Ethiopian airliner down in Africa
Join Date: Feb 2015
Location: The woods
Posts: 0
Originally Posted by GordonR_Cape
A feel augmentation system that takes AOA as input, what could possibly go wrong? I'm sure others can give scenarios where this would be a bad idea...
I previously pointed out that any system that takes a small input such as AOA, and produces a large output (whether feel or trim), is inherently undesirable and unpredictable. MCAS creates a semblance of smoothness by applying trim over a 10 second interval. Applying such a delay into elevator feel would create all kinds of feedback lags, and the solution could be worse than the problem.
The cost issue with any non-software change is how many years that would take to design, test and certify, and what production would be done in the meantime? Boeing have painted themselves into a corner, and the only way out seems to be double or quits (to mix metaphors).
I previously pointed out that any system that takes a small input such as AOA, and produces a large output (whether feel or trim), is inherently undesirable and unpredictable. MCAS creates a semblance of smoothness by applying trim over a 10 second interval. Applying such a delay into elevator feel would create all kinds of feedback lags, and the solution could be worse than the problem.
The cost issue with any non-software change is how many years that would take to design, test and certify, and what production would be done in the meantime? Boeing have painted themselves into a corner, and the only way out seems to be double or quits (to mix metaphors).
Although I didn’t mention AoA as an input in that post, if you need to know AoA in order to correct elevator feel, then clearly you have to measure it. Done properly (dual input and monitor) and maintained correctly it is a valid measurement as used for years on many aircraft.
You could also ask what could possibly go wrong with a pitot system. There are probably more examples of failure in that system and yet it is universally used.
At some point you have to trust something for your data, so make it as reliable as you can and maintain it well.
As for feedback lags and delays, a spring is a wonderful reliable input which can be linearly or dynamically calibrated, and is used in many control run applications. The ten second interval used by MCAS would not apply - as long as the condition remained, the spring would stay compressed. As the condition decreased, so would the spring force. Qed. There are other force generators available, based on hydraulics or pneumatics although a spring is simple and less reliant on other systems.
Main thing is to have the force generated within the control run rather than by potentially powerful MCAS stab. movement (which still needs AoA input).
B
Join Date: Apr 2019
Location: USA
Posts: 217
Originally Posted by Bend alot
You do understand MCAS is a requirement for flight within certification requirements?
With MCAS disabled (due now to any of a number of reasons) - how is flight within certification requirement limits meet?
Nothing at all to do with any trim event (Important you understand that) but what is flight like outside the certifiable limits?
With MCAS disabled (due now to any of a number of reasons) - how is flight within certification requirement limits meet?
Nothing at all to do with any trim event (Important you understand that) but what is flight like outside the certifiable limits?
There are are a whole lot of systems that can and have failed on numerous aircraft that technically put that aircraft out of its limits for certification. You would never depart with failed engines, failed hydraulic systems, failed tires, etc. However, they do and have failed, and we have procedures in place to deal with those failures.
To your specific question of how one could possibly fly the 737 with a known MCAS failure once they fix the runaway stab thingy....., well how about staying away from the edge of the envelope and not stalling the aircraft?
Last edited by 737 Driver; 27th Apr 2019 at 18:44.
Join Date: Oct 2017
Location: Tent
Posts: 759
Originally Posted by 737 Driver
Point of order.
There are are a whole lot of systems that can and have failed on numerous aircraft that technically put that aircraft out of its limits for certification. You would never depart with failed engines, failed hydraulic systems, failed tires, etc. However, they do and have failed, and we have procedures in place to deal with those failures.
To your specific question of how one could possibly fly the 737 with a known MCAS failure once they fix the runaway stab thingy....., well how about staying away from the edge of the envelope an not stalling the aircraft?
There are are a whole lot of systems that can and have failed on numerous aircraft that technically put that aircraft out of its limits for certification. You would never depart with failed engines, failed hydraulic systems, failed tires, etc. However, they do and have failed, and we have procedures in place to deal with those failures.
To your specific question of how one could possibly fly the 737 with a known MCAS failure once they fix the runaway stab thingy....., well how about staying away from the edge of the envelope an not stalling the aircraft?
Just no mention of it in the fix.
As you are vocal and seems a very competent 737 pilot (certainly seems that way by your posts)
Can you detail how the/you cockpit would have responded after take off of the ET flight with you as Captain and with the 2-300 hr FO - who's job was who's when and why ?
Not a trap question just very interested.
Starting from the first warning you get and when you would accept it as a warning or a failure.
Thanks for detailed answers in reply (others let him reply and answer - he is clearly a very competent pilot on type)
Join Date: Dec 2015
Location: Cape Town, ZA
Age: 59
Posts: 424
Originally Posted by bill fly
Hi Gordon,
Although I didn’t mention AoA as an input in that post, if you need to know AoA in order to correct elevator feel, then clearly you have to measure it. Done properly (dual input and monitor) and maintained correctly it is a valid measurement as used for years on many aircraft.
You could also ask what could possibly go wrong with a pitot system. There are probably more examples of failure in that system and yet it is universally used.
