Uncertainty Quantification of Tiltrotor Download Prediction

Light winds can have a significant impact on hover performance. In this paper, the uncertainty in atmospheric wind conditions at hover on download and thrust are quantified for the Joint Vertical Experiment (JVX) tiltrotor configuration. Mid-fidelity simulations using the Reduced Order Aerodynamics Model (ROAM) in CREATE����-AV Helios are used to support the research efforts. An uncertainty quantification (UQ) framework is developed that integrates both aleatory and epistemic uncertain input parameters using wind profile and rotor collective respectively. The probability density function for wind speed and direction are represented using two different approaches: normal; and a Kernel distribution curve based on measured wind data. Uncertainty propagation is executed using Monte Carlo simulations with an Artificial Neural Network (ANN) that efficiently and accurately models ROAM outputs. The results show that higher uncertainty is present in download relative to thrust. The results further confirmed that when collective is also characterized as an uncertain variable, higher uncertainty in thrust relative to download follows. An envelope representing intervals of the response outputs due to input uncertainties is presented so that the data can then be used to facilitate informed decision making. Nomenclature