Continental 737 Off Runway at DEN

[ChristiaanJ]

Sorry Smilin-Ed... it's sort of a classic expression...

forget is right...
"Even if it ain't true, it's still a great story"

A bit like
"I believe you. Thousands wouldn't".

[theron]

i thought the position of the rudder was odd but...

thanks for the clarification on the photo guys, conclusion being it is unreliable and nothing should/can be taken from it.

but i did let it spark my imagination for a short time, sorry.

[PJ2]

theron;

NOTE: For the sake of discussion only:

Seems a shame to have to put that caution up there but there's a lotta spring-loaded posters, so here goes.

The horizontal stab does appear to be significantly "ND - nose down". That control surface is jackscrewed so wouldn't move with lighter forces such as an elevator or rudder might, and I agree with all comments regarding such surfaces.

The horizontal stabs won't have moved as a result of crash forces as they appear undamaged, but I think it is an open possibility that a hydraulically powered jackscrew-moved control surface, (some but not all slats, some but not all flaps but nearly all horizontal stabilizers), could change position through electrical or hydraulic line compromise during the accident sequence. Unlikely, but possible. In any case, I believe that early in the thread, it was stated that no configuration warning had sounded.

I for one think it is an interesting observation even though it will likely be discounted.

For the record but keeping this as a "discussion on accident investigation" in general, the side-photos of the subject aircraft horizontal stab have about the same amount of vertical "slot" appearing (near the leading edge) as other 737-500's in the normal takeoff config.

[airfoilmod]

PJ2 and Theron

Count me in when discussing anything re:accident investigation. A former career of mine, though not inclusive of aviation. One thing to keep in mind though is observation needs to be kept separate from conclusion until there's enough of it. Even then, the best stance to assume is active scepticism, things frequently are not as they may seem.

Once the a/c departed the runway, and probably well before, it was more than likely experiencing stresses and loads that exceeded its design. 2/3 of the way to its last resting place it hopped a road (look for the "break". in its tracks in the aerials).My best guess is that is where the Hull split in two. At rest, and without seeing it close up, The fuselage appears to hang together with common ventral longerons and aft wing box mounts. Also, Jackscrews are designed for compression, not tension loads, and may have snapped (or not) with the extreme longitudinal flex they may have experienced. That's as far as I dare take it with precious little to go on. The NTSB does exquisite forensic work, metallurgy, part number genealogy, etc., plus they have the bonus of sole possession of the evidence.

If you are interested in this, click on lomapaseo's web site and read his study of TWA 800. Don't ask me why, but forensics is fascinating stuff (to me).

AF

[PJ2]

the best stance to assume is active scepticism

Yes, it is indeed.

Again, I have taken a pretty good look and the stab appears to be in the normal position and strongly suspect it isn't a factor - but good observation, in my view.

Anything on the nosegear? The recorders? For the sake of the crew alone, I hope this has a straightforward cause because if its "wind" that "blew" the airplane (weathercocked it) off the runway (without sufficient rudder or nosewheel authority), this will be a mess for a lot of organizations including the FAA.

[airfoilmod]

Nosegear

Not based on this incident but perhaps germane is the concept of the nosewheel itself. Of the myriad reasons for nose wheels, two come to mind, ground handling visibility and passengers who don't like walking up hill while on the ground. (Or downhill). Not meant to bear more than much lighter weights than the mains, neither are they meant to steer an aircraft when aerodynamic authority exists in sufficient measure to do so. A shortcoupled a/c like the 737 puts especially heavy demand on the nosegear because its arm is shorter than a longer wheelbased a/c like the 727.

So this is to say that relative to 1404, if we are looking at nosewheel performance and crosswinds vis a vis off runway excursion, we are looking at the wrong end of the equation. Nosegear "failure" or poor x-wind performance were the results of prior shortcomings in the takeoff roll, not the cause(s) of it.

I have two biases, my admiration of well built a/c Boeing, AB, etc., and the quality of the people who fly them. This makes any conclusion potentially a stressful one, since ordinarily it's people or aluminum who create incidents.

So at the moment I still am of the opinion that the nose gear and or the xwind had not much to do with the accident, at least at its outset.

Opinion: generally, poorly informed conjecture.

(From airfoil's "dictionary of common terms")

AF

[snowfalcon2]

Not based on this incident but perhaps germane is the concept of the nosewheel itself. Of the myriad reasons for nose wheels, two come to mind, ground handling visibility and passengers who don't like walking up hill while on the ground.

