Development and validation of single-engine general aviation aircraft models within merlin 521 motion-based flight simulators

Date of Award


Degree Name

M.S in Mechanical and Aerospace Engineering.


Department of Mechanical and Aerospace Engineering


Timothy Reissman


As aircraft are developed to be faster and more advanced, the role of the human pilot in the combined pilot-aircraft system dynamics is important to understand. New fields of research aim to identify the ways in which adverse coupling of the pilot and aircraft dynamics manifest and potential causes of phenomena such as pilot-induced oscillations (PIO) and spatial disorientation. This research is considered too dangerous for real aircraft, and flight simulators provide a safer alternative. Therefore, there is a need for simulator models that seem realistic to pilots. I propose a methodology for developing and validating aircraft models within the motion-based Merlin 521 Flight Simulator that provide these perceptibly realistic flight dynamics.This specific flight simulator allows any aircraft to be modeled and simulated through a streamlined user interface, while providing an immersive experience for users within a typical cockpit layout. Realistic physics, clear graphics, and 6 degree-of-freedom motion enhance the physical realism of the simulator. Cessna 152 and Piper PA-28R-201 Arrow III models were developed from pilot's operating handbooks (POH) and aircraft design theory. Model validation relied on feedback from a group of 20 pilots with extensive experience flying at least one of these aircraft. Each pilot flew 8 different routine maneuvers and rated model realism with respect to the actual aircraft using the Cooper-Harper Handling Rating Scale. Simulation data, including control inputs and model kinematics, were extracted after each trial. As each model underwent testing, revisions were made to improve the realism of the model.Model performance was quantified by simulator data and used to confirm model deficiencies noticed by the test pilots. The Cessna 152 climbed at 734 fpm, which exceeded the 700 fpm given in the POH and expected by the pilots for sea level. The model was too efficient for short field takeoff, either breaking ground slower than 54 KIAS or before 725 ft of ground roll.The Piper Arrow III model climbed at 1000 fpm on average, higher than the 840 fpm quoted by the POH but seemed reasonable to the test pilots. It was able to takeoff at shorter distances than the 1800 ft ground roll from the POH and achieve speeds higher than 71 KIAS in shorter ground roll distances. The feedback from the pilots indicated the drag was too low for both models, yet the final versions were approaching the limits of the Merlin 521 to simulate a realistic amount of drag. Despite these deficiencies, the Wilcoxon Signed Rank Test statistically indicated the final versions of both models were perceptibly realistic with respect to the Cooper-Harper scale. The Cessna 152 model achieved significance at the 5% level for aircraft characteristics and pilot effort for all conditions, and the Piper Arrow III model reached significance for all conditions except for the S-Turn and Crosswind Landing. The Mann-Whitney U-Test and Friedman both agreed the final model versions were perceived significantly more realistic than previous versions at the 5% level. Thus, the proposed methodology was statistically justified for producing perceptibly realistic aircraft models.


Aerospace Engineering, modeling, simulation, Merlin 521, aircraft model, validation, Cooper-Harper

Rights Statement

Copyright © 2021, author.