The Practical Application of Grain Noise Models in Titanium Billets and Forgings

For pulse/echo ultrasonic inspections of metal components, mathematical models have been developed which can be be used to assess the likelihood of flaw detection. For example, for various classes of simple defects (e.g., flat cracks, spheroidal inclusions), measurement models [1] based on Auld’s reciprocity relationship can predict the RF echo seen on an oscilloscope when the defect is present at some given location in the component. Other models can predict the average level of backscattered microstructural noise that is seen when no defect is present [2]. Together, the models can be used to estimate signal-to-noise ratios for defects of various types, sizes, and locations [3]. Even when all model assumptions are satisfied, accurate predictions require accurate knowledge of transducer characteristics and pertinent properties of the metal specimen. For the prediction of grain noise levels using the Independent-Scatterer Noise Model [2], for example, the required material properties are the density, wave speed, attenuation coefficient, and Figure-of-Merit (FOM) describing the contribution of the metal microstructure to the backscattered noise. In addition to the above metal properties, the geometrical focal length and effective diameter of the transducer must also be be known. All of these model inputs must be deduced by auxilary measurements