Development of an Advanced, Automatic, Ultrasonic NDE Imaging System via Adaptive Learning Network Signal Processing Techniques
Abstract
A conventional pulse-echo imaging system has been modified to operate with a linear ultrasonic array and associated digital electronics to collect data from a series of defects fabricated in aircraft quality steel blocks. A thorough analysis of the defect responses recorded with this modified system has shown that considerable improvements over conventional imaging approaches can be obtained in the crucial areas of defect detection and characterization. A combination of advanced signal processing concepts with the Adaptive Learning Network (ALN) methodology forms the basis for these improvements. Use of established signal processing algorithms such as temporal and spatial beam- forming in concert with a sophisticated detector has provided a reliable defect detection scheme which can be implemented in a microprocessor-based system to operate in an automatic mode. It is shown that signal features extracted from the detected responses contain diagnostic information which enables the defect to be characterized as to type, orientation, and size. Direct comparisons are made between conventional defect images and results obtained from this study; these demonstrate the improvements obtained with the new approach.
Document Details
- Document Type
- Technical Report
- Publication Date
- Mar 13, 1981
- Accession Number
- ADA097740
Entities
People
- Anthony N. Mucciardi
- James R. Gouge Jr.
- Leo J. O'brien
- Nancy A. Aravanis