The upcoming 3.5.0 release of APM:Plane includes support for QuadPlane - a hybrid plane/multi-rotor that allows for high speed long distance flight with vertical takeoff and landing.
The above video shows a converted HobbyKing Firstar 2000 doing a fully autonomous mission, with VTOL takeoff, followed by automatic transition to fixed wing flight and then a VTOL landing. The planes builder, Jack Pittar from CanberraUAV, is shown with the plane below.
The plane itself was fairly easy to build, with the simple additional of rectangular section aluminium arms on the Firstar wings and 4 quad motors.
One issue we found is that the battery we are using (a 3S 4Ah 65C) doesn't handle the 65A needed for vertical takeoff well, with the voltage dropping to 10V on takeoff. After takeoff completes and it transitions to fixed wing flight it quickly recovers. We will be changing the battery setup for future flights.
A second issue is that the wings tend to twist a bit in flight, especially when yaw is commanded. That greatly reduces yaw authority when hovering. We are still thinking about the best ways to prevent wing twist.
The new QuadPlane support is documented at http://plane.ardupilot.com/wiki/quadplane-support and is included in the 3.5.0beta1 release (I will do a separate blog post on that release shortly).
In the near future we expect to add support for other hybrid frame types, including tilt-rotors and fixed wing aircraft with other types of multi-rotor frames attached.
This QuadPlane follows on from our earlier Senior Telemaster QuadPlane which used two Pixhawk flight controllers.
We built the Firstar QuadPlane to test the new code where a single flight controller board running APM:Plane is used, which makes for much simpler operation. We plan on now building a much larger version, with a 50cc petrol motor for fixed wing flight.
Comments
Thx for the feedback Tridge.
Would you think a dual airpeed sensors would help determine wind direction? Maybe it would suffice to monitor wing airflow, for example at each of the wingtips to determine wind direction? It might not be wind direction but it might be good enough to determine lift due to wind to point the wing and rudder to the right heading?
I'd imagine that it's only necessary to control quad yaw at lower wind velocities where the rudder doesn't produce much force. I think that if one uses the "hovercraft" flight principle I mentioned above, the feathering into the wind will occur by itself due to the rudder surface being pushed by a strong enough wind.
The question is how does the quad respond to rudder induced yaw by the wind, and will it try to fight it using the quad motors to maintain it's own heading? Is it possible to curtail the quads yaw control and let the rudder do the wind direction control without the need for sensors? I'm thinking that if the wind is weak it doesn't matter that much if it is not pointing into the wind and when it is strong the rudder will feather it accurately by itself if the quad motor yaw control allows it.
--
@Gary
I think both tilt-rotor and tail sitters are still some way off, but in the works on Arduplane, and can be tried on PX4 code already. I'm mostly interested in the quadplane config atm but there are some hybrid designs with lifting bodies that interest me though! ;-)
@tridge Have you considered the weather vane approach for NAV_CMD_VTOL_TAKEOFF too? even to the point where it transitions into the wind and only then flies to the next waypoint?
I have a Ruby and that autopilot will hover in strong winds, its amazing. It determines that a loiter circle will get messy so holds the nose into wind and raises and lowers the nose as required to hold position. Looks just like a bird of prey. It would be very interesting to measure the power used in 'assisted hover'
Is there any movement towards tail sitting modes? I realize this method and rotating motors allows the autopilot to stay in the same orientation so is in some ways easier...
@JB and Iskess,
The QHOVER mode will actually weather vane now. With no stick input what it does is try to maintain zero yaw rate. It doesn't try really hard though, and in testing (both in simulators and a real quadplane) it does tend to weathervane. If you find this isn't sufficient in a real flight then please provide a log if you have one and I can have a look at how it could be improved.
Note that basing it on the estimated wind direction would only work if the plane had been doing sufficient forward flight beforehand to build up an estimate of the wind direction. The wind direction is not actually properly observable while holding in a hover. You could try to infer it from the way the plane is drifting, but that will tend to be quite inaccurate for a quadplane except at high wind speed because quadplane aerodynamics don't tend to be very symmetric. They naturally tend to drift a bit.
