Spanair accident at Madrid

[md80fanatic]

I have only one question remaining. Judging by the wreckage, it seems the aircraft arrived at the creek with a high rate of speed. How can a reportedly undamaged aircraft travel off runway for 1.5 km, and arrive at the end with such force? No wheel brakes? No spoilers? Seems to me that this is nearly as important as the initial cause(s). Please correct me if I'm mistaken.

[SE210]

I tend to agree with FE Hoppy.

A lot of speculation to whether various systems failed or not. All ivery nteresting, but not plausible.

The flaps were not correctly set - propably not set at all. A mistake that was not arrested by the config warning.

The likelyhood of a flap fault and a config warning, that did not work is marginal.

It will propably turn out to be crew error - as the Detroit accident.

I think, we should focus on how to avoid these errors. Put focus on the ever more stressfull environment, we work in. Sub Part Q, pressure from the company etc.

Cheers

SE210

[Radix]

@md80fanatic: How? They were trying to fly, not trying to land

[justawanab]

md80fanatic,
It wouldn't matter whether the brakes or spoilers were working or not, at that speed, on that terrain it's going to take a long time to stop from the speed that they were doing at the time. I'd defy anyone to stop even a fast car in one piece from that speed following the same trajectory let alone something that is A.) many times heavier in mass and therefore significantly higher inertia and B.) already partially airborne.

[IGh]

Previous investigative research

From slot #1766, a few posts prior:

"... The likelyhood of a flap fault and a config warning, that did not work is marginal... probably ... crew error - as the Detroit accident."

That possible CAWS /power supply fault, and the P40 C/B fault, was discussed in DTW AAR 88/05 on pgs 52-56. Later, there were three FINDINGS and the P.C. that mentioned the possible faults:

"9. The CAWS unit’s takeoff warning system was inoperative and, therefore, did not warn the flightcrew that the airplane was not configured properly for takeoff.

"10. The failure of the takeoff warning system was caused by the loss of input 28V dc. electric power between the airplane’s left dc. bus and the CAWS unit.

"11. The interruption of the input power to the CAWS occurred at the P-40 circuit breaker. The mode of interruption could not be determined."

"... the probable cause of the accident was the flightcrew’s failure to use the taxi checklist to ensure that the flaps and slats were extended for takeoff. Contributing to the accident was the absence of electrical power to the airplane takeoff warning system which thus did not warn the flightcrew that the airplane was not configured properly for takeoff. The reason for the absence of electrical power could not be determined."

[Smilin_Ed]

Why Not Fly It Off?

My only experience with aircraft close to this size comes from a couple of flights with a manufacturer's pilot who insisted that I just rotate the airplane at Vr and let the plane do the rest. I preferred to fly the airplane off the runway but he insisted that rotating to a specific angle was the way to go. If the Spanair pilots had flown the plane off the runway, they would have instantly realized that something was wrong and would have kept the nose down and out of a stall condition until sufficient airspeed was attained to actually fly. Can an MD qualified pilot comment on this. Why just rotate? Why not fly it off? If it doesn't fly at Vr, then something is wrong.

[NigelOnDraft]

Why just rotate? Why not fly it off? If it doesn't fly at Vr, then something is wrong

Because all your Perf calcs are then invalid

NoD

[NigelOnDraft]

So...correct me if I'm going too far...what we have here is the uncomfortable possibility that - owing to a dicky WOW indicator - a perfectly functional RAT probe was disconnected, along with the TOWS, leading to the loss of the aircraft?

Errrr.... If the Flaps were not set for takeoff, I am afraid there it is probably SOPs / Human Factors that led to the loss of the aircraft.

Things like TOWS are backstops that ideally should work, but it is difficult to attribute the (primary) cause of the accident to the TOWS not working. IMHO. A bit like blaming a faulty airbag for a road accident fatility... it would be better to avoid the accicent in the first place.

NoD

[Smilin_Ed]

Because all your Perf calcs are then invalid

Of course they are invalid and that is just the point. By simply rotating without regard to how the aircraft responds, you are asking it to do something it simply cannot do. By gradually rotating, you will discover that it isn't ready to fly and can keep the nose down until you get enough speed. The Spanair pilots pulled the nose up too far for the actual condition of the flaps/slats. It's no surprise that it didn't fly but if they had eased the nose up and got no correct response, they might have been able to finesse it.

[XPMorten]

-14:24:06, crew called out V1
-14:24:10, CVR records a pilot saying "rotate"
-Data recording signals front wheel sensor change of state
-No sound from TOWS.
-Altitude reached 12,2m (40 feet).
-15 seconds after rotation is initiated stick shaker comes on.
-4 seconds later, aural stall alarm (synthetic voice sounds twice stall, stall)

I'm a bit confused here. Shouldn't the stall warning go off earlier?

