Scare Bus, Die by wire.

One thing I do not quite understand from the posts here: With this wired up incorrectly would the ECAM show the correct direction of movement on the control check or not? When you do the control check you should certainly should check for both movement and the direction of movement of the ailerons.

There certainly have been cases of cross controls in older aircraft. I remember the China North West TU-154 which had the autopilot incorrectly connected. Unfortunately the flight crew elected not to disconnect the A/P and the aircraft crashed. However this possibility should be designed out of modern fly by wire aircraft. The fact that the wiring connections change for aircraft with different serial numbers with aircraft of the same type is a serious design flaw.

Flap 5:
that's my point. And yes, the ECAM indicates the correct deflection associated with the stick input during the check, despite the surfaces actually move opposite.

You would need somebody outside the aircraft to check visually if the aileron deflection is appropriate and consistent with the sidestick input.

Any A 320 rated techi out here?


Rgds
TvB

[This message has been edited by TvB (edited 23 May 2001).]

I can see how aileron could indicate correctly when moving in the oposite sense but what about the spoilers?

I like to know that when I make a control input, the input is transferred through a clever combination of wires, pulleys, cranks, control rods, hydraulic pistons and actuators, toa control surface. I can't knock what I haven't tried, and maybe I belong in the ark, but I prefer computers and software to help and assist, not decide how to interpret what I've demanded.

I may well end up on the "electric jet", but at the moment, I'm glad I'm partnered with my trusty old Boeing.

Full marks to the Airbus crew for the way they handled this one. I only hope that given the same circumstances I would have reacted in the same way as that excellent FO.

I guess I'm just afraid of what I don't understand, and maybe that's why they put me on a Boeing (simple, like me).

Mowgli

Better drive a carburetted automobile then. It would be scary to have a computer 'interpret' your throttle demands. Or would you prefer it if your conventional throttle cable snapped just in the middle of that tricky overtaking manoeuvre with the 40-tonne truck bearing down?

Had it occurred to you that conventional flying controls can be jammed?

The F/O on this occasion reacted as he was trained and isolated the offending sidestick. (Yes all you sceptics - what if the Capt had, say, a seizure at the controls at the same moment on takeoff, his arm locking on his sidestick? I'm sure all A320 operators at least introduce this idea during training)

Mowgli & Sops

You lot are missing the point.
So maintenance screwed up,not unknown to happen,but the very fact it was fly by wire saved them.Not possible with pulleys.

3 putt:
I believe, with all due respect: you miss the point.

To the opposite: FBW almost killed them!

1. We are talking something short as 5 seconds before the wing would have hit the ground! Apparently there was light turbulence, so what if another would have hit the aircraft while flying with takeoff pwr, wingtip 1.6 feet above the ground? Do you know how damm close this is? Do you know how many fractions of an inch of additional false/right sidestick input by the Cpt. would have slammed the a/c right into the ground?

2. Yes, somebody at LHTechnics had screwed up. But: how can somebody design an electrical flight control input connector and use different wire diagrams, even within the same series of airplane? This was an incident (luckily no accident) waiting to happen. Such a device has to be designed dummy-proof and failsafe.

If you look at the famous "chain of events" leading to an accident, how many parameters here did almost make it happen? The well known 5, 4, 3, 2...?


3. And the ECAM indications showed proper aileron deflection while they were deflected to the opposite direction? So far I've not seen any A320,330,340 jockey looking out the window to the rear and checking if the ailerons move according to his input. Besides: it would be against the AI procedure. - So how should somebody catch this fault?

Now, and I know we are becoming hypothetical here, what if both sticks have been rewired according to the wrong wiring list. The f.....g ECAM would still show : OKAY!!!

This glitch has to be fixed ASAP and the good thing about this article might be that the story is out now. Might trigger some special attention in the respective workshops and maintenance facilities worldwide. Apparently the incident happend in March. So I believe AI was well aware of it by mid-April. Checked
company bulletins today: there is NIL! A simple bulletin would do the job in terms of "prevention".

