Airbus A320 flight controls
P. Wezeman
mentioned conplaints of a lack of proper force feedback in the Airbus A320
sidestick control. Could someone who is familiar with the system please
describe it briefly? In particular, does it work by sensing force or by
sensing movement? Roughly how much force does it take to operate it? If the
stick moves, where is the pivot point in relation to the pilots hand and
what is the range of movement and force gradient? I am also interested
in any pros and cons of the device.
Thanking you in anticipation,
Peter Wezeman
Robert Dorsett
The A320 flight control system uses a load-demand control law, based on
something called "C*". In effect, the position of the side stick commands
an aircraft loading. Once commanded, the loading is preserved. The side
stick can be viewed as commanding AOA.
In pitch, the maximum load is 10 daN, and can be rotated +- 16 degrees.
It is oriented 20 degrees forward, and has a sensing threshold of 0.5 daN.
In roll, its maximum load in is 3 daN, out 2 daN. Its threshold is 0.4 daN
in both directions. Maximum deflection is +-20 degrees, and it is oriented
12 degrees inboard.
The joystick is not coupled. That means that the two joysticks are com-
pletely independent critters. Their inputs are algebraically added. If
one pilot pulls full-left and the other full-right, nothing will happen.
There is no backfeed. The joysticks do not reflect aerodynamic forces.
An override capability exists; it's a button in the grip. When one pilot
has overriden another, an arrow pointing to the other pilot flashes in the
glareshield. There is also an aural annunciator. This override, in effect,
"turns off" the other stick, until the pilot tries to use it, again.
The pilot's arm rest can be adjusted for a more comfortable fit, but in
reality, most pilots have reported that the arm rest position doesn't sig-
nificantly affect control.
--
Robert Dorsett Moderator, sci.aeronautics.simulation
r...@netcom.com aero-si...@wilbur.pr.erau.edu
ftp://wilbur.pr.erau.edu/pub/av
Emailia
the only feel the pilot has. The ELAC (Elevator and Aileron Computer) have
complete control of the forces on the ailerons and elevators. The SEC (
Spoiler and Elevator Computers) also have control, interfacing with each
other through another interface computer. The Control stick requires
slightly more pressure than a computer joy stick you may be familiar with.
Most pilots report that they are quite pleased with the operation of this
configuration.
The pivot point is approximately 6 cm from the wrist.
Thanks for asking
Tom Reeves (Airbus Inspector)
Robert Dorsett
> In a recent exchange of postings in rec.travel.air several people have
>mentioned conplaints of a lack of proper force feedback in the Airbus A320
>sidestick control. Could someone who is familiar with the system please
>describe it briefly? In particular, does it work by sensing force or by
>sensing movement? Roughly how much force does it take to operate it? If the
>stick moves, where is the pivot point in relation to the pilots hand and
>what is the range of movement and force gradient? I am also interested
>in any pros and cons of the device.
Oh, I forgot the pros and cons:
Pros: it's dirt-cheap. No back-driving servos, no interlinks between the
two sticks. Easy and cheap to maintain. Probably saves a couple hundred
pounds (hey, they had to get the weight savings SOMEWHERE, to justify the
FBW experiment :-)).
Cons:
1. The lack of interconnectivity makes non-verbal communication between
pilots somewhat questionable. The "algebraic" addition of command forces
means that both pilots may end up reacting to a threat the wrong way. It
is thus a "fuzzy" interface.
This observation usually results in a note that in a real airplane, the
strongest pilot wins. Not really. In a real airplane, the weakest dog
usually "rolls over" and defers to the pilot with the stronger personality.
In any event, the communication is clear: you can't move the stick.
2. Similar constraints can also make training interesting for new pilots.
Since most airlines operating the A320 are ab initio airlines, and since
the A320 is an entry-level aircraft, I'm sure this can make life interesting.
3. The lack of force feedback almost *begs* for protections. The same
forces exist (provided by the nonlinear springs) whether the airplane is
on the ground or in flight.
4. A few papers have shown force-less joysticks to generally be among the
weakest there are for most applications. The most obvious other airplane
using one is the F-16, but it uses something rather different (it's a force-
joystick, not a positional joystick).
