PPRuNe Forums - Spanair accident at Madrid

PPRuNe Forums (https://www.pprune.org/)

- Rumours & News (https://www.pprune.org/rumours-news-13/)

- - Spanair accident at Madrid (https://www.pprune.org/rumours-news/339876-spanair-accident-madrid.html)

This has been covered in a post, looong ago, but here's a recap

If the airplane is in ground mode, here are a few obvious hints the Mad Dog will throw at you. Given their obviousness it's quite an effective system still for it's enginuity in verbally telling a pilot why the plane is not set for takeoff as is the checklist designed to filter out the chance of the crew not knowing.

if the plane is airborne......

1. You will get RAT probe heating on the ground. this is confirmed in the before starting checks as you turn on the anti ice, and set the knob to RAT current to verify it is at 0.

2. When taxiing out and doing flight control checks the "SPOILER DEPLOYED" caution does not illumintate as it should. This should raise a flag immediately, as the flight control check is vital to any crew.

3. The flight mode annunciators would read GO RND instead of TAK OFF when the TOGA buttons are pressed. So if you have you AP / FD system setup correctly, TAK OFF will not appear in the FMA, but GO RND. A definite clue that the plane thinks it's in the air.

These are the obvious ones. Because iether they are part of a checklist that must be accomplished or because the evidence is in the form of green LED's staring you right in the face with wrong info. Info which you verify verbally before every takeoff.

Indeed, if the plane is in airborne mode, the TO warning system is deactivated. This requires no explanation me thinks.

Xander

AOB9;
My (former) airline (I'm now retired), extends flaps to the takeoff position, (as planned and calculated during the cockpit check), after engine start and before taxi as part of the After Start Check.

The only time that procedure might not followed is in winter where the aircraft is to be de-iced first because it is snowing sufficiently so as to contaminate the inner leading edge surfaces (behind the slats) as well as the surfaces of the slotted flaps. The "After Start" check is then completed. The slats/flaps selection is left to just prior to takeoff if there is either freezing precip or wet snow falling - same reasons.

In the light of some comments on the thread, some may wonder about this procedure but in such weather, one is caught either way by risks which must be managed: - contaminated surfaces and/or a checklist item, either of which, if not respected, can produce the same outcome... The SOP is to extend slats/flaps just prior to takeoff in the above conditions, (which of course present their own separate risks and requirements).

However, the Airbus has a robust TOCW system which is tested as part of the Before Takeoff checklist and which will sound an aural warning plus light the Master Warning (red, for non-pilots) in front of both crew members when the thrust levers are placed in the FLEX/MCT or TOGA position. The TOCW is cannot be disabled by pulling circuit breakers.

I have done a max-pitch takeoff in the sim, (sidestick full back to the stop) where the alpha-prot mode maintains a "suitable" AOA (pitch around 25deg) while enabling 20deg-bank turns...once again, with slats/flaps in the takeoff position, (1+f). To be clear, I am unsure how this aircraft would perform in any slats-up/flaps-up takeoff but I'll bet some are trying it out in the sim as we write.

I think all planes will have the same happen to their TOCW system if the plane thinks it is airborne. The TOCW system will cease to work. Dirrefent aircraft will have different ways of telling you directly or indirectly if it is.
On the MD80 it will be the ones mentioned in my previous post. On the A320 it may be that you get presented wtih the landing config list instead of the TO config, maybe accompanied be something more directly.
In short, thinking it is in airborne mode disables all planes' TO warning system. That's more than logical I think.

Xander

We also changed our procedures to call for flaps after the before taxi check, but prior to calling for taxi. Our procedure was changed after it was discovered by FOQA that the flaps were not being set early enough in the process.

TOCW locked out in flight mode in not really the problem, it's the ability to be in flight mode while on the ground. How does this situation come to be (how does flight mode while grounded begin), and why hasn't this problem been addressed (assuming it's been known for ages)?

xander:

Quote:

I think all planes will have the same happen to their TOCW system if the plane thinks it is airborne. The TOCW system will cease to work

With respect, you simply can't make such a generalization. Not, at least, unless you've got Airbus 320/330 time and even then the statement is almost certainly incorrect. Christine works in mysterious ways our wonder to behold...

