Ultrasonic Benchmark Problem 2002: Measurements and Numerical Experiments

Abstract

Measurements and numerical experiments are reported for a sub-set of the problems defined as the Ultrasonic Benchmark 2002 by the World Federation of Nondestructive Evaluation Centers. The backscatter from cylindrical voids of 0.125, 0.5, 1.0, 2.0, and 4 mm radii in an Aluminum block for an incident plane longitudinal elastic wave has been numerically evaluated using one exact and one approximate solution and compared. The exact solution is based on the separation of variables method. The approximate solution is based on Kirchoff's approximation. Within the Thomson-Gray measurement model and the paraxial approximation for the transducer beam, the peak-to-peak voltage is computed as a function of void radius using the exact and approximate backscatter solutions. Using a prescribed reference RF signal waveform, the backscattered RF flaw signals are predicted. Within the Kirchoff's approximation, the backscatter from partially illuminated cylindrical voids is evaluated as a function of the fractional area illuminated. RF signal waveforms were measured for cylindrical voids of 0.5, 1.0, 2.0 and 4.0 mm radii in an Aluminum block under a normal incidence immersion method. The RF signals were also measured, as the transducer was scanned perpendicular to the cylindrical void axis. Measurements were also carried out to obtain the backscatter from solid cylinders of various radii immersed in water. Comparisons between predicted peak-to-peak voltages and measurements are made. Attention is drawn to the transition from 2D backscatter to 3D backscatter encountered when seeking comparisons between predictions and measurements.