Efficient Aerodynamic Modeling Process for a Tilt-Wing eVTOL Using a Mid-Fidelity Computational Tool

Abstract

Urban and regional air mobility might develop into important modes of transportation in the future and are thus currently attracting a lot of attention. For these operations, it is promising to focus on electric vertical takeoff and landing vehicles (eVTOLs). Tilt-wing eVTOLs appear to be especially suited for long range missions in this context of regional air mobility, as they combine hover capability in combination with efficient cruise flight. However, the analysis of these vehicles is a major challenge. This is due to the large design space, the limited availability of empirical data, and especially the complex aerodynamics. Despite this complexity, control concepts require an accurate yet performant aerodynamic model of the aircraft. This work implements a sequential process to efficiently identify mid-fidelity aerodynamic models of eVTOL aircraft for control tasks. The sequential process adapts and extends the RAM process from NASA. The process includes design of experiment, medium-fidelity numerical simulations, model identification, and a decoupling. Design of experiments and medium-fidelity numerical simulation reduce the effort to generate the aerodynamic data. The medium-fidelity simulations are performed using the tool DUST. Stepwise regression is used to fit the data into a reduced-order coefficient model. In addition to the total model, a model decoupled from propulsion forces and moments is identified. This allows arbitrary propulsion models to be added to the decoupled model. That enables to reduce the number of explanatory variables in the identification process. The found process is demonstrated on a tilt-wing configuration similar to the Airbus A3 Vahana. The process developed has proven to be effective. It allows to efficiently identify an aerodynamic model for the eVTOL tilt-wing configuration. The aerodynamic model was validated within the process. Moreover, a model decoupled from the propulsion forces and moments could be identified.