VanBosch
Posts: 32
Joined: Mon Aug 14, 2017 5:18 pm

Re: Boeing 737MAX Grounded Worldwide Q2 2019

Thu Jun 27, 2019 9:25 pm

LX015 wrote:
In all seriousness, I wonder if IAG is now having 2nd thoughts about the LOI for 200 MAXs...


Why would they though? It’s only a LOI, even if the plane never flies again it means they have leverage with Airbus and it doesn’t cause them anything.

Upside is they might get a better deal with Airbus but if not what have they lost?
 
planecane
Posts: 1036
Joined: Thu Feb 09, 2017 4:58 pm

Re: Boeing 737MAX Grounded Worldwide Q2 2019

Thu Jun 27, 2019 9:51 pm

PW100 wrote:
planecane wrote:
PW100 wrote:
Is 3 seconds sufficient to recognize a runaway trim in a stand-alone failure scenario (leaving aside that usually other issues be present at the same time)?
If we look at MCAS, it has a 9 second cycle. Should pilots be expected to disable MCAS after only 3 seconds, or should they allow at least a full MCAS cyce, and take action at the next cycle (after 3 seconds)?
Leaving aside the question whether the intermittent behaviour of MCAS would be recognizable as runaway trim (prior to the accidents).


The expectation is/was that the pilots were expected to recognize a runaway stabilizer (caused by MCAS or otherwise) within 3 seconds and begin the NNC. That's why, to me, it has always been completely irrelevant that MCAS was intermittent. The big failure that was missed is that the same thing that would cause MCAS cause a runaway stabilizer ALSO caused other issues simultaneously. This factor is the reason that I believe the lack of documentation of MCAS and specific training played a huge factor. The training could have been part of the iPad course. Simple instructions that if there was uncommanded nose down trim that the runaway stabilizer NNC should be performed first, regardless of other issues would have gone a long way. The information should have been provided that other alerts (single side stick shaker, unreliable airspeed, etc.) would likely occur simultaneously but that responding to the runaway stabilizer should take priority. This should have been part of the training and the QRH.


I understand what you're saying with the intermittent behaviour.
But then it would make not much sense to schedule MCAS cycle for 9 seconds, if Boeing at the same time expected the crew to shut it down after 3 seconds . . .


The 9 seconds because the stabilizer can only move so fast. It took 9 seconds to move the amount that was determined was needed under a worst case scenario (I'm guessing low speed where the stab has less authority). The pilots aren't normally intended to turn it off.
 
RickNRoll
Posts: 1726
Joined: Fri Jan 06, 2012 9:30 am

Re: Boeing 737MAX Grounded Worldwide Q2 2019

Thu Jun 27, 2019 9:57 pm

OldAeroGuy wrote:
XRAYretired wrote:
Just imagine for a moment spending a few seconds finding the QRH which then tells you to repond within 3 seconds. No it doesnt, nor does the EID because it is ludicrous.

Ray


No need to find the QRH as the response to a Stab Trim Runaway is a memory item, as stated many times on this thread..
I don't think you understand the situation. The problem is that the pilots may do the right action and not see the plane respond quickly enough because the computer is overloaded. They will them think they have not taken the correct action and try something else.
 
User avatar
par13del
Posts: 8679
Joined: Sun Dec 18, 2005 9:14 pm

Re: Boeing 737MAX Grounded Worldwide Q2 2019

Thu Jun 27, 2019 10:01 pm

RickNRoll wrote:
No need to find the QRH as the response to a Stab Trim Runaway is a memory item, as stated many times on this thread..
I don't think you understand the situation. The problem is that the pilots may do the right action and not see the plane respond quickly enough because the computer is overloaded. They will them think they have not taken the correct action and try something else.[/quote]
So to speculate, the ET pilots understood it was MCAS, initiated the appropriate NNC but when the response was not immediate they moved on?
 
hivue
Posts: 1903
Joined: Tue Feb 26, 2013 2:26 am

Re: Boeing 737MAX Grounded Worldwide Q2 2019

Thu Jun 27, 2019 10:01 pm

PixelFlight wrote:
I don't think that the FAA "chose to leak", there are now expected to promptly communicate to the others administrations any important finding on the subject and consequently would loss credibility if there are not the first to announce the finding to the public.


So why did they leak it instead of making just such a public announcement? CNBC is implying that there appears to be stuff currently going on between Boeing and the FAA that we do not know about and that is not business as usual and the old way of doing things. See the link in my last post in the immediately previous part of this thread.
"You're sitting. In a chair. In the SKY!!" ~ Louis C.K.
 
Interested
Posts: 647
Joined: Thu May 19, 2016 12:19 pm

Re: Boeing 737MAX Grounded Worldwide Q2 2019

Thu Jun 27, 2019 10:10 pm

planecane wrote:
Interested wrote:

Great - get the NG back in production and everybody can feel safe again



Will you feel that way if it is determined that the same potential cause of runaway stabilizer (a microprocessor failure) exists on the NG as well?


Yes because NG doesn't have MCAS to contend with

It's a proven safe plane - the only reason anyone is even debating runaway stabilizer and pilots struggling with it is because of Max
 
User avatar
Revelation
Posts: 20953
Joined: Wed Feb 09, 2005 9:37 pm

Re: Boeing 737MAX Grounded Worldwide Q2 2019

Thu Jun 27, 2019 10:13 pm

PixelFlight wrote:
DO-178 define "structural coverage criteria". This document resume well enough the subject on page 12 https://pdfs.semanticscholar.org/ba18/3 ... 3c6f1d.pdf
"
structural coverage criteria have to be achieved depend on the software level:
• Level C: 100% statement coverage is required, which means that every statement in the
program has been exercised.
• Level B: 100% decision coverage is required. That means that every decision has taken all
possible outcomes at least one (ex: then/else for an if construct) and that every entry and exit
point in the program has been invoked at least one.
• 100% MCD/DC (Modified Condition/Decision Coverage) is required for level A software,
which means that:
o Every entry and exit point in the program has been invoked at least once
o Every decision has taken all possible outcomes
o Each condition in a decision has been shown to independently affect that decision’s
outcome (this is shown by varying just that condition while holding fixed all other
possible conditions).
"

"Every line executed" != "All logic verified correct for all range of inputs" and/or "all real-time requirements met".
Wake up to find out that you are the eyes of the world
The heart has its beaches, its homeland and thoughts of its own
Wake now, discover that you are the song that the morning brings
The heart has its seasons, its evenings and songs of its own
 
User avatar
PixelFlight
Posts: 508
Joined: Thu Nov 08, 2018 11:09 pm

Re: Boeing 737MAX Grounded Worldwide Q2 2019

Thu Jun 27, 2019 10:34 pm

Revelation wrote:
PixelFlight wrote:
DO-178 define "structural coverage criteria". This document resume well enough the subject on page 12 https://pdfs.semanticscholar.org/ba18/3 ... 3c6f1d.pdf
"
structural coverage criteria have to be achieved depend on the software level:
• Level C: 100% statement coverage is required, which means that every statement in the
program has been exercised.
• Level B: 100% decision coverage is required. That means that every decision has taken all
possible outcomes at least one (ex: then/else for an if construct) and that every entry and exit
point in the program has been invoked at least one.
• 100% MCD/DC (Modified Condition/Decision Coverage) is required for level A software,
which means that:
o Every entry and exit point in the program has been invoked at least once
o Every decision has taken all possible outcomes
o Each condition in a decision has been shown to independently affect that decision’s
outcome (this is shown by varying just that condition while holding fixed all other
possible conditions).
"

"Every line executed" != "All logic verified correct for all range of inputs" and/or "all real-time requirements met".

