cargil48 wrote:

Guys, hi

Before everybody jumps at me going back to this (in)famous topic, I still have some questions regarding the flight controls in general and how the market has gradually responded by increasing automatization in the FMS. My main point arose after reading an article the subject of which caught my attention: "Will we ever have commercial airplanes without pilots in the cockpit?" Obviously this led my brain reviewing all I had read about the MAX problem we all know about and what the industry has learned from it. Probably thousands of lines have been written already about it, spread everywhere in the internet, but I thought there are still some points maybe worthwhile reviewing and condensing in short answers for all those who are not too familiar with what happed and how it should have been avoided in the first place. First of all, I have some knowledge about aviation and airplanes, generally speaking and this means I'm not a pilot. Not even of an UL...but I have years of R/C big airplanes (up to 1/3d scale), building them and flying them, so I'm familiar with the techniques of flying in general and how planes are made.

Now, recapitulating: "The Maneuvering Characteristics Augmentation System (MCAS) is a flight control law managed by the flight control computer (FCC) and introduced on the 737 MAX to help it handle like a 737 Next Generation (NG), particularly at slow speeds and high angles of attack (AOA)." (from an article on Aviation Week).

Question nr. 1: Isn't every pilot trained from the beginning of his (long) career on until becoming able (or allowed) to seat himself on the nr. 2 seat of the cockpit (not even to speak of the nr. 1...) to fly a commercial plane, that pitch and speed are the most important parameters to have in mind? Every R/C piloting guy knows that if he does not have that in mind he looses his toy!... Being so, when a pilot is flying in a situation of low speed and higher AoA than on normal cruise isn't it "natural" that he keeps every five seconds or so looking at the VSI and at the speed indicator?

Question nr. 2: We all know what redundancy is and that (if I'm not mistaken) every FMS relies on three independent systems which feed the FCC with data to allow the FMS to make the correct decisions? If so, how could Boeing have introduced a new additional system to the FMS of the MAX relying on one probe only??? Wasn't that illegal in the first place?

Question nr. 3: From the many times I flew in the cockpit jump seats (or the seat behind the captain) I know that (at least on the 737) there is an aural signal to catch the crew's attention to something and that there is the so-called "cut-off switch" to silence that alerting signal. And there is also another aural warning (that one I fortunately never heard...) which goes: "Terrain, pull up! Terrain, pull up!". Being so, why had Boeing that trouble of developing the MCAS and simply did not another aural warning in the cockpit saying something like "Attitude! Attitude!"??

strfyr51 wrote:

the MCAS system was an angle of attack based system. It might have worked flawlessly had BOTH AOA sensors had comparators and independent operation. BUT? the VP in charge didn't see that as he was only thinking of the number of airplanes he could turn out assuming they were safe. He then sold those airplanes to carriers who had NO 737 experience. CASE in point? the MCAS airplanes had no problems reported at airlines who had flown 737's previously and they banked that the carriers who were new to the 737 wouldn't know the difference, so they did what it did led down the primrose path. with the MCAS system being tied to a single AOA sensor when anybody who ever worked on a 737 knew the loss of heat function or a bird strike can take out a sensor? and the MCAS was only tied to ONE sensor? they lost redundancy and the MCAS was ineffective as it gave false inputs and warnings. and what was part of the fix? to put switches on the pedestal to cut out the MCAS to hand fly the airplane.

Question nr. 1: Isn't every pilot trained from the beginning of his (long) career on until becoming able (or allowed) to seat himself on the nr. 2 seat of the cockpit (not even to speak of the nr. 1...) to fly a commercial plane, that pitch and speed are the most important parameters to have in mind?

Question nr. 2: We all know what redundancy is and that (if I'm not mistaken) every FMS relies on three independent systems which feed the FCC

1. Did some of the other airlines flying the MAX at those times (2019, I think) experience the same problem as well?

strfyr51 wrote:

cargil48 wrote:

Guys, hi

Before everybody jumps at me going back to this (in)famous topic, I still have some questions regarding the flight controls in general and how the market has gradually responded by increasing automatization in the FMS. My main point arose after reading an article the subject of which caught my attention: "Will we ever have commercial airplanes without pilots in the cockpit?" Obviously this led my brain reviewing all I had read about the MAX problem we all know about and what the industry has learned from it. Probably thousands of lines have been written already about it, spread everywhere in the internet, but I thought there are still some points maybe worthwhile reviewing and condensing in short answers for all those who are not too familiar with what happed and how it should have been avoided in the first place. First of all, I have some knowledge about aviation and airplanes, generally speaking and this means I'm not a pilot. Not even of an UL...but I have years of R/C big airplanes (up to 1/3d scale), building them and flying them, so I'm familiar with the techniques of flying in general and how planes are made.

