M. Daniele, M. Mamino, G. Quaranta

The present work proposes methodologies used to assess the uncertainties for a tiltrotor model flying at low speed with the objective of providing a mid-fidelity flight mechanics numerical model that could be used for preliminary analysis in view of a Certification by Simulation campaign. The proposed technique was conceived to comply with the guidelines for rotorcraft certification by simulation developed by project RoCS. Numeric uncertainty is assessed first considering all possible sources of numerical error and identifying those who have the largest impact on the output of interest. Calibrations and uncertainty quantification analyses are performed using Dakota, an open-source toolkit which provides an interface between different simulation codes. As the first step, sensitivity analysis is performed to gain a better understanding of model parameters and their effect on the system response. Sensitivity indexes are calculated using a Variance Based Decomposition (VBD) technique which summarizes how the uncertainty in the model output can be apportioned to the uncertainty of the individual input variables. Then a calibration under uncertainty is performed using Bayesian Calibration approach. In Bayes Calibration prior distributions of input variables are updated through experimental data, in order to derive posterior distributions which are more likely to reproduce the observed data. Prior uniform distributions have been assumed for all the parameters, with a Gaussian-type Likelihood function implemented by Dakota, whereas reference test uncertainties were treated all as epistemic.

49th European Rotorcraft Forum 2023, Bückeburg, 2023

Conference Paper

englisch

21,0 x 29,7 cm, 11 Seiten

DGLR-Bericht, 2023, 2023-01, 49th European Rotorcraft Forum 2023 - Proceedings; S.1-11; 2023; Deutsche Gesellschaft für Luft- und Raumfahrt - Lilienthal-Oberth e.V., Bonn

NA

Bestellbar

2023