German aircrash, what do we now know............

[GRG55]

Re: German aircrash, what do we now know............

As with most accidents a series of events conspire and converge to create the end result. The air traffic controller was only one factor in this instance.

Automation in aviation is wonderful. But when it is working it requires that we humans suspend any disbelief in it and trust it absolutely. And when it fails, as it inevitably will from time to time (e.g. AF 447), it requires skilled, competent human intervention. The Garmin GPS navigator in my airplane is coupled to the autopilot. If it is flying the approach I have to trust it, especially difficult to accept when the airplane is descending in the clouds and the airport is not yet visible. The alternative is to disable it and hand fly the approach.

The quality of media coverage of aircraft accidents is generally deplorable (sensationalism wins out over factual content every time); the coverage of this Germanwings incident is setting all time new low standards imo.

The Airbus lack of interconnection of the two joysticks, which was a factor in AF 447, is a factor once again. Similar to AF 447 the reports suggest the sequence started after a system failure resulted in the autopilot disengaging and handing control back to the pilots. The flight control software likely entered alternate mode and it appears once again that the pilot and first officers did not know what the other was doing by way of control inputs. I've sent the article to a friend of mine who is an experienced A320 captain to try to get some more detailed insight into how the software behaves in this situation.

AirAsia Found to Have Crashed on Rudder Fault, Crew

December 1, 2015 — 4:02 AM MST

AirAsia Flight 8501’s pilots effectively wrestled each other at the controls as they sought to fight off an electronics-system failure, dooming the plane and the lives of all 162 people on board last year, Indonesian crash investigators found.

A crack in the soldering of the rudder system caused the plane to exit autopilot, then start rolling sideways and upward, according to a report released on Tuesday by Indonesia’s National Transportation Safety Committee. The two pilots then tried to control the aircraft in opposing ways, leading the plane to stall.

“What the captain was doing wasn’t in line with the co-pilot,” head investigator Nurcahyo Utomo told reporters in Jakarta. “The captain pulled while the co-pilot pushed so the recovery wasn’t effective.”...

...As a result of the pilots’ actions, the committee said it’s recommending plane maker Airbus Group SE to have connected cockpit control sticks similar to those in rival Boeing Co. planes. Airbus said in a statement that it’s studying the report’s contents...

[jk]

Re: German aircrash, what do we now know............

The Airbus lack of interconnection of the two joysticks, which was a factor in AF 447, is a factor once again. Similar to AF 447 the reports suggest the sequence started after a system failure resulted in the autopilot disengaging and handing control back to the pilots. The flight control software likely entered alternate mode and it appears once again that the pilot and first officers did not know what the other was doing by way of control inputs. I've sent the article to a friend of mine who is an experienced A320 captain to try to get some more detailed insight into how the software behaves in this situation.

AirAsia Found to Have Crashed on Rudder Fault, Crew

December 1, 2015 — 4:02 AM MST

AirAsia Flight 8501’s pilots effectively wrestled each other at the controls as they sought to fight off an electronics-system failure, dooming the plane and the lives of all 162 people on board last year, Indonesian crash investigators found.

A crack in the soldering of the rudder system caused the plane to exit autopilot, then start rolling sideways and upward, according to a report released on Tuesday by Indonesia’s National Transportation Safety Committee. The two pilots then tried to control the aircraft in opposing ways, leading the plane to stall.

“What the captain was doing wasn’t in line with the co-pilot,” head investigator Nurcahyo Utomo told reporters in Jakarta. “The captain pulled while the co-pilot pushed so the recovery wasn’t effective.”...

...As a result of the pilots’ actions, the committee said it’s recommending plane maker Airbus Group SE to have connected cockpit control sticks similar to those in rival Boeing Co. planes. Airbus said in a statement that it’s studying the report’s contents...

do you have any insight into airbus' resistance to doing this? is it cost? if so, any clue on how much? if not, what?

[GRG55]

Re: German aircrash, what do we now know............

do you have any insight into airbus' resistance to doing this? is it cost? if so, any clue on how much? if not, what?

The Airbus was designed around a quite different philosophy than Boeing, with a heavy emphasis on the computers flying the airplane.

