Spanair accident at Madrid

[one post only!]

One thing that makes me wonder is if this aircraft did stall for whatever reason did the crew at any time attempt to get full thrust?

I wonder because after looking at accident details of the 73 going into the Potomac I always wondered at the time why no-one firewalled the thrust levers. It clearly wasn't flying so why not slam the thrust levers to the wall? This has always made me wonder that if as operating crews we lose something after sitting on a modern jet for many hours.

Obviously after V1 no-one has their hands on the thrust levers and hands don't go back on them till after rotation and climbing away. I remember early in my training on Jets being told by a trainer that I was putting my hands back on the thrust levers too soon after rotation. I was told there was no need to put them on so early, it wasn't required. I remember thinking - what if I need extra thrust??? What if the thing doesn't fly? Does this breed a mindset to leave the thrust levers alone?
However, a while back my company as part of the recurrent sim check did a V1 cut at a hot and high airfield. Several knot split on V1 and Vr and unless you toga'd the thrust levers the thing really didn't want to fly. Sounds like everyone did firewall it.
Also whenever we seem to run windshear stuff, particularly soon after rotation people do toga it. Is this because we make the connection - eng fail - need thrust. Windshear - need thrust. But what happens outside of these cases where you aren't getting flashing lights and bells but the thing isn't flying??

Can't help but wonder if you find yourself in a very unusual situation as these chaps did. Not seeming right but you aren't sure as nothing is clearly wrong, realise little too late that its taking way too long to get to speed, you are eating up runway, thing not getting airborne then stalling - you are going to be stressed and maxed out. As capacity shrinks do you have the capacity to think "f*ck me - firewall the donkeys" or do you grip the control column with both hands fighting it into the air?

I am not by any means saying keep hands on after V1 as we all know exactly why we do that. This event being extremely rare it would never justify it but do you think we do possibly lose that connection between struggling to fly (at very early stage after rotation etc) and slapping some extra thrust on?
This could potentially now be 2 cases where application of extra thrust may have saved the day. Unless of course these guys did do that and it still didn't fly!?!?

Guys that have been on jets for a while, what do you think? Do you think we do lose that instinctive connection or not? Particularly when maxed out due to a VERY abnormal situation? I would personally like to think not but.....

In no way thinking this is cause or contributing factor but this accident has made me wonder. Might be thread creep (might however be relevant) but do you think that unless it falls into a nice pigeon hole failure some crews may not realise in enough time that they need extra thrust? Sorry actually not working on a Sunday seems to have got my mind wandering!!

I'm not a TRI/TRE so don't see many people in the sim, be interesting to hear from you guys what you think as you see a lot of crews doing this stuff.

[ExSp33db1rd]

Does the pilot refer only to VR to rotate or should he/she have an approximate landmark for the rotation point? (I guess Yes)

Actually, No. I'm not aware of any civil operator using physical length of runway to measure if the acceleration is normal, this would need marker boards on every runway, and take-off calculations would have to relate speed to distance for every runway used - possible but not really practical, tho' I believe the RAF did this at some of their airfields - don't quote or shoot me, don't know for sure.

The critical speed is V1, after that there is insufficient length to stop in the remaining length of concrete, and it would be very difficult to judge - if the acceleration felt sluggish - whether to accept the overrun or haul it into the air at the last minute and hope for the best.

My reading of this situation is that there was apparently plenty of runway left after rotation, which would indicate that the crew thought that they had reached VR - rotation speed - and had attemped to get airborne, otherwise they would have held it down as long as possible, or even run off straight ahead, which doesn't gel with the suggestion that they might have been airborne too slowly and stalled - why would they pull it off if there was a lot of runway left, unless they thought they were at the right VR speed ? Which then begs the question - why didn't it accelerate to V2 - if that in fact happened ?

Still a long way to go, we will just have to be patient.

Whilst writing this I missed the post of OnePostOnly, and can accept that his hypothesis has some merit, except that I still feel that they must have reached VR, and started to fly, the question then is why didn't the acceleration to V2 continue, or when leaving ground effect, with some problem that they weren't aware of, even a mistake in computing V speeds maybe, OnePostOnly's theory then took over ?

