The A320 is primarily a fly-by-wire aircraft, with the pilots’ joystick and rudder-pedal inputs not going directly to the flight-control surfaces, but, rather, to the aircraft’s flight-control computers, which then decide what, if any, control-surface deflection is necessary, and send the requisite commands to the hydraulic actuators which move the control surfaces:1

sorcery 101

However, in the event of a complete failure of the aircraft’s flight-control-computer system, the A320 can switch to a backup mode known as mechanical law, which bypasses the FCCs completely.2 In this mode, the only one of the primary flight controls to remain active is the rudder, using an emergency cable mechanism which sends the pilots’ control inputs directly to the rudder actuators; with the elevators, ailerons, and spoilerons all inactive, pitch control is maintained using manual horizontal-stabiliser trim (which uses a similar cable mechanism to send trim commands to the hydraulic motors powering the stabiliser jackscrew) and varying engine power, and the ability to control roll separately from yaw is lost completely (to roll the aircraft in one direction or the other, it has to be yawed in the desired direction using rudder inputs and differential engine thrust; the resulting sideslip then causes the aircraft to roll towards the trailing wing, due to the positive slip-roll coupling generated by the A320’s rearward wing sweep):

look, Ma, no direct roll control

The reason the elevators, ailerons, and spoilerons are unuseable in mechanical law is that, unlike the rudder, they have no direct hydromechanical connections to the cockpit controls at all, being connected only via the flight-control computers (thus leaving them cut off from the pilots’ control inputs in the event of a total failure of those computers). If the A320 did have hydromechanical backup controls for the elevators, ailerons, and spoilerons, the pilots would be able to control the aircraft in each of its three axes separately (rather than with roll dependent on yaw, which provides very sloppy roll control with a significant time delay between the initiation of the yaw and the resulting sideslip-induced roll, and with a small but significant initial rudder-induced rolling moment in the wrong direction), and would be able to do so using the same types of control inputs as would be used in normal, fly-by-wire flight (albeit likely with considerably-increased control forces and a markedly-different force gradient), rather than having to switch to a completely different input style in mechanical law:

it might be a bit tricky routing the control cables through all those sharp 90-degree turns, but I'm sure they can handle it

So why is the rudder the only one of the A320’s primary flight controls to have hydromechanical backups allowing its use in mechanical law?

1: The A320’s flight-control system, and that of all later Airbus aircraft, has several different fly-by-wire flight-control modes. In some of these (direct law, alternate 1/2A/2B law), the pilots’ flight-control inputs in some or all control axes are simply passed straight through the computers unfiltered, with the computers parroting whatever the pilots ask of the aircraft without any of the translation present in the usual operational mode (known as normal law); however, the pilots’ inputs are still passed to the computers first, and only then to the control-surface actuators, rather than going directly aft.

2: The pitch and roll/yaw axes can each also revert to mechanical law separately in the event of a major fault in one of the primary flight controls (for instance, a double elevator jam would cause the aircraft to revert to mechanical pitch control), while still leaving the other in fly-by-wire mode.

  • 1
    $\begingroup$ Failures beyond direct law doesn't really worth investing. The usefulness of mechanical law is debatable, in the sense that if it's needed at all and how useful it is when it's needed. The use of FBW is an indication that the plane is uncontrollable beyond direct law. Instead of spending money and weight on mechanical law, you are much better off adding another layer of redundancy to FBW. $\endgroup$ – user3528438 Apr 8 at 22:21
  • $\begingroup$ I think you may have answered your own question, Sean. I don't really know much about this stuff, but while looking at the graphs, I can understand a little bit. The only question is: are these images from credible sources (e.g., faa, airbus, or an a320 operator)? $\endgroup$ – Air Canada 001 Apr 9 at 1:51
  • $\begingroup$ @user3528438, there was one failure which lead to use of the mechanical backup fairly recently, on Feb 28th 2018. It is the only one I've ever heard of. $\endgroup$ – Jan Hudec Apr 9 at 6:47

The thing is your first picture is not really correct. The links from side-sticks go through the “magic” box, but the links from trim wheels and pedals don't. Instead there is a link from the “magic” box to the pedals and trim wheel. So if the “magic” stops working, there is no change in configuration. The pedals and trim wheels just continue working because they are connected differently.

The side-stick output needs to go through the “magic” box; that's the point of using it. So this way they have a mechanical backup in case all the five components of the “magic” box fail without actually having any extra links. Any other mechanical backup would need extra links, which would negate the advantages of the design for something that so far only seems to have happened once.

| improve this answer | |
  • $\begingroup$ "The side-stick output needs to go through the “magic” box; that's the point of using it." Why would that be the case for the joystick, but not the rudder? $\endgroup$ – Sean Apr 22 at 20:04
  • $\begingroup$ @Sean, I suppose it would make some sense to make the rudder pedals fully electronic as well, but they were easier to connect mechanically and just adding the pilot command with the yaw damper command is closer to the desired outcome (which is probably selecting slip-skid ball target) than it would be for the pitch and roll commands. $\endgroup$ – Jan Hudec Apr 22 at 21:09
  • $\begingroup$ Right, but, if the yaw commands can do just fine bypassing the magic box, then why can't the pitch and roll commands? $\endgroup$ – Sean Apr 29 at 0:01
  • 1
    $\begingroup$ @Sean, because the pitch and roll commands are requests for vertical acceleration and roll-rate respectively, which requires varying the surface deflection according to the actual aircraft response that depends on speed, weight and balance, and because the meaning of pitch and roll commands varies with the aircraft attitude and angle of attack. So it takes logic and a feedback loop to even interpret the commands. $\endgroup$ – Jan Hudec May 2 at 20:50

The mechanical backup is only meant to maintain basic controllability during a temporary failure of the whole fly-by-wire system. As the OP mentioned, the rudder is deemed sufficient for basic roll control, while the horizontal stabilizer is deemed sufficient for basic pitch control even without any stability augmentation. That being said, the mechanical backup is not meant to achieve continued safe flight and landing (I would be really surprised if the mechanical backup meets all of the ARAC minimum controllability guidance); it's expected that at least one flight control computer can be rebooted to take over control after the initial failure (Ref. Airbus A320 Crew Manual).

The reason for relegating the mechanical backup to rudder and trimmable stabilizer, instead of to elevator and ailerons, pretty much has to do with the choice of the passive sidestick as the pitch and roll inceptor. It lacks mechanical linkages to translate its motion to external assemblies, nor does it have a feel unit to provide feedback for the airloads. This leads to significant weight savings and simplicity in design, and is the primary advantage of the passive sidestick.

I suspect that the reason for even having a mechanical backup scheme is due to mistrust in FBW systems for civil aviation when they first came to use. As trust in redundancy and dissimilarity grew, newer generations of Airbus aircraft (e.g. A380, A350, A220) did away with it completely, and instead have an electrical backup servo control loop on each of the axis.

| improve this answer | |
  • 1
    $\begingroup$ I know this sounds a bit far fetched, but what if there was an EMP bomb or something similar that went off? $\endgroup$ – lpydawa Apr 9 at 6:40
  • 2
    $\begingroup$ @Ipydawa Then it's game over. The FADEC is all electronic so you will lose engine control, along with your flight controls, perhaps also IMUs, ADS, etc, depending on how much electrical components are within the standby instrumentation. $\endgroup$ – JZYL Apr 9 at 19:35
  • 1
    $\begingroup$ @JZYL, the stand-by instruments are a display too these days, and IIRC display simply the data from ADIRU 3. $\endgroup$ – Jan Hudec Apr 9 at 19:43

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.