Moving towards a-priori identification of undesirable pilot biometrics for collective bounce instability

Abstract

The interaction between the helicopter vibrations and the pilot involuntary control input, filtered through the biomechanical response of the pilot’s body, can lead to the emergence of adverse, possibly even unstable, feedback loops, which in turn produce a degradation of the vehicle handling qualities. These phenomena are called Pilot-Assisted Oscillations (PAO). One of the most important is the “Collective Bounce”, caused by vertical vibrations of the cockpit inducing an unwanted collective control input. On the rotorcraft side, the main rotor coning mode excitation has been shown to produce a phase margin reduction in the collective pitch-heave loop transfer function. On the pilot’s side, biometrics such as stature, weight, age and sex are known to play a major role, but relatively limited effort has been placed in exploring the effects of their variability. especially exploiting predictive numerical techniques in a virtual engineering framework. This work represents a first attempt at filling the gap. A detailed multibody model of the pilot’s upper body , featuring the full musculoskeletal biomechanics of the upper limbs and a simplified, Component Mode Synthesis representation of the torso, is coupled with a simplified rotorcraft model. that reproduces the vertical dynamics of the vehicle, including the coning mode response. A pseudo-random population of pilots, exhibiting different biometrics, is generated and the corresponding multibody biomechanical models are derived. The population is then simulated in a feedback loop with the rotorcraft dynamics and allowed to evolve, through a genetic (de-)optimization algorithm, towards the individuals most likely to be prone to instability. The result of the (de-)optimization process is the identification of the worst possible pilot biometrics with regard to collective bounce proneness on the modeled rotorcraft.