737 Max revelations

[Ron Attwood 3,465]

I was sent a link to this video. Not being an airliner aficionado, this was all news to me. Loads of you will probably be more au fait with what the chap says.

 

[Chock 17,769]

The gist of what he is saying is true, and it's not really news either, however, he does get a few things incorrect.

For example, in alluding to how deHavilland corrected the issues with their Comet airliner after it had several fatalities when pretty new, he mentions that the fault with the deHavilland Comet was that it employed untried design technology, being a pressurised fuselage, which was subsequently weakened by repeated cycles. Whilst that is somewhat true, there's a bit more to it than that. There were quite a lot of pressurised aeroplanes built long before the Comet, and that in itself is not really what caused the accidents with it anyway. The fact is, with the Comet 1, it was some poor design and manufacturing decisions on the part of deHavilland which were really at the root of the problem, notably the choice to make the windows a (inherently structurally weak) rectangular shape, rather than round, or perhaps square with rounded 'corners'. Also the fact that the design called for the panel rivet holes to be drilled, but when deHavilland commenced manufacture, they punched the holes out instead because it was a speedier process, but one which unfortunately made the edges of the rivet holes microscopically ragged, so they too were prone to tearing, as were the window corners. Thus, rather than being the opposite of the issue with Boeing, i.e. a case of bad luck with a pioneering design and honorably rectifying it, as he suggests, it is in fact a very similar state of affairs to Boeing's issues, i.e. doing stuff on the cheap and using a less than ideal design feature.

But, with regard to the 737 MAX and his suggestion that it is an inherently deadly plane by dint of its design which uses the MCAS software to maintain control, and that they should instead have gone with a modern fly-by-wire design. Whilst you could say there is a hint of truth in that, it is somewhat word not allowed of what many fly by wire designs actually do. Many modern FBW aeroplanes are inherently unstable because that instability makes them aerodynamically cleaner and more fuel efficient, and when this is so, it is the FBW system which - behind the scenes so to speak - constantly tweaks the control surfaces by tiny amounts in order to keep the thing in stable flight when it would in fact go completely out of control without those tweaks, so to suggest that MCAS is some kind of weird system the like of which can be found nowhere else on any other aeroplane, is not true at all.

But even if that were so with MCAS, he neglects to mention the real cause behind at least one of the accidents of a 737 MAX, and that is the fact that the aircraft had a mechanical issue with its angle of attack sensor, and rather than go to Boeing to have the issue resolved, the airline went to some dodgy local air maintenance place, which proceeded to fit a replacement angle of attack sensor which was an old one (for the NG, not the MAX) which they had lying around on a shelf, and then they didn't bother to calibrate it, which meant that the airspeed indicator and the artificial horizon on the accident plane was getting false speed and angle of attack data from that sensor.

And then as if that wasn't bad enough, when the NTSB spoke to that maintenance place as a result of the accident investigation, they lied about having calibrated the AoA sensor and tried to fake some records to show they had done it, but they got found out because they forged some photographs and either forgot, or did not know how to change the metadata on them which proved they were taken after the accident and not during the maintenance procedure on the accident aircraft.

So yeah, what Boeing has done is not ideal, but there's a bit more to it than that.

Edited March 11, 2020 by Chock

[nawarren 83]

2 hours ago, Chock said:

Many modern FBW aeroplanes are inherently unstable because that instability makes them aerodynamically cleaner and more fuel efficient, and when this is so, it is the FBW system which - behind the scenes so to speak - constantly tweaks the control surfaces by tiny amounts in order to keep the thing in stable flight when it would in fact go completely out of control without those tweaks.

I think you are getting confused with Fighter A/C. Commercial aircraft are required to be very aerodynamically stable and should not exhibit instability. If they did they would not get a certificate. That's part of the issue with the MAX and why the MCAS was invented to help mask the aerodynamic instability caused by the bigger engines in a different position. 

[Chock 17,769]

Nope, I'm not getting confused on that. You are correct that unstable FBW control is a desirable feature for combat aeroplanes, but where the tailplane is concerned, it's also a desirable feature on any aeroplane which you are trying to squeeze a higher MPG out of too, which is what Boeing were trying to achieve with the MAX. Although perhaps a more famous example of going for less stability in an airliner to improve fuel economy and using a computer control system to get around having the thing not be twitchy, is the MD-11, which has a computerised stability augmentation system which varies the fuel in a tailplane-located ballast tank and which also drives the horizontal stabiliser too, rather than it having a directly proportional link to the yoke in the traditional sense.

This arrangement was designed to enable the MD-11 to have broadly similar handling characteristics to the DC-10 in spite of the MD-11 having a smaller tailplane, so it could have an easier transitional type rating. And the reason it had a smaller tailplane, despite this theoretically making it less able to counter the rearward main wing upward pivot of the  lift vector (which would have made it twitchy without that computer augmentation), was because a smaller tailplane which does less work in countering the less desirable aspects of lift generation, reduces drag and therefore increases fuel efficiency.

