Using a broadband multisine excitation signal for the data-driven modeling of the unsteady lift force on a pitching wing

Understanding unsteady aerodynamic forces on a lifting surface is crucial for many engineering applications. This work demonstrates the use of system identification techniques for the data-driven modelling of the aerodynamic lift force on a pitching wing. This lift force can be considered as a (nonlinear) dynamic function of the angle of attack. The proposed approach consists of multiple steps: 1) excitation by the means of broadband excitation, 2) pre-processing of the measurement, and 3) data-driven modelling techniques. The considered excitation signal is the so-called multisine (also known as periodic pseudo- random noise). However, dynamic stall is known to exhibit important cluster-to-cluster variations and this is no different when utilizing multisines. Therefore, we cluster the data before the actual data-driven Preliminary proceedings modeling and validation. We show that the obtained models can capture the nonlinear aerodynamic forces more accurately than the classical linear and semi-empirical models.