摘要: |
Abstract The random decrement method (RDM) is often used in the experimental modal identification of existing structures. The aim of this study was to improve the applicability of RDM in the identification of multimodal structural damping, taking into account the reduced quality of the random decrement (RD) signature. This goal was achieved by introducing a new approach (Approach 2) to the classic RDM procedure based on linear regression fitting of the extreme values of the RD signature within its limited range in which the stability of this signature is not less than an acceptable threshold level. To quantify the quality of the RD signature, the concept of a stability ratio of the RD signature is introduced. In a case study analyzed in this paper, it was determined that the required value of the stability ratio is at least 92%. The effectiveness of Approach 2 was compared with the classic approach, Approach 1, based on the least-squares curve fitting technique. For this purpose, experimental modal identification of a 292-m steel cable-stayed bridge excited by daily traffic flow was carried out. The results showed improved accuracy of modal damping estimation using Approach 2 for all 11 identified bridge modes, from approximately 1% to 26%, depending on the mode number. To ensure the effectiveness of the RDM analysis, it was recommended to set the parameters of the filtration process, the number of time segments included in the RD signature averaging process, and the triggering level of these time segments. The present study can be applied to the performance of similar modal analyses of complex engineering structures, the response of which is represented by multimode vibrations. |