Boeing 737 MAX problems

[billvon 2,395]

On 3/22/2019 at 9:40 AM, PhreeZone said:

The Max10 is already in its certification process and it is expected to have all these issues plus a few more. So my guess is that Boeing will wind down production of the Max8, push forward the new certification of the Max10 with these issues resolved and have a few differences in there too and then will as a show of goodwill will toss in the upgrades and optional items to trick the planes out with everything for any of the existing Max8 orders that want to convert.

Updates - 

So there's a software fix out that Boeing wants approved on a fast track.  It does two things; it uses input from two AOA sensors instead of one, and it times out after a little while - meaning that as long as the pilot keeps trimming up every time the MCAS trims down, eventually the MCAS will stop trying.

Meanwhile early results from DFDR data indicates that it was the same problem in both crashes.  (There's been no official statement yet, but officials who have seen the data says it looks like the same problem, and the CVR records one of the pilots telling the other one to get the nose up.)

The question now - will the FAA be OK with another software fix?

[gowlerk 1,905]

51 minutes ago, billvon said:

The question now - will the FAA be OK with another software fix?

Is the aircraft safe and stable with the MCAS shut down? Or will some pilot disable it on a climb out because its annoying. And then have a departure stall.

[billvon 2,395]

12 hours ago, gowlerk said:

Is the aircraft safe and stable with the MCAS shut down? Or will some pilot disable it on a climb out because its annoying. And then have a departure stall.

Nope.  The MCAS feature was added because the aircraft would not pass certification testing without it.  So that will make it harder for Boeing to argue "if they shut it down it's not a big deal."

[gowlerk 1,905]

3 hours ago, billvon said:

Nope.  The MCAS feature was added because the aircraft would not pass certification testing without it.  So that will make it harder for Boeing to argue "if they shut it down it's not a big deal."

Exactly. And part of any fix will be for the system to shut down if it gets inconsistent data from one of the soon to be two sensors. Boeing has orders for 5000 of these aircraft. If they get built Mr. Murphy can pretty confidently predict with his law that bad things will happen to some of them. Covering up inherent air frame flaws with pilot over ruling software is just the wrong way to go in a commercial airliner when it is possible, but more expensive to do the right thing. It was not necessary except as a way of not losing sales to the competition in the short term. Both Boeing and the FAA took their eyes off the ball. 

And the FAA waiting to ground them after the second one fell out of the sky is inexcusable. If they felt data was lacking then they should have grounded them pending data. It is unheard of these days for airliners to crash like this. And the coincidence of both of them being this brand new model is more than enough of a data point. This whole affair stinks to high heaven.

[gowlerk 1,905]

For the last few days paid items have been showing up in my FB newsfeed from an organization called "Airline Ratings". Every one of them promotes a link to an article about how wonderful the Boeing certification process on the Max 8 was and how unfair any criticism of them or the FAA is. Now, where do you suppose the money for this is coming from? I wonder if they will be hiring Royreader next.

 

https://www.airlineratings.com/news/boeing-not-rush-737-max-design/?fbclid=IwAR0NTz58j5_hl-ooge0dbEKi_zQS1KIcMB9kPsVU1GMc9ycKS29Brv_eWvE

[baronn 111]

Only talks about the time they put into re-developing this aircraft. Not a single mention about the instability problems with adding the new motors. Airbus designed the Neos to accommodate these. The 737 never did. The MCAS is a band aid on a mortal wound. Will they get it recertified? Eventually. How? My money is on the fastest, least expensive way possible. That will involve payoffs wherever necessary and silencing the critics. Expect to see folks start to disappear.....

[billvon 2,395]

On 3/30/2019 at 10:32 AM, gowlerk said:

Exactly. And part of any fix will be for the system to shut down if it gets inconsistent data from one of the soon to be two sensors. Boeing has orders for 5000 of these aircraft. If they get built Mr. Murphy can pretty confidently predict with his law that bad things will happen to some of them. Covering up inherent air frame flaws with pilot over ruling software is just the wrong way to go in a commercial airliner when it is possible, but more expensive to do the right thing. It was not necessary except as a way of not losing sales to the competition in the short term. Both Boeing and the FAA took their eyes off the ball. 

