Pages: 1 [2] 3
Author Topic: This is why you don't bindly disable DTCs  (Read 11593 times)
nyet
Administrator
Hero Member
*****

Karma: +604/-166
Offline Offline

Posts: 12232


WWW
« Reply #15 on: March 14, 2019, 03:19:28 PM »

Thanks, littco.

I think what really bothers me, though (from my history of staring at ME7L logs lol) is this:

What I don't understand is which sensor is "wrong".

Right (green) clearly can't be right, it's showing negative alpha during climbout.

But if left (red) is right, then MCAS is properly triggering, since red is showing a much higher AoA.

The only reasonable answer is that neither reading is right. Right is reading too high, and left is reading too low. Really strange.

Do you have any further thoughts on that?
« Last Edit: March 15, 2019, 01:05:20 PM by nyet » Logged

ME7.1 tuning guide (READ FIRST)
ECUx Plot
ME7Sum checksum checker/corrrector for ME7.x

Please do not ask me for tunes. I'm here to help people make their own.

Do not PM me technical questions! Please, ask all questions on the forums! Doing so will ensure the next person with the same issue gets the opportunity to learn from your experience.
Blazius
Hero Member
*****

Karma: +89/-40
Offline Offline

Posts: 1277



« Reply #16 on: March 14, 2019, 03:26:27 PM »

Thanks, littco.

I think what really bothers me, though (from my history of staring and ME7L logs lol) is this:

Do you have any further thoughts on that?

Wow a pilot on nefmoto , thats nice Smiley





One of those sensor is reading right, most likely the green one , when when on ground my guess  is that the nose is slightly downward ( -5 or so) [ or upward if the red one is right - but I dont think you will have 20 degress angle on even if the front is raised with the new added front landing gear ]

and the other one being offset by a constant value as it still follows the other one exactly just with a set difference.
Logged
nyet
Administrator
Hero Member
*****

Karma: +604/-166
Offline Offline

Posts: 12232


WWW
« Reply #17 on: March 14, 2019, 03:28:45 PM »

But the green one is showing negative angle of attack on climb... is that even possible? Even with flaps extended?

AoA is angle of attack with incident airmass, not pitch attitude, so I'd expect them to be zeroed on the ground (unless there was some strange crosswinds or updraft/downdraft) with some spring assisted centering (they're physical vanes as far as i know)

and the other one being offset by a constant value as it still follows the other one exactly just with a set difference.

But not constant on log start and later leading up to takeoff...
« Last Edit: March 14, 2019, 03:32:17 PM by nyet » Logged

ME7.1 tuning guide (READ FIRST)
ECUx Plot
ME7Sum checksum checker/corrrector for ME7.x

Please do not ask me for tunes. I'm here to help people make their own.

Do not PM me technical questions! Please, ask all questions on the forums! Doing so will ensure the next person with the same issue gets the opportunity to learn from your experience.
Blazius
Hero Member
*****

Karma: +89/-40
Offline Offline

Posts: 1277



« Reply #18 on: March 14, 2019, 03:35:11 PM »

But the green one is showing negative angle of attack on climb... is that even possible? Even with flaps extended?

AoA is angle of attack with incident airmass, not pitch attitude, so I'd expect them to be zeroed on the ground (unless there was some strange crosswinds or updraft/downdraft) with some spring assisted centering (they're physical vanes as far as i know)

But not constant on log start and later leading up to takeoff...

idk, some fuckery going on , prolly maintenance or configuration issue. If these sensors have to be calibrated after every flight then maybe they didnt do it or something again, apparently Lion Air was cutting costs on maintenance.
Logged
nyet
Administrator
Hero Member
*****

Karma: +604/-166
Offline Offline

Posts: 12232


WWW
« Reply #19 on: March 14, 2019, 03:38:02 PM »

Incidentally, a previous crew noticed the problem and logged it (IIRC). One or both sensors were serviced, i think the one on the left (i'd have to look it up)
Logged

ME7.1 tuning guide (READ FIRST)
ECUx Plot
ME7Sum checksum checker/corrrector for ME7.x

Please do not ask me for tunes. I'm here to help people make their own.

