Surface Roughness and Ultrasonic Materials Characterization

Abstract

Ultrasonic materials characterization relies primarily on measurements of three quantities: (1) backscattered power from the microstructure, (2) frequency-dependent attenuation of the transmitted coherent wave, and (3) shifts in the sound velocity. Surface roughness tends to randomize the phase of the reflected and transmitted waves, which can significantly distort the ultrasonic measurements. For example, uncorrected surface roughness effects can cause substantial underestimation of the porosity in cast aluminum samples.¹ There are three basic effects of surface roughness which should be accounted for in evaluating the data measured by an unfocused transducer. First of all, the coherent reflection and transmission through the interface are attenuated by scattering at the rough surface. Second, the incoherent background scattering from inherent material inhomogeneities in the sample is relatively weakly affected by surface roughness. It does not decrease in proportion to the coherent signals and sometimes even increases slightly. Third, incoherent backscattering at the rough interface produces a slowly decaying tail in the reflection from the front surface, which adds to the incoherent background scattering, especially at small depths below the surface. It has been shown that, for the purposes of NDE, all three effects can be accurately described by the so-called phase-modulation technique. In this simple approximation, explicit analytical results can be derived for (i) the surface roughness induced attenuation of the coherent signals, (ii) the incoherent material noise, and (iii) the level of the additional incoherent surface noise. In this paper, we shall review the above mentioned three main effects of random surface roughness on ultrasonic materials characterization and present simple theoretical predictions and experimental data in order to quantify the resulting changes in the measured parameters.