GB1332898A - Nondestructive examination of materials - Google Patents

Abstract

1332898 Ultrasonic testing UNITED KINGDOM ATOMIC ENERGY AUTHORITY 12 Oct 1970 [15 Oct 1969] 50782/69 Heading H4D In the ultrasonic testing of materials employing the pulse echo technique the nature of the defect producing a selected echo, i.e. whether a single inclusion or a cluster of inclusions close together, is determined by sweeping the ultrasonic test frequency over a range of frequencies and determining the relationship between the amplitude and/or phase of the echo signals and the test frequency. The invention is based on the fact that, over the swept frequency range, the amplitude and phase of the echo signals from a cluster of inclusions fluctuate widely and rapidly whereas echoes from a single inclusion exhibit a smoother and slower fluctuation. Fig. 1 shows one embodiment in which an ultrasonic transducer probe 11 coupled to the material 12 to be tested by means of a rubber pad, by paste or by water (using an irrigated probe), is fed from an oscillator 13 pulsed by a generator 16 and the frequency of which is swept over a predetermined range by a sawtooth generator 15. Pulse generator 16 produces pulses of 10 Ásec. duration spaced by 1 m.sec. periods and the frequency of oscillator 13 and the range of sweeping is determined by the nature of the material under test. In operation, echo pulses from a region (depth) of the material 12 determined by a delay device 17 are gated during the space periods of the pulses to an integrator 19 and a sample and hold circuit 22 the output of which is displayed on a chart recorder 32. Fig. 4 shows the type of variation produced in an experimental arrangement in which the "material" comprised a steel ball in water (curve C) and a bundle of steel wool in the water (curve D). In a modification (Fig. 2) the elements of Fig. 1 (suffix "a") are duplicated but the delay device 32 introduces a slightly longer delay than device 17a so that the position of the echo signal fed to sample and hold circuit 34 corresponds to a region in the material 12a slightly further from the probe but still within the bounds of the reflecting flaw. The outputs of sample and hold circuits 22a and 34 are then normalized by capacitive couplings 36, 35, multiplied in 37 and integrated in 38 which will produce an output if the echo signals in the two channels originate from a single inclusion (correlation between the two sets of echoes) and no output if the echoes originate from different reflecting points in a cluster of inclusions (little or no correlation). The change of phase of the echoes with increasing frequency is shown by curve G (Fig. 5) for echoes from a single inclusion and by curve H for echoes from a cluster of inclusions.