Nonlinear Identification of Transient Tank Vertical Sloshing Motions

In an effort to reduce the fuel burn of commercial jet aircraft, much effort is currently being devoted towards reducing the loads that aircraft experience in-flight due to gusts and turbulence using a range of passive and active methodologies. The EU H2020 SLOW-D project is undertaking a series of fundamental experiments to gain an understanding of vertical sloshing and the added damping that these motions provide, and to apply them for future aircraft wing designs. Recent transient and harmonic experiments undertaken as part of SLOW-D have shown that the added damping from the vertical sloshing motion is a function of the motion amplitude, filling level and excitation frequency. Further, the transient tests have shown that the added damping to a single DOF system comprises three different physical sloshing regimes, whereas the harmonic tests show the importance of the Froude number. In this paper, a selection of the experimental transient data sets from vertical sloshing experiments are analysed using nonlinear identification techniques based upon NAR and NARX models. It is shown how a continuous time model can be identified that accurately predicts the variation of the damping with amplitude and frequency. Surrogate models of identified parameters are developed to enable parametric studies.