Please understand that flight control software require level A per DO-178: 100% MCD/DC (Modified Condition/Decision Coverage)
 
User avatar
PixelFlight
Posts: 508
Joined: Thu Nov 08, 2018 11:09 pm

Re: Boeing 737MAX Grounded Worldwide Q2 2019

Thu Jun 27, 2019 10:37 pm

hivue wrote:
PixelFlight wrote:
I don't think that the FAA "chose to leak", there are now expected to promptly communicate to the others administrations any important finding on the subject and consequently would loss credibility if there are not the first to announce the finding to the public.


So why did they leak it instead of making just such a public announcement? CNBC is implying that there appears to be stuff currently going on between Boeing and the FAA that we do not know about and that is not business as usual and the old way of doing things. See the link in my last post in the immediately previous part of this thread.

I have to agree with you: the FAA did not communicate on this issue. Boeing did it by given credit to the FAA. Strange.

EDIT: In fact the FAA did communicate: https://www.faa.gov/news/updates/?newsId=93206
"On the most recent issue, the FAA’s process is designed to discover and highlight potential risks. The FAA recently found a potential risk that Boeing must mitigate. "
Last edited by PixelFlight on Thu Jun 27, 2019 10:42 pm, edited 1 time in total.
 
User avatar
7BOEING7
Posts: 3038
Joined: Fri Oct 12, 2012 5:28 pm

Re: Boeing 737MAX Grounded Worldwide Q2 2019

Thu Jun 27, 2019 10:40 pm

kalvado wrote:
OldAeroGuy wrote:
kalvado wrote:

But what exactly are the symptoms to be determined within 3 seconds? Sts is another source of trim movement without direct pilot command, and it happens often. So how to ensure response to actual runaway without many false alarms?


As Mentor Pilot says in his video, STS is making control forces easier. Runaway Stab is making it more difficult to control the airplane. If you're having to hold an increasing amount of uncommanded column force, it's a Runaway Stab.

Watch the video:

https://video.search.yahoo.com/yhs/sear ... tion=click

That is a great symptom, but is it actually detectable within 3 seconds and differentiated from turbulence offset or gradient of force in maneuver?
I can imagine 3 seconds response to engine warning sound on takeoff, but I suspect force feeling is largely a closed loop which can be consciously understood only when things are very wrong. I know where I want my car to go, and I don't really notice small glitches like a piece of stone under the wheel. It takes curb or a good pothole to start noticing


When you're hand flying the 737 (if the airplane is in "trim") you'll notice immediately just like a pothole. With the autopilot is flying, if somebody is paying attention, might be a couple of seconds --if you're asleep at the wheel when the autopilot disconnect warning goes off you'll figure it out.
 
Thorkel
Posts: 21
Joined: Mon Dec 15, 2014 2:38 pm

Re: Boeing 737MAX Grounded Worldwide Q2 2019

Thu Jun 27, 2019 10:46 pm

PixelFlight wrote:
BoeingVista wrote:
I disagree with that assessment its quite possible that Boeing has just run out of processing power for the amount of tasks the computer is required to perform.

Freefall wrote:
I think they will need change both (software and hardware) upgrades probably needed because now MCAS software needs to deal with more situations/scenarios and for that more processing power and memory will be needed.

RickNRoll wrote:
The problem is that the processor is an old one that does not appear to have enough power to run more complex time critical calculations.

StTim wrote:
I suspect that what happened was processor overload rather than BSOD type failure.
The "out of processing power" is very hard to believe in the context of a DO-178 compliant software development, especially for the transition from MCAS v1 to MCAS v2. From what have been released so far, the differences between the MCAS v1 and MCAS v2 is just a couple of relatively simple conditions. Even on a 30 years old processors those condition will not add more than a fraction of millisecond of processing time. It's not possible to properly certify for flight operation a computer + software combo with a such insanely small processing margin time.


One of the main reasons why these sort of old processors are used are where you need to argue about hard real-time guarantees. Modern processors, with multi-level cache hierarchies, Out Of Order Execution, branch prediction, etc can be quite difficult to actually build an analysis around a Worst Case Execution Time. Typically they’ll execute very quickly for most things, but the worst case time can be bad and for a hard real-time system thats an issue. It’s one reason why you still see lots of 68XXX, PPC 60X series or early intel 80x86 series processes in flight control systems for both civil and military aircraft. They’re slow, but they’re predictable.

The other thing, in terms of timeliness, is that if these chips are associated with real-time constraints is they won’t be running anything approaching a ‘normal’ OS. They’ll be no pre-emotive multitasking or anything like that (again, doesn’t help for worst case execution times). It’ll be single task, or you’ll have something like a cyclic scheduler where each application has a fixed share of processor time. If it is a cyclic scheduler, there may be options to increase the ‘share’ given to MCAS - there may be unused capacity, or they may be able to reduce the share given to other apps on the processor.
 
OldAeroGuy
Posts: 3870
Joined: Sun Dec 05, 2004 6:50 am

Re: Boeing 737MAX Grounded Worldwide Q2 2019

Thu Jun 27, 2019 10:49 pm

aerolimani wrote:
The way I read this section is that it is defining the power and capability that a trim system must have. To my reading, it's determining how much capability the system needs to have so that the pilot can achieve a fast enough response from the system. In other words, the 3 seconds is defining what the system needs to be capable of, not what the pilot's response time needs to be. This section also defines how the forces should not behave erratically as the system moves through its trim degrees.

I will post here the opening explanation for the section you are quoting. Not anywhere in this whole section, and certainly not in the explanation, is there mention of the actions of an automatic trim system, or its expected characteristics as regards pilot response time. I think it's important to quote the section's opening explanation, as it gives context to everything else contained in the section.

33. Out-of-Trim Characteristics - § 25.255.
a. Explanation. Certain early, trimmable stabilizer equipped jet transports experienced “jet upsets” that resulted in high speed dives. When the airplane was mistrimmed in the nose-down direction and allowed to accelerate to a high airspeed, it was found that there was insufficient elevator power to recover. Also, the stabilizer could not be trimmed in the nose-up direction, because the stabilizer motor stalled due to excessive airloads imposed on the horizontal stabilizer. As a result, a special condition was developed and applied to most part 25 airplanes with trimmablestabilizers. With certain substantive changes, it was adopted as § 25.255, effective with amendment 25-2. While these earlier problems seem to be generally associated with airplanes having trimmable stabilizers, it is clear from the preamble discussions to amendment 25-42 that § 25.255 applies “regardless of the type of trim system used in the airplane.” Section 25.255 is structured to give protection against the following unsatisfactory characteristics during mistrimmed flight in the higher speed regimes:

(1) Changes in maneuvering stability leading to overcontrolling in pitch.
(2) Inability to achieve at least l.5g for recovery from upset due to excessive control forces.
(3) Inability of the flightcrew to apply the control forces necessary to achieve recovery.
(4) Inability of the pitch-trim system to provide necessary control force relief when high control force inputs are present.


