MS18.10 - System Identification and Damage Detection

Automated operational modal analysis (AOMA) of civil engineering structures and suspension bridges based on vibrational monitoring data is now regularly performed to extract the modal features of these structures, on which structural health monitoring is performed and the state of the infrastructure is inferred. AOMA algorithms have evolved since their serious beginnings around 2005 improving both in performance and in automation. However, in the authors’ opinion based on extensive benchmarking work of multiple AOMA algorithms, none has reached a level of automation and performance which can be considered sufficient and reliable enough for high quality SHM work. Many AOMA algorithms meant for bridge applications (Magalhaes 2009, Reynders 2012, Zhang 2014, Neu 2017, and Yang 2019) are based around two clustering steps using first k-means and then hierarchical clustering, using slightly different parameters and/or workflows to achieve their results. This work proposes an improved automatic OMA based around the framework proposed by the above-mentioned algorithms and their strong points. The k-means spurious pole clearing step is replaced by a gaussian mixture model clustering method, more adapted to the data structure at hand. Hierarchical clustering is maintained and fully automated with data-based cut-off distance calculation. Physical modes are then separated from their mathematical counterparts in an automated way. Comparison and testing of the algorithm is done using real-world full-scale bridge monitoring data.

In recent years, renewable energy sources utilization has boosted the use of wind energy, resulting in more offshore wind farms. Wind industry companies demand the safe operation of Offshore Wind Turbines (OWTs) and use Structural Health Monitoring (SHM) strategies to guarantee safe operation. Commonly used SHM strategies rely on System Identification (SI). SI identifies the modal properties of a structure, like the natural frequency, damping ratios, and mode shapes. Tracking the structural modal properties over time can detect anomalies that indicate damage. Different methods to identify modal properties exist, of which Experimental Modal Analysis (EMA) and Operational Modal Analysis (OMA) are the most well-known. EMA relies on input-output measurements and OMA on output-only measurements. Input data are, however, hardly available for civil structures. Therefore, OMA approaches are advantageous for the SI of OWTs because no input data is required. This research focuses on OMA techniques applied to a two-bladed OWT in operational conditions. Two-bladed OWTs are appealing in finding the most profitable way of generating wind energy. However, identifying and tracking modal properties of an operational two-bladed OWT is harder for three reasons. Firstly, unlike three-bladed OWT, the modal properties of a two-bladed OWT change depending on the azimuthal angle of the blades. The two-bladed OWT becomes a time-variant system due to the varying modal properties. SI of a time-variant system is problematic. Secondly, environmental and operational conditions cause potential misidentification of modal properties because these loading conditions violate the fundamental assumptions of OMA. Thirdly, an OWT has closely-spaced modes that are hard to identify separately using OMA techniques. The Transmissibility-based OMA (TOMA) approach can overcome some of the mentioned limitations. TOMA is a relatively new and promising technique that makes no assumptions about the nature of the loads. Therefore, environmental and operational conditions no longer violate fundamental assumptions, which reduces the possibility of misidentification of modal properties. In this research, TOMA is investigated to show the capabilities and limitations of the approach for an operational two-bladed wind turbine. A representative model of an operational two-bladed OWT is developed to generate dynamic responses. The model incorporates environmental loads (aerodynamic and hydrodynamic), an operational load (harmonic), and the changing azimuthal position of the blades. The model simulates dynamic responses for different rotor speeds of the blades. Multiple OMA techniques, including TOMA, are implemented and applied to the responses to identify the modal parameters. The identification focuses on the first and second natural frequencies and mode shapes because these modes are closely-spaced and are mainly excited by the operational and environmental loads. The identified modal properties are compared with the known modal properties. The applicability of the implemented OMA techniques is assessed. A comparison is made between the suitability of the different approaches. Moreover, the research highlights the conditions in which the use of TOMA is beneficial. Topic: System Identification and Damage Detection Author details: Daan ter Meulen Aelbrechtskolk 5A02 3025 HA Rotterdam +31631069177 daantermeulen@hotmail.com

Safety-critical composite structures are prone to barely visible damages which do not manifest themselves in the global structural response. Acousto-ultrasonic techniques have been utilized to identify these damages, but the problem is extremely challenging on multiple fronts. Com-posite waveguides with high-density cores and/or discontinuities are difficult to model for their dispersion characteristics. Damage identification in such structures thus requires, on one hand, monitoring and mapping of the acoustic signals collected from structures to damage characteristics and, on the other, a physics-based understanding the structural acoustic charac-teristics of composites. Damage identification in composites, including characterization of damage types, for practical applications need to merge the information from data-driven learning as well as modelling in-formation to give a robust/reliable identification framework. This paper focuses on a Bayesian framework for inverse identification of damage in compo-site waveguides based on experimental and modelling data. The dominant propagating wave modes in damaged composite waveguides is obtained as eigenmodes of the damaged compo-site waveguide. The obtained dispersion characteristics of the modes are assimilated into the Bayesian identification framework which enables online identification of damage characteris-tics from acoustic signal features. The proposed identification framework will be the basis of a holistic model-informed and data-driven autonomous monitoring of safety-critical structures.