Rudder experiments with flex-wing hang gliders

Interesting experiments: adding a controllable rudder and other yaw devices to 4 flex-wing hang gliders

Steve Seibel
www.aeroexperiments.org

October 13, 2006 edition

 

Overview:

I've added an experimental controllable rudder to these hang gliders: Wills Wing Spectrum, Airborne Blade, Wills Wing Skyhawk, Wills Wing Raven.

As another form of experimental yaw control, I've also experimented with deploying wingtip drag devices of various sizes from the wingtips of these same hang gliders. These wingtip drag devices were simply small drogue chutes that were deployed in flight to trail behind the wingtips.

In all cases, deflecting the rudder to the left (or deploying a drag device from the left wingtip) caused the nose of the glider to yaw to the left in relation to the actual direction of the flight path and relative wind, creating a right-to-left sideways component in the airflow, so that a yaw string (telltale) remained deflected to the left for as long the rudder was deflected (or for long as the wingtip drag device was deployed).

In most cases--with the only exception being in the low-airspeed, VG-tight corner of the flight envelope-- deflecting the rudder to the left (and thus creating a right-to-left sideways component in the airflow over the glider) created a roll torque toward the right. This is "backwards" from the roll torque that we create when we deflect the rudder to in most "conventional" aircraft. However, this is consistent with the "backwards" roll torque that we observed during experiments with the modified Zagi with a large amount of anhedral, as described in "Interesting experiments: Zagi RC glider with variable anhedral/dihedral geometry, and rudder".

In the experiments with the flex-wing hang gliders, this "backwards" roll torque in response to the rudder, or "negative coupling between yaw (slip) and roll", was always much stronger at high airspeeds (low angles-of-attack) than at low airspeeds (high angles-of-attack). We also saw the same dependence on airspeed or angle-of-attack in the experiments with the modified Zagi-- when we gave the Zagi enough anhedral to create a "backwards" roll torque in response to the rudder, or a "negative coupling between yaw (slip) and roll", this negative coupling was always strongest in the high-airspeed, low-angle-of-attack part of the flight envelope. With just the right amount of anhedral, the modified Zagi could be made to exhibit a "backwards" yaw-roll coupling over most of the flight envelope but a "normal" yaw-roll coupling in the low-airspeed, high-angle-of-attack corner of the flight envelope.

All of this is consistent with our understanding of the interplay between the aerodynamic effects of sweep and anhedral as described in "Competing effects of sweep and anhedral".

It is quite interesting that the flex-wing hang gliders tested in these experiments proved to have enough anhedral to create a "backwards" roll torque or "negative coupling between yaw (slip) and roll" over most or all of their flight envelopes. It is also quite interesting that the "backwards" roll torque or "negative coupling between yaw (slip) and roll" was stronger with the VG off than with the VG on. We'll explore this in more detail in "Interpreting in-flight observations: roll torque created by the combined effects of anhedral and sweep in flex-wing hang gliders, VG off versus VG on, high airspeed versus low airspeed".

Details:

In what we'll call the "type 1" experiments, I deflected the rudder (or deployed a wingtip drag device), and then made roll inputs as needed to stabilize the glider's heading as well as the glider's bank angle, and then noted the roll input that I needed to maintain to prevent the glider from rolling to the left or the right as the glider flew along a straight-line (non-curving) flight path. In other words, I noted the roll input that I needed to maintain to prevent the flight path from curving to the left or the right.

In all cases except the Airborne Blade with VG full on, when I deflected the rudder to the left (or deployed a drag device from the left wingtip), I found that I needed to steadily maintain a left roll input to neutralize the roll torque from the rudder or drag device, so that the bank angle remained constant at whatever bank angle was required to prevent the flight path from curving. In other words, deflecting the rudder to the left (or deploying a drag device from the left wingtip) created a right roll torque. This roll torque was much more pronounced at high airspeeds (low angles-of-attack) then at low airspeeds (high angles-of-attack).