The FAA's language above is talking about the pilot's ability to recover from a runaway of the trim system, in either the nose up or the nose down direction. I'm not sure of the sense of your quoted sentence below

"The way I read this section is that it is defining the power and capability that a trim system must have."

The FAR 25.255 requires that the pilot be able to overcome the impact of a runaway stabilizer and the state it will leave the airplane in after a 3 second runaway. For instance, if the airplane is at Vmo/Mmo at the moment that the stabilizer runaway begins and the pilot takes no action for 3 sec and then begins to control the airplane via the elevators, the airplane must be able to recover from the overspeed and not exhibit any unusual flight control behaviors.

To make sure the airplane can meet this requirement, trim speed might need to be set at a low rate to make sure the airplane isn't badly out of trim after 3 sec. Is that what you mean?

By the way, here's the exact FAR language:

Sec. 25.255

[Out-of-trim characteristics.]

[(a) From an initial condition with the airplane trimmed at cruise speeds up to VMO/MMO, the airplane must have satisfactory maneuvering stability and controllability with the degree of out-of-trim in both the airplane nose-up and nose-down directions, which results from the greater of--
(1) A three-second movement of the longitudinal trim system at its normal rate for the particular flight condition with no aerodynamic load (or an equivalent degree of trim for airplanes that do not have a power-operated trim system), except as limited by stops in the trim system, including those required by Sec. 25.655(b) for adjustable stabilizers; or
(2) The maximum mistrim that can be sustained by the autopilot while maintaining level flight in the high speed cruising condition.

The 3 sec is in (a)(1)
Airplane design is easy, the difficulty is getting them to fly - Barnes Wallis
 
OldAeroGuy
Posts: 3870
Joined: Sun Dec 05, 2004 6:50 am

Re: Boeing 737MAX Grounded Worldwide Q2 2019

Thu Jun 27, 2019 11:05 pm

RickNRoll wrote:
OldAeroGuy wrote:
XRAYretired wrote:
Just imagine for a moment spending a few seconds finding the QRH which then tells you to repond within 3 seconds. No it doesnt, nor does the EID because it is ludicrous.

Ray


No need to find the QRH as the response to a Stab Trim Runaway is a memory item, as stated many times on this thread..


I don't think you understand the situation. The problem is that the pilots may do the right action and not see the plane respond quickly enough because the computer is overloaded. They will them think they have not taken the correct action and try something else.


I'm unclear about your statement. Isn't the microprocessor problem related to MCASv2.0? And how would it impact manual electric stab trim and the stab cutoff switches on prior versions of the MAX?
Airplane design is easy, the difficulty is getting them to fly - Barnes Wallis
 
User avatar
tenHangar
Posts: 15
Joined: Tue Jul 11, 2017 2:39 pm

Re: Boeing 737MAX Grounded Worldwide Q2 2019

Fri Jun 28, 2019 12:10 am

SteinarN wrote:
From Aviation Week

It seems in fact like the latest problem is related to FCC overload. The article says the THS is slow to react to manual trim inputs from the thumb switches when simulating a trim runaway. Thereby forcing the aircraft into a steep dive before the THS finally starts reacting to the thumb switches. It is not clear if this problem can be fixed by a software update or if it requires time consuming hardware changes.....
so... this is separate from the the MCAS software issue?
 
hivue
Posts: 1903
Joined: Tue Feb 26, 2013 2:26 am

Re: Boeing 737MAX Grounded Worldwide Q2 2019

Fri Jun 28, 2019 12:24 am

PixelFlight wrote:
In fact the FAA did communicate:


That communication communicates nothing. The FAA had to say something officially after the preliminary details were leaked -- apparently by one of their own.
"You're sitting. In a chair. In the SKY!!" ~ Louis C.K.
 
User avatar
aerolimani
Posts: 1134
Joined: Tue Jun 18, 2013 5:46 pm

Re: Boeing 737MAX Grounded Worldwide Q2 2019

Fri Jun 28, 2019 1:24 am

OldAeroGuy wrote:
aerolimani wrote:
The way I read this section is that it is defining the power and capability that a trim system must have. To my reading, it's determining how much capability the system needs to have so that the pilot can achieve a fast enough response from the system. In other words, the 3 seconds is defining what the system needs to be capable of, not what the pilot's response time needs to be. This section also defines how the forces should not behave erratically as the system moves through its trim degrees.

I will post here the opening explanation for the section you are quoting. Not anywhere in this whole section, and certainly not in the explanation, is there mention of the actions of an automatic trim system, or its expected characteristics as regards pilot response time. I think it's important to quote the section's opening explanation, as it gives context to everything else contained in the section.

33. Out-of-Trim Characteristics - § 25.255.
a. Explanation. Certain early, trimmable stabilizer equipped jet transports experienced “jet upsets” that resulted in high speed dives. When the airplane was mistrimmed in the nose-down direction and allowed to accelerate to a high airspeed, it was found that there was insufficient elevator power to recover. Also, the stabilizer could not be trimmed in the nose-up direction, because the stabilizer motor stalled due to excessive airloads imposed on the horizontal stabilizer. As a result, a special condition was developed and applied to most part 25 airplanes with trimmablestabilizers. With certain substantive changes, it was adopted as § 25.255, effective with amendment 25-2. While these earlier problems seem to be generally associated with airplanes having trimmable stabilizers, it is clear from the preamble discussions to amendment 25-42 that § 25.255 applies “regardless of the type of trim system used in the airplane.” Section 25.255 is structured to give protection against the following unsatisfactory characteristics during mistrimmed flight in the higher speed regimes:

(1) Changes in maneuvering stability leading to overcontrolling in pitch.
(2) Inability to achieve at least l.5g for recovery from upset due to excessive control forces.
(3) Inability of the flightcrew to apply the control forces necessary to achieve recovery.
(4) Inability of the pitch-trim system to provide necessary control force relief when high control force inputs are present.


The FAA's language above is talking about the pilot's ability to recover from a runaway of the trim system, in either the nose up or the nose down direction. I'm not sure of the sense of your quoted sentence below

"The way I read this section is that it is defining the power and capability that a trim system must have."

The FAR 25.255 requires that the pilot be able to overcome the impact of a runaway stabilizer and the state it will leave the airplane in after a 3 second runaway. For instance, if the airplane is at Vmo/Mmo at the moment that the stabilizer runaway begins and the pilot takes no action for 3 sec and then begins to control the airplane via the elevators, the airplane must be able to recover from the overspeed and not exhibit any unusual flight control behaviors.

To make sure the airplane can meet this requirement, trim speed might need to be set at a low rate to make sure the airplane isn't badly out of trim after 3 sec. Is that what you mean?

By the way, here's the exact FAR language:

Sec. 25.255

[Out-of-trim characteristics.]

[(a) From an initial condition with the airplane trimmed at cruise speeds up to VMO/MMO, the airplane must have satisfactory maneuvering stability and controllability with the degree of out-of-trim in both the airplane nose-up and nose-down directions, which results from the greater of--
(1) A three-second movement of the longitudinal trim system at its normal rate for the particular flight condition with no aerodynamic load (or an equivalent degree of trim for airplanes that do not have a power-operated trim system), except as limited by stops in the trim system, including those required by Sec. 25.655(b) for adjustable stabilizers; or
(2) The maximum mistrim that can be sustained by the autopilot while maintaining level flight in the high speed cruising condition.