Now, recapitulating: "The Maneuvering Characteristics Augmentation System (MCAS) is a flight control law managed by the flight control computer (FCC) and introduced on the 737 MAX to help it handle like a 737 Next Generation (NG), particularly at slow speeds and high angles of attack (AOA)." (from an article on Aviation Week).

Question nr. 1: Isn't every pilot trained from the beginning of his (long) career on until becoming able (or allowed) to seat himself on the nr. 2 seat of the cockpit (not even to speak of the nr. 1...) to fly a commercial plane, that pitch and speed are the most important parameters to have in mind? Every R/C piloting guy knows that if he does not have that in mind he looses his toy!... Being so, when a pilot is flying in a situation of low speed and higher AoA than on normal cruise isn't it "natural" that he keeps every five seconds or so looking at the VSI and at the speed indicator?

Question nr. 2: We all know what redundancy is and that (if I'm not mistaken) every FMS relies on three independent systems which feed the FCC with data to allow the FMS to make the correct decisions? If so, how could Boeing have introduced a new additional system to the FMS of the MAX relying on one probe only??? Wasn't that illegal in the first place?

Question nr. 3: From the many times I flew in the cockpit jump seats (or the seat behind the captain) I know that (at least on the 737) there is an aural signal to catch the crew's attention to something and that there is the so-called "cut-off switch" to silence that alerting signal. And there is also another aural warning (that one I fortunately never heard...) which goes: "Terrain, pull up! Terrain, pull up!". Being so, why had Boeing that trouble of developing the MCAS and simply did not another aural warning in the cockpit saying something like "Attitude! Attitude!"??

the MCAS system was an angle of attack based system. It might have worked flawlessly had BOTH AOA sensors had comparators and independent operation. BUT? the VP in charge didn't see that as he was only thinking of the number of airplanes he could turn out assuming they were safe. He then sold those airplanes to carriers who had NO 737 experience. CASE in point? the MCAS airplanes had no problems reported at airlines who had flown 737's previously and they banked that the carriers who were new to the 737 wouldn't know the difference, so they did what it did led down the primrose path. with the MCAS system being tied to a single AOA sensor when anybody who ever worked on a 737 knew the loss of heat function or a bird strike can take out a sensor? and the MCAS was only tied to ONE sensor? they lost redundancy and the MCAS was ineffective as it gave false inputs and warnings. and what was part of the fix? to put switches on the pedestal to cut out the MCAS to hand fly the airplane.

kalvado wrote:

Question nr. 1: Isn't every pilot trained from the beginning of his (long) career on until becoming able (or allowed) to seat himself on the nr. 2 seat of the cockpit (not even to speak of the nr. 1...) to fly a commercial plane, that pitch and speed are the most important parameters to have in mind?

Problem MCAS was solving was improper feedback onto the yoke. In 737, that feedback is provided by a specialized hydraulic computer, and I would assume modifying that would be the proper way of dealing with the problem. For whatever reason Boeing didn't do that. Basic problem from testing was "plane doesn't feel right" in a corner of envelope, defeating training approaches.

Question nr. 2: We all know what redundancy is and that (if I'm not mistaken) every FMS relies on three independent systems which feed the FCC

not on 737. IIRC 2 independent systems without majority vote.

1. Did some of the other airlines flying the MAX at those times (2019, I think) experience the same problem as well?

Total of 3 events, 2 sequential on 1 plane. First crew on Lion plane - actually a jumpseater who had a luxury to analyze situation as things progressed - found the solution. Overall MAX had laughable amount of flight time, like 100k hours, by the time it was grounded.

cargil48 wrote:

Question nr. 2: We all know what redundancy is and that (if I'm not mistaken) every FMS relies on three independent systems which feed the FCC with data to allow the FMS to make the correct decisions? If so, how could Boeing have introduced a new additional system to the FMS of the MAX relying on one probe only??? Wasn't that illegal in the first place?

Question nr. 3: From the many times I flew in the cockpit jump seats (or the seat behind the captain) I know that (at least on the 737) there is an aural signal to catch the crew's attention to something and that there is the so-called "cut-off switch" to silence that alerting signal. And there is also another aural warning (that one I fortunately never heard...) which goes: "Terrain, pull up! Terrain, pull up!". Being so, why had Boeing that trouble of developing the MCAS and simply did not another aural warning in the cockpit saying something like "Attitude! Attitude!"??

cargil48 wrote:

Many thanks to all of you, fantastic amount of detailed answers here!

