Numerical Investigation of Aeroservoelastic Rotorcraft-Pilot Coupling

This paper describes the current status and achievements of a research activity for the investigation of rotorcraftpilot coupling (RPC) governed by aeroelastic issues, to single out possible sources and occurrences of the phenomenon, focused on the determination of possible design solutions towards RPC-free helicopters and rotorcraft in general. A brief presentation of the numerical analysis methodologies used by the two research teams introduces the description of the benchmark used in this analysis, which is completed by a summary of the results obtained in recent investigations. 1. BACKGROUND AND SCOPE The interaction between the aircraft and the pilot represents a potential threat to the airworthiness of any kind of flying vehicles. The history of aircraft has seen the appearance of unfavorable Aircraft-Pilot Coupling (APC) from the very beginning, when aircraft controllability was the issue to overcome to achieve human flight. Subsequent progress in aircraft design led to higher performance and broader flight envelope; at the same time, the amount of workload for the pilot and the crew has been reduced by providing substantial automation in aircraft control. However, the need for the crew to deal with other tasks reduced the time share of their attention that is dedicated to flying the aircraft. This, in critical flight conditions, gives room for unfavorable APCs when some unexpected event triggers an abnormal aircraft behavior. Unfavorable APCs are defined as “. . . inadvertent, sustained aircraft oscillations which are a consequence of an abnormal joint enterprise between the aircraft and the pilot. . . ” [1] This means that the presence of the pilot plays a fundamental role in the phenomenon, but other players participate: the vehicle, the mission task, and a trigger, namely an abnormal event that alters the coupled vehicle-pilot system’s conditions, initiating the adverse phenomenon. APCs are undesirable and sometimes hazardous phenomena associated with less-than-ideal interactions between the pilot and the aircraft. Although not always catastrophic, they can range in severity from minimally affecting operational missions, by degradating the capability to perform a certain mission task within the requirements, to loss of aircraft or lives. The available literature mainly focuses on the investigation of problems that directly involve an active participation of the pilot: the so-called Pilot-Induced Oscillations (PIO) [2–4]. There are clear examples in the history of aircraft and, although less reported and only recently documented, of rotorcraft [5,6], resulting in the so-called RotorcraftPilot Coupling (RPC) phenomena. The resulting oscillations may endanger the safe execution of a mission task and, unless stopped, cause severe damage to the vehicle, or even its loss. Furthermore, both fixed and rotary wing aircraft are potentially prone to adverse interactions that involve the pilot in a purely passive, or semi-active, manner: the so-called Pilot-Augmented Oscillations (PAO), where the pilot acts as sort of a mechanical impedance between dynamically and aeroelastically induced vibrations of the body, and the resulting inputs that are inadvertently fed into the control system. Also this type of problems is more commonly investigated for fixed wing aircraft [7]. Occasionally, problems related to the semi-active behavior of the pilot may occur as a consequence of a pilot’s reflexive attempt to counteract the aircraft behavior beyond the pilot’s bandwidth, thus somehow overlapping with the upper boundary of the frequency spectrum of the active pilot’s case, which is usually does not exceed 1 Hz. While fixed-wing aircraft dynamics, with few notable exceptions typically related to large size aircraft, are typically well-separated from rigid body motion, resulting in aeroelastic vibrations well beyond the bandwidth of the pilot, helicopters and rotorcraft may show significant interactional dynamics at very low frequencies, resulting from the coupling of: • the rigid body rotorcraft dynamics; • the rotor dynamics, including rotor aeroelasticity; • the dynamics of the pilot; • the dynamics of the airframe, which, under specific circumstances, may result in unstable behavior. In this paper, no specific investigation of the physiology of the pilot is pursued; however, it is assumed that, in analogy to active PIO, the compliance of the pilot in assisting the PAO may change as a consequence of variations in the workload, in the level of attention, in the specific mission task element (MTE) or in the available cues. It is the change in the pilot’s impedance that triggers the instability; subsequently, the onset of the adverse oscillations by itself alters the level of attention of the pilot and changes the cues the pilot mainly focuses on, potentially worsening the behavior. This activity has been performed by the Universities “Roma Tre” and “Politecnico di Milano” in the