A significant portion of aging aircraft nondestructive inspection (NDI) applications exist due to the frequent occurrence of maintenance induced damage. The maintenance of Boeing 737 fuselage lap joint structures, for instance, includes the removal of sealant which can result in scribing of the fuselage skin panels. Scribing is a precursor to crack generation and, therefore, it is imperative that this damage is detected and repaired. Cracks emanating from scribes may grow under the lap joint requiring the inspection of crack defects in a second or third layer of a multi-layered structure.

Wyle is developing a nondestructive inspection (NDI) system based on a Giant Magneto-Resistance (GMR) eddy current approach [1]. Unlike conventional eddy current coils, GMR sensors enable crack detection of high sensitivity deep into the test structure and are, unlike conventional ultrasonic testing, not limited to first layer inspection.

The multi-layer lap joint structure inspection has proven to be a difficult problem; previously fielded ultrasonic phased-array approaches for inspection were considered overly burdensome in their complexity of setup and use. Furthermore, they require the expensive removal of the aircraft paint and decals prior to inspection to ensure proper transducer coupling. The Wyle GMR system probes the multi-layer aluminum airframe by inducing electrical eddy currents via an alternating magnetic field transparent to paint and therefore requires no significant preparation of the aircraft. Wyle has worked with Boeing to package this technology into a handheld instrument that relies on limited expertise by the user to easily perform this time intensive inspection in a simplified go/no-go output.

Scribes and cracks will alter the flow of the eddy currents and therefore the magnetic field they produce which are detected by the GMR sensor. However, eddy current, as well as any NDE method, are susceptible to non-critical sources of indication responses due to the structural geometry. Recent advancements toward understanding these sources of noise in a lap joint, recognizing them and separating their responses from the target response are presented.