Air Asia flight from Indonesia to Singapore vanishes

[StevieTNZ]

http://www.nzherald.co.nz/business/news/article.cfm?c_id=3&objectid=11386630

A Singaporean navy ship has found the crashed AirAsia plane's fuselage, a 30-metre-long section with a wing attached, in the Java Sea, authorities said.

Images taken by a remote-controlled vehicle from the ship showed parts of the plane's wing and words on the fuselage, Singapore defence Minister Ng Eng Hen said on his Facebook page. He said Indonesian search officials have been notified so they can begin recovery operations.

 

[StevieTNZ]

http://www.nzherald.co.nz/world/news/article.cfm?c_id=2&objectid=11389541

The AirAsia plane that crashed into the Java Sea last month climbed faster than "normal speed" and stalled, Indonesia's Transport Minister has revealed.

 

[Greg Bernhardt]

http://www.nzherald.co.nz/world/news/article.cfm?c_id=2&objectid=11389541

Can avionic experts here explain to me how a plane climbing quickly can stall? Can't they recover from the stall being some 36k ft up?

 

[russ_watters]

Can avionic experts here explain to me how a plane climbing quickly can stall?

A stall is what happens when the angle of attack (angle between the wings and the direction of flight) is too high and airflow detaches from the top surface of the wing, causing a large increase in drag and a large decrease in lift.

Climbing too quickly is a prime cause of stall. At the start of the climb, you have a lot of speed and the plane doesn't need a high angle of attack to maintain its direction of flight. The steeper the plane climbs, the more speed bleeds-off and the wider the disparity between direction of flight and angle of attack. This continues until stall if no action is taken to correct it (or, rather, an attempt is made to keep the nose-up attitude). In other words, in a steep climb (say, 30 degrees), the plane starts off fine, with the nose pointed up 30 degrees and the plane's path taking it upward at 30 degrees (zero angle of attack). The plane's direction of flight will start to level off as speed bleeds off, but the pilot has to bring the nose down himself, otherwise the plane eventually comes to be traveling forward, but with the nose still up at 30 degrees (now a 30 degree angle of attack). Failure to bring the nose down causes the stall.

Can't they recover from the stall being some 36k ft up?

They should, but recall from the Air France crash that spectacular pilot error can cause the plane to remain in a stall until it falls out of the sky. The airplane will scream at you to stop what you are doing, but it will not wrestle the controls from you (perhaps it should?). How much altitude is needed to recovery depends on the severity of course -- but I would think 10,000 feet would be enough.

Not that this was necessarily pilot error. There are other things that can cause a pitch-up and stall, including cargo shift (in turbulence?) and a stuck elevator. Both have happened before.

 

[berkeman]

What instrument failure could cause the excessively high climb rate readout? Even if they were climbing hard, that climb rate looks wrong...

 

[russ_watters]

What instrument failure could cause the excessively high climb rate readout?

Wasn't the climb rate based on ground based radar data?

Even if they were climbing hard, that climb rate looks wrong...

Well...clearly "wrong" insofar as it probably caused (or was caused by the same thing that caused) the crash.

 

[berkeman]

Wasn't the climb rate based on ground based radar data?

Oops, yes you're right Russ. The article I was reading was mostly about the black boxes, but re-reading it, it does say the climb data was from radar. Thanks.

 

[russ_watters]

FYI:
In a severe thunderstorm, an updraft alone can cause a 6,000 fpm (68 mph) climb rate:

Did you know that in order for a thunderstorm to produce hail the size of a dime, the approximate upward wind speed needed is 37 mph? golfball size hail requires an updraft of approximately 56 mph. Three inch diameter hail, just bigger than a baseball, has upward wind speeds of 100 mph. If the air is rushing upward at 50 to 100 mph, you had also better be on the lookout for possible tornadoes.

http://www.erh.noaa.gov/box/updraft.html

 

[Bystander]

Keep in mind that at 36k the flight "envelope" has become quite restricted due to air density and temperature --- doesn't take a whole lot of control surface movement to exceed the airframe capabilities.