[777fly]

Xcoote,

Your ignorance of aircraft handling techniques, aerodynamics and mechanics indicates that you cannot be a commercial pilot, surely?

I have no experience of MD80 operation, but if the free air stall speed is 200kts, it will fly at speeds well below that figure while in ground effect.

For other posters, ground effect has nothing to do with air density.

[XPMorten]

Some MD80 stall numbers.
At high gross weights, there is a penalty of around 30 kias
on stall speed not having slats deployed.

XPM

[777fly]

XPMorten:

Thanks for the graph. From that I deduce that at max weight, the crew would be looking for a flap 15 climb speed of about V2+10, or about 165 kts, after rotation? The slat retracted stall speed is given as 172kts. The aircraft would certainly have been able to fly in ground effect, with a clean wing, with that small margin.

[Seat62K]

Many of these posts show the degree to which Anglophones are hampered in trying to understand a situation where so much of the available information is in another language.
"Justme69" has commented on the amount of runway used by the aircraft. A friend told me yesterday that El Pais reported that the aircraft used all the runway. This was, I think, reported to the press by the head of the autonomous government in the Canaries. My friend's Spanish is very good (much better than mine) but obviously care needs to be exercised here over the translation. Is there anyone who is bilingual able to comment on this? I don't think I've read elsewhere on this thread that the aircraft used all the runway (could have missed it, though).
P.S. Apparently, El Pais also reported that the wind direction changed shortly before the attempted take off and that the Spanair aircraft attempted its take off with a tail wind. There seems to have been a comment from flight deck crew of another aircraft (taking off, landing or moving on the ground at Barajas) about this.

[NigelOnDraft]

Seat62K (and others I might say)

A friend told me yesterday that El Pais reported that the aircraft used all the runway.

Please just look at the basic facts

The aircraft started it's takeoff run around the start of 36L. It accelerated, rotated, got airborne, landed back on the runway, veered to the right, then went off the runway, then went off the surface alongside the runway, then went down into the ravine, and once there broke up / decelerated / caused an "accident site". Some of that is conjencture / rumour / assumption, but in general seems a fair summary?

That accident site, by pretty near certainty, is approximately 80% of the way along 36L, and displaced to the right. To end up there, yet use all the runway, it must have turned round and come back from the N? Do you really believe that

I would suggest therefore the "rotate" point, whether or not intentional, was roughly halfway down 36L, and very unlikely >70%.

Re Density Altitude: Unliklely to be major factor IMHO. DA is more significant in landing, where it alters your energy management on approach, flare technique (TAS/RoD), and most significantly LDR. Whilst you could try and translate that into takeoff, in practice it is "looked after" in the Perf Calculations, one of the longest runways in Europe, a SH aircraft, twin engined, and takeoffs being critical in V speeds / IAS terms, which are not affected by DA.

NoD

[ExSp33db1rd]

Nigel on Draft - which brings me back to what I have just suggested, i.e the aircraft appears to have got airborne with plenty of runway left ? If so, it presumably must have reached the computed VR, or else why would the crew have tried to haul it into the air below VR with no reason to ?

If so, then something happened immediately it was airborne, to stop it accelerating ( if that is what happened ) or, maybe the computed VR was in error - wouldn't be the first time. I know nothing of present day FMS systems, but I would think that maybe there would be some way to make a gross error check built into the system, or are V speeds still extracted manually from the QRH and written on the back of an envelope - or similar !! which does leave room for error,but again, the VR might have been correct, but the aircraft wouldn't fly because of incorrect flap /slat configuration, as has also been suggested, and I started my comments in reply to a question as to whether or not VR was arrived at by a speed reading alone, or a physical runway position, but it wouldn't matter how the VR was confirmed if, at that point the aircraft was incorrectly configured.

Again - patience, Gentlemen.

[FlyGooseFly!]

Question: Could the ground maintainance engineers have touched a wire (by accident) that connects to the Reversers when 'Quickly' cutting off the fault??


Answer: Almost certainly not. That's not the way airplanes or mechanics work.

