STKNRUD

I have seen another post here about this accident. This article, however, seems to add some interesting detail..


http://www.popularmechanics.com/tech...ce-447-6611877

Maybe after reading it someone can explain to me why the onboard computer system "averages" the stick inputs of the left and right seat. Is this some kind of computerized CPM, i.e....if the captain and first officer can't agree on stick inputs, "Hal" will just average them out? I am sure some explanation or circumstance exits that makes this desirable but I can't think what it would be. Maybe they should instal a stick at the jump seat so someone can break the tie [:@]

rcjetsaok

My Feedback: (7)

Pilot(s) error caused the crash of this airplane. The Pitot tubes did indeed ice up and gave airspeed misinformation. The Auto pilot disengaged, pitot heat was turned on. At that piont the ice melted and the Pilots crashed a perfect, fully operating Aircraft. They did not know or understand what had happened.

Danno

STKNRUD

I know it was all pilot error. I am just curious why the onboard computer "averages" the stick movement between the left and right seat.

jzuniga

My Feedback: (7)

Because it's a modern day Airbus! In the old school one's we fly (A300/310) we don't have any of the modern FMS and or fly by wire systems of the latest generation Airbus'.

Z

Turbotronic

Averaging stick movements must be the dumbest thing ever done to an aircraft. Clearly these modern systems are designed by software engineers with no comprehension of being a pilot or seem to have any idea of man machine interface psychology. This will remain one of the most dangerous issues with aircraft and already in motor vehicles as well. Yes training and vigilance will overcome this but is that not the opposite of the design aim? These pilots were confused into a dumb confused state by overwhelming changes in mode of operation of the flight system, nothing of which was clear and concise. Very convenient to blame them alone. They cannot defend themselves now can they?
AB

darryltarr

My Feedback: (2)

Danno, with all due respect your comments are NOT entirely correct.

The intricate details behind this accident are VAST and remember; for any accident to occur there is always a multitude of events that have to be in sync - explained in the Swill Cheese Reason Model by James Reason.

Of course, the pilots are always the last line of defense and its ALWAYS VERY EASY TO POINT FINGERS AT THE CREW, but trust me - no flight control system, by design, should leave the flight crew guessing or confused by going into some other LAW. This is NOT an AIRBUS bashing, many other manufactures are also guilty of this concept, however, I do believe that Pilots should always have full control and authority at all times (unless when aerodynamically impossible).

David Gladwin

Precisely, Darryl.

And if you guys want to know more about this accident start here :

http://www.pprune.org/rumours-news/4...versation.html. There is also another HUGE thread on the AF447 accident.

Sorry Turbotronic, to suggest that Airbus engineers have no idea of actually flying the machines for which they design the flight control software is, to be polite, most unwise !

(Clearly these modern systems are designed by software engineers with no comprehension of being a pilot or seem to have any idea of man machine interface psychology.)

To see the huge growth of the Airbus fleets around the world suggests you are WAY off the beam.

Regards,

David Gladwin.

STKNRUD

To those defending the engineers, etc, please try answering my question. Why does the system average the stick inputs? No one has explained the reason behind this very fundamental PIC issue.

FalconWings

My Feedback: (57)

....otherwise the pilots may end up fighting each other for control? I agree, just average it up and make them both happy.

FalconWings

My Feedback: (57)

BTW, stop referring to A320/A330/A340 operators as pilots. They are Data Entry Technicians. That's the reason they place a lable on the side stick "For Emergency Use Only" and there is another one in the throttle quadrant labled "Do Not Touch Unless Instructed by FMC".

STKNRUD

I hope your joking!

FalconWings

My Feedback: (57)

.........I don't know, you'd have to ask an Airbus Engineer, but there are none in this forum. Actually, I think the Flight Control System is Siemens GmB.

Per Strommen

Inputs on the side sticks are not averaged, they are added. Meaning that both sticks are "live" and whoever is stronger wins. Just like any Boeing.

rgburrill

With all respect this system was designed by FRENCH engineers. God forbid someone be put into the primary decision making position.

S_Ellzey

The following are the comments from a 777 pilot after getting his rating on the A320 on the side stick controllers:

"A visitor once asked me (was it you?) what happened if the captain moves the side stick to the left and the copilot moves it to the right? My answers were not very clear then, as I was not rated on the A320 yet.

Well, since the side sticks are only "wired", they are not mechanically interconnected like the old conventional planes. The right side cannot feel the movement from the left stick. What happens is that, when there are opposite deflections, their inputs are algebraically added (total sum) with the maximum limit corresponding to the movement of one side stick. If the captain pulls full left, and his copilot pulls full right, the net effect is zero.

So, the last pilot to click on an override push-button on the side stick obtains control (an indicator light in front of the other pilot signals this fact). The potential exists for the pilots to *fight* over the control of the side sticks. Rather than the "strongest" pilot winning, the one with the fastest thumb will win! "

Erik R

My Feedback: (32)

I have been resisting responding,but too much uneducated bloviating here.The answer is,only one person is supposed to move the stick at a time.There is a red override button on each stick to transfer control from the other side,primarily for use in a sidestick failure,causing uncommanded control inputs.I have approx 5000 hrs in the 320,am not a big Airbus fan,and haven't flown it in over 10 years,but what would you have them do?In a conventional yoke scenario,if both pilots are trying to input controls,they end up fighting each other,with literally the strongest winning.Similar result to the Airbus fly by wire system.

