Asymmetrical UAV/AP - Page 11 - RC Groups
Thread Tools
Mar 01, 2012, 12:42 PM
Lift is cheap - Drag sucks
Tom Harper's Avatar
Should be able to fly again Saturday. Hopefully the wind will settle down a bit by then.

Been working on the data logger so may get some flight data. Also, I'd like to try DPATE's idea of a tip plate on the right wing only. Interesting to see what that would do to the roll.

Sign up now
to remove ads between posts
Mar 01, 2012, 01:00 PM
Grad student in aeronautics
I look forward to the results!
You can call me David :-)
(unless there are too many other Davids.....)
Apr 21, 2012, 06:53 PM
Lift is cheap - Drag sucks
Tom Harper's Avatar


Beautiful flying day at last.

Removed the landing gear for the second flight. Hand launch is easy, but pitch is much more sensitive with the CG moved further back. Looks neat and flys great with no gear.

Flew a couple of circles under full Automatic. Handles well but needs tuning. High Roll gain resulted in some hunting.

Brought it home in one piece. Need to plank the underside and extend the skids to keep the dirt and rocks out of the fuselage. Will also add some width to the elevator. Powell says there's not enough control at low speed.

Some clean up and tuning and we will be back to photographing trails.

Apr 21, 2012, 10:29 PM
internet gadfly
nmasters's Avatar
I'm glad you're getting results you like. Have you determined the position of the CG relative to the leading edge of MAC? There's hardly ever any advantage to placing the CG more than 30% aft of the leading edge at MAC. The tail heaviness at low speed you describe could indicate either not enough tail authority or, more likely, too short a static margin.

Apr 22, 2012, 02:54 PM
Lift is cheap - Drag sucks
Tom Harper's Avatar

I was referring to the fact that the cg moved back when the gear was removed. It had been far forward. Now it is about 20% back from the LE.

Apr 22, 2012, 05:39 PM
Lift is cheap - Drag sucks
Tom Harper's Avatar


Planked the underside and extended the Spruce runners.

Will cover it with silkspan and Nitrate dope.

Apr 28, 2012, 01:40 PM
Lift is cheap - Drag sucks
Tom Harper's Avatar

Auto Flights

Ran some autopilot tests this morning with interesting results.

In manual the model is still reluctant to turn left, even with considerable left thrust.

Right circle under automatic control:

Speed: 60 KMH
Current: 7 Amps
Throttle: 40%
Roll bias: -2.49

Left circle under automatic control:

Speed: 60 KMH
Current: 17 Amps
Throttle: 62.5%
Roll bias: 6.85

Roll bias is an integrated error that the autopilot feeds in as an offset. It looks like the motor is hauling it around to the left. Any thoughts?

Everything went smooth. We could have flown a longer, non-circular path, but decided that data logs and an airplane in one piece was a better deal.

Apr 29, 2012, 12:58 PM
Lift is cheap - Drag sucks
Tom Harper's Avatar

Like Sparky Paul says - designs are developed

As a sport flyer, I'd leave it alone. But, I'm interested in the details and this model has a job to do. So, the problems are:

1) Used bullet connectors for the 4000mah Lipo. The spring tension on the male bullet weakens over time. Even when fully inserted it can lose conduction under high current loads. We noticed this in an occaisional brief loss of control and the autopilot stopped logging data after a power interruption. Will change back to Deans connectors.

2) The chord lines of the wing and stabilizer sections are set at zero incidence relative to the thrust line. At low speed there is not enough elevator control to flair the glide. The solution has been to maintain speed on landing, but a little more control won't hurt. Will increase the elevator chord by 50%.

3) The stabilizer has a tapered planform so the spanwise spar is tapered. The elevator is top hinged with tape. The taper varies the hinge radius spanwise from base to tip. That causes the tape hinge to stretch and distort over time. The cure is to center hinge the elevator. Not yet sure what to do.

4) The elevator is not an even match to the stab which leaves a ridge on the underside. Will rebuild the elevator for a better fit.

5) The motor sometimes picks up dirt on landing. Working on a cowl made from one of those little Coke cans.

6) The two graphs below show the pitch and roll angles required to automatically loiter in a circle. The center of the circle is a the lon/lat of a pre-programmed waypoint. The first record is shorter because of a power glitch.

