Introduction - a resume of ultrasonic weld inspection
Ultrasonic testing is routinely used for imaging of subjects in both the medical and industrial fields. The history of the development of the technique shows that industrial usage precedes medical practice (Webb, 1988): the first NDT pulse echo examinations were made possible due to the electronic advances made during the Second World War, whilst the results of the first examinations of living tissue were not published until 1950 (Wild, 1950).Ultrasound undergoes refraction and reflection at interfaces between media of different acoustic refractive indices. Hence a crack, which has a lower density, and hence a lower refractive index, from the surrounding weld material will refract and reflect incident ultrasound, whilst the interface between the weld and the parent metal, if correctly fused, should allow wave propagation with minimal interference. Nearly all ultrasonic NDT uses the 'pulse-echo' method of detecting flaws in specimens, ignoring the lower amplitude refracted signals and concentrating on only the reflected energy. The basic principle is to move a small ultrasonic probe or transducer across the surface of the parent plate adjacent to the weld.
Probes with different beam angles can be used to detect planar defects at varying angles of orientation. A defect oriented normal to a beam angle will produce a significant detector signal, whereas with a beam angle 20 or 30 degrees away from the normal, the reflected signal may be very small, or non-existent. The maximum amplitude of the deflected signal is only reached when the defect surface area in the beam equals the beam area Hence the size of the reflected signal can only give an poor indication of the actual size of the defect.
As well as possible reflections from defect signals in the weld region, large echoes are received from the different faces of the welded components geometry - particularly from the surface (plus couplant) and the backwall. This is illustrated in Fig. 1 which shows the expected series of A-scans as a 60¡ probe is moved over a steel block containing a crack like defect. In position (a) only the surface and backwall echoes are visible. As the probe moves over the edge of the defect then the defect signal becomes visible - however as the probe beam moves entirely over the defect the backwall echo may be blocked completely, as shown in position (c). In practice, reflections off the backwall may used in a such a way as to provide enhanced inspection coverage.
Most ultrasonic testing is undertaken manually with an NDT inspector moving a probe over the surface of the component in a raster fashion between the full and half skip distances along the length of the weld, whilst observing the reflected signal on a flaw detector oscilloscope. Clearly for large components, this procedure may be lengthy, tedious and subject to manual error - hence, where possible, automatic scanning techniques, are being used with increasing regularity.
Shaun Lawson