Perhaps in this context also the fact that the nosewheel configuration gives the airplane positive directional stability on the ground, as the main wheels are aft of the CG, while the tailwheel configuration is directionally unstable. But this is just a sidenote.

[airfoilmod]

snowfalcon2

Absolutely. It is this directional stability ("good steering") that must be actively abandoned by the pilot when transitioning from ground vehicle to flight, actually well before rotation. A two purpose vehicle has pronounced weaknesses at transition. Some aircraft actually jettison part of the undercarriage on takeoff, since it becomes "useless" after takeoff. Modern a/c have finely tuned parameters of performance to save weight without compromising safety.

The compelling statement a passenger made about "We were flying, then landed again" could be interpreted to mean the a/c rotated and settled or derotated onto the nosegear. Once lifted, replanting the nose is problematic, as the ng may be off track and the a/c may be crabbing or scrubbing the mains (skidding), producing sideloads the nosewheel was not intended to absorb. In any case, after some (very low) ground speed had been attained, trying to unwind the roll and rely on the nose gear for directional control a second time doesn't usually have a good result, and may be why it failed (If it did, indeed). AF

[misd-agin]

airfoilmod -

A passenger stating, "we were flying, then landed again", might turn out to be a perfect example of how invalid the observations or comments of non-experts can be when it comes to accident investigation.

From the reports I've read the airplane never took off(119 kt maximum). Have any reports mentioned that rotation was started or attempted? I havn't seen that comment.

Personally, I'm a little sceptical of the 'gust of wind' theory but we'll see what developes.

[Romeo India Xray]

airfoilmod

A shortcoupled a/c like the 737 puts especially heavy demand on the nosegear because its arm is shorter than a longer wheelbased a/c like the 727.

Spot on in relation to this accident, but does not always hold true. The following is for point of interest:

Shorter wheelbase is true for most varients however can only be applied to the shorter base 3/5/6/700 varients of the 737 (ignoring the pygmy original varients). When you get up to the 400 you are close to 10' of the 727 100 wheelbase (41'4" compared to 53'3"). By the time you are in an 800 you are down to a couple of feet and by the time you are in a 900, you are past the 727-100 by over 3 feet, and chasing the 727-200 by less than 7 feet (737-900, 56'4", 727-100 53'3", 727-200 63'3").

RIX

[airfoilmod]

misd-agin

No doubt, and I try to be careful with all information. Consider, though, that this pax was an eyewitness, and the only more valuable testimony comes from an expert eyewitness. To discount the account of this person out of hand would be unwise. (Unless of course, this passenger was a journalist.)

Even though the witness may lack the language of engineering or aeronautics, his or her account is valuable.

If I sounded in any way conclusive or definitive above, I apologize, but all the actual first hand information needs to be collected. The testimony may be useful as corroborative (For instance to pin the FDR info). Or it may be useless, but it cannot be ignored (as yet).

AF

RIX Excellent point. I'm not sure my comments might not be misleading, suggesting the type is challenged in some way that other a/c are not. Such is not the case. Shortcoupling isn't in any way suggestive of an inferior a/c. Engineering and construction allowances consider the scale of a type. My point in general was to emphasize the extra sideload on a short wheelbase, I should also have said the strength of the structure compensates for greater loading.

[RatherBeFlying]

"we were flying, then landed again"

Along with the possiblity of an aborted rotation on the rwy, the off-rwy excursion had the a/c ascending rising ground before the taxiway, levelling over the taxiway and finally descending with the terrain countour.

That sector of the ground track possibly produced an illusion of a takeoff to some of those in the back.

The bump after crossing open air after the service road may also have contributed the illusion of a landing.

Add to that the alteration of time perceptions common in accidents -- "it all happened in slow motion" is a common statement from witnesses and participants.

[lomapaseo]

A passenger stating, "we were flying, then landed again", might turn out to be a perfect example of how invalid the observations or comments of non-experts can be when it comes to accident investigation.

Agree

At the speed they were going any large bump might give the illusion that they were airborne. I really don't see the significance in this event once the speed is known

[AirRabbit]

Originally Posted by BOAC

- there is another thread running on x-wind take-offs (one of many!), so please excuse the 'drift', but let's 'nail' this here and now to avoid further confusion on THIS thread.

While running along a reasonable runway surface, the a/c will wish to 'weathercock' into wind (in this case to the left) - that is how a/c are designed. With suitable steering inputs, initially nosewheel steering and then rudder, it will remain straight AND maintain the centreline without ANY need for aileron application due to rudder input and any gusts will only accentuate the yawing tendency. So, in the Denver case, we would expect a bit of right rudder on the runway, left aileron 'to taste' to stop the upwind wing lifting in the x-wind and that, sadly, is as far as we can go for this unfortunate crew.