Avatar2go wrote:

The original sensor had an intermittent fault, that had manifested over a period of 30 days, but only at altitude, and never caused an MCAS activation. The replacement sensor was refurbished, and had been miscalibrated at the shop. It was bad out of the box, but was not tested as per the installation instructions.

The maintenance staff thus reset the alarms and restored the aircraft to service, as they were accustomed to doing. So the next crew would experience the exact same issue the next day, and that flight would be fatal.

Since the aircraft was new and under warranty, Boeing would have overnighted a new AoA sensor at no charge. But the airline did not want to lose a day of service, so used the refurbished sensor instead.

cargil48 wrote:

Avatar2go wrote:

The original sensor had an intermittent fault, that had manifested over a period of 30 days, but only at altitude, and never caused an MCAS activation. The replacement sensor was refurbished, and had been miscalibrated at the shop. It was bad out of the box, but was not tested as per the installation instructions.

The maintenance staff thus reset the alarms and restored the aircraft to service, as they were accustomed to doing. So the next crew would experience the exact same issue the next day, and that flight would be fatal.

Since the aircraft was new and under warranty, Boeing would have overnighted a new AoA sensor at no charge. But the airline did not want to lose a day of service, so used the refurbished sensor instead.

Flying is a science, even in the actual times of quite a lot of automatization. But when maintenance isn't done as it should, one can have the best airplane and flown by the best pilot but to no avail... The weakest link is always the one to break...

N1120A wrote:

3) The Lion airplane was not maintained properly, as noted. The ET plane was ordered without the optional second AOA sensor and indicator light - safety articles that should have not been optional, but also should have been ordered.

N1120A wrote:

4) The ET crew treated the erroneous MCAS activation as trained - by doing a runaway trim procedure. That didn't work, however, and the airplane still killed them.

N1120A wrote:

What likely would have saved them would have been an extremely counter productive procedure out of the very high ADD airport - reducing thrust to reduce the AOA to a point where MCAS would stop pushing the nose down. They then could have engaged the autopilot and kept flying, as MCAS deactivates with the autopilot on. That said, it shouldn't have even gone that far.

N1120A wrote:

2) Unlike with Lion, there has never really been a doubt about the quality of pilot training and manual flying ability at ET, where pilots frequently operate in challenging, visual operations.

4) The ET crew treated the erroneous MCAS activation as trained - by doing a runaway trim procedure. That didn't work, however, and the airplane still killed them. What likely would have saved them would have been an extremely counter productive procedure out of the very high ADD airport - reducing thrust to reduce the AOA to a point where MCAS would stop pushing the nose down. They then could have engaged the autopilot and kept flying, as MCAS deactivates with the autopilot on. That said, it shouldn't have even gone that far.

kalvado wrote:

Runaway trim procedure corrected for MCAS had some fine print optional actions - which turned out to be absolutely vital. Apparently, Boeing didn't try MCAS runaway in the sim after Lion crash, as simulator has a correct representation of the stab blowback effect. See Mentourpiont video for example. Neither they have engineers who understand aerodynamics of 737 by now (a really hair raising thing to say!)

Avatar2go wrote:

N1120A wrote:

2) Unlike with Lion, there has never really been a doubt about the quality of pilot training and manual flying ability at ET, where pilots frequently operate in challenging, visual operations.

It's true that Ethiopian has a much better safety record than Lion Air. However, both BEA and NTSB noted pilot training deficiencies for this accident. But EAIB refused to allow that finding in their report, as required by ICAO, leading to BEA and NTSB issuing their own reports that document the deficiencies.

4) The ET crew treated the erroneous MCAS activation as trained - by doing a runaway trim procedure. That didn't work, however, and the airplane still killed them. What likely would have saved them would have been an extremely counter productive procedure out of the very high ADD airport - reducing thrust to reduce the AOA to a point where MCAS would stop pushing the nose down. They then could have engaged the autopilot and kept flying, as MCAS deactivates with the autopilot on. That said, it shouldn't have even gone that far.

This is a misconception. The AoA sensor was destroyed at takeoff, so no pitch attitude could have disabled MCAS. Also, the autopilot will not stay engaged in the presence of invalid airspeed and altitude alarms, generated by the inability of the air data computer to correct for AoA. The crew tried several times to engage it, without success.

