Armavia A-320 accident report now available

[punkalouver]

http://www.bea-fr.org/anglaise/actualite/sotchi/tmp/sotchi_060503_final_report.pdf

[A4]

Sobering reading. Alarming similarities to the Gulf/Bahrain accident - go-around / missed approach at night with lack of visual cues. Somotogravic illusion leading to inapproprite control inputs and CFIT......

Recommended reading.

A4

[Ladusvala]

Also interesting to read about the lousy CRM and non adhearance to SOP which contributed to the accident.

Fatigue was mentioned several times in the report but unfortunately not as a contibuting factor, as far as I could see.

[Sobelena]

And as well as the above, the dreaded get-in-alitis factor. Nevertheless, the whole chain of events was set off by a rather odd national procedure which allows ATC to break off your approach on the basis of weather readings! Just who is flying the aeroplane?!

[Manutara]

If you read Page 1 of the report you'll see it's just an English translation of the original report in Russian.

[Basil Smallpiece]

Remarkable how many fully airworthy Airbus aircraft have impacted the surface because the crew failed to understand or interpret the flight modes on display.

I am convinced that the multiple automatic mode changes that occur particularly when operating at the edge of the envelope; the fixed detent thrust levers and the non-interconnected joysticks significantly contribute to pilot disorientation.

Is the A380 cockpit any different or will such disasters also be possible?

If I were an A380 Flight Manager I would decree that the aircraft only be flown using all autopilots, even for take-off, GPWS and TCAS. Non-precision and visual approaches would be banned.

[A4]

OK Basil - I can think of two. GulfAir and ArmAvia. I do not know if you fly the Airbus - I suspect not. If you read the report it is patently obvious that the instruction to discontinue the approach was very poorly executed. I'm not having a go at the crew or the obvious CRM issues - but the facts are that the aircraft was operated in a totally non standard way in contravention of the FCOM.

There may well have been "mode confusion" - but that's not unique to Airbus and I'm not going to turn this into a A versus B debate - somotogravic illusion is not type specific!!! Personally I think the non moving THR LVR's is a red herring. In 10 years it has never caused me a problem as there are aural and visual cues to tell you what's happening.

I think the primary confusion was the overspeed warning because the flaps had not been retracted on the go-around - hence retarding the levers (albeit briefly). The crew went from a low energy situation ("SPEED SPEED SPEED"), which they reacted to correctly, to an overspeed (due flaps) in a short period of time - confusing yes, but entirely avoidable IF they had followed the standard procedures/FCOM. The susequent actions upto impact point to a total breakdown of SA.....

It's easy I know sitting here but it does re-enforce that you MUST be prepared for the unexpected. IF you are told to "turn right heading 240° and climb 2000'" whilst fully configured on the ILS at 3 miles just EXACTLY HOW are YOU going to achieve that? This is something I put to my trainees and if anything it emphasises that whilst in essence it's straightforward you do need to THINK about what you are going to do and the order in which you are going to do it.

As with all accidents / reports, I hope we can all learn something from this one - as I said earlier, recommended reading.

A4

[TheKabaka]

Hi Basil,

You have clearly not been on the flight deck of an airbus. Your assertions are rubbish, with proper training the bus is an excellent instrument or visual flight platform for manual or automatic flight.

I too wish the thrust levers moved but in the instictive way that fully pushing them forwards gives TOGA thrust and pulling them all the way back gives you idle they are perfectly satisfactory.

[Basil Smallpiece]

A4 - here are a few that I can think of:

300/600: China Airlines Nagoya

310: Kenya Abidjan, Aeroflot Siberia, Tarom Bucharest

320: Air France Basle, Indian Bangalore, Air Inter Strasbourg,

330: Test Toulouse

Plus countless other non fatal incidents...

