Premium
Improving identifiability of structural damage using higher order responses and phase space technique
Author(s) -
Peng Zhen,
Li Jun,
Hao Hong,
Nie Zhenhua
Publication year - 2021
Publication title -
structural control and health monitoring
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 1.587
H-Index - 62
eISSN - 1545-2263
pISSN - 1545-2255
DOI - 10.1002/stc.2808
Subject(s) - identifiability , structural engineering , acceleration , truss , noise (video) , hammer , vibration , computer science , control theory (sociology) , engineering , acoustics , physics , control (management) , classical mechanics , machine learning , artificial intelligence , image (mathematics)
Summary This paper proposes a novel approach for improving the identifiability of structural damage by using higher order responses and singular spectrum analysis (SSA)‐based phase space topology (PST) technique. SSA is employed to decompose the structural dynamic acceleration response into several low‐dimensional components. Then, the change of PST (CPST) reconstructed from the decomposed component with higher order responses of structures under intact and damaged states is utilized to identify the structural condition changes. Numerical studies on a planar truss structure are conducted to investigate the accuracy and reliability of the proposed approach. The dynamic acceleration responses of the truss structure corresponding to intact and several damage scenarios are utilized for signal decomposition by using SSA and damage detection with CPST. The effects of white noise, pink noise, and modeling uncertainties on the accuracy of damage detection are investigated. The proposed method is also validated through experimental studies on a Tee‐section prestressed concrete beam model. Vibration acceleration responses under hammer impact loads are measured under intact and damaged states, which are used for the damage detection with the proposed approach. Damage patterns identified by using the proposed approach under different loading levels agree well with the observed crack distributions. Both numerical and experimental results demonstrate that using the higher order response components decomposed by SSA and the proposed method is sensitive to damage and reliable for structural damage detection.