An experimental investigation on the dynamic glaze ice accretion process over a wind turbine airfoil surface

Icing events, particularly under precipitation-icing conditions in which high-liquidity glaze ice tends to form, could pose significant threats to the safe and effective operations of wind turbines in cold and wet environments. During the glaze ice accretion process, the wind-driven unfrozen water was coupled with the growth of ice structures and difficult to be quantified and characterized. In the present study, we introduced a Digital Image Projection (DIP) technique to quantitatively measure the unsteady water runback behaviors and dynamic ice accretion process under typical glaze icing conditions over a highly-cambered wind turbine airfoil surface, i.e., the pressure-side surface of DU91-W2-250 airfoil, in the Icing Research Tunnel at Iowa State University (ISU-IRT). DIP measurement results were found to be able to successfully capture the time-resolved three-dimensional information of the water transport behaviors over the ice accreting surface of the airfoil model during the glaze icing processes. The stumbling motions of the rivulet flows were observed during the icing processes, coupled with an increasing fluctuations induced by the underneath ice roughness. The forces acting on the rivulet flows were analyzed, and a theoretical model based on the force balance was built to predict the rivulet flows. The effects of the incoming airflow velocity on the glaze icing process over the airfoil surface were also studied, and it was found that, as the incoming flow velocity increased, the runback rivulet flows would move farther downstream and became thinner and narrower due to the increased aerodynamic stress acting on them. The rivulet bulged shape was found to have an inverse relationship with the inflow velocity squared. In addition, the ice accreted over the airfoil surface during an icing process was quantitatively decoupled from the unfrozen wind-driven water, and the quantitative results can be used to validate and optimize current ice accretion models. (C) 2019 Elsevier Ltd. All rights reserved.