Benchmarking of Aerodynamic Models for Isolated Propellers Operating at Positive and Negative Thrust

Abstract

Operating a conventional propeller at negative thrust results in the operation of positively cambered blade sections at negative angles of attack, leading to flow separation. Consequently, accurately simulating the aerodynamics of propellers operating at negative thrust poses a greater challenge than at positive thrust. This study offers a comprehensive assessment of the aerodynamic modeling capabilities of numerical methods, spanning low to high fidelity, for computing propeller performance across both positive and negative thrust regimes. Low-fidelity methods, namely, blade-element momentum and lifting line theories, effectively predict propeller performance trends at positive thrust. However, they fail to capture trends at negative thrust beyond the maximum power output point due to the neglect of three-dimensional flow effects. Both steady and unsteady Reynolds-averaged Navier–Stokes (RANS) simulations with y+<1 perform well across both positive and negative thrust conditions, with errors below 2% for both thrust and power magnitudes near the maximum power output point. Lattice-Boltzmann very-large-eddy simulations (LB-VLESs) with y+≤10 exhibit excellent agreement with experimental data with less than 1% error near the maximum power output point but with significant computational costs. Conversely, LB-VLESs with y+≥15 offer a more economical approach to capture general trends with the computational cost of the same order as unsteady RANS. However, wall models introduce errors in modeling flow separation, leading to a 16% overestimation of power magnitude near the maximum power output point. The results highlight the necessity of using tools with increased fidelity levels when considering propeller operation at negative thrust compared to the conventional positive thrust regime.