I wish I could be that certain or positive - those of us with longer memories could tell of non-return fuel valves having their nibs deliberately ground off to facilitate their fitting the wrong way round - resulting in a Vanguard ( Aer Lingus ? ) crash. Bad airframe repairs leading to catastrophic failures too. Just because the aeroplane industry is well regulated and inspected does not remove the "that'll do" attitude in everyone.

I'm also quite surprised by the number of self professed experts who post here who are prepared to state catagorically that this or that cannot possible happen only to be completely refuted by actual facts later on.

.

[Zypher737]

What if the crew where trying to follow an engine out procedure to avoid terrain, which is a factor in Madrid; For my company it would be a left turn normally, but what if Spanair's is a right turn?

I would say the aircraft suffered some sort of engine failure on T/O, got airborne and started right turn too early, below V2, possibly reducing the speed further towards the stall speed, ending up in a stall, with no additional engine power available resulting to the only option left being a descend.

A further contributing factor would be the unhelpful tailwind conditions.

[justme69]

I'm a native spanish speaker, living in the Canary Islands. English is a foreign language to me.

Some politicians spoke of the airplane using up all available runaway before "taking off". This is obviously way exagerated, since as we all know, the airplane was briefly airbone and crashed some (gross wild estimate) at least +200m short of the end of the runaway, possibly as much as +600m away from the end of flat terrain.

Skid marks attributed to the fly show the plane landing on the strip for a short while, then turning (hard, about 45º) right and getting off of it by the side into adjacent terrain with still significant flat terrain in front (significant but definetly way short to stop).

A more credible account of someone metioning how the security video was explained to him by some other politician that saw it, says that the last portion of strip had the ground painted in a different color at the 600m mark off the official end of the runaway, but fails to mention if that part was just visible on the video or the airplane was actually still on ground trying to become airbone when that area was reached.

Regardless, it does seem the airplane took somewhat more airstrip to become airbone than usual, thus the politicians insisting that the plane "obviously looked like it lacked 'power' to take off".

About the wind direction and intensity, pilots flying nearby reported anywhere from 2-3 knots up to 9 knots of tail wind for that flypath. I'm personally assuming 7 knots from available information and definetly 9 at most. Wind seemed low and therefore "inconsistent", changing directions a bit, increasing, decreasing, etc. Outside air temperature was said to be 30ºC, but many pilots keep insisting that temperature just above the paviment can be as high as 10º more at midday. Regardless, Madrid heat during summer midday plus airport sea level elevation seem to call for air density on the low side of things.

Of course, reports (and reporters) have said all kinds of non-sense "24/7" since the accident. I'm dissmissing some of the obvious mistakes and exagerations that are being said and trying to report in english here those informations that seem to have more than one credible source commenting on them, while producing literal translations on those terms that are ambigous (such as talk about "available power" or "airplane falling" or "moving side-to-side") and more reasonable translations when the facts are clear and do not contradict the picture (like when they say there was a problem with the heater automatic on-off switch to then say that the probe, instead of the heater, was disconnected, which is probably a reporting error when it can be inferred in other reports that it was the heater that was disconected, hence the comment on fit-to-fly under no ice formation danger).

I do find english newpaper reports, such as those from the UK press, to be much more ambigous due not so much to slightly wrong translations (they usually fit the exact same reports from the spanish press word-by-word, as it should be, since the source is the same) as to take the comments either out-of-context or out-of-sequence (like with the survivors account of the happenings). I've seen plenty of that.

Survivors speak of things such as the airplane taking-off, moving (or falling) side-by-side soon thereafter, one wing moving way down, falling and crashing against the ground. Obviously their memories of the whole event is, on their own account, not all that clear. But two of the survivors talk roughly of the same event in the sense of "airbone", "side-to-side movement", "falling" or "wing falling" and they both kind-of-lose memory of the exact impact events until they wake up amid the debri. I don't think their recalls from the point when the ground was first hit on can be much trusted and actually slightly contradicted each other subtly as written by the reporters regarding hearing noise (i.e. one news report may lead to believe that a survivor heard a bang noise after the abrupt side-to-side movement but before falling while it's more likely it was heard as the plane hit the ground).

