Hardness detection in Friction Stir Additive Manufacturing (FSAM) based on the CEEMDAN algorithm and laser ultrasonic Spatially Resolved Acoustic Spectroscopy (SRAS)

Abstract

FSAM is an emerging high-value additive manufacturing technology. However, the variation in hardness of the formed parts due to manufacturing parameters limits its rapid development. Currently, there is a lack of effective non-contact, non-destructive testing methods for hardness in this manufacturing process. To address this, this paper employs laser ultrasonic testing (LUT) technology and uses SRAS to extract sound velocity by exploiting the center frequency of the narrow-band surface acoustic wave (SAW) generated by a pulsed laser, thereby characterizing hardness. SRAS signals from the surfaces of additively manufactured parts often suffer from significant noise. To address this, the Complete Ensemble Empirical Mode Decomposition with Adaptive Noise (CEEMDAN) algorithm is applied. This algorithm decomposes the noisy raw signal into a series of Intrinsic Mode Functions (IMFs). Subsequently, the modes primarily containing the signal are selected using the correlation coefficient for reconstruction, thus achieving de-noising of the raw signal. Simulation and experimental results show that this algorithm can extract the corresponding center frequency and SAW velocity from SRAS signals with a low signal-to-noise ratio. Comparison with hardness test results indicates that the SAW velocity and center frequency effectively characterize changes in sample hardness. In linear regression analysis, a coefficient of determination ($R^2=0.982$) is exhibited.