And the FAA waiting to ground them after the second one fell out of the sky is inexcusable. If they felt data was lacking then they should have grounded them pending data. It is unheard of these days for airliners to crash like this. And the coincidence of both of them being this brand new model is more than enough of a data point. This whole affair stinks to high heaven.

Boeing actually just said this about their software fix:

It “took until now because we wanted to get it right. Rushing it is the wrong thing to do.”

Right.  Because they would never rush something to market, and go with a kludge instead of a redesign, just to beat Airbus.

 

[billvon 2,395]

4 hours ago, baronn said:

Only talks about the time they put into re-developing this aircraft. Not a single mention about the instability problems with adding the new motors. Airbus designed the Neos to accommodate these. The 737 never did. 

To be fair, the Neo aircraft were made to fly like the older A320's via software, just like Boeing tried to do.  And in both cases, the attempt to force that similarity was so they could justify not retraining pilots.  The difference is that Airbus has relied on software for stability for decades so it really was nothing new.  For Boeing it was a new system, and they got it wrong.

[gowlerk 1,905]

13 hours ago, billvon said:

To be fair, the Neo aircraft were made to fly like the older A320's via software, just like Boeing tried to do.  And in both cases, the attempt to force that similarity was so they could justify not retraining pilots.  The difference is that Airbus has relied on software for stability for decades so it really was nothing new.  For Boeing it was a new system, and they got it wrong.

I'm looking for information on those Airbus adjustments. I can only find vague references to the fly by wire system needing to be "re-tuned". It is really hard to know just how much the basic flight characteristics of either one need to be compensated for by computers. It's clear that the re-engine of the Airbus product also affected it's flight, but not what aspects were affected.

[billvon 2,395]

2 hours ago, gowlerk said:

I'm looking for information on those Airbus adjustments. I can only find vague references to the fly by wire system needing to be "re-tuned". It is really hard to know just how much the basic flight characteristics of either one need to be compensated for by computers. It's clear that the re-engine of the Airbus product also affected it's flight, but not what aspects were affected.

Right.  Could be small, could be large.  My point was that the Airbus flight control system ALREADY "retunes" the aircraft so that all their aircraft fly about the same.  And when everything is working it does that very well.  When it has to revert to a simpler control law due to lost sensors or instruments, it can have problems caused by pilots not used to how the aircraft flies without computer assistance.  (i.e. AF447.)

It's also worth noting that the 777 has a similar system, although it "appears" to the pilot to be a regular direct-law Boeing system.  However, when pitch/roll/load limits are exceeded, the pilot feels an increasing pressure that tries to correct the out-of-limits condition.  In the case of the 777 system, a design goal was that the pilot can always overpower the computer (ironically.)

It's likely that had this been implemented on a 777 this would have been a non-issue, since it would be a slight modification to a system that was already doing something very similar - and which was designed to be overridable by muscle power alone.

[gowlerk 1,905]

Ralph Nader, yes, that Ralph Nader (still alive) is weighing in and calling for a clean sheet redesign and for the Max to be discontinued. Apparently he has a personal connection to one of the crash victims.

 

https://www.npr.org/2019/04/04/709999296/ralph-nader-calls-for-recall-of-boeing-737-max-jets-after-his-grandniece-was-kil

[ryoder 1,381]

Scroll down to the 33-page preliminary report: https://www.cnn.com/2019/04/04/africa/ethiopian-airlines-full-preliminary-crash-report-intl/index.html

 

[SethInMI 145]

That is really detailed report.

On that website I could not get good resolution on the charts and graphs at the end, but i found the actual pdf and attached it, and you can see there was some crazy shit going on there. They had the stab trim set to cutout as they were supposed to, then had trouble manually trimming and then managed to do some manual trims then 5 sec after a manual trim an an auto-trim movement pushed the nose way down and cratered the plane. 

