Investigation of dynamic stall on a cross-flow turbine blade using modal decomposition

A cross-flow turbine blade, that rotates on an axis perpendicular to the flow, undergoes a complex dynamic stall cycle due to the non-sinusoidal variation of relative flow speed and angle of attack. Furthermore the blade passes through flow speeds much lower than the freestream on the downstream end of the turbine, hence significantly altering the stall cycle. Modal decomposition is performed on high resolution flow fields from large-eddy simulation to explain the stall process and to estimate the relative flow as a step towards reduced-order modeling of the turbine. Distinct flow regimes are explored such as a high angle of attack deep stall regime, a moderate angle of attack regime with optimal power generation, and an intracycle variation of angular velocity that drastically alters the relative flow. Common modes based on velocity and vorticity fields are extracted across these regimes to compare the corresponding temporal coefficients and hence explain the stall dynamics. Additionally, the modes can isolate the relative flow variation from influences such as flow induction and vortices created or encountered by the blade. Relative flow estimation methods using mode temporal coefficients, or probing the reconstructed low-order flow field, are compared for accuracy.