Yesterday Was the first successful transition flight of my Nano Quadplane. After limited success a few years ago with two separate pixhawks on the VTOL Anaconda, I decided it wasn't feasible to use multiple Pixhawks/Pilots for anything other than a tech demo.
Now with all of the work done by the Devs for creating the quadplane code (awesome work gents!), I was inspired once again to make something that worked as seamlessly as imagined back then. I knew there would be challenges and a large margin for failure, so I chose a small EPP plane to minimize damage and rebuilding. Turns out I don't think I've ever seen a quadplane that uses Arduplane/pixhawk this small, so I thought I would share this with everyone. It took about a week to make and probably took it apart and back together no less than 10 times...seriously there is so much crammed in the fuselage I don't think there is room for air.
In this flight I was running Arduplane 3.7, and have yet to test some of the advanced features like weathervaning and pusher motor position hold assist. Using old batteries so I can only hover for about one minute, so no time for autotune. All gains were done manually the old fashioned way, though they still aren't perfect. Now that I am confident and familiar with the code, I will try again on a much larger model (2 meter+ wingspan) with decent flight time and payload carrying ability.
Nano VTOL Full Specifications-
AUW: 610g
Endurance with VTOL Takeoff/Landing: 12 Minutes
Autopilot: Pixhawk 1 with GPS, 900mhz Telemetry, & Power Module
Firmware: Arduplane 3.7.0
Airframe: RMRC Skyhunter Nano RTF
Battery: Turnigy Nano-tech 4s 850mah 45C
Pusher Motor: Multistar 2206-2150kv
Pusher ESC: 3DR 20A w BEC
Pusher Prop: 5x5 Tri-Blade
Lift Motors: Emax 1306 4000kv
Lift ESCs: DYS 20A
Lift Props: 3x3 Quad Blade
FPV: RMRC 25mw 5.8ghz all in one Tx + Camera
Also if anyone would like the file to 3D print the booms send me an email: Joseph@suavx.com
Thanks again to the Devs and Testers who made this possible-
Comments
Great discussion here guys!
Just wanted to add the arm files for those who wanted to try this themselves:
Pascal, you are right. But it's still much, much, more efficient than carrying a 4 death weights. Plus, the props in horizontal fight act as an additional aerodynamic load. Not to mention 4 ESCs, 4x3 heavy wires and so on. Regarding the pitch of the propeller,... well ALL (drone) propellers are designed for let's say one "condition" as soon you change the speed of the propeller is already outside the designated optima envelope. Another point that works for the tilting motors design is the sweet spot of BLDCs. Usually BLDC works best (most efficient) at it's 50% capacity or less. And since in the aeroplane mode the wings do provide a lot of help, motors can run in their optimal thrust/consumption curve.
My additional few cents :-)
Nice design, Dorjano, however, this solution is not very efficient either, as you're still using 4 motors for propulsion in airplane mode. Just think about the power used during this airplane mode. Furthermore, the propellers used (size/speed/pitch) for hovering are not at all optimized for airplane mode. AND, instead of carrying one additional propulsion motor of let's say 50- 70 grams, you're including two additional servos plus the mechanics. What's the gain..?
Regarding the control of the displayed Bixler Quadplane, it seems that the transitional mode is just very basic switching from one state to the other, resulting in the (as seen in the video) attitude and altitude fluctuations of the plane.
Just my few cents...
Hi Joseph
One question though.... What are the four motors doing during a horizontal flight? In a aeroplane mode. Let me guess... nothing. A death weight, right?
This configuration is no good. You should use the same propulsion for VTOL and for horizontal flight. Though there is a better way/firmware - use tilting of the propulsion system. Check this video to see what I mean: https://www.youtube.com/watch?v=e7QNbU71lsA
Dave - I'd say the XUAV Talon isn't the best airframe for a quadplane, it will do it, but it is a pig of a thing and puts any good sized props where they get interfered with by the elevators. I don't think it is ideal either for bracing the wings to prevent twist.
Joseph - great work, I was just looking at that airframe the other day - maybe now I'll pull the trigger
Great job and prove of the potential of the current autopilots. Congrats!
Mario- Thank you!
Galaeta- There were no mods to the wing, looks just like a stock nano skyhunter but with a bunch of extra wires. The quad booms were glued on with shoe goop type glue.
Dave- Yes I have an old v1 talon that I've always wanted to convert. Never had the time either. The problem with the cg in the quadplane is that it needs to be correct for both modes of flight. If it were nose heavy, it would be ok in airplane mode but in quad mode there would be an imbalance in the power output to maintain level between the fore and aft lift motors.
Good work. I like it!. May you kindly post awing reinforcement photo? How about the wing root twist?
Gtz.
Thanks Tridge! Let me know if you have any experimental changes for the code, that's what this thing is for! It crashes really well so no worries if something doesn't work as expected. I certainly proved that with my regular fixed wing nano skyhunter. This thing is pretty tough-
-
1
-
2
of 2 Next