AF447: Flight Data Recorder Recovered

Status
Not open for further replies.
O

outrun

Joined
Nov 26, 2002
Messages
1,715
Location
Texas & BWI Area
Today, I am elated. The internet pictures showed the exterior case to be in remarkable condition with Honeywell and Flight Recorder perfectly legible.

1. USJFCOM & Seal Team 6 resolved with the Bin Laden question

2. The supposed unrecoverable FDR from Air France has been found at the bottom of the Atlantic.

Pending:

3. The FDR data will be recoverable.

4. Cockpit Voice Recorder will be found with its data intact.

5. The crash causality is rectified, learned from, and never repeated

The aircraft likely hit the water intact as some qualified sources are indicating. The means of a controlled death under suspected "level flight conditions at impact" bothers me to comprehend.
 
Causality in the AF 447 crash is pretty straightforward: The Flight Control System over-rode pilot inputs due to failure of the pitot static sensors (that AB had already started to replace) in severe weather.

You have to understand how the engineers designed the A-330: In Normal Law, the envelope protections are designed to keep the stupid pilots from exceeding the design limits of the plane. You can't stall it, overspeed it, over bank it or overstress it. The control stick inputs are fed to 7 computers that decide how much to move each flight control surface. The pilot input is just that: an input. The computers do the deciding.

IF the computers are fed bad data - say, from an iced-up pitot probe, then they will be placing the aircraft in what they think is a safe regime but what is in actuality a dangerous regime.

For example: pitot static sensors detect a low airspeed condition with engine thrust set at a normal level. Computers push the nose down to regain speed. If the airplane was already at normal speed (but the computers can't tell, because of the bad data from the pitot probes) then the nose down pitch will cause the airplane to overspeed, exceeding its desing limits. All the pilot stick input in the world will not over-ride the computers...

Conversely, if the computers detected an overspeed (that didn't really exist), they would pull the nose up to slow the aircraft...and even if the pilots push forward on the stick that will not over-ride the computer envelope protection...and 33,000 feet later, an airplane that has stalled itself due to bad data and remained that way despite the crew's valiant attempt to control it hits the water intact...

There is no mechanical connection between pilot and aircraft surface, except for the trim wheel. You can fix the probes so that the computers get good data...but you are never going to fix a design philosophy in which engineers have placed ultimate authority in the operation of the plane on a set of computers...

Easy to comprehend.

Hard to accept.
 
The A320, A330, and A380 series aircraft are FLY BY SUGGESTION.

I am completely serious - you make a control input, and the airplane's flight control computers will decide what the control surface response will be under normal law.

I flew the A320 for several years and it performed well, but you always hoped the flight control computers didn't degrade while in flight.

My favorite airplane is the 757. Fantastic power, right blend of automation, and great landing characteristics, i.e., stable, excellent braking, and it has a very comfortable cockpit.

757 Guy
 
Originally Posted By: Astro14
There is no mechanical connection between pilot and aircraft surface, except for the trim wheel. You can fix the probes so that the computers get good data...but you are never going to fix a design philosophy in which engineers have placed ultimate authority in the operation of the plane on a set of computers...

Easy to comprehend.

Hard to accept.


At 33,000 feet, the pilots, just like the computers, rely on the pitot sensors for speed information. If the readings were wrong, in all likelihood the pilots would want to "correct" the plane speed just like the computers would.

In addition, the pilot input is not just an input. Under normal conditions, the pilot does make the decision. For example, if the pilot wants to climb to 34,000feet, the computers obey. Only in overriding situations, in which the computers think the pilot's action will put the plane in a dangerous regime, the computers will take over and "make the decision".

It all comes down to the questions of whether humans are more prone to errors than machines. In the vast majority cases, computer overrides make sense and save lives. When bad data are fed into the system, bad decisions will be made, either by the pilots, or the computers, or both.
 
