Designing/making a VTOL (aka V-280 Valor, Bell XV-15, or the V-22 Osprey)

[JasonK]

My eventual plan is to make a VTOL similar to one of the above. My initial sketches are below (very very rough).

Items I still need to work out (before I go in to all out build mode):

• Most motors I see are focused on fast small blades, which has 2 issues (below). I need to see if I can find a solution for a larger prop setup or if I need to work around this (larger prop might be harder for the servo to make small adjustments needs for the flight controller to keep things stable)
  • At scale, these craft have fairly large propellers compared to the airframe size
  • Unless the motor assemblies are pushed out beyond the wing, to much thrust would be lost over the wing without the large prop size
• Flight Controller board, I don't know of anything that can handle this design. However I am not overly worried about this situation, I have enough programming/engineering experience, that I am confident that I can write/extend from existing code something that will allow the needed controls.
• Determining the needed servo/engine/esc/props, needed - which I hope to be able to work out at least partially in reverse via the target mass. My desired target mass of 249g may make this extra challenging. I may find that I need to go up to 2-3x that to make this work out.
• I have thoughts on how I could get by with 100% vectored thrust for all of hover, transition, and forward flight (so less servos needed), however, this does mean no control in forward flight unless the propellers are moving, and given the current design, might have thrust angle issues (vertical center of mass vs center of thrust offset), so I might still want to have full fight control surfaces - which means 2-4 more servos (depending on if I can put the alerons under command of the tilt mechanism or not).
• Deailing with the scale wing size, I suspect the WCL is going to be fairly high without making the wing much larger then the reference craft.

 

[quorneng]

Nice project but I do wonder if you have under estimated the problems the flight controller board will have to take into account.
The stability requirements are different between hover and conventional flight.
Not easy to arrange for a smooth change over in the function of the electronics during the transition phase.

 

[Aireal Anarchist]

here is a build log from a guy on rcgroups that just built a vtol from a C1Chaser wing ....maybe this will be of some help or interest

https://discuss.ardupilot.org/t/c1-chaser-tvbs-conversion-instructions/55917

 

[JasonK]

Nice project but I do wonder if you have under estimated the problems the flight controller board will have to take into account.
The stability requirements are different between hover and conventional flight.
Not easy to arrange for a smooth change over in the function of the electronics during the transition phase.

I absolutely expect some issues there. I do have an Electrical Engineering degree and do expect some issues to be worked out. but that is why this is an interesting project to work on.

 

[JasonK]

here is a build log from a guy on rcgroups that just built a vtol from a C1Chaser wing ....maybe this will be of some help or interest

https://discuss.ardupilot.org/t/c1-chaser-tvbs-conversion-instructions/55917

Thanks, this looks useful

 

[JasonK]

Did a first pass on some mass, entering my desired mass (259g) and guess of a wing span of 24", gives me a ~5.25 cord to get a WCL under 10. This makes the windspan:cord a bit lower then what I can make out from the reference Images from the v-280 Valor. Using that wing area (135 sq in) and a first pass on the various sizes of the rest of the airframe would get me to ~160 g from just the foam needed to build it which would only leave 90g for electronics + battery + everything else. Given the electronics I have, I don't believe that is in the realm of possible (I am still waiting for my H pack, as that looks like a good option to work with until I figure out a way to get 'scale' props).

If I drop my wing area down to 100 sq in with a 5 in cord, that would push my WCL up to 15.2 - racer territory. this would get my foam mass down to ~120g (scaling down the initial calculation) which would leave 130g for electronics... perhaps doable, however flight times aren't promising at that.

If I up my wing area to 175 sq in and my total mass to 1lb (yes my units are all over the place), that gets me a WCL of 11.9. Foam would be approx 210g of about 450g. Defiantly looks like scaling up in size is gaining me some useful characteristics.

 

[JasonK]

well after all that rambling I am getting started. I have my power pack H in hand (which the bits were lighter then I was expecting), some foam and a flight controller board on its way.

I am going to start out with some level flight tests (as the FC won't be here for a while).


Started cutting out a wing, fuselage, etc for a build to test my Power Pack H and some forward flight stuff for my vtol design (don't have a flight board planed out yet, so starting with a test bed and some forward flight tests).

