A330 Air France GIG - CDG AF447 31ST MAY 2009

David,

It's funny how all the air speed pilot tubes were replaced 11 days after the crash on the whole fleet.
This was an Air France A340-300 that overshot the runway in Toronto in 2005. I remember that day, storm came up quickly ,high winds, hail and lightening. I wonder if there air speed was registering correctly when they brought it down. There is a Discovery Channel show on this one. (Air Disasters)
Air France Flight 358 - Wikipedia, the free encyclopedia
Dave
 

David Morton

Lifetime Supporter
Sounds like a case of diversion to anywhere rather than Toronto would have allowed the pilot to demonstrate is mental prowess instead of demonstrating his inability to land successfully in demanding weather. I've seen what the macho attitude can do to people who forgot they were public transport pilots. There is scope for cowardice
in todays litigious society, and most who display this trait usually live to fight another day.
For cowardice - also read discretion.
AF seem to be having an unlucky roll of the dice.
 
Last edited:
A post script to this accident may be interesting. An email (secondhand) from a friend who flies 330's (I believe it was for Air Canada?) described a very similar set of failures, evidently due to icing of the pitot static system, while flying back across the Pacific from Asia. They lost almost all their flight instruments, auto flight controls etc.. He said he survived only because it was day VFR, and if it had been night, weather etc., he would not have made it. They eventually recovered most of their flight functions (although the primary airspeed never came back).
Maybe pilot error is not a good term for this situation...more like "not recoverable in the given conditions"?
If you all remember, we recently lost a billion dollar B-2 (crash on take off, both pilots ejected) because of a frozen pitot static system.
Personally, there's a lot to be said for old jets like the 727 :eek:)
 

David Morton

Lifetime Supporter
Mike - I totally agree with your comment "not recoverable.....
There, but for the grace, we have all trodden at some time.
I still have a small handwritten hardbacked notebook that I compiled on various Boeing types with all types of quick fixes but on the back pages, which could easily be clipped to the checklist on the yoke, was power settings/attitudes/configs/speeds.
I never needed it.
As an aside, I had a scan through it last night and I had a fix on how to get more fuel on a 737.200 than the tanks allowed , and how to unblock a group of 'shutdown' passenger toilets on the 747/400 using ice cubes....both of which saved a diversion.
 
Guys, the pitot static system of any aircraft can be a weak point. Boeing or Airbus. The Birgenair Air 757 is a case in point.

However, while the 340 crash at Toronto can be laid at the feet of the pilots, the 330 crash poses allot more questions than it gives answers. The pitot static system is just one of the issues. The main question in my mind is what were the pilots doing in a thunderstorm over the ITCZ. If there is one type of weather that will kill you its that.
 

David Morton

Lifetime Supporter
I have to say this is a route I am very familiar with as during my time on the -400 as I did at least one South America (Sao Paolo, Rio, or Buenos Aires) every month. The ITCZ was invariably crossed in the dark both ways and it was a fact of life. It was there and you had to cross it. It was similar going down through darkest Africa where the ATC was even worse than SAL. Proficiency in using the weather radar is not something the airlines seem to be good at and they just seem to gloss over it during conversion courses assuming you have done it in a previous life. In my case ,
I came from a radar background (from the R.A.F. Nimrods) and found it very simple
to operate. However, I've come out of the bunk after my rest period (Normally timed to allow me to be on the pole approaching and crossing the convergence zone ) and found my colleagues struggling with the radar and not really understanding what they were looking at.
Jon, It's a fact of life -you have to cross it and you have to get on with it but it's a matter of doing it safely.


(SAL = Cape Verde Islands)
 
Last edited:
Gent's,

I think alot has to do with fly by wire and extreme weather conditions. Electronics and lightening strikes don't mix. With 447 a combination of faulty electronics and the weather, more experienced pilot not at the controls. Toronto was the same. The plane was was said to be hit by lightening on the desent. Of course less experienced pilot was on the throttles and reverse thrust was not started until 17 seconds after they were on the ground. Even though they landed halfway down the runway they could have got it stopped safely.
The sad thing is most of the high time, multy type rated pilots are a thing of the past.
Dave
 

David Morton

Lifetime Supporter
You may well be right - High time pilots are usually somewhere hear the top of their seniority lists and therefore are much more salary intense (expensive) than a new kid out of training. The Authorities such as the old CAA and FAA would normally apply a restriction on how many type ratings a particular person can be proficient on and the more complex the aeroplane the more restrictive they seemed to be. I was limited four in the early 80's with instructing, examining, and air testing which soon was reduced to two in the early 90's - (public trasnport types - not light aeroplanes).
They finally came up with (I think) three years and then a type was actually no longer valid if you hadn't flown it in the three year period and if you wanted to revalidate then you had to do a conversion course with written exams again .
I've been out of it for a while now (2002), and in all honesty, I'm happy to be out of it.

ps - I never did fly by wire (FBW) - just real control cables with Power Flying Control Units.
They do seem to be having their FBW problems which is a bit of a worrry.
Even BA had a 777 wrecked recently and I don't think they have got to the bottom of that one yet though the skipper is now out of work.
 
