On 2023-02-22, Fons Adriaensen wrote:
> And in many cases the aircraft may very well be unstable in that axis:
> if left alone, the roll angle will slowly increase.
Actually, most modern aircraft are stable in the bank axis as well.
Part
of why they have swept wings, bent wings and wingtips and the like, is
to this regard. (Part of: most of it has to with approaching transonic
flight. But not all.)
The thing is though, and as you say below, the pilot won't feel
anything
weird when approaching a spiral. The built in stability of the
airplane
will keep everybody in their seat at 1g acceleration perpendicular to
the floor, evenas the airplane banks to something approaching 90
degrees, and loses all of its lift. Then it just falls, sideways.
When that happens, you're in what's called a "death spiral", because
it's extremely difficult to recover from the condition, and you
typically don't even know you've entered one. When you do, you as a
pilot are already in a state of spatial disorientation; you
*literally*
don't know which way is up and which down, and since the plane is by
now
basically half-way inverted, with now absolutely no lift, losing
altitude like a falling rock, you as the pilot have very little
possibility of correcting.
*Technically*, in *theory*, you often *could* recover, if you have
enough altitude, speed and sturdiness of airframe; even I have run it
through in a game. But in practice, recovery from a well developed
death
spiral is mostly beyond human ability. Especially once you lose
height,
because at low altitudes, already going nose down, you can't even
convert high air speed/energy into a corrective manoeuvre before you
hit
the terrain, and there will only be seconds to lose.
This is then why the pilot flying is supposed to only look at the
instrumentation, and why there are auditory warnings about bank angle
on
the modern jets. The Swedish commercial midsize Boeing pilot, Mentour,
on YouTube, is first rate in explaining all of this stuff.
Okay, so, finally, how would you recover from a well developed death
spiral, presuming you realized you were in one? Well, the optimum way
would be to use all of the airfoils at the pilot's control at the same
time to convert kinetic and potential energy of the frame into first
1)
orientation, and then 2) into safe height in level flight.
The optimum control trajectory going there is universally wild, so
that
you can't even practice for it in a simulator. It can even be chaotic,
in the true mathematical sense. Many of the attempts at automated
recovery I known of literally crashed on that point; you can't do
optimum control here, because it leads you into an unstable
calculation.
Instead, you have to have your algoritm flying off the optimum path,
in
order to keep a stability margin. (Knowing how much off the optimum
path
it should be, and what a stability margin even *is*, is to date an
unknown as well. It's difficult to quantify.)
So, how would I fly out of a death spiral, suddenly and against
expectation fully knowing I was in one? Fully knowing which way, how
fast, at which height, I and my aeroplane was going? Well, obviously,
I
would have to regain lift, evenas I was falling. I'd use ailerons to
gain "level flight" evenwhile falling. While that was done, I'd yoke
up,
no matter the orientation of the airframe (assuming I wasn't downright
inverted), in order to gain altitude and *true* level flight. I'd put
the engines in idle and maybe spoil the airfoil, for want of
structurally sound airspeed and the g-forces which necessarily come
after a recovery from a spiral. Something like that.
Even if I did all of that *just* right, I'd probably contact terrain.
All on-board would be lost. Because recovering from a death spiral,
once
it's started and developed well, is pretty much an inhuman feat. It's
almost impossible for a computer to do, as well. The many algorithms
which have been tried out, taking control away from the human pilot,
none of them have been shown to do any good either.
Fons, this is one of the other things I follow. Amateurishly, but I
still do.
https://www.youtube.com/watch?v=b7t4IR-3mSo
https://www.youtube.com/watch?v=JhzaogGQNFU&t=1056s
https://www.youtube.com/watch?v=o7N0pshAC0k
Now the fun thing is, as my pilot friends say, you can actually *even*
do a full, level, aileron roll on all of the AirBus jumbojets.
Definitely not recommended, might lose your licence, and you'd have to
disengage a number of safety systems, then flying by eye and touch
alone. But I'm told all of them *can* do showflying manoeuvres if need
be. My nerd friends even claim to me, if need be, the things might be
capable of autonomous inverted flight; I'm not too sure if any flight
manufacturer actually ever went so far, but if one did, it surely
would
be AirBus. Damn, woman, stop it right *here*!
> A non-zero roll angle means that part of the lift force generated by
> the wings is now sideways.
Now, once we've handled conventional aircrafts, let's take on fighter
aircraft. The fifth generation ones, like the F22 and the F35. Their
fly-by-wire flight control surface work in concert, and often in a
fully
different way. For example, instead of there being separate spoilers,
in
order to introduce a limited airfoil separation/stall, the primary
control surfaces introduce a time-limited and measured stall in a
servo
loop. And then not, again.
Those things basically fly themselves, strategically stalling some of
the control surfaces when need be. For example, in order to
automatically come back from a leftward death spiral, the fighter will
make its right back control surface stall, losing lift momentarily,
and
so roll the aircraft as well as make it pitch up, lose airspeed
because
of the increased drag, and so gain an amount of lift.
Fons, I like this kind of analysis. Optimum control, and where its
limits lay. This stuff is *fun*!
> That - and not the rudder - is what makes the aircraft make a turn.
The rudder is also a fun turn — pun intended. It's even "fun" in the
death spiral recovery manoeuvre. Because you cannot efficiently come
out
of a spiral using only ailerons. While coming down, in order to
rectify
the spiral, you actually have to apply a lot of rudder as well.
Otherwise you'll end up a whole lot more down, with disastrous
airspeed.
> The vertical component of lift is reduced, and a pitch-stable aircraft
> will just by itself increase its airspeed to restore it. It can do
> that only by going down at that same time.
True.
> Unless you watch the horizon or the attitude indicator, you will
> not be aware that this is happening.
True. Whence the 178 seconds above. Also, "spatial disorientation in
aviation". This Youtube channel of mine, "Mentour", has done quite a
number of features on just this thing. He's a commercial pilot, and
even
has access to flight simulators. See above even for him inverting his
native 737 in one.
> As the roll angle increases, the g-force will apparently remain
> vertical (relative to the aircraft) but increase as well.
Actually the g-force does not increase at all. That's why the death
spiral is so nasty: you don't feel anything at *all*, evenwhile you're
going nose down into the ground.
Much of that is because of the intrinsic stability of the aircraft.
Because the stability means the craft wants to stay at 1g towards the
floor. While it stays that way — no matter its actual attitude — you
won't feel anything off even if the thing is inverted in a barrel roll
—
a nice and harmless aerobatic movement — or in a death spiral — with
at most two seconds to die.
> And at some point you will notice that you are pinned down in your
> seat and unable to move - you are effectively in a centrifuge, way too
> fast, going down, and the g-forces will be so high that they can break
> up the aircraft.
This only happens once you gained too much airspeed and try to recover
by pulling up on the yoke. True, if you're already there, not much can
be done to recover. But at least don't then pull up the yoke too fast
in
order to break the airframe. At max do something like a "gentle" 5g
curve, and if you then manage to not crash into the terrain, level off
and apply some spoiling.
Ah, you too think about this. Hmm. 8)
> To recover: > > 1. Reduce power to idle.
Preferably as soon as you know you're losing altitude. Because you'll
be
trading potential energy for kinetic energy/speed from the get go.
This
is also why I mentioned fighter jets and dog fighting from the get go:
that energy count-down (or up) is how dogfighting has been counted
from
the start. It's how dogfights are won, and the energy management is
also
how planes are either crashed or landed safely.
>> 2. Bring the wings level. This has to be done gently, to avoid even
>> more mechanical stress.
Yes. However, this is difficult to do once you went into spatial
disorientation, your synthetic horizon is at something like 120
degrees,
and you descend at a about a five kilometres per minute, from an
altitude of, say, a generous ten thousand feet. Within a thick cloud
cover, with all of your instruments yelling at you at the same time.
>> 3. As the wings return to level, the excessive speed will put the
>> aircraft into a steep climb.
What is "level", here? In a death spiral, the optimum recovery will
take
you through a route where you'll *definitely* not be level. Your nose
will be looking down, at an airspeed which is *way* over your craft's
design limits. That will also take place well after you can laterally,
in ailerons, balance the aircraft; as such, even a very little take on
the ailerons, or the rudder, the yoke, would immediately either stall
some control surface, or made better, tear each of them apart. And you
don't really know what is "level" hear, either; even your
instrumentation is probably fucked up already; believe you me, no
inertial thingy ever survives the kind of vibration an aircraft
induces
on itself when put into a multiple g's acceleration, combined with a
wide stall.
>> Let it happen but keep the pitch angle under control.
Exactly so. "Let it happen." Many of the worst accidents on record
have
happened because pilots fought their planes, instead of "going with
the
flow" which a plane, designed to be statically stable from the start,
would have done by itself. For example, (ya'll, prolly not Fons) take
a
look at: https://en.wikipedia.org/wiki/Pilot-induced_oscillation .
>> You will regain some of the lost altitude, and airspeed will
>> decrease.
Recovery from a near miss death spiral is still more involved. Because
you might have to operate the aircraft at structural load, and do a
recovery from a prolonged stall over all of the airframe. You might
actually have to "fly" your airframe over a minute in a full stall
over
every part of it, and then try to regain aerodynamic control. "After
sinking, flying, and shaking like a rock from a cannon."
It can be done. But nobody teaches you how to do this, and in fact, I
don't know of *one* algorithm which has flown this route.
> 4. As you approach normal airspeed, bring back power and level off.
That should be obvious, then. It's that third stage before "Profit"
which always slights the eye. ;)
> Ciao,
Moro.
--
Sampo Syreeni, aka decoy - de...@iki.fi, http://decoy.iki.fi/front
+358-40-3751464 <http://decoy.iki.fi/front+358-40-3751464>, 025E D175
ABE5 027C 9494 EEB0 E090 8BA9 0509 85C2
_______________________________________________
Sursound mailing list
Sursound@music.vt.edu
https://mail.music.vt.edu/mailman/listinfo/sursound - unsubscribe
here,
edit account or options, view archives and so on.
