https://www.vox.com/policy-and-polit...ump-tweet-call

EWR764

Let's please not turn this thread political and start discussing the interplay between the Trump administration, the FAA and Boeing, as I am very much enjoying the technical discussion and that's an easy way to run this thread off the rails.

After reading all of the MAX-related ASRS reports (with the understanding that compilation is by no means exhaustive) I can report that it does not show a trend involving the same systems. There were a few entries the media briefly attempted to spin as "evidence" of related complaints, but they were as comparable as apples and oranges to what's been mentioned as causal in the Lion Air crash. The only similarities come at a very broad, technically-insignificant level.

surram

I agree EWR764. You have made very good, detailed points and they are very useful. Really appreciate it. I was merely citing that article to explain the angst of Ethiopian Airlines, and why US was not the preferred destination for the black boxes. This in my mind (top of my head) is very unusual...as the US/NTSB is the most trusted and has the best technology and tools to investigate such events.

Anyway, your point noted...and that will be the end of this topic.

mduell

These are bad examples of DAL A systems, because they're not DAL A systems.
FMS is generally level C, or maybe level B sometimes, but certainly not level A as it does not meet the requirement for severity of failure
conditions ("Failure conditions that would prevent continued safe flight and landing" for Catastrophic requiring level A). Similarly, braking systems are level B.