The ultrasonic wave scatters in materials when it encounters microstructural discontinuities, such as grain boundaries and defects. The scattering cross-section depends on the ratio of wavelength to scatterer size. Inversely, if the scattering cross-section in material is measured nondestructively, material properties may be estimated. In this study, the relations between microstructure and the ultrasonic attenuation and FOM(Figure-of-Merit) of low carbon steels have been investigated. Average grain size, grain size distribution and carbon content of the alloys were taken as microstructural variables.
The effects of grain size distribution on ultrasonic attenuation and FOM were assessed in pure iron. The slope of frequency-attenuation curve plotted in log-log scale was in the range of 1.29 ~ 1.65 for specimens with average grain size larger than 80 μm. This result implies that stochastic scattering was dominant in these specimens. On the other hand, the slope for specimens with average grain size smaller than 80 μm was in the range of 1.88 ~ 2.22, indicating that Rayleigh scattering was dominant in these specimens. The scattering from some large grains contributed mainly to the scattering of the specimen. The measured attenuations and FOMs were compared with the values calculated by LPM theory and general scattering model, respectively. The absolute values were different but their trends were in good agreement with each other. The attenuation increased linearly with average grain size for specimens with average grain size larger than 80 μm, which confirmed the stochastic scattering as being the dominant scattering mechanism in these specimens.
In case of hypoeutectoid steels, the increase in the carbon content reduced the scattering because of reduction in average ferrite grain size. The attenuations and FOMs of as-rolled specimens were lower than those of normalized specimens. To find the cause of lower scattering in as-rolled specimens, the effects of preferred orientation and prior austenite grain size on the scattering were studied. Since X-ray pole figures of as-rolled specimen and a specimen austenized at 1200℃ did not show any appreciable difference, the preferred orientation was not thought to be responsible for reducing the scattering. Although direct measurement of prior austenite grain size in ferrite-pearlite steels is not possible, the grain size may be estimated in the specimens containing 0.160 wt% and 0.206 wt% carbon since the proeutectoid ferrites and Widmanstaetten side plates were found in the microstructures of the specimens. As normalizing temperature increased, the prior austenite grain size increased. A linear relationship was found between ultrasonic attenuation and prior austenite grain size of the specimens. The variation in proeutectoid ferrite grain size which depends on prior austenite grain size appears to be responsible for the scattering in hypoeutectoid steels.
In eutectoid steels, the colony, which is a region of constant layer orientation, acts as the scatterer because the ferrite/cementite spacing is much shorter than the ultrasonic wavelength. Since the difference in colony sizes between all specimens was very small, the attenuation and FOM did not vary with normalizing temperature. The ferrite and prior austenite grain sizes did not affect the ultrasonic scattering in eutectoid steel where the scatterer was the ferrite/cementite colony.
FOM of specific material is related to its scattering characteristics, which is in turn determined by grain size distribution. It can therefore be used as a material dependent parameter, which can represent the minimum detectable flaw size or the inherent signal to noise ratio in the material. Since FOM values have been reported only for few metals, the FOM measurement itself may be meaningful. The FOMs of pure irons, hypoeutectoid steels, and eutectoid steels were measured to be 0.01 ∼ 0.05 ($cm^{-1/2}$), 0.01 ~ 0.06 ($cm^{-1/2}$), and about 0.03 ($cm^{-1/2}$), respectively.
Electromagnetic-acoustic transducer(EMAT), a noncontact ultrasonic transducer, was fabricated to evaluate nondestructively microstructure during material processing. The velocity and attenuation of shear wave were measured by the acoustic resonance method. While measured ultrasonic velocity did not show a correlation with average grain size of ferrite-pearlite steels, the attenuation at 5 MHz showed a good correlation with the size. The yield strength of ferrite-pearlite steels could be evaluated within the accuracy of ± 5 kg/㎟ by the ultrasonic attenuation measurement.