This is what bothers me and it is symptomatic with Airbus Industrie and their birds.

So thanx God there was a very good trained crew, a F/O very, very quick and awake and only this one problem, no further wake, and it all happend with an airline and in a country where they didn't hide the event under some rug or somewhere else...

Suggest you doom merchants and Luddites have a look at TvB's reply on page 2. It is possible that FI's editorial team may not have let all pertinent facts through.

------------------
Bus429 - the pilot's pal!

sj

>If you look at the famous "chain of events" leading to an accident, how many parameters here did almost make it happen? The well known 5, 4, 3, 2...?<

Break the chain by visually checking control surface movements following the mtx. Surely applies to any a/c, fbw or pulleys particularly if it's the control system you've been d1cking with.

I think this illustrates not that computers are infallible - we all know they are not, rather we have come to expect them to be. Many millions of trouble-free hours can induce complacency. Yeah, the design could be better, but simply checking the work would have prevented this particular chain.

Bad practices, good crew.

SOPS and Screwjack,
Any of you want to start a string on B737 rudder-hardovers, exploding centre tanks, or the Egyptair 767 "suicide"? Boeing design at its best? Or what about Concorde's flimsy fuel tanks?
Don't get me wrong, I think all of the above are great aircraft and I hope they'll keep going for as long as they can, really.
In this Incident AI obviously messed up the design, yes. But try to get a slightly more open mind about aircraft design please. We all think the fact that the ECAM didn't show the fault is outrageous. But WHY didn't it? If the system only measures deflection, how does it know if the direction is correct, or doesn't it? Believe me, the designers really were not thinking "oh, it'll never happen..." It's a bit more complex than that. Sometimes the reasons why things are designed a certain way aren't immediately obvious... Any AI people out there with an answer?

Quite right - why is it that a genuine thread is so often hijacked by a bunch of folks who know nothing about the subject they are harping on about?

Screwjack - you have simply demonstrated in a particularly extravagant way how very little you understand about this technology. 'Slammed into the ground' - go back to reading the Daily Mail

Sidestick failures and inadvertant inputs are covered in the AI course very early on.

Do we really have all the facts about this incident? As I said before and E cam reiterated - what about the spoilers?

Having recently come to the Airbus from Naval Aviation where our control deflections were always visually confirmed, I have always been concerned by this aspect of the flight control system. Since I cannot see ANY of the the controls move from the the flight deck, I have developed my own personal check. When I do my walk round I look at the actual control surface positions and compare them to what the ECAM flight control page displays. If the display agrees with what I have seen, then I trust it. For what its worth.

The Airbus 320 flight control surfaces are unpowered until engine start - are you sure this is a valid check? Curious as to what Airbus would have to say about this. Also, the control deflections are easily checked on the ECAM screen. At my company, it is SOP for the captain to call out his sidestick movement, and the f/o must verify the actual movement is what is expected.

M14P, Frederick and Oilhead,

1. I don't read the Daily Mail!
2. I believe you get me wrong here.

Of course no a/c will be ever a 100 % failsafe and there are numerous things the engineers and with AI specially software designers could not think about at the time of designing their blueprints. That's okay and nobody blames anybody for that.

BUT:
once a glitch or serious item has been identified, I expect from a key player like Boeing or Airbus to inform the rest of the community in an appropriate and timely manner. A simple bulletin, describing the facts of the incident will do it and subsequently everybody out there will try to avoid it. So here we have a case of a serious design deficency which was never adressed since the first FBW Airbus had been built a decade ago.
And it applies to the entire family (320,330, 340).

The AI design tends to give pilots and even some mechanics a wrong impression of precission and safety. Sitting in the cockpit you can't see all surface deflection, while you perform your flight control check. The SOP says PF performs the stick inputs, PNF checks the ECAM and then vice versa. So everybody tends to believe what is indicated on the screen is correct and corresponds with the actual deflection.

