Vibration-Based Damage Detection in Flexible Risers Using Time Series Analysis

Abstract

In this paper, a statistical pattern recognition method based on time series analysis is implemented in flexible risers. This method uses a combination of Auto-Regressive (AR) and Auto-Regressive with eXogenous inputs (ARX) prediction models. The flexible riser model used in this paper is experimentally validated employing a proposed numerical scheme for dynamic response of flexible risers. A modal-based damage detection approach is also implemented in the flexible riser model and its results are compared with the ones obtained from time series analysis. The numerical results show that the time series analysis presented in this paper is able to detect and locate structural deterioration related to fatigue damage in flexible risers. Finally, considering the case study results presented in this paper, the presented AR-ARX prediction model works better than the modal-based damage detection method.

Output-Only Structural Health Monitoring for Deepwater Risers: Experimental Study of Wavelet Modified SOBI and Distributed Force Index Algorithm

“…This is also a frequency domain method based on output data, which is not applicable to deepwater risers with varying natural frequencies. Riveros et al 10 suggested that a statistical pattern recognition method with a combined model with auto-regressive (AR) and auto-regressive with exogenous inputs (ARX) works better than modal distribution method. However, in deepwater situation, the inputs are unknown, which means the method is not suitable.…”