Last edited:
David,

Most of my friends that are pilots either went the military route or the hard way building up times on all sorts of aircraft at there own expense. Picking up jobs here and there to help pay for there training. I would rather have these types at the controls of my flight. But like you said with the economies of scale the way are it's all about the money.
Dave
 
David M, your right. The radar is one area that the conversion course never really addresses. However as pilots of an A330 on that route, they should know how to use it and that they are flying in an area of atmospheric instability. The theory they flew through a thunderstorm over the ITCZ, and this lead to icing of the pitot static system and all the subsequent failures is one that concerns me greatly. It gives rise to one question, Why are you flying in a thunderstorm?

Davis S, the problem with Toronto was that the pilots approached the runway high and fast and with a tail wind. They landed half way down what is quite a short runway for the type of aircraft. And the runway was flooded. The fact that reverse thrust was not deployed is not significant, other than as an indication of what was going on in the flight deck.

The main problem was that with their touchdown point being so far down the runway, the aircraft could not have stopped in the distance remaining. Even if they had done everything else correctly they would still have gone off the end. The accident was always going to happen when they lost their option to divert.

http://www.tsb.gc.ca/eng/rapports-reports/aviation/2005/a05h0002/a05h0002.pdf

As for type ratings, nowadays your only allowed one. Unless the types are very similar. For example A318/19/20/21, B757/767. You can have the A330 and A320 but you have to have flown both with in the last 35 days.
 
Last edited:

David Morton

Lifetime Supporter
Jon,
Thanks for that. The authorities have obviously tightened up, and so they should but it doesn't seem to address the FBW issue. Nobody in his right mind intentionally flies in a thunderstorm especially with passengers. Maybe it is the Thales pitots that are involved and giving problems. Pitots and Pitot Static Systems have been around for years and still they are capable of killing people.
 
Running off the end of a wet runway in windy conditions is probably never going to go away completely, regardless of FBW or whatever, because it is so easy (and tempting) for the pilot to fall into the trap of high and fast, and subsequently long. You add airspeed for the winds, the gust factor, then maybe 5 knots for mama and the kids…now you start to drift above glide path…then worst of all you try to hold it off for a squeaker landing…The situation is then exacerbated by ego (I can salvage this), which is reasonably likely given that the pilot didn’t have the experience or judgement to have avoided the situation in the first place. The best solution once recognizing that one is high and fast with a wet runway is to go around, obviously, but that requires considerable maturity. Very difficult thorn for Airlines, instructors, FAA and powers that be to solve.

Re. Fly by wire, a continuing sore point having two reasonable but opposing views is just how much authority do you give the computers. One camp argues that a computer can save a pilot from himself (no hers in this forum, right?), “protecting” him from exceeding some limit value by mistake. This accounts for the spectrum of pilots, acknowledging some are not as good as others and can use the help. The other camp says pilots should have final authority over the aircraft, because they know in certain situations they may have to exceed a limit. They want command, not a vote. Case in point, LTC. Jake Jacobson, F-16 high time squadron commander, gunnery range at Nellis AFB, Nevada. He rolls in for a bomb pass, gets in a little too close and low, and needs less than one more ‘g’ to pull out (lots more detail but not germane). He hit the ground with only a few degrees of down vector. The computer limits the aircraft g to prevent overstress. McDonnell Douglas (I worked there in flight test and cockpit development for a few years while furloughed) uses IMHO a more elegant approach. The F-15 also has a g-limit warning, and a (approx) 10 lb. Stick-pull dead zone at the limit, but if the pilot pulls harder (the natural no-thought reaction to imminent death), he over-rides the limiter and can then pull the wings (in the Eagle it’s the tailplane) off. My kind of jet.
The FBW systems which prevent stall or overspeed, for instance, as found in Airbuses, respond to inputs of airspeed and angle of attack from the air data computers/adirus or whatever and have the authority to command the airplane over pilot inputs. I don’t know what, if anything, a pilot can do to override this, nor how much time and thought it requires if an override is possible, but I can imagine if it’s night, I’m in the clouds, there’s severe turbulence, the instruments begin to fail, the airplane starts to pitch up (airspeed erroneously increasing due to frozen pitot static system) or down (the opposite), I’m not sure whether it really is pitching up or down, …Oh s&*t oh dear.

BTW, UAL only allows it’s pilots to fly one type (757/767 considered one type) and only one seat except for instructors/check airmen. American allows the two with conditions, I believe.
 