I recall a daily proceedure with my former European airline where the flight controls were checked every morning before the first flight with a mechanic outside the a/c, hooked into the intercom. Some egghead found out that this delays the operation by 6 minutes and since it is not performed any more...

- I wonder how the folks in Seattle have resolved the issue with their B777, but I tend to believe such a glitch can't happen here, at least it would be sensed and a warning triggered.

But this is really not that important.Let's face it: It is the "ATTITUDE" of Airbus Industrie, gentlemen!

When it comes down the line I don't see a "responsible reaction" in dealing with their product. In the past they were pretty fast in blaming the pilots for each and every mishap (latest was the Northwest A 320 take-off accident at Detroit some weeks back) instead of turning their attention inside and ask themselves: "what could we do to avoid this situation in the future?".

When it comes to information in terms of product support you get a bunch of pages, revisions and so forth on lavatory lights and other stuff I don't consider as important for flying the a/c, but you won't get a bulletin stating the incident, the findings, what really happened inside their computer brains and what you should try to avoid.

We all know that safety in our job is the no 1 priority and everybody tries to perform at his best. This includes prevention and the tool for prevention is information. I believe the industry has a massive problem, when it comes to this.

Have a good one

If the flight controls have been disturbed, then isn’t it normal to have a requirement for a flight control check entered in the log book? This has to be cleared before flight and requires an outside observer on headset to confirm the surface movements versus pilot inputs. If this isn’t standard practice, then maybe it should be.

Regarding the ECAM display...DoctorA300 wrote that the amount of aileron deflection is measured directly but the direction of deflection is determined from the position of the sidestick?

Well how is direction determined when the aileron deflection is commanded by the AP and not by the sidestick?



[This message has been edited by stagger (edited 24 May 2001).]

SJ
(I've read your post very carefully)

You should really watch your blood pressure! and yes I do know how close "1.6 ft" is.

Let's not get into the AI versus Boeing nonsense(it's been covered and what are they going to do?take all A320 a/c out of service)
I simply made the point that the design of the fbw is such that the f/o was able to split the stick and thus recover the a/c.Now I would imagine that this is a simple statement to understand.The fbw saved them,it would appear mtx let them down.
Are you suggesting that it is better to find yourself at the controls of a Boeing or Cessna after such an alleged mtx release?
A flight control is a flight control and it does'nt care if it is bolted to an AI or a boeing.It does exactly what it says on the tin.

Case One, When exactly do you do your walkround?

bonajet,
nothing was wrong with the flight controls in that incident. the a/c came back to FRA with one ELAC inop (according to MEL you can conduct a flight with this handicap...). At FRA a troubleshooting was conducted and in the process they found one pin on the sidestick connector bend. therefore the connector was replaced, rewiring a new connector according to the respective (in this case 'wrong') wiring list.

The controls were then checked inside the cockpit with the ECAM indications and found okay. nobody had reason to believe that actually there was a problem with the ailerons. No work was performed there.

does this clarify it a bit?

rgds

tvb

You people make us sick...this is a public forum where all Ppruners are welcome to post their RATIONAL comments/observations/ideas on ALL topics..so why jump on 411A..I happen to know that person was an aviation professional, when most of you unionizing crybabies were choking pablum down your mothers' back...we all learn from each other, even from the most unlikely sources...so if you're too good and knowledgeable to acknowledge/ignore 411A's posts, get out of the cockpit before you kill somebody....

Just to make it clear where I stand, I enjoy flying the 320, but I have no intention of letting the same thing happen to me if I can help it.

Oilhead: AI SOPs call for the PF to move the controls and PNF to check deflections on the ECAM F/CTL page. My point is how do you know that what ECAM tells you is correct?

screwjack: Yes, AI's attitude is a problem. Mechanics doing visual checks at dispatch would be great, but I don't expect that to be implemented (cynic).

