RC Groups

RC Groups
    VTOLs
        Discussion The Cube

#1 ilektron Feb 02, 2013 12:29 AM

The Cube
 
1 Attachment(s)
I'm on a kick on finishing things right now, so I figured I'd finally finish off my flying borg cube.

I still need to add feet, so it doesn't just flop over when the motor spins up, and finish mounting the FC. My controller board needs some work on the attitude estimation, so I think this will make a good testbed to practice with. It should be fairly sturdy and be able to take a crash. Even better, I should be able to transition to regular flight, once I figure out a way to mark orientation without detracting from it.

It is a 12x12x12" cube, with a single DT750 and some other cheapo stuff I had laying around. Hopefully I'll be able to get it in the air in the next week, and then on to my tilt-rotor!

#2 ilektron Feb 02, 2013 10:31 AM

4 Attachment(s)
Everything is done except for mounting tweaking my autopilot.

I would love it if this thing could fly like a plane beautifully.

Question for those out there that have built single-copter-ish stuff, how difficult is it to get these things to hover? I fear that the airframe will be to reactive for an autopilot, but I know that I've seen it done. Does they take a lot of tweaking to get it to get off the ground?

#3 matwelli Feb 02, 2013 02:10 PM

Looks very cool, love the idea.
Not sure that you have enough area on those two vanes to control yaw........but I hope you do.
How far away from the vertical cog are the vanes ?
Best way I have found to tune (mine was a dual - not single copter) is to run a piece of cord horizontally, with a bit of slack, and attach the frame to that just above its cog (I use a bit of carbon tube running through the frame) then you have a way to tune it without damaging anything

The first 10 seconds of this video show what I am trying to say

Spazzy second flight (3 min 44 sec)

#4 ilektron Feb 02, 2013 09:51 PM

I'll have to make some rig to tune it. The CG is about 1/3 chord. The vanes are a little less than 1/4 chord, if I remember correctly. My worry isn't yaw, I fear it will be too reactive in the other axis. I had to replace the the rudder servo, but I found my FC and will start fixed the code for it.

#5 Peter Seddon Feb 03, 2013 06:40 AM

Be prepared to adjust the CofG towards the bottom. Make sure it is balanced about the longitudinal axis.

I don't think your legs are long enough for the air to escape properly. I would go for much much longer legs.

There is a great deal of discussion here http://www.rcgroups.com/forums/showthread.php?t=1454102 and
http://www.rcgroups.com/forums/showthread.php?t=1516652
You will have to read through to find the bits that are relevant to your design.

Peter

Quote:

Originally Posted by ilektron (Post 24016434)
I'll have to make some rig to tune it. The CG is about 1/3 chord. The vanes are a little less than 1/4 chord, if I remember correctly. My worry isn't yaw, I fear it will be too reactive in the other axis. I had to replace the the rudder servo, but I found my FC and will start fixed the code for it.


#6 ilektron Feb 04, 2013 10:29 AM

Quote:

Originally Posted by Peter Seddon (Post 24018810)
Be prepared to adjust the CofG towards the bottom. Make sure it is balanced about the longitudinal axis.

I don't think your legs are long enough for the air to escape properly. I would go for much much longer legs.

There is a great deal of discussion here http://www.rcgroups.com/forums/showthread.php?t=1454102 and
http://www.rcgroups.com/forums/showthread.php?t=1516652
You will have to read through to find the bits that are relevant to your design.

Peter

Thanks for the tips. I'm hoping I can pop it off the ground before the leg length makes it too much of an issue, but I fear you may be right.

If I adjust the CG down, won't it compromise normal flight?

#7 matwelli Feb 06, 2013 01:39 AM

Quote:

Originally Posted by Peter Seddon (Post 24018810)
Be prepared to adjust the CofG towards the bottom
Peter

why ? the higher the CG , and further from the control vanes, the more control the vanes will give, thru experiments i have found with tilting motors 70+mm seperation is a minimum, and with vanes almost double that to be really effective

dualcopter (1 min 19 sec)

#8 ilektron Feb 06, 2013 09:54 AM

Quote:

Originally Posted by matwelli (Post 24048667)
why ? the higher the CG , and further from the control vanes, the more control the vanes will give, thru experiments i have found with tilting motors 70+mm seperation is a minimum, and with vanes almost double that to be really effective

That is what I would intuit. The extra distance from CG would give more leverage, aka more torque available to rotate the aircraft.

I had to pull a servo and need to replace it. Hopefully this weekend I'll figure out if I have large enough vanes

#9 corocopter Feb 06, 2013 12:39 PM

Two things to keep in mind with regards to CG placement.

1. As already mention the further the distance the control surface(s) are from the Cg location the more control effectiveness is increased (distance x control surface volume = control force moment. Increasing control surface volume (size) and or the distance, increases control effectiveness. If the control vane is too close to the CG you may experience control reversal.