Did the aircraft actually get airborne WITHOUT the STALL warning going off?
OR, does it mean the acf rolled for 15 sec on MLG (NLG in the air)
before lifting off - then - was airborne for 4 sec before it stalled?

During these 19 sec (15+4), the acf would have moved about 1500m!

Is the stickshaker/stall warning activated by WOW on the NLG or MLG?

XPM

[justme69]

Here are the links to the rest of the pages of the DRAFT for the PRELIMINARY report by the accident's commission (CIAIAC) that should be finalized (it's incomplete, it's only a draft) in about a week by around next saturday/monday.

I'll try to post translations of "importants parts" that may have not been discussed yet.

http://img204.imageshack.us/img204/6...pboard0sm0.png
http://img232.imageshack.us/img232/1...pboard1ja3.png
http://img222.imageshack.us/img222/7...pboard2uq5.png
http://img146.imageshack.us/img146/1...pboard3hb7.png
http://img232.imageshack.us/img232/5...pboard4kp6.png
http://img228.imageshack.us/img228/8...pboard5iq6.png
http://img518.imageshack.us/img518/5...pboard6cr2.png
http://img515.imageshack.us/img515/5...pboard7ey2.png
http://img228.imageshack.us/img228/3...pboard8gq7.png
http://img204.imageshack.us/img204/3...pboard9up9.png
http://img232.imageshack.us/img232/1...board10ij1.png
http://img204.imageshack.us/img204/5...board11fg4.png

[NigelOnDraft]

Of course they are invalid and that is just the point. By simply rotating without regard to how the aircraft responds, you are asking it to do something it simply cannot do. By gradually rotating, you will discover that it isn't ready to fly and can keep the nose down until you get enough speed. The Spanair pilots pulled the nose up too far for the actual condition of the flaps/slats. It's no surprise that it didn't fly but if they had eased the nose up and got no correct response, they might have been able to finesse it.

If you do this technique... then fine.. but at many airfields you will be off the end of the runway

All airliners I have flown are clear - it is a 3 degree/s rotate to a determined Pitch Attitude. There is no scope for allowance as to whether you forgot the Flaps / the 10T of cargo / the 15K tailwind.

In addition, and as may have happened here, the aircraft even with the wrong performance / configuration will "takeoff" into ground effect - the problems are thereafter.

NoD

[bia botal]

Ok for sure the main reason for the accident has(most likely) been discovered, what is of interest to me is the human factors that lead to the crew taking off with zero flap, because as sure as night follows day this is not going to be as clean cut as it appears, the crew skipped through a vital part of a check list, but why? I am sure all of you who have been in the industry for a while, will recall a time when a change to a everyday event lead to something being missed in a check, these guys found them-selfs, late, taxi to runway, return to stand, engineers called, pax wanting to get off, hot day, and from what i have heard from a friend of one of the crew, under extreme pressure from the company to get the flight underway despite reservation from the crew(has anyone else heard this). Do we know if they even had time to go over the MEL and interpret what happens once the ground/air CB is pulled, because from what i have read here it sure sounds like one hell of a complicated issue to me. remember people it has been proved time and again that it takes 7 holes in the swiss cheese to course an accident. Ground/air CB pulled and flaps not set are but two.
where are the rest and will we actually find out.

[justme69]

-Total weight at "first attempted" takeoff 141,862 pounds.
-Max temperature registered by RAT after the "repair": 30ºc (Before the "repair", it reached 105ºc).
-1080 liters of fuel were added after that "RAT heater repair" (RAT heater was disconnected)
-Wind measured inmediately before time of TO by the plane: 210º, 5 knots.
-14:23:19 Brakes released - TOWS alarm never heard after this point (i.e. never heard on CVR recording). Flaps sensors indicated 0º from that point on.
-14.23.28 DFDR (data flight recorder) registers acceleration.
-A few seconds earlier, engines had commenced to thrust, developing 1,4EPR and raising til 2,0EPR max during the time the airplane remained on the ground.
-14:24:06 crew calls V1
-14:24:10 crew calls Rotate
-nose gear strut signals change of state to air mode (i.e. front wheels off the ground)
-14:24:25 stick shaker activates (i.e. stall alarm)
-14:24:29 synthetic voice alarm registered in CVR ("stall, stall").
-Max elevation reached: aprox. 40 feet.
-Aprox. 1950m of runaway to become airborne (or initiate rotation, I don't have it clear myself).
-Initial elevation (attitude) subjectively noticed by witnesses: too "smooth" (i.e. NOT high enough).

