Altimeter 'had role' in air crash

[muriel_volestrangler]

Source: BBC

Investigators have said a faulty altimeter played a role in the plane crash near Amsterdam's Schiphol airport last week that killed nine people.

Dutch Safety Board chairman Pieter van Vollenhoven said the plane was landing on automatic pilot and the problem with the altimeter led to a loss of speed.

He said the aeroplane had twice before reported problems with its altimeter.
...
At a news conference in The Hague, Mr Van Vollenhoven said the plane had been at an altitude of 595m (1950ft) when making its landing approach to Schiphol airport, although the altimeter recorded an altitude of around ground level.
...
The automatic throttle controlling the two engines was closed and they powered down. This led to the plane losing speed, and stalling.

Read more: http://news.bbc.co.uk/1/hi/world/europe/7923782.stm

[tangent90]

1. I don't believe this.

Altimeters are and have always been just about the most reliable instrument in an aircraft.
Also pilots can tell the difference between an altitude of zero and 1900 feet especially in broad daylight and clear weather.
(implied duh)

[bdab1973]

7. I've seen them break before, and/or be off by a significant amount of altitude..

[tangent90]

8. Do you mean radar altimeters? I have never seen a Kollsman with that kind of malfunction

without a pitot/static failure. I'm not rated in that kind of a/c but I think it is unlikely it has no error-correcting signal from
a barometric instrument. And again, how could the crew not notice that kind of discrepancy? It was mid-morning in VMC.

[mn9driver]

2. They are probably talking about the radar altimeter.

I've never flown this particular model, but if the airplane was set up to autoland, the radio altimeter indicating zero might have triggered the throttle reduction as a "flare" or "landing" mode. This still would require the pilots to not observe the mode change or throttle reduction, or react to it.

Again, I'm just guessing since I haven't flown a Boeing in 15 years. The Airbus starts talking to you when it thinks it's about to land, but I don't know about the new 737's.

[rickford66]

3. I agree about radar altimeter

I work on flight simulators and have worked on most commercial and private aircraft sims. The radar altimeter is one of the main inputs to the flight computers and the signal becomes valid around 2000 ft. Some invalid situations are detected, but not all.

[remedy1]

4. There is more than one radar altimeter.

This would have to be multiple failures.

Something is pretty strange here. If both of the RA's failed, it should have been alerted to the crew by a "no autoland" EICAS (or Autoland status) message. The aircraft will not autoland with both RA's inop. Additionally, the crew should have noticed the throttles retarding, and the AS decreasing. That airplane got pretty slow before impact. I can't imagine how that would have gone unnoticed.

The EGPWS would have been inop. That should have shown up as an EICAS warning also.

(The Boeings do have CAWS call outs for altitude.)

[regnaD kciN]

5. More to the point...

...if any of the altimeters controlling autoland was faulty (i.e. giving the wrong altitude), shouldn't the annunciator have been counting down the altitude as "500 feet...400 feet..." and so on. It would seem to me that the pilots should have noticed that the aircraft was nowhere near that height, which would be enough to turn off the autopilot and hand-fly the plane to touchdown.

:wtf:

[remedy1]

6. Correct.

The airplane was at the initial approach altitude of 2000 feet.

At 1950 feet the left RA apparently malfunctioned and recorded an altitude of -8 feet (the altitude it would have shown if the a/c was on the ground.) If the system recognized the incorrect RA, the annunciation (IIRC) would have been the RA readout on the Captain's primary flight display turning amber with a line thru it. (The F/O's display would have remained correct as his RA display is from another RA system.)

That (erroneous) info was apparently passed on to the engaged autopilot/autothrottle system, which retarded the throttles. Normally, this would be no big deal, as it is easily corrected by simply disconnecting/overriding the autothrottles and assuming manual control.

The crew may not have noticed the anomaly initially as the aircraft may have been starting to descend on the glideslope, and it would not have been unusual for the autothrottles to retard to maintain the selected approach airspeed during the initial descent. However, in addition to the throttles retarding, the crew received a landing gear warning, as the EGPWS (enhanced ground proximity warning system) assumed the aircraft was too close to the ground without the landing gear extended. (It apparently received input from the malfunctioning RA.) This may have distracted them momentarily.

Regardless, as the airspeed decayed below the selected approach speed, and continued to decay as they configured the aircraft for landing, the crew should have been aware of the situation and correcting for it. As the aircraft approached stall speed, the stick shaker would have activated.

Why the crew delayed correcting the deteriorating airspeed situation remains to be seen.