[Engine3firehandle]

Shall I start my long list for Boeing, Douglas or Bombadier etc.. ?
The problem is not Airbus or Boeing. It is humans that are not unfailable and are on the wrong track mentaly. And the pilots might not have had enough or poor training.
The differences between airline quality is in one part, how much they spent on crewtraining. But will never be accounted for.
I would wish, I could see a benchmark about crewtraining compared to accidents.
But that is an almost impossible task. The more training you give as an operator the less accidents you can expect.
That is the truth, not that the FMA is not good. Which you will find, by the way, also on other aircraft.
Me, flying presently an American Heavy aircraft and been flying Airbuses and Fokkers as well. I can tell you, the system is basic, easy and almost foolproof.
Fly an engine failure at T/O on an A346 and just do absolutely nothing, yes nothing. The aircraft will continue climbing in a very shallow turn with about 8-10 NM diameter and return to the old place with constant height gain.
Do it in cruise, do again absolutely nothing, shallow descent and straight flight. One wing slightly lowered by 2degrees bank.
Try it in the SIM and you will be surprised.
But pilots are from their personality more conservative then the average human being and more unable to make changes in their personal habits and opinions.
Don´t be afraid about airbus. They build very nice aircraft and that is good and brings Boeing to the point, that they got to change from old a/c to newer technologies and better a/c to compete.
S***, again that Boeing-Airbus thing

[A4]

Ok Basil,

Please correct me if I'm wrong.

Air China, Nagoya. Inadvertant activation of go-around mode by FO. Crew subsequently tried to fight the aircrafts pitch/thrust (go-around) leading to a stall.

Kenya - no recall

Tarom - similar to Air China but "rescued"?

Aeroflot - Siberia..... Don't let your son fly the aircraft..... situation not recoverable.

Air France - Basel. Been done to death. Lot's of conspiracy theories. Bottom Line - if you fly a jet at 30' towards a forest (instead of the 400' as briefed) don't forget that the engines will take a while to spool up FROM IDLE.

Bangalore - OP DES (Thrust to idle) selected on approach. High sink rate developed with intervention by crew too late. Flight directors were not being operated in accordance with FCOM.

Air Inter - Mode confusion. 3.3° desired (approx 900fpm) 3,300 fpm selected.

A330 - Toulouse. Go around with simulated engine failure during ALT* - leading to stall?


The A300, 310 accidents were essentially pilot error. Procedures not being followed. The aircraft was performing "as instructed".

The Basel (Habsheim), Bangalore and AirInter accidents were in the early days of the FBW Airbus series. Without a doubt mode confusion played a part in the AirInter and this was addressed by Airbus with a new FCU design. The Habsheim accident - well read the timeline, the aircraft did exactly as it was told to do. Bangalore. Two inexperienced Capts. Check Capt was NOT authorised by Airbus to instruct. Deselection of ONE flight director (no no on Airbus) cause inappropriate thrust mode.

A330 - Accident during test flight. I think it had somethingto do with no low speed protection during ALT* with oei?


For me the over-riding factor that comes out of the above is you MUST understand what the aircraft is doing and be FULLY familiar with the FCOM. If you accidentally trigger go-around mode (old Airbus and (Boeing?)) DON'T FIGHT IT - especially if the trim is automatically winding NOSE UP. Just go with it - swear under your breath - vector round again. NEVER try to rescue a bad approach.

Mode confusion. When you select a mode on the FCU (MCP on Boeing?) then (hopefully!) you expect confirmation not only from the FMA (Flight Mode Annunciator) but also the secondary indicators (OP DES - I expect to see the thrust reduce to IDLE and a pitch down.) If I select FPA -3° I expect to see 700-900 fpm V/S - if I see 3,000 fpm then something isn't right.

It all comes down to Training. It is the reponsibility of Trainers, AND Trainees, to ensure that modes, and there consequences, are fully understood. A thorough understanding and full appreciation of what you EXPECT to see will make your life a lot easier - particularly close to the ground in a non-normal scenario which isn't the time to scratching your head and asking questions.

If the weather is cr@p - slow down!!! Take your time!! Don't try to be "slick".
If you rush and screw it up you just going to be up in the wx longer.

No aircraft is perfect - and the Airbus is usually singled out for criticism. If you are trained properly and make sure you understand the aircraft and its systems (which is a licence requirement) then the Bus is very capable and very easy to fly.

