Ethiopian Airlines Crash/ B737MAX grounding

bryangiggsy · 15-03-2019 11:47AM

Seems incredible to me that there should be a 3rd system flying the plane in the background . Should only be auto pilot or manual .

Bob24 · 15-03-2019 12:53PM

deandean wrote: »

Ouch. More similarities with the Lion Air crash... https://www.breakingnews.ie/world/pilot-of-stricken-ethiopian-plane-called-in-emergency-soon-after-take-off-911153.html

Friday’s report cites “a person who reviewed air traffic communications” from Sunday’s flight saying controllers noticed the plane was moving up and down by hundreds of feet, with its speed appearing unusually fast.
Satellite-based data showed that both the Ethiopian Airlines and Lion Air planes flew with erratic altitude changes that could indicate that the pilots were struggling to control the aircraft. Both crews tried to return to the airport.

“The decision to send the flight recorders to France was seen as a rebuke to the United States, which held out longer than most other countries in grounding the jets. The US National Transportation Safety Board sent three investigators to help French authorities.”

Thin line between helping and monitoring ...

Nijmegen · 15-03-2019 01:48PM

bryangiggsy wrote: »

Seems incredible to me that there should be a 3rd system flying the plane in the background . Should only be auto pilot or manual .

Never get on an Airbus, so - Fly it in normal law and try do a barrel roll and see what happens. The issue is probably not correctly informing pilots what's there and how to work with it, understand what it's doing, and move around it.

cnocbui · 15-03-2019 04:58PM

Boeing have said they will have a software update for the 737 Max in 10 days time. https://www.cnbc.com/2019/03/15/boeing-shares-rise-on-report-that-company-will-roll-out-software-upgrade-for-737-max-in-10-days.html

That suggests to me they have been working on it for some time, which also implies they knew perfectly well that there was a significant problem.

Quin_Dub · 15-03-2019 05:03PM

cnocbui wrote: »

Boeing have said they will have a software update for the 737 Max in 10 days time. https://www.cnbc.com/2019/03/15/boeing-shares-rise-on-report-that-company-will-roll-out-software-upgrade-for-737-max-in-10-days.html

That suggests to me they have been working on it for some time, which also implies they knew perfectly well that there was a significant problem.

They've been working on it a while - I saw an article the other day saying that it would have been released several weeks ago but had been delayed by the US Government shut-down (delayed FAA/Regulatory sign-off perhaps??) so that means they've been working on it since shortly after the Lion-Air crash at least , if not longer.

mikel97 · 15-03-2019 05:04PM

Yes it looks like they knew but Trumps shutdown delayed it.

https://www.salon.com/2019/03/15/boeings-safety-update-to-737-max-was-reportedly-delayed-due-to-government-shutdown/

Shn99 · 15-03-2019 05:06PM

cnocbui wrote: »

Boeing have said they will have a software update for the 737 Max in 10 days time. https://www.cnbc.com/2019/03/15/boeing-shares-rise-on-report-that-company-will-roll-out-software-upgrade-for-737-max-in-10-days.html

That suggests to me they have been working on it for some time, which also implies they knew perfectly well that there was a significant problem.

They did know, and this software update was originally due out in January

Hotblack Desiato · 15-03-2019 05:21PM

Trump's ego trip cost 157 lives

Candamir · 15-03-2019 05:41PM

Nijmegen wrote: »

Never get on an Airbus, so - Fly it in normal law and try do a barrel roll and see what happens. The issue is probably not correctly informing pilots what's there and how to work with it, understand what it's doing, and move around it.

At least an airbus wont try to kill you while you try to roll it!

(And it’s pretty easy to force it into alt law, the roll is a bit trickier to pull off well)

Limpy · 15-03-2019 05:46PM

cnocbui wrote: »

Boeing have said they will have a software update for the 737 Max in 10 days time. https://www.cnbc.com/2019/03/15/boeing-shares-rise-on-report-that-company-will-roll-out-software-upgrade-for-737-max-in-10-days.html

That suggests to me they have been working on it for some time, which also implies they knew perfectly well that there was a significant problem.

