freeflysystems.com/careers Follow rctestflight on Instagram: / rctestflight_ Support rctestflight: / danielriley Join my discord server to talk about building stuff: / discord
The wings should have their center of drag in line with the center of mass of the craft to avoid introducing the pitch moment that caused the oscillations. This is tough because the wings would have to be mounted even farther outboard to avoid colliding with the rotors. If the front/rear motor pairs were moved more frontward/rearward instead, clearance for the wing could be introduced without elongating it and the available pitch moment from the motors pair would be increased, allowing it to more effectively counteract any pitch-wise disturbance.
that plus these wings are dead they have no control surfaces to combat aerodynamic forces like you would have in any real or model rc aircraft, adding this would likely fix it. ditch the homemade wing and get one for a rc plane that has all the control surfaces built in and intergrate it into the drone so that its software begins operating them to keep it in line with its stabilization programing. Also there's another way to improve efficiency , by changing out the rotors for the torodial design they not only improve flight BUT are near whisper quiet making the drone hardly noticable
Yes. It should have neutral static stability so the rotors have authority. I.e. the aero body must be orientation agnostic with respect to the relative wind and generate no torques about the cg. I don't know how close one can come to this in reality. None of these designs ever come close though because they don't even try.
The other thing to consider is the fact that, not only is the wing causing a pitch moment due to induced forces, the center of mass is now shifted higher if not above the motors creating an unstable system. Even just adding a ballast below the quad could theoretically make the motor control play a little nicer.
I had a thought in a slightly different direction. How about 2 smaller wings, one in front of the front motors and staggered vertically but forward and out of the prop wash path, and the rear staggered slightly below the vertical rotation axis of the rear motors, behind the rear motors, again out of the wash path. So front and rear wings. This would of course require a landing gear to prevent awkward takeoff. But with the same pivoting wing with rotation stops would certainly have fair upward lift without the interference with prop flow paths. It should also equalize the load decrease across the motors..
The wing should be placed in the same way as it would be on an airplane, at the center of mass or further back. to me it looks like the quad is fighting against the instability of its back heavy nature. "A front heavy plane flies badly, a back heavy plane flies once" - Some kerbal space program enjoyer
For years we have been privately testing this exact concept with over 300 test flights on DJI and other drones and we filed a patent in 2019. All the issues mentioned in the comments have been resolved and patented in US Patent US20210107637A1 "Universally attachable hinged wing and vlos aid for multirotor drones". We are scheduled for the release of our product line soon.
Move the wings down to the center of mass of the drone, to avoid having it add a speed-pitch feedback making it unstable. Control the wing AoA with a servo so that it's always correctly oriented relative the direction of travel through the air. Perhaps add a few degrees of dihedral to the wings to get a bit more stability. Use an airfoil section optimized for the very low speeds seen here, something appropriate for maybe 1-3 million Re.
The "dihedral=stability" equation is not so clear cut. the way dihedral works is by rolling the plane in the direction opposite sideslip at positive AOA, but with this arrangement that just adds more complexity for the control system to fight against.
All the issues mentioned have been resolved and patented in US Patent US20210107637A1 "Universally attachable hinged wing and vlos aid for multirotor drones". We are scheduled for the release of our product line soon.
At Warwick university I saw a drone/fixed wing hybrid that had a stationary delta wing which was vertical when on the ground and then the drone completely turned horizontal during free flight thus turning into a fixed wing plane with a rotor above and below each side of the wing. I imagine this required custom software to calculate the best angle of attack for different speeds.
The RC aircraft X-Vert seems to be what you are describing, but with only a pair of motor. It uses pre-programmed maneuvers to transition to fixed wing and back again, while also changing the flight controls to be the standard for each.
Tailsitters are efficient for long range, although as a camera drone the body interferes with the downward field of view at low speeds or hovering, and cannot fly sideways at higher speeds. The concept in this video has been patented in US patent US20210107637A1 "Universally attachable hinged wing and vlos aid for multirotor drones" and we are scheduled for the release of our product line soon.
