Without taking into account the special probes used for specific testing problems, the number of probes listed by the manufacturer is rather bewildering. A few brief directives will therefore be given to facilitate selection.
For testing fine grained materials by the contact method it has been found that testing
frequency wave lengths from 3 to 1mm in steel permit, under favourable conditions, detecting of flaws of approximately 0.5mm and larger i.e. sizes where the designated flaw usually
begins to be applicable. With these frequencies the most common diameter of the standard probe for daily use is 8 to 10mm.
Using this standard probe a skilled operator will be able to solve a great many problems. When, however, skill alone cannot give the answer, other frequencies will have to be used: where scattering in the material in the form of ‘grass’ masks the display on the screen, lower
frequencies such as 1 and 0.5MHz may have to be used, although in general this lengthens the close-range interference zone by a factor of 2 and 4 respectively, while proportionally reducing the detectability of small flaws. Even lower frequencies may have to be used to make penetration at all possible in materials with strong acoustic attenuation, such as castings, plastics and wood.
Frequencies above 10MHz are more rarely required because excessive sensitivity to minute flaws is by no means always desirable, and particularly also because these thin crystals, when used in the direct contact methods are not very resistant even if fitted with a protective layer. This high frequency range is mainly reserved for indirect coupling where fragile crystals such as lithium sulphate can be used.
Diameter of probe and frequency determine the form of the sound beam which in the case of the standard probe can be characterised by a near field in steel of 30 to 130mm, and an angle of divergence of 10˚ to 30˚.
A long near field means long range, i.e. high sensitivity for small flaws at great depth if the attenuation of sound in the material is neglected. The length of the near field determines the decrease in sensitivity with distance: to bring this decrease to a minimum the length of the near field should where possible, be not less than approximately one third of the maximum flaw distance. In the case of large specimens the condition cannot be fulfilled because, although the length of the near field increases with the square of the diameter of the probe, the standard probe constitutes a practical limit since most test pieces lack sufficiently large contact faces. On the other hand, higher frequencies, owing to the strongly increasing attenuation are ruled out.
Thus, while for long ranges, probes of maximum distance are used, the smaller probes of e.g. 10mm and less at frequencies from 2 to 15MHz are used only for testing ranges up to 100mm. Compared with the larger probes, the smallest probes in this range have the advantage of more accurate lateral flaw location because of their narrower beam in the near field. They are used perforce where the coupling surface of the test piece is too small for the larger probes. Accurate lateral flaw location usually makes a large angle of divergence undesirable.
However, when the problem is the detection of smooth obliquely orientated flaws which might easily be overlooked completely if a pencil-like beam is used, and if the specimen cannot be scanned for such flaws by oblique beaming at all possible angles, it is preferable not to attempt an exact flaw location but rather to use a wide-angle beam e.g. a 10mm diameter probe at 2MHz with an angle of divergence of 20˚.
The recommendation of a definite frequency and a definite probe which have been used successfully with a definite instrument for solving a given testing problem does not
necessarily lead to similarity favourable results when an instrument of different manufacture is used, it is therefore always advisable first to try out the most suitable probe. The reason for this is that a particular instrument with a probe, may for instance have a substantially higher sensitivity at 0.5MHz compared with that at 1MHz than the other instrument. Thus while in the case of the first instrument a given test should preferably be carried out at 0.5MHz, the result
in this case of the second instrument may be better when using 1MHz. At both ends of the frequency range viz. the high and the low, the amplification of the instrument or the
sensitivity of the probes usually decreases. Unfortunately no standardised data concerning this point is as yet available.
When either the shape of a flaw or the proximity of small discontinuities in the same plane, e.g. inclusions given rise to multiple echo signals, the signals from the extremities of the region concerned must be used for size estimation. These may be very much smaller than those obtained from more favourably orientated parts of the flaw surface and are often found at closely similar ranges. In these cases, special care must be taken in the observation of echo signals.
The soundest practice is to take the 20dB drop from the last echo seen. High-resolution probes should be used wherever practicable. When the procedure is employed on curved plate or on pipework the geometry of the method is essentially unaltered for measurements in the vertical plane. The length of the defect is that as marked on the surface, and particularly with thicker sections may not be the actual size.
With undressed welds it is advisable to use some other datum related to the weld centre line. The possibility of error through accidental displacement of the probe makes it advisable (at least for newcomers) to use a probe guide such as a magnetised steel strip.