What happens then? Can it be done where the plane currently sits. Would pilots need new training, what time scale are we talking from when software is rolled out.

Edit: just read it may take 6 months to certify the fix.

plodder · 15-03-2019 06:10PM

It's going to be interesting to see the nature of the software fix. It will draw a lot of attention to the problem that MCAS was originally trying to solve, and if/to what extent the fix affects the normal operation of it.

Bob24 · 15-03-2019 06:49PM

Hotblack Desiato wrote: »

Trump's ego trip cost 157 lives

Based on the above post, the Trump card seems to be a good one for Boeing and the FAA to escape their responsibilities :-s

A government shutdown doesn’t mean no-one is working - if they knew there was a life threatening situation and the shutdown would slow down the fix, they could either have assigned emergency ressources to get it done or grounded the plane.

cnocbui · 15-03-2019 07:07PM

I wouldn't be so sure about those emergency resources, The air traffic controllers weren't getting paid and Canadian ATC workers were doing whip rounds and were sending them Pizzas.

Bob24 · 15-03-2019 08:48PM

cnocbui wrote: »

I wouldn't be so sure about those emergency resources, The air traffic controllers weren't getting paid and Canadian ATC workers were doing whip rounds and were sending them Pizzas.

There is no way I would believe that no emergency staff was available at the FAA and that if a critical potentially life threatening issue had been identified with a fix pending validation, there was no way either to ground the plane until validation is complete or to do whatever was required to get the fix through ASAP. And Boeing was certainly not affected by any shutdown and could also have made decisions.

L1011 · 15-03-2019 08:49PM

You can believe what you want - but anyone who was in was effectively voluntary as they weren't getting paid.

It may have had no impact on the potential fix at all; but it screwed up lots of other stuff.

Bob24 · 15-03-2019 08:54PM

L1011 wrote: »

You can believe what you want - but anyone who was in was effectively voluntary as they weren't getting paid.

Of course, I get that and it shouldn’t be that way (although my understanding is that in most cases it wasn’t exactly voluntary work as they actually got paid later on, which obviously is only right).

But it doesn’t change the fact that they were there and meant to deal with any urgent/critical issue. And again Boeing was also in a position to make decisions and didn’t suffer any shutdown.

I’m not in the business of defending Trump here - I don’t care about that. But touting him as potentially responsible for the lack of fix seems like an easy way for the actual decision makers in this matter to disctract public attention from their own choices and responsibilities.

Odelay · 15-03-2019 09:02PM

Hotblack Desiato wrote: »

Trump's ego trip cost 157 lives

No. Boeing knee they had an issue. They could have grounded the planes until the fix had been approved. As much as I hate to say it, this has nothing to do with Trump.

Tomrota · 15-03-2019 09:45PM

plodder wrote: »

It's going to be interesting to see the nature of the software fix. It will draw a lot of attention to the problem that MCAS was originally trying to solve, and if/to what extent the fix affects the normal operation of it.

Does anyone know if this kind of design is normal on a modern aeroplane? From reading a multitude of different sources, I get the impression that the plane’s design wasn’t altered to accommodate the new changes (eg. The engine) that they wanted to make. And due to this, the plane has a really skewed centre of gravity and is limited by software.

But how does this software cause the plane to nose dive? Just one malfunctioning sensor? This design seems bonkers to me.

Would the A320neo, for example, have a similar design and software systems in place? If so, what makes it different from the 737Max?

Hotblack Desiato · 15-03-2019 11:58PM

Odelay wrote: »

They could have grounded the planes until the fix had been approved.

Two chances of that happening. The FAA didn't consider it unsafe to keep the MAX flying until ET302 went down - or rather until lots of other CAAs started grounding it...

As much as I hate to say it, this has nothing to do with Trump.

If it wasn't for the shutdown, would ET302 have had the fix in place? Maybe, maybe not.

Hotblack Desiato · 16-03-2019 12:18AM

Tomrota wrote: »

Does anyone know if this kind of design is normal on a modern aeroplane? From reading a multitude of different sources, I get the impression that the plane’s design wasn’t altered to accommodate the new changes (eg. The engine) that they wanted to make. And due to this, the plane has a really skewed centre of gravity and is limited by software.

