Fly by wire

Discussion in 'Instruments / Avionics / Electrical System' started by StRaNgEdAyS, Oct 22, 2003.

Help Support HomeBuiltAirplanes Forum by donating:

  1. Oct 22, 2003 #1

    StRaNgEdAyS

    StRaNgEdAyS

    StRaNgEdAyS

    Well-Known Member

    Joined:
    Oct 20, 2003
    Messages:
    817
    Likes Received:
    3
    Location:
    Northern NSW Australia
    I have been looking and looking for some online resources to give me some info on electic controls...
    I kinda like the idea of having a real joystick and servos rather than running cables to the control surfaces, and I guess it will probably save a bit of weight as well.
    I am keen to check them out but have yet to find a site that goes into any deatail about them.
    with the time I reckon it's gonna take me to get to completion, I should be able to squirrel away the needed cash to purchase it when it copmes time to install it.
     
  2. Jan 10, 2004 #2

    pylon500

    pylon500

    pylon500

    Well-Known Member

    Joined:
    Dec 26, 2003
    Messages:
    270
    Likes Received:
    2
    Location:
    Taree Airport Australia
    I was thinking along these lines about ten years ago, and was hoping to be the first in Australia (if not the world) to have a fly by wire ultralight complete with auto-pilot.:gig:
    After reading articles in old 'HomeBuilt Monthly' magazines, I saw the use of model servos to drive trimmer tabs on long-Eze and BD-4's.
    I then saw in model mags the new 'auto stabilizing' systems for R/C choppers and thought, combine the servo's to drive trim tabs on the actual control surfaces with the built in gyro system slaved to an R/C joystick in the arm-rest of the cockpit and Whallah! sidestick fly by wire with auto pilot for less than about $1,000AUD :D
    Never finished the aeroplane I was going to incorporate it in, but was planning to have all balanced 'push-rod' controls so the servo's could move the controls fast enough so as not to be chasing itself all the time :rolleyes:
    I suppose some of the R/C yacht winch servo's are strong enough to drive the controls directly.
    As for electric trims and the like, I think those 'MAC' servo trims are way too expensive, and with a bit of work you can make your own.
    I fitted electric flaps to my ultralight by using the motor, gearbox and clutch from a cheap 12volt battery drill, a length of threaded rod and a two way toggle switch from Tandy
    You can even get cheap electric drills with an instant stop feature to avoid over-running a position setting.
    The attachment shows the flap motor installation in my Stollite, the belcranks were for a 'flapperon' system (since removed)
    I'll get there one day,
    Arthur.
     

    Attached Files:

  3. Jan 10, 2004 #3

    StRaNgEdAyS

    StRaNgEdAyS

    StRaNgEdAyS

    Well-Known Member

    Joined:
    Oct 20, 2003
    Messages:
    817
    Likes Received:
    3
    Location:
    Northern NSW Australia
    Yeah, I've looked at those winches as well. There are a number or high energy servo's on the market that would do the job well. The trick I reckon will be rigging a joystick to drive them.
    I like the idea about the drill motor for the flaps!:D
     
  4. Jan 10, 2004 #4

    Midniteoyl

    Midniteoyl

    Midniteoyl

    Well-Known Member

    Joined:
    Sep 4, 2003
    Messages:
    2,417
    Likes Received:
    500
    Location:
    Indiana
  5. Jan 11, 2004 #5

    StRaNgEdAyS

    StRaNgEdAyS

    StRaNgEdAyS

    Well-Known Member

    Joined:
    Oct 20, 2003
    Messages:
    817
    Likes Received:
    3
    Location:
    Northern NSW Australia
    Nice presentation, I must admit it looks quite impressive, but then cost does look like it is going to be the big decider.
    I imagine this system, once developed for sale, is going to be pretty expensive!:eek:
     
  6. Jan 12, 2004 #6

    orion

    orion

    orion

    Well-Known Member

    Joined:
    Mar 3, 2003
    Messages:
    5,801
    Likes Received:
    133
    Location:
    Western Washington
    Fly-By-Wire has been of interest to the general aviation community for some time however, to date, the cost and difficulty of building such a system in a reliable and cost effective manner has been elusive. The Beechcraft presentation is glitzy and somewhat filled with a bit of BS but if they really mean to develop what they present, be sure it will cost as much as a good used airplane.

