Viability of Coda Wave Interferometry for Fatigue Monitoring of Asphalt Mixtures

Abstract For asset management, various non-destructive techniques are used to assess the structural or functional integrity of the pavements to help determine the correct maintenance and rehabilitation (M&R) treatments to prolong the useful service life of a pavement. Few studies have evaluated coupled non-invasive damage tests for asphalt materials in a laboratory setting. In this study, ultrasonic measurements were coupled with tension-compression testing using a dedicated pulse-receiver data acquisition system to characterize fatigue-healing tests. Time signals were acquired periodically (i.e., 10-minute intervals) throughout the fatigue-healing test and then analyzed in the time-domain using the coda wave interferometry (CWI) technique. Two windowing procedures were proposed, including Method 1, which incorporated a simple statistical approach and Method 2, which was derived from the diffusive energy envelope paired with an adaptive window length stretching technique. After windowing the time signals, the CWI procedure was completed using a dynamic time-warping (DTW) algorithm. CWI measurements using both windowing selection methods were demonstrated to be capable of distinguishing loading and rest periods during healing tests. The CWI technique was also sensitive enough to accurately differentiate loading at the lower and higher strain levels used in this study. Notably, Method 2 CWI results were able to maintain lower signal-signal variability and capture the “non-linearity” or curvature observed in traditional fatigue tests typically attributed to bias effects.