4
$\begingroup$

I'm looking to build an ultralight aircraft.

The video above shows a good example of how one is controlled with wires or with metal fulcrums/levers, but I am interested if they can be controlled electronically.

My idea of this is that it would work exactly like an R/C plane, but instead of being controlled with a radio, it will be controlled with a joystick through a wired or wireless signal.

I saw the other post that talked about a wireless fly-by-wire system and they said it could be unreliable, but that's not exactly what I'm wondering. I want to know how I can get the supplies to do it and look at the options.

How strong do the servos need to be for such an application? Obviously the rudder, elevators and ailerons will be light, but not as light as an RC plane's, so I can't use servos from a local hobby shop.

Improve this question
$\endgroup$
2
  • $\begingroup$ Possible duplicate of Would a wireless fly by wire system be practical? $\endgroup$
    – Greg Hewgill
    Jan 23, 2018 at 0:53
  • 2
    $\begingroup$ Regarding your follow-up question: wireless transmitters are prone to failure, but in the case of an RC aircraft, nobody dies so the risk is acceptable. I'm not saying you shouldn't attempt this project, but wear a parachute :) $\endgroup$
    – Sanchises
    Jan 23, 2018 at 8:38

2 Answers 2

Reset to default
7
$\begingroup$

First I'd like to say you are thinking outside the box, not a bad thing...

However for an ultralight aircraft this is impractical. The first problem you have is that most ultralight aircraft don't have electrical systems. Adding one may put you over the weight class for an ultralight. Servo's, batteries, joysticks, etc are also not very "ultralight".

The second thing you need to think about is that R/C's work because a single servo controls two surfaces in the case of ailerons, so they are synchronized. The aerodynamic forces become much, much larger as you scale up, so you will need individual (or multiple) servo's per control surface. Then you need to think about how to synchronize them, and calibrate them.

Next comes in a question of failure modes. If your elevator servo fails, what do you do? What about your electrical system? Dead battery? These FBW systems work on large aircraft because they have multiple redundancies (including traditional cable controls). You won't have the space or weight budget to do that.

It just comes down to trying to shove a complicated system into an uncomplicated aircraft.

And for a little note about "large scale" FBW systems. Usually these systems use hydraulic servo valves, which means that the aircraft includes a hydraulic system (or 3), which are electrically or cable actuated. These can provide sufficient force (along with heavy counterweights) to move the control surfaces. The servo's that you would need would not be available in electrical form since they would need to be powerful, long stroke, and fast. There are servo's that are used in some aircraft for things like flaps, but they are not fast.

Improve this answer
$\endgroup$
5
  • $\begingroup$ That makes sense. Thanks for the comment. The weight and calibration problems seem like they'll be the biggest with this. I still have some questions though... I guess I can't put a paragraph it just posted my comment... So my first question is, how big and heavy does a battery have to be that only controls servos? (answered, thanks) As far as the aileron problem, I could attempt to calibrate the servos so they alternate at the same time, or I could ditch the ailerons and go with a rudder and elevator only style, what problems could that bring up in an ultralight? $\endgroup$
    – TrapAlcubierreDrive
    Jan 23, 2018 at 2:15
  • $\begingroup$ Another question I have is what does a large scale RC plane use for "servos" some of these must get up to the weight of some ultralights right? Like this one (dpccars.com/gallery/var/albums/Large-Scale-RC-Planes/…) $\endgroup$
    – TrapAlcubierreDrive
    Jan 23, 2018 at 2:23
  • $\begingroup$ Large scale RC's use these types of servo's and they really aren't in the same weight class. The rudder/elevator only would be difficult to control roll, or difficulty with landing in cross-wind. That may be a topic for a new question though (too big for a comment). $\endgroup$
    – Ron Beyer
    Jan 23, 2018 at 2:35
  • 1
    $\begingroup$ "most ultralight aircraft don't have electrical systems" Depends on the definition of ultralight. I'm training in an ultralight, and besides being a two-seater with enough of an engine to cruise comfortably (with me and my instructor) at some 75-ish KIAS, it definitely has electrics; radio, auxilary (electrical) fuel pump, outside lights, FLARM, transponder, pretty much everything you might need in a small GA plane for VFR flight. I suspect you're going by the US definition of ultralight, but don't see anything from the OP that indicates their location, so that assumption might be premature. $\endgroup$
    – user
    Jan 23, 2018 at 15:37
  • $\begingroup$ I think I saw on YoutTUbe somewhere a few years ago showing a state-of-the-art all electrical servo for fighter jet for a fly-by-wire system. Looked like a big steel box with a massive rod and ball connector coming out of it. Obviously not available for an ultralight. $\endgroup$
    – DKNguyen
    Jul 21, 2020 at 14:37
3
$\begingroup$

DC servo's would be best because they can be operated directly from a battery. The higher the DC voltage you can supply the better: 24 VDC is preferable over 12 VDC because of the greater power loading and higher deflection velocity you can obtain.