3 putt: Banter aside, before I get in. Our 320s are fairly old and the ailerons, elevator and rudder are almost always deflected from neutral when the jet is shut down. I have noticed that on the new jets being delivered the ailerons are almost always neutral at shut down (drat).

Its not perfect, but until AI or my company comes up with a better system, its all that I can do personally to make sure that I stay alive. Any better ideas wellcome.

TvB

From what you say It would appear that the ECAM only shows the sidestick position and not the actual aileron position. If that is the case there is little point in doing a control check.

When I place the sidestick to the left I know that it is to the left! What I need to know is where the control surfaces are! I was always told not to look at the flap lever but to look at the ECAM because it used feedback from the flaps to show the actual position of the flaps. If what you say is true then the feedback to the ECAM is not from the control surfaces but just from the sidestick, which is of little use.

Edited for typos

[This message has been edited by Flap 5 (edited 24 May 2001).]

TvB – my point is that if the flight controls are disturbed then a check is done with an outside observer watching the results. This is done if the sidestick was worked on as well as it is part of the flight controls. If the FCDC is swopped then the new unit is checked via the MCDU.
I am just surprised that there was no visual check requirement as there would have been in our airline. Was it not a requirement in LH?

Are we not talking about two different displays here? We look at two displays on taxi out on the bus. One for stick deflection, and one for surface deflection. You can see excactly what the ailerons are doing!

I can certainly see what the flight controls AND side stick are doing on the ground. In the air I can (electronically) see what the flight controls are doing too.

There is an ECAM Flight Controls Page which shows the position of the surfaces, not the sidestick. That page is either manually selected, or automatically appears on taxi out as you commence the full and free check. That page also shows ELAC and SEC status (FWIW).

Then there is the Primary Flight Display (artificial horizon in the old days, except this does a bit more) (PFD) Ground Display mode. After the first engine is started, a white box appears on the PFD. It shows the deflection of either sidestick, but not the resultant control deflection. It disappears after rotation. Then within the box there appears a little Maltese cross thingy. We call it the "side stick orders indication". It shows the combined displacement of Captain's and First Officer's side sticks. This way one can see the commanded side stick deflection during take off and landing roll.

Oilhead

[This message has been edited by Oilhead (edited 25 May 2001).]

Please excuse my lack of knowledge of civilian operating procedures, but I can't understand why a check of control surfaces was not done prior to take off. Are there reliable indications of stick movement in the cockpit? Or could a control check from the first officer have suggested the control surfaces worked correctly on the ground?

This check wasn't carried out in a hawk a few years ago (aileron cables not connected post maintenace) with tragic consequences.

This said, excellent effort by the FO and quick thinking no doubt saved countless lives.

TvB, are you sure the connector was at the sidestick and not the rack connector at the rear of the ELAC?
The sidestick wiring is relativeley simple with good access to work. Why would anybody rewire the capt's sidestick connector then move accross to the F/O's seat to check it out???
The pins on the rack connector are more susceptable to damage and more likely to be damaged troubleshootinfg an ELAC fault, the access is not as good, neither is the visibilty to seee what your doing so more likely venue for a f*** up.

Thank you Case One!!!

Finally someone who understands my concern.
Happy Landings and a long life.

Screwjack

bonajet, flap 5 and pengineer!


From what I got through the research of this incident:


It was the CM1 (cpt) sidestick connector, a 560 pin connector (4 segments with 140 pins each).


A flight control check on F/O's side was performed by the mechanic, but only inside the cockpit, using the ECAM indications.


A total of four wires (2 for the roll and 2 for the associated control channel for roll) were reversed in polarity. This apparently made the ELAC believe everything is okay and consistent. Indication on the screen would show proper movement of aileron according to desired actual input but actual movement would be opposite!?!


To my knowledge the AI logic and computers do not compare flight control input/output vs. actual aileron deflection (sensed at the aileron).


