Dynamic Modeling of Wind Turbine Gearboxes and Experimental Validation

Rune Pedersen

Research output: Book/ReportPh.D. thesisResearch

476 Downloads (Pure)

Abstract

Grinding corrections are often applied to gear teeth, which will alter the load distribution across the tooth. Grinding corrections will also change the load sharing between neighboring tooth pairs, and in turn the gear mesh stiffness. In this thesis, a model for calculating the gear mesh stiffness is presented. The model takes into account the effects of load and applied grinding corrections. The results are verified by comparing to simulated and experimental results reported in the existing literature. Using gear data loosely based on a 1 MW wind turbine gearbox, the gear mesh stiffness is expanded in a Fourier series and combined with a simple, torsional multibody model. Under the assumption of constant angular velocity of the gears, the methods of time-varying modal analysis are applied to this system. This investigation is carried out in order to evaluate the potential of the time-varying modal analysis in relation to gear dynamics. A multibody model of two complete 2.3MWwind turbine gearboxes mounted back-to-back in a test rig is built. The mean values of the proposed gear mesh stiffnesses are included. The model is validated by comparing with calculated and measured eigenfrequencies and mode shapes. The measured eigenfrequencies have been identified in accelerometer signals obtained during run-up tests. Since the calculated eigenfrequencies do not match the measured eigenfrequencies with sufficient accuracy, a model updating technique is applied to ensure a better match by adjusting the model parameters.
Original languageEnglish
Place of PublicationKgs. Lyngby, Denmark
PublisherTechnical University of Denmark
Number of pages108
ISBN (Print)978-87-904-162-63
Publication statusPublished - Jun 2010
SeriesDCAMM Special Repport
NumberS113
ISSN0903-1685

Fingerprint Dive into the research topics of 'Dynamic Modeling of Wind Turbine Gearboxes and Experimental Validation'. Together they form a unique fingerprint.

Cite this