Post mortem on an IEC connector

Conventionally, a fixed wing pilot sits in the left seat. This is a hangover from airfields having a left hand circuit for fixed wings, so on the circuit leg turns, the bank is in the direction that the pilot has a view of the ground and is able to see that he does not overshoot his turn points. Obviously, that does not apply with airport 'straight in' long finals approaches, but I don't think that you really want to be having one flight deck seating convention for one plane, and the opposite for another.

As far as the robot knowing more than a human pilot, on paper that might be true. But sometimes, complex tasks like flying require dynamic 'outside the box' thinking to handle unforseen circumstances, and that is where the experience and flexible thought processes of an experienced flight crew, might just make the difference.

With the A330 incident, AF investigators have today announced that the automatic ACARS error messages were streaming events of "inconsistent height" and "inconsistent speed", which they think may have been due to the automatic throttles cycling as a result of the heavy turbulence which the pilot had declared he was encountering, using the ACARS manual text messaging option. Presumably, if that was what was actually occuring, it would not have been desirable, and the pilot would have been aware of it, so is this an example of a total fly by wire control system that the pilot cannot disengage, and operate manually ?

The trouble is that once you've thrown away the yoke and other manual controls, there's no going back. I don't have a basic problem with a fly by wire system, but I think that the option of over-riding it in exceptional circumstances, and when agreed by both crew, is the sensible one. If you totally lose the computer systems, or have a total electrical failure on one of these planes, then that's it. You are screwed every which way, and you are going to die. If you have a similar failure on a plane which has a triple redundancy hydraulically linked set of controls, then provided that the fluid resevoirs retain some system pressure, there's a good chance that the pilot is going to be able to at least make a controlled descent, and possibly even a successful landing.

Arfa

I heard it was pitot icing which isn't supposed to happen on heated pitots which all icing condition certified aircraft should have. Take away that info gather by the pitots and you disable the aircraft's ability to process flight dynamics. Pitot readings indicated a lower than normal airspeed, the engines throttle up and the aircraft then flies too fast for the turbulent conditions causing a catastrophic failure of the airframe. Decompression occurs and the airframe breaks apart at FL 330.

The thing is though that the FBW systems on a fighter aircraft are not quite the same as on a civilian airliner. Military aircraft are designed with having the pilot being able to throw them around the sky in a tactical manner in mind. This is the reason that they don't actually 'fly', and the reason that a computer system is needed to interpret the stick inputs from the pilot, and analyse the many conditions that prevail at that moment, and then give the control surfaces the appropriate input to make this unstable missile, do what the pilot wants it to.

On the other hand, a jet airliner is not an unstable lump, and does fly all on its own. Responses for inputs are absolutely predictable. Given that we are talking about the safety of several hundreds of civilians here, I really would prefer that if anything went wrong with the automatic decisions of the FBW, or even if there was a massive systems failure, that the pilot at least had a fighting chance of being able to manually control the aircraft, instead of just sitting in his seat waiting to hit the ground ...

I can see the point of FBW systems on fighter aircraft, but the reason for having them on civilian airliners, eludes me. It's all very well saying that they eliminate pilot error, but it seems that they are also responsible for 'machine error' accidents. I was talking to my aviator friend again today, and he says that there have been many low altitude incidents where the robot has got it wrong, resulting in wing waggles and roller coaster rides. The thing is, the general public don't really hear about them, as they are not castrophic events.

Apparently, Airbus are now speeding up the replacement of Pitot heads on all their aircraft. Something to do with the anti-ice heater not being good enough. Makes you wonder whether the error messages that were sent on the ACARS, followed a pattern that had been seen before, and was known to be associated with Pitot tube icing ...

Arfa

built

co-seat,

Surely they've reached the 21 century and don't rely on pitot tubes alone. Surely X,Y,Z GPS with over-the-land interpolated speed which should agree with the pitot figures when wind speed taken is taken into account.

Yeah. Lots of interesting stuff coming out now. From today's newspaper --

" ..... investigators said that the Airbus's pilots had no idea how fast they were flying after sensors iced up and the computer went haywire. The computer sent 24 error messages and was flying without autopilot or cruise control. ..... it was not clear whether the autopilot had been switched off, or had stopped working. Mets said the jet flew into a freak storm with

Taking N.Cook's point about the Pitot being only one input. This is true, with there being secondary inputs from the INS and from GPS. However, as the Pitot data is fundamentally 'mechanical', it is considered the data of primary reliability. Presumably, the fact that it was (possibly) giving erroneous readings, should have led to the data from the other two sources being taken into account, and a two out of three decision made on that basis, but there may be a software conflict situation arising out of that. I don't know if that aircraft has one or two tubes, but if it has two, and they were both icing and giving the same wrong reading, that may have been enough to screw the decision making process. My flying chum says that as well as ramping the engines up, if the flight control systems decide that the airspeed is still not going up, then it may well put the nose down as well in an effort to increase the speed. Then you are in a full power dive.

The ACARS error messages are purely for advance maintenance purposes, and are simplistic in that if the pilot switches a system off - such as the autopilot or the auto cruise control - this will immediately trigger exactly the same 'error' message as if the system had failed on its own. There is no distinction between a failure and a deliberate disconnect. Either way, it is an 'error'. The bad thing is though, that if the system is then switched back on (or recovers from a temporary fault), this does not trigger a different message to say it's back on. It is just an error that has now gone away again.