The factors 1-4 have been touched upon in the literature and this and other
newsgroups since 1986 or so. All have been cited as justification for the
simulation of conventional control columns by Boeing in the 777.
Brad Gillies
"P. Wezeman" <pwez...@blue.weeg.uiowa.edu> wrote:
> In a recent exchange of postings in rec.travel.air several people have
>mentioned conplaints of a lack of proper force feedback in the Airbus A320
>sidestick control. Could someone who is familiar with the system please
>describe it briefly? In particular, does it work by sensing force or by
>sensing movement? Roughly how much force does it take to operate it? If the
>stick moves, where is the pivot point in relation to the pilots hand and
>what is the range of movement and force gradient? I am also interested
>in any pros and cons of the device.
>
I am a mechanic for an airline which operates A320s and have "played" with the
system alot.
The sidestick is for all intents and purposes a joystick. It is basically a
series of transducers hooked up to a stick to provide input to a computer.
The feel provided by the stick is artificial. There is a series of "snubbers"
which give the stick artificial weight. The harder you push the stick the
more it resists (like the shock absorbers on a car).
The stick provides no tactile feedback to the pilot, so he has to be visually
aware of his situation. For example if the aircraft for sake of argument had
an aileron which was not deflecting normally there would no less and no more
resistance on the stick. This also is noted when there is no hydraulics
available. The stick feels the same with hydraulics as it does without.
The force required to move the stick is roughly equivalent to any madern
joystick such as the thrustmaster. Except that the stick resists rapid
movement for obvious reasons.
Icannot comment on the force gradient as it changes with the force applied to
the stick.
The pros:
roomier flightdeck in front of the pilots
pilot and F/O get table trays to eat lunch on (or laptop computers)
more game like
The cons:
Not tactile feedback to pilot
decreased situational awareness
more game like
Hope this is what you wanted.
Brad
C. Marin Faure
Dorsett) wrote:
> In article <airliner...@ohare.Chicago.COM> "P. Wezeman"
<pwez...@blue.weeg.uiowa.edu> writes:
> > In a recent exchange of postings in rec.travel.air several people have
> >mentioned conplaints of a lack of proper force feedback in the Airbus A320
> >sidestick control. Could someone who is familiar with the system please
> >describe it briefly?
> The joystick is not coupled. That means that the two joysticks are com-
> pletely independent critters. Their inputs are algebraically added. If
> one pilot pulls full-left and the other full-right, nothing will happen.
I had heard the left-plus-right-equals-nothing formula was the case on the
original version of the system, but that it has subsequently been
changed. Imagine this scenario: both pilots suddenly become aware of
another airplane rapidly approaching on a collision course. The
instinctive reaction of most pilots is to turn toward the window that is
next to him. The captain slams his stick left, the first officer slams
his stick right. Result: the plane continues straight ahead into the
collision. I believe the very real possibility of something like this
occuring caused Airbus to have the stick authority software rewritten, but
I don't know how the priorities are assigned, if they have, in fact,
changed it.
C. Marin Faure
author, Flying A Floatplane
Brad Gillies
>The joystick is not coupled. That means that the two joysticks are com-
>pletely independent critters. Their inputs are algebraically added. If
>one pilot pulls full-left and the other full-right, nothing will happen.
>
>There is no backfeed. The joysticks do not reflect aerodynamic forces.
>
>An override capability exists; it's a button in the grip. When one pilot
>has overriden another, an arrow pointing to the other pilot flashes in the
>glareshield. There is also an aural annunciator. This override, in effect,
>"turns off" the other stick, until the pilot tries to use it, again.
The pilot in the left seat always has priority if both override buttons are
pressed so the captain can always have control of the aircraft.
This is to prevent both pilots from taking different action with a resultant
nothing.
Robert Dorsett
> pressed so the captain can always have control of the aircraft.
> This is to prevent both pilots from taking different action with a resultant
> nothing.
I have not been able to find anything which suggests this. A recent FCOM
from a major carrier operating the airplane, the original Aeroformation
training materials, and a variety of academic papers published by Airbus
engineers all suggest that the two seats have a peer arrangement. There is
no bias in favor of either seat in terms of basic control inputs, and as
far as the override button, the last one to hit it gets it.