A4

[kingair9]

Kenya - no recall



http://aviation-safety.net/database/record.php?id=20000130-1

Crew did not react properly to a stall warning after t/o:

The investigation showed that the pilot flying reduced the pitch attitude but did not apply TOGA thrust on the engines

imho MEL is mixing up accidents of A300/310 series with the computer logic of the 320 series which is very different to the conventional 300/310 series.

[Hand Solo]

If I were an A380 Flight Manager I would decree that the aircraft only be flown using all autopilots, even for take-off, GPWS and TCAS. Non-precision and visual approaches would be banned.

That'll be interesting when you have to visit an airfield with no operating ILS! Of course that won't happen because wuth autopilot only take offs you won't be taking off! Now, lets talk about banning all manual flying on that deathtrap the B744. CX and BA have had a couple of interesting go-arounds, must be the aircraft to blame.

[TheSailor]

Hello,

Just for a note:

Error in the dates in the beginning in the history chapter...
2 May .... 23 -24 May .... ??

Salute http://photobucket.com/albums/v509/Bebermaur/th_bye.gif

[Kraut]

@Basil

You never flew an AIRBUS! Your point......Non-precision and visual approaches would be banned................

is totally off! Non precision approaches with an AIRBUS (flown acc FCOM) is always a pleasure to execute!

Visual approaches: peace of cake, if you know what you are doing.

As A4 mentioned, you got to know the flight modes!

[Basil Smallpiece]

Jawohl and know the FMS modes as well like the Hapag Lloyd 310 Vienna and Air Transat 330 Azores crews who both incredibly ran out of fuel!

[Hand Solo]

And what exactly did the Air Transat FMS have to do with running out of fuel? Seems to me like Basil has an axe to grind. If we're going to start talking about fuel exhaustion best stay away from those 767s too.

[Kraut]

@ BASIL

Running out of fuel and CFIT events are not really comparable!
And here we are talking about CFIT.

[RoyHudd]

Folks, please stop rising to Basil's silly baits. The person is no pilot, clearly evidenced by his first "contribution"..Take-Off with all autopilots, ditto GPWS & TCAS....good grief, this website is plagued by amateurs who seem to love pretending to be commercial pilots.

[Brian Abraham]

From the report
While intervening, the co-pilot had not pressed the take-over pushbutton, therefore both pilots’ control inputs were added and averaged. Such dual piloting is prohibited by the A320 FCOM. At that time the actions of the two pilots were not coordinated. The Captain twice moved the side stick half-way backwards , possibly, reacting to the EGPWS, but at the same time the co-pilot was inadvertently making nose-down inputs, which might have led the Captain to believe that the aircraft response to the control inputs in the pitch channel was not adequate.
My bolding. Without a A v B argument developing would the Bus drivers care to comment on this aspect. Is it an area/scenario you cover in the sim (Train like you expect to fight, fight as you have trained)? Man/machine interface aspects?

I realise they were likely not following procedure (co-ordinating their efforts) here as well, but I ponder the reason why the copilot failed to press the button if he realised things were going pear shaped..

[Hand Solo]

Yes it is something covered in the sim. If you are going to take control with the other guy handling you always press the priority button or your inputs will be summed. Later models of aircraft even give you an ECAM Dual Input warning to cater for a failure to press the button.

[Gnadenburg]

Brian

The use of the take-over push button is now a well established SOP. Some historical incidents which have made this so, include an occassion in Sydney in the early 90's when there was a RWY infringement. The Capt initiated a GA and took over from the FO on short finals. However, the F/O still had stick input and GA performance was compromised.

Having the Side Sticks algebraically summed presents other problems in what you refer to as the man-machine interface.

For example, if you overspeed the aircraft, there is a 1.75G pull-up. Problem being, the pilots possibly panicked input is added to this aswell!Two pilots panicked inputs, now summed together, and you run quickly up to the aircraft's 2.5G limit in an overspeed scenario.

I see the report makes mention of what can be described as speed tape confusion. An anecdotal problem with Airbus PFD's.

