No revelations here, just some thoughts after spending a year designing, building and testing a VTOL plane.
I finally settled on a tilting-tri design with the 2 front motors doing VTOL yaw and lift/propulsion. I partly wanted the challenge of building/tuning a tilt system. I coudnt bring myself to design a system that took 4 stationary motors for a ride with one pushing. One idea was to have all 3 tilting but decided against that.
So now its flying quite well and its no surprise that using the same motors for lift and propulsion requires a compromise. I already had 8 spare MT4008-600 motors so designed the plane around them. Im using a heavy Li-ion 4s4p 21700 pack and the plane was printed in ASA - not lightweight. So, the plane is on the ragged edge for spare VTOL lift and requires a careful choice of props, at least on the tilt motors.
I have found CAMCarbon 15x8 quite good so far. (I could go 6s and have lots of spare).
Of course choosing a prop pitch that lifts well is going to impact max forward speed and this can be seen in the graph. The system is capable of pushing 20A through each motor in VTOL flight BUT during this speed-ramping test, 100% throttle will only push about 12A through each motor in forward flight. RPM at hover is around 4500 and in cruise 100% throttle gives 6500 and 26m/s. This is approaching the theoretical 8500 for this Kv motor and battery.
Anyway, I would say that the design has exceeded expectations and can slow cruise around at 17m/s at 6.5A and my ‘usual’ cruise of 20m/s at 11.5A so well over and hour on the 18Ah battery.
Wondering about my next project. I have some DualSky MR5015MR-7 going spare so… Perhaps I will try taking 4 stationary motors for a ride and use a powerful pusher for speed?
Any comments welcome.
A VTOL will always be a compromise somewhere, but it looks like you’ve done a good job with this one.
For some inspiration for you next one have a look at the Quantum Systems Trinity F90. It has 3 tilt motors, but the two forward shut off once the transition is done and forward propulsion is from the rear high mounted motor.
Of course that layout (Trinity F90) has the advantage that the rear lifting motor is not carrying much of the weight so the prop/motor can be selected with less regard for its static lift and more for cruise. I see that many companies/universities go for the static quad layout with one pusher. Probably they have gone through very detailed research before choosing that system so there must be some advantage, not least of which is no tilt servos, a component which has caused me some trouble. I would be more inclined to choose this system if there was a way to align the stationary props. I know this is available on some large motors. Ive been experimenting but could do with some help from a coder/electronics guy.
This ‘Parcelcopter 4.0’ is also interesting. They probably do it with a ‘screwjack’ in each side and dont rely on the tilt for yaw (just a guess). Looking at the rear prop in this pic, its likely the rear 2 motors are shut down in cruise.
The forward props on the Trinity fold so that keeps it simple. Less drag than non-rotating prop on a quad/pusher. But, yes, the servos.
Funny you mention that, our school also has a quad/pusher VTOL. It’s hard to compare because the planes are very different, but overall I would say the Trinity is a more efficient design over all. It’s almost silent as it goes over in fixed wing mode. Our Quad/pusher (think Sky Fury) I don’t think is as efficient, but we chose it because the system allowed us to do much more than just survey.
About silent. This is mine on a large waypoint circle at about 60mts height. Yaapu chatting on a table 10 mts away and a chicken in a farm 300mts away are louder.
I bet there was little concern for performance and the priority was given to mechanical simplicity .
There are a few commercial solutions that do that, IIRC one is from T-Motor. AFAIK people glue magnets to the motor rotor and somewhere close that help props align in the correct orientation.
A year ago my first ‘self briefing’ specified being able to deliver 2 x one pint sized beer cans (or similar object), stored in the belly autonomously at high speed. (cant imagine why ). Since then there was considerable ‘mission drift’ and we ended up with a cruiser. My first ‘sketch’ on the right and what we ended up with on the left. Still quite keen to make a delivery system though and kinda wish there was an amateur competition for such a thing.
I can not think of a VTOL that is not a classic 4 stationary motors and one pusher.
I need it because i discharge the whole battery dedicated to the pusher motor when flying.
And when it is time for landing and i am somewhere i need vertical landing, there is no juice left.
So i use a dedicated battery for vtol operations.
I do not see a way to use a dedicated battery using tilting motors vtol
Cant be too difficult to arrange a system that increases pitch when the prop is forward. Just need a hollow shaft on the motor and a ‘cam’ on the fixed servo unit. Had a quick brainstorm in cad. Can use folding blades available in many sizes. I was looking at RC helicopter tail units but nothing really suitable and the prop slot is the wrong way.
Great idea - or you would have try to build a constand speed prop
with mechanical control (centrifugal weights).This should also work independently of the tilt angle servo.
You could be on to something here. The prop on a tilt rotor will always be a compromise between vertical and horizontal performance, but if you could make a viable variable pitch system that could have a big difference for VTOLS.
For the last year I have been working more or less on the same things, and if I may chime in a few ideas here:
Tricopter seems at first glance to be a very promising thing, and I built a 4 kg prototype which flies well. However, when I started playing with props and optimizing the props for cruise (which one should do before building protypes), I ran into the same problem as outlined above - it is virtually impossible to optimize for both.
The idea of variable pitch prop is really a dead end, I though about it, and probably the best way for trying to implement it would be to take helicopter swash plate system.
But: these systems are complicated and prone to failure
the helicopter RPM is much lower, so one would have to see how it would work with RPMs which are 3…8 times higher
The mechanism is heavy and fragile, and this mechanism would have to be mounted on tilt motor, which would make the whole thing even heavier.
