Bayesian time domain approach for damping identification and uncertainty quantification in stay cables using free vibration response

The dynamic characteristics, particularly damping, play crucial roles in the design, monitoring, and vibration control of stay cables. The free vibration test, commonly used for damping identification of stay cables, demonstrates notable technical advantages in specific mode excitation and improving the signal-to-noise ratio. Nevertheless, challenges arise from uncertainties associated with ineluctable ambient vibration influences and noise contamination in measured responses. In recent years, there has been a growing interest in Bayesian modal identification methods due to their remarkable ability to quantify uncertainties. However, further advancements are needed in the time domain, particularly when addressing the relatively short duration of free vibration response data. This is essential to minimize the impact of background ambient vibration and enhance the accuracy of modal identification. This paper proposes a novel Bayesian modal identification approach specifically crafted to address free vibration data affected by ambient vibrations, aiming to achieve high-precision modal identification and effective uncertainty quantification from free vibration response of stay cables. Additionally, a speedup algorithm is used for determining the most probable values (MPVs) to enhance computational efficiency. The efficacy of the proposed approach is validated through simulated scenarios, encompassing both a single-degree-of-freedom structure and a finite element model of a stay cable. Furthermore, the effectiveness of the proposed method is validated through its application in a laboratory test involving a scale model of the stay cable, as well as in a field test of actual stay cables of Sutong Bridge. The results demonstrate that the proposed method achieves high accuracy in identifying modal damping ratios, and effectively quantifies uncertainties in the identification results. •A novel Bayesian approach for modal identification using free vibration data.•Addressing uncertainties linked to ambient vibrations and noise in responses.•High-precision modal identification and effective uncertainty quantification.•A fast computational algorithm determining the most probable values.•Successful validation using diverse simulated, experimental and field datasets.