On aircraft which had tough flaps and slats with lots of settings, including some which would be detrimental to lift when in cruise if set at certain positions, pilots often did a similar thing to take the load off the tailplane by reducing the wing's lift in order to gain some fuel economy, in spite of this not being recommended practice. The B727 was an aircraft you could do that on although it did make it more unstable. It is believed that the crew of TWA Flight 841 in 1979, which was a 727 that went out of control and actually broke the sound barrier in a dive before managing to recover, were indulging in that practice, although they never admitted to having done so and they wiped the CVR before they landed! They claimed that erasing the CVR was something they always did after a landing since at that point they knew they were down okay, and they did it on that occasion from force of habit, which is a bit hard to believe although it is difficult to prove otherwise, so they got away with it. 🤣

Edited March 12, 2020 by Chock

[ESzczesniak 128]

2 hours ago, nawarren said:

I think you are getting confused with Fighter A/C. Commercial aircraft are required to be very aerodynamically stable and should not exhibit instability. If they did they would not get a certificate. That's part of the issue with the MAX and why the MCAS was invented to help mask the aerodynamic instability caused by the bigger engines in a different position. 

Stability is some issue with more powerful low hung engines, but it’s not inherently unstable. It increases the thrust-pitch couple whereby power changes cause the aircraft to pitch. Virtually all aircraft have this behavior to some degree. Retrimming is alway required after a power change. It’s the magnitude that changes. 
 

The major reason for the MCAS was to make the thrust-pitch response “feel” like he 737NG so they wouldn’t need an extensive certification phase as a “new airframe”. It instead was streamlined as a modification to an existing airframe.

The 777 in fact uses a not entirely dissimilar concept for its C*U FBW law.  Here, the FBW introduces expected pitch changes based on speed to make the aircraft feel as its flying like a conventional aircraft. 
 

The issue with the MAX and the MCAS is that, by default, it relies on a single AOA sensor. These have come a long way, but at their heart still rely in vane sensors prone to erratic readings. An option for a redundant second AOA sensor is available for a cost, but should have been default. Relying on a single sensor is never a smart idea. 
 

Pilot training is also an issue. Cursory review of the MCAS documentation clearly state what it does and the advisories that it’s active. In a runaway trim situation, the system is either interrupted by use of the yoke elevator trim, or fully disabled with the stab trim cutout switches. It boggles my mind that when a basic pilot mantra is “trim, trim, trim”, even for GA, that these pilots never touched the trim once when the aircraft was in “unstable” flight. 

[overspeed3 410]

Boeing is now the new poster child for corporate greed - which may be a well-worn cliche for American business, but they certainly earned that title.  The only mitigating circumstances for producing a supposedly flawed aircraft like the Max, was the fierce competition from Airbus, which received European subsidies to prop-up their company's ability to compete with Boeing:  A vicious cycle?   

But, if the 737 Max was so unwieldy, and with insufficient pilot  training, how did thousands of domestic and global flights manage to fly and land safely over the years with seeming little serious incident prior to the two crashes?  Did all those pilots and 1st Officers dummy-up?  

[Biggles2010 343]

1 hour ago, overspeed3 said:

But, if the 737 Max was so unwieldy, and with insufficient pilot  training, how did thousands of domestic and global flights manage to fly and land safely over the years with seeming little serious incident prior to the two crashes?  Did all those pilots and 1st Officers dummy-up?  

The 737max  was only in service for 2 years before the first crash. Perhaps the previous flights were just lucky. Remember the DC10 had a similar history, with many flights before the fatal crashes caused by faulty design of the locking mechanism for the cargo doors, allowing them to unlatch in flight. Similarly the Trident was in use for some time in Europe, before a fatal crash, for which one of the causes was the early retraction of the leading edge slats, which could be done without triggering a warning system. In this case there were other major factors too, but it showed that a design fault, that seems so obvious now, could be allowed to enter service and take some time to show up.

 

 

[threexgreen 3,139]

18 hours ago, ESzczesniak said:

Pilot training is also an issue. Cursory review of the MCAS documentation clearly state what it does and the advisories that it’s active. In a runaway trim situation, the system is either interrupted by use of the yoke elevator trim, or fully disabled with the stab trim cutout switches. It boggles my mind that when a basic pilot mantra is “trim, trim, trim”, even for GA, that these pilots never touched the trim once when the aircraft was in “unstable” flight.