Do not PM me technical questions! Please, ask all questions on the forums! Doing so will ensure the next person with the same issue gets the opportunity to learn from your experience.
nyet
Administrator
Hero Member
*****

Karma: +604/-166
Offline Offline

Posts: 12232


WWW
« Reply #20 on: March 14, 2019, 03:56:41 PM »

Interesting post on /. https://tech.slashdot.org/comments.pl?sid=13577684&cid=58274744

Quote
It has fuselage from the 737, Engines from the 787, flight controls from the A320 (the famous plane where the automation led to crashes. Now 737 Max has taken that mantle)

Boeing should have done a clean sheet design as a replacement of 737 instead of putting engines so big on an airframe meant for much smaller engine.

They created and unstable plane and tried to fix it in software.

While this is an approach often used in fighter jets which are deliberately made unstable so that they can change directions easily its not something you do on a civilian airplane. A civilian pilot does not have the reflexes of a fighter pilot to fix things if the computer is misbehaving

To recap the plane was too small for the engines they wanted to put on it. So they put the engines in a cantilevered position so now the center of thrust was significantly away from the centre of gravity and the plane had a tendency to pitch up and stall. To avoid this they added MCAS which would pitch the nose down in case of a stall detection. To detect the stall they used the AoA sensor and in a freshman Fault Tolerant Computing bug depended on only one sensor when they had 2. They made the warning light showing the AoA sensor is broken an option (only American signed up for this option which is probably why American hasnt had a crash). Then to make things worse they didnt tell the pilots. Also in the NG if the auto trim was runaway pulling back on the yoke would disengage the auto trim. With MCAS they changed this. The auto trim would only disengage for 10 seconds and then MCAS would add more trim and it would keep adding more and more trim till the pilots could not counter even if they pulled the yoke all the way back. Again a software bug. Further to make things worse THEY DID NOT TELL THE PILOTS ABOUT THIS CHANGE. So the yoke maneouver does not work so the only maneovour that works is disengaing the trim using the 2 cutoff switches but this only disengages the Auto trim. If the plane is already nose down it doesnt go back to normal trim. Now you have to pull back on the yoke which was not working till a moment ago or spin two manual trim wheels to get the trim back. All this is happening close to ground as MCAS only engages at low speeds found at takeoff.

Boeing could have avoided this in many ways

1) Build a clean sheet design which is stable with the larger engines
2) Failing that build a MCAS which is fault tolerant with multiple sensors or can be countermanded by the pilot by pulling back on the yoke (This is what they are doing now with the software fix). Not ideal for if the pilot is really flying badly now he can stall the plane
3) Failing that tell the pilots about the MCAS system, the change in the yoke behaviour and have them go through difference training.

They did not do 1 as it would cost too much money
They did not do 3 as they wanted to avoid airlines having to train pilots making the plane easier to sell. One of the reasons there are 5000 737 Max orders is that it needs no crosstraining to fly (officially)
They did not do 2 because of sheer laziness or stupidity in the engineering team

So the Engineering team is now fixing their error No 2. But the Exec team's error No 1 and the Marketing team's error no 3 are still not fixed.
« Last Edit: March 14, 2019, 04:06:01 PM by nyet » Logged

ME7.1 tuning guide (READ FIRST)
ECUx Plot
ME7Sum checksum checker/corrrector for ME7.x

Please do not ask me for tunes. I'm here to help people make their own.

Do not PM me technical questions! Please, ask all questions on the forums! Doing so will ensure the next person with the same issue gets the opportunity to learn from your experience.
littco
Hero Member
*****

Karma: +52/-7
Offline Offline

Posts: 903


« Reply #21 on: March 14, 2019, 04:02:49 PM »

Thanks, littco.