Right now my weight is looking about like this which gives me some room to work with for my final design:

Foam	85g
2x Motor/ESC/Prop	24g
Receiver	8g
Power Dist Board (guess)	10g
4x Servo	20g
Motor mounts & Glue\tape	unknown
Total	187g + unknowns

I Still need to build/design some nacelles for the motor mounts. a Basic set for the plane flight tests and then some special ones for the rotating motors.

My cutting plan right is as follows:
Fuselage:
Top/Bottom: 32mm x 508mm
Sides: 40mm x 508mm
Wing: 220mm x 508mm, with double bevel @ 90mm, score at 125 & 150 (Ailerons and servo slots TBD)
Spar: 25mm x 508mm (Servo slots TBD)
H-Tail: 210mm x 130mm with 2 70mmx70mm angle cuts for shape
V-Tail: 105mm x 130mm with 1 70mmx70mm angle cut for shape (1/2 H-Tail)

I still need to work out the following for cutting:

• electronics access
• fuselage doubling for around the wing joint area
• nose
• polyheadral wing tips for the plane flight mode initial testing.

I picked up a flight controller board, looks like they are generally flash-able with different software. I see Betaflight, which is focused on quads and inav, which is focused on autonomous flight - at least to an extent. I don't see anything obvious in either to handle the VTOL behavior changes (and the glut of features in both that aren't what I need make the code bases rather large to work through to figure out the bits I need). If someone actually read this far and has any suggestions or communities that would be good to check in with, please let me know.

 

[JasonK]

The initial cuts laid out.

 

[JasonK]

more cutting notes:

V/H-Stabilizer -> control surfaces bevel cut at 30mm from back edge
Stabilizers -> 70mm overlap with fuselage
H-Stabilizer - bottom of fuselage
V-Stabilizer - top of fuselage

fuselage cuts to connect H-Stabilizer
cut 70mm from bottom layer of fuselage
4mmx70mm notches in sides of fuselage

Aileron cuts: 30mm from edge, 140mm wide, 30mm deep
servo cutouts -> 30mmx30mm cut out 30mmx130mm from corner (also cut spar for this space)

I ended up just putting in some cross braces, instead of the whole top of the fuselage as I was going to need access to the internals while working, this ended up putting my weight with glue, sticks, and rubber-bands down to 74g (+5g for the remain bits of the top)

Picture is with 50g of clay in the nose to get it balanced at the 25% of the main wing (31.25mm). In toss tests, it flies nicely, amost like throwing a dart, however it does go fairly fast compared to my tiny trainer (to be expected given the _much_ smaller wing and over all reduced cross section). I definately need to reinforce the spots that hold the wing in place, they take quite a bit of stress if the wing hits something.

 

[JasonK]

After banging my head on lots of ideas, I ran across pitcherons. Which if I do that with dual motors (which was expected for this design), I can get all the need control behaviors for hover, slow flight, full flight out of just 2 motors and 2 servos. I am going to first build this over 250g to make sure the idea works, then look at what I could to to make it sub-250g if it works well.

 

[JasonK]

Some renders of my current proposed idea:

 

[JasonK]

Found this: https://design215.com/dcal/toolbox/tubing-calculator to help with calculating strength of the tubes.
Considering using a 8x6x?? tube for the main spar and a pair of 10x8x?? tubes for the 'rotating' bits. However this might be to tight to rotate smoothly.

Recalled goodwinds.com and looked up some carbon fiber spars there and am considering this:

.375" x 24" Pultruded Carbon Tube
OD 9.52 mm
ID 7.37 mm
Length 61.0 cm

and this

.281" x 24" Pultruded Carbon Tube
OD 7.14 mm
ID 4.57 mm
Length 61.0 cm

which can both be cut to the needed lengths. that gives 0.23mm of 'space' for some lubricant to ease rotation, which shouldn't cause any appreciable slop, but make the parts fit.

 

[quorneng]

Following with interest.
Would it not be advisable to make a simple test rig to explore the functionality of the hovering capability of the control board before even considering building the air frame?
Small steps.

 

[JasonK]

Following with interest.
Would it not be advisable to make a simple test rig to explore the functionality of the hovering capability of the control board before even considering building the air frame?
Small steps.

to an extent yes, however, the air-frame will change both the moment of inertia and the aerodynamics of the craft. So yes verifying that a bi-copter setup can be hovered would be useful (however it has been done, so I know it can be done), but isn't something I would want to tune/etc with because of the meaningfully different behavior.