While in general I agree with what Astro has posted, I'd also note that in this case the theory is based on the limited information that was sent via the ACARS system. That info suggests the plane did come out of Normal Law with all the safety overrrides and programing, and went into alternate law without those protections (which is what the system is programmed to do in the event of all bad data - of course, the computer has to figure out what is good and bad...)

Heres to hoping the recovered FDR and hopefully the CVR as well contain readable and useful information that will assist in determining exactly what happened.

Remarkable enough that the wreckage was located and items are being recovered in this area.

I also hope this doesn't turn into an A vs B thread (Aviation enthusiasts know what I mean...) Both makes have their pluses and minuses. Neighbor is an A330 Pilot and loves the bird. Also flew the 757 and enjoyed that one too. Different, but both good.
 
Originally Posted By: 757guy
My favorite airplane is the 757. Fantastic power, right blend of automation, and great landing characteristics, i.e., stable, excellent braking, and it has a very comfortable cockpit.


Most pilots' opinions I've come across seem to agree that the 757 is the 'performance car' of the passenger airline industry. I never did like how it looks, unproportionately long with an odd nose shape. But I've heard that it flies very well.
 
There is no override for the FCS? I am having a bit of difficulty understanding how we have given total and uninterruptable control to a computer system with no possibility of override?

Even more difficult for me to handle is (if the crash happened as you all have described) is the thoughts that were going through the pilots heads as they fought to their last breath to take control of the plane.
 
In the Hornet, you could over-ride the FCS limits by hitting the paddle...always appreciated the ability to overstress if I needed it...like a low pull-out...where the overstress will hurt less than the ground impact.

On the AB - there is no over-ride. There is an argument that the multiple redundancy of 7 computers with multiple pitot static and 3 ADIRU inputs that are continuously compared actually creates a safer situation than a pilot in direct control...I've made my philosophical stand on that point...

I think you could get the AB to go to Direct Law (out of Normal Law) by turning off one FCC and two ELACs...but I can't recall...in Direct law, there are fewer (but still some) envelope protections...and no, the pilot can't override them. Stick input goes to computers - they decide.

In most airplanes that I've flown, you could fly with a pitot static failure using known pitch and power setting. For example, a 2 degree nose up and 3,000lb per engine fuel flow on a Tomcat would get you a 300 - 325 knot cruise...now, try doing that in a thunderstorm...and Chuck Yeager himself would be hard pressed...

One instrument that we had in the Tomcat and Hornet was AOA - alpha. With a complete pitot static failure, you could fly alpha, with known power, and be just fine...in fact, in the Hornet, alpha was often preferred over KCAS for profiles like max range or max endurance...but alpha isn't displayed in most airliners, just the Flight Path Vector...and I wonder what the FPV was showing on AF 447 when the pitot-static went TU...

But if you have a complete pitot-static failure in the AB, I think that alone puts you in direct law and now, you are flying based on erroneous information...and the flight envelope protections still apply...the plane will still over-ride pilot input to enforce the envelope protections...It may simply not be possible to fly a known pitch and power in this case...
 
On Airbus AC in Direct Law the controls go from the stick to the control surfaces unmodified. Control sensitivity depends on air speed and there is no auto trim. There are no protections provided in Direct Law, however overspeed and stall aural warnings are provided.

If the flight controls degrade to Alternate Law, Direct Law automatically becomes active when the landing gear is extended if no autopilots are engaged. If an autopilot is engaged, the airplane will remain in Alternate Law until the autopilot is disconnected.

For kicks and gigs in case of complete electrical flight control signals are lost you can "fly" the AC by mechanical means. IE manual trim to provide vertical speed and rudder for lateral control, provided you have hydraulic power. If you lose electric and hydraulic power you are effectively turn into a lawn dart.
 
OK - thanks...been about 7 years since I flew the "Bus"...

I will be very interested in what the FDR on AF 447 shows in the end...
 
CVR has now also been recovered. Here's to hoping that at least one, if not both the CVR and FDR, are still readable after all this time at the ocean floor.

