Rotor Blade Design Framework for Airfoil Shape Optimization with Performance Considerations
Luke D. Allen, Joon W. Lim, Robert B. Haehnel, Ian Dettwiller · AIAA Scitech 2021 Forum · 2021
View Video Presentation: https://doi.org/10.2514/6.2021-0068.vid This work introduces a framework for automated rotor blade airfoil design optimization based on helicopter performance. The framework combines two computational workflows, each created using the Galaxy Simulation Builder (GSB) software package. First, the airfoil parameterization code ParFoil is used to generate a database of morphed airfoil geometries, with aerodynamic properties predicted by ARC2D (via C81Gen). The airfoil database is used to generate a surrogate model for airfoil performance coefficients based on ParFoil parameters. The second workflow utilizes the surrogate model to perform design optimization on a portion of a rotor blade. Optimization is carried out using GSB and the integrated Dakota numerical optimization library. This approach provides users with a variety of optimization algorithms and access to the Department of Defense Supercomputing Resource Center’s (DSRC’s) machines. The framework is demonstrated using Dakota’s multiobjective genetic algorithm (MOGA) to perform a multiobjective, constrained optimization of the tip region of the standard UH-60A main rotor blade. The problem is formulated such that the power coefficient is minimized for forward flight and hover, simultaneously, while subject to a constraint on the rotor pitch link load. The airfoil thickness and thickness crest position of the outboard SC1095 airfoil are the only design parameters used in this example study. Analysis of select points from the Pareto-optimal set shows reductions in main rotor power requirements across a full range of forward flight speeds. The power coefficients for hover and forward flight with advance ratio μ=0.3 are reduced by up to 0.90% and 3.47%, respectively. Improvements of up to 7% are predicted for higher flight speeds approaching μ=0.4. Furthermore, it is shown that the predicted pitch link load can be reduced by as much as 19.2% without incurring a penalty on rotor performance.