Another interesting point is that the aircraft took on every last ounce of fuel that it could, before leaving Rio. Not so much that it was overloaded out of spec, but much more than it would have needed for the flight with divert margin. So this might mean that the pilot was aware of the massive storm on his proposed routing, and was covering for the possibility of having to go around it. However, all of this extra fuel meant that he could not attain his filed flightplan cruise altitude of FL370, and was instead, at a radio approved FL350.

If it does turn out to be down to Pitot icing, and this was a known problem, I wonder if a charge of corporate culpability will be levelled at Airbus ?

That one that you mentioned where the tail fell off, I seem to recall that was due to a stress fracture in the composite which was repaired with a horseshoe of ally rivetted on over it. This is a valid repair technique for an ally skin fracture, but not for composite. I believe that there are no approved techniques for repairing this kind of problem, in this kind of material. Apparently the aircraft that has just gone down, was involved in a ground collision a couple of years back, when its wing hit the tail of another aircraft, doing that tail serious damage. I wonder if any unseen latent damage was done to the Airbus's wing, that then failed in the turbulence / excess power conditions ?

Arfa

Returning to lightning , this news report from yesterday - perhaps there should be a return to thermionic valves for safety reasons

Fishermen rescued in thunderstorm Poole lifeboat Sgt Bob Martin The new Poole RNLI inshore lifeboat took part in the rescues

Two fishermen had to be rescued after they became cut off and were unable to return to shore during a thunderstorm off Dorset.

Relatives of the men, who had been out fishing all day, alerted the coastguard when they had not returned by midnight.

The Poole RNLI lifeboats located the men's motorboat and dinghy in the harbour. Their radio, phone and mobiles had been put out by the electric storm.

They were transferred to the lifeboat and taken back to shore unhurt.

Gavin McGuinness, volunteer helmsman, said: "They did the right thing staying put. It was the worst conditions I have seen for some time.

"The visibility was poor, the rain was hitting us like ball bearings out of a machine gun, sheets of rain interspersed by bolts of lightning made conditions horrendous".

"Huge bolts of lightning bouncing of the water all round the harbour, lighting up the night skies, it was very close to the boats."

He said the torrential rain had also blanked out the radar.

I'm going by military specs of the F117 which NEEDS 3 working pitots or it cannot be flown. Don't know how many pitots a 330 has but certainly the fly by wire systems have similarities, this is just my assumption remember. For immediate control of the aircraft, the computers need at lease 3 axis of information for fundamental control. Since the pitot has been around forever I again assume they play a very large role in the guidance systems.

That may be the case but for me all bets are off when a joystick is used to maneuver the aircraft regardless of it being military or otherwise. I wonder how 'flyable' the A380 would be if it were under total human control? The A380 needed more lift from the wings but couldn't make them longer because of how most of the airports are designed. So they engineer winglets to cast of vortices that kept the last 50 feet or so of the wings from providing lift. You'll notice them (winglets) retrofitted on a lot of aircraft. So I guess who exactly knows what these guys have up their sleeves.

My flying oppo says that there are speed inputs from the inertial nav and GPS, but that the pitot input is the speed data of primary reliability, precisely because it is a simple device, and a 'known', as you say, for many many years.

I put the lunchtime news on today to see if there had been any further announcements. In the section "coming up in today's programme", they showed pictures of the tail floating in the ocean, with divers standing on it, and made some comments about that being found, and the body count now being up to 29. Then they said words to the effect of "investigators are still focusing on erroneous data from external speed sensors, being the cause of the 'accident' ". However, when it came to the actual news item playing some time later in the show, there was absolutely no further mention of this, and almost all of the item concentrated on pictures of the wreckage, and the base where bodies are being taken. Strange. I wonder whether the original comment was just a leftover from yesterday when they were saying this, put in to time-pad the excerpt, or whether the broadcast was edited on the fly to take out this aspect for some reason ...

Arfa

Well GPS is certainly not infallible. And inertia components I would assume input data for pitch, yaw, accel/decel. Inertial guidance (gyroscope) is a basic component of any autopilot regardless if it's a glass cockpit or not.

Interesting. But media coverage usually is not too technical but rather concentrates on visuals like wreckage and bodies for ratings.

an almost total waste of time , as far as technical info , More 4 prog last night was about lightning. Somewhere in the visuals (unexplained) was a distant piece of film/video of a plane being struck by lightning and passing round to the other side to continiue the arc to another bit of cloud

No, it's not *that* rudder problem which was on the MD80 and a lubrication (or lack of) issue.

The 737's problem was associated with actuator reversal. Pilot commands right rudder and gets left, commands left rudder and gets right. Planes kept falling out of the sky killing all on board, one pilot survived and told them what happened yet Boeing still said it

Happened on 747's too

Yet another 747, 9th May 1976 near Madrid

Yes, in October 2001 near NYC

There is no element of response on such aircraft, you often have to move controls in completely the opposite direction and at a different rate to what you might perceive to be the right one :)