T.O. 33B-1-1
4-66
4.5.8.3
Metal Smearing.
Flowing of surface metal may result from machining operations, abrasion during service, or by deformation during
assembly or disassembly of an aircraft or component. The depth of smearing in nonmagnetic materials and its
metallurgical effects will rarely exceed 0.002 to 0.003 inch. At normal crack detection frequencies, the metallurgical
changes created by smeared metal may not affect eddy current response. However, metal build-up and depressions
associated with the smearing create changes in lift-off. With meter type instruments this change may exceed lift-off
compensation and cause irrelevant indications. With impedance plane analysis instruments detection of flaws will not
be affected, if sufficiently separated in phase angle from the lift-off angle. In ferromagnetic steel, eddy current
penetration is very shallow, and any blemish of the surface increases the difficulty of inspection.
4.5.8.4
Scratches, Gouges, And Pitting.
Scratches, gouges, and pits may result in eddy current signals similar in magnitude to those from cracks. As test
frequencies increase, the sensitivity to scratches tends to increase, because the eddy current field is more concentrated at
the surface.
4.5.8.5
Compensation Of Meters For Surface Conditions.
Because eddy current inspection is sensitive to a number of material and inspection conditions, many indications
require additional investigation to separate cracks from other variables. In some cases when testing with meter type
instruments, separation of cracks from increases in lift-off may be obtained by changing the amount of lift-off
adjustment. Increasing the amount of lift-off adjustment or intermediate layer reduces the signal from lift-off or
reverses its direction. Reduction of the amount of lift-off adjustment increases the magnitude of the response from lift-
off.
4.5.8.6
Rate Of Deflection.
Rapidity of response with a meter or impedance plane display instrument is also a means of evaluating indications.
When traversing a crack, a quick rapid deflection is obtained. Variations in conductivity, gradual thickness changes,
out-of-round holes, and variations in edge-to-probe spacing provide a slow, gradual change in measured response. The
inspector should be aware of the rate of change in response from cracks, as contrasted to the rate of signal change from
slow changing material properties or test conditions.
4.5.8.7
Estimation Of Crack Size.
Cracks have the three dimensions of length, width, and depth. All three of these dimensions have an effect on the eddy
current response from the flaw. In general, the length of the flaw can be related to the distance over which a signal
above a specified level is obtained. When the crack is perpendicular to the surface, the approximate depth of the crack
can be determined from the eddy current indication. With meter type instruments this will usually be related to the
signal amplitude. With impedance plane analysis instruments the depth can be determined by the phase angle and
amplitude of the indication. When using a meter type instrument to measure small cracks, amplitude correlates better
with crack area rather than crack depth. The width of the crack also influences the magnitude of the indication. With
impedance plane analysis instruments, the signal shape, phase, and amplitude can be used to estimate the depth and
area of the crack. Generally, the crack dimension of greatest interest is the depth. With meter type instruments this is
the most difficult dimension to measure, and estimation of crack size should only be applied to those situations where
crack geometry is similar and considerable data has been developed relating crack size to meter response. With
impedance plane analysis instruments, crack depth measurements in nonferromagnetic materials can be performed
when based upon the phase response.
4.5.9
Effect Of Scan Rate And Pattern.
4.5.9.1
Meter Instruments.
When manual scanning is performed using meter type instruments without a recorder, the amplitude of response
decreases with increasing scanning speed because of the damping effect of the meter. In the same manner, the ends of a
crack can be established by the point at which the amplitude of deflection falls below a predetermined level.
Mechanized scanning rates with meter instruments and recorder output are limited to the response rate of the recorder.
Scan speeds must always be maintained below the maximum response of the detection system in order for the system
sensitivity to remain linear and within calibration.
