Two scaling methodologies are presented to address the dissimilitude normally experienced when attempting to measure global aerodynamic loads on a small scale wind turbine rotor from a full scale reference. The first, termed direct aerofoil replacement (DAR), redesigns the profile of the blade using a multipoint aerofoil optimisation algorithm, which couples a genetic algorithm (GA) and XFOIL, such that the local non-dimensional lift force is similar to the full scale. Correcting for the reduced Reynolds number in this manner allows for the non-dimensional chord and twist distributions to be maintained at small scale increasing the similitude of the unsteady aerodynamic response; an inherent consideration in the study of the aerodynamic response of floating wind turbine rotors. The second, the geometrically free rotor design (GFRD) methodology, which utilises the Python based multi-objective GA DEAP and blade-element momentum (BEM) code CCBlade, results in a more simplistic but less accurate design.
Numerical simulations of two rotors, produced using the defined scaling methodologies, show an excellent level of similarity of the thrust and reasonably good torque matching for the DAR rotor to the full scale reference. The GFRD rotor design is more simplistic, and hence more readily manufacturable, than the DAR, however the aerodynamic performance match to the full scale turbine is relatively poor.