Once it is airborne, it WILL tend to be blown downwind (with the air mass), counteracted by its 'yawed' angle into wind and should track the runway centreline again. Now it does not need the right rudder to counter the 'weathervane' force as it is no longer constrained by physical contact with the runway (now flying in a moving body of air), so it is normal to remove both the right rudder and the left aileron (known in the trade as 'crossed controls') in a co-ordinated fashion to leave the a/c in a balanced flight tracking the centreline but pointed into wind.

I am not usually this "picky," but I feel I must offer the following ... I completely agree with your comments down to the point where you say “…and should track the runway centerline again…” The first part of that sentence is absolutely correct, and where I differ with your comments begins at the “tend to” part. As you partially acknowledge, the aircraft will at first “tend to be blown downwind (with the air mass),” but the truth is that after becoming airborne, and the cross-control inputs are removed, the aircraft will continue to operate within the air mass, and, unless a further turn into the direction from which the wind is coming is made by the pilot, the aircraft will not track the runway centerline extended. It is a very rare circumstance that the weathervane response of the aircraft would be sufficient to exactly counter the air mass movement, and it is so rare that it is almost never seen in actual operations. And the only reason I say “almost never” is to give some credence to the off-chance that somewhere, some circumstance set may meet the specific requirements (including having all the stars aligned and the pilots’ having the proper facial expression and proper number of toes and fingers crossed) that would result in the aircraft actually tracking the runway centerline extended due only to the weathervane of the aircraft in response to the crosswind conditions. For some reason, this fallacy of thought has persisted despite the numerous aerodynamic demonstrations made to the contrary.

[airfoilmod]

Pickier?

I read Boac's comments differently than you, Air Rabbit, and think he is precisely correct. After breaking ground, crossed controls can be returned to an airborne set to include crabbing into the wind, neutralizing the air mass' attempt to move him downwind, and off line. In this format, the wings are level, the heading is such as to nail the a/c on centerline, extended, and thepax (and other pilots on board) won't be wondering why the a/c is still partially banked.

BOAC: "...counteracted by its.. angle into the wind"

pickier still: no one has mentioned that one other reason (besides keeping the upwind tip down) for left bank into a left xwind is to alter the vertical vector of lift so the lifting moment is constantly "pulling" the a/c upwind, thereby keeping the track centered. If in a crossing wind you want to keep the nose directed at the desired heading, you must bank. (important when wheels are on the ground) You can do this airborne, but most pilots choose to keep the wings level and instead crab. Which, after all, is however slightly, flying sideways.(Don't tell anyone).

AF

[AirRabbit]

Originally Posted by airfoilmod

Pickier?

I read Boac's comments differently than you, Air Rabbit, and think he is precisely correct. After breaking ground, crossed controls can be returned to an airborne set to include crabbing into the wind, neutralizing the air mass' attempt to move him downwind, and off line. In this format, the wings are level, the heading is such as to nail the a/c on centerline, extended, and thepax (and other pilots on board) won't be wondering why the a/c is still partially banked.

I agree that once airborne, the crossed controls can be returned to an airborne set to include crabbing into the wind – with the wings level. However, with all due respect to you and everyone else here who believes this, unless you add an additional turn into that wind, and again return the wings to level, you will, in almost every single event, not have a sufficient “crab” angle to effectively neutralize the movement of the aircraft with respect to the ground. Of course, any crab angle into the wind will partially neutralize the movement of the aircraft with respect to the ground – but the weathervane response will not, in and of itself, be sufficient to maintain a ground track equal to the runway centerline extended.

[BOAC]

AirRabbit is quite correct - I was trying to 'KISS'. I was not trying to suggest that no turn into wind was required - it is obvious it must be, since the a/c was initially pointing along the centreline while on the runway (on a good day) and using its 'yawed' angle into wind was to indicate the final result when airborne.

In real life, by the time the un-cordinated BOAC has returned his controls to neutral I generally find that even without "the proper facial expression and proper number of toes and fingers crossed" (apart from the steely blue narrowed eyes, of course) the a/c heading is such that I finish up tracking the required track like a god..................I have never quite worked out how it happens, it just seems to work? I guess it stems from the removal of the right rudder slightly ahead of the left aileron as 'AFM' is probably suggesting - but I am frightened to analyse it more. I cannot see AFM's second para, however.

Anyway - way off topic now!

