Great Stuff Kevin, thanks for your insight. I have a couple of points
I slightly disagree with however further down in your post. :-)

Kevin Horton wrote:

You've mixed up two different accidents here. The 330 at Toulouse was a
loss of control due to the aircraft (on autopilot) going way below VMCA
with one engine at idle and the other at full take-off thrust. The sat
and watched until it was too late to recover.

Yeah I mixed those up. Thanks for keeping me honest. Those were both
IMHO over-reliance in airbus automation accidents IIRC. I saw this
frequently with new-to-airbus co-pilots who would stare at the PFD
trying too figure out why the last button push on the FCP didn't do
anything. Instead of disconnecting everything and regaining control.

After you got yourself behind the power curve, however, for whatever
reason, I'm essentially talking about old-school guys like me who were
used to flying non-FADEC machines capable of "overboost." If you got
into trouble, because you were stupid, in say the previous generation
of Boeing products: You could always push up and call for power far
in excess of limiting max GA epr or N1, N2, EGT limits. (but maybe
that's because like you say: old eng's didn't operate so close to the
surge/stall margin.) It's unlikely the engines were going to fail
like a piston or super/turbo charged engine might. Those old buckets
would warp. The blades might creep and stretch and the engines might
have to be scrapped (at say 5 mill a copy.) But you had a better
chance of clearing the trees by going all way the to the mechanical
stops (physical wire to the FCU Hydr/Mech linkage) than you do now
with a Throttle resolver / PFM/MEC/ FADEC arrangement. The airbus
test pilot may think he's called for Jesus power, but FADEC will not
let him have it.
This may have saved me a couple of times in my career flying 60's gen
aircraft overseas. You smash everything to the wall and only slightly
pull back on the engines that are "barking." (compressor stalling.)
ATC would steer you into mountains in those days in some places.
(more war stories.)


The accident you are referring to was the A320 at Mulhouse-Habsheim. The
pilot did a very low (30 ft AGL) pass with the thrust at idle. The speed
decreased til he was at full aft stick, riding on the AOA limiter just
above the stall.

I haven't flown any FBW. But we had the predecessor AOA system on
the A310 which had a A/T "alpha floor" mode (Vls) which would not
allow you to command (not select) a speed slower than 1.2 Vso. Check
pilots would scare the **** out of themselves relying on this system,
and come back and rewrite the manual! This also lead to a bunch of
documented (AWST) vertical tailslides at third world airlines where a
little turbulence knocked the A/S below alpha floor for a second.
Throttles (sometimes asymmetrically) would in about six seconds from
(flight) idle reach G/A thrust: locked into what the french call
Thrust Latch mode: meaning if you disconnected A/T's and manually
retarded them, and let go, they would re-engauge themselves (without
your permission) and smoothly place you back up to full pwr again.
(New guys never noticed the uncommanded re-power up. They would
fixate on the airplane departing altitude and start ****ing around
with the pickle switch trim: which was active!) The auto pilot would
respect redline on flaps at all costs. It would pull the airframe up
into a 90 degree body angle and then stall the machine into an airshow
tailslide just like Art Shoal used to. Even if you disc the A/P on
the pull up it's too late. The machine has insuficient down elevator
authority now to arrest the pull up(cuz nugget ran the tailplane down
and "Auhhto" overshot it the back the other direction to get even;
there was no aural stabilizer-in-motion sound in a/p trim so nobody
noticed the comming set-up!) Great French design! Hang on Grandma!
We called these man vs. machine incidents/accidents.

bushes when he was looking down on them as he descended, were actually
trees that were higher than he was. He couldn't raise the nose, as
the fly-by-wire (FBW) was already on the AOA limiter, so the only way to
climb was to get more airspeed. He slammed the thrust levers forward, and
the FADEC accelerated the engine on its normal acceleration schedule.

Turbine engines run more efficiently if they are running close to the
surge line (i.e almost ready to compressor stall). But the engine has to
come closer to the surge line to accelerate. So the closer you run to the
surge line the slower acceleration you'll have.

FAR 25.119(a) requires go-around performance to be calculated using the
thrust that is available 8 seconds after a throttle slam from idle.
Manufacturers want the engine to run as efficiently as possible, but they
don't want to take a hit on the AFM go-around performance. So, they
typically design the fuel controls to allow full go-around thrust to be
reached in just less than 8 seconds from a throttle slam from idle. I've
done tests to check the acceleration on many transport category aircraft,
and the result is usually somewhere between 7 and 8 seconds, and this is
the same no matter whether the engine has a FADEC or an "old fashioned"
hydro-mechanical fuel control unit.

Were the engines in flight idle? Were you guys pulling the ground
sensor breaker? Ground idle takes longer. Older High Bypass designs
eg: the GE CF6-80 series only take about six seconds from flight idle
to reach GA thrust IIRC, but still cannot over boost. But it's more
like twelve seconds in profile mode (slow spool up looking at FMS
parameters.) So I remembered it wrong. I think he tried to change alt
with Level Change and Profile mode engaged first, and when nothing
much happened (norm) he smashed the thrust levers to the wall and ate
wood. But Kevin, I'll concede the acceleration argument to you.
(Older designs were even slower (something like fifteen seconds to
spool up (e.g. GE CF700's aft-fans.) You could bust altitudes
descending, if you didn't lead with the throttles a couple thousand
feet before level off.


So don't blame the FADEC for the A320 accident at Mulhouse-Habsheim. It
was caused by a pilot who had way too much confidence in the low-speed
protections of the FBW.

Yep, you're right. FADEC by itself didn't put him in the trees. But
most accidents have "a chain" of factors that cause the accident. If
you can break any one of the factorial links the accident would not
happen. I submit the inability to get over-boost power was just one
of those links. Another was a FBW AOA limit that cannot be
temporarily sacrificed to clear obstacles.

Fortunately the FBW prevented him from raising
the nose, as then the aircraft would have stalled, any many people would
probably have died. As it was "only" three live were lost.

Well I have to disagree with this. We train annually now to fly below
stick shaker to escape microburst wind shear ground contact on t/o.
We will go below stall speed (bugged) momentarily in ground effect
will full power to avoid contact. We don't care about airspeed. We
only look at V/S. If we didn't do this, some dry microbusts would
kill us. Risking a stall is always better than contact with hard
objects. (remember impact g-force energy goes up exponentially with
speed) (besides: most jets don't break fast, they burble and
pre-buffet a bit first. After a positive rate is obtained and we're
still alive, then we fly on intermittent stick shaker (way higher deck
angles than FD/AP AOA limits) until about 1000 ft AGL. AOA FBW
autopilots never fly at speeds this low to escape terrain to my
knowledge. But I'd have to ask an A320 driver to be sure. The other
thing that bugs me about that machine is not being able to bust into a
45 degree bank. (Another thread for that one.)

But keep in mind that if you'd advocated these advanced techniques
twenty years ago, they would've pulled your ticket. :-(

(but you'd still be alive.) :-) Now's its req FAA training. Note:
These techniques vary widely from airline to airline and change from
Chief Kahuna to Chief Kahuna. YMMV. For the record I think FADEC is
great. Do you want it in your GA airplane?

Cheers,

pacplyer