Location

San Diego, CA

Start Date

1-1-1985 12:00 AM

Description

An algorithm is developed to characterize the type and size of a flaw by matching the power spectrum of the flaw with those of the known ones. The algorithm is based on the inversion method developed by Fertig and Richardson for ellipsoidal flaws at low and intermediate frequencies. The present algorithm assumes spherical flaws, but has no restriction in frequency range. The general procedure is to send a wide band ultrasonic pulse upon the flaw and record the reflected waveform. Assuming small acoustic signal level so that linearity holds, the scattering cross-section of the flaw in the direction of the receiver transducer can be obtained by deconvolution with a back wall waveform. For each of the possible flaw types, the most probable radius is estimated by minimizing the “distance” between the measured scattering cross-section and the calculated scattering, cross-sections of that flaw type. The probability for each flaw type is estimated using the values of the minimum “distances” and their second derivatives at the most probable radii of the possible flaw types on the assumption of Gaussian probability distribution.

Book Title

Review of Progress in Quantitative Nondestructive Evaluation

Volume

4A

Chapter

Chapter 3: Theoretical and Applied Inverse Methods

Section

Ultrasonic Applications

Pages

589-601

DOI

10.1007/978-1-4615-9421-5_66

Language

en

File Format

Application/pdf

Share

COinS
 
Jan 1st, 12:00 AM

Identifying Spherical Voids and Inclusions by Matching the Power Spectra

San Diego, CA

An algorithm is developed to characterize the type and size of a flaw by matching the power spectrum of the flaw with those of the known ones. The algorithm is based on the inversion method developed by Fertig and Richardson for ellipsoidal flaws at low and intermediate frequencies. The present algorithm assumes spherical flaws, but has no restriction in frequency range. The general procedure is to send a wide band ultrasonic pulse upon the flaw and record the reflected waveform. Assuming small acoustic signal level so that linearity holds, the scattering cross-section of the flaw in the direction of the receiver transducer can be obtained by deconvolution with a back wall waveform. For each of the possible flaw types, the most probable radius is estimated by minimizing the “distance” between the measured scattering cross-section and the calculated scattering, cross-sections of that flaw type. The probability for each flaw type is estimated using the values of the minimum “distances” and their second derivatives at the most probable radii of the possible flaw types on the assumption of Gaussian probability distribution.