Secondly, there was no mention that I can find, if the aircraft was software upgraded to have the automatic reinstatement of FD's on GA.

[A4]

Quote: For example, if you overspeed the aircraft, there is a 1.75G pull-up. Problem being, the pilots possibly panicked input is added to this aswell!Two pilots panicked inputs, now summed together, and you run quickly up to the aircraft's 2.5G limit in an overspeed scenario.

True. BUT at least the Airbus can be manouvered very roughly straight to those limits with out fear of stall. If both pilots pull back, and one also closes the THR LVR's the speed will decay very quickly and before you know it ALPHA FLOOR will kick in (assuming Normal Law and A/THR available)

Quote: Secondly, there was no mention that I can find, if the aircraft was software upgraded to have the automatic reinstatement of FD's on GA.

This is not relevant to this accident. The FD's were on (G/S and LOC) If the LVR's had been pushed to TOGA, even momentarily, the SRS G/A TRK would have engaged. Presumably they pulled heading to initiate the turn and the FD's would default back to V/S and HDG with the current rate of V/S. If the AP is then disconnected and the aircraft flown with disregard to the FD's that's when problems can really start with very confusing PFD presentation. So, set the approriate mode, READ THE FMA and FOLLOW the FD accordingly.

For the benefit of others, the later Airbus's have a feature whereby if the FD's are off and TOGA is selected they automatically reappear in "Go around" mode. On the older Bus's it would require re-selection of FD's, then possibly from TRK/FPA to HDG / v/s mode and then a brief selection of TOGA to activate Go Around mode (SRS G/A TRK). This does require discipline from PF and PNF in a high stress situation.

A4

[ZBMAN]

If the LVR's had been pushed to TOGA, even momentarily, the SRS G/A TRK would have engaged. Presumably they pulled heading to initiate the turn and the FD's would default back to V/S and HDG with the current rate of V/S.

Agree with what you say A4, however, correct me if I'm wrong, but the SRS doesn't default to VS during a GA. SRS disengages (automaticaly) at GA acceleration altitude or if you get ALT*, if the crew selects another VERTICAL mode, or if the crew selects a speed (then SRS reverts to OP CLB with triple click etc). To default back to VS when selecting HDG they woul need to be in CLB NAV.

You are right about the "pop up" FD's on the later models. On the old buses a GA from a FD off approach was challenging to says the least, especially as a the approach would be GA prone to start with. (GIB springs to mind)

[A4]

Hello ZBM,
What I was trying to say was this: You are established on the ILS, AP engaged with SPEED G/S LOC on the FMA. ATC then tell you to "turn right heading 240° and climb 3000' ". If you now pull HDG and nothing else the FMA will read SPEED V/S HDG (with a triple click) The V/S will be the V/S you had at the moment you pulled HDG. IF you then hit TOGA you will get MAN TOGA SRS G/A TRK.

You are correct to say that SRS disengages at ALT* during the GA.

Hope that clarifies it.

A4

[ZBMAN]

Thanks A4, now I see your point!:ok:

[Gnadenburg]

• Specific features of speed indication on the PFD, especially speed limitations for the given aircraft configuration that are shown as the red bars at the top of the speed indication strip. One may imagine the influence of the reflex acquired in training, for example, in response to a TCAS warning when the pilot is anxious to avoid the displayed red part of the instrument scale, which may result in the instinctive forward movement of the side stick, especially when the pilot is in a state of psycho-emotional strain. This version is substantiated by the fact that the pilot was monitoring the flight speed and its limitations (VFE) that depended on the aircraft configuration and retracted the high-lift devices in a timely manner, and the control inputs on the side stick coincided with the moments when the current speed was getting close to the limit value.

This is the first official recognition of speed tape confusion I've noted. But I have seen the problem on Airbus before the advent of TCAS- questioning the "one may imagine" theory?

A4

Yes. Intermediate GA dispels the no FD theory.

[SIC]

Airbus has great safety features built into the control system -

BUT I think there are too many modes on that PFD. Boeing PFD by comparison has less information and is clearer.

