Rebooting Your Airbus (After All The Screens Go Dark)

Rebooting Your Airbus (After All The Screens Go Dark)
April 24, 2006
By Russ Niles,
Newswriter, Editor

Cures aside, pilots of Airbus A320-series airliners are getting new guidance on what to do if the screens on their electronically biased aircraft go blank. "Checklists will be streamlined so re-booting of power is quicker," an Airbus spokesman told the London Daily Mirror after Britain's Air Accidents Investigation Branch released a report on an incident aboard a British Airways A319 last October. The plane was carrying 76 passengers to Budapest from London when most of the electronic displays went blank. The crew was able to bring everything back online in 90 seconds and the passengers were blissfully unaware of the glitch.

The incident brought to light five similar instances on Airbuses. In the October incident, the plane was over southern England when the crew heard an audible "clunk." Five of six screens went out, the intercom and radio failed, the autopilot and autothrottles disengaged and most of the cockpit lights went out. The captain took over the controls and flew night VFR (fortunately it was a clear night) while he and the first officer sorted out the power failure. The flying pilot's task was further complicated by the fact that the backup analog instruments aren't lit. The AAIB has issued a series of safety recommendations but its final report isn't finished yet.

That's an inflight entertainment system -- nothing to do with flight instruments.

Normally flight instruments have to make an MTBF requirement that off-the-shelf general operating systems casn't make. They use embedded systems.

Avidyne, whose system is in light general aviation planes including the Eclipse and Adam very light jets, uses a Windows NT kernel only and builds the rest of its system. In the Eclipse it is your only interface to many systems, circuit breakers, pressurization, antiskid.

The NT Kernel is stable enough for Part 23 aircraft certification (GA) but not good enough for Part 25 (Airline).

The systems used in airliners are made by SAGEM or Rockwell Collins or other similar prime contractors, and every component is very expensive. This Airbus failure sounds more like an electrical system miswired (maybe by one of the low-cost offshore repair stations airlines are using, some of which are just dreadful), than any failure in the "computer bits."

The incident with Air Transat (low-budget, ill-maintained Canadian charter outfit) gliding into the Azores was caused in part by the pilot not thinking through why he was going through fuel. He was pumping it overboard... not realising that, he pumped the fuel from the side with no problems across to the leaking one. He displayed incredible skill AFTER that, but you're supposed to exercise your superior judgment so you don't need to show off your superior skills. (This guy was a two time loser on the judgment score, he had lost his licence once for drug smuggling). But he saved the ship and the souls on board, so all's well that ends well.

The other Canadian glider came because of Canuck political correctness, the English/French language barrier, and dumb government regulations about metric measures that required everybody involved to do weight and volume conversions inter-measurement-systems with a pencil. Institute dumb procedures, get dumb results (an immutable law that extends far beyond the tiny population of Canada that clings to our northern border).

Finally, flashlights. Wags say that pilots use flashlights as a handy container for the flat batteries in their flight bags. I've dispensed with the nine pound Mag-lite and instead carry a couple of head lights -- a few bucks at Wal-Mart, and they have white and night-vision-friendly red LEDs. They run on AAAs and last a long time. I change them at the start of the snow season, and carry fresh spares. I check the light before every afternoon or night flight. And yeah, there's a handheld GPS in there, although it's older than dirt.

d.o.l.

Criminal Number 18F

AA 587 had nothing to do with the composite construction. It had to do with the unintended consequences of improved pilot training. the pilot flying had just been to an upset training course, where he flew an Extra 300, an airplane made in germany that is stressed for +/- 12G.

He applied vigorous rudder (as one does in an Extra during takeoff, landing and maneuvering, and one normally does not do in a jet). As a result, the rudder of the airbus was stressed beyond its failure limit, of about (IIRC) 1.6G.

Why so low? Because the certification requirement for lateral loads is only 1.5G. Before this accident, not one ATP (top pilot rating) in a thousand knew this, and not one A&P/AI (top mechanic) did either. The engineers who designed the thing knew it. They took another one and chucked it into a test fixture to see where it let go -- and it let go right where the one in the mishap aircraft did.

You can find similar mishaps (in terms of overstress failure) with other aircraft. For example, a Boeing 720 or 707 got caught in a lateral jetstream off Mt. Fuji once. The ust failed the vertical stabilizer to the left, where it failed the left horizontal down and away. Stripped of the tail downforce, the plane pitched down violently enough to fail the wing outer panels down and away, and the plane tumbled down to the ground.

