737 crash at Schipol

[Archer]

On TWA800 it was (probably) a spark from a fuel level indication unit that triggered the explosion.

Ben, that is indeed the 'Polderbaan'. Somehow BBC at first used an old image on their site which didn't show this runway even though it has been in service since 2003. They have rectified that now I see. The overall consensus seems to point toward a loss of power on both engines causing a stall at low level, this is based mostly on eye witness testimony. Yesterday evening a newsteam showed a piece of footage filmed inside a research simulator re-enacting that scenario. They did point out that it was speculation but it did fit the facts. The accident investigation has started so we'll have to await the results of that before we can draw any conclusions.

[Nigel H-J]

From Chris Trotts' writing it appears that he may be correct, but not speculating yet.

Quote from BBC World News:

Witnesses on the ground described seeing the plane appear to glide through the air, having lost all propulsion, before hitting the ground and breaking into three pieces.

Nigel.

[Techy111]

Latest.....

http://news.bbc.co.uk/1/hi/world/europe/7913538.stm

Tony

[WhisperJet]

Here's the outcome. Looks like the radio altimeter read wrong and the system reduced thrust...
http://www.flightglobal.com/articles/20 ... hange.html

nick.

[Archer]

The autothrottle reduced thrust to idle when the a/c was at 1950 ft, it mistakenly thought it was in the flare mode because of the RadAlt reading of -8 ft. The autopilot reacted to this by increasing pitch to follow the glidepath, this led to a decay in airspeed leading to a stall at 450 ft. Throughout this all the crew apparently didn't notice the low thrust setting or the low speed. Sad. :-(

[TSR2]

In my limited knowledge of these things, is there not some sort of warning if the Alt and Rad Alt are significantly out of sync? Also Is the autopilot not inteligent enough to railise that it isn't in the glidepath at all?

[WhisperJet]

Throughout this all the crew apparently didn't notice the low thrust setting or the low speed.

Where the hell did he have his hands?
He must have recognized the thrust lever folding back, no?

:brick:

Moreover I ask myself once again if there's someone up there playing his satanic tricks to the aviation world? Kind of "Let's make it three stalls now, it's been a long time. All three on approach and all three in a few months timeframe. But at least let's be fair to the manufacturers, we'll make it an A320 at Perpignan, we'll make it a Dash8 at Buffalo and we'll end up with a 737 at Schiphol."

Man, gimme a break!

:-(

Nick

[forthbridge]

Throughout this all the crew apparently didn't notice the low thrust setting or the low speed.

Where the hell did he have his hands?
He must have recognized the thrust lever folding back, no?

They apparently did (Throttles had been set to 'retard') - pilot advanced them twice but did not hold them in position and they reverted to idle.
They also knew about one of the radalts being out. For whatever reason they did not take steps to disable the autothrottle or hit TOGA - but they were aware of the reducing speed.

[cstorey]

We are all still speculating. However, getting down the back of the drag curve is in fact one of the most insidious problems there is in an approach. In the 70s I used to fly an aeroplane that was notorious for this, and I used to demonstrate its habit to trainees so that they would never forget just how frighteningly undramatic it was - with a normal stable approach set up, at say 700 feet /min a 3 to 5% reduction in thrust was applied and the nose pitched up no more than 2 degrees from the stable approach attitude. This produced a reduction in IAS of no more than about 8 knots . Within about 15 seconds the aircraft was now descending at about 3000 feet/min with absolutely no indication of this unless one looked at the VSI

No, it shouldn't happen , but it is easy to understand how it can happen if there is any distraction in the cockpit

Chris

[Prop Jockey]

In my limited knowledge of these things, is there not some sort of warning if the Alt and Rad Alt are significantly out of sync? Also Is the autopilot not inteligent enough to railise that it isn't in the glidepath at all?

Hi Ben,

The Rad Alt is measuring height above terrain directly below the aircraft, whilst the altimeter is reading height or altitude above which ever datum has been set on the preesure scale - so they're doing different jobs and providing different information. Imagine an approach over high ground to an airfield in a valley - your Rad Alt reading height above the ground below, which in turn is higher than airfield your landing at, so you Altimeter set to airfield QFE will give a reading potentially quite different (higher) from the Rad Alt.

As for the autopilot - yes it is in normal circumstances when operating in Appr logic - but it seems that what may have happened is the Rad Alt reading triggered the autopilot to go into 'Flare' logic. So the autopilot will do its thing - and the plane lands. In normal circumstances, by the time you get to this point there would be no need to have a fail safe scenario where the autopilot said 'Actually, I've executed the Flare logic, you know what, speed is decaying and I don't seem to have landed yet - i'll do something about it'. In the Flare logic, position of the glide slope is no longer important, because the aircraft is transitioning from following the GS to trying to make a reasonably graceful arrival on terra-firma - and its about to go whizzing past the GS antenna anyway.

Cheers

Rich