The 3 sec is in (a)(1)

I now understand what is actually happening in this discussion. The original document shared is this one, which is what I was reading: AC 25-7C Flight Test Guide For Certification Of Transport Category Airplanes. https://www.faa.gov/documentLibrary/med ... -7C%20.pdf

I now realize that you are quoting the PART 25—AIRWORTHINESS STANDARDS: TRANSPORT CATEGORY AIRPLANES, as found here: https://www.ecfr.gov/cgi-bin/text-idx?n ... 1.0.1.3.11.

However, I do maintain that a pilot reaction time of 3 seconds is still just an extrapolation from this text. It's rather a causation vs correlation situation. One could write a guideline such as this, based on a 3-second reaction time. However, there is no definite path leading the other direction, from reading this guideline to assuming a 3-second reaction time.

Until someone can show me something in official certification documentation which actually says that pilots are expected to respond to a runaway stabilizer within 3 seconds, then I am going to assume this 3-second concept was Boeing's. I feel that it was a very poor choice to design MCAS around the standard of a 3-second reaction.

As with my TCAS example, I feel that a 3-second reaction is reasonable, in response to the direct instruction of a visual and aural command. As to runaway stabilizer, a situation which first requires diagnosis, especially if it comes with a host of other alarms, I think 3 seconds is inappropriate. It's somewhat mind-boggling to me that Boeing would design around a 3-second reaction time.
 
kalvado
Posts: 1814
Joined: Wed Mar 01, 2006 4:29 am

Re: Boeing 737MAX Grounded Worldwide Q2 2019

Fri Jun 28, 2019 1:45 am

7BOEING7 wrote:
kalvado wrote:
OldAeroGuy wrote:

As Mentor Pilot says in his video, STS is making control forces easier. Runaway Stab is making it more difficult to control the airplane. If you're having to hold an increasing amount of uncommanded column force, it's a Runaway Stab.

Watch the video:

https://video.search.yahoo.com/yhs/sear ... tion=click

That is a great symptom, but is it actually detectable within 3 seconds and differentiated from turbulence offset or gradient of force in maneuver?
I can imagine 3 seconds response to engine warning sound on takeoff, but I suspect force feeling is largely a closed loop which can be consciously understood only when things are very wrong. I know where I want my car to go, and I don't really notice small glitches like a piece of stone under the wheel. It takes curb or a good pothole to start noticing


When you're hand flying the 737 (if the airplane is in "trim") you'll notice immediately just like a pothole. With the autopilot is flying, if somebody is paying attention, might be a couple of seconds --if you're asleep at the wheel when the autopilot disconnect warning goes off you'll figure it out.

How would you describe speed trim on the same scale?
If you will, the question is if it feels different (speed? duration?); and runaway has 50% chance to be in the same direction as automatic trim anyway.
I assume it has to be felt to some extent; so it may increase the threshold for "trim behaves strange" feeling
 
ltbewr
Posts: 14179
Joined: Thu Jan 29, 2004 1:24 pm

Re: Boeing 737MAX Grounded Worldwide Q2 2019

Fri Jun 28, 2019 1:50 am

I wonder if the way the certain controls on the 737 MAX series were set up, like the stab trim, were intentionally designed to prevent excessive and sudden movements to prevent disaster. For example, the 11/13/2001 flight AA587 where the pilots used excessive rudder controls after an encounter with wake turbulence shortly after takeoff causing the vertical stabilizer to break off. That crash led to retraining of pilots and adjustments to rudder controls to limit excessive inputs.
 
User avatar
QuarkFly
Posts: 368
Joined: Mon Aug 15, 2016 4:20 pm

Re: Boeing 737MAX Grounded Worldwide Q2 2019

Fri Jun 28, 2019 2:17 am

Thankfully this Frankenstein-ish stunted MAX aircraft is finally getting the scrutiny it deserves!

No mass produced airliner that will be flying in 2050 should have patched kludges like MCAS and force pilots to turn off perfectly working electric trim to deal with an anarchic manual trim wheel because of software or sensor problems. And now, when finally more justified analysis of the MAX is occurring, even more problems are smoked out...no surprise.

Thankfully international regulators beyond the lapdog FAA are also on the case here. Boeing should commit to replace the MAX before 2030 with an all new Fly-By-Wire narrow-body. They are flush with cash and stock-buybacks can be put on hold for a decade. And if MAX need to stay on the ground for a year or two to fix things correctly...so be it.

Shame on Boeing...the whole MAX exercise increasingly looks like corporate malfeasance.
Always take the Red Eye if possible
 
User avatar
BoeingVista
Posts: 1991
Joined: Mon Jan 12, 2009 9:54 am

Re: Boeing 737MAX Grounded Worldwide Q2 2019

Fri Jun 28, 2019 4:04 am

Boeing and FAA initially disagreed on severity of "catastrophic" 737 software glitch.

https://theaircurrent.com/aviation-safe ... re-glitch/

So Boeing still don't really get the meaning of "safety is our first priority", it means not releasing a software fix with another deadly flaw, or minimising deadly flaws in the MAX by redesignating the risk as lower than its actual effect on the flyability of the aircraft via paperwork. The FAA rated the "glitch" as catastrophic where as Boeing rated it two rungs lower as "major"; FAA stuck to their guns for once and Boeing fell into line.

When fixing an aircraft you don't want to design in the next crash


“An average line pilot wouldn’t know what the hell they’re doing,” said one of the people, who said the uncommanded movement was not an erroneous activation of the MCAS function, which has been at the center of the twin crash investigations. The FAA’s view ultimately won the day. Boeing in a statement said it “agrees with the FAA’s decision and request, and is working on the required software.”

But the fix is expected to add months to the jet’s return to service, clouding an already complicated outlook for the 737 Max.
BV
 
zhetenyi1973
Posts: 5
Joined: Thu Apr 04, 2019 7:08 pm

Re: Boeing 737MAX Grounded Worldwide Q2 2019

Fri Jun 28, 2019 4:19 am

StTim wrote:
The other thing, in terms of timeliness, is that if these chips are associated with real-time constraints is they won’t be running anything approaching a ‘normal’ OS. They’ll be no pre-emotive multitasking or anything like that (again, doesn’t help for worst case execution times). It’ll be single task, or you’ll have something like a cyclic scheduler where each application has a fixed share of processor time. If it is a cyclic scheduler, there may be options to increase the ‘share’ given to MCAS - there may be unused capacity, or they may be able to reduce the share given to other apps on the processor.


This is simply not true. It looks like you never heard of OSE which is an Real-Time OS and even Linux is capable to run RT applications. Of course you have to patch the kernel for it.
In an RT OS you can define hard priority between processes meaning a higher priority process is scheduled immediately if it wants to run. Of course you can starve out lower priority
processes if you are not careful. You can even have higher priority processes than interrupts, meaning the kernel will note the interrupt, but the handler code will be run later.
Most probably something similar happened here, so the solution is simply to increase the priority of the process which handles that particular data. Or the message dispatcher has to
prioritize these kind of data/messages over other data/messages in case there is only a simple loop in the code.
 
