Air France Jet Crash: Are Commercial Jets Safe Against Lightning?

[FredGarvin]

The storm causing a break up is a good idea considering what happened in the Rockaways to an A320. That happened due to rapid course corrections. I wonder what they think could happen with 100 mph shear winds hitting the tail section?

I don't know that aircraft's systems at all. I wonder what systems are routed through the tail end.

Ugh. What a way to go.

 

[russ_watters]

Especially since it's about the only other cause of no-warning loss of an aircraft in cruise.

Given that the lightning stike thing thing hasn't happened in 40+ years, any other resonable scenario must be considered. There are at least a couple of other reasons why a plane might suddenly plummet from a seemingly normal cruise:

-Fuel tank explosion, a la TWA 800
-Fire, a la ValuJet

 

[russ_watters]

The storm causing a break up is a good idea considering what happened in the Rockaways to an A320. That happened due to rapid course corrections.

...breaking off the vertical stabilizer.

I wonder what they think could happen with 100 mph shear winds hitting the tail section?

I'm also wondering if a microburst hitting a plane at cruising speed could cause enough of a vertical (negative) g-force to rip off the wings. That would have the effect of a sudden -20 degree angle of attack change.

I don't know that aircraft's systems at all. I wonder what systems are routed through the tail end.

Certainly sensors associated with rudder and elevator position at the very least. If the flight control computer doesn't see an input from one of them, could it revert to another operating mode?

 

[mheslep]

...No one has mentioned the "B" word yet.

I thought the fact that the flight avionics had the opportunity to radio home about several electrical problems made that somewhat unlikely.

 

[russ_watters]

I thought the fact that the flight avionics had the opportunity to radio home about several electrical problems made that somewhat unlikely.

Not really - the ones that have happened before haven't completely torn apart the plane, just damaged it enough to make it unflyable. Ie, Pan Am 103.

[edit] Just read up on TWA 800. That one was apparently, that catastrophic: the plane just abruptly disappeared, electronically. The flight data recorders were intact and simply stopped recording and the transponder stopped tranmitting.

For Pan Am 103, I was wrong, though: the wiki says that while the explosion only punched a 20 inch hole in the fuselage, but the secondary effects, structural damage and aerodynamic forces were enough to rapidly tear apart the plane... but not violent enough to kill the passengers. It's a little disturbing to think about, but in most such cases (and in the Shuttle Challenger!), the passengers were almost certainly killed by impact with the ground, not by the explosion/breakup of the aircraft.

[edit2] TWA 800 was particularly gruesome. Though the explosion was much bigger than Pan Am 103's, the damage was more localized and complete -and not huge, meaning the damage didn't affect the rest of the plane at all. Ironically, this led to a very similar crash scenario: the nose separated from the fuselage and fell intact and the body with the wings continued intact in a separate piece. Most of the passengers surely survived the explosion and those in the back 3/4 of the plane rode a burning but still flying piece of airplane until it pitched up enough to rip off the wings (probably only a few seconds), then fell, in flames. It must have been horrible.

 

[mgb_phys]

What about the wind shear and breakup idea? Would the tail ripping off (like that one in 2001) cause such flight control failures? Are these flight control failures certain computer failures or could they be the computer's necessary reaction to a change in aircraft controllability or loss of sensors (whether electrical or physical)?

Flight587 was an airbus A300. The pilot used full rudder in flight which you aren't supposed to do, the vertical stabilizer failed at about twice it's design load. After this Airbus fitted software to all it's fly by wire systems which stop the pilot breaking the aircraft. This is a little controversial as some traditionalist pilots claim it could stop them recovering in a very extreme situation.

From what I understand, there are two debris fields, several miles apart. And 4 minutes to crash from 35,000 feet is pretty quick, about 100 mph.

4 minutes is the time between error messages, ie between the autopilot disengaging and the assumed failure of the cabin. The time to impact is unkown

I thought the fact that the flight avionics had the opportunity to radio home about several electrical problems made that somewhat unlikely.

Not necessarily, if something ripped a hole in the body destroying major bits of the avionics+control system the system could have sent the error messages before the cabin lost pressure and the ACARS failed.

I'm also wondering if a microburst hitting a plane at cruising speed could cause enough of a vertical (negative) g-force to rip off the wings.

Microbursts don't rely stress the airframe like that. If a bunch of air the plane is sitting in suddenly accelrates downwards the airframe goes with it there is no net stress on the wings. Microbursts are only a danger when you are near the ground - where suddenly being thrown down 1000ft might be bad news if you are only 900ft up!

