An Automated Methodology for Flight Simulation Model Calibration to Manufacturer and Flight Test Validation Data

Abstract

Flight simulation models require calibration to available aircraft data to accurately reproduce real aircraft performance and handling characteristics. In practice, this calibration is typically carried out using high-effort system identification methods or iterative, manual tuning, both of which demand significant time and engineering resources. This work presents an automated two-stage calibration approach for physics-based general aviation aircraft models developed within the PEACE/MADCASP framework. The first stage calibrates aerodynamic, propulsion, and propeller characteristics to steady-state climb and cruise performance data from the aircraft's Pilot's Operating Handbook. The second stage extends calibration to flight test data, addressing aircraft attitude, control surface deflections, and dynamic response characteristics. The methodology is demonstrated on a Cessna 172 Skyhawk model using data provided by Frasca International, with results showing agreement within prescribed performance tolerances derived from FAA Part 60 Level 5 FTD requirements. A supplemental steady-state calibration of a Cirrus SR20 model provides an initial demonstration of the methodology's generality.