Use of Surrogate Modeling for Preliminary Aircraft Landing Gear Design

Early-stage aircraft design involves exploring a wide range of input parameters while simultaneously considering their effects and associated uncertainties. In this study, surrogate modeling techniques are applied to the preliminary design of aircraft landing gear. Specifically, Gaussian Processes (GPs) are utilized for regression, interpolation, optimization, and uncertainty quantification. The focus centers on modeling the nonlinear damping function of a simple landing gear mathematical model, incorporating fixed and uncertain parameters such as aircraft mass, unsprung mass, and static pressure. By levering GPs a prediction is made on the time response of various outputs, such as the tire forces onto the ground. By setting a design objective as minimizing the reaction factor (normalized ground load) the peak force experienced by the aircraft can be minimized by tailoring parameters related to the landing gear design, such as the stiffness or damping characteristics. Additionally, trends and relationships between these accelerations and the model inputs are explored. Information extracted from these relationships can be used to drive factors and the overall behavior of the landing gear system, which in turn can help engineers make important design decisions. The presented results are preliminary but serve as valuable guidance as to how critical design decisions can be made early in the design cycle. Furthermore, this research contributes to the development of robust surrogate models that aid engineers in making informed choices during the early stages of aircraft design.