High reliability damage imaging under non-uniform environmental temperature variations based on modified dynamic time warping

Abstract

In practical applications of Lamb wave monitoring, the environmental temperature fields usually vary non-uniformly over the monitored structures, which can inconsistently affect the wavepackets of every diagnostic signal and reduce the reliability of final detection results. A number of temperature compensation methods have been reported. However, under the complicated non-uniform temperature variations, many methods are deteriorated in their performance or require large amount of Lamb wave data. To effectively improve the reliability of Lamb wave damage detection under non-uniform temperature variations, modified dynamic time warping (MDTW) is presented in this paper. In MDTW, with the flexible time alignment ability of dynamic time warping (DTW), the non-uniform temperature variation effects on Lamb wave signals can be compensated just requiring one set of baselines. Furthermore, to preserve the damage information in the compensated signals, the warping adjusting mechanism is introduced to resolve the temperature over-compensation problem of traditional DTW. This makes MDTW a robust non-uniform temperature compensation method for high reliability Lamb wave detection. After the basic principle of DTW is analyzed, MDTW is proposed and numerically validated. Then, a MDTW-based Lamb wave imaging method is developed for high reliability damage imaging, in which a triple-suppression of non-uniform temperature variations can be accomplished. Finally, the efficiency of MDTW and MDTW-based imaging is demonstrated by the temperature experiment of A0 mode Lamb wave damage imaging on an aluminum plate with its temperature field non-uniformly varied by 13℃∼43℃.