So why did they give the pilots upset training? Simple, there have been cases where airline pilots, many of whom have no aerobatic experience, have done the wrong thing when a plane upset. An example was the 737 crash at Aliquippa. IF it was a momentary hardover caused by a rudder actuator reversal (a controversial judgment), the plane could still have been recovered. The pilot in command at least had no aerobatic experience. Both pilots held the yokes fully back and the plane remained stalled all the way down.

Upset training would have trained those pilots to unload the aircraft (pitch to remove G loads, which would have unstalled the wings), rolled wings level ("step on the sky" was what I learned) and gradually recover the aircraft. In that case it might have made the difference, and that's why airlines pay for their pilots to go and play with an EXTRA for a few days.

As far as the whole Airbus versus Boeing thing is concerned --

1. Both companies try to build safe a/c

2. Both companies comply with the same rules

3. Airbus a/c are mostly made in England, France and Germany but often have major US content (radios, engines, big-$$ stuff).

4. Boeing a/c are mostly made in the USA buthave increasing amounts of foreign content (engines from England or Europe, parts made overseas. For instance, the fuselage of the 787 Dreamliner is made on the cheap in China; that decision idled hundreds of workers in Wichita and Seattle).

5. All manufacturers use more nonmetallic composites as more is learned about composites. Still, there's a lot of metal in any aircraft. Airbus is trying to save weight more by using exotic metals, like lithium, these days.

6. A principal advantage of composites over aluminium alloys is that composites are not subject to fatigue as nonferrous metals are. Google "Aloha 232" to see why this matters.

7. A second principal advantage is that composites are not subject to corrosion. Aluminium alloys tend to be subject to terrible corrosion -- if you recall your periodic table, it's a pretty reactive element.

8. A principal advantage of aluminum is after 100 years of building aircraft structures from it, it's a very mature technology. We know how to inspect, test and repair it well. We're still learning a lot about composites in service, and there's really no way to learn that except put them into service.

d.o.l.

Criminal Number 18F

We're still learning a lot about composites in service, and there's really no way to learn that except put them into service.

And one thing we're learning is that Airbus composites disintegrate in hydraulic fluid.

You obviously know quite a bit about this stuff, but I'm not completely ignorant either. I've got a couple of Lancairs in my hangar, and spent quite a bit of time studying composite technology, so I'm not against composites in general or anything. And I'm not some kind of "buy American" geek that's against Airbus merely because they're built in France (and other places).

One problem with composites is they're very difficult to inspect for damage. Thus the inspection on the A320 rudders was a "knock" test, where a mechanic knocks on the part and listens to what it sounds like. As far as I know, there's no automated, or objective way to do this test. Just a trained ear, and apparently the damage must be significant to show up then. I know there are methods to imbed sensors into the part and ultrasonic map it at the factory so that later damage will be detected. But I don't know how many manufacturers are doing this.

I've heard the story about the rudder reversal on AA587. The problem is there was supposed to be some type of control system to prevent this from occuring. The airplane is fly-by-wire. Did this system fail? I'm unclear on this.

One fact we can't dodge is that if the news hit the fan that there was some kind of inherent fault in Airbus aircraft it would seriously damage the airline industry, and foreign relations with Airbus countries would be a little hot too. So there is a very large motivation to pin an unknown problem on the pilot while the real problem is quietly delt with, or even swept under the rug.

Because of the recent revelations on two Airbus incidents involving delaminations and hydralic fluid, one of which stressed the vertical stabilizer attach points nearly to the same extent as AA587, I can't completely discount the idea that things aren't all they seem to be.

While my reply was addressed to your post, elements of it were aimed at some of the other posters here. Sorry; I do that rather than make thirty short posts.

And one thing we're learning is that Airbus composites disintegrate in hydraulic fluid.

That's really overstating the case rather strongly. (It's the equivalent of responding to Aloha 232 by saying "Bonded 737s come unglued and fall apart" -- an element of truth in it).

One problem with composites is they're very difficult to inspect for damage.

Right, I was getting at that when I mentioned that we're still discovering ways to inspect these materials -- and failure modes we hadn't anticipated. That's routine engineering.

