Damage detection in a 3D wind turbine tower by using extensive multilevel 2D wavelet decomposition and heat map, including soil-structure interaction. (June 2021)

• Record Type: Journal Article
• Authors: Mehr Motlagh, Mohsen; Bahar, Arash; Bahar, Omid
• Is Part Of: Structures. Volume 31(2021)
• Page Start: 842
• Page End: 861
• Publication Date: 2021-06
• Subjects: Wind turbine tower -- Damage detection -- Multilevel 2D wavelet decomposition -- Soil-structure interaction -- Heat map -- Computer-aided engineering; Structural engineering -- Periodicals; 624.1
• DOI: 10.1016/j.istruc.2021.01.018
• Languages: English
• ISSNs: 2352-0124

Abstract

Abstract: The wind-turbine tower is the structure's supporting pole and accounts for 15% of the total cost. Damage to it may cause the structure's catastrophic failure. Advances have been made to conditional monitoring systems for the blades and mechanical components inside the nacelle, while less attention has been given to the tower's health-monitoring efforts. Moreover, the interaction between soil and structure may be significant. This article describes and numerically demonstrates the multi-level 2D wavelet decomposition's extensive analysis to identify a damaged location in a wind-turbine tower with and without soil-structure interaction. The finite-element model based on the NREL 5-MW reference onshore wind turbine is established and verified. This method uses the 3D mode shape to detect the tower's damaged location. A novel algorithm was introduced to accelerate and facilitate the damage finite element modeling process of wind turbine towers. An application code was provided to determine the correct 3D directions of the turbine tower meshes. In order to perform extensive calculations, a new algorithm named "Automatic Wavelet Analysis Aid" used to the extraction and automation of the whole wavelets families and their wavelets. Additionally, the conversion of the cylindrical coordinate system to the Cartesian coordinate system was used, and the relationship between the detection of damage accuracy for this proposed transformation is presented. Finally, the heat map illustrations of data and envelope curves are used as a roadmap for damage detection.