Collect data from V22 Osprey threads.

[OwenN]

Has anyone done the whole thing using Ardupilot?

I saw a good one on RC-groups from 2014 that had interaction between flight surfaces and motor tilt, and gyro vertical adjustment of motor tilt.
This stuff would go right into my project-canard version with the props between the wings.

Has anyone collected all the hot tips on this stuff?
There seem to be several current Osprey v22 threads.

KeiTora's blog - RC Groups
I will try to follow this thread and see how it went.

I want to use a separate descent program that flicks motors up at low throttle, normal flying speed, then aerobrake at high angle of attack

down to hover speed, gradually ramping up motor throttles. Once hover is reached, then I want to switch the motors back to linkup to flight surfaces.

Can this be done within Ardupilot?. Does the brand of flight controller make a difference? I want to use a Matek 765.

I have also seen a scheme to add a "back end" arduino processer to handle some of this stuff. would this be neccessary?

RC VTOL - V4 - YouTube

 

[JasonK]

I have also seen a scheme to add a "back end" arduino processer to handle some of this stuff. would this be neccessary?

No, Tom did that because he couldn't update his firmware and/or couldn't handle reprogramming it himself, use a different Flight controller and possibly custom firmware build of your flight software of choice and you wouldn't need that extra step.

 

[OwenN]

Here is a list of threads about V22 Ospreys from RC Groups. Most of them seem to be quite old. The standards of flight controllers and ardupilot , etc, have advanced a lot since then. I am reading them all, to see if there are any handy tips.

If Ardupilot can't handle it, I can't really customise it!
Has anyone done a successful implementation just using Ardupilot, or some other non-modified firmware system?

Treating Ardupilot as sort of a "black box", and further processing its output seems easier.

Will Ardupilot install on an Arduino board? Do I need to learn how to program the Arduino? - I only got as far as Pick Basic, and a tiny bit of C+++.- I don't really want to properly learn that. It takes a lot of practice to debug it and to learn all the rules, and to get familiar with the main library routines.

It may be possible to pinch some Ardupilot code to make a back-end control mixer, and to have several take-off and landing protocols.

A more basic-like language would be nice.

I will have a bit of a read-up on the Arduino board.

https://www.rcgroups.com/forums/showthread.php?3290853-Bannana-Hobbies-V-22-to-OAV-Conversion
https://www.rcgroups.com/forums/sho...ue-Bicopter-with-full-transition-capabilities
https://www.rcgroups.com/forums/showthread.php?2156225-Design-Questions
https://www.rcgroups.com/forums/sho...ional-VTOL-with-transitional-flight-both-ways
https://www.rcgroups.com/forums/showthread.php?1796683-Lockheed-Martin-Various#post23666453
https://www.rcgroups.com/forums/sho...i-Tilt-Rotor-Plane-System-Design#post28375279
https://www.rcgroups.com/forums/showthread.php?2283575-OpenAero2-Roll-Not-Working#post29924811
https://www.rcgroups.com/forums/showthread.php?3423979-OpenaeroVTOL-for-a-dummy
https://www.rcgroups.com/forums/sho...-with-transitional-mixers-(perfect-for-VTOLs)
https://www.rcgroups.com/forums/sho...-V22-with-real-transitional-flight-controller
https://www.rcgroups.com/forums/showthread.php?3216021-RIX-V22-Osprey-Rebuild#post41019203
https://www.rcgroups.com/forums/sho...le-depron-V22-VTOL-(with-plans!)#post33817826
https://www.rcgroups.com/forums/showthread.php?1557619-Tilt-rotor-fixed-wing-(Osprey)
https://www.rcgroups.com/forums/sho...TOL-Tiltrotor-(Osprey-V22)/page5#post29361454
https://www.rcgroups.com/forums/showthread.php?2057761-VTOL-Tiltrotor-(Osprey-V22)#post26923864

Also, I have read that the basic bicopter version of the osprey suffers from unintended pitch-down, maybe from ground effects.
Is this a real effect, and how could I mitigate that? "Pop" to 1.5 meters on launch?
How much backwards tilt do I need
at the motors?.

I am planning on 10 degrees.

At least I have a good spec on motors, props, thrust to weight, battery. (4S, 2.5 thrust to weight, 9 inch props, 1400g plan weight,
2805.6 1800KV motors-gives good pitch speed)

By my calculations, it should do a good "warp speed" launch, under 2 seconds to full flight speed.