Because of the enormous cost of safety and certification regulations aircraft design tends to be evolutionary (the enormous cost and time invested by Boeing to create an entirely new airplane, the B787, shows what I mean). Airbus, starting with a clean sheet, saw an opportunity to do something quite different from Boeing, which they felt (correctly at the time) was tied to evolving their existing platforms. The heavy use of software to control the airplane was intended to produce a safer aircraft and to reduce overall operating costs for Airbus customers as through a high degree of commonality between models (flight control software) for pilot flight training and servicing. Once a pilot was trained to fly one Airbus model this commonality allows them to be trained to fly all other models at less incremental cost because so much of the computer systems management is common.

This would be such a major change that no manufacturer, Airbus or any other, could do it easily or without a lot of careful consideration. It may require a complete or partial recertification. And it would probably mean a major change to the training syllabus for every single Airbus pilot worldwide. Not trivial.

"Resistance" isn't really the issue. It's making sure any such change doesn't introduce other unintended, undesirable consequences.

[GRG55]

Re: German aircrash, what do we now know............

The Airbus lack of interconnection of the two joysticks, which was a factor in AF 447, is a factor once again. Similar to AF 447 the reports suggest the sequence started after a system failure resulted in the autopilot disengaging and handing control back to the pilots. The flight control software likely entered alternate mode and it appears once again that the pilot and first officers did not know what the other was doing by way of control inputs. I've sent the article to a friend of mine who is an experienced A320 captain to try to get some more detailed insight into how the software behaves in this situation.

AirAsia Found to Have Crashed on Rudder Fault, Crew

...

A number of similarities to AF 447:

1) Sequence started with a known, uncorrected technical issue with the airplane (rudder travel limiter circuit fault)
2) The less experienced First Officer had control of the airplane
3) The airplane was no longer flying in Normal Law mode on the flight computers (the computers prevent abnormal conditions such as a stall, overspeed, and limit pitch and bank attitude).
4) While in Alternate Law mode (aircraft has issues and unable to fly itself, but the computers continue to provide warnings to assist manual flight) the aircraft entered an upset condition that required pilot input to correct
5) Inappropriate control inputs, including max or near-max up elevator, which caused the airplane to ascend rapidly and then stall (eerily similar to AF 447)
6) The pilots were unable (or incapable) of recovering the aircraft after it stalled (once again, like AF 447, appears to be a CRM issue, not a completely unrecoverable aircraft due to mechanical or structural failure)

The main difference - the flight crew appear to have initiated the switch from Normal Law by pulling the circuit breaker to the Flight Augmentation Computer in an effort to reset it to deal with the alarms from the failed rudder travel limiter circuit. Shutting off or re-booting in flight one of the computers that is flying the airplane would seem to be tempting fate.

[Chris Coles]

Re: German aircrash, what do we now know............

My viewpoint is slightly different, as I am not a airline pilot, but am what can fairly be described as a senior glider pilot, (flying Open Class gliders to gain my Diamond Distance and Goal - never got my height Diamond), having been taught to fly under the guidance of Derek Piggot, then CFI at Lasham Gliding Society, starting in 1972 and today am not currently in full flying practice.

It is my understanding that when there are more than a single pilot in the cockpit, the convention is for the senior pilot to call: "I have control" and for the other to say: "You have control" and that that ends the debate as to who is flying. Yes, I can well understand that, in very stressful circumstances, that process breaks down. Again, having witnessed an accident where another seemingly senior glider pilot crashed simply because the situation he found himself in took him beyond his understanding of what was wrong; and he froze at the controls, (airbrakes opened as he took off with a passenger for an air experience flight in, (if my memory is correct a Janus, http://en.wikipedia.org/wiki/Schempp-Hirth_Janus), and, having been towed back around the airfield and dropped by the tug over the launch point to allow him to land straight ahead, he did not either level the wings or pull back on the stick, which was all that was needed to prevent the crash), which resulted in broken everything but both he and his passenger shaken but relatively undamaged.. That particular accident caused the creation of a new international rule; if a tug pilot sees airbrakes open he will now waggle his rudder as a signal to the glider pilot to close his airbrakes.