[airfoilmod]

AR

I don't think weathervaning has anything to do with this. Weathervaning is an unfortunate term used to describe an unqualified pilot's attempt to let the A/C fly him (her). The Vertical Stabilizer doesn't articulate, the Rudder does. No separate turn is necessary into the wind (with bank) to establish a nose that will counteract xwind. If you relax the x-ed controls smoothly, an additional "turn" is unnecessary. Are you saying it's necessary to establish straight and level (allowing the wind to displace the a/c downwind) before then initiating a new turn into the wind followed by relaxed aileron to end up with your crab angle?

Relaxing opposite rudder allows the aircraft to turn into its established bank. The opposite Rudder was what was preventing the turn to begin with. Stop the turn when you like the crab angle. Am I missing something? Wouldn't be the first time.

AF

[Romeo India Xray]

The Vertical Stabilizer doesn't articulate, the Rudder does. No turn is necessary into the wind (with bank) to establish a nose that will counteract xwind. If you relax the x-ed controls smoothly, an additional "turn" is unnecessary.

aerofoilmod, I am with you on this. Have been doing it in every cross wind for the last 17 years - since I was training for my first licence in fact. Of the 737, it is in my opinion "conventional" in all normal and abnormal runway situations. Now the Trident, that was a little more interesting due to the offset nosewheel, but I only flew the sim of that so not a real comparison.

Ah well, off to work now Only a 2kt x-wind tonght, but just got the METAR and -16 is a bit brisk ....

RIX

[AirRabbit]

I certainly don’t wish to get into a long, drawn-out discussion on this – although it has generated into a minor one admittedly … I’m fully aware of the relationship between the Vertical Stab and the Rudder, and, actually, not only do I very much like your description of some pilots letting the A/C fly them, I also agree with it.

To answer your question about establishing straight and level before initiating a turn to establish the crab angle … No, that is not what I’m saying.

Here’s what I know …

1) Given to their own proclivities, most aircraft will definitely try to “weathervane” in response to the wind conditions encountered after takeoff (normally well within ground effect) – but each aircraft will try to turn only by an amount equal to or less than what it would take to counter the amount of relative crosswind – note, that is relative crosswind, and not the amount of crosswind. As you undoubtedly understand, the difference is the wind referenced to the aircraft. Of course, the true wind is wind relative to a fixed point on the ground, not the aircraft.
2) I also said by an amount “equal to or less than” … so … how much “less than?” It depends on several things: If I recall correctly, the slower the rotation rate; the slower the rate of climb, the greater the thrust, the higher the velocity, the greater the aircraft mass, and the lower the altitude … all contribute to the smaller amount of weathervaning. And, oppositely, the higher the rotation rate, the higher the rate of climb, the lower the thrust, the lower the velocity, the lower the aircraft mass, and the higher the altitude … all contribute to the greater amount of weathervaning.
3) This still leaves the “true wind” ... the air mass through which the aircraft is operating. The higher the aircraft climbs, the less ground friction affects and slows the wind speed. The greater the speed of the air mass, the greater the turn required to maintain a runway centerline extended.
4) In most cases, in the initial portion of a crosswind takeoff, the crosswind exerts its greatest force on an airplane's vertical tail and aft fuselage, thereby causing the airplane to want to pivot the nose into the wind, if not prevented by the pilot. However, there are exceptions to this. Some airplanes (examples include the Rockwell Sabreliner and the McDonnell Douglas DC9 series) have more side area in front of the main gear – the pivot point – than behind it and the same crosswind wound tend to pivot the tail into the wind ... again, if not prevented by the pilot. While this may still be considered “weathervaning,” it is exactly opposite of what is described in number 1 above.

After participating for 2 years in a 3.5-year program (consisting of multiple demonstrations and multiple aircraft) aimed at determining what should be the required clearances issued to pilots departing when a reasonably sturdy crosswind is evident – particularly with departing aircraft on parallel runways – here is what I’ve learned:

After departing with a crosswind, after allowing the airplane to settle into a coordinated condition, after removing all cross control previously applied, there will be a “crab angle” into the wind established. This crab angle will reduce the amount of downwind drift. But in every case, the crab angle achieved by the airplane alone was insufficient to maintain a ground track over the extended runway centerline. Therefore, departing pilots are advised to either “maintain runway heading,” or when departing from parallel runways, they are advised to “maintain runway heading until [some predetermined altitude or geographic reference] and then fly [a heading 15 – 20 degrees away from the runway heading – with the left side turning left and the right side turning right]” because it’s virtually impossible for departing aircraft to consistently fly over the ground track of the extended runway centerline.