DenverTed
Posts: 243
Joined: Wed Mar 27, 2019 11:12 pm

Re: Boeing 737MAX Grounded Worldwide Q2 2019

Fri Jun 28, 2019 4:37 am

So this isn't MCAS, it is something in the MAX software, which is different from the NG software. And the manual trim wheel recovery is free and clear for the NG and the MAX, or where does the FAA stand on that right now? I guess the NG is flying, so they must not be too concerned with the safety of the manual trim wheel.
 
User avatar
767333ER
Posts: 959
Joined: Wed Jun 15, 2016 5:14 am

Re: Boeing 737MAX Grounded Worldwide Q2 2019

Fri Jun 28, 2019 5:23 am

I think it’s really pathetic at this point that some people still try to stick up for Boeing and their botched plane. I get some people like Boeing and the 737 and yeah the 737NG has quite the positive track record, but face the music this thing is a piece of junk, at this point there’s no two ways about it! I get that they may fix it ant it may be fine one day, but right now as it is junk is really the best way to describe it.
Been on: 732 733 734 73G 738 752 763 A319 A320 A321 CRJ CR7 CRA/CR9 E145 E175 E190 F28 MD-82 MD-83 C172R C172S P2006T
 
14ccKemiskt
Posts: 85
Joined: Fri Nov 26, 2010 10:46 pm

Re: Boeing 737MAX Grounded Worldwide Q2 2019

Fri Jun 28, 2019 6:22 am

PixelFlight wrote:
EDIT: In fact the FAA did communicate: https://www.faa.gov/news/updates/?newsId=93206
"On the most recent issue, the FAA’s process is designed to discover and highlight potential risks. The FAA recently found a potential risk that Boeing must mitigate. "


It is important to the FAA that THEY discovered something. They are under severe pressure as well and they really want to show that their own processes are able to find anomalies.

I wonder how the relationship between Boeing and FAA really is nowadays. Are they just maintaining a facade of beeing master and servant now or are they behind closed doors still in that "coziness" together?
 
RickNRoll
Posts: 1726
Joined: Fri Jan 06, 2012 9:30 am

Re: Boeing 737MAX Grounded Worldwide Q2 2019

Fri Jun 28, 2019 6:23 am

DenverTed wrote:
So this isn't MCAS, it is something in the MAX software, which is different from the NG software. And the manual trim wheel recovery is free and clear for the NG and the MAX, or where does the FAA stand on that right now? I guess the NG is flying, so they must not be too concerned with the safety of the manual trim wheel.
I am guessing that is why some one in the FAA leaked this. They didn't like the way Boeing was able to get away with this without anyone knowing what they have been doing behind everyone's back.
 
oschkosch
Posts: 198
Joined: Mon Oct 29, 2018 3:41 pm

Re: Boeing 737MAX Grounded Worldwide Q2 2019

Fri Jun 28, 2019 6:24 am

Maybe somebody can do a total summary of all currently known problems when we start the Q3 Thread? Would definitely be useful, and it is only 3 days until Quarter 3 starts :-)

Gesendet von meinem SM-G950F mit Tapatalk
 
marcelh
Posts: 648
Joined: Wed Jun 19, 2013 12:43 pm

Re: Boeing 737MAX Grounded Worldwide Q2 2019

Fri Jun 28, 2019 6:52 am

BoeingVista wrote:
Boeing and FAA initially disagreed on severity of "catastrophic" 737 software glitch.

https://theaircurrent.com/aviation-safe ... re-glitch/

So Boeing still don't really get the meaning of "safety is our first priority", it means not releasing a software fix with another deadly flaw, or minimising deadly flaws in the MAX by redesignating the risk as lower than its actual effect on the flyability of the aircraft via paperwork. The FAA rated the "glitch" as catastrophic where as Boeing rated it two rungs lower as "major"; FAA stuck to their guns for once and Boeing fell into line.

When fixing an aircraft you don't want to design in the next crash


“An average line pilot wouldn’t know what the hell they’re doing,” said one of the people, who said the uncommanded movement was not an erroneous activation of the MCAS function, which has been at the center of the twin crash investigations. The FAA’s view ultimately won the day. Boeing in a statement said it “agrees with the FAA’s decision and request, and is working on the required software.”

But the fix is expected to add months to the jet’s return to service, clouding an already complicated outlook for the 737 Max.

Very disturbing to read Boeing is acting.
 
AirwayBill
Posts: 135
Joined: Fri Mar 29, 2019 8:37 am

Re: Boeing 737MAX Grounded Worldwide Q2 2019

Fri Jun 28, 2019 7:15 am

The "fixing" of the MAX at this point just seems like they are trying to fix portion of a steel bridge badly damaged by rust with nothing but duct tape and WD40. First, it simply does not work, and second, the whole bridge is at stake of collapsing.

The more I read about Boeing executives' reactions, the more I question their actual mental health and their ability to find a positive outcome to this situation. It also makes me wonder (since we opened this can of worms), what other designs could have been impacted so far by their extreme negligence and covered up effectively, with, luckily to them, no crash to unveil the possible hidden problems.

These are probably the darkest days of commercial aviation in a few decades, and the nightmare is not over.
 
User avatar
JerseyFlyer
Posts: 1369
Joined: Fri May 25, 2007 7:24 pm

Re: Boeing 737MAX Grounded Worldwide Q2 2019

Fri Jun 28, 2019 7:22 am

This seems to be a test of what happens when a microprocessor is broken, i.e. a test of adequate redundancy, which neither Boeing nor FAA tested for previously. Puts both in a bad light.
 
User avatar
cv990Coronado
Posts: 340
Joined: Thu Nov 08, 2007 4:38 pm

Re: Boeing 737MAX Grounded Worldwide Q2 2019

Fri Jun 28, 2019 7:47 am

Surely after all these developments, it will be necessary for the MAX to need a separate simulator?
SSC-707B727 737-741234SP757/762/3/772/WA300/10/319/2/1-342/3/6-880-DAM-VC10 TRD 111 Ju52-DC8/9/10/11-YS11-748-VCV DH4B L
 
xmp125a
Posts: 233
Joined: Mon Mar 11, 2019 6:38 pm

Re: Boeing 737MAX Grounded Worldwide Q2 2019

Fri Jun 28, 2019 8:02 am

SteinarN wrote:
From Aviation Week

It seems in fact like the latest problem is related to FCC overload. The article says the THS is slow to react to manual trim inputs from the thumb switches when simulating a trim runaway. Thereby forcing the aircraft into a steep dive before the THS finally starts reacting to the thumb switches.


Let's remember the cockpit recording from the ET crash. "Manual trim is not working". We debated ad nauseam whether they meant manual manual trim (wheel) or manual electric trim (switch).

This could be huge, if the problem was in the original MAX firmware, and was discovered only now. And is basically impossible to prove or disprove that the actual processor behaved this way when ET had problems trimming -, as in all real-time embedded systems. Sole fact that it exists is a disaster for Boeing (remember Toyota lawsuit?).
 