So I guess the thought is they turned the trim back on to try and trim it manually up, and then the auto-trim kicked in and pushed it down. But they could have used a trim wheel to trim it up instead of the switches? 

prelim 737 max.pdf

[billvon 2,395]

6 hours ago, ryoder said:

Scroll down to the 33-page preliminary report: https://www.cnn.com/2019/04/04/africa/ethiopian-airlines-full-preliminary-crash-report-intl/index.html

 

Hmm.  The surprising thing there is the failure of manual pitch trim.  Per the CVR, the pilots were apparently not able to trim the nose back up after they shut down the pitch trim motors, even with the manual trim wheel.  That could lead them to turning the pitch trim motors back on (to get the nose trimmed up via the thumb switch) which of course gave the MCAS the chance to pitch it right back down again.

[PhreeZone 15]

I believe that the manual trim can only be done below certain airspeed, above it the forces are too high to allow for manual inputs since you are fighting against the forces on the tail.

[ryoder 1,381]

19 minutes ago, PhreeZone said:

I believe that the manual trim can only be done below certain airspeed, above it the forces are too high to allow for manual inputs since you are fighting against the forces on the tail.

 

https://theaircurrent.com/aviation-safety/vestigal-design-issue-clouds-737-max-crash-investigations/

And since I can't copy/paste from that page, a very interesting point is attached as an image.

BTW "Mentour Pilot" is a Swedish airline pilot / instructor, and I've watched some of his videos recently. His Youtube channel is here:

https://www.youtube.com/channel/UCwpHKudUkP5tNgmMdexB3ow

 

Edited April 5, 2019 by ryoder

[pchapman 261]

That issue of trim and control column forces has long been a background issue for some aircraft in general. You don't want the trim to be effective both slow and fast, but then it is easy for it to have too much authority at higher speed. So it might overcome full back elevator, or cause pitch forces to be too high to realistically hold and fight against the trim. A small plane might have just a small trim tab that's easy to counteract, but on airliner jets with large speed ranges the whole stabilizer (with the elevators attached) moves, so it can have a lot of authority.

If an aircraft is trimmed nose low and picking up speed, to lighten the trimming, one would have to unload the stabilizer, such as by slowing down, or by unloading the elevator, by not pulling back hard on it. But that's just what one can't do easily when there's too much speed or too low.

Looks like they might have had to get on the issue quite quickly to avoid getting into the trouble they did. But it'll take some careful reading of the flight recorder traces, and understanding of control column and trim wheel forces, to sort this all out.

I'm starting to wonder if Boeing had really simmed this very well after the first accident. When alarms are going off, and one stick is shaking because of the messed up AoA sensor, real world pilot reactions aren't going to be instant. We're still caught up with this whole issue of it on the one hand being a really messy situation, but on the other one should supposedly be able to get ahead of it by killing the electric trim ... if one can do that before the stabilizer trim & flight angle get to be too bad.

And what the heck is going on with Boeing's AoA sensors & associated electrical system? They had confidence in it during the design phase, and the sensors are by Rosemount, who have built AoA sensors for a long time for many aircraft. Although I have no idea about any associated black boxes, it would likely be an electrical failure somewhere in the Boeing & Rosemount & any other supplier's system, not the actual vane on the outside of the plane somehow jamming in a weird position. With a large fleet of aircraft out there, does one really expect a failure in the fleet every 6 months?  It would still be bad enough on non Max 737's, where it could presumably still cause a stick shaker to go off because of the bad data.

Going by the graphs in the prelim report:

When the Ethiopian Airlines pilots lost the plane, they really lost it. Although they had some altitude loss early in the flight when the emergency started (and had Ground Proximity Warning System alerts), they fought it and then managed a slow climb for a long time. Then they got one last automatic nose down trim of maybe 4 seconds, which despite hauling back on the sticks even further than during rest of the flight, pitched the nose from above to below the horizon. There were no more pilot input electric pitch trims, and the thing just nosed over -- they lost their roughly 5000' above ground in 15 or 20 seconds, pitching down to -2g's and 45 degrees dive.