Lufthansa did not answer the FI request for comment. No addition information was therefore obtained in ref. to their specific procedures how to perform this flight control check after changing the connector.
But I believe they do it quite to the book, what ever the book says (they are Germans for Christ's sake!!!).


One source at LH told me that there is no requirement for a specific flight control check simultaneously inside/outside after this kind of repair. Could not get this confirmed as AI did not comment either. Possibly different company procedures might apply for different operators, but I really don't know.


Maybe those of you with access to "basis" are able to retrieve more information on this incident.


What some of you guys might find interesting though is what I got officially from Boeing in this regard (- of course I was curious how they have secured this important feature at their FBW B 777 -). Here is the answer from Seattle, WA:


"In the B 777 the EICAS will detect a single fault, as it was the case with the LH A 320. It (the computer-system) would have a problem to detect multiple faults (which is a common known problem with artificial intelligence systems).


The EICAS would generate an error message as it compares the pilots inputs verses the actual outputs. Each pilot input is translated into 3 independent signals. The computer also compares these 3 signals for plausibility and that they all agree.


Further it compares the ACTUAL surface deflection (left aileron up if control wheel input to the left, right aileron down and vice versa) to the desired control input.


Further: there is a specific mandatory maintenance task after performing ANY rewiring of ANY components. A special computerized test equipment is used to detect wiring faults along the line and they will be detected with this equipment.


Last there is a required visual check outside while somebody inside the cockpit performs the flight control check after repair to insure appropriate deflection an free travel of any control surface."


BTW: it took them (Boeing) 16 hours and 42 minutes to answer the questions, with a night in between!!!


Have a great weekend


TvB
(independent) editor aviation
www.aviationsafetyonline.com



[This message has been edited by TvB (edited 05 June 2001).]

Ouch!

I worked for QF for a while some years ago and their SOP was, before service, for the Flight Crew to visually check all the flight controls whenever any flight control system had been disturbed by maintenance.

This involved the F/E to take a headset (with a very long lead) and chat over the service interphone whilst the Pilots operated all the control surfaces.

We all thought this was a bit of a yawn in theory. I had to do it on a 747 S.P. on only one occasion. We discovered that the spoilers were not operating as advertised! (The most inboard spoiler was not doing it's thing, I've forgotten the detail, but it is different to a standard 100/200/300) By the way we only were doing this check because some other flight control had been disturbed. We relayed this info to maintenance who stated that it had probably left Boeing rigged in this manner! Nevertheless the A/C was dragged back into the hanger for investigation.

I hope QF still have this check as part of their SOP's, and it would certainly have done the trick in preventing LH's incident. Unfortunately my airline do not have this check requirement.

Well done to the Lufty crew.


Regards
Exeng

Kegworth - one engine on fire, other one thought to be and shut down. Subsequent proposals included CCTV to enable pilots to see what's happeneing behind them. Universally rubbished. How many dead was that?

A320 - pilots can't see control surfaces move for viusal check, airlines can't afford 6 minutes for an outside person to watch and confirm movement in desired sense. CCTV would do the trick, without time penalty, wouldn't it? How many have to have near misses like this one, and some die if the FO isn't as quick, before CCTV is though maybe to have some cost- and safety-benefits?

Just a thought from an amateur. (I was a motor industry beancounter - and no, I did not stop safety-related costs being incurred, I encouraged them because bad safety costs too much in the long run. And I fly gliders - visually check movement of control surfaces before every flight, with external helper if I can't see, and check very carefully after every maintenance episode - but then, what do I know?)

Chris N.

How does it go...

If you think safety is expensive, try having an accident!

