The semi-elliptical slot, shown here, is often used to model fatigue
cracks, stress-corrosion cracking, and other anomalies. For this
reason, it is one of the canonical shapes that VIC-3D® includes in
its library of flaws.
In "Benchmark Problems for Defect Size and Shape Determination in
Eddy-Current Nondestructive Evaluation," J. Nondestructive Evaluation,
Vol. 15, 1996, D. J. Harrison, L. D. Jones, and S. K. Burke have
performed careful experiments to provide eddy-current impedance data and
experimental parameters for use in a series of benchmark problems for
verification of theoretical inversion techniques. Data for four slots
were taken under computer control by scanning a probe coil along the
line of the axis of the slots in steps of 1.0 mm for each of sixteen
frequencies, spaced at equal logarithmic intervals between 0.25 kHz and
50 kHz. One of the slots is semi-elliptical, with length 22.10 mm,
depth 8.61 mm, and width 0.33 mm.
We have taken the impedance data of Harrison et al at 250 Hz,
and inverted it using VIC-3D® as the driver of an inversion
algorithm. The variables to be determined by inversion are the two
semi-axes of the ellipse. For each trial value of the pair of
variables, VIC-3D® computes a solution of the forward problem with
the semi-elliptical flaw, whose semi-axes take the trial value. Because
the semi-elliptical flaw is already in the VIC-3D® flaw library,
this calculation is done quickly.
In this manner we computed the length to be 22.99 mm and the depth
8.21 mm. (We fixed the width to be 0.33 mm.) We consider this result
to be within experimental error, because Harrison et al admit an
uncertainty of ±0.5 mm in locating the center of the slot. This
information is important in modeling the data-scan.