737 Max

Status
Not open for further replies.
Member
Joined 2014
Paid Member
Thank you George. Found a handy youtube video YouTube where a 737 pilot (oddly using a model 747 for illustration) explains it and shows a little video of them spinning. He mentions trim runaway there and scratches the surface of the Boeing design philosophy in that you can see what the autopilot is doing all the time and spot issues before they go non-linear.



Most fascinating for me was how he discussed that you add trim by feel on the column as it was something I had never even thought about. The concept of delicately balancing 70 odd Tonnes flying at 450knots by touch is going to take a little while for me to get my head around :).
 
All expressed concerns here, are valid as you can read.
737 MAX - MCAS

Alas, for the SW changes only, the design/implement/test/verify/approve cycle will take a minimum of two-three months, even if the change procedure had been completed prior to the second accident.

George

In your link, there is a link to a FAA directive, FAA Emergency AD 2018-23-51 -, on the subject.

Based on the FAA directive, my read of it is the FAA (and Boeing would have to be involved) acknowledges a problem with the MCAS system which can be caused by a faulty AOA reading.

BTW, extracted from the FAA directive
"and possible impact with terrain"

But it doesn't actually mention the Lion Air incident. So maybe a smoking gun, but doesn't blame Lion Air on MCAS and AOA.
 
Member
Joined 2016
Paid Member
Thank you George. Found a handy youtube video YouTube where a 737 pilot (oddly using a model 747 for illustration) explains it and shows a little video of them spinning. He mentions trim runaway there and scratches the surface of the Boeing design philosophy in that you can see what the autopilot is doing all the time and spot issues before they go non-linear.



Most fascinating for me was how he discussed that you add trim by feel on the column as it was something I had never even thought about. The concept of delicately balancing 70 odd Tonnes flying at 450knots by touch is going to take a little while for me to get my head around :).

Great link - thanks Bill... except he seems to have a lot of interesting videos! Another time burner! Interesting: YouTube
 
Last edited:
AX tech editor
Joined 2002
Paid Member
Most fascinating for me was how he discussed that you add trim by feel on the column as it was something I had never even thought about. The concept of delicately balancing 70 odd Tonnes flying at 450knots by touch is going to take a little while for me to get my head around :).

Its all a matter of scale. I did the same flying my 1000 pounds ultralight at 90 knots ;-)

Jan
 

6L6

Moderator
Joined 2010
Paid Member
Most fascinating for me was how he discussed that you add trim by feel on the column as it was something I had never even thought about. The concept of delicately balancing 70 odd Tonnes flying at 450knots by touch is going to take a little while for me to get my head around :).

Its all a matter of scale. I did the same flying my 1000 pounds ultralight at 90 knots ;-)

This is precisely the reason that control feel is so important - it must work as expected, by feel - tactile feedback. In the case of the 747 it's the hydraulic 'feel computer' that's making the yoke react the way the pilot expects, where in Jan's aircraft it's actual control cables and the flying surfaces themselves.
 
Last edited:
It turns out there is an "AOA sensor disagreement" safety indicator....but you have to select it as an option at extra cost when ordering the 737.

A safety feature is an optional extra!!!

Good grief, Charlie Brown.


Unbelievable!!!
I hope this safety feature is a built-in one in future, otherwise the control software makes no sense and should be disabled. As i read, there was no clear description in the handbook to handle this task.


Life depends on an optional extra!! Thanks Boeing
 
www.hifisonix.com
Joined 2003
Paid Member
Reminds me of some of those accidents on Air Crash Disaster where the ILS systems at airports were down because they were being repaired. There was one case - one of the Pacific Islands - that had a particularly difficult landing approach but the ILS was turned off. The approach required the aircraft to clear a ridge. A plane load of people on a late night flight slammed straight into the ridge. Final conclusion: ILS not available and pilot disorientation.
 
Last edited:
AOA sensor

lion-3AOA sensor.jpg AOA sensors have been problematic for many years, bug strikes can make them
read incorrectly, icing, etc. Relying on two and picking one randomly seems
to me to be a bit reckless without disconnecting or giving a warning when there are extremely different readings.
.
A former Collins Aerospace employee
 
Last edited:
What about this then?
According to the report, the aircraft’s digital flight data recorder (DFDR) noted a 20 degree disparity in the aircraft’s left and right AOA sensors that persisted throughout the flight. Flight JT610’s pilots started experiencing problems when the aircraft reached an altitude of 2,150 feet. At that point, flaps were completely retracted and the automatic aircraft nose down (AND) trim remained active for 10 seconds. Immediately after experiencing the AND condition, the pilots commanded aircraft nose up (ANU) trim and were instructed by a terminal controller to climb and maintain an altitude of 5,000 feet.

Once the pilots reached that altitude, the automatic AND trim was activated again followed by the flight crew commanding another ANU trim. This condition continued for the remainder of the flight. The pilots then started flying the aircraft manually, at which point the controller noticed the aircraft was descending. Minutes later, the pilots requested to land because of a flight control problem. Shortly after, the DFDR stopped recording and pilots stopped responding to controller communications.
Lion Air Report Analyzes Pilot Struggle With 737 MAX Trim System - Avionics

May have been some bug that wasn't completely ironed out? At any rate, it seems a recurring problem on that particular aircraft.
 
Last edited:
AX tech editor
Joined 2002
Paid Member
View attachment 744930 AOA sensors have been problematic for many years, bug strikes can make them
read incorrectly, icing, etc. Relying on two and picking one randomly seems
to me to be a bit reckless without disconnecting or giving a warning when there are extremely different readings.
.
A former Collins Aerospace employee

I remember the early Apollo flights where the flight computer was triplicated. Results from computations were compared continuously, and if one of the three disagreed with the other two, that one was taken off line.

Not 100% fail-safe, but a good step forward.

Jan
 
I remember the early Apollo flights where the flight computer was triplicated. Results from computations were compared continuously, and if one of the three disagreed with the other two, that one was taken off line.

A similar triple system is used on Airbus, or used to be when I knew someone who ceded for them.

Many years ago I worked on a ‘2 out of 3’ system for monitoring critical chemical processors. We supplied dozens of these systems which were fed with triple sensors. If there wasn’t total agreeemt, an alarm was sounded.

Same as I described a few pages back:

On the machines I work on, the critical controls require 2 or more sensors to agree before the signal is classified as valid. The systems are also only given 10% total authority over the whole of the controls. A 'miscompare' either invokes a 'voting' system where the outlier is rejected, or the system is invalidated and is flagged as being in a fault condition.

That is why I just can't understand why this system would be designed to rely on the validity of one data source.
 
Status
Not open for further replies.