Discriminating Damage from Surface Wetting via Feature Analysis for Ultrasonic Structural Health Monitoring Systems

Benign environmental effects, including temperature and surface condition changes, can adversely affect the performance of an ultrasonic structural health monitoring system because ultrasonic waves are sensitive to such changes as well as to damage. Compared to temperature, the problem of surface condition changes is more difficult to formalize because of their variability in terms of type, location and extent. In this paper, we systematically investigate the effects of surface wetting with a simplified experiment where well controlled water drops are added onto the surface of a specimen in conjunction with damage. The approach is to find selective features which are sensitive to damage but insensitive to the applied surface wetting. Features considered are derived from the time domain ultrasonic signals and their matching pursuit decompositions. Experimental results show good performance in detecting damage in the presence of the applied surface wetting, and provide a set of differential features with potential for dealing with the problem of surface wetting for ultrasonic structural health monitoring systems.