At some point you have to trust something for your data, so make it as reliable as you can and maintain it well.
As for feedback lags and delays, a spring is a wonderful reliable input which can be linearly or dynamically calibrated, and is used in many control run applications. The ten second interval used by MCAS would not apply - as long as the condition remained, the spring would stay compressed. As the condition decreased, so would the spring force. Qed. There are other force generators available, based on hydraulics or pneumatics although a spring is simple and less reliant on other systems.
Main thing is to have the force generated within the control run rather than by potentially powerful MCAS stab. movement (which still needs AoA input).
B
Although I didn’t mention AoA as an input in that post, if you need to know AoA in order to correct elevator feel, then clearly you have to measure it. Done properly (dual input and monitor) and maintained correctly it is a valid measurement as used for years on many aircraft.
You could also ask what could possibly go wrong with a pitot system. There are probably more examples of failure in that system and yet it is universally used.
At some point you have to trust something for your data, so make it as reliable as you can and maintain it well.
As for feedback lags and delays, a spring is a wonderful reliable input which can be linearly or dynamically calibrated, and is used in many control run applications. The ten second interval used by MCAS would not apply - as long as the condition remained, the spring would stay compressed. As the condition decreased, so would the spring force. Qed. There are other force generators available, based on hydraulics or pneumatics although a spring is simple and less reliant on other systems.
Main thing is to have the force generated within the control run rather than by potentially powerful MCAS stab. movement (which still needs AoA input).
B
An analogy which I drafted in my comment (but then deleted), explains part of the situation: Imagine a motor vehicle driving down a potholed road, with all the irregularities transmitted to the steering wheel. Normally there is some kind of damper to prevent harsh feedback forces to the driver. Imagine that kind of instantaneous feel in the control column while trying to fly an aircraft?
Given the need for some kind of smoothing of AOA, by definition it would involve a delay or lag between pilot inputs and the feel forces. From a control systems theory viewpoint, delay lags are never a good thing, since they can lead to pilot induced oscillation, and other side-effects.
The opposite scenario, of full-force applied when over a threshold, is harsh and equivalent to a stick-pusher, with its own side-effects, training requirements, and type certification issues.
AFAIK MCAS was designed to avoid both scenarios, by being slow and unilateral. It fulfilled the criteria of not being harsh, nor could it induce short-period oscillations.
Sorry, if I took your comment and ran it to a logical absurdity, but IMO it would not be a simple fix. It really can be hard to explain complex control systems feedback, and I'm sure others can do it better,
Join Date: Jan 2008
Location: Reading, UK
Posts: 13,142
Originally Posted by GordonR_Cape
I am not an expert on flight instrumentation, but everything I have read about AOA vanes on the B737 in this thread, indicates that it is not a particularly well suited input parameter for direct command of flight controls.
However I'm not aware of any stick-pusher that acts on the input of a single sensor, with no redundancy.
Stick-Pusher Philosophy
Join Date: Apr 2019
Location: USA
Posts: 217
Originally Posted by wonkazoo
When I shared what I did last night it wasn't to expiate the guilt of the crews, but instead to try and get people to realize that "blaming" people who are no longer here to defend themselves is a bit disingenuous, and equally it distracts us from the very real root causes of the two incidents.
There is one huge difference between my position and that expressed most recently by 737 Driver, and that is to the responsibility for the outcome of what happened. Note that I say responsibility and not blame.
The result of this sad effort was a system that, if it failed, would basically try to kill the pilot and everyone on board. I say again: MCAS will try to kill everyone on board if it fails.
I simply cannot recall (but am inviting others here to fill in the blanks if you can) another system on a transport category aircraft with a failure mode that defaulted to "I'm going to try to fly the airplane into the ground.
These incidents, indeed the entirety of MCAS' existence are a failure of corporate responsibility aided and abetted by a complete abrogation of regulatory responsibility. All in the pursuit of profits for shareholders..........Those are the responsible parties, and that is what I hope people will look carefully at. It starts with the airplane. Build a safe one and operators will still find a way to muck things up, and crews will still make mistakes. But step one, the most important step, is build an effing safe airplane.
This is a forum for professional pilots. Yes, we could sit back, point fingers, and opine about how badly someone else screwed up. Or we could take a hard look at our profession and ask why multiple crews had such difficulty and/or reluctance in applying some very basic airmanship techniques to resolve an aircraft malfunction that, while being unique and baffling, ultimately did not render the aircraft unflyable.
Last edited by 737 Driver; 27th Apr 2019 at 12:08.
Join Date: Jan 2008
Location: Reading, UK
Posts: 13,142
Originally Posted by wonkazoo
I simply cannot recall (but am inviting others here to fill in the blanks if you can) another system on a transport category aircraft with a failure mode that defaulted to "I'm going to try to fly the airplane into the ground.
Safety Recommendations from the investigation report:
Boeing should improve the reliability of the radio altimeter system.
The FAA and EASA should ensure that the undesirable response of the autothrottle and flight management computer caused by incorrect radio altimeter values is evaluated and that the autothrottle and flight management computer is improved in accordance with the design specifications.