Once airborne:
-slight left roll
-20º right roll
-slight left roll
-32º right roll
-Maximun pitch: 18º

Once on the ground:
-First impact on the ground: tail cone and, almost at the same time, WITH THE RIGHT WING and the capots from the right engine.
-First ground marks: 60m away from runaway, measured perpendicular to the runaway's axis, and at the 3207,5 meter mark, measured from the direction of the runaway (total runaway length in Madrid, some 4400m and is also 60m wide itself).
-It rolled on the ground for 448m until reaching the limit (of that terrain), almost in a straight manner, at an angle about 16º with the runaway.
-Lost contact with the ground (terrain drops there) and recovers contact with the ground 150m away from the peripheral road that surrounds the runaway, which is 5,5m below in elevation from the first terrain touched by the plane.
-It continued on the ground on irregular terrain until it found the creek where the main remains are found, already quite disintegrated.
-Fire started affecting some 12 hectare (120.000 square meters) of terrain with vegetation, most of it to the right bank of the creek.
-Distance from first ground marks til main remains area: aprox 1093m

Other data:
-CVR is 4 channels. Has 40minutes of recording in fair estate, including the critical accident's phase. (Seems it recorded all the way to the end of the accident, so I'll spare the gruesome details).
-QAR was found. Its data will be attempted to be recovered by manufacturer.
-Practically all parts and pieces recovered.
-Engines detached from rest of the plane during the crash.
-Reverser for engine 2 was found some 235m away from the first marks on the ground.
-It was NOT deployed (it was noticed to be locked out out of service for delayed maintenance 3 days before)
-Reverser for engine 1, was deployed.
-No faults suspected from reversers prior to the accident.
-No faults suspected from the engines prior to the accident.
-No faults suspected from tail control surfaces.
-No faults suspected from landing gear.
-ONE OF THE PISTONS THAT ACTUATE FLAPS was extended 12cm (could've been from the impact or after the fall, of course).
-Some parts have not been identified yet: central commander's instrument panel, central instrument panel, both flaps actuators, many other panels and electrical relays.

About the TOWS configuration alarm:
-It's inhibited while in flight by action of the R2-5 relay. This relay energizes from the main suspension strut "sensor" of the front wheels.
-Three other functions are enabled when the R2-5 relay is energized: RAT heater, radio rack cooling, and alternating current AC cross-tie.
-Radio cooler system and AC cross ties are redundently serviced by R2-6 and R2-8 respectively. But TOWS and RAT heater relay exclusively on signal from R2-5.
-TOWS is required, of course, to be operative for flights (i.e. MEL).

Then the discussion we already know about how Spanair required a TOWS test on the first flight of the day and everytime both pilots entered the plane for the first time (i.e. a change of pilots). But in September 1987, MD sent out a recommendation after Detroit's accident for this check to be performed prior to each flight. Spanair, probably, never got that Telex.

[ironbutt57]

Unpublished, improper rotation technique's may very invalidate all the published performance for the particular take-off, in any event the airplane would have probably "flown off", and begun to encounter difficulties when leaving ground effect..I'm sure the crew, as in DTW probably never realized what was happening to them..and I recall reading that some types of cb's that were subjected to many openings and closings during maintenance procedures, were subject to internal failure, and while appearing closed externally, internally they were actually open...any menton of difficulty engaging auto-throttles appear on the FDR as in the DTW accident?

[snowfalcon2]

Apologies if I've overlooked, but I did not find any reference to the slats position in the draft report. Anyone?



ps. justme69 thanks for info.

[borghha]

Indeed, perhaps Spanait never saw the telex with the recommendation, because they weren't operating at the time. But CAIAC rubberstamped their SOP's and apparently didn't notice that Spanair had omitted the pre flight TOWS check...

Lot s of holes lining up here.

Was there a DFDR track of flap retract after the RTG?

[Rainboe]

Some of the responses here are getting a bit amateurish. Some of you should stop arguing technique with NoD. It is the only way to fly a jet. You rotate at a set rate to a target attitude and let the plane do the rest. You do not treat it like a Cessna on a long grass runway! Some of the questions are evidently 'I can't be bothered to read up to now'!

[Smilin_Ed]

Fly The Airplane, Not The Chart

Sorry NoD, the performance charts have gone out the window but at least the aircraft is off the ground.

Exactly. If you use a chart for flaps/slats extended and then takeoff with flaps/slats retracted, your calculations are not valid and you need to fly the airplane, not the chart.

[NigelOnDraft]

scanafrica / Smilin Ed If you (try to) get airborne, and fail, then of course there are measures that you or I might take in a desperate attempt to stay alive

However, that was not Smilin Ed's original point which was <<Why not fly it off?>> i.e. on every takeoff.

Your 500m remaining idea might be great... err except on those days when you are getting airborne within 500m anyway with a correct rotation Or close in obstacles... or...

We have millions of successful takeoffs each and everyday using, in general, the correct rotation technique. Why endanger each takeoff to "cater" for the 1 in a few million times someone forgets something pretty fundamental?

Might I also suggest a read of the NW MD-82 accident report. Without Flaps/Slats the aircraft would climb ~41' in ~5000' from Rotation - correctly configured it would climb 600'. The only solution is to configure the aircraft correctly, not just "try and fly it" with whatever Flaps/Slats you find yourself. IMHO

NoD