Wings can also take a lot of stress, 787 wing being loaded to 150% of it's maximum design load

 

[Phrak]

[edit] Just read up on TWA 800. That one was apparently, that catastrophic: the plane just abruptly disappeared, electronically. The flight data recorders were intact and simply stopped recording and the transponder stopped tranmitting.

[edit2] TWA 800 was particularly gruesome. Though the explosion was much bigger than Pan Am 103's, the damage was more localized and complete -and not huge, meaning the damage didn't affect the rest of the plane at all. Ironically, this led to a very similar crash scenario: the nose separated from the fuselage and fell intact and the body with the wings continued intact in a separate piece. Most of the passengers surely survived the explosion and those in the back 3/4 of the plane rode a burning but still flying piece of airplane until it pitched up enough to rip off the wings (probably only a few seconds), then fell, in flames. It must have been horrible.

I think the official report of TWA 800 is discountable.
The last I heard of it on public channels, it was aired on NBC with graphics supplied by the CIA involving the conclusion of a flame front chassing the plane then igniting the center tanks. The early official story evolved from plausible to ad hoc. It stinks of team spirit.

 

[Phrak]

Commerical oceanic airliners are required to have instruments in thriplets (two for over-land flights).

For instance, there are 3 ADIRUs (Air Data [and] Inertial Reference Units). There are three flight computers. If the first pair disagree, they are shut down, and the third assumes control.

The first indicated failures that, apparently, caused the autopilot program to relenquish control, were from the ADIRUs and the ISISs. These are located in instrument bays somewhere in the main airframe. Probably under the cockpit.

Concerning cabin pressure of the last transmission, in the goof lingo used in commercial aviation, I read "Cabin vertical speed" advisory to mean what you all have thought it to be: a drop in cabin pressure. Thanks for the link MGB...

But the initial failures reported occurred in the ADIRUs and the ISISs. These are equipped with redundant power sources: 120V/400Hz, 28V and 28V battery backup. They are equipped with lightning supression circuits. Can they protect against all overvoltage spikes producted by lightning? I don't see how.

 

[russ_watters]

Flight587 was an airbus A300. The pilot used full rudder in flight which you aren't supposed to do, the vertical stabilizer failed at about twice it's design load. After this Airbus fitted software to all it's fly by wire systems which stop the pilot breaking the aircraft.

I'm not suggesting the pilot caused it, what I'm suggesting is a scenario where even the flight control computer couldn't respond fast enough due to the extreme severety of the wind event. If a plane moving at 500mph hits an area where the wind is pointed downward at 150mph, geometry tells us that it now has an angle of attack of -17 degrees. An angle of attack of -17 degrees at such speeds would cause an enourmous aerodynamic stress.

4 minutes is the time between error messages, ie between the autopilot disengaging and the assumed failure of the cabin. The time to impact is unkown

Good point.

Microbursts don't rely stress the airframe like that. If a bunch of air the plane is sitting in suddenly accelrates downwards the airframe goes with it there is no net stress on the wings.

A rapid acceleration is a g-force. A microburst pushes the wings down quickly, and the wings pull the fuselage down. That's a lot of airframe stress.

Wings can also take a lot of stress, 787 wing being loaded to 150% of it's maximum design load

Define "a lot" and 150% of what?. Just about any plane other than a fighter is capable of destroying itself via aerodynamic forces. 3-4 g's (positive) is not a lot of strength compared to the potential magnitude of the aerodynamic forces.

 

[russ_watters]

I think the official report of TWA 800 is discountable.
The last I heard of it on public channels, it was aired on NBC with graphics supplied by the CIA involving the conclusion of a flame front chassing the plane then igniting the center tanks. The early official story evolved from plausible to ad hoc. It stinks of team spirit.

Conspiracy theory is not allowed here. The official report is the authority on the subject.

 

[Phrak]

Conspiracy theory is not allowed here. The official report is the authority on the subject.

Of course. People never, ever conspire. It is not out human nature. Not even when two are speaking in private. My bad.

