Fidelity is a fundamental concept for crucial decision-making processes during engineering design, test and evaluation of aircraft and rotorcraft. Fidelity, as applied to rotorcraft, proves to be a prerequisite for most human factors research and simulator development programs. This paper reviews the concept of fidelity, its metrics and methods concentrating on rotorcraft applications. The paper stems from the Applied Vehicle Technology (AVT) Panel of the NATO Science and Technology Organization (STO) work of NATO AV-296 concentrating on recent advancements on rotorcraft flight simulation model fidelity prediction and assessment. Starting from the question of what is meant with the term simulation fidelity, the paper will demonstrate that even though the term fidelity can be loosely translated as simulation goodness or faithfulness to reality, when it comes to fidelity assessment, this concept has a multidimensional and multifaceted character depending on the configuration considered. A key message of the paper is that simulation fidelity should be decomposed in different dimensions and attributes with their own metrics. Using multiple fidelity metrics is therefore meaningful for assessing simulation fidelity in general and in particular for the case of rotorcraft.

This paper presents the validation of a Level-D rotorcraft simulation framework by comparing different quantitative model fidelity metrics. The simulation framework uses a non-linear blade element rotor model to simulate the flight dynamics of rotorcrafts with accurate stability and control characteristics identified from system identification techniques. For this study, frequency-domain system identification is used to generate linear state-space 6-DoF quasi-steady models and an automated process is used to adjust the non-linear simulation framework. Regulatory authorities assess performance models for Level-D simulators by comparing time-domain simulation responses with measured aircraft responses for the same set of control inputs. Over the years, new quantitative fidelity metrics were proposed, like the Maximum Unnoticeable Added Dynamics and the Allowable Error Envelopes. This paper will focus on the blade element rotor model hover modeling process to demonstrate its application using a Bell 412 flight test data package. It will also show how this modeling approach can be used to match both Level-D requirements and the alternative fidelity metrics stated above.