Incorporation of Airfoil-Interactional Data to Improve the Accuracy of Stacked Rotor Performance Predictions in the Design Stage

Abstract

In this dissertation, a methodology is presented for lower-fidelity modeling of stacked rotors, with improved accuracy compared to existing lower-fidelity rotor models. This methodology is built on a conventional prescribed vortex wake model of the rotor, with lifting-line airfoil blades. Such a lifting-line model cannot fully capture the aerodynamic interaction between the rotor blades, which are driven heavily by thickness and shape effects. To account for these effects, a high-fidelity 2D CFD code is used to model the airfoil-to-airfoil interaction along the span of the rotor. These 2D CFD loads are then injected into the lifting-line/prescribed wake rotor model, using an iterative technique to account for the changing deflections of the rotor blades. To accurately determine the airfoil-interactional loads along the span of the rotor, it is necessary to have some way to relate the conditions along the span of the rotor in 3D to the airfoil conditions in 2D, and vice-versa. Methods for parameterization of the airfoil system are presented, which account for both the geometry of the rotor/airfoil system and their aerodynamic conditions. Two different methods of relating the airfoil loads back to the rotor are presented, which offer different strategies depending upon the constraints of the underlying rotor model. Any rotor design must include selection of the airfoils on the blades, and stacked rotors are no different. To that end, 2D airfoil simulations are presented, which demonstrate both the necessity of the current methodology, and offer suggestions for future stacked rotor design. 2D airfoil loads are pre-computed (prior to the lower-fidelity rotor simulations) using an established 2D CFD code. Automation of this code allows for rapid generation of large data sets with minimal user input. The combined 2D CFD/3D prescribed wake methodology is presented and validated against recent experimental results. The baseline prescribed wake model is shown to significantly less variation of thrust and power with phase angle than the experiment. Inclusion of lifting-vortex airloads leads to improvements in thrust prediction, but with incorrect magnitude at small phase angles and incorrect power predictions. Only by including the 2D CFD airfoil-interactional airloads are the most accurate results achieved. Multiple inflow and coupling methods are also examined, with discussion of the strengths and limitations of each. Overall, it is believed that the current work should offer the potential for significantly faster and more accurate design of stacked rotors than was previously possible.