 

[russ_watters]

Here's an interesting plane crash where a change of CG due to the pilot going to the bathroom (!), combined with wind shear caused the breakup of a small plane:

It appeared that the pilot's decision to go to the bathroom shifted the weight of the already unbalanced plane and caused it to become unstable and uncontrollable while in an area of strong turbulence. The nose of the plane slowly pitched up and then abruptly dropped just before the crash. In three seconds, the plane rose more than 100 feet with its nose down, and was then rocked by wind shear three times greater than what is defined as extreme turbulence. Within the first 12 seconds that they encountered problems, the G-force shifts rendered the crew and passengers incapacitated and unconscious and caused the breakup of the aircraft in flight.

http://aviation-safety.net/database/record.php?id=20010303-0
Relevant fresh news story:

Ten minutes later, a cascade of problems began: Automatic messages indicate the autopilot had disengaged, a key computer system switched to alternative power, and controls needed to keep the plane stable had been damaged. An alarm sounded indicating the deterioration of flight systems.

Three minutes after that, more automatic messages reported the failure of systems to monitor air speed, altitude and direction. Control of the main flight computer and wing spoilers failed as well.

The last automatic message, at 11:14 p.m., signaled loss of cabin pressure and complete electrical failure — catastrophic events in a plane that was likely already plunging toward the ocean.

"This clearly looks like the story of the airplane coming apart," the airline industry official told The Associated Press. "We just don't know why it did, but that is what the investigation will show."

Now that's a long time from first failure to the likely time of break-up, so my idea of a microburst basically just snapping the plane in half is unlikely. It doesn't mean it couldn't have caused damage that led to a later break up, though.

One fear — terrorism — was dismissed Wednesday by all three countries involved in the search and recovery effort. France's defense minister and the Pentagon said there were no signs that terrorism was involved, and Jobim said "that possibility hasn't even been considered."

So there that one is, Fred.

http://news.yahoo.com/s/ap/brazil_plane

 

[russ_watters]

Interesting article about this crash and a "brutal freak turbulence" theory:

Brutal freak turbulence is the most plausible cause of the crash of Air France Flight 447. If lightning alone caused the crash questions would be asked about the design of the A330, a medium-sized long-range airliner that enjoys a high reputation with the world's airlines...

The best-known case of turbulence causing a commercial airliner crash was when a BOAC flight from Tokyo to Hong Kong went down near Mount Fuji in 1966 after encountering a storm.

All 113 passengers and 11 crew on board were killed and the subsequent inquiry found the probable cause of the disaster was that “the aircraft suddenly encountered abnormally severe turbulence which imposed a gust load considerably in excess of the design limit.”

http://www.blnz.com/news/2009/06/01/Analysis_turbulence_lightning_most_likely_r_Fr.html

From the wiki on that crash:

A U.S. Navy A-4 Skyhawk that was sent up shortly after the accident to search for the wreckage encountered extreme turbulence in the accident area. The cockpit accelerometer display registered peak acceleration values of +9 and -4 g-units, causing temporary loss of control, and leading the Navy pilot to believe his aircraft would also break-up in the turbulence. The pilot regained control and landed safely, but the aircraft was grounded for post-flight inspection by maintenance personnel. Many other aircraft that passed near Mount Fuji that day also reported moderate to severe turbulence.

http://en.wikipedia.org/wiki/BOAC_Flight_911

So that example is of exactly the type of thing I was speculating about. But it also is 40 years old.

 

[Phrak]

Here's my scenario.

I presume Airbus builds it's nose cones from carbon fiber. Why? because of the compound curvature of the surface and Airbus's dedication to reducing fuel costs from any quarter. If they don't build it from carbon, nevermind, I'm up the wrong tree.

Carbon composite has far greater bulk resistivity than aluminum. Electical damage to semiconductors was once identificed as coming from both electrostatic discharge and induced electromagnetic induction (ESD-EMI). Lighning has a great capability for induction. Changing fields on one side of a carbon fiber barrier can transmit to the other side. Simply put, carbon fiber makes a poor Faraday cage. On the other side of the nose cone resides the main avionics systems, as far as I can tell.

I surmise that lightning penatrated the nose and took out the main avionics.

 

[Borek]

Anybody knows if navy submarines are used to help localize black box? They are equipped with the state of the art hydrophones, they can be at the site in a blink (well, much faster than surface vessels) and they are not limited by the weather.

Could be it is not as easy - their hydrophons can be optimized for other frequencies and for other directions, but the idea seems plausible to me.

 

[Borek]

They say pilot slowed down - perhaps too much - before flying into the thunderstorm area.

I wonder where this information came from.