Mike,
Thank's for that. I in the camp where the pilot should have the final say.
Mr. Morton, Jon and Mike I would fly with guys anytime.
Jon,
The point I was trying to make with Toronto is that four years later maybe the data for the airspeed was incorrect due to pitot tube. I do remember that day , the birch trees in my back yard were vertical one minute and horizontal the next. (25" tall) Drastic drop in temperature, Lightening and hail. I give the pilots the benefit of doubt now. The wind shear alone would have made for a white knucle landing. I would go with gut feeling of an experienced pilot over a computer any day.
Dave
 
David,
I agree wind shear would have been the major problem, but these types of conditions are not unexpected in the vicinity of a thunderstorm. Ever since Delta Airlines flight 191 at Dallas we have know about the dangers thunderstorms pose and that is why you're not supposed to fly within 10 miles of them, 20 according to Airbus.

The other thing to remember about Toronto is that it was an A340 that crashed, these were never fitted with the suspect probes as far as I know.

I think for the explanation of these accidents one has to look at Air France. They must be the only airline to have had 3 accidents, 2 of which were fatal, in the last decade. All of which were of modern high performance jets, with an otherwise impeccable safety record. Since 2000 they have lost a Concorde, an A340 (Toronto) and an A330, it makes sombre reading.
 
Jon,
I'm not a pilot but I do know about electronics. The A series cockpits would all share the same platform. That was one of the main reasons airlines went with them. Other than size and layout, the components would be the same across all types. There are software revisions for each type. The periferals would also be common where engineering allowed. This allows airlines to run more than type and not stockpile an ton of spares. Also it made it easier for their pilots to transition between types. Simulator training with one simulator, change the software and layout for each types flight characteristics.
Dave
 
Last edited:

David Morton

Lifetime Supporter
YouTube - France - Mulhouse - Airbus A320 Crash during auto take off

This was the first fully automated Air France Aribus A320 coming to grief. In the industry we jokingly said that the GPWS (Ground Prox Warnig System) on Airbus aeroplanes should have had another 7th mode added which detected Pine Trees
and annunciated "bleep .bleep. PULL UP. PULL UP. PINETREES".
IMHO all that seemed to happen was they strapped some elastoplast (figuratively speaking) around the computers that run these aeroplanes and re certified the A320 for flight with indecent haste. Now it is one great big ball of sticking plaster.
Even the Boeing 737.300 had a few difficulties when it first came in. I can remember one ahead of me going into Gatwick and he was cleared to descend from FL100 to 3000ft and nothing happened. The controller asked if they had heard this clearance and the response was "Yes - we heard it but we can't make the aeroplane understand it " - Such was the early days of the FMCS systems, Fancy Mode Control Panels and EFIS displays.
After the Air France GIG-CDG accident that big ball of sticking plasters will get even bigger - mark my words.


For what it's worth, there is no such thing as an Auto Take -off .....
and this was just a fly by at an air show - with passengers !
 
Last edited:
David M,
I think your absolutely right about the sticking plasters, there were some horror stories about the A320 when it first came out. Most of these problems have been resolved, but with sticking plaster as you say.

David S,
I hear what your saying about the electronics, but its not quite as simple as that. In the A330 accident Its the actual probe on the outside of the aircraft which is the problem, its prone to icing up. Two types of pitot probe are fitted to the A330 and made by different manufactures, Its pot luck which type you have fitted. The reason this was done was to spread the manufacturing risk. Its the same with the ADIRUs (Air Data Inertial Reference Units) fitted and the flight control computers. One of the theories about the A330 accident was that the probes iced up in a THUNDERSTORM and fed wrong info to the ADIRUs, which in turn fed wrong speed info to the flight control computers and the pilots.

The Quantas A340 upset what happened a while ago, that was software related with a glitch in the ADIRU causing the aircraft to pitch violently.

That bundle of sticking plasters is getting bigger all the time.
 
Jon,

The pitot tubes on the 330 are the AA-types made by Thales(3), also on the 340. The ADIRU units are like a big smart network switches that route the data from the senders of the pitot tubes to the Flight Management System. The data is prioratised by the main computers. My main premise here is, in the computer field on data is garbage in garbage out.
Dave
 
David S,
Your spot on, just like any other computer.

There are 2 types of probe on the 330, Thales is one I'm not sure who makes the other. We had half our fleet fitted with the suspect one. Before they were all changed it made interesting conversation on the flight deck at to which probe was which! They have all been changed now apparently.

As for the 330 accident, the chain starts with putting your aircraft into a position where this could happen. I really feel for the pilots of that aircraft, once it had all started to go wrong they were placed in an imposable position. I have seen it in the sim a few times and its not pretty, in a thunderstorm with the aircraft pitching all over the place just makes it 1000X worse. And just about unrecoverable I would think.
 
Back
Top