Uncertainty propagation and management of mixed uncertainties for multi-fidelity multi-disciplinary analysis of propeller with different blade sweep

Abstract

A study on the uncertainty quantification and propagation in a multi-fidelity, multi-disciplinary framework, focusing on aerodynamics, aeroacoustics, and aeroelasticity in propeller blade design is presented. Employing the Lattice Boltzmann Method (LBM) for high-fidelity simulations and the Lifting Line Free Vortex Wake (LLFVW) and Finite Element Analyses (FEA) for mid-fidelity simulations, the study analyzes a 2-bladed, 0.3m diameter propeller using NACA4412 airfoil cross-section, across nine blade sweep configurations. It investigates the effects of two uncertain parameters: freestream velocity, using Interval analysis, and blade tip offset, using Monte Carlo Simulation. The differential analysis between mid and high-fidelity methods shows an uncertainty range of 13.83% to 30.32% for freestream velocity and 2% to 32.48% for blade tip offset, due to the inclusion of the mid-fidelity method. Using the Halton sampling method, it is demonstrated that the uncertainty in the sweep parameter is propagated differently across various performance metrics of the propeller. A backward sweep tends to increase both the mean (by 5-6 %) and uncertainty (by 2-4 %) in all performance parameters, suggesting a potential enhancement in performance but with increased risk. In contrast, a forward sweep reduces mean performance (by 2-3 %) and uncertainty (by 4-6 %) of all parameters except structural deflection, which shows an increasing trend (by 0.5-2 %). This indicates a more reliable aerodynamic and aeroacoustic performance but potentially less efficient operation and increased risk in structural integrity.