Accuracy assessment of Beddoes-Leishman and IAG dynamic stall models for wind turbine applications
Omar Sherif Mohamed,Pier Francesco Melani,Galih Bangga,Navid Aryan,Luca Greco,Alessandro Bianchini
DOI: https://doi.org/10.1088/1742-6596/2767/5/052053
2024-06-12
Journal of Physics Conference Series
Abstract:The study presents a systematic comparison between two of the most-credited dynamic stall models for wind turbine applications: the original Beddoes-Leishman (BL) model and the newly-developed IAG. The scope of such comparison, supported by experimental data, is to shed new light on the actual suitability of current dynamic stall models for their integration into modern wind turbine simulation codes, and on the best practices to calibrate them. Two different strategies are followed for the calibration of the BL model: 1) standard one, compliant with common practices found in the literature; 2) a physics-oriented one, focusing on the constants defining the dynamic stall onset as well as on the parameters governing the duration of the vortex shedding process. The IAG model, initially developed based on the first-order BL formulation and recently improved by reducing the number of constants and removing compressibility effects, is applied instead in its standard form only. The two models are compared across a range of oscillation mean angles, amplitudes, and reduced frequencies. Results demonstrate that the original BL model, although with a challenging calibration process, when properly tuned, can provide a very good description of aerodynamic unsteady loads. While showing consistent results, the IAG formulation appears to be more robust, as it employs fewer constants and extracts most of the needed information directly from the input polar data. The comparison between the calibrated BL and IAG models highlights critical modelling aspects, the computation of drag and determination of the stall onset above all, offering valuable insights for the future development of dynamic stall formulations.