I think what really bothers me, though (from my history of staring and ME7L logs lol) is this:

Do you have any further thoughts on that?
AOA is the the aircraft relative to the airflow, The AOA sensor is basically a wing on pivot , due to gravity they fall down when the aircraft is stationary as the speed increases they will start to lift, same putting you hand out a car window, until they are basically flat to the airflow. this would normally read on the AOA sensor as a small positive number. as you then pitch up on take off the AOA sensor will stay flat relative to the oncoming air but the reading will increase as the plane pivots around the airflow. Typically a pitch up of 20degrees could be seen with a AOA of maybe 4 degrees on the aircraft due to the flaps so 25-30 AOA could been seen. After this the AOA will vary due to speed, plane configuration, c of g etc its dynamic and always changing.

Looking I would say Green is correct. As the plane taxies the AOA sensor is hanging down as no airflow over it, as it accelerates as it takes of it rises into the airflow and increases AOA and then increases with pitch. The RED line not only starts at a high AOA while taxing but also then decreases as it rotates .

A simple error check would be, when aircraft is on the ground, squat switch is active and airspeed is less than 15knts then AOA sensor should read X, if not then error, or if left or right are more than say 5degrees different a warning light is activated. We have it on just about every other system. As far as im aware up until now certainly all the aircraft I have flown the AOA sensor is there as a guide and not needed other than a good airmanship guide. IE optimimal fuel burn or C of G . ITs not like a pitot or gryo which is a critical flight instrument .

Also I believe this is the 1st time Boeing has done the coding for it's aircrafts where before it was outsourced so maybe a lack of experience was to blame.
Logged
nyet
Administrator
Hero Member
*****

Karma: +604/-166
Offline Offline

Posts: 12232


WWW
« Reply #22 on: March 14, 2019, 04:06:37 PM »

Looking I would say Green is correct. As the plane taxies the AOA sensor is hanging down as no airflow over it, as it accelerates as it takes of it rises into the airflow and increases AOA and then increases with pitch. The RED line not only starts at a high AOA while taxing but also then decreases as it rotates

Thank you! This is the last missing piece in my mind.

Quote
Also I believe this is the 1st time Boeing has done the coding for it's aircrafts where before it was outsourced so maybe a lack of experience was to blame.

YGTBFKM

That is... pathetic.
« Last Edit: March 15, 2019, 02:04:56 PM by nyet » Logged

ME7.1 tuning guide (READ FIRST)
ECUx Plot
ME7Sum checksum checker/corrrector for ME7.x

Please do not ask me for tunes. I'm here to help people make their own.

Do not PM me technical questions! Please, ask all questions on the forums! Doing so will ensure the next person with the same issue gets the opportunity to learn from your experience.
nyet
Administrator
Hero Member
*****

Karma: +604/-166
Offline Offline

Posts: 12232


WWW
« Reply #23 on: March 14, 2019, 04:13:07 PM »

As far as im aware up until now certainly all the aircraft I have flown the AOA sensor is there as a guide and not needed other than a good airmanship guide. IE optimimal fuel burn or C of G . ITs not like a pitot or gryo which is a critical flight instrument .

More than that, I think the critical point is that in this case it is also used as an INPUT to a system which can overcome pilot input, with absolutely no plausibility checking. That is truly insane. Literally criminal negligence.
Logged

ME7.1 tuning guide (READ FIRST)
ECUx Plot
ME7Sum checksum checker/corrrector for ME7.x

Please do not ask me for tunes. I'm here to help people make their own.

Do not PM me technical questions! Please, ask all questions on the forums! Doing so will ensure the next person with the same issue gets the opportunity to learn from your experience.
littco
Hero Member
*****

Karma: +52/-7
Offline Offline

Posts: 903


« Reply #24 on: March 14, 2019, 04:17:14 PM »

YGTBFKM

That is... pathetic.

Due to the monumental cockups with the 787 and it's outsourced avionics they brought it all in house.

Logged
littco
Hero Member
*****

Karma: +52/-7
Offline Offline

Posts: 903


« Reply #25 on: March 14, 2019, 04:24:54 PM »

More than that, I think the critical point is that in this case it is also used as an INPUT to a system which can overcome pilot input, with absolutely no plausibility checking. That is truly insane. Literally criminal negligence.

To be fair the Airbus is like that, depending on the stage of flight, the envelope you are allowed to use, is controlled and if the computer doesn't like it , it will override your controls.