The servo's would be dimensioned for torque: they need to overcome the hinge moments of the primary flight controls. These are functions of your airspeed, the control surface area, and the position of the hinge relative to the surface chord. For a static elevator deflection for instance:

$$H_e = C_{h_e} \cdot \frac{1}{2} \cdot \rho \cdot {V_h}^2 \cdot S_e \cdot {\bar c}_e$$

With:

  • $H_e$ = hinge moment (e is index for elevator) [Nm]
  • $C_{h_e}$ = hinge moment coefficient, function of Angle of Attack, elevator deflection and trim tab deflection.
  • $\rho$ = air density in kg/m$^3$
  • ${V_h}^2$ = airspeed around the horizontal tail [m/s$^2$]
  • $S_e$ = elevator surface [m$^2$]
  • ${\bar c}_e$ = mean aerodynamic chord of the elevator in m. Method to construct this is found here - but use elevator/aileron/rudder geometry, not the wing.

From my old lecture book, paper copy only

$C_{h_e}$ is a bit hard to find, old NACA reports should be useful here. Above is an example of the elevator hinge moment coefficients of a Fokker 27, measured in a wind tunnel. If we fill out the equation with example numbers, we could take 0.2 maximum for $C_{h_e}$; $V_h$ = 30 m/s; $S_e$ = 0.05 m$^2$; ${\bar c}_e$ = 0.05 m:

$$\Rightarrow H_e = \frac {0.2}{2} \cdot 1.225 \cdot 30^2 \cdot 0.05 \cdot 0.05 = 0.28 Nm $$

The surfaces will need to be able to withstand the aerodynamic hinge moment at highest aircraft Angle of Attack, then be able to reach the opposite deflection stop in about 1 second. So the specs of the servo's are:

  • max. torque = according to above method, substitute dimensions from comparable existing ultralights or the ones from your own design if you have that already.
  • max. deflection about +/- 30 deg. This is the useful stroke.
  • max. output velocity 60 deg/sec at full torque. You could use a gearbox for smaller servo dimensions, but the output speed after the gearbox must be > 60º/s.

Please make sure that you have a backup means of aircraft controllability for when the battery fails! Because flight control is vital for survival, the failure considerations are more important than the servo dimensioning:

  • The servo could fail hard-over, leaving the control surface in one of its stops and ending the flight in a fall. A mechanical spring can bring the surface back close to neutral position when the servo exerts no torque - you would have to dimension the servo accordingly, so that it can also generate the spring torque. Not using a gearbox is better.
  • Then you may wish to consider what Airbus did with the A320, and leave the rudder actuated by mechanical control, plus make the horizontal tail trimmable.

EDIT

As I understand now, the aircrat will be unmanned during the testing phase. That makes a huge difference: failure mode analysis considers both the chance of a failure, and the consequences. Getting gravely injured because of a flat battery would not be good.

Where to get things is nowadays: the internet. This company makes DC servo's and controllers and sells them on line. Their largest brush type DC servo is about 0.12 Nm, brushless types are smaller and lighter and can be delivered in higher torque values. The torque value from my example is only an example.

It will be possible to use ailerons, use one motor per, and synchronise them. If you would control by rudder and elevator only, you would lean over the wrong way in turns and manoeuvrability would be severely restricted. RC planes have ailerons.

Improve this answer
$\endgroup$
2
  • $\begingroup$ Thank you for the detailed explanation with calculations. That's really helpful, so thanks. .28 Nm is within the spec range of some of the servos that I've been looking at but I think the costs of doing something like this might outweigh the risk. The reason I wanted to do this is to have a safe way to test the ultralight without injuring anyone (thereby using electronic controls and a radio). I might have to make another post asking about that, because it sounds like a pretty crazy idea. Anyway, thanks for your suggestions again, I'm still going to try to make this work, so we'll see. $\endgroup$
    – TrapAlcubierreDrive
    Jan 26, 2018 at 20:36
  • $\begingroup$ If in the testing phase the aircraft is unmanned, that would make a huge difference. With failure modes there are always two things to consider: chance of a failure, and consequences of a failure. You don't want a flat battery to be a death penalty. $\endgroup$
    – Koyovis
    Jan 26, 2018 at 20:52

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .