Semi-Scale VTOL V-22 Osprey

I started this project one and a half years ago, when I built the V22 version 1. It was great at hovering, however in fast forward flight (FFF) it flew terrible. But now, after a lot of tweaking and experimenting, the version 2 has become a very nice plane to fly! It is a great compromise between "scaleness" and simplicity. And it's cheap to build too!

Overview

The construction is as I said rather simplistic: it is built out of 6 and 3mm depron and a single wooden spar. The wing is of one piece and can be rotated 90° around the top-axis (as the real V-22 does on aircraft carriers) for easy transport and storage.

The motors are mounted on a simple tilt mechanism (like the one from David's tricopter v2, but modified for >90° rotation), each of which are driven by one servo. The flightcontroller is a KK2.1.5 flashed with OpenAeroVTOL.

The wings have a Jedelsky-inspired airfoil: the front part is made of 6mm ribs and 3mm cover, and the back is simply one 6mm piece that is also the flaperon. I have chosen this construction to make 90° downward deflection possible.

That's all very nice, but how does it actually work?

In hover flight mode (HF) it's controlled just like any helicopter or multirotor, the flightcontroller does all the mixing for you.

• To go forward it tilts both motors forward
• To yaw it tilts them in opposite direction
• To roll it will use differential thrust (increase the thrust of one motor and decrease that of the other)

Then when you engage slow forward flight mode (SFF), both motors will gradually tilt forwards and the flaperons will slowly go up. They will stop at about 50% inbetween HF and FFF.

In SFF it's literally a mix of a bicopter and a conventional airplane, it will fly at about half the speed of that of FFF. Thus the flightcontroller will halve the control throws it had in HF, and add 50% of those it will use in FFF.

Then when you transition to FFF, the motors will tilt to their most forward position and you'll fly the V22 like any normal airplane. Only, because it has no rudder, it uses differential thrust to yaw.

The build

Specs

• Wingspan: 106,5cm
• Length: 115cm
• Scale: ± 1:15
• Fuselage weight: 243g (incl. one servo)
• Wing set weight: 715g (incl. motors + electronics)
• Battery weight: 191g
• Total: 1149g

• Motors: Hextronik DT750 (not the ideal choice - KV too low)
• Tilt servo’s: Aerostar AS-170MG modded with two 2.2KΩ resistors for bigger throw
• ESC: HobbyKing 20A
• Prop: 10×4.5 SF
• Battery: Turnigy 2200mAh 3S 25C
• Other servo’s: HXT900
• Flight controller: KK2.1.5

The plans

The build plans can be downloaded off of my website. 

There are no build instructions, but it should be “figureoutable”. And if you need any more clarifications or detail photographs, just leave a comment here or on my website! (A lot of questions have already been answered on my website). Note that this is not meant to be a build-and-fly model. It’s more of a buid-and-experiment-tune-and-maybe-fly model.

OpenAeroVTOL settings

I know I'll be asked about this. They're on my website (I did not feel like duplicating all the tables etc).

Flying characteristics

Hovering: It is quite stable and flyable in hover mode. However, sometimes it will suddenly point the nose vertically downwards and starts falling. With enough height, it will recover automatically but I've already had a couple of close calls. I'm not yet sure what causes this. Maybe it’s the PID’s, maybe it’s actually some sort of a stall, or maybe it’s preventable. I’m not sure. It, by the way, does not like to fly backwards or with the tail in the wind. That’s because the horizontal stabilizer will then act like a horizontal destabilizer.

Transitioning: The transition to fast forward flight is very uneventful, you have more than enough control authority the whole time.

Slow Forward Flight: The Osprey actually flies best in between hover flight and fast forward flight. It feels very responsive and predictable.

Forward flight: It has plenty of power during hover flight, yet it feels very underpowered in forward flight. This is because the DT750 (with the propeller I use) has a low maximum airspeed. This can only be fixed by using another motor setup (which I don't want to spend my money on). It's however stable enough for me to try some low-level maneuvering - that's something I could never have done with the first version.

So far, there are 6 flight video’s, 3 of which I've put below.

Conclusion

After a lot of tweaking and practice, it has become a very nice plane to fly. And because it fits in our car in 1 piece, it’s also very fast to set up at the flying field. I first thought this would never be more than an hard-to-fly experimental plane, but it now is a favourite of mine.

Its only real shortcomings are the low KV motors and the unpredictable nose-down tendency in HF. The latter could probably be fixed with a movable-battery, like the first version had. Also, scale motor pods would be really nice. Though these are things for antother time.

Note: This is a summary of the full article on Hangar42. There's also a RCGropus thread.