    A simpler system can be developed of course, but then we must address issues of safety and reliabiltiy also. We must also differentiate between fly-by-wire and digital proportional control.

    The latter is what we mostly see in radio controlled model airplanes. The moving surfaces are controlled by servo mechanisms that actuate the motion in proportion to the pilot's input on the stick.

    Fly-by-wire on the other hand is a combination of the proportional control, an inertial sensor and a logic circuit, which all together do not allow the airplane to be flown out of its controllable flight envelope. In order to do this one has to program what is a relatively extensive amount of code and make the whole thing work for each particular airplane you want to install this into. The system not only has to account for the airplane model, it also has to account for even small variations in the CG position as that will establish different control parameters. Due to the difficulty (and cost), I'll leave that out of this discussion.

    If we therefore assume a simple joystick operation, as we see in RC airplanes, then we have a simpler case and most likely, more affordable. For this application we therefore need several items: First, we need actuators with a position feedback sensor (automatically center themselves each time you turn on the system and allow the proportional control). The actuators need sufficient force capability to deflect the surface you wish to control. A sail winch servo will not have sufficient torque for this (calculate your control surface moments). Most likely you'll need a linear electro mechanical actuator with sufficient force capability and rate of motion. the latter is very important since you have to make sure you can control the airplane in the same manner as with a directly connected stick.

    Then you'll need an actuator controller. This allows the main controller (logic circuit) to correctly move the linear actuator. RC sevos have this system built in but affordable linear actuators do not.

    To control all this you will of course need a joystick. This cannot be a simple video game stick - you need soemthing that is more robust and preferably something that uses a non-contact sensing system that is induction based (as opposed to potentiometer based). Potentiometers are contact systems and thus can short out or wear out.

    You also need something that goes between the joystick and the actuator system. The function of this control box is to analyze the joystick input signals and convert them to motion signals for the actuator mechanism.

    Finally, you'll also need to incorporate a set of induction based potentiometers in order to provide the craft with a trim mechanism. Remeber, here your control surfaces do not float - they're rigidly connected to the actuators so any trim input has to be done by the actuators themselves. The pots allow you to control the actuators' neutral position (another reason you need the position feedback circuit).

    And that's it. The problem though is that there are no off-the-shelf systems available for this that have the finesse that we would need for an airplane. How do I kinow this? About a year ago I went through the exercise for one of our contracts where we developed just this system for a joystick controlled powered parachute. Yes, I did eventually find all the necessary components but what I though would be a relatively quick development, turned out to take nearly six months of virtually full time research.

    Even done on the cheap, if you're even remotely concerned about safety and reliability, figure you'll spend about $700 or more for each linear actuator, about $2,000 for the signal controller and voltage controller, and about $250 to $500 for the joystick. Add in a bit more for the programming of the controller, shielded wiring, hermetically sealed connectors, and you'll have about $6,000 into the prototype system.

    Also, keep in mind that this sytem does not include any backup devices. If the controls quit working for any reason, you're out of luck. There is a way to incorporate a secondary system but that will of course require a second logic circuit as well as actuators that have the capability to overpower and/or over-reach the base system.

    In short, yes this is doable and actually, compared to other normal airplane systems, not all that expensive. But it has to be done very carefuly with safety always at the forefront of your mind.
     
  7. Jan 12, 2004 #7

    StRaNgEdAyS

    StRaNgEdAyS

    StRaNgEdAyS

    Well-Known Member

    Joined:
    Oct 20, 2003
    Messages:
    817
    Likes Received:
    3
    Location:
    Northern NSW Australia
    I think I was referring to a straight joystick operation, but the concept of a true "fly by wire" was appealing enough to have also been in my mind as well. after all the real deal system is what allows us to build quicker, nimbler yet aerodymanically unstable planes that fly and fly well. :D
    I would certainly be nice to have a plane that you couln't fly past it's limits in, but then it's finding those limits that make some things fun isn't it?:whistle:
    I thnk the big questions I had in my mind were, would it be a significant weight advantage? and would it be overly expensive to build?
    Other questions would have been, should it be a force feedback system? and why don't we see much of this in GA?
    OK then so I think I can see why there isn't a lot of it about outside of commercial/military aviation, and I think I can be sure it's gonna cost me..:gig:
    I think the force feedback question would be a matter of personal taste, and wealth (be like doubling the cost I thnk wouldn't it?)
    That only leave the weight question, I have been working on a doable mechanical sidestick, and considering this is supposed to come in under 300kg MTOW, would there be any significant weight advantages in a plane this small?:confused:
     
