A Structural Health Monitoring Approach for Damage Detection in Wind Turbine Blades Based on Compressed Sensing Acquisition of Acoustic Emission Events

Abstract

The wind power capacity installed over the U.S. is growing at a fierce pace. According to American Wind Energy Association reports, between 2007 and 2010, wind power provided more than 35% of all new U.S. electric capacity. Despite the rapid growth in wind energy, there are still significant wind energy resources that are underutilized. The U.S. Department of Energy estimates that offshore wind alone could offer over 4,000 GW of electrical energy which is four times the nation’s current total generation capacity. In order to harness this renewable resource significant logistic challenges must be overcome to keep these systems working in harsh offshore environments. The deployment of a structural health monitoring (SHM) systems will be important for these off-shore wind power plants, which are gaining an increasing interest by investors. Wind turbines come in a number of typologies (e.g. horizontal-axis wind turbine and vertical-axis) and are selected according to the specific environment in which they are designed to operate. Furthermore, wind turbines are complex cyber-physical systems consisting of blades, generators, hydraulics and electronics/computation each of which features different failure mechanisms with unique impact in term of down-time and repair cost. All these systems must be continuously monitored in order to ensure they are operating to their maximum potential. Unfortunately continuous monitoring using conventional analogto- digital converters will result in a large amount of data that will need to be stored and processed. In this work we study the suitability of compressed sensing (CS) for long-term acoustic emission (AE)-based SHM of wind turbine blades. In particular the current work simulates the CS acquisition process as applied to AE data acquired during experimental fatigue-to-failure tests (see Figure 1) performed on a full-scale blade of a horizontal axis wind turbine (HAWT). AE represents a promising technique to develop a monitoring strategy for detecting and possibly localizing the presence of damage (i.e. delamination) in mechanical components. When damage occurs or propagates in a structure an acoustic wave is generated by the rapid release in the internal stress of a material. These AE events can be captured at a relative large