Any pilot knowing his game/plane etc can deal with it - but the reality is that these days when you look at the " lowest common denominator " type pilot ( ie non english speaking undertrained inexperienced etc etc so common in many smaller companies in less regulated places) then a simpler display and mode system WILL help.
I personally took a while to get used to stupid things like ALT* vs ALT etc etc.

[hetfield]

Good point! If I remember correct when A320 went into service words like "crashproofed" or "needs very less training" out of the mouth of AIRBUS representativs come into my mind.

Obviously the opposite is true.

[sudden Winds]

Being an actual Boeing driver and having already started studying the airbus manuals, to become a scarebus driver in a few weeks I can tell you that I have a lot of respect for the kind of thinking put into airbus. There´s nothing wrong with the computers, the flight laws or the FMS (of course you didn´t need me to figure that one out).
The non back driven throttles are not to be considered a conventional set of throttles but a thrust rating panel...
there´s no need for those throttles to move, you just look at the gauges and the rest of the instruments. The FMS is a bit more complicated than our U10.6 at first, but they got all you need and are as useful as any other FMS.
The sidestick is another marvel of technology. Its flight laws are great and will let you fly the a/c no problem to its very performance limit.
I think airbus is difficult to learn at first, because with the first type rating you obtain, you ´re introduced to their philosophy...but once you´ve digested that...transitioning to the rest of the a/c is like a walk in a park. Congrats Airbus...
PS1: my heart belongs to the most beautiful and lovely airplanes in the world...Boeing.
PS2: isn´t it wonderful to be a pilot? you fly both, enjoy both and let them fight indefinitely to be better than the other one...
PS3: one thing I do like about airbus a lot is the amount of information contained in their manuals and their FCTM. It´s great to read them.

[dessas]

Obviously Airbus are still not happy enough with the Russian report...
http://www.flightglobal.com/articles/2007/07/12/215449/french-investigators-want-further-study-following-armavia-crash.html

[FlyGooseFly!]

Posted by Hetfield :-

Good point! If I remember correct when A320 went into service words like "crashproofed" or "needs very less training" out of the mouth of AIRBUS representativs come into my mind.

Yep, and I've heard Boeing guys say the same about the 777 and now the Dreamliner!

Quite frankly, I am amazed at the amount of "stuff" all you modern heavy drivers have to learn and keep at your fingertips - it seems that the more "automatic" systems become, the more difficult it is for humans to interact with them properly at times of stress.

Yet every time someone mentions fully automated flight, a cry goes up that a human pilot will always be needed and reassures the SLF but I wouldn't like to put money on the scenario of disasters caused versus disasters averted.

This investigation is at pains to point out that the situation was recoverable right up to the last five seconds of flight - yet the cockpit crew were consumed by the problem and this proves to be the case over and over again.

I have read far too many "horror comics" that list - in the cold light of day - the sad demise of jolly nice chaps you'd be happy to have round to dinner - making complete horlixxes of the most simple tasks - especially the guys with many hours on type. Give me a clean cut looking young'un who looks like he/she takes their "pubs" to bed to read up.

[Tee Emm]

This investigation is at pains to point out that the situation was recoverable right up to the last five seconds of flight - yet the cockpit crew were consumed by the problem and this proves to be the case over and over again

I know very little about the automatics of the A320. But after reading the Armenia accident report I got the impression that precious time was lost as the crew fiddled with the automatics and seemingly lost situational awareness with a fatal result.

In the A320, is it possible to promptly and with ease to disengage the FD, AP and AT and simply conduct a raw data manually flown go around and then when in a stabilised safe situation introduce the automation again?

[A4]

Tee Emm,

Yes it is.

A4

[Pollution IV]

I agree Airbus is overall a good design, however, its shortcomings quickly become evident in non-standard situations such as faced by these poor chaps.

