Continental 737 Off Runway at DEN

[Green-dot]

41 seconds after brake release. Had the aircraft been sitting idle for a little while between brake release and take-off power? Seems quite long, a 41 second take-off roll before the noises began and not yet at rotation speed at that time?

Another obervation from the picture of the tracks in the snow, where is the nose gear track? I see the 2 mains, abruptly terminating where the engines dig in and a smal spot approximately in the center at the start of and between the engine tracks which may have been the nose gear contacting the ground at that location but no other signs the nose gear had made contact the moment the aircraft veered off the runway. Unless the nose gear was bearing a light load, not enough to create a track, it appears as if the aircraft might have been nose high, in other words, rotation had commenced?

[thcrozier]

3 g's?

I haven't been an active pilot for 15 years, but isn't 3 g's a mighty low threshold to shut down any system in an aircraft?

[Mark in CA]

Green-Dot - perhaps the plane was in a skid and the nose wheel track is coincident with the port main gear? To me, there seems to be signs of snow being shot off to the right, as if the plane was carving like a skier through the turn.

[theron]

@ green-dot

green-dot, are these marks made by the nose gear? or maybe im seeing something that isnt really there

[jackellis]

If you wish to keep your blood pressure down do NOT read Crash Probe Focuses on Risky Maneuver - WSJ.com

Where do they find them?

IMHO this article borders on being irresponsible. On what basis does the reporter conclude "two experienced pilots ... allowed the plane to stray off the runway's centerline in clear weather, and then tried to stop suddenly and lost control as the jet barreled toward its projected takeoff point"?

[newarksmells]

Wouldm't surprise me

if they find the cause to be the aircraft striking an animal and the events going downhill from there.

For those who don't know DIA, it's closer to Kansas than Denver meaning it's not situated in the mountains, it's on rolling flatland about 25 miles northeast of Denver. When the construction started, one of the biggest problems was removing grass snakes and various squirrels, chipmunks, ferrets etc before they could even flatten the land to build the runways.

The airport was was their natural habitat and continues to be so it would come as no great shock if one of these varmints made an excursion to the runway and got caught up in the wheels. It would certainly explain the "bounce" the passengers felt prior to takeoff as well as the rattling and thumping on the CVR...

Having said that, 41 seconds does seem like a long time to be on a takeoff roll unless he held for traffic ahead of him to clear to his satisfaction and avoid any wake turbulance conditions. Anybody what type plane took off immediately before the CAL 737?

[theron]

marks in the ground

the same marks seem to fade in and out further along?

[forget]

3 g's? I haven't been an active pilot for 15 years, but isn't 3 g's a mighty low threshold to shut down any system in an aircraft?

The switch is mounted in the rear section of the aircraft, at a 45 degree incline to the longitudinal axis. The 3G threshold is therefore a combination of the aircraft’s normal and longitudinal accelerations.

45 degree incline to the longitudinal axis. Doesn't this mean the switch needs 6G vertical and horizontally to trip?

[paulthornton]

45 degree incline to the longitudinal axis. Doesn't this mean the switch needs 6G vertical and horizontally to trip?

Looking at what others have said here, I would expect that it would require around 4.25G in either axis alone to trigger the switch (4.25 sin 45 = 4.25 cos 45 = 3).
Think of it like a crosswind at 45 degrees to the runway - you'll have a headwind/crosswind component of 0.707 * windspeed.

Paul.

[overthewing]

Object near runway

Does anyone know what that object is, that's to the right of the track just as they leave the runway?

Is it an airport-related structure, or could it be part of the a/c?

[etesting2000]

Animal

Back about 1980 I saw the results of a turboprop twin HP-137, encountering about a 120 lb doe with #1 at V1. Bent prop, tweaked shaft but enough went through to tear of the main gear door and brake lines. They got it stopped and back to the gate. I'd have to say you would need something bigger than a bear to deflect a 737.

[dl3daz]

G-Forces

Looking at what others have said here, I would expect that it would require around 4.25G in either axis alone to trigger the switch (4.25 sin 45 = 4.25 cos 45 = 3).

Referring to G-forces alone is quite useless. You need one more figure, i.e. duration.
The above mentioned 3.900G for 1/1000sec. would mean a change in speed from of 38m/s. 38m/s for 1ms will have the box travel the distance of 38mm or less than two inches.

Let it drop to concrete floor from 2m altitude and parts of it will be exposed to 3000G easily.

3900G for one second would result in 38.000m/s, or 100.000ft/s.

Just my 2¢,
Falk

[vapilot2004]

The G switch on the data recorders refers to older types that had an actual recording medium. These recorders were manufactured prior to the 1990s and have since been replaced with DFDRs, which have few moving parts and data is recorded direct to memory modules.

The old FDRs and CVRs had moving parts with heads and recording medium...stylus/drum, transducer/wire and later tape/head types. A high G shock could indeed cause the mechanism to jump to another track or lose it's place which could damage recently written data.


