Prediction of rain erosion damage progression using disdrometer rain data: The importance of liquid water content

Abstract

Abstract Wind turbine blade erosion poses a significant challenge to the durability and performance of wind turbines. Modeling of rain erosion damage, considering atmospheric conditions, improves our understanding of the progression of leading-edge erosion on wind turbine blades. In this study, we investigate the impact of varying raindrop characteristics on rain erosion damage development. We analyse 2.5 years of data from a disdrometer, which measures the size and velocity of falling rain droplets, at Risø campus. Various post-processing methods of the disdrometer data are used for estimating representative droplet diameters and fall velocities for each rain event. We compare measured droplet fall velocities with theoretical terminal velocities, revealing a necessity for revising theoretical approaches to raindrop fall velocity for erosion damage modeling. The measured rain rates and representative fall velocities are used to calculate the liquid water content in the air. We introduce a bin-wise summation method for estimating the liquid water content, circumventing the need for representative droplet assumptions. As this method provides the most accurate input for the damage model, we benchmark the other post-processing methods against it and employ it to evaluate bias estimates of associated damage predictions. The largest bias (22%) in accumulated damage is found with an arithmetic mean droplet diameter approach and the smallest bias (-2%) with the median volume estimation method for damage model input. Furthermore, we demonstrate that, for a given rainfall volume, smaller droplets contribute to larger accumulated damage compared to larger droplets.