Also the crew did perform the runaway stabilizer checklist, but were late in doing so, and in the initial confusion, did not reduce throttle from the 94% takeoff setting. This caused the aircraft to overspeed, even while in a steady climb, beyond both the maximum safe operational speed and the maximum airframe rated speed. The throttles remained at 94% until impact.

At those speeds, the manual trim wheels did not work, so while the checklist was successful in disabling the stabilizer motor & MCAS, it was unsuccessful in relieving the high control column forces. The crew then decided to reverse the checklist and re-enable the stabilizer motor. At that point they were overwhelmed by the column forces.

For reasons we will never know, although the crew reactivated the stabilizer motor in order to use electric trim, they did not subsequently use it to relieve the column forces. Nor did they acknowledge or respond to the overspeed warning alerts.

Reducing throttle and airspeed would have enabled the manual trim again. The column & trim forces increase with the square of airspeed.

I point these things out, not to "blame the pilots", but because it's essential to understand the facts of what actually happened. None of us know how we will respond until we are in the moment.

kalvado wrote:

Another aspect of it - you are describing it from the office chair perspective. Entire ET flight was 6 minutes, most of that time one of the pilots would be struggling with 100 lb load on the yoke, the other would be trying to turn the trim wheel - which barely moves under those conditions. Alarm, stick shaker, and no indicator for the root cause...
Yes, there were survivable ways out of their situation. Those became totally obvious within a week from the crash, once the blowback issue was properly understood.

IADFCO wrote:

In particular, one can check whether the assumption of a 4-second pilot reaction time is critical. Thousands or tens of thousands of simulations can be easily run in a day on a modern workstation, and can point out critical situations to be checked by real test pilots in flight.

kalvado wrote:

What would have saved them (possibly) is going by the book, going to unreliable airspeed procedure (pitch and thrust) without raising flaps. Terrain gamble would make it interesting as climb or descent are not guaranteed in that situation. Raising flaps wouldn't cause a problem in NG.

N1120A wrote:

There seem to be many misconceptions here.(...)

kalvado wrote:

N1120A wrote:

3) The Lion airplane was not maintained properly, as noted. The ET plane was ordered without the optional second AOA sensor and indicator light - safety articles that should have not been optional, but also should have been ordered.

ET plane was ordered without AOA indicator, but with non-optional failure indicator. The way software was designed, though, non-optional feature was tied to an optional one, and didn't work without a paid option.

N1120A wrote:

.

ArcticFlyer wrote:

Runaway stabilizer was the correct procedure to follow (which the ET crew did initially) and if executed promptly would have saved both flights. Delayed (or, in the case of JT610, lack of) accomplishment of the procedure resulted in increased control loading which would have required the crews to relax the controls - and quickly give up precious altitude - in order to trim manually. The technique was essentially to let go of the yoke, PM cranks the trim wheel as fast as he can until altitude becomes critical, then PF (and possibly PM as well) pulls up to hopefully establish a climb until you have enough altitude to repeat. When I flew the 200 we trained this technique in the sim but for some reason Boeing no longer prescribed it starting with the 737CL.

cargil48 wrote:

kalvado wrote:

N1120A wrote:

3) The Lion airplane was not maintained properly, as noted. The ET plane was ordered without the optional second AOA sensor and indicator light - safety articles that should have not been optional, but also should have been ordered.

ET plane was ordered without AOA indicator, but with non-optional failure indicator. The way software was designed, though, non-optional feature was tied to an optional one, and didn't work without a paid option.

N1120A wrote:

.

Wow... Things get now really weird, here! I think it isn't necessary to say why. Too many options here for the (airline) customer to opt or to reject parameters (or systems) essential for flight safety.

In addition, in August 2017, Boeing engineers identified that not all 737 MAX 8 aircraft were equipped with an Angle-of-Attack (AOA) disagree alert3 despite intending for it to be standard for the fleet. While Boeing included this issue in updated certification documents in October 2017, it did not directly notify FAA of the issue. Ultimately, in February 2019, FAA reviewed Boeing’s decision and agreed that the alert was not necessary for the safe operation of the airplane

ArcticFlyer wrote:

Unreliable airspeed was not the problem here; every single cockpit instrument in both accident aircraft correctly presented the conditions the crews faced.

kalvado wrote:

ArcticFlyer wrote:

Unreliable airspeed was not the problem here; every single cockpit instrument in both accident aircraft correctly presented the conditions the crews faced.

AoA, while not directly presented to the crew, is a part of an air data set used to compute presented parameters. Actually per AoA disagree procedure:
1 Airspeed errors and the IAS DISAGREE alert may occur.
2 Altimeter errors and the ALT DISAGREE alert may occur.
And at least in ET302 captain's stick shaker was activated.