User avatar
BoeingVista
Posts: 1991
Joined: Mon Jan 12, 2009 9:54 am

Re: Boeing 737MAX Grounded Worldwide Q2 2019

Fri Jun 28, 2019 8:10 am

xmp125a wrote:
SteinarN wrote:
From Aviation Week

It seems in fact like the latest problem is related to FCC overload. The article says the THS is slow to react to manual trim inputs from the thumb switches when simulating a trim runaway. Thereby forcing the aircraft into a steep dive before the THS finally starts reacting to the thumb switches.


Let's remember the cockpit recording from the ET crash. "Manual trim is not working". We debated ad nauseam whether they meant manual manual trim (wheel) or manual electric trim (switch).

This could be huge, if the problem was in the original MAX firmware, and was discovered only now. And is basically impossible to prove or disprove that the actual processor behaved this way when ET had problems trimming -, as in all real-time embedded systems. Sole fact that it exists is a disaster for Boeing (remember Toyota lawsuit?).


Exactly.. and the CDR has some weird traces that could be interpreted as switches being flicked but the tail surface not moving, I think ET302 saw part of this issue.

Also, from the article I posted above it sounds like the FAA went looking for this failure mode and were not really surprised when the aircraft didnt react as it was supposed to, so did Boeing already know about this failure mode and didnt care or did FAA identify a new failure?

Neither answer is good at this point, either way FAA now has to crawl over every aspect of the MAX control logic challenging every one of Boeings assumptions all the way back to the NG.
BV
 
xmp125a
Posts: 233
Joined: Mon Mar 11, 2019 6:38 pm

Re: Boeing 737MAX Grounded Worldwide Q2 2019

Fri Jun 28, 2019 8:11 am

zhetenyi1973 wrote:
This is simply not true. It looks like you never heard of OSE which is an Real-Time OS and even Linux is capable to run RT applications. Of course you have to patch the kernel for it.
In an RT OS you can define hard priority between processes meaning a higher priority process is scheduled immediately if it wants to run. Of course you can starve out lower priority
processes if you are not careful. You can even have higher priority processes than interrupts, meaning the kernel will note the interrupt, but the handler code will be run later.
Most probably something similar happened here, so the solution is simply to increase the priority of the process which handles that particular data. Or the message dispatcher has to
prioritize these kind of data/messages over other data/messages in case there is only a simple loop in the code.


Unless the grandfathered processor is actually too slow for all the new tasks they added in the MAX and realtime tasks starve on their own. You are forgetting this very plausible scenario, since Boeing now plans "another software modification to re-route data across multiple microprocessor chips in the flight control computer".

Source: https://theaircurrent.com/aviation-safe ... re-glitch/
 
xmp125a
Posts: 233
Joined: Mon Mar 11, 2019 6:38 pm

Re: Boeing 737MAX Grounded Worldwide Q2 2019

Fri Jun 28, 2019 8:16 am

BoeingVista wrote:
Also, from the article I posted above it sounds like the FAA went looking for this failure mode and were not really surprised when the aircraft didnt react as it was supposed to


With bot Lion and ET investigation still ongoing and this being unprecedented situation for a long time, perhaps they got a hint from NTSB? "Guys, can you check this next time when you are in MAX sim?". These unexplained traces and voice recordings are certainly being one of the main mysteries the investigation team needs to solve, so I would not be surprised.

AFAIK FAA is on much better terms with NTSB than NHTSA is... (who is fighting NTSB recommendations)
 
xmp125a
Posts: 233
Joined: Mon Mar 11, 2019 6:38 pm

Re: Boeing 737MAX Grounded Worldwide Q2 2019

Fri Jun 28, 2019 8:27 am

par13del wrote:
RickNRoll wrote:
No need to find the QRH as the response to a Stab Trim Runaway is a memory item, as stated many times on this thread..
I don't think you understand the situation. The problem is that the pilots may do the right action and not see the plane respond quickly enough because the computer is overloaded. They will them think they have not taken the correct action and try something else.

So to speculate, the ET pilots understood it was MCAS, initiated the appropriate NNC but when the response was not immediate they moved on?[/quote]

IF the problem was in the original software, not introduced with MAX 2.0. Either way this is potentially worse than MCAS issue:

1) IF the original firmware had the same problem, it throws a very different light on ET crash, where it would mean that it is the yet unknown problem, which may prevented ET crew from saving the plane, even when following the checklist, and obviously there are scenarios that overload the safety-critical CPU (perhaps there are more, and FAA found only one of them)
2) IF the problem was introduced with MCAS 2.0 then it looks disastrous for Boeing since they unknowingly introduced NEW failure mode.
 
User avatar
BoeingVista
Posts: 1991
Joined: Mon Jan 12, 2009 9:54 am

Re: Boeing 737MAX Grounded Worldwide Q2 2019

Fri Jun 28, 2019 9:05 am

xmp125a wrote:

Let's remember the cockpit recording from the ET crash. "Manual trim is not working".


Yes, when the (manual) electric trim switch is flicked the pilots would be expecting the manual trim wheels to spin to the command, this would be the visual and auditory cue that the input was working, under pressure thats a reasonable short hand though not precise enough. In the time critical height limited situation they found themselves in they had to move to the next item / fix.

xmp125a wrote:

IF the problem was in the original software, not introduced with MAX 2.0. Either way this is potentially worse than MCAS issue:

1) IF the original firmware had the same problem, it throws a very different light on ET crash, where it would mean that it is the yet unknown problem, which may prevented ET crew from saving the plane, even when following the checklist, and obviously there are scenarios that overload the safety-critical CPU (perhaps there are more, and FAA found only one of them)
2) IF the problem was introduced with MCAS 2.0 then it looks disastrous for Boeing since they unknowingly introduced NEW failure mode.


Its really difficult to know which one of these is worse for Boeing but they now also have a pissed FAA on their case as this discovery makes it obvious that their original oversight was lacking.
BV
 
xmp125a
Posts: 233
Joined: Mon Mar 11, 2019 6:38 pm

Re: Boeing 737MAX Grounded Worldwide Q2 2019

Fri Jun 28, 2019 9:12 am

PixelFlight wrote:
The "out of processing power" is very hard to believe in the context of a DO-178 compliant software development, especially for the transition from MCAS v1 to MCAS v2. From what have been released so far, the differences between the MCAS v1 and MCAS v2 is just a couple of relatively simple conditions. Even on a 30 years old processors those condition will not add more than a fraction of millisecond of processing time. It's not possible to properly certify for flight operation a computer + software combo with a such insanely small processing margin time.


Why do you assume Boeing's code is DO-178 compliant? With all the cut corners that we witnessed, maybe the hastily added MCAS 1.0 was not tested for compliance.Maybe the MCAS 2.0 fix was not tested. Maybe the last code that was entirely compliant flies in NG.

I agree with the rest. If somehow turns out the processor cannot handle all the tasks, delegated in the MAX, the only feasible way would be to add another processor and transfer all the MAX-specific tasks there. Would of course needed redesigning the FCC, recertification, but this could be done within a year.

While we talk disparagingly about "spaghetti code" in software design, the embedded production hardware often IS spaghetti after everything is optimized - the meaning being it is not modular at all and you cannot take or change just little bit, because any change you made will require some changes to some other component, which will require changes up the chain, etc.
 
