FADEC = complex

Greg Farris wrote:
I am not aware of accidents cause by software failure of Fadecs - perhaps
there have been - but these are certainly rare compared with mechanical
failures of linkages.

Bell Helicopter(s):

http://www.ntsb.gov/Recs/letters/2004/A04_68_69.pdf

Osprey:

http://www.defenselink.mil/news/Apr2...200104093.html

Chinook Helicopters:

Unknown if its FADEC caused actual accidents, but is a suspect.
Uncommanded engine excursions, and false failure indications in early
software revisions.

Airbus:

The most famous of all, the Airbus "low pass" at the French air show,
when the FADEC throttles refused to power up (thinking the plane was
landing) and the plane settled into the trees. Not a bug per se, but
certainly poor software planning and it resulted in changes in fly by
wire thinking.

So in general, yes the failures resulting in deaths seem to be rare.
Failures that result in pilots needing a new set of underwear are a
little less rare. IFSD (In Flight Shut Downs) happen. In one case,
the ECC software kept flopping between power supplies and shut down the
engine. The software was fixed.

There are not enough small plane FADECs out there yet to judge for GA.
Hopefully the software is better tested than, for example, the G1000
that almost messed up NW_Pilot's recent Atlantic crossing!

Kev

"Mxsmanic" wrote

Innumerable automation projects have made this pretty obvious over the
past half-century or so. Today it is generally not considered
something that must be proved, at least by people who design these
systems.

What??? There are automated systems all around you, that are functioning just
fine. It should be easy to prove, it it is such a problem. You do need to
prove it. Cites? Otherwise, utter bull.

You don't need certification to fly safely. Conversely, certification
is not a guarantee that your flight is safe.

Irrelevant to the discussion. The subject is failures of FADEC, and the modes
of failure. Certification is a distractor. If there is a problem with FADEC,
show a cite of said problem. Otherwise, utter bull.

The domain of fly-by-wire is still much more poorly understood than
more traditional systems ... so much so that certification and testing
are still much more trial and error than they are for older systems.
Thus, certification is far less useful for fly-by-wire systems; they
can still fail very catastrophically indeed.

Yu need to get it into your head that we are not talking about fly by wire. Get
it? FADEC is NOT fly by wire.

Even though that is just a distractor, I will refute your statement by saying
that fly by wire is very reliable, and very well understood. There are large
airliners flying all over the place, carring millions of people, and they don't
fall out of the sky. If fly by wire is such a huge probem, and it is so poorly
understood, certainly you can cite a NTSB case where the fly by wire caused a
crash. Most military high performance aircraft also use fly by wire, and once
they leave the test ing and development stages, they don't have a problem
either. Perhaps you can cite a case of fly by wire causing a crash in military
aircraft, post development. If not, utter bull.

Just a reminder, though, that fly by wire is not the subject. FADEC is the
subject. Go ahead, the ball is in you court.

If you can not back up your statements, and continue to argue, you waste
everyone's time.

You are a troll. But that has been established, to my satisfaction, already.
Give up. Go away.
--
Jim in NC

"Capt. Geoffrey Thorpe" The Sea Hawk at wow way d0t com wrote

But, since the primary concern is "torque greater than demand" what the
redundant software / chips typically do is shut things down when a discrepancy
is detected.

But, again, the action taken is generally to reduce power or shut down the
engine to avoid the #1 worry bead.

This is something new to me. What manufacturer has such a concern, of torque
greater than demand?

I wonder why that is such a concern? I had a truck that had a manual
transmission, and it was not unusual to stall it, while trying to get something
to move, that very much was resisting my desire to move it. g It is a case
of torque greater than demand, isn't it, or is that a case of demand greater
than torque?

So why is this a concern to the chip, or are you talking about something else?

I think I might learn something today, after all! :-)
--
Jim in NC

In article ,
says...


There are plenty of experts available, and not just in Paris. Airbus
still chooses to go its own way.



Aren't they just so misguided!
When you think of how simple it would be to correct their errors. . .

Kev,

The most famous of all, the Airbus "low pass" at the French air show,
when the FADEC throttles refused to power up (thinking the plane was
landing)


Oh? Please quote the passage from the accident report that says this.
I'd be really interested. All I know is the passage where it losely
says: The pilot actively and consciously set up the system to
circumvent all the safeguards built into it to make the plane do the
stupid unapproved show-off-maneuver (sp?) instead of preventing an
accident like this as it was designed to. And after doing that, WTF did
he expect?

--
Thomas Borchert (EDDH)

Mxsmanic,

I suppose it needs something to
distinguish itself from Boeing,


Well, if it does, neither FADEC nor FBW are it. Google "any modern jet
aircraft" for the former and "Boeing 777" for the latter.

--
Thomas Borchert (EDDH)

In article om,
says...


The most famous of all, the Airbus "low pass" at the French air show,
when the FADEC throttles refused to power up (thinking the plane was
landing) and the plane settled into the trees.


It is anything but demonstrated that the aircraft systems did anything
unexpected in this accident. The pilot tried to make this claim at the
outset, but ended up being saddled with responsibility for his show-off
manoeuver. The official result of the investigation is that it was the
pilot's fault, and the aircraft has been exonerated of any failure.



Not a bug per se, but
certainly poor software planning and it resulted in changes in fly by
wire thinking.

Much more to change thinking on matters of pilot training. the claim of "poor
software planning" is unsubstantiated.




So in general, yes the failures resulting in deaths seem to be rare.

Yeah, really.

"Jose" wrote

Then ignore him and his threads.

Absolutly not. He spreads untrue statements, and makes statements of his
version that are opposite from what a knowlegeable person just told him. He is
a troll, and deserves to be known as a troll. I will not give up my right to
post as I please, any more that the person that told me to "do not tell me who I
can talk to" willl be silenced.

Then do so correctly, lest you become guilty of the same sloppiness, PLUS
adding noise to noise.

I have not been sloppy. Everything I have posted is true, an born out by other
cites and experts.

That is not necessary. We all know what he really is.

That must not be true, or people must enjoy responding to trolls. People
continue to answer questions, and responding to ridiculous posts like whether a
Barron has an ejection seat.

When he leaves, or conducts himself properly, I will stop. I do not think he is
capable of responsible conduct.
--
Jim in NC

"Mxsmanic" wrote

They are variations on the same theme.

False. Totally different types of systems.
--
Jim in NC

Well, Jose, if that statement doesn't qualify as utter BS, I don't know
what does.

How about this one?

MX: As an autopilot moves the ailerons of an aircraft to maintain heading and
attitude,

Humm. Ailerons control altitude? Must be a delta wing plane. g

Amazing.

The thing that gets me, is that he is arguing completely about autopilots,
instead of FADEC, now. I guess he realizes that he lost that battle, so he
switched subjects. Comparing the automation of FADEC to an autopilot is not a
valid comparison. Get that , MX?

Can anyone say (with a straight face) that MX is not a troll?
--
Jim in NC