Bayesian Model Updating of Structural Dynamics: A 4-DOF Experimental Benchmark Using UncertaintyQuantification.jl

Bayesian model updating provides a rigorous framework for identifying uncertain parameters of structural dynamic systems based on experimental data. In this study, we investigate a four-degree-of-freedom (4-DOF) structure, where acceleration measurements are used to infer the underlying mass, stiffness, and damping parameters. Prior information is derived from estimates based on material properties and geometry, while the posterior distribution is inferred from measured eigenfrequencies and time-domain responses. The model updating is implemented in Julia using the open-source package UncertaintyQuantification.jl, which provides a flexible framework for Bayesian inference and uncertainty propagation. The study highlights the complete workflow from experiment to Bayesian model calibration, demonstrating how open-source tools in Julia enable transparent, reproducible, and extensible structural identification. All data and code are made publicly available to promote open research and benchmarking in uncertainty quantification.