response; the secondary standards are said to be traceable to the primary standard. The actual testing in the field
environment utilizes the secondary standards, and the secondary standards are periodically compared to the primary
standard to assure integrity.
Drilled holes are not recommended to be used as calibration standards. However, drilled holes provide a relatively
simple means of fabricating simulated defects in nonferromagnetic metals. To obtain the most consistent response,
holes should be drilled completely through the part or beyond the effective depth of penetration of the eddy currents. A
variety of diameters may be used to assess the range of sensitivity of an inspection with a minimum of expense and
effort. The main disadvantage of the drilled hole for a calibration standard is that larger drilled holes do not always
produce an eddy current signal with the same characteristics of a crack. Hole diameters smaller than the probe diameter
produce signals similar to a small crack. Estimation of actual crack size shall be based upon comparison to an
Drilled Holes In Ferromagnetic Steel.
Drilled holes in steel do not exhibit consistent responses and are difficult to relate to crack size. The variation in
response can be attributed to the lack of penetration of the eddy current signal in ferromagnetic material. With few
exceptions, drilled holes should not be used for calibration standards for eddy current tests of ferromagnetic materials.
Electrically discharged machined (EDM) notches, in a variety of sizes, shapes and locations, can be placed in almost all
metals. The width of the notch can be held to as small as 0.003 inch, and although far greater in width than most
cracks, this method provides a narrower slot, or notch, than all other fabricating techniques such as saw cuts. Similar
responses are obtained on real cracks. Eddy current meter response can often be nonlinear. This means that as the
flaw depth increases, the meter needle may increase or decrease. This is due to the changing phase of the eddy current
signal with depth; the meter is sensitive to only one phase angle and is not uniformly responsive to all phase angles
presented by the EDM notch. The same situation applies to real flaws in a test part. Linear response can be expected
with the meter when EDM notches and flaw size and depth are less than the probe diameter.
EDM Notches In Ferromagnetic Steel.
The eddy current signal does not penetrate well in ferromagnetic materials because of the shielding effect of the high
magnetic permeability. EDM notches are useful as examples of flaws open to the surface of a part. Surface breaking
cracks are best detected by using a very high frequency (500 KHz and greater) which is not meant to penetrate deeply
into the part. Under these conditions the test provides very high sensitivity to surface flaws in ferromagnetic materials.
Likewise the test provides little if any information on flaw depth.
Probably the simplest method of preparing eddy current standards is by means of a jeweler's saw. With a 7/0 blade,
notches as narrow as 0.007 to 0.008 inch can be made in the edge of a standard. Circular jeweler's slotting saws are
also available for other notch locations. Phase response is similar to that obtained from cracks. However, as notch
width increases, the similarity to a crack decreases.
Calibration standards utilizing machined notches can be used under some tests conditions. However, the response of a
particular probe size and frequency to the notch must be evaluated for its applicability for a test situation.
Choosing Reference Standards For Cracks.
As previously discussed, the primary requirement for eddy current reference standards is that they provide uniformity of
response that can be related to the minimum size crack to be detected. To various degrees, several types of reference
standards may meet this criteria. Consequently, such factors as cost, ease of fabrication, availability, and field
application become prime considerations. Table 4-6 tabulates the various types of reference standards and indicates the
advantages and disadvantages of each, and makes recommendations as to the applicability of each as crack reference