framework of the GARTEUR HC AG-16 cooperative effort for the investigation of specific issues related to active and passive pilot interaction with rotorcraft. Special focus has been dedicated to the passive interaction of the pilot with the rotorcraft aeroelasticity. The activity is still ongoing and will end within 2007, including, among other, experimental investigations of the biomechanical and aeroelastic aspects of the interaction, supported by dedicated flight simulator campaigns performed at the University of Liverpool. This paper describes part of the activity performed in preparation of the experimental campaigns. A more specific investigation of the physiological aspects of the modifications in the pilot’s impedance related to the workload could be part of future efforts, where the very positive experience gathered during the GARTEUR cooperation with other European Universities, aerospace research centers and rotorcraft industries will likely be exploited. 2. PROBLEM DESCRIPTION AND OBJECTIVES The popular Bo105 helicopter (Figure 1), although not specifically known to be prone to this type of problems, has been selected as a test bed for the numerical investigation that represents the subject of this work. The main reason is the public availability of a wealth of general information and technical data on this specific rotorcraft, but another reason it the possibility, owing to its peculiar high controllability characteristics, to introduce potentially adverse couplings with the pilot by means of limited changes in the control chain, including the physiological properties of the pilot itself. In this sense, this work does not address specific problems encountered by a real rotorcraft, but rather the potential for problems of this type within generic rotorcraft, introduced by adequately tweaking interactional parameters. 2.1. The “Vertical Bouncing” Problem Particular attention has been devoted to the so-called “vertical bouncing” problem (Fig. 2), where the cone mode of the main rotor, excited by the vertical motion of the vehicle, couples with the collective control inadvertently introduced by the pilot as a consequence of vertical oscillations of the entire airframe and of vertical oscillations related to the deformation of the fuselage. This mode shows a large potential for unfavorable coupling: • the vertical motion of the airframe can have significant coupling with collective flapwise bending (cone) of the main rotor blades; • the vertical motion of the airframe can also couple with the vertical bending of the fuselage; Figure 1: Bo105 operated by DLR FT at Braunschweig. Figure 2: Unstable vertical bouncing mode shape (in the rotating frame). • the vertical motion of the fuselage, both related to the rigid-body motion and to the fuselage bending causes vertical accelerations of the pilot; • the resulting vertical motion of the pilot may cause collective bar excitation through the pilot’s arm; • the motion of the collective bar changes the imposed pitch of the main rotor blades, resulting in changes in the rotor thrust; • the dynamic change in main rotor blade pitch, even though presumably at low frequencies, may statically excite blade twist, resulting in further amplification of blade aerodynamic loads; • the collective excitation may excite the collective lead-lag mode, which is tightly coupled to drive train modes (this aspect has been ignored so far, due to a lack of data about drive train dynamics). It is clear that the numerical analysis of this problem requires the capability to simultaneously model several aero(servo)elastic aspects of rotorcraft dynamics to a degree of refinement that allows to capture at least the order of magnitude of the relevant couplings. The problem has been addressed by the two research teams using different approaches, as described in the following. 2.2. Analysis Approach The highlighting of this problem is separated in two steps: 1. the development and validation of a reliable aeroservoelastic model of the rotorcraft that allows to consider all the coupling terms that might be relevant in the phenomenon; 2. the selection of parameters that might be reasonably perturbed in order to assess the stability properties of the system and its sensitivity with respect to this specific dynamic event. The model is described in detail in the following; with respect to sensitivity parameters, focus is set on the “gain” of the passive pilot, where the term gain is somehow arbitrarily related to a scale factor to be applied to the pilot model to emphasize the feedback he might inadvertently send to the collective control in reaction to vertical oscillations of the cockpit. 3. MODEL DESCRIPTION The model of the helicopter consists in several subcomponents that closely resemble those of the real vehicle or, occasionally, idealizations of real subcomponents that are required by the analysis procedure. 3.1. Main Rotor The most detailed component, from a structural dynamics point of view, is the main rotor; in fact, t