Air France 447 - On topic only!

[Peter Kesternich]

I think I'm never going to understand how you can avoid a stall if you don't know your airspeed.

Well - EC - there are many things not all of us are meant to understand

[Peter Kesternich]

(...)
Apparently, the ONLY INSTRUMENT that failed was the airspeed indicator (well, all three of them).

They had attitude (artificial horizon), altitude, heading, engine readings, and else.

The added problem to the problem of loosing airspeed information is that not only this information disappeared from the display, but from other various systems that use it to compute their own functions: The autopilot, autothrottle and rudder limiter all stopped working, and the flight control computer reverted from normal law to abnormal alternate law, changing the way the sidestick inputs are translated into control surface commands (not that the plane will roll when you try to pitch up or something like that, the sidestick will wor as usual but the control response would be just a bit different than usual), and removing most envelope protections from the control logic (which shouldn't be such a problem since even today most airplanes are not fly-by-wire and have no envelope protection).

Now put all that in a flight at an altitude where not too much faster can be too fast and not too much slower can be too slow, at night with obscured skies (i.e. absolutely zero visual cue), add a good bunch of thunderstorm, pouring rain and turbulence, put the captain in the crew-rest area, mix with some anxiety and nervousness of the remaining two oveloaded pilots maybe not in their very best day, and I can see things going out of hand.

Well - Gabriel - this is by far the best summary of possible events for AF447 that I have seen in over a year.
My questions still is: would that be enough to really cause an airliner to crash? Okay - many things can go wrong, but with plenty of altitude, once the airpseed indicators failed they SHOULD have been able to get abck to a level and balanced flight after descending a couple thousand feet. The worst outcome in that case should have been a stopover at Dakar. Somehow, I have a feeling that there was more than just frozen pitots and blank ASIs.

[brianw999]

Easy. Just keep the angle of attack below stall.

In other words, if the airplane's trajectory is more or less horizontal and the pitch is more or less horizontal, you are not going to stall.

Total bollocks.

You don't need a high angle of attack to stall. You need AIRSPEED as well as a proper angle of attack to avoid a stall.
The BEA Trident that crashed just outside Heathrow years ago was in a "jet stall" (horizontal stab stalled). It was flying more or less in a stable trajectory and attitude which, with sufficient AIRSPEED would have made it fly but it came down like a lift with no cables attached because it didn't have airspeed.

[EconomyClass]

How about an attitude that you use a lot + a power level you use a lot + a vertical speed you use a lot.

That might have some utility.

Actually, I'm thinking the attitude might be the biggest part of the equation.....

Never mind, I forgot, always point the nose up.

That's my problem. I read something about atitude that says after a certain numbe rof degrees (15?) lift starts to drop. Also, how much more power can you apply when you're at cruising speed?

[Peter Kesternich]

(...) The BEA Trident that crashed just outside Heathrow years ago was in a "jet stall" (horizontal stab stalled). It was flying more or less in a stable trajectory and attitude which, with sufficient AIRSPEED would have made it fly but it came down like a lift with no cables attached because it didn't have airspeed.

Well, Brian, but if I recall correctly, the Trident that crashed at Staines was not flying with a more or less horizontal trajectory and a more or less horizontal pitch. The aircraft was pitched up and trying to climb. It came down because the wings stalled after the leading-edge high-lift devices were retracted prematurely and the crew did not recognize the increasing angle of attack.

I believe what Gabriel wanted to say is that to maintain altitude and point the nose at the horizon you must use a certain amount of power that will keep the aircraft travelling at a speed where it won't stall. That way, flying attitude can keep you out of the stall.

[Peter Kesternich]

(..)Also, how much more power can you apply when you're at cruising speed?

Depends on the type of engine. Rule of thumb: cruise power setting is in the area of 75-90% rpm N1.

[Evan]

That's IMHO uncertain - airspeed affects altitude computations. I wonder what such A330 shows in such case. Perhaps no altitude at all, who knows. It would be worth looking at.
They probably had GPS altitude (and then radalt), better than nothing.