I had a very similar incident in the 320 at a busy airport, A/Plt on, late ngt, ~1900ft, just selected the gear down, when ATC instructed us to go-around giving a 90deg hdg change and clmb to 4000', plus a change of freq. It was winter and the acft was lgt, so upon selection of TOGA the performance was startling. The AP kicked-off as it pitched to ~25deg NU (until I stopped it), and the ensuing call-outs, clean up procedures, radio calls (including change of assigned alt), tfc awareness etc caused a pretty hi workload to say the least. I have done appch go-arounds from intermediate stages on pevious acft (generally hand flown) which all seemed rather routine compared to the Bus dumping it all in my lap at such a crucial stage. The way the automation handled the go-around was completely unsatisfactory and alarmed the passengers and cabin crew - if a pilot handled the acft that badly in a sim check, he/she would certainly be in for a harsh debrief!

My point is, that the automation definitely has limits and can only be trusted under routine circumstances. When a situation like this hits you, the basics of flying- ie a good instrument scan while hand flying, is essential. No amt of extra CRM trng or automation knowledge is really going to help and it was my previous years of experience and trng before joining the airlines, that helped me handle the situation satisfactorily.

Although my Airbus rating trng was very thorough by industry standards, I had never seen an appch discontinued anywhere but at minima, nor had I done recovery from an unusual attitude. Yet 2 airbus acft have been lost in recent years due to mishandling such a sequence. Aviation Regulators worldwide should be heeding the lessons from these accidents and altering their regulatory requirements, thus forcing airlines to adjust/improve their trng. Airlines will not do this on their own as the extra sim trng costs money, so they seek the cheap option of improving CRM/TEM awareness - this alone won't save the day. A meaningfull improvement of hand flying standards in this industry is what is required - afterall, it wasn't the automation that flew the acft into the water.

Heavy reliance on automation and de-emphasis of hand flying skills, as the acft manufacturers and Regulators stringently endorse, is just going to lead to more of these type of accidents. Furthermore, incidents like these are not limited to the lesser airlines in obscure parts of the world. A few years back, my airline had a very similar incident which almost ended in disaster due to the Capt not taking ctl of the acft during an automated MAP. Fortunately, due to an experienced FO and good flt deck discpline (ie direction of attention rather than fighting over the ctls) the acft was safely recovered.

Lets all focus on the real issue of flying skills, not just look to better wizz-bang automation to do the job for us. :ok:

PS: Mr Airbus, please make an acft with a PFD that is at least slightly bigger than a postage stamp!

[A37575]

Having read extracts from the French accident investigation BEA report on the Armavia A320 crash, (Flight International 17-23 July), I was puzzled by the mumbo jumbo terminology.
Example:
"The BEA believes the captain's psycho-emotional state was the result of the unexpected instruction to abort. It says the crew did not expect any more disturbances after the aircraft was cleared to land. The order to stop the descent, which arrived 46 seconds later, was thus completely unexpected and ran counter to the pilots' mental representation of the situation. This destabilised the crew, already annoyed with the controller, in particular the captain, who reacted to this instruction rapidly, and, it appears, without developing any strategy."


One wonders if these words are written by a current pilot or a psychologist. The latter seems more likely. Does a pilot really have to develop a "strategy" in order to conduct a go-around in an A320? And why on earth should a crew become "destabilised" if pissed off with ATC?

[jbsharpe]

I wonder if it isn't a case of lost in translation? :hmm:

[Radix]

To my sense it more or less can be interpreted that the captain behaved reactive where he should have reacted proactive.

Could be a translation problem indeed.

[LEM]

I would like to add just one thought:

In critical situations, pilots react somewhat instinctively, pulling out old habits radicated at the "imprinting" stage of their formation.

Airbus is different.

Better or worse, but it is different.

My point is about the difference from the basics acquired at the very early stages of initial training.

Instead of learning how to fly on a Piper or a Cessna, future Airbus pilots should learn on mini-Airbuses.

After a newly born dog sees her mother for the first time, it's very difficult to make him believe he's a cat, no matter the amount of training to persuade him.

I still think Airbus design is unnatural and alienating.
Alienation in this case is: the lack of direct relationship with elements, ie fixed thrust levers and non interconnected joysticks.

I know training pays off, but when things go to worms, our brain desperately reverts to basics.

The basics learned on a Piper.

LEM