They did trim the aircraft in both cases, manually counteracting MCAS trim inputs. The thing which is striking is that even though they obviously recognized the trim runaway, they never trimmed enough to fully neutralize MCAS trim inputs. While an MCAS (flawed version) nose down trim input lasted for 10 seconds, the Ethiopian pilots trimmed only for about 3 seconds against that. The only exception was the Lion Air Captain who successfully kept the airplane in proper attitude when MCAS was actually activated (which, incidentally, was only active for about 1 minute on that flight). MCAS, even though severely flawed in its first version, was designed to always give priority to manual pilot electric trim inputs, so the pilot can always successfully negate any MCAS input and also return the trim into a neutral position before cutting the electric trim system out.

Personally, I think this is where excessive reliance on automation comes into play. Trimming away forces as a result of thrust changes, flaps extension/retraction, etc. while flying manually is so fundamental in flying most aircraft that it should be absolutely natural for any pilot. In contrast to the US, there is more reliance on automation in other parts of the world and many pilots engage the A/P very soon after takeoff and only disconnect it when the aircraft is fully configured and stable on final. This way nearly all trim inputs to trim the forces away are handled by the automation and you virtually don't have to trim manually anymore and your manual flying skills start to rust. Now you have MCAS trimming and you feel the forces on the control column you are holding immediately. Naturally you will start to counteract and trim the forces away, but these crews stopped trimming halfway through. If you notice that your first trim input was not enough because the aircraft is still out of trim (MCAS isn't trimming again until 5 seconds after the last pilot trim input) then trim some more and repeat that when the aircraft wants to go down again. Obviously, this is just a guess, but I can't see any reason other than a lack of feeling for trimming manually as to why the sort of hesitant trim inputs.

In general, I was a bit surprised at some things he said, given he is a 737 pilot. Most notably his comment on the manual trim system. The Ethiopian aircraft was flying at speeds way beyond V_mo of 370 kts and above instead of a maximum of 250 kts at that altitude. I find it hardly surprising that you can't trim manually at those speeds and aerodynamic loads. The way I understood him was that he was saying it's not possible to trim manually even at low speeds, which raises quite a lot of questions marks: most prominently the fact that the Lion Air crew before the accident flight had the same problems, cut out the trim and actually continued to their destination (whether that was a safe decision is different matter) with manual trim. Based on this guys comment that aircraft would have been significantly out of trim for the entire flight. I've also seen plenty of reports from other pilots who ran into a scenario where they had to trim manually and were actually able to do so.

Overall, there are certainly valid points that suggest going for a new, clean sheet design rather than the MAX. But I find the whole discussion about the MAX being "inherently unstable" is way overblown. The increased nose up momentum at high AoAs due to the large engine nacelles is obviously not a good aerodynamic characteristic and undesirable but it's actually the only difficulty this aircraft has. The reason for all what's going on is a terrible MCAS design and how Boeing appear to have put profit ahead of proper development and therefore safety.

As @Chock pointed out, there are many aircraft that use software solutions to help with aerodynamics, e.g. the A321neo which exhibits an undesirable accelerated nose up pitch on go arounds at certain weights which was fixed by software to counteract that tendency. Sound familiar?

[ESzczesniak 128]

On 3/12/2020 at 4:03 PM, threegreen said:

...In contrast to the US, there is more reliance on automation in other parts of the world and many pilots engage the A/P very soon after takeoff and only disconnect it when the aircraft is fully configured and stable on final. This way nearly all trim inputs to trim the forces away are handled by the automation and you virtually don't have to trim manually anymore and your manual flying skills start to rust. Now you have MCAS trimming and you feel the forces on the control column you are holding immediately. Naturally you will start to counteract and trim the forces away, but these crews stopped trimming halfway through. If you notice that your first trim input was not enough because the aircraft is still out of trim (MCAS isn't trimming again until 5 seconds after the last pilot trim input) then trim some more and repeat that when the aircraft wants to go down again....

Perhaps this is more the point I was going for.  I don't know the accident reports and FDR data personally.  On top of that, I'm only an armchair pilot.  I think a very detail oriented one (PMDG or similar, QRH's available, real word ATC and weather, etc) but still an armchair pilot.  However, I am, prior to a career change, an aerospace engineer (CFD and compressor design) and think I understand the systems well.

As you said, flown manually, the phases of flight immediately after takeoff are full of trim changes.  Even as an armchair pilot I know this.  There's thrust reduction, acceleration height, multiple configuration changes.  You trim to neutral.  You continue to add or subtract trim until neutral control force.  If they were trimming this way, then they should've had at least a five second window the MCAS wasn't sending them in to a nose dive.  Then trim again because the aircraft is nosing down again.  This should have resulted in a nearly continuous window of the MCAS being deactivated due to manual trim inputs.  Beyond this, almost immediately after any failure I encounter in the simulator, the first action is to aviate...take manual control and fly the airplane stick and rudder.  Then, sort out the failure.

There are issues beyond reliance on autopilot systems and manual trim.  But theoretically, this would have been a non-issue in world with solid manual flight skills..