Average grain size evaluation using scattering-induced attenuation of coda waves

Grain size is one of the key microstructural factors affecting the mechanical properties of polycrystalline metal materials. In this study, a novel method for grain size evaluation using ultrasonic coda waves is proposed. Different from conventional bulk wave methods that require a point-by-point scanning of the structure, the proposed method allows for a rapid evaluation of the average grain size of the whole part from a single inspection location using one-pass testing data. A piecewise energy attenuation function dealing with different attenuation mechanisms is proposed to obtain the effective attenuation coefficient of coda waves. A power-law model is constructed to correlate the effective attenuation coefficient with the average grain size. Ultrasonic testing on nickel-based superalloy plate specimens with different average grain sizes is performed for model calibration and method verification. The applicability and robustness of the proposed method are further validated using a realistic turbine disk specimen with an irregular shape and non-uniform grain sizes. Results show that the proposed method yields a reliable and accurate estimation of the average grain size with a maximum relative error less than 20 %.