Being trained or experienced to recognise what is happening is nearly as important as actual flying. I think the negligence is in the lack of training or knowledge as I am sure now if any pilot where to experience the same thing they would automatically turn off the MCAS and I am sure that if neither had crashed AND Boeing had trained for this, nothing would have changed. It has a name, normalisation of deviation, ie making something wrong normal. Now though they will have to change it so it can never happen again.
Logged
vwaudiguy
Hero Member
*****

Karma: +53/-37
Offline Offline

Posts: 2024



« Reply #26 on: March 15, 2019, 11:56:01 AM »

Nyet, I am a pilot by profession and fly commercially, one thing that is not shown that graph is the throttle position ( speed is ). The whole reason Mcas was fitted to these new MAX aircraft was to over come the larger engines which where further forward and higher than the older NG ones, this meant with a change of power (not speed) setting the pitch moment was greater ( up or down ) and thus MCAS was fitted to correct this. Issue with these is that pilot training and conversion from the old aircraft to new ones was a 1 hour conversation including a supposed youtube video. With auto throttles, and the stage of flight they where, IE flap retractions and gear, the possibility for the MCAS to over correct is possible and then big changes in pitch to occur that the pilot due to lack of suitable training where not prepared for, leading to a war between a pilot that wants to 1 thing and a flight computer wanting to do the opposite. The planes wins and result is a non survivable pitch down. It is as much pilot lack of training, the system could of been turned off, as much as it is bad programming for allowing it, ie trip out. It could be the error was noted by the crew but the paper work allowed them to fly, The Mcas was seen , I believe as a flight enhancer rather than a safety critical item. Ie makes the crews life easier.

I agree with the AOA sensors, but generally the air date computer will take both sets of data and for the auto pilot use the sensor relating to the side which is flying, IE of the captain is flying on the left use the left hand AOA sensor, should this fail then the backup will be used, but typically a set limit of tolerance will be built in, ie 6degress then a fault flag will be given, this will be part of the whats called an MEL will give guidance to the pilots if they can fly with the error or not, and in some cases it will allow this.

I am no expert on this particular plane by any means but I know from experience until a problem happens it may have gone unnoticed for years and not been any issue. Computers are only as good as the people who program them or the people that use them.



Thanks for sharing your experience.
Logged

"If you have a chinese turbo, that you are worried is going to blow up when you floor it, then LOL."
nyet
Administrator
Hero Member
*****

Karma: +604/-166
Offline Offline

Posts: 12232


WWW
« Reply #27 on: March 18, 2019, 12:46:51 PM »

Great breakdown here
https://www.seattletimes.com/business/boeing-aerospace/failed-certification-faa-missed-safety-issues-in-the-737-max-system-implicated-in-the-lion-air-crash/
Logged

ME7.1 tuning guide (READ FIRST)
ECUx Plot
ME7Sum checksum checker/corrrector for ME7.x

Please do not ask me for tunes. I'm here to help people make their own.

Do not PM me technical questions! Please, ask all questions on the forums! Doing so will ensure the next person with the same issue gets the opportunity to learn from your experience.
Risky
Full Member
***

Karma: +4/-0
Offline Offline

Posts: 107


« Reply #28 on: March 18, 2019, 02:48:23 PM »

Quote from: Article
According to a detailed FAA briefing to legislators, Boeing will change the MCAS software to give the system input from both angle-of-attack sensors.

Input from both sensors should've been allowed from the start.



Logged
mysman
Newbie
*

Karma: +1/-0
Offline Offline

Posts: 8


« Reply #29 on: March 21, 2019, 08:02:35 AM »

Apparently Boeing did incorporate plausibility checking but it was an optional "upgrade"... sigh.

https://www.nytimes.com/2019/03/21/business/boeing-safety-features-charge.html
Logged
Pages: 1 [2] 3
  Print  
 
Jump to:  

Powered by SMF 1.1.21 | SMF © 2015, Simple Machines Page created in 0.029 seconds with 18 queries. (Pretty URLs adds 0s, 0q)