 

[russ_watters]

Keep in mind that at 36k the flight "envelope" has become quite restricted due to air density and temperature --- doesn't take a whole lot of control surface movement to exceed the airframe capabilities.

If the avionics allow it...

So in an attempt to answer such speculation, here's the control logic for Airbus (A320 in this case) fly-by-wire:
http://www.skybrary.aero/index.php/Flight_Control_Laws

The flight mode of Normal Law provides five types of protection: pitch attitude, load factor limitations, high speed, high-AOA and bank angle. In addition, Low Speed Protection is available in certain phases of flight. Normal Law flight mode is operational from take-off and remains active until 100 feet above the ground during the landing phase. Failure of certain systems or multiple failures will result in degradation of Normal Law to Alternate Law (ALT 1 or ALT2)...

Load Factor Protection automatically limits the control inputs so that the aircraft remains within AOM "g" limitations and Pitch Attitude Protection limits the aircraft attitude to a maximum of 30° nose up or 15° nose down.

High Angle of Attack Protection, which protects against stalling and the effects of windshear has priority over all other protection functions. The protection engages when the angle of attack is between α-Prot and α-Max and limits the angle of attack commanded by the pilot's sidestick to α-Max even with full sidestick deflection.

So, indeed, when working properly, the avionics will prevent stalls and too much up or down pitch. But what if it isn't working properly? Air France 447 was an Airbus A330, which has the same anti-stall protection, yet it stalled for several minutes as it fell out of the sky...

Alternate Law 1 (ALT1) combines Normal Law lateral mode with Alternate Law pitch modes. Low Energy Protection is replaced by Low Speed Stability meaning that the aircraft no longer has automatic stall protection. At low speed, a nose down demand is introduced based on IAS (instead of AOA) and Alternate Law changes to Direct Law. In addition, an audio "STALL" warning is introduced. α-Floor protection is not available so conventional pilot stall recovery action is required.

Load Factor and Bank Angle Protections are retained. High Speed and High Angle of Attack Protections enter Alternate Law mode. Pitch Attitude Protection is lost.

ALT1 control law degradation will result from some faults in the horizontal stabilizer, a single elevator fault, loss of a yaw-damper actuator, loss of slat or flap position sensors or a single air data reference fault. Dependent upon the failure, autopilot may not be available.

In Alternate Law 2 (ALT2), Normal Law lateral mode is lost and is replaced by roll Direct Law and yaw Alternate Law. Pitch mode is in Alternate Law. Load factor protection is retained. In addition to those protections lost in ALT1 (Pitch Attitude and Low Energy Protection), Bank Angle Protection is also lost. In some failure cases, High Angle of Attack and High Speed Protections will also be lost.

As is the case with ALT1, some failure cases that result in ALT2 will also cause the autopilot to disconnecnt. ALT2 is entered when both engines flame out, with faults in two inertial or two air-data reference units, with faults to all spoilers, certain aileron faults or with a pedal transducers fault.

So under alternate flight controls laws, which occur due to certain failures, the types of pilot-induced departures from controlled flight we are discussing become possible. Worse, the "feel" (sensitivity, etc.) of the controls is different, which makes the plane even harder to fly. Air France 447 was in "Alternate Law 2" mode, climbing at 7,000 fpm and at an angle of attack of 30 degrees prior to its first stall. With the co-pilot still applying full back control input with the plane stalled, the angle of attack increased to 40 degrees as it fell out of the sky.
http://en.wikipedia.org/wiki/Air_France_Flight_447

I'm looking for examples of incidents where Airbusses overstressed their airframes due to too much elevator (what could happen at high speed in a microburst)...[edit]

Here's one: A bizarre incident in 2008 where an A330 was at cruising altitude when an "inertial reference unit" failed. Like the one on your phone and video game controller, the inertial reference unit tells the flight control computers the plane's attitude. Instead of reporting that they were flying straight and level, it started sending random, brief, but large pitch indications, which caused the flight control computer to respond with strong elevator input. The pilots were able to land safely.
http://www.flightglobal.com/blogs/unusual-attitude/2008/10/qantas-a330-upset-caused-by-in/

 

[nsaspook]

http://www.pprune.org/rumours-news/...ntact-surabaya-singapore-119.html#post8835002

 

[Greg Bernhardt]

Could the rapid accent be do to a violent updraft? Then problems with a sudden down draft. That is common with micro bursts, no?

