This research explores a method for repairing wind turbine blades made from carbon fiber while they remain installed, eliminating costly disassembly. It addresses the challenges of repairing damaged blades and proposes a practical solution to improve repair efficiency. Through structural analysis, we found that this approach can ensure safety during repairs, providing significant benefits to the wind energy industry.

This study presents a new wind dataset, generated by a convection-permitting regional climate model across North America. By validating the dataset against wind observations, we have demonstrated that this dataset captures the wind patterns over complex terrains more realistically than the European Centre for
Medium-Range Weather Forecasts (ECMWF) reanalysis version 5 (ERA5). Additionally, this study quantifies model uncertainty in wind speed, comparing it against interannual variability, to better inform wind farm siting. 

Tilting wind turbines change the airflow behind them, which can lower energy production in wind farms. This research tested both a traditional physics-based model and a deep learning method to predict these effects. While the traditional model improved with added optimization, it struggled with complex wake patterns. The deep learning approach was faster and more accurate, showing potential for better wind farm design and control with reduced computational cost.