Talk about too much on the plate...

Avatar2go wrote:

After these accidents, the FAA did extensive simulator testing, with real air crews. They were surprised to find the extent to which the crews all handled the same problem differently, and mostly not in specific accordance with their training.

The result of that was the realization that you cannot design for all possible crew responses, there are simply too many. Nor can you design solely for the trained and expected response, as Boeing did. Instead, you have to design for tolerance to crew responses. For the cases where the crew doesn't respond at all, or responds incorrectly.

These are among the changes we now see in the regulatory certification environment. It's now expected that certification documentation will demonstrate that stability of flight will be maintained in the absence of the correct response, within the limits of physics and reason.

They were surprised to find the extent to which the crews all handled the same problem differently, and mostly not in specific accordance with their training.

cargil48 wrote:

Avatar2go wrote:

After these accidents, the FAA did extensive simulator testing, with real air crews. They were surprised to find the extent to which the crews all handled the same problem differently, and mostly not in specific accordance with their training.

The result of that was the realization that you cannot design for all possible crew responses, there are simply too many. Nor can you design solely for the trained and expected response, as Boeing did. Instead, you have to design for tolerance to crew responses. For the cases where the crew doesn't respond at all, or responds incorrectly.

These are among the changes we now see in the regulatory certification environment. It's now expected that certification documentation will demonstrate that stability of flight will be maintained in the absence of the correct response, within the limits of physics and reason.

I'm baffled to read this!

They were surprised to find the extent to which the crews all handled the same problem differently, and mostly not in specific accordance with their training.

Being so... which degree of confidence can one have in the actual aeronautical industry and the standards they apply? Because, friends, let's go straight to the point: What matters in all these problems is not this or that in technical questions, but to give the users of commercial airlines full confidence in the company they use to travel, the respective equipment and all the staff involved in flying an aircraft full of people from A to B. Period.

cargil48 wrote:

Avatar2go wrote:

After these accidents, the FAA did extensive simulator testing, with real air crews. They were surprised to find the extent to which the crews all handled the same problem differently, and mostly not in specific accordance with their training.

The result of that was the realization that you cannot design for all possible crew responses, there are simply too many. Nor can you design solely for the trained and expected response, as Boeing did. Instead, you have to design for tolerance to crew responses. For the cases where the crew doesn't respond at all, or responds incorrectly.

These are among the changes we now see in the regulatory certification environment. It's now expected that certification documentation will demonstrate that stability of flight will be maintained in the absence of the correct response, within the limits of physics and reason.

I'm baffled to read this!

They were surprised to find the extent to which the crews all handled the same problem differently, and mostly not in specific accordance with their training.

Being so... which degree of confidence can one have in the actual aeronautical industry and the standards they apply? Because, friends, let's go straight to the point: What matters in all these problems is not this or that in technical questions, but give the users of commercial airlines full confidence in the company they use to travel, the respective equipment and all the staff involved in flying an aircraft gull of people from A to B. Period.

cargil48 wrote:

This reminds me that tragic accident of AF447 and how it was later seen that nobody in the cockpit knew anything about the last thing to do: Fly by P & P when you don't know what's happening!...

kalvado wrote:

cargil48 wrote:

Avatar2go wrote:

After these accidents, the FAA did extensive simulator testing, with real air crews. They were surprised to find the extent to which the crews all handled the same problem differently, and mostly not in specific accordance with their training.

The result of that was the realization that you cannot design for all possible crew responses, there are simply too many. Nor can you design solely for the trained and expected response, as Boeing did. Instead, you have to design for tolerance to crew responses. For the cases where the crew doesn't respond at all, or responds incorrectly.

These are among the changes we now see in the regulatory certification environment. It's now expected that certification documentation will demonstrate that stability of flight will be maintained in the absence of the correct response, within the limits of physics and reason.

I'm baffled to read this!

They were surprised to find the extent to which the crews all handled the same problem differently, and mostly not in specific accordance with their training.

Being so... which degree of confidence can one have in the actual aeronautical industry and the standards they apply? Because, friends, let's go straight to the point: What matters in all these problems is not this or that in technical questions, but give the users of commercial airlines full confidence in the company they use to travel, the respective equipment and all the staff involved in flying an aircraft gull of people from A to B. Period.