XRAYretired
Posts: 488
Joined: Fri Mar 15, 2019 11:21 am

Re: Boeing 737MAX Grounded Worldwide Q2 2019

Fri Jun 28, 2019 9:21 am

xmp125a wrote:
BoeingVista wrote:
Also, from the article I posted above it sounds like the FAA went looking for this failure mode and were not really surprised when the aircraft didnt react as it was supposed to


With bot Lion and ET investigation still ongoing and this being unprecedented situation for a long time, perhaps they got a hint from NTSB? "Guys, can you check this next time when you are in MAX sim?". These unexplained traces and voice recordings are certainly being one of the main mysteries the investigation team needs to solve, so I would not be surprised.

AFAIK FAA is on much better terms with NTSB than NHTSA is... (who is fighting NTSB recommendations)


Crash Investigations are formal. If such a problem was considered a possible contributory factor, it would be formally tested with the crash flight confirguration.

This does not preclude that it was considered, and if so, the FAA and Boeing would be informed and it would be expected that the sceanario would be tested for the fix configuration, but then it is also expected that certification tests the worst case anyway.

More info required I think.

Ray
 
xmp125a
Posts: 233
Joined: Mon Mar 11, 2019 6:38 pm

Re: Boeing 737MAX Grounded Worldwide Q2 2019

Fri Jun 28, 2019 9:27 am

Amiga500 wrote:
xmp125a wrote:
Amiga500 wrote:
2. Boeing cannot realistically change the computing hardware within 24 months. They would have to re-certify the whole architecture (S/W and H/W). That would be an absolute nightmare scenario.


Let's take a step back.Remember when everyone were shaking their heads how on Earth MCAS V1.0 found its way to production?

A bit of conspiracy teory, but plausible one: what if "proper" MCAS design was tested during MAX development but found to be too demanding for the flight computer? And then MCAS 1.0 with its fatal flaws was born, and seemingly stupid design decisions, like use 1 sensor only, alternate sensor between the flights?

Is that really out of the question?


Ah, nah, can't really believe that.

Trying to do that would leave a paper trail - which would have ended up in folks being prosecuted before now.


I think it is way too soon to claim this. I am not from USA, but if I learned anything from the way various US agencies operate (FBI in particular), absolutely nothing is leaked until the hammer falls, hard. We simply don't know what they found during the investigation. As, again, some tinfoil hat speculation, perhaps there was FBI->NTSB->FAA leak which instructed FAA test pilots to test the scenario which found the latest flaw in MAX code. Or they had been incredibly lucky. End-to-end testing of embedded realtime systems is usually useless, unless you know what are you looking for, which is how I would describe this latest test which revealed the flaw.
 
Amiga500
Posts: 2267
Joined: Tue Mar 03, 2015 8:22 am

Re: Boeing 737MAX Grounded Worldwide Q2 2019

Fri Jun 28, 2019 9:32 am

xmp125a wrote:
Why do you assume Boeing's code is DO-178 compliant?


We already know that v1.0 was not Do-178 compliant.
 
Amiga500
Posts: 2267
Joined: Tue Mar 03, 2015 8:22 am

Re: Boeing 737MAX Grounded Worldwide Q2 2019

Fri Jun 28, 2019 9:34 am

BoeingVista wrote:
Neither answer is good at this point, either way FAA now has to crawl over every aspect of the MAX control logic challenging every one of Boeings assumptions all the way back to the NG.


Another nightmare scenario.

That's gonna add months.
 
ACATROYAL
Posts: 49
Joined: Mon Aug 20, 2007 11:25 am

Re: Boeing 737MAX Grounded Worldwide Q2 2019

Fri Jun 28, 2019 10:01 am

Well the FAA and Boeing identified a new issue with the MAX plane!! Try to fix one thing and something else pops up or goes wrong! It's amazing how the fantastic 737 went from Hero to Zero. They should of stopped at the 737-800 { arguably one of the best planes ever} This is why I won't get near a MAX plane ever again, simply put it's just too much engine for this airframe...Boeing is trying desperately to get this plane to do something it was never meant to do...
 
Interested
Posts: 647
Joined: Thu May 19, 2016 12:19 pm

Re: Boeing 737MAX Grounded Worldwide Q2 2019

Fri Jun 28, 2019 10:15 am

Has anyone actually got anything positive to say about the plane?

I can half understand the defence of it by people and explanations of how we ended up with it etc

But that's all very defensive

But my question is to the people who defend the plane (or anybody for the matter). Is there actually anything to say about this plane that is positive or for people to be proud of or get excited about?

Isn't it just the ultimate compromise of a plane that's proved to be a disaster?
Last edited by Interested on Fri Jun 28, 2019 10:22 am, edited 1 time in total.
 
User avatar
Revelation
Posts: 20953
Joined: Wed Feb 09, 2005 9:37 pm

Re: Boeing 737MAX Grounded Worldwide Q2 2019

Fri Jun 28, 2019 10:17 am

JerseyFlyer wrote:
This seems to be a test of what happens when a microprocessor is broken, i.e. a test of adequate redundancy, which neither Boeing nor FAA tested for previously. Puts both in a bad light.

I've read at least four different accounts so far (CNN, Bloomberg, AvWeek, AirCurrent) and all say the issue is the trim system is not responsive enough, and none say it's an issue with a broken microprocessor, or a lack of redundancy.
Wake up to find out that you are the eyes of the world
The heart has its beaches, its homeland and thoughts of its own
Wake now, discover that you are the song that the morning brings
The heart has its seasons, its evenings and songs of its own
 
Amiga500
Posts: 2267
Joined: Tue Mar 03, 2015 8:22 am

Re: Boeing 737MAX Grounded Worldwide Q2 2019

Fri Jun 28, 2019 10:18 am

Interested wrote:
Isn't it just the ultimate compromise of a plane that's proved to be a disaster?


All designs are compromised in one way or another. Its just the way things are.

If they had taken the extra time/money to fix the tail volume ratio (i.e. make tailplane bigger) then there would not have been an issue.
 
Interested
Posts: 647
Joined: Thu May 19, 2016 12:19 pm

Re: Boeing 737MAX Grounded Worldwide Q2 2019

Fri Jun 28, 2019 10:32 am

Revelation wrote:
JerseyFlyer wrote:
This seems to be a test of what happens when a microprocessor is broken, i.e. a test of adequate redundancy, which neither Boeing nor FAA tested for previously. Puts both in a bad light.

I've read at least four different accounts so far (CNN, Bloomberg, AvWeek, AirCurrent) and all say the issue is the trim system is not responsive enough, and none say it's an issue with a broken microprocessor, or a lack of redundancy.


The CNN report claims there is an issue with the microprocessor that leads to the nose pointing down if the chip fails and this cannot be recovered quickly enough by pilots

It almost seems to suggest that if the pilots could correct the problem quick enough it wouldn't be such a big thing?

I've not quoted directly as I've been banned in the past for quoting directly!
 
Interested
Posts: 647
Joined: Thu May 19, 2016 12:19 pm

Re: Boeing 737MAX Grounded Worldwide Q2 2019

Fri Jun 28, 2019 10:36 am

XRAYretired wrote:
xmp125a wrote:
par13del wrote:
I don't think you understand the situation. The problem is that the pilots may do the right action and not see the plane respond quickly enough because the computer is overloaded. They will them think they have not taken the correct action and try something else.

