Operational monitoring of horizontal axis wind turbines with inertial measurements
Abstract
Operational monitoring of wind turbine rotor blades was proposed in this research as an enabling technology for structural load and damage identification, improved turbine control, and future smart adaptive enhancements of wind turbines. The objective of operational monitoring was to estimate the quasi-static and dynamic loading and deflection of an operating wind turbine rotor. To enable operational monitoring, methods for sensor selection, placement, and fabrication and algorithms for estimating quasi-static and dynamic loading, deflection, and damage for a set of operating wind turbine rotor blades were developed. The thesis of this dissertation is that distributed rotor-based inertial measurements can be used to estimate the state of the operational rotor including its deflection, load, and damage to the blades. It was demonstrated that static deflection of the rotor blade could be estimated by using the centripetal acceleration plane as a fixed reference and the dynamic excitation of the structure could be estimated using the mode shapes from the stationary rotor blade to transform discrete measurements of the rotor response into modal contribution factors. Lastly, the passive alternating weight load during operation could be used to interrogate damage in the rotor blade. A rotating cantilevered Euler-Bernoulli beam model was first studied to examine the feasibility of the proposed deflection and load estimation techniques. Sensor positions were optimized and accurate deflection and load estimators using different numbers of sensors were demonstrated. To examine the dynamic complexities that are observed in a full-scale wind turbine, a lumped parameter, MSC.ADAMS©, Micon 65/13M wind turbine model was updated and used to predict the operational forces, accelerations, and deflections of the rotor blades. Using coordinate transformations and cyclic signal processing, the operational state of the rotor blade was accurately determined in forced response simulations for various wind states (mean, shear, turbulent). The alternating gravitational body force in the span-wise direction of the blade was shown to be sensitive to blade root damage but insensitive to turbulent wind loading. Future work will focus on applying the methods proposed and evaluated in this simulation model to data from operational wind turbines.
Degree
Ph.D.
Advisors
Adams, Purdue University.
Subject Area
Mechanical engineering
Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server.