The National Transportation Safety Board determines that the probable cause of this accident was the in-flight separation of the vertical stabilizer as a result of the loads beyond ultimate design that were created by the first officer’s unnecessary and excessive rudder pedal inputs. Contributing to these rudder pedal inputs were characteristics of the Airbus A300-600 rudder system design and elements of the American Airlines Advanced Aircraft Maneuvering Program.

To answer some of your specifics, you ought to look at some of the reams of .pdf on the Flight 587 Docket Website. But since you may not have time, and I'm familiar with this stuff, let me spin you up.

The airplane is not fly-by-wire. It has a very conventional rudder control system -- pushrods to bellcranks to cables running the length of the fuselage, hydraulic boosted. Absolutely bog-standard. So any pilot inputs are transmitted (within the limits of a conventional cable and boost system) to the rudder.

The DFDR does track rudder and rudder pedal position (and IIRC, also rudder-pedal strain). There was nothing anomalous there.

Finally, there was no delamination in the vertical stabilizer or rudder (while the main wreckage area was badly burnt, these parts were recovered, unharmed but for their separation, from the bay). The stab is held on by pins through six attachment lugs and three separate transverse-load lugs. All of these, or the metal where they attached, failed in overload. (I'm sure you understand the engineering facts of life here -- once one lets go it's curtains, because the remaining lugs only have say 5/6 the strength of the system that just started failing).

One thing we can do is distinguish an overload failure from a contamination-driven delamination. In the same vein that we can tell an overstress failure from a fatigue failure in an aluminium structure. The damage looks different, under a scope if not to the naked eye.

[I]f the news hit the fan that there was some kind of inherent fault in Airbus aircraft it would seriously damage the airline industry...there is a very large motivation to pin an unknown problem on the pilot

While it's true that the accident investigation profession usually looks at the pilot first and hardest, it's not prejudice.. it's profiling. Mostly, planes don't crash, pilots crash planes. And I thought that the judgment here was fair to the pilot flying, FO Sten Molin -- he didn't know that tails are only required to bear 1.5G. I mean, did you? I didn't. He did what he was trained to do. Turned out to be the wrong thing. We can best honor his memory by making sure that mistake doesn't get repeated.

And... while I hear what you're saying... on the 737 rudder hardovers, why didn't they point out that those planes were recoverable (at least two of them were -- one of which was successfully recovered, and one wasn't). If they are so concerned about keeping up appearances, why pin the blame on a design feature of every 737 flying?? But they did.

In every investigation, the parties get to say their piece. The union sends a guy, Pratt or Rolls or CFM sends an engine guy, the airline sends an operations guy, and everybody spins it his own way, not because they're liars by any means, but because they're captive of their frames of reference. Out of this, the IIC and staff have to make a report that is a coherent narrative and provide a decision memo to the full board.

The full board is political to the nines, but the investigators are not. They call 'em as they see 'em. Are they right? Usually. Are they sometimes wrong? Hey, they're human, of course they are. Just like us pilots. Mechanics. And engineers.

Also, one last thing. Not all Airbusen have the composite stab. Most 300s don't (this was a 300-600, its only the 300-600 and, ISTR, 310 that have this tail).

d.o.l.

Criminal Number 18F

[And one thing we're learning is that Airbus composites disintegrate in hydraulic fluid.] That's really overstating the case rather strongly.

It's hard to overstate the fact that a hydraulic fluid drenched rudder disintegrated in flight on March 6, 2005 on an Airbus A310-300 climbing out of Cuba. The FAA report claims that this applied nearly the same loads on the rear Vertical attach lugs as AA587 and I'm thinking it claimed there was damage to the lugs.

If the AA587 rudder was found undamaged, well then that shoots my hypothesis, but you must admit that there would be *significant* motivation to cover up any structural defects in the part.

My experience with Air Force fighters tells me that there's no way an airplane can be kept clean, and that once drenched in hydraulic fluid there's no way to determine how much damage will occur in the future, and probably no way to completely clean all traces of fluid from all voids in the part.

I'm sure existing airframes can be maintained, even if they have to replace contaminated parts. My main concern is that Boeing has learned from this for the 787, and will choose a resin system that eliminates this problem. And/or reformulate hydraulic fluid to eliminate it. It will be impossible to change the fuselage should a line leak, and repair may be impractical.

Disclaimer: Opinions posted on Free Republic are those of the individual posters and do not necessarily represent the opinion of Free Republic or its management. All materials posted herein are protected by copyright law and the exemption for fair use of copyrighted works.