But how does this software cause the plane to nose dive? Just one malfunctioning sensor? This design seems bonkers to me.

Would the A320neo, for example, have a similar design and software systems in place? If so, what makes it different from the 737Max?

Seems to me that Boeing are between a rock and a hard place on this one.

Airbus went in totally the other direction to Boeing and everyone else long ago - basically the pilot flies a computer and the computer flies the plane. People thought this was a very scary idea 30 years ago, but it works well enough and has as good as safety record as anything else, if not better.

Boeing didn't used to have computers making control inputs with autopilot off, but now they do, and 737 MAX pilots who think with AP off that they're the ones in control are finding out that that's not necessarily the case. But they don't have the flight envelope protections that Airbus have, either. Seems they're now caught between two stools, between two incompatible flight control paradigms, two worldviews of how man and machine should interact.

Yes they should really have introduced a whole new aircraft to replace the 737 before now, but that would mean (a) massive investment (b) a new type certificate, and if your customers are going to have to retrain all their pilots and ground crew then what's stopping them buying A320 family? which would be cheaper than a 737 replacement as the A320 development costs were mostly incurred 30 years ago and long since paid off.

Without some very trick, expensive, heavy and probably less reliable extensible main gear, or a very major redesign of the fuselage and wing box, there is no option but to mount the MAX engines even higher and further forward than they were on the 737 NG, whose engines were already in a bit of an odd position anyway. Boeing found that the nacelles on the MAX were generating lift at high angle of attack - they're well forward of the centre of gravity so this means that when you pitch up too much the aircraft generates a tendency to pitch up even more. Not good, which is why they brought in MCAS to counteract this. (some have said that MCAS was to make the MAX feel the same as the NG to the pilot - not strictly true as MCAS is only supposed to kick in in an unusual AoA situation)

Whether FAA were right to allow Boeing to certify the MAX on the same type certificate, using MCAS to counteract an inherent aerodynamic flaw, is very debateable. Clearly there is pressure on them to support "team USA" and one of the FAA's offical roles is to promote aviation commerce, which conflicts with its goal to promote aviation safety.

Airbus are well positioned now to keep on developing the 320 family for years to come until they eventually replace it with a composite fuselage aircraft. The irony is that it'll be easier for them to make any 320neo-neo variants fly just like the existing ones (not constrained by low undercarriage etc. so flexibility in size and shape of engines is there) and also easier for them to use the flight computers to make any differences in flight characteristics disappear as far as the pilot is concerned - without affecting their existing model of how the pilot interacts with the plane.

What seemed like a bit of an odd decision 30 years ago looks very prescient today.

Hotblack Desiato · 16-03-2019 12:36AM

As for a single failed sensor having the capability of bringing down an aircraft, that's just gobsmacking. Someone in the FAA was asleep at the wheel for that one. They didn't even mandate that the AoA sensor disagree warning light was a requirement not an optional extra.

Nijmegen · 16-03-2019 01:20AM

Re the shutdown, I know Delta had to delay the introduction of the A220 because they couldn’t get certification inspections, but Southwest arranged to pay for the inspectors to help them out of a similar jam. https://www.wsj.com/articles/southwest-agreed-to-pay-faa-for-inspectors-time-during-government-shutdown-11548804686

vectorvictor · 16-03-2019 09:26AM

Hotblack Desiato wrote: »

Airbus went in totally the other direction to Boeing and everyone else long ago - basically the pilot flies a computer and the computer flies the plane. People thought this was a very scary idea 30 years ago, but it works well enough and has as good as safety record as anything else, if not better.
how man and machine should interact.
...
What seemed like a bit of an odd decision 30 years ago looks very prescient today.

Notwithstanding that it also took a few A320's being planted into the ground too before it was properly refined and training improved.

howsshenow · 16-03-2019 10:02AM

Re "Boeing didn't have computers making control inputs with Autopilot off" ...... What about the Stabiliser speed trim system and perhaps the Elevator Mach trim actuator?

Henry Ford III · 16-03-2019 10:40AM

Why was the 737 Max put into service if it was known to have issues?