The FAA and EASA should ensure that the undesirable response of the autothrottle and flight management computer caused by incorrect radio altimeter values is evaluated and that the autothrottle and flight management computer is improved in accordance with the design specifications.
Join Date: Nov 2004
Location: UK
Posts: 406
Originally Posted by wonkazoo
Boeing designed and placed into service an airplane with an active control system that had unilateral control over the horizontal stabilizer, with enough authority to place the airplane in an unrecoverable state if just a single component failed. Further this system gave no indication to the pilots that it was operating, or when malfunctioning that it was operating in error. Additionally this system, which was created solely to increase the amount of force required to pitch up the aircraft at high AOA used the most critical part of the airframe to do this minimal task, instead of using a passive system that had no control authority.
The result of this sad effort was a system that, if it failed, would basically try to kill the pilot and everyone on board. I say again: MCAS will try to kill everyone on board if it fails.
I simply cannot recall... another system on a transport category aircraft with a failure mode that defaulted to "I'm going to try to fly the airplane into the ground. If you line up all the dots and pull two switches at the right moment I will let you live. Otherwise you die... Oh, and BTW I'm also going to fail concurrently with three or four other systems, which actually will alert you to their issues, unlike me, who will sit here quietly winding your trim forward until you get to the point where you cannot wind it back. Sorry about that!!" (It's also worth noting here that the Emergency AD that was put out only gave instruction on what was essentially an enhanced trim runaway. There is no actual way (that I have seen) for a pilot to actually determine if MCAS is malfunctioning. At best you are to stop the resultant (trim runaway) and remain in ignorance over the state of MCAS. WTF?? A system with complete authority over the horizontal stab and you have no way of knowing anything about it. Failure modes, operational status, errors, nothing. Just "If the airplane is trimming down (for whatever reason) and you don't want it to pull the console switches." Really??)
These incidents, indeed the entirety of MCAS' existence are a failure of corporate responsibility aided and abetted by a complete abrogation of regulatory responsibility.... {But} in the end, the only entity who both could have designed a safe airplane, and who not only failed but by all appearances worked to conceal their failure through omission, was Boeing. And the agency that looked the other way was the FAA.
Those are the responsible parties, and that is what I hope people will look carefully at. It starts with the airplane. Build a safe one and operators will still find a way to muck things up, and crews will still make mistakes. But step one, the most important step, is build an effing safe airplane.
Warm regards,
dce
The result of this sad effort was a system that, if it failed, would basically try to kill the pilot and everyone on board. I say again: MCAS will try to kill everyone on board if it fails.
I simply cannot recall... another system on a transport category aircraft with a failure mode that defaulted to "I'm going to try to fly the airplane into the ground. If you line up all the dots and pull two switches at the right moment I will let you live. Otherwise you die... Oh, and BTW I'm also going to fail concurrently with three or four other systems, which actually will alert you to their issues, unlike me, who will sit here quietly winding your trim forward until you get to the point where you cannot wind it back. Sorry about that!!" (It's also worth noting here that the Emergency AD that was put out only gave instruction on what was essentially an enhanced trim runaway. There is no actual way (that I have seen) for a pilot to actually determine if MCAS is malfunctioning. At best you are to stop the resultant (trim runaway) and remain in ignorance over the state of MCAS. WTF?? A system with complete authority over the horizontal stab and you have no way of knowing anything about it. Failure modes, operational status, errors, nothing. Just "If the airplane is trimming down (for whatever reason) and you don't want it to pull the console switches." Really??)
These incidents, indeed the entirety of MCAS' existence are a failure of corporate responsibility aided and abetted by a complete abrogation of regulatory responsibility.... {But} in the end, the only entity who both could have designed a safe airplane, and who not only failed but by all appearances worked to conceal their failure through omission, was Boeing. And the agency that looked the other way was the FAA.
Those are the responsible parties, and that is what I hope people will look carefully at. It starts with the airplane. Build a safe one and operators will still find a way to muck things up, and crews will still make mistakes. But step one, the most important step, is build an effing safe airplane.
Warm regards,
dce
It counters the apologists' excuses for Boeing's abominable design, and it should silence the criticisms of those who repeatedly try to blame the pilots of those 2 condemned flights / aircraft.
I am sad that it has taken so long for 'dce', 'wonkazoo' to come here to sum this up so perfectly, along with the important lessons for everyone from sharing such a vivid and honest account of their own '***** or bust' moment all those years ago. It takes a special mind to admit to failings with such openness and honesty, and we are all the richer for it. The question is, when will Boeing and the FAA similarly dig deep enough to find similar humility and honesty, and to redress the very clear failings?
Thank you, 'dce' - respect!
Join Date: Apr 2019
Location: USA
Posts: 217
Originally Posted by Bend alot
Can you detail how the/you cockpit would have responded after take off of the ET flight with you as Captain and with the 2-300 hr FO - who's job was who's when and why ?
Not a trap question just very interested.
Starting from the first warning you get and when you would accept it as a warning or a failure.
Not a trap question just very interested.
Starting from the first warning you get and when you would accept it as a warning or a failure.
