Impact of material properties on ultrasonic waves

Although several researches in the literature (Carreón, Barrera, Natividad, Salazar, & Contreras, 2016; Freitas, Albuquerque, Silva, Silva, & Tavares, 2010; Gür & Tuncer, 2005; Mutlu, Oktay, & Ekinci, 2009; 2013; Palanichamy, Vasudevan, Jayakumar, Venugopal, & Raj, 2000; Stella, Cerezo, & Rodríguez, 2009; Vijayalakshmi, Muthupandi, & Jayachitra, 2011) cover the impact of microstructure of polycrystalline metals on ultrasonic wave propagation, their focus principally remains on metallurgical evaluations where the main targets include grain sizing and microstructural characterization. Generally for grain sizing, the tests are carried out in the Rayleigh region of scattering where the wavelength of ultrasonic waves is significantly larger than the mean size of grain diameter and hence the ultrasonic wave scattering is a function of the third power of grain size. Furthermore, most of the aforementioned studies employ longitudinal waves due to the simpler generation and application of this wave mode. In brief, further research is essential to explore the effect of microstructural evolutions on the outcome of longitudinal and transversal ultrasonic inspections and their capability in flaw characterization. It is also noteworthy that the ultrasonic properties of each material grade (more specifically each steel grade) are affected differently by its characteristic microstructure. Therefore, each steel grade should be studied individually and the effect of variations in the ultrasonic properties of these materials on the inspection outcome should be assessed.

Ultrasonic velocity, attenuation and backscattering noise characteristics are the most important parameters studied in the literature. As discussed by several researchers (Papadakis, 1965; Ploix, 2006), the intrinsic attenuation of materials is due to scattering and absorption. The latter is negligible in polycrystalline materials as compared to the scattering effect which results from deviation and reflection of the propagating wave at acoustic impedance discontinuities. Scattering not only decays the travelling wave energy, but it also causes grass noise via the random reflection of a small percentage of wave energy back to the receiver (Shull, 2002). This noise can be captured in the pulse/echo ultrasonic inspections where it can easily mask the signals from small flaws. Thus, it is believed that the combined effect of scattering induced attenuation and grass noise adversely affects the detection capability (Feuilly, Dupond, Chassignole, Moysan, & Comeloup, 2009; Guo, 2003).

Papadakis (Papadakis, 1970), who is one of the main researchers in the ultrasonic field, evaluates the effect of austenitization temperature of quenched SAE 52100 steel on the longitudinal and transversal wave attenuation. The author mentions that the scattering phenomenon is the main contributor to the ultrasonic attenuation in steel. It is also added that the attenuation principally depends on the grain size as well as the elastic moduli of the substructures of the grains. As also reported by Feuilly et al. (Feuilly et al., 2009), the attenuation increases monotonically with the grain size and frequency in the Inconel® 600 alloy. In duplex stainless steels, the ultrasonic velocity is found to be a function of both the phase content and the grain size in different heat treated samples while the attenuation is reported to be directly correlated with the grain size (Vijayalakshmi et al., 2011). Palanichamy et al. (Palanichamy, Joseph, Jayakumar, & Raj, 1995) indicate that transversal waves are more sensitive than longitudinal waves in terms of dependency on the grain size of austenitic stainless steels. They also observe that ultrasonic velocity shows lower sensitivity than attenuation to the variations in the microstructural features and residual stresses.

Apart from the grain size, the attenuation and the velocity of both longitudinal and transversal waves are found to change as a function of phase transformation induced by different heat treatments (Freitas et al., 2010; Gür & Cam, 2007). Both of these studies mention that due to the lower crystal lattice distortion and dislocation density, the ultrasonic waves propagate more rapidly in the ferrite-pearlite microstructure than in the martensitic one. In contrast to Freitas et al. (Freitas et al., 2010), Papadakis (Papadakis, 1970) indicates that lamellar pearlite and tempered martensite are considered to be the most and the least attenuating phases among all. This is confirmed by Kumar et al. (Kumar, Laha, Jayakumar, Rao, & Raj, 2002) who affirm that the martensitic microstructure possesses low elastic anisotropy as a consequence of randomly oriented martensite laths breaking the prior grain volume into fine regions. In both of these studies, the prior austenite grain size is found to be the main contributor to the attenuation. Additionally, the presence of ferrite phases in a martensitic microstructure is reported to induce a significant increase in the attenuation (Papadakis, 1970). The results obtained by Stella et al. (Stella et al., 2009) also imply that heat treatment processes can affect the acoustic properties of materials through the change in the fraction of phases present in the microstructure. According to their conclusions, the variations in the microstructure of specific steel alloys may not be observed by only evaluating the ultrasonic velocity; however, the attenuation and power spectrum of reflected ultrasonic waves could convey useful information concerning microstructural changes. In another study, Kruger and Damm (Kruger & Damm, 2006) propose that since the longitudinal wave in austenite travels in a noticeably lower velocity (approximately 5600 m/s at room temperature) than in martensitic and ferritic structures (approximately 5900 m/s), so the variations of ultrasonic velocity could be employed for online estimation of the volume fraction of austenite in the microstructure during the cooling of low alloy steels.

Other researchers concentrated on establishing correlations between the mechanical properties and ultrasonic parameters. In some references (El Rayes, El-Danaf, & Almajid, 2015; Kumar et al., 2002; Lin et al., 2003), hardness was reported to be inversely correlated with ultrasonic velocity in low alloy, ferritic, and martensitic steels whereas in another study (Bouda, Benchaala, & Alem, 2000) on medium carbon steels, ultrasonic velocity changes proportionally with hardness. However, Hsu et al. (Hsu, Teng, & Chen, 2004) were not able to demonstrate a particular relationship between ultrasonic velocity and hardness of tempered CA-15 martensitic stainless steel. These authors reveal that ultrasonic attenuation directly rises with hardness and tensile strength due to the fact that these mechanical properties are enhanced as a consequence of the presence of carbides (scatterers) at the martensite grain boundaries. More recently (El Rayes et al., 2015) indicate that, in the 9-12% Cr martensiticferritic steel, ultrasonic attenuation increases with hardness because hardness elevation is caused by the higher content of martensite and dispersed carbide in the microstructure; the latter promotes more scattering interfaces and higher anisotropy resulting in ultrasonic waves to be further attenuated.