Accession Number : ADA180648

Title :   A Signal Processing Study of Ultrasonic Nondestructive Evaluation of Materials.

Descriptive Note : Final rept. 1 Oct 85-30 Sep 86,

Corporate Author : SOUTHEASTERN MASSACHUSETTS UNIV NORTH DARTMOUTH DEPT OF ELECTRICAL AND COMPUTER ENGINEERING

Personal Author(s) : Chen,C. H.

Report Date : FEB 1987

Pagination or Media Count : 97

Abstract : In this research the use of modern signal processing and digital pattern recognition in ultrasonic NDE of materials is examined. The high resolution Burg's maximum entropy spectral analysis is shown to provide vital information on the differences among various defect geometries in an aluminum plate. A number of important features computed from the power spectrum have been identified, which in conjunction with correlation matching and by using multi-transducer data can effectively classify the test specimen into three major categories: flat, angular cut, and circular holes and to determine the angles in a multi-stage classification procedure. Size estimation is still a problem that requires further study. Three estimation is still a problem that requires further study. Three deconvolved results. The research clearly shows that the effectiveness of ultrasonic spectroscopy as pioneered by O. Gericke can be significantly improved by using modern digital signal processing and digital pattern recognition.

Descriptors :   *NONDESTRUCTIVE TESTING, *ULTRASONIC TESTS, ALUMINUM, METAL PLATES, CORRELATION, MATCHING, DIGITAL SYSTEMS, PATTERN RECOGNITION, MULTIPLE OPERATION, TRANSDUCERS, CIRCULAR, HOLES(OPENINGS), HIGH RESOLUTION, DEFECTS(MATERIALS), SIGNAL PROCESSING, ENTROPY, SPECTRUM ANALYSIS, POWER SPECTRA, ESTIMATES, SPECTROSCOPY, ULTRASONICS

Subject Categories : Test Facilities, Equipment and Methods

Distribution Statement : APPROVED FOR PUBLIC RELEASE