In the case of Ethiopian, however, the answer is quite straightforward. After Lion Air, the existence and potential failure modes of MCAS were made public. This was a very hot topic around my airline as we operated the MAX. We have our own internal message board (not unlike PPRuNe), and different thoughts were kicked around. While there were some pilots who said they would refuse to fly the MAX until a more permanent fix was in place, the general consensus was that there were some basic techniques that could be used to mitigate the threat.
Since MCAS was inhibited by either the A/P or the flaps extended, the primary defense was to make sure you had one or the other. That is, on takeoff one would engage the A/P first and then retract the flaps. On landing, keep the A/P on until some amount of flaps were extended. If during takeoff, but before A/P engagement, you were to experience anything that looked like a failed AOA or unreliable airspeed, then don't retract the flaps. By applying these techniques, MCAS would never have an opportunity to activate. Whether these kind of discussions occurred at Ethiopian is currently unknown, but the information was available to process.
A better question would have been, "How would you have handled this malfunction as a Lion Air pilot who had no knowledge of MCAS?"
I think by now you may have gathered that I'm a hands-on type of pilot, so the answer is pretty much the same one I've been giving all along. Fly the aircraft.
By the numbers then: Stick shaker. WTF?! Check my power (increase as necessary), check my attitude, check my configuration. Is it flying or is it wallowing? If it is wallowing, keep the nose down and accelerate. If its flying, probably a false indication, continue the climb, call for the gear. Cross check instruments. I've got my hands full, so ask my FO to read off what he sees on all three airspeeds. At 400 feet check my roll mode, have FO ask for straight ahead if appropriate and declare emergency. If by now I've determined we have unreliable airspeed, memory items except I'm going to keep takeoff power and 15 degrees pitch until 1000' where I set 10 degrees and 80% N1.
Now I do absolutely nothing except climb to a safe altitude with flaps hanging. Once at a safe altitude, we proceed slowly and methodically through the NNC for Airspeed Unreliable. Quite frankly, I don't know if I would ever retract my flaps in this scenario (and hence no MCAS issue) because I'm going to return to the departure airport for landing.
If I ever did retract the flaps, it would only be after I had a stabilized aircraft. If MCAS then kicked in, I seriously doubt I would let the trim run continuously for 9 seconds before I did something about it. When you do a fair amount of hand-flying, trimming is like breathing. You hardly think about it. Controls get heavy, trim. Apply thrust, trim. Reduce thrust, trim. Enter a turn, trim. Rollout of turn, trim. If MCAS activated, it would probably take a few cycles of back and forth before I realized that something was amiss, but I have never been reluctant to trim as necessary. Eventually I would have made my way to the runaway trim NNC, but would have done so from an in-trim state.
This is not to say that everything would have been executed flawlessly (i.e. good chance I would forget to call for gear initially), but then that's why you want an experienced First Officer to back you up. Personally, I feel that the Ethiopian Airline policy of placing low-time pilots in the right seat of a passenger airline borders on the criminally negligent.
So, to flog the topic one more time..... When presented with an undesired and/or unexpected aircraft state, it is absolutely crucial for the pilot flying to be prepared to revert to basic airmanship skills. Set the pitch. Set the power. Trim the aircraft. Monitor the performance. Adjust as necessary. Get to a safe altitude. Stabilize the aircraft. And then work the problem.
None of this requires "sky god" or "test pilot" level of skill. It simply requires making the conscience decision that you are going to fly the aircraft with the tools that are readily available.
Join Date: Apr 2019
Location: USA
Posts: 217
Fly the damn aircraft...
Originally Posted by maxter
737 driver: Do you not train for those failures in a sim. I think that was what was being suggested for MCAS failure, not just an Ipad brief
Over the years and through several different organizations, I have seen different variations and phraseology of these basic principles, but every single one of them started with the same first step - FLY THE AIRCRAFT. This is so fundamental and so important that at one previous employer we used to say that the first three steps of any emergency was 1) Fly the aircraft, 2) Fly the aircraft, and 3) Fly the aircraft.
Sadly, this concept has to be repeated frequently because, for various reasons, pilots keep forgetting this touchstone principle. FLY THE AIRCRAFT. First, last, and always.
Join Date: Mar 2013
Location: France
Age: 36
Posts: 105
Hello all.
i just checked one of the previous aircraft I used to fly, equipped with a stick pusher system, this is what I found :
The push force felt at the control column is approximately 80 lbs, which is adequate for stall
correction, but can still be overpowered by the pilot in the event the system cannot be
disabled.
NOTE: The number 1 and number 2 computer must both agree to push before the stick
pusher can be activated.
The stick pusher will activate when all of the following conditions are met:
• #1 STALL SYST FAIL, #2 STALL SYST FAIL and PUSHER SYST FAIL caution lights are
OFF.
• The airplane is airborne.
• The airplane altitude is greater than 400 feet AGL.
• The airspeed is less than 200 -
4 KEAS.
• The stick shakers have been activated by both systems.
