نوع مقاله : مقالات برتر - IRAST2018
نویسندگان
1 دانشجوی دکتری، دانشکده مهندسی عمران، دانشگاه تبریز، تبریز، ایران
2 دانشیار، دانشکده مهندسی عمران، دانشگاه تبریز، تبریز، ایران
چکیده
کلیدواژهها
موضوعات
عنوان مقاله [English]
نویسندگان [English]
The propagation of incremental cracks in the main components of the structures is an important quandary that can cause irreparable damage in the structures over time. This study proposes a statistical method based on signal processing and machine learning to investigate crack location and propagation in reinforced concrete beams. The assumed beam with different boundary conditions is modeled in finite element software (ABAQUS) and incremental cracks are defined according to Extended Finite Element Method (XFEM) in two default locations along the length of the beam. Then, the incremental distributed load is assigned on the beam and the structure response is obtained through the loading time interval. The responses of the beam are used as a database to process with Continuous Wavelet Transform (CWT) and the wavelet coefficients are extracted in the same nodes that structure responses were. The distribution of wavelet coefficients has an effective and valuable implication to investigate the crack location and propagation. In order to consider the variation of wavelet coefficients during crack propagation, Principal Component Analysis (PCA) is used as a machine learning method. The study indicates that the first few components are highly correlated with the location of the damage. The results show that the proposed algorithm has recognized the crack propagation and localized it with high accuracy.
کلیدواژهها [English]
)