Dutch Roll Coupling

Dutch roll coupling is another case of a dynamic loss of control of an airplane because of an unusual combination of lateral-directional static stability characteristics. Dutch roll coupling is a more subtle but nev­ertheless potentially violent motion, one that (again quoting Day) is a "dynamic lateral-directional stability of the stability axis. This coupling of body axis yaw and roll moments with sideslip can produce lateral – directional instability or PIO.”[744] A typical airplane design includes "static directional stability” produced by a vertical fin, and a small amount of "dihedral effect” (roll produced by sideslip). Dihedral effect is created by designing the wing with actual dihedral (wingtips higher than the

wing root), wing sweep (wingtips aft of the wing root), or some combi­nation of the two. Generally static directional stability and normal dihe­dral effect are both stabilizing and both contribute to a stable Dutch roll mode (first named for the lateral-directional motions of smooth-bottom Dutch coastal craft, which tend to roll and yaw in disturbed seas). When the interactive effects of other surfaces of an airplane are introduced, there can be potential regions of the flight envelope where these two contributions to Dutch roll stability are not stabilizing (i. e., regions of negative static directional stability or negative dihedral effect). In these regions, if the negative effect is smaller than the positive influence of the other, then the airplane will exhibit an oscillatory roll-yaw motion. (If both effects are negative, the airplane will show a static divergence in both the roll and yaw axes.) All aircraft that are statically stable exhibit some amount of Dutch roll motion. Most are well damped, and the Dutch roll only becomes apparent in turbulent conditions.

The Douglas DC-3 airliner (equivalent to the military C-47 airlifter) had a persistent Dutch roll that could be discerned by passengers watch­ing the wingtips as they described a slow horizontal "figure eight” with respect to the horizon.

Even the dart-like X-15 manifested Dutch roll characteristics. The very large vertical tail configuration of the X-15 was established by the need to control the airplane near engine burnout if the rocket engine was misaligned, a "lesson learned” from tests of earlier rocket-pow­ered aircraft such as the X-1, X-2, and D-558-2. This led to a large sym­metrical vertical tail with large rudder surfaces both above and below the airplane centerline. (The rocket engine mechanics and engineers at Edwards later devised a method for accurately aligning the engine, so that the large rudder control surfaces were no longer needed.) The X-15 simulator accurately predicted a strange Dutch roll characteristic in the Mach 3-4 region at angles of attack above 8 degrees with the roll and yaw dampers off. This Dutch roll mode was oscillatory and stable with­out pilot inputs but would rapidly diverge into an uncontrollable pilot – induced-oscillation when pilot control inputs were introduced.

During wind tunnel tests after the airplane was constructed, it was discovered that the lower segment of the vertical tail, which was oper­ating in a high compression flow field at hypersonic speeds, was highly effective at reentry angles of attack. The resulting rolling motions pro­duced by the lower fin and rudder were contributing a large negative dihedral effect. Fortunately, this destabilizing influence was not enough

to overpower the high directional stability produced by the very large vertical tail area, so the Dutch roll mode remained oscillatory and stable. The airplane motions associated with this stable oscillation were com­pletely foreign to the test pilots, however. Whereas a normal Dutch roll is described as "like a marble rolling inside a barrel,” NASA test pilot Joe Walker described the X-15 Dutch roll as "like a marble rolling on the out­side of the barrel” because the phase relationship between rolling and yawing were reversed. Normal pilot aileron inputs to maintain the wings level were out of phase and actually drove the oscillation to larger magni­tudes rather quickly. The roll damper, operating at high gain, was fairly effective at damping the oscillation, thus minimizing the pilot’s need to actively control the motion when the roll damper was on.[745]

Because the X-15 roll damper was a single string system (fail-safe), a roll damper failure above about 200,000 feet altitude would have caused the entry to be uncontrollable by the pilot. The X-15 envelope expansion to altitudes above 200,000 feet was delayed until this problem could be resolved. The flight control team proposed installing a backup roll damper, while members of the aerodynamic team proposed removing the lower ventral rudder. Removing the rudder was expected to reduce the direc­tional stability but also would cause the dihedral effect to be stable, thus the overall Dutch roll stability would be more like a normal airplane. The Air Force-NASA team pursued both options. Installation of the backup roll damper allowed the altitude envelope to be expanded to the design value of 250,000 feet. The removal of the lower rudder, however, solved the PIO problem completely, and all subsequent flights, after the initial ventral-off demonstration flights, were conducted without the lower rudder.[746]

The incident described above was unique to the X-15 configura­tion, but the analysis and resolution of the problem is instructive in that it offers a prudent cautioning to designers and engineers to avoid designs that exhibit negative dihedral effect.[747]