Maybe have the rod going through the wings be closer to the center of mass of the wing so it's easier to reach the edge case. This way, the drone uses less energy to keep the wing up
With a properly tuned flight controler taking the wing into account, I think your preliminary data shows the wing it might have a significant efficiency gain in forward flight. I think for certain mission profiles you're onto something. As always great video Thanks
agreed. i have been watching his videos for years and its always fascinating yet informative along with his uniquely deep voice being chill to listen to lol
rctestflight your research is inspiring and see where you are going. Based on your RPM data showing distinct differences between 4 rotors perhaps prudent to try positioning 4 separate wings behind the thrustline (beneath) of each rotor. So as each rotor thrust will behave independently and observe any change in stall threads attached on low pressure surface of each wing in full throttle forward cruise. In turn cumulative lift may show where lift disparities exist. Good job of messing around so much with this, excellent!
I'd really like to see this experiment repeated with the anomalous motor issue fixed. Like you said having one motor completely idled is really going to mess up the quads ability to stabilize itself and it might be a lot better without that problem present.
How about the sort of angle-of-attack control tail/elevator thing you used on your rigid wing sailboat? I think that would be a lot smoother than the hard stop on the wing angle.
Or an tuned progressive elastic resistance to the AOA? Like 3 different surgical tubing lengths in parallel, that way the AOA would be a function of the airspeed. You could still have a hard stop and the period of the oscillations may be longer as well. The tuning might be tricky, but maybe with force meter measurements on a hard stop for several AOA/airspeed combos you could tune the elastic dampening. Great video!
No way bud. The wing should be centered on the CG. Any aerodynamic stability, positive or negative, is going to fight the pid controllers. Aerodynamic stall is desirable in some cases and a normal part of the maneuvering flight regime. An AoA sensor should feed the FC to inform cruise efficiency. Get all these nasty aerodynamic flipperon things out of here. This is a quadcopter!
What if you put the wing fixed perpendicular to the rotors and in flight go full throttle and angle the quadcopter 90° forward and use it as a conventional aircraft. But then basically you've converted it into a VTOL
Very interesting, even to a non mechanical/non engineer amateur drone enthusiast like me. I subscribed after watching your first Dyson vacuum engine powered airplane flight . The more I watch, the more I'm hooked on watching your RU-vid videos. You're a very smart young whipper snapper. And as I watch this video, it makes me proud to be from the Pacific Northwest. ( grew up in Bellevue) but live in downtown Seattle now. Keep up the good work!
A tail-sitter craft with four props in a quadcopter orientation, can easily make the transition to forward flight. The camera gimbal can be held on a swinging (pitch) axis. Keeping the entire camera assembly below the craft as it transitions back and forth between forward flight.
I love the idea and it apparently works a bit, extend the wings out beyond the rotors and mount it to the bottom of the frame of the unit. maybe make the size of the wings about 1/4 of the size that you have on there now. just a thought from my head.
I'd love to see you push this idea a bit further. an idea I had would be to have 2 wings across with 2 rotors set into each wing. That might give you more stability. in any case, love the vids and cant wait for the next one!
Finally! I've been periodical searching online to see someone do this... I've been thinking of it as an attachment to dji drones that already have long flight times!
Great idea but I think it would need a rewrite to the controller. Combining what Zipline does with being able to take off and land anywhere would be great
The drone is missing some kind of more effective pitch control like in standard airplanes, i think this is causing the oscillations. Also fins below the rotors maybe would prevent the rotation of the airflow, which could boost efficieny as well
I agree. I think the flight controller program would need to be designed for this hybrid design and also include servo control for wing pitch, and then it should be able to effectively counter the instability.
Great video, thanks a lot for this story, and for the already thorough analysis. We have similar ideas, that we should try in the next months, I'm looking forward to being able to report on our own progress.
Late to the party, but I think the performance would be better with the wing mounted just below the props instead of above them. It'll also be a more compact overall package.
the down wash of the props would keep pushing the wing down aswell as messing with the airflow over the wing so idk if it would be better, would like to see it tested still
Here's my silly idea: Instead of large solid foam wings, use small spools of fabric and pairs of linear tracks to create the airfoil. Winches made with continuous servos will be used to extend the fabric, while the spool will be tensioned back by coil springs. This could provide more adjustments besides rotation, as there's total control over the airfoil's surface area. For example, to retract it partially or entirely when hovering or in turbulence. Wait a minute, did I just reinvent sails and repurpose it as a motorized kite...? 🤔
I actually made something a lot like this a few years ago, but for a 4' quad style base, and rather than 1 wing in the middle, I had 2 wings front and back. They were automatically put level with 1 servo that was linked to both of them. It kinda worked for the lift aspect but the control algorithm had a ton of trouble with it. I never ended up getting a chance to fix that though.