So to speculate, the ET pilots understood it was MCAS, initiated the appropriate NNC but when the response was not immediate they moved on?


IF the problem was in the original software, not introduced with MAX 2.0. Either way this is potentially worse than MCAS issue:

1) IF the original firmware had the same problem, it throws a very different light on ET crash, where it would mean that it is the yet unknown problem, which may prevented ET crew from saving the plane, even when following the checklist, and obviously there are scenarios that overload the safety-critical CPU (perhaps there are more, and FAA found only one of them)
2) IF the problem was introduced with MCAS 2.0 then it looks disastrous for Boeing since they unknowingly introduced NEW failure mode.


Nice theory, but the FDR does not seem to support it. Thumb switch signal is reported, MCAS can be shown to be inhibited and at least some pitch trim effect on most of the commands. It would have to be a very strange effect.

If we knew the test result details it would help e.g. did the processor register the command and not act on it or not register at all. Was MCAS inhibited or continue to drive the stabiliser nose down?

Ray[/quote]

Apologies for not having more detail.

But I seem to remember posters discussing in one of the crashes that every time the pilot(s) tried to trim they seemed to arrive at the exact same number??

And posters were thinking this was too co-incidental to happen randomly?

Could this new flaw be anything to do with that end result?

Or am I off track?
 
User avatar
Revelation
Posts: 20953
Joined: Wed Feb 09, 2005 9:37 pm

Re: Boeing 737MAX Grounded Worldwide Q2 2019

Fri Jun 28, 2019 10:38 am

Interested wrote:
The CNN report claims there is an issue with the microprocessor that leads to the nose pointing down if the chip fails and this cannot be recovered quickly enough by pilots

It almost seems to suggest that if the pilots could correct the problem quick enough it wouldn't be such a big thing?

So, if your phone is too slow, is your phone broken, or is there a problem with the apps you've loaded or even the phone's OS?

This problem appears to be more like the later, not the former.

If the microprocessor was broken it would need to be replaced, and it seems no one is suggesting that any longer.

And from what we read yes, the issue will be resolved once the system is responsive enough.
Wake up to find out that you are the eyes of the world
The heart has its beaches, its homeland and thoughts of its own
Wake now, discover that you are the song that the morning brings
The heart has its seasons, its evenings and songs of its own
 
XRAYretired
Posts: 488
Joined: Fri Mar 15, 2019 11:21 am

Re: Boeing 737MAX Grounded Worldwide Q2 2019

Fri Jun 28, 2019 10:43 am

xmp125a wrote:
par13del wrote:
RickNRoll wrote:
No need to find the QRH as the response to a Stab Trim Runaway is a memory item, as stated many times on this thread..
I don't think you understand the situation. The problem is that the pilots may do the right action and not see the plane respond quickly enough because the computer is overloaded. They will them think they have not taken the correct action and try something else.

So to speculate, the ET pilots understood it was MCAS, initiated the appropriate NNC but when the response was not immediate they moved on?


IF the problem was in the original software, not introduced with MAX 2.0. Either way this is potentially worse than MCAS issue:

1) IF the original firmware had the same problem, it throws a very different light on ET crash, where it would mean that it is the yet unknown problem, which may prevented ET crew from saving the plane, even when following the checklist, and obviously there are scenarios that overload the safety-critical CPU (perhaps there are more, and FAA found only one of them)
2) IF the problem was introduced with MCAS 2.0 then it looks disastrous for Boeing since they unknowingly introduced NEW failure mode.[/quote]Nice theory, but the FDR does not seem to support it. Thumb switch signal is reported, MCAS can be shown to be inhibited and at least some pitch trim effect on most of the commands. It would have to be a very strange effect.

If we knew the test result details it would help e.g. did the processor register the command and not act on it or not register at all. Was MCAS inhibited or continue to drive the stabiliser nose down?

Ray
 
Thorkel
Posts: 21
Joined: Mon Dec 15, 2014 2:38 pm

Re: Boeing 737MAX Grounded Worldwide Q2 2019

Fri Jun 28, 2019 10:44 am

zhetenyi1973 wrote:
StTim wrote:
The other thing, in terms of timeliness, is that if these chips are associated with real-time constraints is they won’t be running anything approaching a ‘normal’ OS. They’ll be no pre-emotive multitasking or anything like that (again, doesn’t help for worst case execution times). It’ll be single task, or you’ll have something like a cyclic scheduler where each application has a fixed share of processor time. If it is a cyclic scheduler, there may be options to increase the ‘share’ given to MCAS - there may be unused capacity, or they may be able to reduce the share given to other apps on the processor.


This is simply not true. It looks like you never heard of OSE which is an Real-Time OS and even Linux is capable to run RT applications. Of course you have to patch the kernel for it.
In an RT OS you can define hard priority between processes meaning a higher priority process is scheduled immediately if it wants to run. Of course you can starve out lower priority
processes if you are not careful. You can even have higher priority processes than interrupts, meaning the kernel will note the interrupt, but the handler code will be run later.
Most probably something similar happened here, so the solution is simply to increase the priority of the process which handles that particular data. Or the message dispatcher has to
prioritize these kind of data/messages over other data/messages in case there is only a simple loop in the code.


I think you’ve quoted the wrong person here. That’s not quote. I’ll also say that everything I’ve said there is true and I’ve been developing software for real-time safety related systems for two decades. I’m intimately familiar the vast majority of real-time OSs out there, having deployed applications on QNX, VxWorks, LynxOS and several RT Linux flavours.

These RT OSs are used in many circumstances, but are much less commonly used for safety critical flight control systems and most of the time don’t work well (or in many cases at all) on earlier 680XX, early PPC or 80286 or below processors. Lack of MMU/16bit processors/etc. Many of these OSs also did not exist when the avionics systems on the 737s/A320s etc were conceived. With those OSs, you can get much better predictability than a general OS, but still fail to guarantee Worst Case execution Time. You really need the DO178 specific and/or ARINC 653 specific variants - and if it’s ARINC 653, it’ll be a cyclic scheduler (at least for the parent). Cyclic schedulers are simple as are 68040 processors. They’re also predictable and analysable, and have been commonly used for avionics systems.
Last edited by Thorkel on Fri Jun 28, 2019 11:05 am, edited 2 times in total.

Popular Searches On Airliners.net

Top Photos of Last:   24 Hours  •  48 Hours  •  7 Days  •  30 Days  •  180 Days  •  365 Days  •  All Time

Military Aircraft Every type from fighters to helicopters from air forces around the globe

Classic Airliners Props and jets from the good old days

Flight Decks Views from inside the cockpit

Aircraft Cabins Passenger cabin shots showing seat arrangements as well as cargo aircraft interior

Cargo Aircraft Pictures of great freighter aircraft

Government Aircraft Aircraft flying government officials

Helicopters Our large helicopter section. Both military and civil versions

Blimps / Airships Everything from the Goodyear blimp to the Zeppelin

Night Photos Beautiful shots taken while the sun is below the horizon

Accidents Accident, incident and crash related photos

Air to Air Photos taken by airborne photographers of airborne aircraft

Special Paint Schemes Aircraft painted in beautiful and original liveries

Airport Overviews Airport overviews from the air or ground

Tails and Winglets Tail and Winglet closeups with beautiful airline logos