Steering with gyros

[tia20]

Hi there,

I have had an idea regarding model gliders being controlled by internal gyros rathern control surfaces: my idea is to mount gyros (the ones which actually have roting discs in them) on servos and rotate them about an axis perpendicular to the axis of rotation of the discs inside the gyro. by conservation of angular momentum one should be able to control roll about any axis of the plane without moving any control surfaces.

what does everyone think. does it make sense. has anyone done anything similar? will I be able to induce great enough a torque on the plane by rotating the the axis of the gyros?

cheers

iver

[Sparky Paul]

How large must the rotor be?
How is the torque on the rotor axis transmitted to the airframe?
What keeps the rotor spinning?
How much more than a normal r/c system would this weigh?

[tia20]

those are the questions I am asking myself too.
will do some rough calculations later today and let you know.

Iver

[green66]

Hi tia20,

The physics of the concept are valid, and it's used in low-inertia (space) applications but, at first glance, I think it'll be difficult / unrealistic to employ in a glider because of the magnitudes involved.

If you're trying to eliminate the drag of deflected control surfaces, eliminating ailerons would probably yield the greatest benefit, since they not only create drag, but also mess up the airfoil profile every time they're used. Problem is that a glider carries high inertia in its roll and yaw axes due to the relatively large span, thus will require high torque which means a large change in angular momentum must be generated to have a meaningful effect.

Since ang momentum is the product of the rotor's polar mass moment of inertia and its rpm, either the rotor would have to be low mass but large in diameter, or small but heavy, or spinning very fast. Large things don't fit in airplanes, heavy things are bad for airplanes, and spinning things require energy = motor, battery, controller etc = weight / complexity. And if the rotor isn't spinning constantly, i.e. just during control inputs, it would require a powerful motor and lots of current to spin it up to speed quickly :( (which would also cause a torque along an undesired axis).

For a glider, the concept would probably be more workable for controlling the pitch axis, especially for high aspect ratios which are more pitch-sensitive, but that application is also a stretch, esp since a torque would have to be constantly generated to overcome the wing's pitching moment. If the main benefit of an inertial system is to reduce the drag of a control surface, the gains will likely be small when considering that a) control hingelines can be easily sealed, b) all-flying (tail) surfaces have no hingeline, and c) the drag only occurs during a control input, which is relatively infrequent and subtle for gliders.

You could test a prototype system w/o flying (crashing) with a plane suspended by string at its CG. You'll see an effect, but I believe the control response from devices small and light enough for a glider will be too sluggish to be practical.

[tia20]

Hi Green66

thanks for your input. Many of the issues you mentioned I had already considered if not in such a clear cut way.

I will have to do some calculations in order to make out how adverse the effects mentioned will actually be. I will let you know about my outcome and how I got there.

any idea of what kind of commercially available gyro I could use? for starters I was thinking of using an electric motor with a disc attached to it and oriented by servos.

cheers

Iver

[green66]

any idea of what kind of commercially available gyro I could use? for starters I was thinking of using an electric motor with a disc attached to it and oriented by servos Yeah that's about the simplest/cheapest way to proto the thing and get some physical sense of what's going on. R/C boat/car/heli shops will have flywheels that might be useable as a rotor if you don't want to make one.

[kallend]

It won't violate the 1st or 2nd Laws of Thermodynamics.

Which means that since you're not using work done by gravity (glider) or the powerplant (powered plane) to rotate about the various axes, the work will all have to come from the system powering the gyros.

Have you calculated the work needed to, say, roll a typical glider through 90 degrees?

[Salto]

Actually, the work comes from the servos moving the gyros, not from the motors spinning the gyros.

Unlike conventional wings and tails with control surfaces which in combination act as "force amplifiers", the gyro system will have to provide the entire control force.

Interesting project.

Graham.

[tia20]

?Hi there,

I finally did some "back of the envelope" calculations regarding the sizing of the gyro. The steps I went through are as follows:

Angular momentum = I * Omega

where omega is the anular velocity of the gyro.

The torque exerted by the gyro being turned at thetadot perpendicular to its axis of turning is:

T = thetadot * I * omega

The torque required to roll the airplane I assumed (on the basis of a standard F3B glider) to be 10N (the total lift generated by the wings is approx 20N. When doing a turn the outer half of the wing, on the outside and inside of the turn, gain and loose 5N respectively. since they are 2m apart - on average - the torque required is 10N for a relatively sharp turn.)