[FlyingOfficerKite]

When determining performance on light single and multi-engined aircraft most of my mentors stated that it was unlikely you would achieve the performance stated in the manual as these figures related to a company test pilot carrying out the tests in a new aircraft in ideal conditions at known weights. Not too much of an issue, when flying privately in the UK at least, if the performance achieved does not match the performance published in the manual. This performance is factored for average aircraft and pilot performance.

Likewise jet transport aircraft are extensively tested by company test pilots carrying out the tests in a new aircraft, at known weights, with a manufacturer keen to prove to airlines that their aircraft performs as well as or better than its rivals.

When flying public transport on a daily basis pilots refer to standard performance tables, based on standard passenger weights in aircraft whose power output in taken as read. I have never known of any definitive testing of jet airliners at any other than theoretical weights. Who knows if jet engines always provide optimum power? Who knows if ALL the fleet will achieve comparable performance? Who knows if the performance figures calculated by the airline will be achieved in practice under all conditions in all aircraft? Has this EVER been proved? It would be interesting to take the actual weights of the passengers and compare them to the calculated standard weights to see if there is a significant variation.

The figures are taken as gospel, but unbeknown to most passengers the figures used to calculate performance are THEORETICAL - not ACTUAL passenger weights or PROVEN aircraft performance data. When was the last time YOUR aircraft was ramp tested to ensure it could achieve its book performance. Probably NEVER.

So a relatively old Spanair aircraft, flying out of a hot and high airfield, what performance was actually achieved?

It may be proved that the performance issue was not a determining factor in this accident, maybe only a contributory one. But next time you fly remember no one knows ABSOLUTELY what the weight of the aircraft is or can GUARANTEE that it will perform EXACTLY as stated in the aircraft manual. Safety factors built in to the calculations ensure aviation is a safe environment in which to operate and that aircraft don't fall out of the sky.

However, many aircraft have crashed in the past because of lack of this absolute knowledge of weight and performance and they will continue to do so in the future.

There are a lot of articles on the web calling for a change in the standard weight of passengers. Remember as an adult female in the UK and Europe the 'standard weight' is 70kg - That's about 11 stone. Take off hand luggage, which isn't weighed, the weight of your clothes, handbag and other items and the airlines probably assume the 'average' woman weighs around 9 stone. Similar calculations for men would result in a 'weight in your birthday suit' of 12 stone. So next time you fly and see people stuffing max size hand luggage in the overhead locker and see 'the fat family' sat next to you - or you weigh more than 12 stone as a man or 9 stone as a woman, the pilot will NOT have accounted for this excess weight.

Just do a quick quiz when your flying - look at the passengers, guess their weight and work out the excess. Makes you think?

Also check out Google to see the articles calling for a revision to these THEORETICAL weights!

Any male passenger over 12 stone and any female passenger over 9 stone and any child under the age of 14 who weighs more than 5 stone, with hand luggage is compromising air safety.

[parabellum]

Whilst I have no doubt that performance figures are initially obtained by company test pilots on new aircraft it has always been my understanding that the test figures are factored to allow for an average line pilot, as well as other margins concerning the aircraft.
A manufacturer would be leaving themselves wide open to being sued if they only published test pilot figures for a company ops manual and I doubt very much if their product liability insurers would allow them to do so.

[Joetom]

Good info Nigel o d, tks.
.
Reading the last few posts, it's pointing my mind that maybe aircraft got off the ground with ground effect and then lift was less than required.
.
My thoughts with all involved.

[west lakes]

Looking at the distances and taking previous posts at face value

One quoted take off distance for 36L is 4446m
Looking at Google Earth a rough measure shows from the southern end: -
350m of "white" coloured surface
3090m of "black" top surface ending just past the piano keys for 18R and level with the first left hand turn off (possibly high speed) heading north
1000m of light coloured surface, ending just before a perimeter road and with another left hand turn off.

The above added together equal 4440m

A quote from an Iberia Captain makes reference to the skid marks and wheel tracks, possibly from this flight, in the grass to the right, at or about the first left hand turn off

A map, posted from the press, shows the a/c veer off location as this

[HarryMann]

If so, then something happened immediately it was airborne, to stop it accelerating ( if that is what happened ) or, maybe the computed VR was in error - wouldn't be the first time.

Two things here:
Hot weather
Reversing winds reported (light and variable?)

We hear words of windshear; would prefer to hear of potential heavy thermal activity. In the above conditions its quite possible when a big bubble is kicking off upwind (in front) to have huge quantities of air sucked in, creating tailwinds of 20~30 kts at ground level - to a light aircraft that can accelerate quickly, an exciting ride, to an airliner that can't you can be taking off and losing airspeed as fast as you are accelerating (vs the ground) over the first 50~150 feet...

a third? Max weight take-off
a fourth? sightly premature rotation (based on low AUW estimate)
a fifth? Lower thrust, see below...

Outside air temperature was said to be 30ºC, but many pilots keep insisting that temperature just above the paviment can be as high as 10º more at midday.

Yes, that is just about possible but at only a metre or two above tarmac, known as a super-adiabatic layer, effectively giving a density inversion. This layer wants to be airborne but sticks to the ground especially within standing crops, and ultimately becomes so unstable that someone running through the field or a car passing by will trigger it. Sailpane pilots know all about this... and when it does go a million cubic metres of air can bubble up and has to be replaced - hence strong incoming winds to its core - so you don't want a big one going off upwind as you depart!

Runway markers?
After several accidents in the 60's and 70's I thought it had become routine to time acceleration to V1, maybe someone can explain - perhaps it's the massive increase in low-speed thrust that fan-engines introduced removed the need. But it was thought much easier to implement than runway markers at the time - it also reflected any combination of engine thrust deficiency and underestimated TOW.

Yes, I am not an airline pilot

[Seat62K]

Thanks "justme69" for your post above.
Looking at the aerial photograph of Barajas on www.elpais.com, the latter portion of the runway does appear to be of a different material/colour. Am I right in assuming that aircraft are not expected routinely to use it? If so, this might explain the report that the Spanair aircraft used "all" the runway.

Apropos earlier comments about emergency services' access to the ravine, I was in Madrid on Wednesday and live coverage on television showed emergency vehicles descending into the ravine. Couldn't say how easy it was for them to do so, though.

[NigelOnDraft]

After several accidents in the 60's and 70's I thought it had become routine to time acceleration to V1

Never heard this at all...

Please understand that on a VERY long runway, it is unlikely that a SH twin engined aircraft will ever have a "Stop V1" less than VR. Their manual might produce a "V1", but that may well be a "Go V1" or somewhere in between...

Whilst not an MD-80 pilot, and more than willing to be corrected by one, I fly A319/320/321 around Europe and rarely will a genuine "Stop V1" be produced - exceptions of course ABZ and similar

NoD

[HarryMann]

Nigel, thanks... I'll reword that then.

In the 70's I became aware (and understood the reason was some slow acceleration aciidents) that either a suggestion had been made or it had actually been introduced by some airlines - the practice of 'timing' the roll up to V1.

I imagine if it was tested, trialled or used in anger, that a pre-V1 decision point of speed Vs time would have been implemented. That is, V1 itself would not have been the timed point...

Remember also that at that time, a 3rd crew member, namely flight engineer was common, who perhaps would have been the nominee for such a task...

Remember also that at that time, airliners had low-bypass engines and did not have the phenomenal acceleration they do today - everything was a good degree more marginal I imagine.

[bsieker]

Originally Posted by FlyingOfficerKite

However, many aircraft have crashed in the past because of lack of this absolute knowledge of weight and performance and they will continue to do so in the future.

Would you mind substantiating that claim by a number of concrete examples?

The standard tables and calculations include a big safety margin, and even a bit overweight, and with a bit less-than-rated power will ensure safe operation.

If the intakes or tailpipes are not badly deformed, identical engine types will produce the same thrust at the same EPR value. With less-than perfect engines it may vary with N1, and with fuel flow, but not with EPR.


Bernd