D. H. Klyde and D. G. Mitchell, Recommended Practices for Exposing Pilot-Induced Oscillations or Tendencies in the Development Process, presented at USAF Developmental Test and Evaluation Summit, Woodland Hills, CA, November 16-18, 2004. (AIAA 2004-6810, STI-P-647)It has now been more than a century since the first powered flight took place on the dunes of Kill Devil Hills. An often overlooked element of those remarkable first four flights was the occurrence of longitudinal axis pilot-induced oscillations (PIO). Looking back it is now easy to see how the inherent instability of the 1903 Wright Flyer design resulted in PIO with either Orville or Wilbur at the controls. Since that time, aircraft have grown in capability and complexity. For example, the systems designed to control the aircraft have evolved from reversible cables and pulleys where the force of the pilot's inputs directly manipulates the control surfaces to modern fly-by-wire designs where software interpretations of the pilot's inputs manipulate the control surfaces. There has been, however, one constant through this evolution - the ongoing occurrence of pilot-induced oscillations, sometimes with catastrophic results. In the last 30 years almost every military or commercial aircraft designed has experienced PIO, either in the development process or in operational flight. If recent history is any indication, PIO will certainly continue to occur in the future. In this paper, status of relevant PIO criteria will be given, on-board detection and alleviation schemes will be described, and flight test methods will be reviewed. Finally, recommended practices will be outlined to expose PIO tendencies, if they exist, so that the catastrophic events can be minimized or eliminated.

D. H. Klyde and D. G. Mitchell, Investigating the Role of Rate Limiting in Pilot-Induced Oscillations, presented at AIAA Atmospheric Flight Mechanics Conference, Austin, TX, August 11-14, 2003. (AIAA 2003-5463, STI-P-611)From the Wright Flyer to fly-by-wire, the phenomenon of pilot-induced oscillations or PIO has been observed on prototype, experimental, and operational military and commercial aircraft. The introduction of irreversible control systems with surfaces driven by powered actuators brought many benefits along with increased system complexity and the introduction of additional nonlinearities. Chief among these nonlinearities are the hardware and software rate limits associated with the control surface actuators. Basic sizing tradeoffs conducted in the design process set the maximum rate of an actuator, while software rate limits are introduced to prevent overdriving the control surface when loads or structural limitations exist. As demonstrated in this paper, when operating as intended, there are usually no ill effects associated with rate limits, however, certain conditions can lead to a highly saturated condition. This results in the sudden introduction of significant added phase lags to the pilot-vehicle system. In many cases the end result is often PIO or other related loss of control events. One of the earliest well documented PIO events involving rate limiting occurred on the first flight of the X-15. This event is significant in that common linear systems analysis techniques do not reveal a susceptibility to PIO. The analysis of X-15 Flight 1-1-5 and the results of a more recent flight research program are presented in detail in this paper.

Elimination of PIO by design remains an elusive goal. Throughout the history of powered flight, as manual flight control systems have morphed and evolved, so too have the pilot-induced oscillations that often plague them. In the mid-1990’s, the US Air Force declared as part of its Unified PIO Theory program that the cumulative reduction of PIO shall be 80% via criteria, 99% via evaluation methods, and 99.95% via detection and compensation. This too remains an elusive goal. One reason lies in the lack of consensus regarding the definition of exactly what is a PIO. The intent of this paper is to coalesce the many definitions of PIO into a single entity. Then newly developed analysis techniques are applied to a flight test database to expose the unique characteristics or signature of PIO, at least as they apply to a representative mission operation — the probe-and-drogue aerial refueling task.

D. H. Klyde and D. G. Mitchell, A PIO Case Study - Lessons Learned through Analysis, presented at AIAA Atmospheric Flight Mechanics Conference and Exhibit, San Francisco, CA, August 15-18, 2005. (AIAA2005-5813, STI-P-661)Regardless of aircraft size, mission, or operational envelope, pilot-induced oscillations, or PIO, continue to plague new designs as well as those with years of fleet service. The primary interacting components of PIO remain pilot, vehicle, and trigger. The recent proliferation of unmanned aerial vehicles, that by their name eliminate the pilot, is one way to handle the problem of PIO. It is clear, however, that piloted aircraft will continue to be a mainstay well into this second century of powered flight. Considering PIO triggers, it is the opinion of the authors that the elimination of all possible triggers is not possible as there will always be some unanticipated event that can lead to PIO. Thus the trick remains designing aircraft that are resistant to PIO regardless of the presence of a trigger or the intended or unintended actions of the pilot. To this end, handling qualities engineers have worked hard to develop criteria that reliably expose susceptibility to linear and nonlinear forms of PIO. An undesirable side effect is that the user community has come to rely on these criteria to provide the required “go/no go” answer. Often the desire to get such an answer is so high that criteria may be applied in an unintended manner leading to questionable results. Although proper application of criteria is a vital step in the process, what has been lost, to a large extent, is an ability to systematically analyze the PIO susceptibility of a given configuration. In this paper a case study is presented as a means to describe various PIO analysis techniques. The analysis methods described herein are by no means comprehensive, but instead build upon the fundamental techniques for characterizing pilot-vehicle systems. Results are compared with flight test data and lessons learned are documented.