Statistically, risk of dying in a crash on the drive to the airport is close, if not higher, than the risk of flight itself.
And it was long since known (to everyone except FAA) that humans are pretty unreliable under stress, nothing new here. In most cases, there is enough wiggle room to handle suboptimal responses.
Specific feature of MCAS (or infamous Toyota acceleration) is that company broke many rules to get the product out of the door, without realizing it's not a Hollywood movie and happy ending is not guaranteed.

cargil48 wrote:

Being so... which degree of confidence can one have in the actual aeronautical industry and the standards they apply? Because, friends, let's go straight to the point: What matters in all these problems is not this or that in technical questions, but give the users of commercial airlines full confidence in the company they use to travel, the respective equipment and all the staff involved in flying an aircraft full of people from A to B. Period.

GalaxyFlyer wrote:

kalvado wrote:

cargil48 wrote:

I'm baffled to read this!


Being so... which degree of confidence can one have in the actual aeronautical industry and the standards they apply? Because, friends, let's go straight to the point: What matters in all these problems is not this or that in technical questions, but give the users of commercial airlines full confidence in the company they use to travel, the respective equipment and all the staff involved in flying an aircraft gull of people from A to B. Period.

Statistically, risk of dying in a crash on the drive to the airport is close, if not higher, than the risk of flight itself.
And it was long since known (to everyone except FAA) that humans are pretty unreliable under stress, nothing new here. In most cases, there is enough wiggle room to handle suboptimal responses.
Specific feature of MCAS (or infamous Toyota acceleration) is that company broke many rules to get the product out of the door, without realizing it's not a Hollywood movie and happy ending is not guaranteed.

The statistics might be true, but in my 70 years, I’ve known precisely one person killed in a car accident and, at last count, 15 in plane crashes, both civil and military.

GalaxyFlyer wrote:

I didn’t say they were statistically valid, just personal experience. I have no idea how many hours the hundreds of pilots I’ve known flew, but it’s a pretty big number. Certainly the miles driven anmong those people is in the millions. And my experience is shared by many brethren.

True, the fatal accident rate outside 121 is approximately that of vehicle rates, but comparisons are hard.

flipdewaf wrote:

cargil48 wrote:

This reminds me that tragic accident of AF447 and how it was later seen that nobody in the cockpit knew anything about the last thing to do: Fly by P & P when you don't know what's happening!...

How do you know when you don’t know anymore?

Fred


Sent from my iPhone using Tapatalk

cargil48 wrote:

flipdewaf wrote:

cargil48 wrote:

This reminds me that tragic accident of AF447 and how it was later seen that nobody in the cockpit knew anything about the last thing to do: Fly by P & P when you don't know what's happening!...

How do you know when you don’t know anymore?

Fred


Sent from my iPhone using Tapatalk

Because I read carefully the transcripts of the CVR after the "black boxes" were found and analyzed. It also was said that while the captain had gone to the rest room, the second-in-command and the 1st officer were doing each one its own procedure trying to get the plane under control. And, commenting that, someone on the pro side said that what I quoted above, P&P being the safest thing to do when something really serious occurs with the complete instrumentation. As far as - I do add - the A/P is disengaged but in the AF447 case it was automatically due to sudden lack of info from the pitot tubes.

flipdewaf wrote:

cargil48 wrote:

Because I read carefully the transcripts of the CVR after the "black boxes" were found and analyzed. It also was said that while the captain had gone to the rest room, the second-in-command and the 1st officer were doing each one its own procedure trying to get the plane under control. And, commenting that, someone on the pro side said that what I quoted above, P&P being the safest thing to do when something really serious occurs with the complete instrumentation. As far as - I do add - the A/P is disengaged but in the AF447 case it was automatically due to sudden lack of info from the pitot tubes.

Not what I mean, how does one recognise that they (them selves) don’t know what’s happening? Not how do we analyse another’s behaviour in hindsight.

rt23456p wrote:

Even on the Ethiopian airline flight, the pilot already turned off elec hydraulic(which MCAS was tied to), and attempt to manually trim the aircraft back from diving, which was impossible due to the state of the aircraft. If there was a separate switch for MCAS, he would have saved the plane.

flipdewaf wrote:

cargil48 wrote:

flipdewaf wrote:

How do you know when you don’t know anymore?

Fred


Sent from my iPhone using Tapatalk

Because I read carefully the transcripts of the CVR after the "black boxes" were found and analyzed. It also was said that while the captain had gone to the rest room, the second-in-command and the 1st officer were doing each one its own procedure trying to get the plane under control. And, commenting that, someone on the pro side said that what I quoted above, P&P being the safest thing to do when something really serious occurs with the complete instrumentation. As far as - I do add - the A/P is disengaged but in the AF447 case it was automatically due to sudden lack of info from the pitot tubes.

