Abstract

Blades are one of the main and more costly components of wind turbines and, besides, are difficult to inspect and maintain. Due to operational conditions, wind turbine blades (WTBs) are prone to suffer structural damage, chiefly because of fatigue or extreme loading. Therefore, it is crucial to monitor the state of health (SoH) of these components by implementing robust damage diagnosis and prognosis methods, for identifying the damage and estimating the remaining useful life (RUL), respectively. To study the degradation effects due to fatigue and extreme loads on WTBs, a testing system for small-scale blades was designed and built. The testing machine is able to induce sinusoidal loading with varying amplitude and frequency, then generating damage caused by repetitive loads. Test protocols consisting of extreme (close to resonance) and fatigue-type loading were designed to progressively damage the WTBs, and impact and pull-back tests were conducted at different stages of damage. The free vibration response data measured by an accelerometer array during impact tests is used to identify the modal properties of the blades using an output-only system identification algorithm. The variations of the identified modal properties are then correlated to the physical damage using modal-based damage indices.