 

[L Edge]

JasonK
You really need to look at the aerodynamics of things-- for instance-- wing is vertical and a short burst of air hits the wing and fuselage which will pitch the plane. How will get the fuse get back to horizontal with your movements especially with a V-tail? You need a large area of elevator to help you.

Sidenotes, alot of Rc vtols that work well use a dihedral to help the stability during the transitional stage. Otherwise, the wing will stall it out.
The CL-84 solved the problem by adding a vert prop in the tail. Haven't seen any fly with just your setup. You also need to build super light.

I turned my lemon into lemonaid and had fun with it. It will fly up to about +30 and after that, stalls out.
I used a gyro that always kept the fuse level to get it to do this. Have another video showing how it always stalls and rolls over.

Best suggestion is to leave wing horizontal and add bi-copter setup at the wingtips.

 

[JasonK]

@L Edge,
the control system should work just fine for forward flight:

and vertical flight would be similar to this:

So the remaining questions are, as you said, will the wing tilting have a big or small impact on the lift off flight mode and the transition.

Worse case situation, I am wrong and spent a few days of spare time designing and building it.

 

[L Edge]

If you look at the bicopter video at 1:35, as he said, going backward, it goes freaky. That is what i said above. Notice, he never took off from the ground. Those are the problem areas as well as the transitional stage.

I tried another approach where the bicopter works (servos on leading edge where motors/prop controlled all 3 axis) where I have to ascend quickly (overcome ground effects including wind) but the failure was with any wind on landing. It was like Horizon's X-Vert.

As Quorneng said, do a simple test rig to see what happens. I wonder also if you are going to get any binding (rotation) and does it mean more weight to the rig. You will find that you will need dihederal with your setup to maintain stability.

 

[JasonK]

If you look at the bicopter video at 1:35, as he said, going backward, it goes freaky. That is what i said above. Notice, he never took off from the ground. Those are the problem areas as well as the transitional stage.

yup, it has some issues in the tune, likely related to the aerodynamic difference between forward/backwards flight, There are other videos of people who got the same craft working just fine with a lower CG and possibly a few different configs.

I tried another approach where the bicopter works (servos on leading edge where motors/prop controlled all 3 axis) where I have to ascend quickly (overcome ground effects including wind) but the failure was with any wind on landing. It was like Horizon's X-Vert.

yup, there are going to be aero dynamic issues to solve, not that worried about it.

As Quorneng said, do a simple test rig to see what happens. I wonder also if you are going to get any binding (rotation) and does it mean more weight to the rig. You will find that you will need dihederal with your setup to maintain stability.

test rig won't solve the aero dynamic issues of the shape of the plane (meaning motors, servos, batteries/FC) - I have to test it with the expected aerodynamic form.
dihederal impacts 'plane mode' stability, would have no impact on hover, give an FC's ability to turn something unstable stable, I don't even know how much of a help this would be in transition.
I was already planning on the using bearings, so I don't think any rotational binding will be much of an issue.

have you seen this? because the key differences are rotating the full wing instead of just the motors (yes this makes wind behave different in hover) and using pitcherons instead of 'traditional' control surfaces.

Also note: in hover mode, the elevator/rudder/ailerons/etc do basically nothing to help with control of the craft, even if it gets upset by the wind, the thrust vectors of the motors and amounts of thrust is all that matters for controlling the attitude. in hover mode, there isn't any meaningful airflow over the tail, so it can't be used to stabilize it. If I had ailerons, with the wing tilt, they could help, but they aren't used with pitcherons.

Also, the above are just a few examples. there are plenty of examples were things similar have been shown to work, just not with my exact combinations of stuff. And worse case, it doesn't work.

 

[Zephyr1]

You're going to want to get a KK2 board and flash some openaerovtol. You can buy the board pre-flashed for about $30 and then there is a massive forum on RCgroups detailing the use of the board. It's what I'm using for a a similiar vtol project

 

[JasonK]

You're going to want to get a KK2 board and flash some openaerovtol. You can buy the board pre-flashed for about $30 and then there is a massive forum on RCgroups detailing the use of the board. It's what I'm using for a a similiar vtol project View attachment 177925

1 - that craft looks intresting, how big/heavy is it? is there any more details on it?

I already have a F411 wing FC sitting on my bench.. but haven't seen openaerovtol yet, will look it up.