Quantifying orientation-specific damage using diagnostic imaging

Different from damage of relatively smooth edges such as through-holes and delamination, the orientation-specific damage (e.g., crack or notch) often exerts strong directionality, and the scattered elastic waves may not be captured efficiently by sensors at certain locations. This situation poses a challenging issue to elastic-wave-based damage identification techniques. In this study, influence of orientation of damage with sizable length in a particular dimension on Lamb wave propagation was scrutinised, in conjunction with an active sensor network. Based on the established correlation between damage parameters (location, shape, orientation and severity) and extracted signal features including both the time-of-flight (ToF) and signal intensity, a probability-based diagnostic imaging approach was developed. With the approach, orientation-specific damage can visually be highlighted in a probability image. Effectiveness of the approach was experimentally demonstrated by predicting a through-thickness crack and L-shape crack in an aluminium plate.