That brings me back to my main point; as I see it, many fully trained airline pilots have never been taught to fly both into and out of every possible event. A very good example being unusual attitudes in clouds. As a glider pilot, I was taught every possible event, such as the many ways to stall an aircraft; and that simple statement encompasses sometimes the most amazing, almost unbelievable situations, where the slightest deviation from the expected can result in the aircraft spinning incipiently. Yes, once learnt, easily foreseen and just as easy to retain control, and I must add, with a smile on your face. But if never shown the possibility and then faced with the unexpected, a pilot can freeze at the controls.

Airlines should always bring on trainees from internationally recognised gliding clubs such as Lasham, who will have shown them situations well beyond their expectation, that will stay with them for the rest of their flying careers.

Airbus aircraft should be able to easily create a system that follows the "I have control" command, to switch the use between the two side sticks; making the other inoperative; or to switch the feedback of the command to the secondary user, so that they will know what inputs and feedback are being processed by the flight computers. Yes, the Boeing way will involve a direct connection between the two side sticks, but that will almost certainly be very difficult to create.

Many airline pilots have never been taught to fly an aircraft in every possible situation, where glider pilots are always taught such as they have to apply flight control without an engine in every circumstance. I will go to my grave believing that if any of the pilots here described had been taught to fly a glider first; neither of these accidents would have happened.

As for the German copilot; would it have been possible to gain access to the main computer inside the aircraft to switch it off?

[GRG55]

Re: German aircrash, what do we now know............

My viewpoint is slightly different, as I am not a airline pilot, but am what can fairly be described as a senior glider pilot, (flying Open Class gliders to gain my Diamond Distance and Goal - never got my height Diamond), having been taught to fly under the guidance of Derek Piggot, then CFI at Lasham Gliding Society, starting in 1972 and today am not currently in full flying practice.

It is my understanding that when there are more than a single pilot in the cockpit, the convention is for the senior pilot to call: "I have control" and for the other to say: "You have control" and that that ends the debate as to who is flying. Yes, I can well understand that, in very stressful circumstances, that process breaks down. Again, having witnessed an accident where another seemingly senior glider pilot crashed simply because the situation he found himself in took him beyond his understanding of what was wrong; and he froze at the controls, (airbrakes opened as he took off with a passenger for an air experience flight in, (if my memory is correct a Janus, http://en.wikipedia.org/wiki/Schempp-Hirth_Janus), and, having been towed back around the airfield and dropped by the tug over the launch point to allow him to land straight ahead, he did not either level the wings or pull back on the stick, which was all that was needed to prevent the crash), which resulted in broken everything but both he and his passenger shaken but relatively undamaged.. That particular accident caused the creation of a new international rule; if a tug pilot sees airbrakes open he will now waggle his rudder as a signal to the glider pilot to close his airbrakes.

That brings me back to my main point; as I see it, many fully trained airline pilots have never been taught to fly both into and out of every possible event. A very good example being unusual attitudes in clouds. As a glider pilot, I was taught every possible event, such as the many ways to stall an aircraft; and that simple statement encompasses sometimes the most amazing, almost unbelievable situations, where the slightest deviation from the expected can result in the aircraft spinning incipiently. Yes, once learnt, easily foreseen and just as easy to retain control, and I must add, with a smile on your face. But if never shown the possibility and then faced with the unexpected, a pilot can freeze at the controls.

Airlines should always bring on trainees from internationally recognised gliding clubs such as Lasham, who will have shown them situations well beyond their expectation, that will stay with them for the rest of their flying careers.

Airbus aircraft should be able to easily create a system that follows the "I have control" command, to switch the use between the two side sticks; making the other inoperative; or to switch the feedback of the command to the secondary user, so that they will know what inputs and feedback are being processed by the flight computers. Yes, the Boeing way will involve a direct connection between the two side sticks, but that will almost certainly be very difficult to create.

Many airline pilots have never been taught to fly an aircraft in every possible situation, where glider pilots are always taught such as they have to apply flight control without an engine in every circumstance. I will go to my grave believing that if any of the pilots here described had been taught to fly a glider first; neither of these accidents would have happened.

As for the German copilot; would it have been possible to gain access to the main computer inside the aircraft to switch it off?