The only thing I have difficulty to understand is why Boeing in the first place didn't connect the MCAS system to both AOA, exactly as stick pusher was in my previous aircraft.
now, not writing as long post about it, but the mcas issue is not only a design problems ... We can clearly see a tendency in the world of pilot to loose confidences in manual and basic flying skill in the modern aviation world. I would like to believe authority will not only react to fix the MCAS and that it, it is applicable to all the aviation world and instead of have more and more automation because pilot are more and more crap, we should maybe put back the pilot in the center of the loop and ask / help / train them to have their skill up to the task.
i just checked one of the previous aircraft I used to fly, equipped with a stick pusher system, this is what I found :
The push force felt at the control column is approximately 80 lbs, which is adequate for stall
correction, but can still be overpowered by the pilot in the event the system cannot be
disabled.
NOTE: The number 1 and number 2 computer must both agree to push before the stick
pusher can be activated.
The stick pusher will activate when all of the following conditions are met:
• #1 STALL SYST FAIL, #2 STALL SYST FAIL and PUSHER SYST FAIL caution lights are
OFF.
• The airplane is airborne.
• The airplane altitude is greater than 400 feet AGL.
• The airspeed is less than 200 -
4 KEAS.
• The stick shakers have been activated by both systems.
The only thing I have difficulty to understand is why Boeing in the first place didn't connect the MCAS system to both AOA, exactly as stick pusher was in my previous aircraft.
now, not writing as long post about it, but the mcas issue is not only a design problems ... We can clearly see a tendency in the world of pilot to loose confidences in manual and basic flying skill in the modern aviation world. I would like to believe authority will not only react to fix the MCAS and that it, it is applicable to all the aviation world and instead of have more and more automation because pilot are more and more crap, we should maybe put back the pilot in the center of the loop and ask / help / train them to have their skill up to the task.
Join Date: Jun 2003
Location: UK
Posts: 2,524
Originally Posted by 737 Driver
some very basic airmanship technique
We get it. You think they were sub-standard. You would have used your greater "airmanship" to keep the basic parameters in line and therefore the aircraft flying.
Your lack of appreciation of human factors in a real scenario with real line pilots in the real world is what you should address. After-the-event-heroism is ever so easy.
My suggestion is to slice the event up into small time slices. In each slice - look at the information presented, the probable solution, the required actions and the workload. Then in the next slice, base your model on a modification of the previous slice - increasing or decreasing the factors, and extrapolating the data flows. So if it starts with a stall indication, then keep the "it's a stall" flowing, until (given the workload estimate) the information is sufficient to demonstrate to the pilots that it's not a stall. Doing this, you will discover that the pilot with reduced workload will process new information earlier.
In my model, the captain wanted to get the thing cleaned up, then reduce the workload, then work out what was wrong. They cleaned up - he used the a/p to reduce the workload, and as they were starting to diagnose the aircraft bit them. His workload reduced his SA (along with the stick shaker) to a critical point. The pilot failed due to excess workload, not due to lack of "airmanship".
Join Date: Jun 2003
Location: UK
Posts: 2,524
Originally Posted by formulaben
I am not rude and I asked 3 times...thank you for finally answering that you have no clue how much more training is required.
This is about:
Systems knowledge
Recognition
Recovery
Systems knowledge
A document/ipad/manual will give the systems knowledge.
Recognition
If the new software were to have an aural alert "MCAS" whenever it triggered, then this would make recognition easy. Because if you are in a phase of flight that is normal, and the machine cries "MCAS", then you know it's a fault and quickly flick the switches. This part would not require sim training.
If the new software does not alert you, then recognition has been demonstrated by two crews as difficult, and therefore this would require sim training. Maybe 3 or 4 scenarios based on different phases of flight and different erroneous inputs.
Recovery
The recovery may require sim training regardless of the above. If you're in a benign state and MCAS activates (and you recognise) then recovery is not hard, and has been previously trained. But what if you are in an expeditious descent (say, 330 knots) and MCAS triggers? You flick the switches. But now you are probably in a slight overspeed, pointing the wrong way, holding against the trim and are faced with hand-trimming. I would suggest that I would like to practice that in the sim, and I'd like the crew of the Max I am in the back of to have also practised that - before I get on board.
So however you look at it, sim training is required. Probably just 1 hour per crew, but 1 hour none the less.
Anything less, I'm not going.
Join Date: Apr 2008
Location: Paris
Age: 71
Posts: 275
Originally Posted by threemiles
A single sensor failure of any kind on any airliner happens every day and is a no-brainer. Except on the MAX, thanks to Boeing and FAA.
Thanks. Can't say it better.
Thanks. Can't say it better.
The design of the 737 Max MCAS with two AoA sensors present and only one input used is *wierd*
Edmund
Join Date: Mar 2018
Location: Central UK
Posts: 560
Originally Posted by Bend alot
Perfect answer!
Just no mention of it in the fix.
As you are vocal and seems a very competent 737 pilot (certainly seems that way by your posts)
Can you detail how the/you cockpit would have responded after take off of the ET flight with you as Captain and with the 2-300 hr FO - who's job was who's when and why ?
Just no mention of it in the fix.