 

[Mech_Engineer]

Anybody knows if navy submarines are used to help localize black box? They are equipped with the state of the art hydrophones, they can be at the site in a blink (well, much faster than surface vessels) and they are not limited by the weather.

Could be it is not as easy - their hydrophons can be optimized for other frequencies and for other directions, but the idea seems plausible to me.

I don't hydrophones will be very useful in finding a black box under water, unless it sending out a sonic signal. What they need is a very sensitive RF scanning setup (assuming the black box is transmitting an RF signal), and possibly side-scanning sonar (although that assumes there are large pieces of the plane sitting at the bottom of the ocean, rather than small debris fields).

 

[mgb_phys]

I'm not suggesting the pilot caused it,

I meant that it wasn't the wind that blew the stabilzer off, the pilot turned it full to the sie while flying along at several hundred mph.

what I'm suggesting is a scenario where even the flight control computer couldn't respond fast enough due to the extreme severety of the wind event.

It doesn't try, the software detects very rapid movements and let's them happen - it's better to be moved off course or off altitude briefly than waste fuel and stress the airframe by fighting every gust.

If a plane moving at 500mph hits an area where the wind is pointed downward at 150mph, geometry tells us that it now has an angle of attack of -17 degrees. An angle of attack of -17 degrees at such speeds would cause an enourmous aerodynamic stress.

It's possible to have damage caused by very localised win shear, where one wing is being pushed up and one being pushed down. Small planes have been flipped over by this in say wake turbulence. But generally the change in pressure happens on scales larger than the wing chord so the entire lifting surface is being pushed down - which is much lower stress.

The 150% wing test always struck me as a slightly silly figure. It's 150% of the intended operating maximum - of course if you really design the operating maximum to be the worst conditions it could encounter then there is no need for 150%. But if you consider the maximum to be the maximum for normal use then 150% is too small a margin for some of the situations it could get into.

 

[mgb_phys]

I don't hydrophones will be very useful in finding a black box under water, unless it sending out a sonic signal. What they need is a very sensitive RF scanning setup (assuming the black box is transmitting an RF signal), and possibly side-scanning sonar (although that assumes there are large pieces of the plane sitting at the bottom of the ocean, rather than small debris fields).

It sends out a sonar signal - the designers did consider that aircraft will fall into the sea, RF doesn't go through miles of seawater very well.
The problem is that the sea is 4km deep with a rocky bottom, at that depth sidescan will tell you nothing.

 

[mgb_phys]

I presume Airbus builds it's nose cones from carbon fiber. Why? because of the compound curvature of the surface and Airbus's dedication to reducing fuel costs from any quarter. If they don't build it from carbon, nevermind, I'm up the wrong tree.

Almost all nosecones are made from some sort of composite. They house the weather radar and putting your radar antennae inside a faraday cage has performance issues.

Changing fields on one side of a carbon fiber barrier can transmit to the other side. Simply put, carbon fiber makes a poor Faraday cage.

True but irrelevant, the individual avionics and the wiring is very well shielded.
The main issue with composite airframes and lightning is electrically bonding panels together, if there is a break in electrical continuity between a panel that is hit and an adjacent one you get arcing which leads to damage.

On the other side of the nose cone resides the main avionics systems, as far as I can tell.

The avionics racks are under and behind the cockpit as on most planes. This was the cause of a loss of a 737 a few years ago where a blocked coffee machine had been dripping coffee onto them for years unnoticed until they shorted out. The original 737 also had a problem with toilet cleaner leaking into them. They are better protected on later models.

 

[Borek]

I don't hydrophones will be very useful in finding a black box under water, unless it sending out a sonic signal.

That's what they do.

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

Edit: mgb already answered while I was googling.

 

[DaveC426913]

This was the cause of a loss of a 737 a few years ago where a blocked coffee machine had been ripping coffee onto them for years unnoticed until they shorted out.

I thought that's what regular maintenance checks were for!

 

[mgb_phys]

I thought that's what regular maintenance checks were for!

You would have thought so!

Actually I had slightly misremembered. The coffee drain took out all the power on a Quantas 747 but it made a safe landing http://catless.ncl.ac.uk/Risks/25.02.html#subj1

 

[Phrak]

It's possible to have damage caused by very localised win shear, where one wing is being pushed up and one being pushed down. Small planes have been flipped over by this in say wake turbulence. But generally the change in pressure happens on scales larger than the wing chord so the entire lifting surface is being pushed down - which is much lower stress.

