Wind turbine blade failure can be catastrophic and lead to unexpected power interruptions. In this paper, a Structural Health Monitoring (SHM) algorithm is presented for wireless monitoring of wind turbine blades. The SHM algorithm utilizes accumulated strain energy data, such as would be acquired by piezoelectric materials. The SHM algorithm compares the accumulated strain energy at the same position on the three blades. This exploits the inherent triple redundancy of the blades and avoids the need for a structural model of the blade. The performance of the algorithm is evaluated using probabilistic metrics such as detection probability (True Positive) and false alarm rate (False Positive). The decision time is chosen to be sufficiently long that a particular damage level can be detected even in the presence of system sensor noise and wind variations. Finally, the proposed algorithm is evaluated with a case study of a utility-scale turbine. The noise level is based on measurements acquired from strain sensors mounted on the blades of a Clipper Liberty C96 turbine. Strain energy changes associated with damage from matrix cracking and delamination are simulated with a finite element model. The case study demonstrates that the proposed algorithm can detect damage with a high probability based on a decision time period of approximately 50-200 days. NOMENCLATUREN B D Average of strain energy one-step increment, [ J/step] B k D Strain energy one-step increment, [ J/step] E D Young's Modulus of a composite blade, [GPa] E 0 D Degraded Young's modulus, [GPa] E EH D Young's modulus of an energy harvester, [GPa] g D D Degradation parameter, [-] k d D Decision time, [step] k f D Last time step of the given time frame, [step] k s D Statistical evaluation time, [step] m D Number of divided windows from the evaluation time, [-] n D Integer, [-] p FP D Possibility of False Positive, [%] p TP D Possibility of True Positive, [%] r .ij/ k D Residual of i, j blades at time k, [ J] T D Threshold, [ J] v D Variational term in B k , [ J] V D Volume of energy harvester, OEm 3 W .i/ k D Strain energy of the EH in the blade i, [-] t D Step time, [sec/step] ı k D Random Variable in the residual, [ J] D Strain of a composite blade, [ -strain] 0 D Strain of a damaged composite blade, [ -strain] 2 D Variance of ı k , [ J 2 ] 2 k D Variance of r k , [ J 2 ]
Structural health monitoring of wind turbine blade mechanical performance can inform maintenance decisions, lead to reduced down time and improve the reliability of wind turbines. Wireless, self-powered strain gages and accelerometers have been proposed to transmit blade data to a monitoring system located in the nacelle. Each sensor node is powered by a strain Energy Harvester (EH). The amplitude and frequency of strain at the blade surface (where the EH is mounted) must be sufficient to enable data transfer. In this study, the strain energy available for energy harvesting is evaluated for three typical wind turbines with different wind conditions. A FAST simulation code, available through the National Renewable Energy Lab (NREL), is used to determine bending moments in the wind turbine blade. Given the moment data as a function of position along the blade and time (i.e. blade rotational position), strain in the blade is calculated. The data provide guidance for optimal design of the energy harvester.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
customersupport@researchsolutions.com
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.