« TTG’s Review of Jack Hanson’s Second Novel, “Forlorn Hope” | Main | Boehner should read the constitution and the law. »

21 January 2015

Comments

Feed You can follow this conversation by subscribing to the comment feed for this post.

SAC Brat

There are Airworthiness Directives and Service Bulletins for pitot masts and angle of attack vanes, which can give erroneous data. A non US or European carrier may not be required to implement them depending on their regulating group.

http://www.faa.gov/regulations_policies/airworthiness_directives/search/?q=A320&makeModel=&type=Current&sort=effectiveDate&direction=desc&filter=pitot

A problem with many pilots is that often they can't tell when the computers are lying to them, or resolve a display of bad or unreliable data. This model airplane has three flight data systems and a standby display, but you need a pilot to observe that the engine power setting is not correct for the altitude, attitude and speed. Also, typically this airplane transmits many reports automatically during the flight for trend monitoring and will also transmit for a fault or value out of parameters.

BabelFish

Pat, this will be a real puzzle, I agree. They need to find and look at the pattern of the metalurgical results of the impact. That will tell them much about the speed, attitude, forward motion (or not) of the aircraft as it impacted the water (and if it did so in one piece or had already broken up). I know that flat spins scared the bejesus out me when I was taking my flying lessons. My instructor could get in and out of them in a Skyhawks. I would have surely died.

Babak Makkinejad

That is all fine but any additional data may not help; the pilots may not have time to react quickly enough - it may not be humanly possible.

Should there be instead an automated sub-system that would take control of the aircraft from the pilots in response to sensor readings?

SAC Brat

The pilots should be ahead of the airplane so they can overide the automation. If the airdata system gets contaminated and shows a higher or lower than normal reading the computers will follow the readings if within parameters, either by reducing power through the autothrottle system or by going into stall protection mode. To train the pilots to that standard costs a lot of money and could lead to a smaller pool of pilots available.

To successfully ditch an airplane a lot of variables have to go in your favor, such as safely putting the airplane down, configuring the airplane properly (this airplane model has a ditching selection button to close some cabin exhausts), depressurizing the cabin so the doors can be opened, etc.

Mike C

Babak-

The short answer to your question is "yes, if."

How would an automated subsystem be able to react to its environment if the hardware it depended upon to sense the environment was faulty? I urge you to look into the details of the Air France Flight 447 crash.

turcopolier

All

Looks like it went into a spin or just dove for the water. He was at 38000 feet when the stall happened. I am surprised he could not have gotten out of his predicament. He had a lot of air. pl

VietnamVet

Colonel,

This crash is very similar to the Air France 447; in both cases, Airbuses flew into equatorial thunderstorms. The Air France crash sequence started when the radar was set incorrectly and it inadvertently flew into a thunderstorm. The sensors iced over and auto-pilot dropped control onto a rookie co-pilot who panicked and pulled-up on the controls and stalled the plane. In this case the pilot requested a higher altitude but was denied by air traffic control due to too many aircraft in the vicinity and it flew into the storm.

The Air-Asia crash should be watched to see if Airbus follows Boeing’s apparent lack of involvement in the MH-370 and MH-17 investigations. Is this the third case where corporate needs and global political spin deny the loved ones and the flying public knowledge of the actual causes of the deaths of all the souls on board?

I assume it is next to impossible to pull a swept wing transport plane out of a flat spin no matter the altitude. Google found this internet quote “No one is taught a flat spin recovery in a Commercial Airliner”.

scott s.

A couple things: first it would be valuable to know the vertical velocity component at the time of water impact. I have seen many comments to the effect that water will behave "like a solid" on impact, but ISTM that needs to be qualified by the impact energy in the aircraft. Second is that it would be valuable to know the forces on the flight crew during the upset. We assume the crew could execute various actions, but things like high G loading or vibration could reduce or eliminate the ability of the crew to evaluate and react to whatever inputs they were getting. Since it appears there is good data from both CVR and FDR we should have plenty of data to work with.

Walrus

Firstly, commercial jet aircraft generate their performance by cruising very close to the speed of sound (>0.9 Mach) at altitude, any sudden upset can produce catastrophic results.

Secondly, these aircraft are designed NOT to stall, ever, because the recovery of those aircraft is not guaranteed under most conditions, even the application of full power, etc.

Thirdly, my opinion, and I think now of an increasing number of pilots and operators, is that Airbus has been too smart for its own good in applying and implementing its automation philosophy.

To borrow a quote from Pprune.org:

" I think They hit an energy enhancing event (ie updraft etc). the plane reacted by trying to hold altitude, OVERSPED, then pulled up, trimmed up, then gave up and the pilots suddenly were in alternate 2 or direct law and forgot to manually trim nose down - stall.........

................. How many Airbus incidents have had confusion over what the laws were meant to be doing in the control environment? And, yes I have nearly 5000 hours flying the euro version."

To put that another way, when Airbus automation "gives up" the drivers are going to have to be way more experienced than most are, which was reflected in Airbus call for more "stick and rudder" basic training in the aftermath of the Air France Atlantic crash where the confused copilot held full back side controller all the way from 30,000 ft into the water.

---which begs the question of the lethal problem of the side controllers not being mechanically interconnected which means that one pilot can't tell what control inputs the other is applying....