I'll have to read the rest of the report later.

[ryoder 1,381]

WaPo: Additional software problem detected in Boeing 737 Max flight control system, officials say

https://www.washingtonpost.com/world/africa/ethiopia-says-pilots-performed-boeings-recommendations-to-stop-doomed-aircraft-from-diving-urges-review-of-737-max-flight-control-system/2019/04/04/3a125942-4fec-11e9-bdb7-44f948cc0605_story.html

But later Thursday, Boeing confirmed to The Washington Post that it had found a second software problem that the Federal Aviation Administration has ordered fixed — separate from the anti-stall system that is under investigation in the two crashes and is involved in the worldwide grounding of the aircraft. That additional problem pertains to software affecting flaps and other flight-control hardware and is therefore classified as critical to flight safety, said two officials with knowledge of the investigation.

[pchapman 261]

1. The "second problem found" - The WaPo article has Boeing saying it is minor and easily fixed. Well, obviously we'd like a second opinion.

2. I've looked more through the preliminary report on the Ethiopia crash.

It looks like the crew turned the electric trim system back on. That's weird.

They initially had the problems, discussed turning off the electric trim (Stab Cut-out switches), and did so. MCAS even tried to trim nose low using the electric trim after that point, but nothing happened to the actual trim because the electric trim was disabled. Good, they had followed procedures!

They they keep flying OK for quite a while. Although it would be a high stress situation with back pull on the controls, stick shaker, and later overspeed clackers too for a small overspeed (since they left a lot of power on from the early climb).

There's a discussion where the first officer says manual trim not working --  But did he actually try to use the wheel and not try to mash the trim button on the electric system that wasn't working now? Was the 360hr pilot really familiar with forces that could be on the manual wheel? Did he actually fold out the handle to increase leverage?

Anyway, over a minute later something unexpected happens, that could only happen if the electric trim were turned back on, assuming no bizarre electrical faults we don't know about)

The data shows a couple very short upward trim activations of the electrical trim buttons. The actual pitch trim position is hard to see on the graph, but it looks like it moved up a little -- but only a little, as the activations were very short. Then the MCAS kicks into effect one last time, trimming significantly more nose down. That's when the pilots lose effective control and nose over. 

So someone, without discussing it in the cockpit, turned electrical trim back ON, but only used it to make a couple short and ineffective up trims.

I figure the first officer tried to fiddle with things, and turned the trim back on without discussing it, made short useless up trim activations, didn't immediately turn electric trim back off, so they then lost control as MCAS kicked in again with a bunch of down trim.

That's my best interpretation with current data, although who knows who actually did what.

Was the electrical pitch having trouble getting the stabilizer to actually pitch, given the air loads at speed with columns pulled back? Who knows but the aircraft was flying OK, not the much different from earlier when trim worked fine. And the trim attempts were very short, too short to have much effect.

Although the Airworthiness Directive from last time mentions keeping the Cut-out switches off for the rest of the flight, I could see that turning electric trim back on temporarily might be a way to help the situation:  Turn the electric trim back on, hold electric trim UP until control forces are neutral again, and select the Cut-out switches again to stop MCAS from then pitching the plane down again.

 

 

 

 

 

 

 

 

 

 

[SethInMI 145]

4 hours ago, pchapman said:

So someone, without discussing it in the cockpit, turned electrical trim back ON, but only used it to make a couple short and ineffective up trims.

Yeah, that is the feeling of the commentators that I have read. I can see it happen, and it would have been ok as I read it if they had just kept pushing the electric trim switch until the aircraft was flying level, then turned the stab trim off, as the MCAS won't run when the trim switches are pressed. But unfortunately it was not a sustained effort, more like a desperate "nothing is working, let's just try things"

The manual even says to do just that, use the trim switches to level the stab, then turn it off. But I can understand in that situation it would be really hard to think that through.

[gowlerk 1,905]

Boeing is starting to face reality.

https://ca.finance.yahoo.com/news/boeing-cuts-737-max-output-wake-two-deadly-012447162--finance.html

 

 

[ryoder 1,381]

Mentour Pilot just posted another video...to explain why his famous video about manual trim in the 737 Max was taken down.