Not what I mean, how does one recognise that they (them selves) don’t know what’s happening? Not how do we analyse another’s behaviour in hindsight.

Fred


Sent from my iPhone using Tapatalk

kalvado wrote:

Re AF447: If you have to resort to 15 year old case as an example of how bad things are.... Are they really bad at all?
Can you recall any automobile problems more than 10 year old? Mind you, road mortality in US is on par with gun mortality, and is way way higher than air accident mortality...

cargil48 wrote:

flipdewaf wrote:

cargil48 wrote:

Because I read carefully the transcripts of the CVR after the "black boxes" were found and analyzed. It also was said that while the captain had gone to the rest room, the second-in-command and the 1st officer were doing each one its own procedure trying to get the plane under control. And, commenting that, someone on the pro side said that what I quoted above, P&P being the safest thing to do when something really serious occurs with the complete instrumentation. As far as - I do add - the A/P is disengaged but in the AF447 case it was automatically due to sudden lack of info from the pitot tubes.

Not what I mean, how does one recognise that they (them selves) don’t know what’s happening? Not how do we analyse another’s behaviour in hindsight.

Fred


Sent from my iPhone using Tapatalk

That is a long story because it is necessary to go back to when the events happened. The flight crew suddenly had a blank panel in front of them, because the FMC lost the signals from the pitot tubes. It was at night and they were beginning to get into turbulence due to entering the intertropical convergence zone and something never happened before happened suddenly: Due to that sudden absence of input from the pitot tubes, the FMC was from one moment to the other not able to generate data to the FMS and made all the navigational panels get dark. That is where that thing enters which you ask: "how does one recognize that they (them selves) don’t know what’s happening?" Because they suddenly lost all navigational parameter, the A/P and A/T disengaged and they both felt "alone" without knowing what to do (because AF never trained its pilots to such a situation).

Starlionblue wrote:

cargil48 wrote:

flipdewaf wrote:

Not what I mean, how does one recognise that they (them selves) don’t know what’s happening? Not how do we analyse another’s behaviour in hindsight.

Fred


Sent from my iPhone using Tapatalk

That is a long story because it is necessary to go back to when the events happened. The flight crew suddenly had a blank panel in front of them, because the FMC lost the signals from the pitot tubes. It was at night and they were beginning to get into turbulence due to entering the intertropical convergence zone and something never happened before happened suddenly: Due to that sudden absence of input from the pitot tubes, the FMC was from one moment to the other not able to generate data to the FMS and made all the navigational panels get dark. That is where that thing enters which you ask: "how does one recognize that they (them selves) don’t know what’s happening?" Because they suddenly lost all navigational parameter, the A/P and A/T disengaged and they both felt "alone" without knowing what to do (because AF never trained its pilots to such a situation).

It's easy to judge the pilots in hindsight at zero airspeed and zero altitude. In the AF447 case, they were confronted by something they had never seen before, at night, in stormy weather, most likely when fatigued. It's a bit like going into the simulator at three in the morning. There's no way you'll perform as well as you would mid morning.

The screens did not "go dark" and the aircraft did not "lose all navigational parameters". That's not how the instruments and displays work. Speed data was lost. IIRC, everything else, including, inertial navigation and GPS were both still working, attitude indication was still working, and so was altitude. If a type lost all the displays from a simple proble failure the type would not be certified.

Even disregarding pitch and power, the pilots failed to do that most crucial thing when something happens and the situation is otherwise stable. Sit on your hands and think for a moment. There are very few situations in a modern airliner where an instinctive and rapid response is advisable.

flipdewaf wrote:

Starlionblue wrote:

cargil48 wrote:

That is a long story because it is necessary to go back to when the events happened. The flight crew suddenly had a blank panel in front of them, because the FMC lost the signals from the pitot tubes. It was at night and they were beginning to get into turbulence due to entering the intertropical convergence zone and something never happened before happened suddenly: Due to that sudden absence of input from the pitot tubes, the FMC was from one moment to the other not able to generate data to the FMS and made all the navigational panels get dark. That is where that thing enters which you ask: "how does one recognize that they (them selves) don’t know what’s happening?" Because they suddenly lost all navigational parameter, the A/P and A/T disengaged and they both felt "alone" without knowing what to do (because AF never trained its pilots to such a situation).