The first graph is for a right circle. The model goes into a steep bank angle until it aligns itself with the circumference of the circle identified by the first waypoint. Once on the circumference path, the roll oscillations damp out and the pitch angle is near zero. The circle radius is 100 m. The power setting for 60kph is 40%.

The second graph is a left circle of 125 m. The model acquires the circumference of the circle but the roll oscillations do not damp out and a positive pitch angle is maintained. The power setting for 60kph is 65%.

It will take some time to understand this.

Comments welcome.

Last edited by Tom Harper; Apr 29, 2012 at 01:13 PM.
Apr 29, 2012, 02:34 PM
Ascended Master
Sparky Paul's Avatar
Tom, stretching out the left circle, the period appears similar to the right circle, but something occurs at the settling point which disturbs the plane and gets the oscillation going again.
What's the heading at that point?
Apr 29, 2012, 03:48 PM
Lift is cheap - Drag sucks
Tom Harper's Avatar



I added heading/10 to make it fit.

It looks like it was weaving around the circle. I haven't been able to plot it on Google Earth yet.

Apr 29, 2012, 05:42 PM
Ascended Master
Sparky Paul's Avatar
I was wondering if the abrupt change from a 360 heading to 0 had some influence on the roll... I don't really see that there now.
Apr 29, 2012, 06:25 PM
Lift is cheap - Drag sucks
Tom Harper's Avatar

Atto does a 16 point least squares smoothing so it somehow handles the 360/0 problem.

Any other thoughts?

I think there is a clue in the increased pitch angle. That could be to make up for the loss of lift caused by the aileron deflection. Also the huge increase in power. The Autopilot increases power to maintain the same airspeed as in the right turn.

Apr 30, 2012, 11:19 AM
Lift is cheap - Drag sucks
Tom Harper's Avatar

Epiphany in the shower

I studied the planform long and hard and felt like all of the rigging should have yielded a left turn. So, I let it rest for a while and then it began to make sense:

1) Oscillation - It is critically damped in a right turn but oscillates in a left turn. Well, duh! All of that left thrust increases system gain to the left and critically damps to the right.

2) Power - So, why the power increase? Speed, weight and altitude are constant so lift must be constant. In a left turn the single left side aileron decreases lift on that wing panel. Two things must happen in response: the autopilot must increase angle of attack to maintain lift and it must correct for adverse yaw. That results in a significant drag increase and a compensating increase in power. More power = more system gain resulting in oscillation.

3) Cure - Remove the left thrust - Add a right side aileron - Increase rudder area.

Also I need to make sure that the rudder servo is correctly phased. If out of phase it would add to the problem.

Apr 30, 2012, 12:17 PM
HerkS's Avatar
Tom, you are way beyond me with the autopilot and all at this point.

But just lurking along, I wondered if the plane may be trimmed out with significant yaw. If it is flying sideways because of either the rudder trim setting, or off-center lateral drag distribution, it seems to me that it could cause some of these issues.
Apr 30, 2012, 12:35 PM
Lift is cheap - Drag sucks
Tom Harper's Avatar


The flight log doesn't show yaw, but I agree that's probably part of the problem. I believe it's going to be a process of elimination. From the above I better understand the implications of the one sided aileron. I'll remove the left thrust, add the right aileron, move the cg to the half span point and go from there.

Last edited by Tom Harper; Apr 30, 2012 at 12:49 PM.

Thread Tools

Similar Threads
Category Thread Thread Starter Forum Replies Last Post
Discussion Asymmetrical Flybar Travel on the Big Lama - Related to Death Dives? rampa202 Coaxial Helicopters 52 Nov 26, 2011 10:58 PM
Discussion AP Full Arducopter UAV Gasser Build. ben1101 Aerial Photography 6 Oct 31, 2011 06:56 AM
Build Log AP UAV like plane , using 10g BL Prime_8 Foamies (Scratchbuilt) 9 Jun 10, 2011 01:51 AM
Small UAV for AP? New project! icebear Aerial Photography 47 Jun 12, 2005 09:51 PM
Asymmetrical AP-plane out of depron century_series Foamies (Kits) 31 Sep 09, 2004 03:49 AM