You're flying an airliner at cruise speed with a 5 degree angle of attack and enter a region of air with 150 of so mph vertically downward velocity, changing your angle of attack by -17 degrees. Now you are flying with negatively loaded wings at an angle of attack of -12 degrees nearing or surpassing the negative stall CL. The lift coefficient would be around -1.0 accellerating the aircraft downward at 1.4 gees. This is roughly the behavior of a wing built on the 747A315 section.

 

[Phrak]

Almost all nosecones are made from some sort of composite. They house the weather radar and putting your radar antennae inside a faraday cage has performance issues.

Good to know. I only had common sense to go by.

True but irrelevant, the individual avionics and the wiring is very well shielded.
The main issue with composite airframes and lightning is electrically bonding panels together, if there is a break in electrical continuity between a panel that is hit and an adjacent one you get arcing which leads to damage.

'Very well shielded.' What does this mean? How well shielded and against what kind of electrical faults; static discharge, or induced current discharge? How many volts and how many volts per seconds compared to the values lighning obtains? It's relevant until this values are know, then perhaps still relevant.

The avionics racks are under and behind the cockpit as on most planes. This was the cause of a loss of a 737 a few years ago where a blocked coffee machine had been dripping coffee onto them for years unnoticed until they shorted out. The original 737 also had a problem with toilet cleaner leaking into them. They are better protected on later models.

I looked for quite some time to find where avionics 'n stuff were located in the A330 or any Airbus. I came up empty. Thanks for the useful info.

 

[DaveC426913]

I looked for quite some time to find where avionics 'n stuff were located in the A330 or any Airbus.

: backs very carefully away from Phrak, deletes all correspondence linking us :

 

[Astronuc]

Perhaps out of this event, someone will design a cushion for black boxes that will allow them to float rather than sink - something along the lines of the cushions that were used for the Rovers landing on Mars. It would seem advantageous to have black boxes in a buoyant protective cushion.

 

[russ_watters]

I meant that it wasn't the wind that blew the stabilzer off, the pilot turned it full to the sie while flying along at several hundred mph.

I know, I wasn't suggesting that either - the pilot overreacted to an attitude change due to wake turbulence.

It doesn't try, the software detects very rapid movements and let's them happen - it's better to be moved off course or off altitude briefly than waste fuel and stress the airframe by fighting every gust.

If the gust is a 100mph updraft (the media is reporting they were in the area at the time) and the computer doesn't react fast enough, the wings get ripped off.

It's possible to have damage caused by very localised win shear, where one wing is being pushed up and one being pushed down. Small planes have been flipped over by this in say wake turbulence. But generally the change in pressure happens on scales larger than the wing chord so the entire lifting surface is being pushed down - which is much lower stress.

You're still not getting what I'm saying: the wind does not need to be up on one side and down on the other to rip the wings off.

Entering a localized 100 mph up or downdraft hitting the wings causes a very rapid acceleration, whether the pilot/computer reacts to it or not. Sure, the wings can withstand such a distributed load and rapid acceleration - but they are connected to an airplane. The wings cannot accelerate the airplane up/down that fast without ripping the wings off at the root.

We aren't talking about mere rough air here, where you can look outside the airplane and watch the wings flap up and down a few feet when flying through it. This is a large, rapid, and sustained, change in the direction of the relative wind across the wing: a large change in angle of attack and large change in the lift generated.

Did you look at the plane crash I gave info about where an A-4 flew into the area later and experienced spontaneous +9 and -4 g accelerations and a loss of control due to the updrafts? The airliner that crashed was ripped apart by the effect of the updrafts alone.

It isn't all that rare that sever turbulence thows people against the ceiling of an airliner.

In any case, different news sources are talking to different experts and thus favoring different theories. USA Today's expert is theorizing about exactly what I said - an updraft tore apart the plane. The CNN expert is favoring the lightning theory.

 

[russ_watters]

Perhaps out of this event, someone will design a cushion for black boxes that will allow them to float rather than sink - something along the lines of the cushions that were used for the Rovers landing on Mars. It would seem advantageous to have black boxes in a buoyant protective cushion.

That's not enough - they also need to find their way out of the plane. There have been some articles about that in the news due to this crash.

 

[russ_watters]

CNN now says that at least one debris field isn't debris from the plane: http://www.cnn.com/2009/WORLD/americas/06/04/plane.crash/index.html#cnnSTCText

They are also saying some pilots reported seeing something consistent with a plane exploding and dropping to earth.