One of the main benefits with a 4+1 (really any VTOL that has separate propulsion dedicated for VTOL vs forward flight) outside of mechanical simplicity of no tilting parts is that you can really optimize each propulsion system for its regime. Most tilt-rotor systems fair towards using picking a propulsion system that is optimized for forward flight - which makes hovering in VTOL mode very inefficient. If your mission requires you to hover for a decent amount of time (i.e. not just a quick takeoff and landing), then you really need to make sure your hovering performance is top-notch. Thus, even though you are carrying around the extra weight/drag of an additional propulsion system in ether mode - the increase in efficiency you get from dedicated propulsion systems for each flight regime leads to improve performance and flight time. You will see a lot of delivery drones go this route since they often hover for long periods of time during the delivery phase.
In a tricopter, if you optimize for cruise, you will end up with a relatively small prop with high pitch to diameter ratio, and both of these things are very bad for hovering. To hover with such a prop, you would have to increase the motor power and ESC to cover losses due to inefficiency, and the extra weight of these (motor, ESC, and extra battery) quickly reduce overall cruise efficiency as well. The only way to to get better results with tricopter is to use different front and rear propulsion systems, like Trinity
Basically, you can use a smaller rear motor which is further back, so as to reduce overall power requirement for that motor, then, you can deal better with inefficiencies of that motor for hover.
The best thing of Trinity style is that you can get two propulsion modes - one very economic cruise on rear motor, and the other high rate and efficient climb on two front motors, and also redundancy if the rear motor fails, to come back.
My thoughts on increasing the cruising efficiency of 2 or 3 tilt rotors is to switch voltage in flight as most of the inneficience will come from controller switching loses.
I see the hovering fase not relevant to the overall efficiency of endurance Vtols and the safety is easily implemented with good battery sensors.
Implementation require a custom battery switcher and an isolated controller supply.
Similarly delta /wye switcher or even parallel /serial winds can be done at very low weights as they are not going to do any switching under load.
That is only marginally true. By the way, implementing a safe battery switcher for the currents VTOL uses involves pretty complicated electronics.
And, since there will be always a mosfet in the circuit, the losses will be permanent during the whole flight in plane mode.
Basically, you would have two batteries in series (like 4S on top of which a 2S is connected), and you have to have two switches to connect either intermediate connection (4S) or the higher connection (6S). This involves using P channel mosfets which are usually less efficient than the N channel mosfets. Now, the switchover is also going to be complicated. If your timing is bad, and both of the switches are activated, you will shorten the upper battery. I am not sure how critical a too long delay would be during the switch over on the ESC.
My two cents worth based on some builds I’ve done…
The best compromise for me is the 4+1 configuration as it addresses some considerations:
Hybrid power - Since your VTOL time is normally just a couple of minutes or so, you can have a tremendously smaller LiPo for that and use high capacity Li-Ions for your FW flight. You maximize FW flight time without compromising the landing power needed in VTOL.
Bigger and higher pitch prop for FW flight - The FW motor will only spend a few seconds of max burst during the transition and for the rest or most part be running at 50% or less power, so you can install a bigger high pitch prop. I built one that normally as a FW aircraft used an 11x7 prop on a 900 kV motor; as a VTOL, I ran a 13x8 on the stock motor and ESC that barely warmed the motor and ESC. I could’ve probably pushed a square 13x13 if it was available.
Redundancy - If your FW motor fails, you can still land with the quadmotors as your failsafe.
Endurance - You can have a gas/fuel FW engine that can provide not minutes but hours of endurance flight.
If aerodynamic penalty is a concern, I normally turn on the ESC brake to prevent the quadmotors from spinning in FW flight, especially during turns. I also read somewhere that T-Motor produces motors that will automatically algin with the booms to help reduce the drag during FW flight.
I agree, except with redundancy: relying on 4 motors to fly back is virtually zero redundancy. For long range flights which may go 50…100 km, if main motor fails, the 4 hover motors with 50 % battery charge at best will give you 7…10 km flyback range.
Folding props may be one solution for aerodynamic drag reduction, the other option is to design either mechanical or electronic device for turning the props to be aligned with the wind.
In fact, I have a simple, but crazy idea for doing so electronically - turning the motor at slow speed is very easy, it is a kind of very crude ESC. It would switch the three lines step by step so that the motor would turn maybe in a second, until a sensor detects that the prop is aligned. Now, the tricky part is how to combine with the main ESC. I think that one can put the two ESCs in parallel. The auxiliary ESC would detect when the main ESC is off (there would be no voltage on the motor lines), and only then it would begin to act. The main ESC must be disconnnected electronically, so no braking could be used on the main ESC.
It is quite obvious that if the main ESC becomes somehow active, the aux ESC will fry in miliseconds. To prevent major damage, the cables leading to aux ESC must be extremly thin so that they would burn open first in case of a programming accident.
Turning the prop to align it can also be done through the main ESC, but that would involve reprogramming the ESC which could be very tricky, and would make the ESC less reliable. I probably would go the other route, with auxiliary ESC.
I should’ve been more explicit… the redundancy is if your FW motor goes bad and whether it glides or stalls but descends to Q Assist altitude, it can recover and land with the hopes of avoiding or minimizing damage to the aircraft.
I’ve already seen a YT video of someone already doing that. The ESC pulses until it aligns with a marked sensor. T-Motor has the same solution already available. Check this proof of concept video.
My guess would be that T motor solution would be quite expensive. What I have in mind is an add on module which could be used with any ESC. And the video is quite old, 3 years!
If you are in the mountains, a safely landed plane which is 20 km from you is as good as one that crashed to smithereens.
.
). Since then there was considerable ‘mission drift’ and we ended up with a cruiser. My first ‘sketch’ on the right and what we ended up with on the left. Still quite keen to make a delivery system though and kinda wish there was an amateur competition for such a thing.