  8. Jan 12, 2004 #8

    Midniteoyl

    Midniteoyl

    Midniteoyl

    Well-Known Member

    Joined:
    Sep 4, 2003
    Messages:
    2,417
    Likes Received:
    500
    Location:
    Indiana
    I wouldnt think that by the time you had your actuators and backups, controller and backup, and of course, a battery backup, that the weight would be low.. The torque tubes used now dont weight much at all. The weight 'savings' on larger aircraft is due to replacing hydraulic systems - which do weight.

    Force-feedback, to me, would be a must. Alot of flying comes from 'feeling' the forces at work on your plane and correcting for it automatically. Kinda hard to hold a 30-45 degree bank with no feedback in the system.
     
  9. Jan 12, 2004 #9

    orion

    orion

    orion

    Well-Known Member

    Joined:
    Mar 3, 2003
    Messages:
    5,801
    Likes Received:
    133
    Location:
    Western Washington
    Yes, all this does weigh a bit but it's actually not as bad as one might think. But for a 300 kg airplane? Probably be best to stick with the torque tubes.

    The all up weight on the powered parachute came to something under 40 pounds, including battery. For an airplane it might be less since you would not need as long an actuator as you need for the parachute application. However you do need four actuators (three if you join the ailerons together). The actuator size is also a function of the force required so there might be a weight issue there also. But, on the other hand, a small plane will not need as large of an actuator.

    The force feedback control system is actually not as necessary as you might think. Many people find this feature useful on ground based simulators however, this is usually because there is no other sensory feedback. In an airplane though, there are several other inputs you have for the senses, in addition to your instruments, so the stick feedback is not as important. The joystick however is spring loaded so there is a bit of artificial feel anyway.

    You can also use a strain-gauge type system such as is used in some fighters. This measures the strain encountered in a rod to which the stick is mounted. The stick itself has only slight motion but the signals are actually generated by the gauges attached to the shaft to which it is attached. The amount and rate of deflection is a function of the force the pilot enacts onto the stick.

    This actually has several advantages but in my mind one of the main ones is that you cannot inadvertently cause a control input since if you accidntly hit it with your hand, the strain will be minimal and thus the system will not respond.

    A joystick on the other hand can easily be accidently deflected, which will take a bit of getting used to by the pilot. One scenario might be getting your hand in the vicinity of the stick during significant turbulence.
     
  10. Jan 12, 2004 #10

    StRaNgEdAyS

    StRaNgEdAyS

    StRaNgEdAyS

    Well-Known Member

    Joined:
    Oct 20, 2003
    Messages:
    817
    Likes Received:
    3
    Location:
    Northern NSW Australia
    I would have thought that force feedback would be a must as well, but it appears from discussion with a friend of mine who actually flies heavies, that it probably isn't once you have familiarised yourself with the way the plane behaves and the inherent "lightness" of the controls. A bit like going from manual steering to power steering in a car I guess.
     
  11. Jan 12, 2004 #11

    Midniteoyl

    Midniteoyl

    Midniteoyl

    Well-Known Member

    Joined:
    Sep 4, 2003
    Messages:
    2,417
    Likes Received:
    500
    Location:
    Indiana
    Even power steering has feedback and its deliberately designed in. In GM's 'drive-by-wire' concept, the major complaint was lack of feedback, thus force feedback is now installed.

    Airliners with fly-by-wire are flown by computers that constantly adjusts the ailerons, elevators, and rudder to maintain the altitude and attitude that the pilot wants, and provides a smoother ride for the passengers but isolates the pilot from all the 'feel' and creates a huge dependence on the instruments. Not really a bad idea on a plane that big that flys that high, but I rather think it would make a huge difference in a light plane, especially once you hit turbulence. I know for me, at least, that trying to fly over mountains without being able to feel the stick 'twitch' in my hand would be disconcerting.
     

Share This Page

arrow_white