More interesting I think is the landing transition for NAV_CMD_VTOL_LAND. I originally made that try to hold the heading from the landing course, because I thought that is what users would want (if you plan a landing on an east-west course I thought it would be good to hold that heading so the plane lands the way the landing is setup). I found that as I upped the wind in the simulator that the quadplane became quite unstable in the landing if it was a cross-wind, so I changed it to use the same system as QHOVER, and it now weathervanes the tail around so it is pointing into the wind for the landing (if there is sufficient wind to overcome the yaw resistance). That works really nicely and landings in high wind are quite stable.
Cheers, Tridge
I second the weather vane mode and have mentioned it elsewhere before. I think Tridge is fairly busy though! ;-)
With a quadplane it consumes a fair bit of power to maintain position and heading out of the wind, unlike a standard quad without wings, the quadplane has to fight any lift the wings and rudder are producing whilst the quad motors tilt the aircraft. The issue is that tilting the quadplane in any direction like a normal quad, even forwards, results in negative lift being produced by the exposed wing surfaces, which in turn requires even more power from the quad motors to maintain altitude. The plane starts fighting the quad in a regressive loop.
The other advantage of pointing the aircraft into the wind is that the plane motor can be used to counter the wind to give lateral positional control in exchange for a small amount of power, requiring the quad motors only to provide enough lift to maintain altitude. In fact using the plane motor results in some extra lift generated by the wings which in turn reduces the load on the quad motors, depending on the wind.
Ideally, if the wind is significant enough to produce some lift from the wing surfaces (ie above a setpoint/level where the quad power consumption is consuming more to hold position out of the wind in hover), the weather vane mode would use the quad motors to point the nose into the wind and possibly even increase the angle of attack of the wings to produce more lift in a hybrid hover mode like a bird of prey. I think this will considerably improve quadplane hovering in windy conditions, but is unlikely to be that helpful at low wind speeds, say under 4-5m/s.
There are only limited reasons to maintain heading in hover, typically on a quad this is because one wishes to point a camera in a particular direction. On a quadplane using a camera on a turret would likely result in better quadplane hover performance, without impeding camera view angles..
I think in principle the quadplane should be operated as a airborne "hovercraft with unlimited altitude", rather than be flown like a quad.
That means that the quad never does anything more than maintain altitude at a level attitude, and every horizontal axis movement command is translated into the aircraft yawing to that heading and applying the forward plane motor to move laterally. The plane motor could even potentially be used in reverse thrust as well for example to counteract small gusts. Being able to calculate exact realtime wind vector would likely result in the most efficient use of available lift components to achieve the desired pathway. Maybe differential dual airspeed sensors could work? I'd like to see a single multidirectional ultrasonic airspeed sensor one day!
Originally I proposed that the plane and quad modes do not actually "transition" from one to another per se, rather that the user can "tune" what level of control is applied to whichever control mechanism, be that quad motors or plane surfaces and horizontal motor, which leaves the possibility for lifting body type hybrid quadplanes with minimal wing surface areas and resulting low drag. Imagine a lifting body design with tiny high stall "winglets" that can VTOL with minimal quad components. In a new breed of quad racing one could gain a very substantial amount of speed similar to a pylon racer, but still have the low speed high-G cornering ability of a quad! :-)
Overall I think a quadplane does not have to adhere to conventional flight "physics" in the sense that it has to fly as either a plane or a quad, rather that it can be designed to seamlessly fly using both, depending on the lift state of it's individual components.
Anyways that's my two cents.
Regards
JB
How do you feel about changing the QHOVER so it doesn't try to hold heading, and instead simply weather vanes into the wind so the quad isn't fighting the tail. Or even better, use the wind direction calculation to actively maintain a heading into the wind.
Hello Tridge
Thank you for your quick response.
if I understand correctly I replace the Flight Mode in the parameter list.
I use Mission Planner 1.3.34.2,
where can i da insert the commands?
greeting
Rainer
@Daniel, I don't think tiltrotor will be very difficult but I haven't had time to work on it yet. I'll certainly announce when it is done, or if you are interested in helping with the code then let me know and I can give you some pointers.
-
4
-
5
-
6
-
7
-
8
of 10 Next