You are referencing a couple of different issues here Chris. The first ("who has control") is what i was referring to in item 6 of my post above:

A number of similarities to AF 447:

6) The pilots were unable (or incapable) of recovering the aircraft after it stalled (once again, like AF 447, appears to be a CRM issue, not a completely unrecoverable aircraft due to mechanical or structural failure)

...

CRM is an acronym for what was originally called "Cockpit Resource Management" (now more generically called "Crew Resource Management" as the techniques pioneered in the aviation industry are being applied in other spheres including medical trauma teams). Present day CRM training traces its roots to the aftermath of the December 1978 United Airlines Flt 173 accident at Portland, Oregon, where a highly experienced crew allowed the aircraft to run out of fuel while they were all preoccupied trying to diagnose an apparent gear extension problem on the approach to landing.

This crash, and its similarities to the 1972 Eastern Airlines L-1011 accident in Florida (where the crew allowed the plane to fly into the ground while they diagnosed an apparent gear extension failure - which turned out to be a burned out indicator lamp), was the catalyst to create formal training procedures to deal with cockpit communications, task distribution and decision making in a team environment where human error can result in catastrophic outcomes.

It is now part of the formal training and re-currency curriculum at virtually every commercial air carrier worldwide, and it has had a measurable positive influence on safety statistics. However, it seems to work better in environments/societies where critical thinking is openly valued and it is permissible to question authority - the cultural differences between North America/U.K./Germany and, say, greater Asia seem to be showing up in the accident statistics. Commercial crews flying mechanically sound state-of-the-art airplanes into the ground still happens - Gulf Air 072, where the crew flew a perfectly good A320 into the waters of the Persian Gulf on a crystal clear night, and Asiana 214, where the crew allowed their B777 to hit the seawall short of the runway threshold at San Francisco on a clear day, are a couple of examples of the breakdown of CRM in the cockpit.

The second point you raise is recovery from unusual attitudes. This was a subject of much debate after the AF447 accident.

The debate centres around the fact that the computer systems on modern commercial aircraft "protect" the aircraft from being put into any such unusual attitudes under normal circumstances. As long as everything is okay the computers fly the airplane precisely and it cannot be stalled or overbanked - as my A320 captain friend emailed me earlier this week when we were conversing about the Air Asia accident report, in Normal Law mode "Just point and fly". As a result a great deal of the type rating training is learning and demonstrating proficiency with the computer systems management of the airplane, and as you noted above this is calling into question a potential lack of emphasis on basic flying skills in the event the computers are unable to fly the plane.

There is nothing significantly wrong with the Airbus aircraft or its systems. It has to be flown with knowledge and respect. But then even a Cessna 150 trainer or a glider will kill you if you don't fly it with an appropriate level of respect.

I'll close with a related anecdote. My brother is a former military pilot and fighter pilot instructor who now flies the Boeing 787 Dreamliner. The full-motion simulator for the B787 is arguably the most advanced and capable commercial aircraft simulator in existence. When he was sent to the first type rating session on the simulator the aircraft was programmed to be flying very light (small fuel load, no passengers or baggage). The instructor suggested the pilots "play with the airplane" to get a feel for the simulator before they started the formal proceedings. My brother promptly rolled the simulator upside down and held it in inverted level flight, which was enough to disorient the instructor.

[Chris Coles]

Re: German aircrash, what do we now know............

You are referencing a couple of different issues here Chris. The first ("who has control") is what i was referring to in item 6 of my post above:



CRM is an acronym for what was originally called "Cockpit Resource Management" (now more generically called "Crew Resource Management" as the techniques pioneered in the aviation industry are being applied in other spheres including medical trauma teams). Present day CRM training traces its roots to the aftermath of the December 1978 United Airlines Flt 173 accident at Portland, Oregon, where a highly experienced crew allowed the aircraft to run out of fuel while they were all preoccupied trying to diagnose an apparent gear extension problem on the approach to landing.

This crash, and its similarities to the 1972 Eastern Airlines L-1011 accident in Florida (where the crew allowed the plane to fly into the ground while they diagnosed an apparent gear extension failure - which turned out to be a burned out indicator lamp), was the catalyst to create formal training procedures to deal with cockpit communications, task distribution and decision making in a team environment where human error can result in catastrophic outcomes.