Wanderer78 · 16-03-2019 10:59AM

Henry Ford III wrote:

Why was the 737 Max put into service if it was known to have issues?

I'd imagine many aircraft have known issues, before and maybe even after gaining airworthiness, minor problems I'd imagine, but I'd say there's a lot of pressure getting them into the sky, and a lot of pressure placed on aviation authorities to do so, it shouldn't happen of course, but.....

Roger_007 · 16-03-2019 12:17PM

Hotblack Desiato wrote: »

As for a single failed sensor having the capability of bringing down an aircraft, that's just gobsmacking. Someone in the FAA was asleep at the wheel for that one. They didn't even mandate that the AoA sensor disagree warning light was a requirement not an optional extra.

The problem with all software is that there are always bugs. This usually because the programmers are not able to predict every possible scenario and what action to take in each event. Programs usually work fine 99% of the time, or even 99.999% if it's very good.
In real life it is not possible to predict everything that might occur. If a system relied totally on the perfect functioning of just one sensor, then that is a bug, and a very large one. Sensors can and do malfunction. It's a fact of life. Obviously there should be redundancy in sensors for critical systems. Relying one one sensor, if that's what happened, seems negligent in the extreme.
Also I fail to understand why this MCAS was not integrated with the autopilot. Either the pilot is flying the aircraft manually or is not. If the autopilot if switched off then why is there another system still on overriding the pilots instructions.

cnocbui · 16-03-2019 01:08PM

Roger_007 wrote: »

The problem with all software is that there are always bugs. This usually because the programmers are not able to predict every possible scenario and what action to take in each event. Programs usually work fine 99% of the time, or even 99.999% if it's very good.
In real life it is not possible to predict everything that might occur. If a system relied totally on the perfect functioning of just one sensor, then that is a bug, and a very large one. Sensors can and do malfunction. It's a fact of life. Obviously there should be redundancy in sensors for critical systems. Relying one one sensor, if that's what happened, seems negligent in the extreme.
Also I fail to understand why this MCAS was not integrated with the autopilot. Either the pilot is flying the aircraft manually or is not. If the autopilot if switched off then why is there another system still on overriding the pilots instructions.

The MCAS stays on because it's supposed to make the flying characteristics of the plane as similar as possible to the earlier versions so it wouldn't require re-certification. It's main use is where the plane would likely be manually flown - takeoff and climb. At high angles of attack, the new engine configuration creates lift over the engine nacelle, forward of the wing, causing the nose to pitch up even more. It's supposed to alter the manual flying charactristics of the plane, which is why it isn't disengaged with the autopilot.

The short answer and reason for MCAS is money. Boeing wanted to sell as many as possible by making it as attractive as possible to ever cost-conscious airlines by producing something that appeared to be an almost direct swap for planes they were already using. They also wanted to avoid the considerable cost to themselves of fixing the pitch-up problem by aerodynamically altering the plane to overcome it - essentially a new plane design and all the costs of that and getting it certified.

The FAA are more culpable than Boeing as a core element of their very job is to stop Aircraft manufacturers prioritising costs over safety and making them do things properly.

The law suits are going to be substantial and won't be easily fixed with a software update.

cnocbui · 16-03-2019 01:19PM

Hotblack Desiato wrote: »

...

Whether FAA were right to allow Boeing to certify the MAX on the same type certificate, using MCAS to counteract an inherent aerodynamic flaw, is very debateable. ...

Very good summary, however, I think it would be safe to say allowing the MCAS kludge was debatable when they were thinking about it, but it's no longer debatable that they did the wrong thing.

Padraig Mor · 16-03-2019 02:22PM

Maybe I'm reading things wrong, and obviously we have no definitive answers to either crash yet, but - to me - it seems like there isn't really a specific problem with MCAS per se (other than lack of pilot training re it), but rather its implementation - specifically that it *appears* that MCAS was triggered by faulty data sent by a single AOA sensor. If the AOA data had been set up properly, is it not likely that there would not have been crashes?

Ethiopian Airlines Crash/ B737MAX grounding

Comments