Operational modal analysis of two typical UHV transmission towers: A comparative study by fast Bayesian FFT method.

• A field measurement of two typical UHV transmission towers is implemented by performing a structural health monitoring (SHM) system. • The fast Bayesian FFT method is applied for the dynamic analysis of the UHV transmission towers. • The reconstructed signals are used to validate the accuracy approximation of the Bayesian approach. • The results can be a reference for wind resistance design of the UHV transmission towers. The ultra-high-voltage (UHV) transmission tower is currently the most advanced transmission tower in the world. The UHV transmission line is mainly composed of tangent towers and tension towers with different structural forms. This is interesting from a dynamic point of view, since the two typical structures may behave differently in terms of dynamic characteristics. In this study, a comparative study on operational modal analysis (OMA) of these two typical UHV transmission towers under ambient test is carried out by a fast Bayesian FFT method. Prior to the test, the selected frequency bands of modeling are determined by the sensitivity analysis. A structural health monitoring (SHM) system is installed on the towers during the maintenance period to monitor the acceleration response. The dynamic properties of the towers are obtained by the recently developed fast Bayesian FFT method and compared, the accuracy approximation of the Bayesian approach is verified by frequency domain information reconstruction. Furthermore, the damping ratio results as well as other voltage levels are compared with different nations codes. It is found that the stiffness of the tension tower is relatively higher than that of the tangent tower. The first-mode damping ratio results of tower in all directions exceeds 2.59% and the vertical cross-arm direction of tension tower reaches 4.54%. The research results can provide a reference for the wind resistance design for UHV transmission tower, and can be used in model updating in the future works. [ABSTRACT FROM AUTHOR]