Outage NDE packages still close with “ECT OK” while the instrument was never zeroed on the current shoe. Frequency was left from the last alloy family. A subsurface indication at the fir-tree root waited for the next start.
Eddy current owns coil coupling and calibrated response, not waving a probe. Geometry, lift-off, and reference standards decide whether a crack is a blip or invisible.
Gate before you clear a row
| Item | Why | Fail action | |------|-----|-------------| | Same shoe / frequency as procedure | Wrong setup misses depth | Reconfigure; re-run standards | | Calibration on notch standards | Drift hides small cracks | Recal; do not proceed | | Lift-off check on curved airfoil | Gap kills sensitivity | Reseat; use correct shoe | | Technician cert + blade map | Unmapped skips are silent misses | Complete map before sign-off |

Discipline
Log probe ID, frequency, and standard set on every blade row. Do not carry yesterday’s setup into a different alloy or coating without requalifying. Coating thickness changes are MoC for ECT—not a footnote.
The probe moves. The calibration decides whether that motion was an inspection.
