Real-time damage assessment of civil structures using fiber grating sensors and modal analysis

摘要

In this effort, long gage fiber Bragg grating sensors are used to provide dynamic strain measurements in conjunction with a damage detection algorithm to provide a real-time assessment of the health of a civil structure such as a bridge or a building. Past research in the area of nondestructive damage detection in structures has shown that modal frequencies and modal damping ratios may not be sensitive enough indicators to detect damage, while changes in mode shapes are significantly more sensitive to structural defects or local damage, with the greatest change occurring in the vicinity of the structural defect. In general, local damage corresponds to a local loss of stiffness of some structural element(s). Therefore, damage identification methods based on changes in vibration mode shapes such as changes in mode shape curvature have encountered some success. However, the changes in mode shape curvature are traditionally obtained through double numerical differentiation of "noisy" mode shapes identified using accelerometer data. Thus, the identified mode shape curvatures may be too noisy to identify (i.e., detect, localize, and quantify) low to intermediate level of local structural damage. On the other hand, the effective modal macro-strain vectors used in this effort, which are equivalent to curvature mode shapes, are identified directly from the long-gage fiber grating sensor measurements. Thus, they are in general much less noisy and therefore significantly more reliable than the curvature mode shapes identified using accelerometer data. This is exactly where lies the advantage of using long-gage fiber optic deformation sensors for structural health monitoring and global damage identification of structures.