There are a number of reasons why one does not want to assign bias to a
seat:
- training operations frequently put a training captain in the right
seat, with a trainee in the left.
- It violates the Pilot Flying/Pilot Not Flying paradigm (which assigns
distinct responsibilities to each pilot, each of whom are (usually) trained
to the same technical proficiency level). In this paradigm, the PF is
directly in charge with the airplane's flight guidance; the PNF handles all
other activities (radio, navigation, checklists, etc). Since the PF is "in
the loop", if we're going to assign priority to either pilot, the PF should
have it. The PNF will almost certainly have his head down working on the
FMGS.
- Situations arise in scheduling when two captains are in a cockpit. Who
sits in the left seat is often determined by professional courtesy, but
the aircraft commander is often determined by a coin flip.
SO, I kind of doubt this. But since the Airbus software can be revised,
I welcome any corrections. Please cite your source. :-) I'd also be
interested in hearing from pilots or mechanics for airlines where this
does and does not exist: could even be a stupid customer option. :-)
P. Wezeman
replies to my posting about the operation of the A-320 side stick.
Mr. Reeves wrote, "The side stick can be viewed as commanding A.O.A."
Lift being proportional to the square of airspeed at a given angle of
attack, would I be correct to infer that if a given position of the side
stick gives an acceleration of one G at a certain airspeed, then the same
position of the side stick at twice that airspeed would give an acceleration
of four Gs? ( I realize that the system would not actually allow the loading
to go this high.)
In roll, does a given deflection of the side stick command a given
roll rate regardless of airspeed, or does it command a given control
surface movement, i.e. a higher roll rate at a higher speed?
Thank you again,
Peter Wezeman
Francis JAMBON
fau...@halcyon.com (C. Marin Faure) wrote (=E9crivait)=A0:
> I had heard the left-plus-right-equals-nothing formula was the case on the
> original version of the system, but that it has subsequently been
> changed.
I don't know if the "side stick priority logic" has been changed but I
have found these informations from a rather old FCOM :
A320 Flight Crew Operating Manual - FLIGHT CONTROLS - CONTROLS AND INDICATORS
1.27.40 - P 3 - REV 16 - SEQ 002 - "For training only"
--------------------
| LATERAL CONSOLES |
--------------------
SIDESTICKS
----------
Sidesticks, one on each lateral console, are used for manual pitch and
rool control.
They are springloaded to neutral.
When the A/P is engaged a solenoid operated detend locks both side sitcks
in the neutral position. If the pilot applies a force above a given
threshold (5 daN in pitch, 3.5 daN in roll) the stick becomes free and A/P
disengages.
(picture of one sidestick deleted)
Side stick priority logic :
---------------------------
- When only one pilot operates the sidestick :
His demand is sent to the computers.
- When the other pilot operates his sidestick in the same or opposite
direction both pilots inputs are algebraically added. The assition is
limited to single sitck maximum deflexion.
A pilot can desactivate the other sick and take full control by pressing
and keeping pressed his priority takeover pushbutton.
If a takeover push button is pressed more than 30 seconds the system will
latch, thus allowing the push button to be released without losing priority.
However, at any time, a desactivate stick can be reactivated by momentarily
pressing its takeover push button.
If both pilots press their takeover pushbuttons, the last pilot to press will
get the priority.
Note: If an auto pilot is engaged, the first action on a take over push button
----- will disengage it.
In a priority situation
-----------------------
- A red light will come on in front of the pilot whose stick is desactivated
- A green light will come on in front of the pilot who has taken control,
if the other stick is not in the neural position (to indicate a potential
and unwanted control demand).
Note: If, on ground at take off application, one stick is desactivated,
----- the <<take off CONFIG>> warning is triggered.
--- end of FCOM page ---
Hope it helps. Don't ask me where I've found the FCOM.
Francis
--
Francis JAMBON | Equipe Ingenierie des Interfaces Homme-Machine
<Francis...@imag.fr> | Human-Computer Interfaces Engineering Group
Informations sur/on WWW -> http://www-lgi.imag.fr/Les.Personnes/Francis.Jambon/
Telephone / Phone : (+33) 76 63 59 70, Telecopieur / Fax : (+33) 76 44 66 75
___________ CLIPS-IMAG, B203, BP 53, 38041 Grenoble cedex 9, France ___________
Jerry Schaefer
of a commerical grade jet would automatically and instinctly turn to the
RIGHT as would the oncoming airplane. This is what is taught to pilots.