TOCW in flight is a huge problem, usually as the result of faulty prox sensors. Aircraft believes it is on the ground in flight. T/O configuration horn sounds, believing flaps and slats / leading edge devices are not configured for T/O. Outflow valves open and pressurization is rapidly lost. Mode S lost due to aircraft believing it is on gdn. The list goes on.

The information from the judiciary investigation continues not giving any probable causes. There is obviously not some "major event" that can be plainly observed to have caused the accident.

700 pages, no probable cause found by the judge so far.

Crew workload the past few weeks was normal. Judge wants to question the mechanic technician that did the probe heater "repairs" again as he didn't mention the first time about the reverser having been locked out of service 3 days before (presumably by two fellow technicians). The judge wants to know if he noticed what state the reversers were in.

Another (unspecified in the article) maintenance issue appears a few days before on a flight to/from Barcelona.

July 26th the airplane was taken back to maintenance area while taxiing to t/o from Palma de Mallorca because the pilot felt the front landing gear was making more noise than usual. Mechanics checked for 36min and didn't find anything wrong. The pilots and technicians agreed it was probably caused because the gear was too new (I guess it had recently been replaced/overhauled). Flight continued without incidents to Copenhagen. (Maybe the valve/sensor that feels the weight to indicate "ground mode" was involved?)

So the judge is waiting for more information from the technical investigators to figure out the probable cause and any (if at all) criminal responsabilities.

It seems that only analysis of the FDR (and sound recorder to an extend) can really shed enough light to figure out what most likely happened.

Trim.Flaps.Spoilers

PJ2:

Quote:

As such, I personally would have nothing further to say with respect to "killer items" as this thread covers them sufficiently

Thanks for your further clarification and patience. I will read the books you note. I think I might have misunderstood your original drift. Depending on the final outcome of the investigation of course, it is a little discouraging to know that nothing more can be done to eliminate human factors in such regard. Statistical significance having been understood.

philipat;

Quote:

Depending on the final outcome of the investigation of course, it is a little discouraging to know that nothing more can be done to eliminate human factors in such regard.

I understand perhaps, but to remain discouraged is to remain discouraged with what is and will never change, and not just in aviation.

The goal to "eliminate" human error is a mistaken goal. Recognition of that fact alone is what can make aviation safer for as soon as one sets aside any idea that perfection can be achieved (or that one actually has it already!, and believe it or not, such hubris exists within aviation today), one can make progress in flight safety, first by being on one's guard and second by building anticipatory/preventative procedures/systems which are at once robust, but not brittle.

The term "brittle" means that a design must be at or close to perfection before it will work properly. The opposite term might be "resilient" but I'm not sure that entirely captures the full idea that pursuing perfection so that mistakes don't occur, is itself, a mistake.

Judge has watched 18 video recordings concerning the accident (I'm assuming many of those of witnesses nearby AFTER the crash itself, but including the 7 seconds airport recording the captures the exact moment of the accident).

Preliminary technical reports received by the judge by technical investigators concentrating on the engines confirm the engines neither exploded nor catched fire prior to the accident. Most of this conclusion comes from analysis of the airport surveillance video recorders that captures the accident and witnesses.

Most spanish press gives this scenario as certain: Pilot tried to take off until the last possible moment, but when he touched ground again he deployed the only working reverser (left) while the right one couldn't be used as it was locked-out-of-service for maintenance.

6 of the crew members on board were on active duty. Another 4 crew members were flying avant.

Sitting next to 1C seat of crew survivor flight attendant Antonia Martinez was another Spanair pilot, an airbus commander. The flight attendant couldn't hear anything from the cabin (she never was there) but she recalls the comments of her partner sitting right next to her (the airbus pilot) as soon as take-off commenced: "Something weird is happening here" (referring to the way he perceived the take-off maneuver).