As you are vocal and seems a very competent 737 pilot (certainly seems that way by your posts)
Can you detail how the/you cockpit would have responded after take off of the ET flight with you as Captain and with the 2-300 hr FO - who's job was who's when and why ?
Vee one, Rotate! Stickshaker sounds on rotation. Master Caution and other assorted warnings.
You immediately hit TOGA and check pitch attitude. Almost immediatey see from the aircraft's performance that it is climbing normally and feels normal. You're already somewhat sceptical about the stall-warning. You Tell PNF to crosscheck airspeeds.He does so and his and the SBY agree. You now know the warning is false. "Gear UP" "You have control, FLY the aeroplane Bloggs! climb maintain 10,000ft (at Addis), level at 200Kts do not retract flaps, my RT". Scan panel and assay the warnings. Looks like unreliable airspeed is the first to deal with. Memory items - Assure yourself the aircraft is indeed flying normally. Monitor the FO, tell him to reduce power if necessary, you don't need 5000fpm at this point. Once comfortable-ish locate of the stickshaker cb and pull it. (it was drilled into us at my 737 conversion that you need to be able to find that one quickly just to reduce stress levels once it is deteremined to be false).
Pan/Mayday call for radar circuit to land 15 mile finals to give us time.
I really don't think I'd retract flap in that situation. I don't know what's caused the stickshaker and associated stall warnings but I do know the airframe seems to be flying as normal. I'd like to think I had the wit not to disturb anything that might affect that and leave flap where it is. I'm going to need it in a few minutes time for landing, why retract? It's just addng another potential unknown into the mix and I'd simply rather avoid that altogether. Accuse me of being wise after the event if you like, but my first thought on reading the sequence of events was "why did they move the flaps?"
Let's assume I did raise the flaps and MCAS set off on it's tricks. By now the stickshaker has been shut up and we can think. We level at 10,000 clean and speed 250. Bloggs says the bloody trim keeps running forward and he has to keep motoring it back. This is weird. What the heck's going on? You watch it for a couple of cycles and it seems the automatic trim is doing things that are unasked for and unwanted. How is this difficult to contain - even if your company hadn't told you about LionAir, Boeing's STC and you were such a hermit you didn't read the news or discuss tech in the crewroom?
"Bloggs, we're going to revert to manual trim, tell me when you're in trim and I'll select the cutout switches. (I might even take control for a minute or two to see what he is experiencing.)
There isn't much of an existential threat in any of this, is there? Apart from a little sphyncter exercise in the first ten seconds after rotation once the stickshaker is silenced this can all be done in a conversational tone of voice, which is how I prefer to manage malfunctions.
Bloggs flies the circuit. Capt runs checks, briefs landing, talks to No1 and pax, takes control and lands normally if perhaps a bit overweight.
UNLESS - the big unless - you don't stop the stickshaker beating your brains to mush and you don't forget to fly the aeroplane (so runaway airspeed doesn't happen). I think this step is critical in creating an environment where logical thought can once again occur. No overspeed, trim still operable even if you did select cutoff switches before getting it somewhere cloise to normal.
That done, you land. No one dies. No hysterical outbursts in the media shrieking about Boeing's corporate greed and murderous incompetence, or the wicked complicity with the FAA and crooked approval standards. MCAS gets fixed relatively quietly - probably still with a fleet grounding - and all returns to normal. Most importantly, the airframe is intact and they can see what ent wrong right away.
Boeing perhaps adds a stickshaker cutoff switch to the panel somewhere...to aid those whose instructor wasn't as punctilious as mine. Thank you Al.
Having flown extensively with 200hr cadets fresh to the line in a major UK airline I'd not be much concerned about my colleague's (in)experience. Of course I can't comment on those from other training environments. In my experience he is probably as good as you at flying the aeroplane accurately, probably has at least as good if not better tech and systems knowledge so use him as a voice-activated autopilot as Mr Boeing expects you to do. At my conversion course we were told that Boeing's design philosophy was that the aeroplane was designed to be operated by a competent Professional pilot and an (insert name of continent) PPL. It does that as advertised. Ultimately a sucessful outcome hinges on that, and that alone. If other nations employ different standards of crew training and operating standards then this premise becomes faulty (through no fault of Boeing's) and all bets are off.
All the above only relevant If you stick to the correct procedures and employ a tad of (swearword alert) A!*[email protected]***p.
You are certianly in a different pickle if your company hasn't troubled to pass on the latest Boeing safety advice or promulgated lessons learned from the previous accident, but then why wouldn't any pilot on the same type take the trouble to find out himself - as if you could avoid learning all about it in the media - so how could they not have had an inkling that this was all deja-vu? I know Ethiopia is a bit insular but for heaven's sake!
A proper Western investigation into Ethiopian's operating and training standards would make interesting readng but that's not going to happen. We'll be lucky to get an unmolseted CVR which would tell us what really went on. I fear the final report may not tell us much useful at all.
Last edited by meleagertoo; 27th Apr 2019 at 15:17.
Join Date: Feb 2009
Location: Virginia
Posts: 944
I find it fascinating to watch highly qualified and experienced folks take nearly opposite views of the crews’ and Boeing’s respective contributions to the accidents.
