ECCOMAS 2024

Optimal Sensor Placement Methodology for Adjoint-based Weakening Localization in Structures

  • Warnakulasuriya, Suneth (Technische Universität Braunschweig)
  • Antonau, Ihar (Technische Universität Braunschweig)
  • Airaudo, Facundo (George Mason University)
  • Löhner, Rainald (George Mason University)
  • Antil, Harbir (George Mason University)
  • Wüchner, Roland (Technische Universität Braunschweig)

Please login to view abstract download link

Structures can change their material properties during operation due to several reasons, such as aging or corrosion. With the availability of robust sensors and advanced numerical methods for structural analysis, the establishing of Digital Twins of structures is possible and highly promising. Therefore, it is important to develop methods to finally have a digital representation (Digital-Twin) of the aforementioned structures with the actual properties. One of the important constituents of Digital-Twinning is the identification of the actual system parameters in an efficient way. This requires placing different types of sensors in specific strategic locations. This work focuses on developing and testing novel sensor placement methodologies for displacement and strain sensors using the adjoint approach for weakening detection. Firstly, sensor value sensitivities with respect to material properties for prospective/possible sensor locations are computed using the adjoint approach. These sensitivity distributions are used to identify best possible sensor configurations to improve the material properties identification. A set of selected and representative structural problems are studied with the newly developed sensor placement methodology. The effectiveness of the methodology is discussed by applying it on weakening detection methodology proposed in work done by Airaudo et al (2023).