Now rearranging:

I = T / thetadot / omega

as practical values I assumed omega = 10000 revs/min i.e. approx 1000rad/s and thetadot 10rad/s (roughly 2 revolutions per second).
Hence omega and T cancel.
Therefore

I = 1/omega

Since I = mk^2

m = 1/omega/k^2

for a round lamina k^2 = r^2 /2

hence:

m = 1/ 500/ r^2

a pracical r would be 5cm. hence

m = 0.8kg

well, that pretty much shows that a gyro is not going to be a practical means of inducing roll.

Note: The installation of one gyro would make the installation of two more necessary since yaw moments would be generated on roll, due to a gust for example, while roll forces would be induced when yawing. depending on the orientation of the gyro pitch could also be coupled.

the overall weight of such an installation would then be of the order of about 3-5kg depending on the exact power consumption and appropriate batteries.

oh well, shame that, I thought I might be onto something. but then there is always hope for high end technology. If one used a carbon fibre wheel at say 10 000 revs/min and a maximum thetadot of say 50rad/s (10rev/s) the weigth (negelcting motors) would be reduced by a factor of 50.

sorry for the bad equations. guess I cant post a latex pdf file ;-)

thank you all for your inputs

Iver

[tia20]

just to answer the point about the work:

again, assuming the required moment to roll the aircraft is 10N and a roll of 90degrees:

Work = torque * angle = 10 * pi /2 ~15J

remember, a one volt battery of 1Ah stores 1*60^2 J = 3600J
hence, a 10volt pack, with 1000mAh (1Ah) would have 36 000J

have I done a mistake?

I think the above shows that the main power consumtion is going to be due to friction in the bearings of the gyro, rather than the servos....

anyone got an idea what a typical electrical rating of a, say, 0.1kg gyro would be?

Iver

[kallend]

just to answer the point about the work:

again, assuming the required moment to roll the aircraft is 10N and a roll of 90degrees:

Work = torque * angle = 10 * pi /2 ~15J

remember, a one volt battery of 1Ah stores 1*60^2 J = 3600J
hence, a 10volt pack, with 1000mAh (1Ah) would have 36 000J

have I done a mistake?

I think the above shows that the main power consumtion is going to be due to friction in the bearings of the gyro, rather than the servos....

anyone got an idea what a typical electrical rating of a, say, 0.1kg gyro would be?

Iver

I think you may underestimate the work required. most of it will come from aerodynamic forces. For example, the upgoing wing sees a reduced AOA, which tends to oppose its motion, while the downgoing wing sees an increased AOA, which also opposes its motion. The AOA change will vary inversely as the airspeed for a given roll rate, but the force generated will vary as airspeed^2, so overall the force needed will vary as airspeed.

Ditto with pitch for a plane with CG in front of the neutral point.

A system like this works fine for spacecraft, but they don't have aerodynamic forces to contend with.

[coaxial-gyro]

I'm led to believe a German bus works within these methods , with a great big gyro disk ! But that was uncovered at the pub , so does anyone know more ?! were all intreasted .
I've always understood the gyro qulitys need more attention then the one off .....,
( Ooh ,..then my one year old duaghter spoke. )
shall explain the world off tommorow . buy.

[Salto]

I think you'll find the bus uses a gyroscope to store energy to be used for accelleration, allowing the electric or IC engine to run at more constant load, and therefore more efficiently.

Graham.

[AustinTatious]

Actually, the work comes from the servos moving the gyros, not from the motors spinning the gyros.

Unlike conventional wings and tails with control surfaces which in combination act as "force amplifiers", the gyro system will have to provide the entire control force.

Interesting project.

Graham.


Yeap.. think abotu that for a second... the gyro will not try to change the orientation of the plane with any more force than the servo is applying to the Gyroscopic disk.

TO get a rough idea of how effective this would be, try using one servo to roll your whole plane back and forth!

Also, you CAN overpower gyroscpic precession. If you are applying GOBS of force to the gyro, you may exceed what the gyro can deal with.. the answer woudl be to spin the disk faster and/or make them heavier.


I think things like Wingarons and full flying stabs is a much better way to reduce control drag.