J. Schaefer
-
JERRY SCHAEFER DHQ...@prodigy.com
C. Marin Faure
(Jerry Schaefer) wrote:
> I would presume that any good pilot, and further presume that the pilot
> of a commerical grade jet would automatically and instinctly turn to the
> RIGHT as would the oncoming airplane. This is what is taught to pilots.
You would hope this would be the case, but studies have shown that most
pilots' IMMEDIATE instinctive reaction is to turn toward the window next
to them. This would apply in a near panic situation. Given at least a
couple of seconds to let training take over, the initial instinctive
reation would in most cases be overruled by the training which says bank
right.
Pete Mellor
writes:-
> I don't know if the "side stick priority logic" has been changed but I
> have found these informations from a rather old FCOM :
>
>A320 Flight Crew Operating Manual - FLIGHT CONTROLS - CONTROLS AND INDICATORS
>1.27.40 - P 3 - REV 16 - SEQ 002 - "For training only"
and transcribes that page of text. I referred to a later copy of
the FCOM to check. The whole manual that I referred to was at
REV 19, and the page that Francis quotes is at REV 18 in that version.
The text is identical except for two lines:-
If a takeover push button is pressed for more than 40 seconds
the system will latch, ... ^^
(changed from "30 seconds"), and:-
However, at any time, a deactivated stick can be reactivated by
momentarily pressing the takeover push button on either stick.
^^^ ^^^^^^^^^^^^^^^^
(changed from "... pressing its takeover pushbutton.")
The effect of the first change is obvious.
The second one is puzzling me. It would seem fairly sensible that a
crew member whose *own* stick was deactivated should be able to
reactivate it by pressing its *own* takeover pushbutton, and the
REV 16 text states this clearly. However, the REV 18 text implies
that, having deactivated FO's stick and taken control, the Captain
would reactivate FO's stick by a subsequent action on the *Captain's*
takeover pushbutton (or vice versa, reversing the roles of the two
crew members).
Questions:-
1. Do these changes mean that the logic has been modified, or were
they made because the manual was originally incorrect compared
to the existing logic?
2. Have I understood the second modified statement correctly, or
is this another example of the practical application of the
"fuzzy logic" for which the FCOM is famous?
In answer to C. Marin Faure's original question, I am not aware
of any change to the "left-plus-right-equals-nothing formula", nor
of any alteration to the symmetry of the sidestick logic. There
are no Operations Engineering Bulletins (OEB) relating to the
sidesticks in my copy of the FCOM.
> Don't ask me where I've found the FCOM.
Ditto! :-)
Pete
----
Peter Mellor, Centre for Software Reliability, City University, Northampton
Square, London EC1V 0HB, UK. Tel: +44 (171) 477-8422, Fax: +44 (171) 477-8585
E-mail: p.me...@csr.city.ac.uk
Pete Mellor
relating to sidestick priority logic:-
> However, at any time, a deactivated stick can be reactivated by
> momentarily pressing the takeover push button on either stick.
> ^^^ ^^^^^^^^^^^^^^^^
>
>(changed from "... pressing its takeover pushbutton.")
> ...
> ... It would seem fairly sensible that a
> crew member whose *own* stick was deactivated should be able to
> reactivate it by pressing its *own* takeover pushbutton, and the
> REV 16 text states this clearly. However, the REV 18 text implies
> that, having deactivated FO's stick and taken control, the Captain
> would reactivate FO's stick by a subsequent action on the *Captain's*
> takeover pushbutton (or vice versa, reversing the roles of the two
> crew members).
I have since spoken to two members of BALPA, who confirmed that the
second interpretation is correct, and that this is what the pilots
wanted. I would guess that this required a software modification.