I wonder if folks might get closer to consensus if the judged the crews based on what did happen, and Boeing on what could have happened. As SLF, I have to accept that the chaos in the cockpit would have been beyond anything I’m capable of imagining. But I’m still skeptical that a crew trapped in a tug-of-war with the control column shouldn’t have worked out that they should apply nose-up trim – and keep applying it until things got better or it became completely obvious that it wasn’t working.
At the same time, I’m assuming that the AOA probe failure could have happened in IMC. At least I haven’t seen anyone explain why it couldn’t. It seems perfectly reasonable that a crew faced with a stick shaker for no obvious reason and a display showing the horizon rising above the flight path for no obvious reason might hesitate before making major nose-up control and pitch inputs. And it sounds like it could have become too late pretty quickly.
Am I missing something?
I wonder if folks might get closer to consensus if the judged the crews based on what did happen, and Boeing on what could have happened. As SLF, I have to accept that the chaos in the cockpit would have been beyond anything I’m capable of imagining. But I’m still skeptical that a crew trapped in a tug-of-war with the control column shouldn’t have worked out that they should apply nose-up trim – and keep applying it until things got better or it became completely obvious that it wasn’t working.
At the same time, I’m assuming that the AOA probe failure could have happened in IMC. At least I haven’t seen anyone explain why it couldn’t. It seems perfectly reasonable that a crew faced with a stick shaker for no obvious reason and a display showing the horizon rising above the flight path for no obvious reason might hesitate before making major nose-up control and pitch inputs. And it sounds like it could have become too late pretty quickly.
Am I missing something?
Join Date: Sep 2007
Location: Derby
Posts: 51
Originally Posted by HundredPercentPlease
Getting a bit bored of multiple posts slagging off these pilots.
We get it. You think they were sub-standard. You would have used your greater "airmanship" to keep the basic parameters in line and therefore the aircraft flying.
Your lack of appreciation of human factors in a real scenario with real line pilots in the real world is what you should address. After-the-event-heroism is ever so easy.
My suggestion is to slice the event up into small time slices. In each slice - look at the information presented, the probable solution, the required actions and the workload. Then in the next slice, base your model on a modification of the previous slice - increasing or decreasing the factors, and extrapolating the data flows. So if it starts with a stall indication, then keep the "it's a stall" flowing, until (given the workload estimate) the information is sufficient to demonstrate to the pilots that it's not a stall. Doing this, you will discover that the pilot with reduced workload will process new information earlier.
In my model, the captain wanted to get the thing cleaned up, then reduce the workload, then work out what was wrong. They cleaned up - he used the a/p to reduce the workload, and as they were starting to diagnose the aircraft bit them. His workload reduced his SA (along with the stick shaker) to a critical point. The pilot failed due to excess workload, not due to lack of "airmanship".
We get it. You think they were sub-standard. You would have used your greater "airmanship" to keep the basic parameters in line and therefore the aircraft flying.
Your lack of appreciation of human factors in a real scenario with real line pilots in the real world is what you should address. After-the-event-heroism is ever so easy.
My suggestion is to slice the event up into small time slices. In each slice - look at the information presented, the probable solution, the required actions and the workload. Then in the next slice, base your model on a modification of the previous slice - increasing or decreasing the factors, and extrapolating the data flows. So if it starts with a stall indication, then keep the "it's a stall" flowing, until (given the workload estimate) the information is sufficient to demonstrate to the pilots that it's not a stall. Doing this, you will discover that the pilot with reduced workload will process new information earlier.
In my model, the captain wanted to get the thing cleaned up, then reduce the workload, then work out what was wrong. They cleaned up - he used the a/p to reduce the workload, and as they were starting to diagnose the aircraft bit them. His workload reduced his SA (along with the stick shaker) to a critical point. The pilot failed due to excess workload, not due to lack of "airmanship".
But your model is not what is contained in the QRH for the Boeing 737. There are actions for a stick shaker on lift-off and neither involve engaging the auto-pilot or retracting the flaps. One would hope any pilot flying the aircraft would be familiar with this. Certainly in EASA countries, jet upset and stall recovery has been one of the hot topics and recurrent training for the past few years has certainly included elements to improve pilot recognition and recovery from such events. To focus on MCAS is wrong, there is much more to these accidents than that and 737driver is right, pilot training is one of them.
Psychophysiological entity
Join Date: Jun 2001
Location: Tweet Rob_Benham Famous author. Well, slightly famous.
Age: 81
Posts: 4,945
wonkazoo #4412
I thought, a very good post with an up to date overview of the situation in general.
737Driver #4438
Over the years since the 447, I've become seriously concerned about the lack of aerodynamic feel the Children of the Magenta line seem to display. I was instructed to take the BAC 1-11 to the push every other base check and remember with astonishing clarity how that aircraft felt. Hand flying to and from the cruise was taken for granted. No sims then.
A British Eagle training captain took a fellow FO and me to STN for some Viscount familiarization. One of the fun things was going over the end of the runway at 1,500' (yes, on QFE) and landing off that. We'd both been on it a year by then and it was just a fun thing to do. I can not imagine being restricted to simulated flight to get to know a type. I can not imagine being in the RHS never having really pulled the aircraft about. Mind you, we flew a lot of empty sectors back then which did not please the bean counters.