 

[OwenN]

Reading on the OAV conversion, I see they have a pot on the ground transmitter controlling tilt, as well as elevator linkage to tilt.

This avoids negative attack angle when flying forward in part-hover mode. He also has a 3-mode switch for normal flight, 20 % tilt, and full hover.
how would you connect all this to the ground transmitter?
Does it have current loop inputs? I need to have a look at the specs for the raw ground transmitter.
How many channels do they normally have?

How do you get these back out at the other end? What is the S-bus inverter cable? does this go from the receiver to the
FC?
Do modern systems have BECs or is that all done in the flight controller?
(they are supposed to cut motor current on low battery, leaving everything else powered).
<edit> never mind. I now have 10 more videos to watch!

 

[JasonK]

If Ardupilot can't handle it, I can't really customise it!

I have had openvtol recommened to me, but that seems to be 'old' and might not be able to target current flight controller boards

Will Ardupilot install on an Arduino board? Do I need to learn how to program the Arduino? - I only got as far as Pick Basic, and a tiny bit of C+++.- I don't really want to properly learn that. It takes a lot of practice to debug it and to learn all the rules, and to get familiar with the main library routines.

current flight controllers are a solid order of magnitude more powerful then a basic arduino board, there are a few boards that can be programed with the arduino that can run a custom written flight controller, but I do not believe ardupilot is going to run on any of them.

It may be possible to pinch some Ardupilot code to make a back-end control mixer, and to have several take-off and landing protocols.

A more basic-like language would be nice.

to my knowledge all flight controller stuff is written in c/c++, this is high performance embedded software, which is almost exclusively done in some version of c or c++.

Also, I have read that the basic bicopter version of the osprey suffers from unintended pitch-down, maybe from ground effects.
Is this a real effect, and how could I mitigate that? "Pop" to 1.5 meters on launch?
How much backwards tilt do I need
at the motors?.

I am planning on 10 degrees.

I full suspect your going to need more then 10deg of motor deflection from vertical. My build I am putting in around 40, if my servos did a full 180deg of motion, I would have it 45-50 available deflection for vertical and 45-40 deflection available for forward flight.

The only thing I could see for unexpected pitch down if the GC/motors/etc are all lined up just right, is in 'slow forward' flight as the angle the motors would be at would cause a pitch down effect, so if the elevator doesn't have enough authority at the slow speeds, you would end up in a more 'hover' like situation with the motors tiled at a very strong pitch down position.

 

[JasonK]

Reading on the OAV conversion, I see they have a pot on the ground transmitter controlling tilt, as well as elevator linkage to tilt.

This avoids negative attack angle when flying forward in part-hover mode. He also has a 3-mode switch for normal flight, 20 % tilt, and full hover.
how would you connect all this to the ground transmitter?

With stuff light this, the ground transmitter can be rather 'dumb', the flight controller is doing all the mixing and complicated stuff on board. your going to be sending at least 6 channels worth of data to the flight controller -> Throttle, pitch, yaw, roll, 'mode', 'armed'. were 'mode' controls the blend between hover <-> forward flight. This could be on a 3 way switch (so only 3 settings) or on something with a full slider input (giving full control on the exact mix between the 2). 'armed' is to enable/disable the Flight controller's gyro/stabilization logic, so you can power up the craft without the flight controller turning on the motors while your setting it up.

and you might want more channels then that, for example you might want to be able to switch between an angle mode (were the pitch and yaw inputs become, what angle to tilt the craft to, instead of how fast to rotate around that axis) and rate mode (how fast to rotate around the given axis).

and if you want any 'auxilery' stuff like runing lights that you can turn on/of, retractable landing gear, etc. those will need more channels.

 

[OwenN]

About 8 should do- 3 attitude, 1 throttle, 1 variable motor tilt, mode-selection-3 or more positions could be on one channel??
arm....

I don't really need to "do" horizontal hover . This project was intended to do belly launch, straight up and down.

Horizontal hover is a nicety for the imaginary passengers, and probably a good thing for a heavy cargo copter.

More tilt back is a problem because the props may hit the wing, and it requires large "V" cutouts in the main nacelles,
which are horizontal and slung under the wings.
I arranged minimal wing clearance with the motors straight up.

I only added the motor tilt because of the motor location-back towards the aerodynamic center.