Jean-Marie Dany
have found these informations from a rather old FCOM :
A320 Flight Crew Operating Manual - FLIGHT CONTROLS - CONTROLS AND
INDICATORS
1.27.40 - P 3 - REV 16 - SEQ 002 - "For training only"
and transcribes that page of text. I referred to a later copy of
the FCOM to check. The whole manual that I referred to was at
REV 19, and the page that Francis quotes is at REV 18 in that version.
The text is identical except for two lines:-
If a takeover push button is pressed for more than 40 seconds
the system will latch, ... ^^
(changed from "30 seconds"), and:-
However, at any time, a deactivated stick can be reactivated by
momentarily pressing the takeover push button on either stick.
^^^ ^^^^^^^^^^^^^^^^
(changed from "... pressing its takeover pushbutton.")
The effect of the first change is obvious.
The second one is puzzling me. It would seem fairly sensible that a
crew member whose *own* stick was deactivated should be able to
reactivate it by pressing its *own* takeover pushbutton, and the
REV 16 text states this clearly. However, the REV 18 text implies
that, having deactivated FO's stick and taken control, the Captain
would reactivate FO's stick by a subsequent action on the *Captain's*
takeover pushbutton (or vice versa, reversing the roles of the two
crew members).
Questions:-
1. Do these changes mean that the logic has been modified, or were
they made because the manual was originally incorrect compared
to the existing logic?
That is correct the last guy that pushes the button is always the winner, if
ou keep it maintained more than 40 seconds you kill the other sidestick
2. Have I understood the second modified statement correctly, or
is this another example of the practical application of the
"fuzzy logic" for which the FCOM is famous?
In answer to C. Marin Faure's original question, I am not aware
of any change to the "left-plus-right-equals-nothing formula", nor
of any alteration to the symmetry of the sidestick logic. There
are no Operations Engineering Bulletins (OEB) relating to the
sidesticks in my copy of the FCOM.
The sticks react algebrically when they are actionned at the same time if
you don't push the priority button which is also the instinctive disconnect
of the autopilot. So the captain could push his stick to the left and the
first officer to the right nothing will happen until somebody press the
priority button. Strange !!
Brad Gillies
> I don't know if the "side stick priority logic" has been changed but I
> have found these informations from a rather old FCOM :
>
I ran through the logic test from the Maintenance manual for the sidestick
logic last night and this is what I found...
Aircraft is on short term lease from an Company in Britain so the logic could
be different.
1) Pushing either sidestick button will transfer priority to the associated
side. Aural warning "Priority Left (right)".
2) Holding the side stick for 40 seconds deactivates opposite stick
3) Pushing button on deactivated stick, reactivates stick.
4) holding button on deactivated stick switches priority to that stick.
5) Moving one stick right and one stick left results in no flight control
deflections.
6) same as above for all flight control surfaces. except rudder.
This was done according to the MM and passed the test.
Any questions?
--
-----------------------------
Brad Gillies Br...@io.org
AME (Canada), A&P, PPASEL
HTTP://www.io.org/~bradg
-----------------------------
If it is Blue, Frozen and falls
through your roof, It is not a
Meteorite.
peter (p.j.) ashwood-smith
>logic last night and this is what I found...
>Aircraft is on short term lease from an Company in Britain so the logic could
>be different.
>
>1) Pushing either sidestick button will transfer priority to the associated
>side. Aural warning "Priority Left (right)".
>2) Holding the side stick for 40 seconds deactivates opposite stick
>3) Pushing button on deactivated stick, reactivates stick.
>4) holding button on deactivated stick switches priority to that stick.
>5) Moving one stick right and one stick left results in no flight control
>deflections.
>6) same as above for all flight control surfaces. except rudder.
>
>This was done according to the MM and passed the test.
>Any questions?
I'm sorry but this stuff scares me silly. This all seem so convoluted and
unlikely to be correctly remembered by a flight crew in an emergency. Compared
to a simple mechanical linkage between the two devices which incorporates
feed back of what each pilot is doing and with what force and with no need
for silly averaging algorithms etc. All for what ... to save a few pounds of
cabling or whatever to link the controls.
Digital is not always better.
--
Cheers,
-------------------------------------------------
Peter Ashwood-Smith | email: pet...@bnr.ca
Bell Northern Research | work : (613)763-4534
Ottawa, Ontario, Canada | home : (819)595-9032