She continues in her declaration to the judge: "The airplane couldn't climb, like if it wasn't able to cope with the weight of the plane, it zig-zaged laterally and I thought it was trying an emergency landing. The lateral moves were very abrupt, the airplane was out of control".

She didn't see any fire. She found herself in the creek's water undeneath some probably metal sheet of fuselage and someone was pulling her hair and shoulders trying to help her.

She confirms the pilot didn't address the PAX during the accident (obviously not enough time or a priority).

Last known communciation from him to air controller Enrique. A. who lost visual contact with the plane at 14.24 (until flight voice recorder communications are recovered and anaylized). The pilot repeated the parameters given by the controller to take-off and acknowledges having permission to TO.

Judge declares work fatigue not a factor, as both pilots were under 60 hours in 20 days.

One of the survivors declares the "sudden roll movements" so abrupt that in 2 seconds he could see through the window "landing strip, sky, landing strip again". He found himself still strapped (buckled up, I guess) to his seat with one of his arms trapped underneath a piece of fuselage.

Most survivors accounts of the accident are very imprecise. They all talk of normal feeling take off, sudden weird left and right movements and vibrations, and then crashing.

Initial rumors about the left engine catching on fire came from an airport worker (no idea what they are called in english, it's usually the guys that drive around the "follow me" cars or perhaps make indications to the airplane to turn right/left with those funny flashlights). He is one of only two sources for evidence, though, about such a thing, which wasn't captured by the video recordings etc.

The judge received the testimony of a flight attendant from a nearby plane incoming from Guayaquil. She didn't see the airplane herself but declares to have heard a comment by her flight's commander: "He is eating the strip" (referring to the Spanair airplane using almost all available runaway, although remember at least another 600m of paviment was available). She said the pilots commented on that engine "flashing" and the MD-82 "falling like a leaf".

Many witnesses speak of abrupt roll but also of some yaw (after leaving the ground).

Air traffic controller in charge of the flight declared to his boss: "The airplane I just authorized to TO, I think it fell down. The nose went up, but couldn't lift-off, started to roll side-to-side and then I couldn't see it anymore".

The control tower supervisor said she had heard some vague talk about an Spanair flight with engine problems earlier in the day. But evidence to that doesn't show up anywhere.

The weight of the airplane according to ground papers was 64.263kg. 11.189kg of fuel. Most cargo has been destroyed in the accident, only a few pieces have been recovered.

22 pieces of luggage belonged to the crew. 115 pieces to PAX. Plus undetermined carry-on.

It carried 400kg of fresh-fish cargo on a refrigerated container. Also some cargo for shoe manufacturer Timberland. A live animal (a dog, didn't survive the accident, has been retrieved).

If inop TRs are found to be a factor at Madrid and Sao Paulo, as pax I'd like to see any TR inop situation make an aircraft unserviceable

22/04 - Difficult to scan 1200+ posts, but a previous one has already stated that aircraft are certified with wheel brakes only, thrust reversers are a bonus: without them a/c meet regulatory specifications, with them they exceed it.

If you expect TOTAL safety in any mechanical device operated by man, then stay in bed - where you will eventually die anyway, - sorry to sound callous but this is a real World governed, and not too badly in fact, but economics.

That's not trying to minimise the suffering, just being realistic.

Any news about the FDR condition ?

Also CVR was supposed to be in good shape - surprisingly no mention so far in official briefings

Quote:

Most spanish press gives this scenario as certain: Pilot tried to take off until the last possible moment, but when he touched ground again he deployed the only working reverser (left) while the right one couldn't be used as it was locked-out-of-service for maintenance.

How can the Spanish press be certain about anything regarding this accident at this stage with an investigation barely underway in which FDR and CVR data are still under investigation? Looking at recently posted pictures, there is very little left of the aircraft and regardless of claims made early on in this thread, i have not seen any pictures of anything resembling a piece of wing, let alone one which would show retracted slats or flaps.