Amazing that they were found and recovered at all given the remoteness and depth.
 
The only plane my father loves more then the 757 is the F-14 he flew in the Navy. But out of the 757, 767, DC10, and 777, he likes the 757 the most. Don't even get me started about fly by wire and computer mitigated control surfaces. He wasn't all that thrilled when he moved to the 777 for those reasons, but made the choice as 757's no longer fly the New York route out of SAN and if he was going to commute to LA, might as well go international.
But I'm getting off topic. Hopefully they get some meaningful data off the recorder.
 
Originally Posted By: Astro14
I will be very interested in what the FDR on AF 447 shows in the end...


Well, lucky you, and about 690 million other interested people, not the least of which is a plethora of overpaid lawyers.

According to the Wall Street Journal online as of today (May 16th) they were successful in downloading ALL of the flight data recorder information, as well as the last two hours of the cockpit voice recorder information.

The FDR download apparently was more straightforward than they expected, given the two years under intense pressure and being immersed in saltwater. In fact, the BEA described it as being in "excellent condition," given the above environment.

The CVR data download required more effort, including the drying and cleaning of certain memory boards and internal computer chips in order to be able to download the data.

According to the Bureau d'Enquêtes et d'Analyses, or BEA, it will take several weeks to conduct an in-depth analysis of the retrieved data, at which point an interim report will be issued, likely sometime in the mid-summer.
 
Last edited:
looking forward to the report. A330's are used by Qantas for long domestics and shorter internationals here, and I know an A330 has had an incident over WA.
 
just fro interest sakes, how do the control systems on an A330 family plane compare to a 777? Surely the 777 also has near fly by wire controls? (educate me)
 
Both have fly by wire controls, meaning the pilots commands to an input device (sidestick, traditional yoke, etc...) are tranlated into electrical commands by computers that activate the various control surface mechanism versus having cables or other devices to put things in direct connections.

The difference between the Airbus and the Boeing is in the design philosophy in the computers. The Airbus computers take all the data they recieve and when at the edges of the planes capabilities will keep the commands to the control surfaces within limits that will not harm the plane.

The Boeing system will more or less allow full command by the pilot, so that the commands input into the system will be carried out by the control surface. The idea or arguement on this is that the pilot knows better what he needs in an emergnecy situation than the plane does, and his commands should be fully executed.

There are signifcant merits to both arguments, with neither case being completely flawless.
 
Originally Posted By: MNgopher
There are signifcant merits to both arguments, with neither case being completely flawless.


This I fully agree.
 
Quote:
The Boeing system will more or less allow full command by the pilot, so that the commands input into the system will be carried out by the control surface. The idea or arguement on this is that the pilot knows better what he needs in an emergnecy situation than the plane does, and his commands should be fully executed.

There are signifcant merits to both arguments, with neither case being completely flawless.


Neither are completely flawless, but don't you think the pilot, with more flying experience than the computer (which has only control law sofware), should make the final decision? I do and that's why I favor the Boeing FCS.

Until the computer can be programmed with highly redundant Artificial Intelligence, I favor full pilot command, real-time human intelligence.
 
Last edited:
That interim report didn't take long after all!

Apparently several Airbus executives injured themselves rushing to the cameras and microphones (I'm joking) to say that the analysis of the data thus far showed no system malfunctions on the part of the A330.

They somewhat implied the focus will be on the actions (or lack thereof) of the two pilots of the ill-fated A330.

That said (way too quickly on the part of Airbus, in my humble opinion), there exists way too much data that still begs for in-depth analysis for Airbus to rush to the media with such a statement.

Considering a criminal investigation is underway involving this mishap, Airbus will not have the final word as to the cause and the in-depth analysis will likely occur.

However, after recalling the findings of the French judicial system surrounding the Concorde crash, one can't help but wonder how much politics will influence the end result.
 
Status
Not open for further replies.
Back
Top