Good job crew, getting passengers out before the inferno breached the cabin!

[ankh]

Good job all around.

Someone mentioned being glad no aircraft was on that taxiway -- would aircraft ever hold at the takeoff end location, say immediately after having landed? I'd expect aircraft would be holding at the other end.

What struck me looking at the picture was -- if the aircraft had left the runway very slightly earlier, almost as fast, it looks like it could have hit that fire station. Anyone know what the ground contours are? Looks like the fire station is on landfill, up at runway height?

[capeverde2008]

Hard Disk Facts

Bringing a bit more IT technobabble to the 'Black Box' recording scenario....having not seen the inner workings of this model or type of drive used (if in fact a HDD is used).

Following might be helpful.

Yes, a 'parked' drive platter head is able to withstand huge 'G' forces, however, in use and recording, this is a hard parameter to determine before the platter head touches the disk platter in use.

For example, the disk head in use has a 'flying speed' over the platter from up to about 15,000rpm and is 'flying' mere microns above the drive platter - based on air pressure inertia of the spinning platter. (That's why Hard Drives have a 'breathing hole' of approx. .5mm in them). That 'balances' the hinged arm of the platter read head at micronic levels based on platter spin speed.

Under probably relatively low 'G' force (in use) the platter head can contact the hard drive platter (which it's not meant to) and this causes 'thermal asperity', i.e. a heating of the magnetised platter and damages the magnetic data to a greater or lesser extent depending on how long the head is in contact with the platter. This could render the data unreadable permanently or due to the physics laws of 'sticktion' be recoverable and read on 'forensic-type' equipment. i.e. sometimes data is available after initially being unreadable.

As a useable example, if one accidentally knocks off an external HDD (built to similar specs to an a/c HDD Black-Box recorder) from a surface as low as a couple of feet even onto a carpeted surface while it is in use - it can be rendered unuseable. If the drive was off at the time....in all probability it would be unharmed. It's all to do with whether the platter head is 'parked' safely off the drive platters that it can withstand tested levels of thousands of 'G' per microsecond.

'In use' drive heads are largely untested (as far as I'm aware, after many years in the IT tech field) for 'G' resistancy other than purely and relatively small lateral or longitudinal shocks (not a combination of complex lateral/longitudinal forces combined). A pre-determined (unknown to me) level of 'G' shock will produce an electronic command for the drive head to 'park' itself in a 'locked' secure place at the side of the platter. This speed of action varies from drive to drive and is relational to temperature and air pressure (drive head 'flying speed'). It is only a matter of microseconds for this 'park' status to occur (if triggered correctly), however, it can relatively easily cause the drive to stop working/recording data under loads of a few 'G'.

Hope this is helpful and not too IT 'nerdy'

[etesting2000]

I just took a look at Google Earth and it shows that only the paved areas are flat. Everything else is drainage, ravines and rolling low hills. Older image as 34L is under construction

[Green-dot]

@ theron,

green-dot, are these marks made by the nose gear? or maybe im seeing something that isnt really there

the same marks seem to fade in and out further along?

Those marks may or may not be from the nose gear. Such marks exist adjacent to eachother as well along the track. For instance, in the second picture you posted, to the left of your marked areas just below the red encircled 2 where the tracks are about to cross taxiway WC, you can distinguish several of such marks splitting up into seperate tracks adjacent to eachother. On the other side of taxiway no such tracks at all. Maybe they are just a pattern in the frozen soil.

I would have expected the nose gear track to have stood out more pronounced as the aircraft was decelerating to some degree (putting at least some load on the nose gear) since the NTSB mentioned use of thrust reversers.

Mark in CA,
With the (aircraft) nose gear in a sideslip and being covered by the left main gear track, wouldn't that have resulted in nose gear skid marks when the aircraft crossed taxiway WC?

[TonyWilliams]

I know there are a lot of photos, but this one is a little clearer and sharper.

http://extras.mnginteractive.com/liv...RASH_GRAPH.pdf

Look at the debri/snow pattern after the wheels past through... direct 90 degree crosswind.... with peak wind reported at 36 knots 4 minutes after the accident.

As has been reported, the -500 model with winglets has a 34 knot x-wind published speed. Certainly not the cause, but more likely a mitigating circumstance than I previously thought.

[edmundronald]

3g shutdowns ...

As an electronics engineer who hasn't been active for years, I do find 3g shutdowns by design a bit strange.

It's pretty easy these days to design a device which can stand much more; in fact I'd be willing to bet that your average flash-RAM ipod packed in a few inches of polystyrene would stand a LOT more. Of course a more heat-resistant shock absorber might be preferable for airplane use.

But then, the good thing about a non-working flight recorder is one can blame the aircrew


Edmund

[Swedish Steve]

A 100,000+lb aircraft bouncing on a squirl? Perhaps.

I pushed back a B777 in the summer and we forgot to remove a rubber chock behind the mainwheel.
We didn't notice it until we passed it in the tug!