This work presents the parametrization of a wind turbine CFD/BEM model, intending to match experimental measurements under a wide range of operating conditions. The goal of this work is to build up a generic way of CFD parametrization.
A typical characteristic of a numerical model is that it can be calibrated to match a specific operating condition, but once the situation changes, the match becomes not even satisfactory. Such a numerical model cannot be used to predict the behavior of the system under a new condition. The reason for the characteristic mentioned above is typically multiple errors of several system parameters. The wrong system parameters jointly lead to the correct result for the operating condition used for model calibration but certainly fail to deliver a high-quality model.
In this work, the system parameters of interest are the lift and drag coefficients of an airfoil of a scaled wind turbine. The power and thrust coefficients of the turbine have been measured under more than one hundred operating conditions. 
A BEM model of the turbine is used to calibrate the lift and drag coefficients by minimizing the difference between the model outputs and experimental measurements. The calibration model is so good that its errors are almost always smaller than the measurement uncertainties. The calibrated model also has minor errors for several new operating conditions that are not used for calibration.
A more critical target of this work is to let CFD results match well with experimental measurements. Due to the high computational cost of CFD, the lift and drag coefficients cannot be calibrated in the same way as for the BEM model. However, the polars obtained through the BEM model can be directly used for CFD. A series of SOWFA CFD simulations of the turbine under different conditions have been conducted. The actuator line method is adopted to model the blades of wind turbines while the immersed boundary method is employed to model the nacelle and tower. In a SOWFA simulation, the body force smearing parameter epsilon influences power and thrust of the turbine. After calibrating the epsilon distribution along the blade span, CFD and BEM results show high consistency, although small differences are existent. All CFD simulations adopt the same epsilon distribution, regardless of the operating condition.
In summary, this work presents a way of obtaining a high-quality numerical model of a wind turbine. The model works well for a wide range of operating conditions. The calibration of CFD parameters is quick and efficient because it relies solely on a BEM model that is not computationally costly. Once the model is correctly built up, it can predict the behavior of turbines under new operating conditions.