A ranking of the strength of the "backwards" roll torque created by deflecting the rudder (or deploying a wingtip drag device), from greatest roll torque to least roll torque:

Airborne Blade with VG off--Wills Wing Skyhawk--Wills Wing Raven--Wills Wing Spectrum--Airborne Blade with VG on

All the gliders except the Airborne Blade with VG full on did show a definite "wrong-way" roll response to the rudder or wingtip drag device throughout the entire flight envelope, except in some cases at speeds very near stall (below min. sink).

In the particular case of the Airborne Blade with VG full on, the rudder (or wingtip drag device) did create a mild roll response in the "normal" direction at low airspeeds. Pulling the bar perhaps 6 to 10 inches in from trim created a neutral roll response to the rudder (or wingtip drag device), and the bar had to be pulled in rather far to create a strong "backwards" roll response to the rudder (or wingtip drag device).

(Note that the bank angle and the required roll input are completely separate things--for example if an aircraft is banked to the right of wings-level, it's quite possible for the pilot to need to apply either a steady left roll input, or a steady right roll input, or no roll input, in order to hold the bank angle constant.)

With all the gliders, the bank angle that was required to make the glider fly in a straight line when the rudder was deflected (or the wingtip drag device was deployed) was very near wings-level. With all the gliders, when the rudder was deflected to the left, a very slight left bank was required to yield a straight-line (constant-heading) flight path. With all the gliders, when a drag device was deployed from the left wing tip, a very slight right bank was required to yield a straight-line (constant-heading) flight path. (The reasons for these particular flight characteristics will be explored at length elsewhere in the Aerophysics Exploration Pages.)

In what we'll call the "type 2" experiments, I deflected the rudder (or deployed a wingtip drag device), and then made roll inputs as needed to stabilize the glider's heading as well as bank angle, and then I relaxed my roll input and noted in what direction the glider tended to roll. In all cases the results of the "type 2" experiments were consistent with the "type 1" experiments. For example, we noted that all the gliders (except the Airborne Blade with VG full on at low airspeeds) showed a "wrong-way" roll response to the rudder or wingtip drag device throughout the entire flight envelope. Deflecting the rudder to the left, or deploying a drag device from the left wingtip, created a net roll torque to the right. In all these cases, when I deflected the rudder to the left (or deployed a drag device from the left wingtip), and then made a left roll input as necessary to stabilize the bank angle and keep the glider flying in a straight line, and then relaxed the left roll input, the glider did in fact roll into a right bank. This caused the flight path to curve to the right, so that the glider ended up in a tightening right turn, with an increasing right bank angle. The rate of the "wrong-way" rolling motion was strongly dependent on the airspeed, with higher airspeeds creating much higher roll rates. In the particular case of the Airborne Blade with VG full on at low airspeeds, when I deflected the rudder to the left (or deployed a drag device from the left wingtip), and then made a right roll input as necessary to stabilize the bank angle and keep the glider flying in a straight line, and then relaxed the right roll input, the glider slowly rolled into a left bank, and ended up in a left turn.

In these "type 2" experiments, the reason that I first deployed the rudder or wingtip drag device, and then stabilized the glider's heading, and then relaxed my roll input to see how the glider would respond to the auxiliary yaw device, was that deflecting the rudder or deploying the wingtip drag device involved some significant physical exertions on my part, which were likely to produce accidental roll inputs. It wasn't normally possible to avoid making any roll inputs on the control bar while deflecting the rudder or deploying the wingtip drag devices.

In the particular case where enough VG was applied to give the glider a very mild "normal" or positive response to the rudder or wingtip drag device while flying at trim with the wings level, the glider would then show a "backward" or negative response to the rudder or wingtip drag device when the wings were not level, which would generally cause the glider to end up in a turn that was opposite to the opposite direction of the yaw input. We saw the same dynamics with the variable-geometry Zagi. This makes sense when we consider the following points: 1) For a given trim setting, an aircraft will fly at a lower angle-of-attack in a turn than in wings-level flight. 2) An aircraft with anhedral and sweep will show a less positive, or more negative, coupling between yaw (slip) and roll at low angles-of-attack than at high angles-of-attack.