This one is nearly an hour long, and I have not had time to watch it yet.

 

[Divalent 89]

6 hours ago, ryoder said:

Mentour Pilot just posted another video...to explain why his famous video about manual trim in the 737 Max was taken down.

This one is nearly an hour long, and I have not had time to watch it yet.

 

His explanation for taking the video down is in the first 6 minutes, the rest of the video is answering viewer questions submitted to him.

Short answer is that he realized he was speculating about the cause of an accident that was still under investigation.

(BTW, I find his videos go just as well at 1.5x speed)

[gowlerk 1,905]

More examples of short cut taking in the design process emerging. At the end of the clip is the first sign I've in the American mass media that the flaw in the basic design that made the MCAS necessary is being publicized. The world outside of the USA has been talking about it for a while now. This is likely to cause further erosion of public confidence in the MAX and lead to cancellation of more orders. 

 

https://www.cnn.com/videos/world/2019/04/04/boeing-analyst-737-max-griffin-dnt-lead-vpx.cnn/video/playlists/around-the-world/

[pchapman 261]

From an Aviation Week article, detailing some of the changes:   (my bolding)

Quote

[MCAS]

However, the system is only needed to enhance stability with slats and flaps retracted at very light weights and full aft center of gravity (CG). The aircraft exhibits sufficient natural longitudinal stability in all other parts of the flight envelope without the MCAS to meet the rules. Boeing emphasizes that the MCAS is not an anti-stall or stall-prevention system, as it often has been portrayed in news reports.

The new software load [P12.1] has triple-redundant filters that prevent one or both angle-of-attack (AOA) systems from sending erroneous data to the FCCs that could falsely trigger the MCAS. It also has design protections that prevent runaway horizontal stabilizer trim from ever overpowering the elevators. Boeing showed pilots that they can always retain positive pitch control with the elevators, even if they don’t use the left and right manual trim wheels on the sides of the center console to trim out control pressures after turning off the trim cut-out switches.

Most important, the MCAS now uses both left and right AOA sensors for redundancy, instead of relying on just one. The FCC P12.1’s triple AOA validity checks include an average value reasonability filter, a catastrophic failure low-to-high transition filter and a left versus right AOA deviation filter. If any of these abnormal conditions are detected, the MCAS is inhibited. 

Three secondary protections are built into the new software load. First, the MCAS cannot trim the stabilizer so that it overpowers elevator pitch control authority. The MCAS nose-down stab trim is limited so that the elevator always can provide at least 1.2g of nose-up pitch authority to enable the flight crew to recover from a nose-low attitude. Second, if the pilots make electric pitch trim inputs to counter the MCAS, it won’t reset after 5 sec. and repeat subsequent nose-down stab trim commands. And third, if the MCAS nose-down stab trim input exceeds limits programmed into the new FCC software, it triggers a maintenance message in the onboard diagnostics system.

They're also planning more sim sessions for pilots -- Getting practice with dealing with the sort of situation where the stick shaker is going off and the airspeed value is unreliable, making things a bit chaotic in the cockpit. (Airspeed gets somewhat wrong if the AOA is wrong, as the AOA on is used to correct smaller errors in the pitot tube readings.)

Some of the new memory items for pilots for situations like that are basic "power and pitch" rules. If you are unsure of airspeed, and your flaps and slats are retracted, then for example set the engines to 75% speed and pitch the nose up at a 4 degree angle. That should give a reasonable climb and airspeed whatever the aircraft weight.  The Airbuses also have stuff like that -- If some sensors are unreliable, there are rules of thumb to use. 

The Ethopian pilots, for example, were distracted enough that they kept the same takeoff power during the whole flight, which put the aircraft at the very high end of the airspeed envelope and even led to overspeed warnings, during flight prior to the final dive. The high speed also made the trim issues and forces on the control wheels worse. It would all have been more manageable at a more moderate speed.