Radalt doesn't do much above 2,500 ft. I think this A330-200 had digital ISIS and not the older analog back-up instruments. Still, I would think, since it is designed to provide flight data when all computer systems are offline, that is was providing them an altitude reading. AFAIK, all that is required for pressure altitude is pressure feed from a static port.

Anyway, it gets back to the old QRH procedure, which requires the PNF to look up a table of weights and flight level ranges and determine a pitch and power setting, and in previous experiences, pilots noted that this is quite challenging when the bus is shaking. The Back Up Speed System should be a mandatory feature on an aircraft that cascades back to the stone age when the pitots fail.

[Evan]

I believe what Gabriel wanted to say is that to maintain altitude and point the nose at the horizon you must use a certain amount of power that will keep the aircraft travelling at a speed where it won't stall. That way, flying attitude can keep you out of the stall.

Pointing the nose at the horizon is a good way to overspeed. Or to sink. When in doubt, the Airbus QRH calls for a healthy (not excessive) amount of pitch and power. (about 5° and CLB setting on the A320)

[Gabriel]

That's IMHO uncertain - airspeed affects altitude computations.

How's that? Not AFAIK.

They probably had GPS altitude (and then radalt), better than nothing.

RadAlt at cruise?

[Gabriel]

Total bollocks.

I beg you pardon?

You don't need a high angle of attack to stall.

Now that's "total bollocks"...

[Gabriel]

That's my problem. I read something about atitude that says after a certain numbe rof degrees (15?) lift starts to drop. Also, how much more power can you apply when you're at cruising speed?

Let's just say that the attitude and power that the plane had before the airspeed indicators failed were perfectly Ok to keep the airplane cruising straight and level.

[Evan]

Let's just say that the attitude and power that the plane had before the airspeed failed were perfectly Ok to keep the airplane cruising straight and level.

There was speculation that they were decelerating due to turbulence. If they lost airspeed data while changing airspeed, and were distracted by all the system failures, working that problem, they might have let the speed drop below stall speed for the AoA they were flying. With no autoflight and in alternate law, there's no alpha prot and no alpha max. In the confusion, they may have forgotten that too.

[Gabriel]

There was speculation that they were decelerating due to turbulence. If they lost airspeed data while changing airspeed, and were distracted by all the system failures, working that problem, they might have let the speed drop below stall speed for the AoA they were flying. With no autoflight and in alternate law, there's no alpha prot and no alpha max. In the confusion, they may have forgotten that too.

That would have been the case if they had tried to compensate the tendency to descend due to the reduced poswer with pulling up, and hence increasing the AoA. The airplane by itself won't stall if you just choke the throttles and let it go. Especially in alternate law where the plane won't try to keep 1G with the stick centered as in normal law. It will descend, though.

Oh, and there's no such a thing as a "stall speed for the AoA they were flying".
Oh#2, and stickshaker would have been operational (just a guess?)

[kris]

How's that? Not AFAIK.

I think A330 static ports aren't 100% static, the pressure needs to be corrected WRT airspeed. I wonder what happens when the airspeed isn't known.

RadAlt at cruise?

I just meant they shouldn't have hit water unexpectedly. Since they were in one piece prior to hitting the ocean, and presumably in some sort of level flight... possibly with wrong altitude information, dark night without any visual cues - but there is radalt.

[Evan]

Ugh, ok, what I mean is the speed below which the airplane cannot create enough lift to sustain its weight in 1g flight. So not stall per se, just not sufficient lift either. So they might not have noticed an increasing sink rate amidst the confusion of working system failures. And then what? perhaps they caught on and brought the nose up before adding power (or spooling up), at a low airspeed, and presto, stall.

The airplane by itself won't stall if you just choke the throttles and let it go. Especially in alternate law where the plane won't try to keep 1G with the stick centered as in normal law. It will descend, though.

And it did.