In what we'll call the "type 3" experiments, I replaced one of the rudder control lines with a bungee cord so that I could pre-tension the line in flight. Then, when I released the restraining line, the rudder would quickly snap to the full-deflected position. The purpose of this experiment was to allow me to look at the net roll torque that arose as the nose was actively yawing to a new orientation in space, both with respect to the actual direction of the flight path and airflow at any given moment, and in relation to the external world. This active yawing motion would create a difference in airspeed between the two wingtips that could affect the glider's tendency to roll to the left or to the right when the rudder was deflected. However, I found that the yawing motion that arose when the rudder snapped to full deflection was so modest and so brief that it was impossible to assess the roll torque arising from this yawing motion. It was only several seconds after the rudder was deflected that the roll torque created by the rudder became apparent, as described in the "type 1" and "type 2" experiments. However, when I carried out the equivalent of the "type 3" experiments in a Zagi RC glider modified to have anhedral and a controllable rudder, I could create a much more vigorous "active" yawing motion, and in no case was the direction of the roll torque arising from a rapid "snap" of the rudder to full deflection different from the direction of the roll torque arising from steadily holding a smaller rudder deflection, at any given airspeed. In other words, in the Zagi experiments, at some given airspeed, if a steady, modest rudder deflection to the left made the glider roll to right, then a rapid snap of the rudder to full left deflection had the same result. However it is undoubtedly the case that some aircraft with very mild anhedral, that normally show a very mild right roll torque in response to a left rudder input, can be made to "snap roll" to the left in response to a strong left yawing motion that greatly increases the airspeed of the right wingtip and decreases the airspeed of the left wingtip.

In what we'll call the "type 4" experiments, which I only carried out in the Wills Wing Spectrum and Airborne Blade (with VG both loose and tight) I tensioned the rudder in the centerline position so that it served as a fixed vertical fin, and rolled the glider quickly from wings-level into a steep bank several times. Then I released the rudder lines to let the rudder float freely in the airflow, so that it could not create any yaw torque, and rolled the glider quickly from wings-level into a steep bank several times. I wasn't able to detect any definite change in the gliders' roll responsiveness, or any definite change in the amount of sideslip due to adverse yaw that I saw in the center-mounted yaw string during the rolling maneuvers.

In what we'll call the "type 5" experiments, which I only carried out in the Wills Wing Spectrum and Airborne Blade (with VG both loose and tight), I deflected the rudder to the left and locked it there, which made the center-mounted yaw string stream to the left. I then rolled the glider quickly from wings-level into steep left and right banks several times, while noting the roll rate and the yaw string deflection. The results were always consistent with what I saw in the "type 1" and "type 2" experiments. For example, in the Wills Wing Spectrum and the Airborne Blade with the VG loose, since I had to maintain a steady left weight-shift roll input to prevent the glider from rolling to the right (with the rudder deflected to the left) at all airspeeds, I found that I could make the glider roll to the right faster than I could make the glider roll to the left. With the Airborne Blade with the VG tight at low airspeed, the result was the opposite: I could make the glider roll to the left faster than I could make the glider roll to the right. With all the gliders regardless of VG setting, since the yaw string deflected to the left when the glider was wings-level (with the rudder deflected), I saw that adverse yaw created a much larger deflection of the yaw string (toward the high side of the turn) when I rolled the glider to the right than when I rolled the glider toward the left.

More content will be added to this page in the future, including photos of the wingtip drag devices. For now, peruse the "Pool of images for the Aerophysics Exploration Pages" for photos of these devices.

We'll further explore the significance of these observations in "Interpreting in-flight observations: roll torque created by the combined effects of anhedral and sweep in flex-wing hang gliders, VG off versus VG on, high airspeed versus low airspeed".

 

Advance to "Interpreting in-flight observations: roll torque created by the combined effects of anhedral and sweep in flex-wing hang gliders, VG off versus VG on, high airspeed versus low airspeed"

Skip to "The main cause of adverse yaw during rolling motions: the 'twist' in the relative wind"

Up to the Aerophysics Exploration Pages contents page

Up to the Aeroexperiments site map

Copyright © 2004 aeroexperiments.org