Uncertainty Quantification in Gust Loads Analysis of a Highly Flexible Aircraft Wing

In this work, a typical industrial gust loads process has been adapted for use with a nonlinear aeroelastic system. This gust loads process is carried out to determine the worst case “1-minus-cosine” loads on different flexibility variations of a representative, high-aspect ratio, UAV wing, which is subject to uncertain atmospheric and structural properties. Polynomial chaos expansion-based techniques are then used to replicate the map from input uncertainty to the output uncertainties of particular quantities of interest. The results obtained from the nonlinear analysis are compared to linear results to understand what effect nonlinearities have on the propagation of uncertainties through the system. Significant differences in the linear and nonlinear outputs are found using deterministic analyses alone, but additionally, the uncertainty bounds also show differences. For example, for some quantities such as static loads or angle of attack the uncertainty bounds are higher for the nonlinear system than for the linear system. However, the opposite is seen for most of the gust loads, apart from root torque loads on the most flexible case, indicating that the effect of nonlinearities on uncertainty propagation can not be easily generalised. Furthermore, the necessity for round-the-clock gusts for highly flexible aircraft is highlighted.