 

[Borg]

Could the rapid accent be do to a violent updraft? Then problems with a sudden down draft. That is common with micro bursts, no?

I think that is the general assumption since the ascent was approximately 3 times its normal capabilities.

 

[Greg Bernhardt]

Irianto had more than 20,000 flying hours under his belt, having worked for another airline in Indonesia for 13 years before that, and had been a Indonesian Air Force pilot for a decade prior.

http://www.cnn.com/2015/01/29/asia/airasia-disaster/index.html

The captain sounds like a rock star pilot. Hard to imagine pilot error to such extent.

 

[Borg]

http://www.cnn.com/2015/01/29/asia/airasia-disaster/index.html

The captain sounds like a rock star pilot. Hard to imagine pilot error to such extent.

AirAsia co-pilot at controls when plane crashed

 

[Greg Bernhardt]

http://news.yahoo.com/co-pilot-controls-airasia-plane-crashed-investigator-081344469.html

He said the captain sat on the left and acted as "the monitoring pilot".

You can't tell me the captain just sat there as the plane goes down :)

But I guess there could have been too much turbulence to get to his controls

 

[Borg]

You can't tell me the captain just sat there as the plane goes down :)

But I guess there could have been too much turbulence to get to his controls

I would think that a handoff during a crisis situation might not be the easiest or smartest thing to do.

 

[russ_watters]

I think that is the general assumption since the ascent was approximately 3 times its normal capabilities.

I wouldn't go quite that far. Yes, updraft/downdraft is very possible here, but the media is giving people the wrong impression of the "normal capabilities" of a jetliner. The bottom line is that while it isn't normal, it is within the capabilities of a jetliner to do that on its own (without assistance from an updraft).

I'm still leaning toward this mostly being an Air France style pilot error.

It is annoying though how they keep releasing little bits of disjointed information.

 

[russ_watters]

You can't tell me the captain just sat there as the plane goes down :)

Again: Air France. The plane has side-stick fly-by wire controls (the pilot can't easily see what the co-pilot is doing) so the pilot may not know he's doing the opposite of what the copilot is doing. And the confused airplane averages their inputs.

The Air France captain's last words were something to the effect of "What? No!" When the copilot revealed what he'd been doing, seconds before the plane hit the water.

 

[russ_watters]

I would think that a handoff during a crisis situation might not be the easiest or smartest thing to do.

That was indeed part of the problem in Air France 447: the pilot said he had the controls, but the copilot never let go.

 

[Greg Bernhardt]

Again: Air France. The plane has side-stick fly-by wire controls (the pilot can't easily see what the co-pilot is doing) so the pilot may not know he's doing the opposite of what the copilot is doing. And the confused airplane averages their inputs.

Airbus has had 6 years to figure out a solution, but nothing has changed?

 

[russ_watters]

Airbus has had 6 years to figure out a solution, but nothing has changed?

Such things happen slooooowly, and unfortunately the control stick location was a big part of the problem that can't easily be changed. And I'm not sure if any software chances were made...I need to go back through the report sometime.

Still, the copilot's error was so spectacularly bad I would have hoped all pilots would receive a day of training on it. Not sure if they did.

[Edit]. I said "see" above, but feel is equally important. If you are struggling with mechanical controls it is easy to see and feel when the pilots are fighting each other, but in a fly-by-wire system with no feedback, it takes almost no force to move the stick and you can't feel what the other pilot is doing or how the plane is responding. I think newer ones have feedback while older ones don't.