We are a two-man team that just completed the first major phase of our yearlong APM 2.5 tiltrotor project. We wanted to keep things in the dark until we knew it had potential. Here is a quick summary, but if you’re interested in VTOL aircraft, we’d really like you to read our detailed project summary (attached PDF) and give us some feedback.
Just some of many major accomplishments include:
Design and manufacturing of a tiltrotor on a relatively low budget.
- Utilized many free software applications to make educated design compromises
- Utilized 3D printing to create complex flight control components
- Slow, methodical engineering to identify problems and solutions before first flight
Stable hover with conversion to 23 m/s full airplane flight
- Light hover maneuvering completed in STABILIZE
- Hovering tested in LOITER
- Thrust vector can be set at any intermediate angle and flown carefree all the way to airplane mode
Merged several significant portions of Arducopter and Arduplane software on APM 2.5 hardware
- Heading strategies for quad and plane merged with unique tiltrotor control
- Seamless blend of servo actuated aerodynamic surfaces, and 2 brushless motors to achieve stable flight at any speed and intermediate thrust angle
Tiltrotor 1 Prototype Stats:
- All-up Gross Weight = 2.4 KG
- Motors= 2x G15 810kv with 14 inch CF props / single 3300 mah battery
- OGE Hover Power Required - 40 amps
- Conversion Power Required – Thrust vector at 30 degrees from vertical 10 m/s = 20 amps
- Airplane Power Required - 23 m/s = 30 amps
We have lots left to do, but would love to get some feedback to help guide us.
Here is what we have planned next:
Autonomous Flight
- Merge Copter and Plane navigation strategies to suit a tiltrotor.
Pixhawk
- We’d love to incorporate the new capabilities of EKF, spline waypoints, terrain following ect someday. The APM 2.5 has not limited our development at all, and is still our bread and butter.
Tiltrotor 2?
- We learned A LOT in a year. There are numerous things we can make better on this prototype, or improve upon with the “next generation” model. We’d like to build a bigger, better, more efficient tiltrotor with all the lessons learned. This is where we need some feedback from the community. Please read the detailed design summary (attached).
Of course we are not going to brag about designing a tiltrotor and NOT post a video.
Tiltrotor_1_Design_Summary.pdf
This barley summarizes a year of work in 10 minutes.
Comments
Larry - It is interesting to note that the young fellow in the UK employed a 'clear wing tip' like Team Tiltrotor used initially.
If you pay attention to what he has done with the KK2 settings and his TX settings, he has distributed the 'control' of the transition between two devices (actually 3 but I will get to that at the end). His was definitely a 'try it and see' approach and is worthy of more consideration for a simple approach to the problem. With TX expo/rates and mods to the KK2 PIDs, he worked out a way to hover, forward fly, hover. It is a well done and interesting demonstration.
'Meatware' does the rest and if the presence of his Trex Dominator 700 heli (seen in one video) is any indication, he has trained his eye-hand feedback loop well. That is the 3rd 'processor' in his control system.
-=Doug
Really cool stuff. There is a kid over in the UK that did something similar. I wish I had the time to work out one for mysel!
http://flitetest.com/articles/vtol-tilt-rotor-v2
Rick- Thanks for checking out our tiltrotor. I don't know enough about your aircraft to tell you if your 2836 motors and 10x6 props will be a good choice for your v-22 tiltrotor. It will probably have sufficient the power to hover, but you may not be optimizing your efficiency.
Unfortunately I don't think you will be able to fly your tiltrotor with just minor changes to arducopter quad PID's. As your rotors change angle they will no longer control a single axis. For example a "propeller driven" roll input at 45 degrees thrust angle will result in both roll and yaw outputs. This will reek havoc on your PID strategy and you will be sending roll error to on output that will result in an undesired a yaw correction. It only gets more complicated when you introduce the yaw PID because arducopter software will attempt to maintain a compass heading, so when you transition to airplane mode and input a roll command, arducopter software (as written for a quad) will try to achieve a roll angle and keep your nose pointed at a compass heading. I can assure you that this will end badly for an airplane/tiltrotor. I will let a developer talk to a BI-COPTER software availability ( I am not aware if any Bi-COPTER code availability) . But either way you will run into the same problems with BI-COPTER software, once you change your thrust vector. You may hover well with Bi-COPTER, but you will not be able to transition to airplane with the current quad PID architecture.
You should not need to gear the throttle with thrust angle. We conveniently hovered at 50% throttle and also had very good airplane mode cruise at 50% throttle, so typically throttle was only used to adjust rate of climb. At a given thrust angle you will achieve some trim speed. Depending on your thrust angle, you will want to retain the ability to climb or descent when your rotors are providing the lift. As you transition and the wing provides the lift, the throttle will primarily control speed and pitch angle will control climb / descent. Tiltrotor flight control is not easy! Hopefully we can help get you an easier solution in the future.
The osprey build is a beautiful display of craftsmanship. I'd be devastated the first time it crashed. I may have missed it somewhere in his blog, but if he plans on fixing brushless motors in the nacelles and simply rotating both nacelles it probably won't work (well). He will need to install swashplates (like the Rotormast) or attempt what we did and "mock" longitudinal flapping. I did see a brief discussion he posted about using fluid and a printer pump to move the CG around, but that doesn't solve a lot of the problems that come with tiltrotors. Either way, it is an impressive build. We had no idea so many tilt rotors were being made!
Jake - yeah that might be an issue but, I think he is willing to let that part of it go. ;)
On the other hand, how cool would a propane/LPG powered tilt rotor be?
Best save that pondering for another thread.
He is currently showing step-by-step build of another Osprey. Templates and hot-wire cut foam construction.
-=Doug
Amazing, lots of respect in achieving what you have so far. Up the VTOL!
Team Tiltrotor here's the link to Rick's build. OSPREY
That link is his member profile and blog listings. Slide down the page for more pics like below.
He is more of a airframe type of guy and has been looking for help in the area you have made great progress in.
He lives in British North America -- aka Canada. ;) Thus the maple leaf on the Osprey.
Now if Rob L. and this guy got together...
-=Doug
@Lazy- If you want to know more about the technical details than the attached PDF described, please ask away.
@Jake- Thanks for the sound advise. There is no doubt we would not be here without this communities hard work. Thanks for pointing us at these ESC's.
Interesting...Did you show more technical details?
It is sort of outside the scope of this blog post to discuss pay for coding, but it is definitely possible.
-
1
-
2
-
3
of 3 Next