Effects of similarity measures and assignment methods on mode pairing for the application of timber plates

Correctly pairing experimentally and numerically determined mode shapes is crucial for successful model updating. It ensures that the updated model accurately reflects the physical behavior of the structure. This study investigates the two main steps applied for successful mode pairing. First, the correlation between the model and experiments is analyzed using different measures of similarity. Second, based on the computed correlation, a variety of strategies for a correct assignment of the mode pairs is studied. Here, an approach to iteratively combine the mode pairs showing the maximum similarity value in the similarity matrix, an extension additionally using the auto-similarity matrix, the Hungarian method, and a clustering-based approach are investigated. To study the efficacy of the various approaches, the study incorporates an application involving a timber plate. Thus, the effects of employing different similarity measures and pair assignment methods are demonstrated, providing insights for future studies related to mode pairing and model updating.