A Data-Driven Method for Average Grain Size Assessment using Ultrasonic Guided Wave

Abstract The grain size of polycrystalline metals has been one of the important material property parameters of interest in aerospace and industrial fields. As a nondestructive, convenient, and low-cost approach, the nondestructive evaluation technique based on ultrasonic guided waves has the potential to replace traditional destructive methods such as optical metallography for rapid assessment of grain size. Currently, several studies have demonstrated that multiple features of ultrasonic bulk waves can be used for the quantitative assessment of material microstructure parameters. However, bulk wave methods have limitations in the face of structures with large sizes commonly found in engineering practice. Despite the excellent properties of guided waves for rapid scanning of large areas, there is still a lack of discussion on the use of guided waves for grain size assessment. This paper proposes a data-driven model for guided-wave grain size assessment. The amplitude attenuation, coda wave energy attenuation, and cross-correlation decay are used as three grain-size sensitive characteristic features to construct a linear response surface model. The developed model is validated by a series of GH742 alloy specimens with different average grain sizes.