Validation of mathematical models for helicopter flight simulators past, present and future challenges

Abstract

At the heart of a flight simulator resides the mathematical representation of aircraft behaviour in response to control inputs, atmospheric disturbances and system inputs including failures and malfunctions. While this mathematical model can never be wholly accurate, its fidelity, in comparison with real world behaviour, underpins the usefulness of the flight simulator. The present paper examines the state of the art achieved in validating mathematical models for helicopter simulators, addressing the strengths and weaknesses of the present European standard for the qualification of helicopter flight simulators, JAR FSTD-H (previously JAR-STD-1H/2H/3H). Essential questions are examined, such as: What is the required model fidelity to guarantee a simulation is sufficiently representative to be fit for purpose? Are the tolerances set in the current standards fine enough that they lead to only minor changes in handling qualities? What is an acceptable tuning process for the simulation? What is the effect of modelling fidelity on the overall pilot control strategy? What is the relationship between the settings of the simulator cueing

Practises to identify and prevent adverse aircraft-and-rotorcraft-pilot couplings—A ground simulator perspective

“…The physical fidelity approach to simulators based on designing and measuring simulator physical components can also be seen in the Federal Aviation Administration (FAA) categories of flight simulators used for training [68]. However, more recently, there has been a trend to shift simulator fidelity from physical fidelity towards 'perceived fidelity' or "cognitive fidelity" [10,77,80], i.e. "the degree to which the device can induce adequate human psychomotor and cognitive behaviour" for a given task and environment.…”

“…Although this mathematical model can never be wholly accurate, its fidelity, in comparison with the real vehicle behaviour, determines the usefulness of the flight simulator in any but especially A/RPC research. Many papers have been written concerning the required model fidelity to guarantee that a simulation is sufficiently representative to be fit for its intended purpose, for example [80] for helicopters. Also, regulatory authorities have produced functional performance standards -for fixed wing aircraft JAR-STD 1A [65] and for helicopters JAR-FSTD H [66] standards in Europe and FAA AC 120-40B [69] and FAA AC120-63 [70] standards in the United States of America.…”

Practices to identify and preclude adverse Aircraft-and-Rotorcraft-Pilot Couplings – A design perspective

“…However, more recently, there has been a trend to shift simulator fidelity from physical fidelity towards 'perceived fidelity' or "cognitive fidelity" [10,77,80], i.e. "the degree to which the device can induce adequate human psychomotor and cognitive behaviour" for a given task and environment.…”

“…Although this mathematical model can never be wholly accurate, its fidelity, in comparison with the real vehicle behaviour, determines the usefulness of the flight simulator in any but especially A/RPC research. Many papers have been written concerning the required model fidelity to guarantee that a simulation is sufficiently representative to be fit for its intended purpose, for example [80] for helicopters.…”

An Investigation of Task Specific Motion Cues for Rotorcraft Simulators

“…Whilst ICAO 9625 recognizes the need for task specific fidelity requirements and defines four fidelity levels for training devices (none, generic, representative and specific) it does not provide any new fidelity criteria. Pavel et al 3 examined the current standards and reported that "it is still not clear whether meeting the standards will guarantee a simulation sufficiently representative of the real world, such that the simulator is fit for purpose". Whilst the paper focused on flight model fidelity, it was recognized that simulator visuals and motion characteristics all contribute to the pilot's perception of the fidelity of the simulator and that the subjective assessment process needed further refinement to identify and correct deficiencies in the overall fidelity of the simulator.…”