If two pilots are trying to fly the jet at the same time,there is a fundamental breakdown in what we call crew resource management.In any abnormal situation,the first thing the captain needs to do, is assign who is flying the jet,which is the obvious number one priority.If he was manipulating the controls while the F/O was still the understood flying pilot,thats a human error.He needed to clearly communicate that he was taking control of the aircraft.Again,not a big Airbus fan,but the system works well.My carrier had at least 5 similar incidents occur on the 330,and all were recovered with minimal attitude/altitude excursions.

As far as "laws" go,the progression is normal,alternate,then direct law.No major difference in control per se.Each progressively degraded law basically offers less protections from things like overspeed or stall profiles.As has been stated before,a great many factors contributed to this tragedy,human factors being one facet.God bless the victims and their families.

STKNRUD

I didn't intend to start a "bash" the Airbus thread. IMHO, the over-engineered design of the FMS is just the logical end to something started a long time ago with saving money (not that saving money is bad). First they got the FE pilot qualified in the 60's so they could eliminate the third pilot. Then the development of FMS allowed them to eliminate the FE. Now we have a "logical" extension of making the FMS more and more sophisticated that, IMO, is resulting in some un-intended consequences. Stick input averaging/adding, is an example. Does stick priority systems and their resulting 'overrides', 'panel alerts', etc. really contribute or does it add more complexity and pilot acknowledgment/response requirement that can have unintended consequences?

I am not suggesting that this one item caused the accident or even was a major contributing component. But the article makes it appear that two pilots were attempting to fly the same stick with the FMS deciding what to do with the offsetting inputs. Not-with-standing some design purpose, I can't see the off-setting benefit to the complexity and possible consequences. Call me a "stone-age pilot" that predates FMS but is it possible we have only added un-necessary complexity to some systems? Have we created a sequence of never-ending flight automation additions only because "the FMS can do it" or because we have become trained to believe that all automation is good because it makes the pilot's job easier?

STKNRUD

ErikR

Regarding CRM, amen!

But when you automate stick priority through a FMS, you have to provide for a great deal of possible circumstances such as pushing the "takeover" button on the sticks at the same time which takes control from the other stick. The FMS makes the decision that the last one to push their button has control. So CRM is in deed the issue but the FMS is attempting to provide a substitute for "I have the airplane". The complexity, alerts, overrides, etc. only makes a bad situation worse and weakens CRM, IMO.

rcjetsaok

My Feedback: (7)

Just take the Pilot's completely out of the Cockpit, and when all those ones and zeros get all tangled up and people die, then there is no question what went wrong !!! I agree that the pilots were not completely to blame. They were blind sided by a senerio that none of the of the engineers could have ever imagined. Given this type of situation with this technology, Apollo 13 would have never made it home.... We have out smarted ourselves.


Danno

Turbotronic

Exactly right, I am also a software engineer and have seen studies of how poor design leads to unexpected and (to the human operator) completely illogical output. Poorly designed software systems can just about exibit limitless sometimes untestable outputs. I recently drove an bluemotion VW that played boss on who controls engine response/shutdown and acceleration. Not so good on them UK roundabouts! I could visualize the stupid code in the system and it infuriated me. And I am not Airbus or French bashing either. It is a universal problem. With most first adopter gizmos these days coded in places like indian sweatshops I lose interest, its just too painfull to endure sometimes.

Quikturn

My Feedback: (12)

WOW! What a frightening read. I blame the 2 copilots for this one. Yes, the Airbus may be complicated and have some goofy systems but these were not competent pilots and they did not know the airplane.

A frozen pitot tube at altitude would likely make the airspeed indicator act like an altimeter which is to say, as you climb and gain altitude; airspeed indication increases. This situation can trick you into a stall if you're going on airspeed indication alone. One question that goes unanswered is what possessed the copilot to climb in the first place? And where was the acting captain to correct him and get back to the assigned altitude? What possessed the copilot to pitch up 18 degrees (at 37,000 ft!)? Also, it sounds like the pitot heat was not turned on initially. Every airplane I have flown requires pitot heat to be on for flight at all times. Frozen pitot tubes may have caused the autopilot to disengage but lack of airmanship caused the rest.

Kometfreak

To me it looks like basic stick and rudder skills were flat out ignored. Known power at known pitch equals known airspeed and rate of climb or descent.

jzuniga

My Feedback: (7)

I too have read the report. The captain was in the back taking his required FAA rest. Unfortunately for him, it appears he did not get to the cockpit in sufficient time to turn the situation around.

Z

eddieC

My Feedback: (2)

Forgetting to put the pitot heat on has downed airliners since at least the '60s. I have a simple procedure in turbines: Pitot heat on taking the runway, turned off clearing the runway after landing.

David Gladwin

[quote]ORIGINAL: eddieC

.I have a simple procedure in turbines: Pitot heat on taking the runway, turned off clearing the runway after landing.

I have a better one !

Piot heat ON in pre start checks EVERY time, OFF in shutdown checks !

That way the pitot heads are hot enough to cope with ANY condition at all times, even on the ground, Boeing checklists refer.