Comparison of wind turbine blade models through correlation with experimental modal data

R. Janeliukstis*, R. Riva, E. Di Lorenzo, M. Luczak, S. C. Yeniceli, S. H. Madsen, B. Peeters

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingArticle in proceedingsResearchpeer-review

Abstract

In this work, wind turbine blade numerical models have been developed with two different finite element software – DTU Wind Energy HAWCStab2 with Timoshenko beam elements and MSC NASTRAN with solid elements. A correlation analysis has been performed comparing both of these models with experimentally measured data in order to validate the models. For the experimental part, we have performed the experimental and operational modal analysis of a 14.3 m long composite blade clamped at the root. The natural frequencies have been extracted from the measured acceleration responses with three different methods – PolyMAX, MOESP and CWT and compared with the ones computed by the two software, while the mode shapes were correlated with MAC and COMAC metrics. Although the frequency values are in good agreement, the ordering of the 5th flapwise mode shape has been swapped with the 1st simulated torsional mode in the solid model, while this was not the case with mode shape ordering in the beam model.
Original languageEnglish
Title of host publicationProceedings of ISMA2020 and USD2020
PublisherKU Leuven
Publication date2020
Pages3507-3514
ISBN (Electronic)978-908289311-3
Publication statusPublished - 2020
Event2020 International Conference on Noise and Vibration Engineering and 2020 International Conference on Uncertainty in Structural Dynamics - Virtual event
Duration: 7 Sept 20209 Sept 2020

Conference

Conference2020 International Conference on Noise and Vibration Engineering and 2020 International Conference on Uncertainty in Structural Dynamics
CityVirtual event
Period07/09/202009/09/2020

Fingerprint

Dive into the research topics of 'Comparison of wind turbine blade models through correlation with experimental modal data'. Together they form a unique fingerprint.

Cite this