Hidden corrosion detection with guided waves

Book · Mon Dec 01 04:00:00 EST 1997

OSTI ID:548555

Rose, J L; Jiao, D; Pelts, S P; Barshinger, J N; Quarry, M J ^[1]

---

1. Pennsylvania State Univ., University Park, PA (United States). Dept. of Engineering Science and Mechanics

A corrosion detection feasibility study with ultrasonic guided waves is presented. Rather than use normal beam ultrasonic techniques that move point by point over the structure being examined, guided waves are more global in nature allowing examination of regions over 2--3 cm on up to 4--5m with reasonable efficiency in corrosion detection and in estimating average wall thickness between the two transducers. A sample problem in aging aircraft inspection for fuselage thinning is studied and used to illustrate the guided wave inspection procedure and potential for classification and sizing analysis. Special sensitivities to certain kinds of defects are possible with guided waves as a result of wave structure changes across the thickness of the structure by simple mode and frequency changes. A variable angle beam probe and multiple array comb type transducer are considered for flexibility in mode selection and isolation on the dispersion diagrams in order to achieve the special wave structures required for corrosion detection and also for reasonable penetration power. Going beyond the initial feasibility study of guided wave application for the detection of corrosion, a number of tools are developed that can be used in future work to more accurately detect, size, and classify corrosion types in various structures. Boundary element modeling analysis, for example, allows the calculation of reflection and transmission factors for a particular mode and frequency input to a particular pitting or corrosion type defect. This can be followed by detailed feature analysis for subsequent use in neural net algorithm development for corrosion analysis. The physically based features from guided wave analysis can then be used to provide insight and guidance into the algorithm development program using pattern recognition and neural nets.

• OSTI ID: 548555
• Report Number(s): CONF-970332--
• Country of Publication: United States
• Language: English

Related Subjects

36 MATERIALS SCIENCE
AIRCRAFT
CORROSION PRODUCTS
DETECTION
DIAGNOSTIC TECHNIQUES
EXPERIMENTAL DATA
NEURAL NETWORKS
PATTERN RECOGNITION
THICKNESS
ULTRASONIC TESTING
ULTRASONIC WAVES
WAVEGUIDES