If pilot incapacity to keep ahead of the automatics is identified as one of the causes of this accident, then I suggest you follow this rule: "If its not Boeing, I ain't going". "More automation" is not the answer

turcopolier

walrus

"...these aircraft are designed NOT to stall, ever, because the recovery of those aircraft is not guaranteed under most conditions, even the application of full power, etc." IMO, no technology can be expected to perform perfectly in every circumstance. pl

BabelFish

From my light plane training: A pilot is not a chauffeur. If an air traffic controller asks you to do something that is not advisable, the pilot in the left hand seat (fixed wing) is in command. He/she has the right and responsibility to say something like 'I'm not going into that weather and I am turning back if you can't steer me to a safer route'.

I agree with Walrus' entry. IMO, these unfortunate pilots were 'behind' the airplane, reacting to the latest thing it did and not anticipating it's next move. They probably did not know how to recover their aircraft from a stall, in the conditions that occurred.

The Moar You Know

"IMO, no technology can be expected to perform perfectly in every circumstance."

This is true, and sadly something which aircraft designers (especially the idiots who designed the F22 and F35) have forgotten. However, Airbus is a lot more reliant on some technologies which CAN be, shall we say, somewhat problematic. The fly-by-wire systems are great until they hit a situation that they've not been programmed for. At that point, the airplane is supposed to toss control back to the human pilot. That doesn't always happen correctly, see:

http://en.wikipedia.org/wiki/Airbus_Industrie_Flight_129

Here's two summaries of Airbus "flight law" conditions:

http://www.airbusdriver.net/airbus_fltlaws.htm

http://en.wikipedia.org/wiki/Flight_control_modes

I will direct your attention to the fact that "mechanical backup" requires hydraulic power to run, which as far as I know the Airbus commuter size jets do not have. I assume the larger ones do.

There is also the possibility that the aircraft could have lost its tail as happened in this incident. It's largely glued on (I am not at all kidding):

http://en.wikipedia.org/wiki/American_Airlines_Flight_587

Now, all that being said, I like flying Airbus aircraft, the seats are an inch wider, and when you're 6' 2" and pushing 225, every inch counts. And despite everything I just posted they are very safe. But they are very different from aircraft that some of us may have learned to fly on. My father is, even in retirement (been flying for 49 years) still training Airbus pilots and pilots of other fly-by-wire jets, and he's never expressed any reservations about them at all. Like with any other aircraft, the pilot must be vigilant and not get into bad situations like this poor bastard did. The prudent thing would have been to turn around, land at the departing airport, and get fired or disciplined. Better than getting killed.

I don't think I've ever met a single pilot who would contemplate that as a course of action, though. Very goal-oriented people. You should watch them train. Wow.

Walrus

Col. Lang,

Re stalling a commercial swept wing jet aircraft; I didn't make myself clear. One of the compromises involved in commercial aircraft design is that many cannot be recovered from a stall at all, or if they can, require tens of thousands of feet and very fast reactions and precise inputs from the pilot to succeed.They are not designed to ever be stalled.

For that reason they are fitted with stick vibrators, stick pushers, automation, etc., to absolutely prevent the aircraft from getting outside the flight envelope because once outside, there is no easy way to get back inside.

There have been multiple crashes during instruction where the automatics - stick pushers, yaw dampers etc. have been switched off by some idiot instructor, in spite of warnings in the flight manual to the contrary from the manufacturer, "just to see what would happen" with fatal results.

MS2

Below is a link to some official material that basically reinforces what Walrus is saying. The section labelled "NOSE HIGH, WINGS LEVEL" is especially interesting.

http://www.iasa.com.au/folders/Safety_Issues/others/aerodynamic_principles_of_large.html

This source advocates using a 45-60 roll to overcome a tendency of some airplanes to pitch up uncontrollably in some "departure" conditions. No model names are named. The application of power that is taught as part of light plane stall recovery is counterproductive in this scenario.

A major operational difficulty with these augmentation systems is that when things get really hairy, they either disengage themselves or require disengagement to an "alternate law", and then the pilot has to start flying a plane that handles in an unfamiliar way in a hairy situation. That said, the onus is on the pilot to understand that "alternate law" defeats stall protection and holding full aft stick (Air France 447) is no longer acceptable. Simulation of AF447 indicated that their last chance to recover expired at around 10,000 feet.

BabelFish

Yes indeed. I say again, commercial pilots are being trained to the level of air chauffeur. The Asiana Air pilots did not truly understand their airplane and did a CFIT at the San Francisco airport. The Air France copilot was acting out of concert with the command pilot. The aircraft are not designed to recover from a stall.

As sophistication has increased the airplanes are far more reliable but the 707 prototype was taken for a barrel roll, at low altitude. I believe the wings would tear off a current generation airliner if the same maneuver was attempted.

MS2

Barrel rolls if done correctly are nominally a 1g maneuver. I would be more worried about uncovering some corner case in the software than structural failure.

The comments to this entry are closed.

My Photo

August 2020

Sun Mon Tue Wed Thu Fri Sat
            1
2 3 4 5 6 7 8
9 10 11 12 13 14 15
16 17 18 19 20 21 22
23 24 25 26 27 28 29
30 31          
Blog powered by Typepad