You essentially made a canard aircraft without the rest of the air frame and gave it some alternative form of P thrust. Thinking about how it all interacts is making my head spin, so keep going and make more iterations!
For a future video, try mounting the wing below the rotors, this may seem more unstable but it's highlighted to be just as viable as above aerodynamically by the drone pendulum fallacy. Having the wings below will allow them to rotate without the consequence of hitting the rotors, potentially allowing you to bring the CoL closer to the CoG (although, I'm not sure whether the downwash from the rotors will impact this).
Nice work. Look at the scorpion/stargazer freewing UAV and design your wing the same way. Use a pitch neutral airfoil (flying wing airfoil) and you can remove the stops and have the wing self align regardless of airspeed. Aurora flight sciences also had a vtol with this idea many years ago
I am currently completing my senior design and my project is a fixed wing (piper cub style) with vertical take-off and landing capability and a forward pulling traditional style motor. This is interesting because it's such a different approach from the previously mentioned Warwick university example. I would see this design being helpful at very slow speed but testing this to a slow flying fixed wing VTOL would be interesting to see what design is more efficient. could be useful for package delivery applications... maybe... as you mentioned high wind would be extremely detrimental to this design.
Cool idea. We went through this with the UAVforge spy drone project. Fixed wing VTOL planes do exist. Motors are mounted on the wings, but the whole wing can rotate so it effectively becomes a bi-copter.
The way you mounted the wing high above the quad puts the center of lift ahead of the CG when the quad pitches to translate forward. Thats almost certainly the main cause of your pitch instability in fast flight. The front motors are decreasing in speed to the point of idling because pitch moment from the forward positioned wing is so large it's taking all the load off them.
I like the approach as I've made similar concepts some time ago. I used servo-like stabilized AoA of the wing to maintain constant angle regardless copter pitch and current sensor to get real power consumption. I notified significant gain in efficiency, however I dealt a lot on control algorithms as copter's autopilots does not like a wing to be attached and provide lift. It is bit misleading to autopilot, you increase the forward speed and you generate lift so it needs to decrease the thrust having less authority to control pitch, roll and yaw.
Love it!!! one thing we didn't see was the wind direction, which could have explained the no thrust motor. and as other said, think the wings need to be closer to the center of mass, or maybe double them up as a large biplane?
Agreed with all the comments about the CG vs center of drag, but something else that might be worth trying is no stop on the AoA and then some kind of tail boom and stabilizer attached to the wing in the same way those drone sailboats get the perfect AoA on their sail
Whenever I move to Washington in the future I hope I get to see your inventions getting tested around Seattle. How fun would that be to see some device and think "oh that's just rctestflight"
The best example to take inspiration from for fixed wing efficiency & Quadcopter type VTOL (vertical take-off & landing ) performance will be a US made military aircraft called V-22 osprey . Hope some makes a RC drone based on it
Great stuff.i would like to see it done on a 7" long range quad. With iNav. Also maybe add wing pitch controlin flight to speed up optimizing wing angle.
I did not expect that! That is really interesting! What if you put the wing further back and attached a rear wing with a stabilizer? It is like a whole plane attached but with a fixed angle. I think that would be more efficient and stable with it attached. Please correct me if I am wrong.
Build one you can fly yourself around, everyone seems to place the pilot at or above the rotors, on a hard landing the pilot could be thrown down onto the rotors, and if the pilot is at rotor level a blade could snap and impale them, i'd like build my own with a fuel engine with two counter rotating props for the main lift, and four battery rotors for control. really enjoy your videos
Check out the British designed Fairey Rotodyne from the late 1950s. A single rotor was used for take off and landing and which was then then put into auto-rotation during level flight. Lift also came from 2 short span conventional wings where the propeller engines were mounted. The system worked well but the biggest con was the noise made by the rotor tip jets on take of and landing. Not a good point as the aircraft was intended to be operated from city centres. There was a hush kit in design when the project was cancelled but by then it was too late to get any more government funding :(
Very cool video. Very informative. Do you think if you elongated the front/back arms for the rotors (leaving a gap at the body) and a narrower wing attached to each side in line with the height of the rotors would alleviate the oscillating? With the setup you used for this experiment did you determine the flight time or distance available for this drone?
