Eddy current array coverage and pitch

An eddy current array is a row of small coils fired in a controlled sequence so that one pass covers what would otherwise take many passes of a single probe. Two numbers describe the geometry: the pitch, the centre-to-centre spacing of the elements, and the active width, which for a single row is (N - 1) x pitch + element width. Everything outside that width is not examined, so the pass plan follows directly from it.

Adjacent elements cannot be driven at the same time without coupling into each other, so the instrument multiplexes: the elements are divided into groups and the groups are fired in turn. Each element is therefore refreshed only once per multiplex cycle, at a rate of acquisition rate divided by the number of groups. That divided rate, not the headline instrument specification, is what sets the maximum scan speed for a required data density.

Transverse resolution is set by the pitch. A flaw lying midway between two element centres couples less strongly to both and returns a reduced amplitude, so the calibration must be performed with the reflector in the worst position between elements, not conveniently under one. Pitch also sets the smallest transverse feature the array can resolve as a separate indication - features closer together than roughly one pitch merge.

Between passes an overlap is applied so that the reduced-sensitivity edges of the array are covered by the adjacent pass. Ten percent is a common minimum; without it the joint between passes becomes a strip of unqualified sensitivity running the length of the component. Use this to plan pass counts and index offsets before mobilising, and to check that the scanner encoder and instrument settings can actually deliver the data density the procedure calls for at the speed being proposed.

active width W = (N - 1) x pitch + element width
effective advance per pass = W x (1 - overlap/100)
number of passes = ceil( scan width / effective advance )
per-element refresh rate = acquisition rate / multiplex groups
max scan speed = per-element refresh rate x required axial sample spacing