It can just yank straight up with minimal motor tilt, and hover tail-down using elevons for yaw.
In this flight mode, the nose is generally always up past 50 degrees, and nose pitch-down shouldn't occur.

It is cured on launch by air drag on the wing plane area center, which is offset from the props when they are raised.
You could fly an Osprey the same way, with large excess thrust. I think an osprey likes best to fly with slight
motor tilt, as wing loading is very high.

Also, could someone explain the usage of "Mean Aerodynamic Chord", where does it start and end, and how long is it?
I know how to find a mean aerodynamic center.
I assume that 35% of the mean aerodynamic chord is behind the mean aerodynamic center, and is still a stable balance point?
Is this only with a small tailplane, or does it apply to Canards as well?

 

[JasonK]

if your planing on having vertical take of, like an Osprey, you either need to have sufficient thrust vectoring (the angle of tilt) or full variable pitch rotors like a helicopter.

regarding CG calculations/etc, there is a nice tool here: https://www.ecalc.ch/cgcalc.php
it also explains most of the terms in question.

 

[OwenN]

I have seen that before, but it doesn't explain Canards very well. What I have been calling the mean aerodynamic center, here it is called the neutral point. The MAC is the mean aerodynamic chord of the main wing only.
I need to find how to trim a canard.
There is quite a big gap between the neutral point and the MAC 25% point, or AC for the main wing.

I could trim anywhere between the main wing AC, the neutral point, and beyond.
If beyond, the front wing just carries a proportion of load beyond its size proportion to the main wing. I have another PDF to read here which may enlighten me further.

Re: takeoff, if the thrust line goes through the COG, I just need to go straight up.
Everything will come into line through drag.

If there was no drag, it would reach 40 mph in about 2 seconds, when the prop tilt can be locked.

In the real world, probably a little slower.
However, in that speed range, drag is only a small proportion of the thrust, once the aircraft lines up with the airflow.

Hovering is only an attempt to seem realistic for a scale osprey aircraft. If it has a good motor-battery combo,

it only needs 2 x 9 x 5 props to lift 1.4-1.6 kg (3.5 lbs) very rapidly!

With this kind of thrust, you can do loops and rolls, possibly most 3D manoeuvres, with an Osprey model-a bit different from the usual
"flight of a pregnant duck!"
It really is a cross between a quadcopter and an airplane!

You may need to use a bit of hover tilt to compensate for the high wing loading.

-the props are not scale to the real thing, though.

With the big 1800 KV motors, (54 amps at 4S) you could use 12 inch props and get the same thrust, top speed and less amps at hover.
You don't need to worry about overloading the motor if you stick to the correct voltage range.

My 4000 mAh 30C battery would only last 2 minutes if I stayed in hover, maybe 10 minutes in cruise flying.

 

[OwenN]

This is diverging a bit from collecting Osprey tips, but:

I have found a PDF on Google on how to design a Canard, where the front wing is well forward.

1) For glide stability, the cog must be a little in front of the neutral point.

This also moves forward under the influence of fuselage lift.

2) the front wing usually has a high aspect ratio, and is more cambered than the rear wing.
it may also have a positive angle of attack relative to the rear wing.
This ensures that the front wing stalls before the rear wing, yet is still capable of a high wing loading,
to support a forward COG.
-----

My design violates most sensible principles, and will rely on:
1) gyroscopic trimming;
2) Propeller blow-over effects giving the main wing a high tolerance to stalling.
3) a lot of thrust, to assist stall recovery.

I shall experiment a bit with COG trim, to see if my calculation of the neutral point
is a survivable place for the COG when under gyro trim control.

I could fly slowly and close to the ground. It is most likely that I will find find an intermediate speed where the front wing stalls,
and the front pitches into the ground.

I can avoid this by flying at a 50 degree or so angle, and relying on plane drag and
prop-assisted control surface authority, as well as prop-assisted "hover".

Pitch-down tests can be done at a higher altitude to allow room for acceleration and recovery.

Another possible result could a be a blow-over, where the front pitches up and stalls. It shouldn't proceed to a complete blow-over and loss of control.

I have made the front elevator surfaces very large, so blowback tendencies can be cancelled quickly by the gyro trim.

 

[JasonK]

at this point, it sounds like you don't plan on building a Osprey/vtol in behavior.

 

[OwenN]

Yes I have become interested in yet another arrangement, so will be abandoning the tilt-motors for now.
I am going back to my other thread about wing loading.