As the craft rotates forward, your wing is moving ahead of the CG and thus creating a pitch-up moment. If you set it further back, you would do better at higher speed (but worse at lower). Ideally you would have a way to move it back in response to the crafts pitch-angle to keep the center of lift closer to the CG.
Could you use glider wings pivoting at the center of mass of the rotorcraft? You might be able to make those clearances work with the downside of being much wider.
Thanks for the interesting test. As already mentioned in other comments, the wing position was not close enough to the center of gravity for that flight phase. My assumption is that you could save the 24% much easier by using a more efficient propulsion concept: Larger propellers and adapted motors (lower KV, more torque). This also increases the area exposed to the wind and decreases stability in rough conditions - but much less than with wings added as in your example.
i think this is the application of tilt rotor aircraft. Being able to rotate the propellers forward allows the payload to stay in the same orientation while flying vertically and laterally.
Thanks for another great video! I was wondering, have you ever thought about giving a drone a body that acts as a lifting body? Or any other shape that generates lift?
Great video! How about an H frame with a small winglet on each arm pulled close to the body? Not sure you could get them out of the downdraft of the motors though.
There is a profesional drone with a fixed wing simular to this setup. But the wing is just behind center of mass (as it would be in a normal aircraft. So the motors can just thrust and do normal adjustment. But of cause. It does.have a set of software for hovering and one for forward flight. Its not only considerally faster in wing mode, it can also fly for twice as long, increasing the range quite a lot.
I'd be interested to see how it performs with the wings starting outside of the rotor area so that there would be less interference. As someone else mentioned: it would be good to have them lower down as well, although I understand the challenges of achieving that.
Wing is generating lift but the stabilization software as you said doesnt know there is another part generating lift thats why its trying to correct its attitude and osiliating.
That was entertaining! :-) I would build a smaller wing to get a happy medium between efficiency and stability. Also make the carbon rods a couple of feet longer to build the wings further out so there's no wing directly above the rotors. I think the rotors decrease the pressure from under the wing and disrupt airflow so you lose lift and efficiency. I'd love to see a brand new build and test flight. Thank you for making great videos!
Awesome idea to improve drone efficiency! I think the main cause of instability is the aerodynamic pitch instability induced by the airfoil. As soon as the wing starts flying. Every normal aircraft needs to adress this problem. Usualy its done by choosing diferent agles of incident between main wing and tail wing. The simplest way to fix this on your drone would be probably by using a stable airfoil like its used for flying planks. No need for a tail anymore.
I've been thinking a tandem wing biplane with the two wings forming a box that is the frame for the quadcopter (but with negative stagger so the frame would apoear to tilt back when the rotors and wings are pointed vertically) and the rotors blowing over the wings to create lift over part of the wing regardless of AOE. With negative stagger the top-back wing could have lower AoE than the lower-front wing so if the frame pitches too far forward the back wing loses lift and if it pitches too far back the front wing stalls so it stays relatively level even without control inputs. Adding extended wings beyond the box with substantil dihedral to reduce rolling then all the maneuvering is via differential thrust so no aerodynamic controls are needed and the same contols work in forward flight as vertical.
Perhaps a reflexive airfoil for stability when freely mounted like that, 23015? That should allow mounting at its center of lift. Or, using a reflexive airfoil, try rigid mounting the wing, testing for optimum AOA at desired speed.
Hi I was wondering if you ever thought about making a video on how to use ardu pilot because I have been wanting to make a autonamas drone but don’t know how to and I see that you use ardu pilot on a lot of your projects so must be a good choice for an autonamas drone. Thanks
I am curious if there are modern thermal sensors or for example an old array from a 90s to 2ks digicam (which are basically heat mappers) are feasible to detect uplifting streams for an efficient glide and fallback for corrections to the pure drone-mode?
Wow, this idea looks great, but I would suggest to mount the wing further rearwards and with a fixed pitch angle because when the center of lift of the wing is behind the CG the wing actually provides positive pitch stability so it wont get into those oscillations. Also would negative angle of attack not cause the drone to crash because it would limit the maximum forward pitch of the quad. I really would like another video of this concept with some design improvements.
I was thinking the same thing about wing location. When the drone tilts forward, the wings move forward causing lift in the front so the drone tilts back, shifting lift to the back etc etc. I have no expertise though so was looking for a comment that was similar.