2. Since your craft has lifting surfaces you must maintain the CG location ahead of the Mean Aerodynamic Center, (MAC) which would be at 25 % aft from the leading edge of your current shape. I would strongly suggest you set the CG location at exactly 22.5 % aft from the leading edge, otherwise you'll have a overly sensitive craft if the cg is further aft, or it will be nose heavy if further forward (under responsive).

Anti-torque control will require off-set deflection of at least 2 vanes at all times.

I would "second the motion" to suspend the craft from a string system for testing.

Later make sure your landing gear legs are extra long, ground effect will completely destroy the ability to control the craft near the ground.

You could hand launch and dump it into tall grass to land if you want to save the weight of landing gear.

Kelly

#10 ilektron Feb 06, 2013 01:35 PM

Quote:

Originally Posted by corocopter (Post 24052713)
Two things to keep in mind with regards to CG placement.

1. As already mention the further the distance the control surface(s)are from the Cg location the more control effectiveness is increased (distance x control surface volume = control force moment. Increasing control surface volume (size) and or the distance, increases control effectiveness. If the control vane is too close to the CG you may experience control reversal.

2. Since your craft has lifting surfaces you must maintain the CG location ahead of the Mean Aerodynamic Center, (MAC) which would be at 25 % aft from the leading edge of your current shape. I would strongly suggest you set it at exactly 22.5 % aft from the leading edge, other wise you'll have a overly sensitive craft if the cg is further aft, or it will be nose heavy if further forward (under responsive).

Anti-torque control will require off-set deflection of at least 2 vanes at all times.

I would "second the motion" to suspend the craft from a string system for testing.

Later make sure your landing gear legs are extra long, ground effect will completely destroy the ability to control the craft.

Kelly

That is a great answer. Thanks for your expertise. I don't think I'll be able to get the CG that far forward in the current config. However, it will be flown with stabilization at all time. I guess the question will be, how fast is my autopilot compared to the airframe. Will it be controllable? That is the question. Is there a threshold before the airframe will become completely uncontrollable or are there too many factors to determine 'controllability'?

#11 Brandano Feb 06, 2013 01:53 PM

Stabilization can only affect control dampening. While it will make the plane slower to depart, it will also make it less responsive to control inputs. Unless you include air sensors in the control loop, gyro stabilization alone can't prevent control departure.

#12 ilektron Feb 06, 2013 02:24 PM

Quote:

Originally Posted by Brandano (Post 24053380)
Stabilization can only affect control dampening. While it will make the plane slower to depart, it will also make it less responsive to control inputs. Unless you include air sensors in the control loop, gyro stabilization alone can't prevent control departure.

Yes, but at what COG position can you reasonably avoid control departure under most circumstances?

Even your average airplane is underactuated. As long as the flight profile stays within certain bounds, you can be relatively sure that the aircraft will respond in a certain manner to a given input. Where are these thresholds when it comes to CG?

#13 Brandano Feb 06, 2013 03:00 PM

I think that these low aspect ratio flying wing types can be flown successfully with a cg as far back as 33% of the MAC, at least at moderate speedsit will be nervous, but should return to a straight path if you let go of the sticks.

#14 corocopter Feb 06, 2013 05:07 PM

Quote:

Originally Posted by ilektron (Post 24053214)
That is a great answer. Thanks for your expertise. I don't think I'll be able to get the CG that far forward in the current config. However, it will be flown with stabilization at all time. I guess the question will be, how fast is my autopilot compared to the airframe. Will it be controllable? That is the question. Is there a threshold before the airframe will become completely uncontrollable or are there too many factors to determine 'controllability'?

Here is a solution for fixing the CG Location.

Flip the motor/prop over, making it a pusher. This way the battery/controller/avionics are on top. Also put the servos up there too.

Your craft has an aspect ratio of 1:1 which is the same as a Pizza Box Flyer design which I've flown extensively using nitro power.The absolute most aft CG location without gyros was 24.5 % of MAC, but that was very touchy!

Don't rely upon silicon micro chips for the initial design criteria, use good design logic to start with for test flights, you maybe able to reach the 33% MAC that Brandano mentions after you get the P and I terms perfect, but at start up be conservative.

Your craft is far to heavy, get it down under 500 grams, you have too much lifting surface.

If you flip the motor over, make it a mid engine and cut out all excess foam, the CrazyBall design is a good example.

As already mentioned on this thread, check out the Japanese Ball Drone thread, especially the last dozen pages. I've posted a great deal of info for this sort of craft here that may also be of help.

This is a excellent project, please keep us informed and good luck.

Kelly

#15 corocopter Feb 06, 2013 05:43 PM

Flying Can UAV design
 
Here you go, check out post #332 of the Flying Can UAV...

http://www.rcgroups.com/forums/showt...&pp=10&page=34

Note: pusher design using carbon fiber polished sheet stock, the flight performance is spectacular.

A tube is stronger and lighter than a cube/box shape, easier to shape/roll to the correct size and has the correct amount of lifting surface(s).

I priced this carbon sheet stock and it was very expensive, perhaps very thin fiberglass sheet stock would approximate the performance at a much lower cost, or maybe foam would work for the initial testing?


All times are GMT -5. The time now is 02:05 PM.