Yaw-roll oscillations in flex-wing hang gliders
August 8, 2007 edition
I've found that some flex-wing hang gliders will enter into, and sustain, yaw-roll oscillations in the absence of any pilot roll input.
Here is one procedure I followed that produced these oscillations: I pulled the control bar in to at least the mid-chest area and then I "locked" the control bar in this position, keeping my body exactly on the centerline of the control bar.
Here is an alternate procedure I followed that produced these oscillations: I pulled the control bar in to at least the mid-chest area and then I kept the control bar in that position in the pitch direction while exerting no roll forces--I allowed myself to "float" from side to side of the centerline. (In some cases I closed my eyes to be sure that I was not giving unconscious roll inputs in response to changes in bank angle.) When I exerted no force on the control bar in the roll direction, I never found myself tending to hang more than a few inches from the centerline. This procedure produced the same results as when I "locked" myself in a fixed position in roll as well as pitch.
In either case, some gliders would slowly enter into a series of increasing yaw-roll oscillations, where the nose would describe a "horizontal figure-8" on the horizon as the flight path curved upward or downward in concert with the changes in bank angle. Typically the glider would start to bank in one direction, the glider would sideslip (presumably due to adverse yaw from the rolling motion) as the bank angle changed increased, the flight path would curve downward (because the airspeed and lift force lagged behind the increase in bank angle), and then at some point the nose would be rising again and the bank angle would be decreasing, until the nose had risen above the horizon and the bank angle had rolled through wings-level and into a reversal, and the process would continue in the new direction.
Sometimes these oscillations would involve rather dramatic bank angles and slip angles. This was particularly true with my Icaro Laminar R-12.
No glider entered into these oscillations when flown with the bar near trim-- the bar always had to be pulled in significantly. When the bar was near trim, all the gliders would generally tend to slowly drop one wing and enter into a descending banked turn. The more pulled-in the bar, the greater the tendency for these yaw-roll oscillations to start, and the more pronounced they would become.
With all the gliders that were prone to these yaw-roll oscillations, they were always more pronounced with VG off than with VG on. For example in some cases with VG off the oscillations would start when the bar was pulled in to mid-chest but with VG on the oscillations would not start until the bar was pulled in to the beltline.
In some cases when flown at some particular bar position, a glider would not enter into yaw-roll oscillations by itself (especially in smooth air) but if I gave a few reversing roll inputs I could start a yaw-roll oscillation. At this point the oscillation would continue and increase even if I "locked" myself in place on the control bar as described above.
Here are some gliders that exhibited a strong tendency toward these yaw-roll oscillations in the absence of any pilot roll input: Wills Wing Spectrum, Airborne Blade, Icaro Laminar R-12. Note that these gliders are also all prone to yaw-roll oscillations during aerotow, more so with the VG off than with some VG applied (if present).
Here are some gliders that exhibited little or no tendency toward these yaw-roll oscillations in the absence of any pilot roll input, even with the bar well pulled-in: Wills Wing Falcon, Aeros Stealth KPL. Note that these gliders also track very well during aerotow, with little or no tendency toward yaw-roll oscillations.
It is a misconception that yaw-roll oscillations in hang gliders are always pilot-induced. Naturally, yaw-roll oscillations can be damped out by appropriate pilot input, and many expert pilots do this without any conscious effort or awareness, which has given rise to the idea that faulty pilot inputs are the sole cause of the oscillations. But many hang gliders will enter into yaw-roll oscillations by themselves, or sustain any yaw-roll oscillation that the pilot begins, even in the absence of any further pilot input. This is true both in free flight and during aerotow. In fact I suspect it is rare for a hang glider pilot to experience a true "PIO" in roll that is driven purely by the pilot's mis-timed roll inputs. (I have experienced this once, in a Wills Falcon after being used to flying stiffer, heavier gliders, and it felt quite different from the yaw-roll oscillations I've experienced during aerotow or during high-speed free flight or during the experiments described here.)
I suspect that the yaw-roll oscillations we see during aerotow are closely related to the yaw-roll oscillations that we see during free flight at low angles-of-attack (high airspeeds.) In both cases the oscillations are more evident with VG off than with VG on, and in both cases the oscillations are more evident with the bar pulled in than with the bar closer to the position for a best-glide or min-sink angle-of-attack.
As described elsewhere on this website, on a wing with both anhedral and sweep, the wing will show more effective anhedral at low angles-of-attack than at high angles-of-attack. Classic "Dutch Roll" oscillations are generally driven by the interaction between an aircraft's effective dihedral and the aircraft's yaw stability. Since yaw-roll oscillations in hang gliders are most pronounced at low angles-of-attack where effective anhedral is greatest (or effective dihedral is least), this suggests they may have little in common with classic "Dutch Roll" oscillations. What is the "restoring force" that keeps a hang glider's yaw-roll oscillation going? Why doesn't the glider just diverge into a spiral dive? Does aeroelasticity play an important role?
What is the significance of the fact that these yaw-roll oscillations are more prone to occur when the VG is loose than when the VG is tight? Is this due to the way the VG changes the glider's aeroelasticity, or due to the way the VG changes the glider's effective anhedral, or both?
Elsewhere on this website we give experimental evidence, and visual observations of geometry, in support of the idea that a flex-wing hang glider actually has more anhedral with the VG loose than with the VG tight.
At the start of this article we outline two different procedures that both yielded yaw-roll oscillations in some hang gliders. The fact that both procedures produced the same results, and the fact that only minimal muscle forces were needed to keep my body "locked" on the glider centerline in the roll direction, and the fact that my body never tended to move more than a few inches from the glider centerline, suggests that the "pendulum" roll stability effect contributed by a fixed pilot mass was minimal. This also suggests that aerodynamic sideforces created by a sideways airflow component over the wing--which would have acted high above the CG and which would have been the root cause of any "pendulum" roll stability effect that existed when the glider was sideslipping-- were minimal.