It is now part of the formal training and re-currency curriculum at virtually every commercial air carrier worldwide, and it has had a measurable positive influence on safety statistics. However, it seems to work better in environments/societies where critical thinking is openly valued and it is permissible to question authority - the cultural differences between North America/U.K./Germany and, say, greater Asia seem to be showing up in the accident statistics. Commercial crews flying mechanically sound state-of-the-art airplanes into the ground still happens - Gulf Air 072, where the crew flew a perfectly good A320 into the waters of the Persian Gulf on a crystal clear night, and Asiana 214, where the crew allowed their B777 to hit the seawall short of the runway threshold at San Francisco on a clear day, are a couple of examples of the breakdown of CRM in the cockpit.

The second point you raise is recovery from unusual attitudes. This was a subject of much debate after the AF447 accident.

The debate centres around the fact that the computer systems on modern commercial aircraft "protect" the aircraft from being put into any such unusual attitudes under normal circumstances. As long as everything is okay the computers fly the airplane precisely and it cannot be stalled or overbanked - as my A320 captain friend emailed me earlier this week when we were conversing about the Air Asia accident report, in Normal Law mode "Just point and fly". As a result a great deal of the type rating training is learning and demonstrating proficiency with the computer systems management of the airplane, and as you noted above this is calling into question a potential lack of emphasis on basic flying skills in the event the computers are unable to fly the plane.

There is nothing significantly wrong with the Airbus aircraft or its systems. It has to be flown with knowledge and respect. But then even a Cessna 150 trainer or a glider will kill you if you don't fly it with an appropriate level of respect.

I'll close with a related anecdote. My brother is a former military pilot and fighter pilot instructor who now flies the Boeing 787 Dreamliner. The full-motion simulator for the B787 is arguably the most advanced and capable commercial aircraft simulator in existence. When he was sent to the first type rating session on the simulator the aircraft was programmed to be flying very light (small fuel load, no passengers or baggage). The instructor suggested the pilots "play with the airplane" to get a feel for the simulator before they started the formal proceedings. My brother promptly rolled the simulator upside down and held it in inverted level flight, which was enough to disorient the instructor.

So, what is the negative pitch angle to keep an inverted B787 Dreamliner in level flight?

And a further thought, going back to a wonderful part of my flying experience with one very famous, (at Lasham), Charlie Kovac, a Czechoslovakian who had fought in WW2 and brought his survival experience to us ab initio trainees where, when checking us out for air experience flying, asked us to imagine he was the ordinary person that was being shown how to fly for the first time and to thus tell us how he should fly the aircraft. So, when we said push the stick forward, yes, you guessed correctly; he pushed it fully forward and when we then said pull back, he did just that and pulled the glider, a good old K13, right back into an intense stall. His point being, to impart our instructions carefully worded so as to prevent such incidents.

Knowing now that a 787 can be rolled, I would imagine a similar situation where a friendly captain offers the controls to a novice, (yes it has happened with an A320), except that instead of in that crash where the nose was pushed down into a dive, the "friend" tries to roll the aircraft. Another reason to keep the novice well away from the controls in flight.

[GRG55]

Re: German aircrash, what do we now know............

So, what is the negative pitch angle to keep an inverted B787 Dreamliner in level flight?

...

I don't know, but my brother thinks the B787 might be the only commercial airliner that has the performance to maintain constant altitude inverted flight (and I am sure no one will want to subject an expensive actual airframe to that treatment).

[touchring]

Re: German aircrash, what do we now know............

and he froze at the controls,

Perhaps every commercial airline pilot should be tested at Lukla before allowed to fly planes with passengers.

[GRG55]

Re: German aircrash, what do we now know............

Perhaps every commercial airline pilot should be tested at Lukla before allowed to fly planes with passengers.

I heard a story many years ago that Cathay used to require new pilot candidates to fly the approach into Kai Tak on their simulator.

https://www.youtube.com/watch?v=OtnL4KYVtDE

[lakedaemonian]

Re: German aircrash, what do we now know............

A good read, to keep things in perspective (hopefully).

The magic of flying: A pilot shares his passion for one of the wonders of the modern age

Flying is really fun and really safe.

I only wish I had the time(or could make the time) to pursue personal flight more.

For now, paragliding is my personal flight fun, although I once had dreams of something like that Eclipse 500 disaster.