Furthermore i repeat my question from post #1202:

On the MD82, with ATS engaged and TOGA or FLEX selected, in the event of a reverser unlock during takeoff, is the thrust lever automatically retarded to idle by the reverser mechanical interlock system and the ATS disconnected? If such an event occurred after passing 80kt or during rotation would an alert be inhibited until an altitude of 1000ft?

On an aircraft type i worked on several years ago (no MD8x models) this is the case and a thrust reverser unlock alert is inhibited between a speed above 80 kt until 1000ft AGL. If in this accident, a TR unlocked near rotation or at lift-off, and with similar design/certification criteria no alert was presented, how would the crew react to a thrust lever retarding to idle? I assume, if the MD82 has a similar interlock mechanism and alert inbit it must be part of training to identify such an event as a possible uncommanded TR deployment?

Whatever happened, things happened fast and the crew must have had very little time to identify the problem.

Regards,
Green-dot

Yeah, it has been mentioned before. FDR recorder data has been successfully retrieved and it's being verified and analysed by Boeing.

Voice recordings are not of good quality (not fully understandeable), so it's been filtered and enhanced in labs prior to analysis.

Green-dot: yeah, you know how reliable the press is. Since we don't know if take off took longer than normal before or after VR, it is indeed still possible a scenario where a reverser accidentaly deployed around VR. But the flight attendant speaks of the Airbus pilot avan speaking of "something weird" just about as soon T/O was initiated. One would like to assume a pilot with over 10 years of experience would've noticed a deployed reverser before V1.

The press also insists on this airplane not being able to deploy reversers unless on the ground, but I think others here have confirmed that the MD-82 indeed can have them deployed while flying, even voluntarily (so even more in case of a malfunction).

Have there been cases of MD-82's with truly uncommanded reverser deployment under "normal conditions" (i.e. totally smooth operation, no other problems, turbulence, bad weather etc)?

Thrust-reverser related accidents. (from Wiki.)

In-flight deployment of thrust reversers on aircraft not having that function as a design feature has directly contributed to the crashes of several transport-type aircraft:

• On 9 February 1982, Japan Airlines Flight 350 crashed 1,000 feet (300 m) short of the runway at Tokyo Haneda Airport following the intentional deployment of reverse thrust on two of the DC-8's four engines in an apparent suicide attempt, resulting in 24 passenger deaths.

• On August 29, 1990, a Lockheed C-5A crashed shortly after take-off from Ramstein Air Base in Germany. As the aircraft started to climb off the runway, one of the thrust reversers suddenly deployed. This resulted in loss of control of the aircraft and the subsequent crash. Of the 17 people on board, only 4 survived the crash.

• On 26 May 1991, Lauda Air Flight NG004. The Boeing 767-300 aircraft suffered an uncommanded deployment of the No. 1 thrust reverser, which caused the airliner to stall and crash.[1] All 213 passengers and 10 crew were killed.

• On October 31, 1996, TAM Linhas Aéreas Flight 402. The Fokker 100 crashed shortly after take-off from Congonhas/São Paulo International Airport, São Paulo, Brazil, striking an apartment building and several houses. All 90 passengers and 6 crew members on board died. Three people were killed on the ground. The crash was attributed to the uncommanded deployment of a faulty thrust-reverser on the right engine shortly after take-off.

I knew of several accidents most likely caused by an accidental reverser deployment, but being, from my understanding, an active system that requires of hydraulic assistance to be deployed (under normal, not extreme conditions such as strong vibrations, etc), I always thought it would be quite "impossible" for one to deploy "just like that".

Again, I understand it becoming loose or whatever from bangs, vibration, extreme weather, etc, but thought that under normal conditions an ACTIVE rather than passive type of malfunction would have to be necessary to make them engaged.

If it is a design problem that allows for a malfunction to trigger them, wouldn't it have been solved by now? Nobody would fly a thing that under some "normal" circunstances could enter reverse w/o warning ...