On the loss of vision thread, someone piped up and said, that's what the bloke in the RHS is for. Just go somewhere and land. Yeh, right. Good luck with the weather.
I'd have no problem stepping onto a MAX with 737Driver at the helm. But, he is pre-sprung to react to the smaller anomalies, but I'm still concerned by that 'noise' on the thumb switch trace, instead of a clear sustained ANU command. As mentioned, I'm still concerned about another 'ghost in the machine' computer anomaly.
I'd still step onto a flight was several of the experienced bods on this thread, but I wouldn't discount the chance of the craft delivering a WTF! moment.
I thought, a very good post with an up to date overview of the situation in general.
737Driver #4438
By the numbers then: Stick shaker. WTF?! Check my power (increase as necessary), check my attitude, check my configuration. Is it flying or is it wallowing? ( my bold ) If it is wallowing, keep the nose down and accelerate.
A British Eagle training captain took a fellow FO and me to STN for some Viscount familiarization. One of the fun things was going over the end of the runway at 1,500' (yes, on QFE) and landing off that. We'd both been on it a year by then and it was just a fun thing to do. I can not imagine being restricted to simulated flight to get to know a type. I can not imagine being in the RHS never having really pulled the aircraft about. Mind you, we flew a lot of empty sectors back then which did not please the bean counters.
On the loss of vision thread, someone piped up and said, that's what the bloke in the RHS is for. Just go somewhere and land. Yeh, right. Good luck with the weather.
I'd have no problem stepping onto a MAX with 737Driver at the helm. But, he is pre-sprung to react to the smaller anomalies, but I'm still concerned by that 'noise' on the thumb switch trace, instead of a clear sustained ANU command. As mentioned, I'm still concerned about another 'ghost in the machine' computer anomaly.
I'd still step onto a flight was several of the experienced bods on this thread, but I wouldn't discount the chance of the craft delivering a WTF! moment.
Join Date: Feb 2019
Location: shiny side up
Posts: 431
I would like to see a brand new FMS, not version 14 with all the legacy bs hiding in the background.
A good flow chart of the connectivity and flowpath with the if/then clearly shown. Now there is not only the if/then, but the legacy but if, but if, but if, then.
B 737 driver....In regards to the MCAS and AP, there seems to be some differences noted.
From Pilot reports:
Wind and mechanical turbulence was noted. Careful engine warm times, normal flaps 5 takeoff in strong (appeared almost direct) crosswind. Departure was normal. Takeoff and climb in light to moderate turbulence. After flaps 1 to "up" and above clean "MASI up speed" with LNAV engaged I looked at and engaged A Autopilot. As I was returning to my PFD (Primary Flight Display) PM (Pilot Monitoring) called "DESCENDING" followed by almost an immediate: "DONT SINK DONT SINK!"
I immediately disconnected AP (Autopilot) (it WAS engaged as we got full horn etc.) and resumed climb.
Another report:
After verifying LNAV, selecting gear and flaps up, I set "UP" speed. The aircraft accelerated normally and the Captain engaged the "A" autopilot after reaching set speed. Within two to three seconds the aircraft pitched nose down bringing the VSI to approximately 1,200 to 1,500 FPM. I called "descending" just prior to the GPWS sounding "don't sink, don't sink." The Captain immediately disconnected the autopilot and pitched into a climb. The remainder of the flight was uneventful.
Concur, but, as we have all noticed, the ac these days are getting pretty slippery, not to mention a lot more thrust.. its not as easy...been a bit surprised on some short finals.
A good flow chart of the connectivity and flowpath with the if/then clearly shown. Now there is not only the if/then, but the legacy but if, but if, but if, then.
B 737 driver....In regards to the MCAS and AP, there seems to be some differences noted.
From Pilot reports:
Wind and mechanical turbulence was noted. Careful engine warm times, normal flaps 5 takeoff in strong (appeared almost direct) crosswind. Departure was normal. Takeoff and climb in light to moderate turbulence. After flaps 1 to "up" and above clean "MASI up speed" with LNAV engaged I looked at and engaged A Autopilot. As I was returning to my PFD (Primary Flight Display) PM (Pilot Monitoring) called "DESCENDING" followed by almost an immediate: "DONT SINK DONT SINK!"
I immediately disconnected AP (Autopilot) (it WAS engaged as we got full horn etc.) and resumed climb.
Another report:
After verifying LNAV, selecting gear and flaps up, I set "UP" speed. The aircraft accelerated normally and the Captain engaged the "A" autopilot after reaching set speed. Within two to three seconds the aircraft pitched nose down bringing the VSI to approximately 1,200 to 1,500 FPM. I called "descending" just prior to the GPWS sounding "don't sink, don't sink." The Captain immediately disconnected the autopilot and pitched into a climb. The remainder of the flight was uneventful.
Over the years since the 447, I've become seriously concerned about the lack of aerodynamic feel the Children of the Magenta line seem to display.
