Optimal Sensor Placement for Damage Localization in High-Fidelity Digital Twins: Development of Benchmark Scenarios
Please login to view abstract download link
Sensor placement is a critical step in constructing digital twins, as the accuracy of damage identification depends strongly on the spatial distribution of sensors. This study presents a manufactured numerical benchmark for evaluating optimal sensor placement strategies in structural damage identification. The optimal sensor locations are selected from a predefined set of admissible positions by solving an optimization problem. The benchmark problem involves a two-dimensional plate with a central hole subjected to a single loading scenario. Different finite element meshes are considered to investigate the influence of discretization on optimal sensor placement and damage identification performance. Sensor configurations with varying numbers of sensors are examined to assess the trade-off between measurement density and identification accuracy. Structural damage is modeled as a localized reduction of Young’s modulus at the element level. The inverse problem of damage identification is formulated as a numerical optimization problem and solved to recover the spatial distribution of material degradation. Several quantitative measures are introduced to evaluate the quality of the identified damage fields, including accuracy, localization capability, and robustness with respect to sensor configuration. The proposed benchmark provides a systematic framework for comparing sensor placement strategies and supports the development and validation of digital twin methodologies for structural health monitoring.