It's easy to judge the pilots in hindsight at zero airspeed and zero altitude. In the AF447 case, they were confronted by something they had never seen before, at night, in stormy weather, most likely when fatigued. It's a bit like going into the simulator at three in the morning. There's no way you'll perform as well as you would mid morning.

The screens did not "go dark" and the aircraft did not "lose all navigational parameters". That's not how the instruments and displays work. Speed data was lost. IIRC, everything else, including, inertial navigation and GPS were both still working, attitude indication was still working, and so was altitude. If a type lost all the displays from a simple proble failure the type would not be certified.

Even disregarding pitch and power, the pilots failed to do that most crucial thing when something happens and the situation is otherwise stable. Sit on your hands and think for a moment. There are very few situations in a modern airliner where an instinctive and rapid response is advisable.

Yes, first thing to do is take a deep breath…instinct is often wrong.

Out of curiosity do you ever do simulator sessions when tired or deliberately not functioning to understand effects for human factors first hand?

Fred


Sent from my iPhone using Tapatalk

Starlionblue wrote:

flipdewaf wrote:

Starlionblue wrote:

It's easy to judge the pilots in hindsight at zero airspeed and zero altitude. In the AF447 case, they were confronted by something they had never seen before, at night, in stormy weather, most likely when fatigued. It's a bit like going into the simulator at three in the morning. There's no way you'll perform as well as you would mid morning.

The screens did not "go dark" and the aircraft did not "lose all navigational parameters". That's not how the instruments and displays work. Speed data was lost. IIRC, everything else, including, inertial navigation and GPS were both still working, attitude indication was still working, and so was altitude. If a type lost all the displays from a simple proble failure the type would not be certified.

Even disregarding pitch and power, the pilots failed to do that most crucial thing when something happens and the situation is otherwise stable. Sit on your hands and think for a moment. There are very few situations in a modern airliner where an instinctive and rapid response is advisable.

Yes, first thing to do is take a deep breath…instinct is often wrong.

Out of curiosity do you ever do simulator sessions when tired or deliberately not functioning to understand effects for human factors first hand?

Fred


Sent from my iPhone using Tapatalk

In our operation, we do not do such sim sessions. But AFAIK there have been numerous studies on this. Also sims when the pilots have been drinking alcohol, to study the effects.

IADFCO wrote:

Starlionblue wrote:

flipdewaf wrote:

Yes, first thing to do is take a deep breath…instinct is often wrong.

Out of curiosity do you ever do simulator sessions when tired or deliberately not functioning to understand effects for human factors first hand?

Fred


Sent from my iPhone using Tapatalk

In our operation, we do not do such sim sessions. But AFAIK there have been numerous studies on this. Also sims when the pilots have been drinking alcohol, to study the effects.

Yes, there is plenty of literature on the topic.

In a graduate flight dynamics class the professor showed us the effects of coffee and alcohol on the pilot using Bode plots. In a nutshell, coffee increases the pilot's gain, and alcohol introduces a delay. Both are destabilizing to the coupled pilot-aircraft system. If the Bode plot of the response to pilot inputs (e.g., p to lateral stick or q to lon) doesn't look right, e.g., it is flat between, say, 0.5 and 10 rad/sec, the whole thing can go unstable. BTW, this is what happened with lots of airplanes early in the development, e.g., F-16, Tornado, YF-23, etc. Handling qualities specs guard against this for military airplanes, but not for commercial airplanes.

flipdewaf wrote:

IADFCO wrote:

Starlionblue wrote:

In our operation, we do not do such sim sessions. But AFAIK there have been numerous studies on this. Also sims when the pilots have been drinking alcohol, to study the effects.

Yes, there is plenty of literature on the topic.

In a graduate flight dynamics class the professor showed us the effects of coffee and alcohol on the pilot using Bode plots. In a nutshell, coffee increases the pilot's gain, and alcohol introduces a delay. Both are destabilizing to the coupled pilot-aircraft system. If the Bode plot of the response to pilot inputs (e.g., p to lateral stick or q to lon) doesn't look right, e.g., it is flat between, say, 0.5 and 10 rad/sec, the whole thing can go unstable. BTW, this is what happened with lots of airplanes early in the development, e.g., F-16, Tornado, YF-23, etc. Handling qualities specs guard against this for military airplanes, but not for commercial airplanes.

I’d guess there was plenty of studies but first hand it’s maybe enlightening for pilots but not that useful.

That’s very interesting about the control rates the increased feedback time (changed human response) makes the control loop unstable. It makes perfect sense when it’s pointed out but not something one would immediately think about.

Fred


Sent from my iPhone using Tapatalk