Location, location, location. The biggest problem I have with model flying. I live in south west London where space is generally at a premium, and what opens space there is seems to be the domain of dog walkers. I have a few sites near me where I can fly, but they are a reasonable drive, or I have to get there super early to have the place to myself. (Having the place to my is definite pre-requisite, especially when testing out new aircraft).
During lockdown it was incredibly frustrating, if only I had field at the end of my garden! The idea of a VTOL was born, I would be able to do the development and hover tests in my garden. I set about doing the research and ordering the parts I would need. By the time I had what I needed lockdown was over. Nevertheless, the idea of being able to do some flying in my garden had taken hold.
First up, VTOL is HARD. Even with the amazing open source software that is OpenAeroVTOL, and the build logs from people like Tom Stanton there is a lot to get right (and wrong).
I settled on modelling the Leonard AW609, the V22 Osprey has been done a few times and, lets be honest it's not a very nice aircraft to look at. The AW609 has had a troubled development, lasting nearly 20 years, going through various different companies from Bell and Boeing to AgustaWestland, (who became, Leonardo).
The mission is simple, design an RC aircraft modelled loosely on the 609 that can take off vertically, hover, transition to forward flight and back. I say 'loosely' as it is difficult to really go scale on this, either the props would need to be massive, of the rest of the aircraft tiny!
The only way to make this realistically possible is using a flight controller running a piece of software called OpenAeroVTOL. It is freeware programmed by Happy Sundays to take care of the stabilisation of the aircraft much like a standard flight controller does. The twist though is that it manages the mixing of controls through the transitions from the hover to forward flight. That is to say, during the hover roll is controlled by asymmetric thrust, but in forward flight the same asymmetric thrust would control yaw. Its good enough for NASA, so its good enough for me!
The software runs on the kk2 flight control board. To control the aircraft, we have 2 motors in nacelles at the end of the wing. The nacelles are actuated by a servo each. Because the trust line is above the CoG we are able to control pitch with just these 2 motors (that's the theory anyway). Two more servos control the flaperons, which as well as acting as flaps reduce the wing area that blanks the prop during the hover. The flight controls are completed by a stabilator on the tail, controlled by a single servo.
So far I have managed some hover testing, 15 seconds is the longest hover in the confined space of my garden,. These tests have yielded lots of 'learning', a redesigned nacelle and few other tweaks, I'm still a way off testing forward flight but so far I'm ahead of the 20 year development curve...