JJM: - Well that was true in the past but today this "partial remaining risk" is extensivly covered by insurance deals. The only looser might be the airline if it happens to be a small operator and not one of the big players. Did anybody go down or out of business recently because of a crash? Singapore? Egyptair? Air France? Swissair? American? I don't think so. Therefore I tend to believe that some egg heads really do take a fatality into account (only from time to time) when they do their budget cuttings. Let's hope that the "professionals" will stay smart, qualified and fast enough to jump into this gap when required and the manufacturers find ways to eliminate the rest.


Change? Well only if the "consumer" would demand it, and he really likes to be shuttled for as low as 250 bucks half around the world (and back)... so what?




[This message has been edited by screwjack (edited 27 May 2001).]

Oilhead is correct, and Doctor A300 and a few others are mistaken. The aileron position info (direction and amount of deflection) displayed on the flight controls ECAM page of course is that of the ailerons! Four position sensors called LVDTs look at the ailerons. This is quite independent of the sidestick positions. This display would not have been affected or reversed because of any miswiring of a sidestick. The additive positions of the sidesticks is displayed until rotation on the PFD.
I think Pelican put his finger on it. On control checks during taxi out the pilots saw spoiler movement and aileron movement in response to commands from each sidestick,correctly displayed on ECAM. It just happened to be in the wrong direction for one stick. Really quite easy to miss at a very busy time and so unexpected. Every pilot does control checks, thousands of times. It is always correct isn't it? Except for this oh so rare occasion.

Bummer, avoman.

Where do I find the info about the LVDT's in the AOM? Chapter 1.27 doesn't give me a clue about this matter...

AVOMAN, STAGGER
I have to appologize, I wrote about the AIL. Indication, but mixed in the SIDE stick indication on the PFD.
AVOMAN, The Signal from the LVDT´s for the AIL. pass through first the SDAC´s and then the DMU´s. And as far as I can interpret the AMM the LVDT´s signal is converted to an 8 bit ARINC429 signal in the SDAC, the polarity (UP or DOWN) is derived from something else, which I have not been able to figure out yet.
But thanks for pointing out my mistake.
Brgds
Doc

So just to reiterate then:

The sidestick order symbol would have been correct but the ailerons would have indicated reverse sense on the F/CTL page (exactly as most sensible people would believe)

If this is the case are we looking at a chain of human error insofar as an engineering procedures error was then followed by a crew error?

So now we can all stop wailing and moaning about the hidden agenda/conspiracy theory run from Toulouse and admit that the Airbus FBW series are darned good aircraft after all.

DoctorA300, Don't the LVDT bits go through the FCDCs instead of the SDACs? I thought the FCDC sent their info straight through to the DMCs and then onto ECAM Flt Cntrls? Would seem more logical to me. I might be totaly wrong! Thanks for shedding some light on the mixup between the ECAM Flt Cntrls page and the sidestick position indicator on the PFD. You got me totaly confused there. M14P, Airbus great aircraft indeed, but different wiring orders for same connector on different aircraft of exactly the same type? Accident waiting to happen. AI has some explaining to do.

Frederick,
You are so right, LVDT´s go to the FCDC.
Brgds
Doc

Guys,
Not in the aviation industry, but the gist of what I saw here is that the sidestick was wired to a 320-200 specification, no A320-200 in particular, which turned out to be wrong. Had it ben wired to the specification of that particular aircraft, it would've been fine? Now, as for people saying this is a serious gaffe on AI's part, I have this to say. Do you honestly think every 737-400, A330-300, etc are identical in every technical detail? No, didn't think so. If airline maintenance has a proper document control system in place, these issues would be picked up. At the end of the day, the schematics used did not correlate to the serial number of the aircraft, this can happen anyone unless there are proper procedures in place. Any upgrades/changes by a manufacturer presumably go through numerous forums for an approval, and are then added to equipment configuration details. These equipment configuration details are captured in subsequent documentation, which is why it is so important to use the correct reference material for the exact aircraft being worked on. Is this even remotely the case, or am I miles off? Just curious, not trying to pi** off the professionals, but I'd like an insight into the workings of these issues...........