High‐Resolution deconvolution techniques and their applications in ultrasonic NDE

The technique of ultrasonic pulse echo measurement is a very cost‐effective method of non‐destructive evaluation (NDE). These measurements are, however, usually masked by the characteristics of the measuring equipment and the propagation paths taken by the ultrasonic pulse. With a proper modeling of the pulse echo, these effects can be reduced by deconvolution. In this paper, several deconvolution algorithms that have been popularly used in other areas such as seismic explorations are revised and adapted to ultrasonic NDE applications. Simulation results are presented to support the feasibility of using these algorithms to extract impulse responses from the ultrasonic pulse echoes for use in defect classifications. The performances and computational complexities of these algorithms are analyzed and compared. In addition, application of these deconvolution techniques to B‐scan enhancement is also discussed.

[1]  Andrew T. Walden,et al.  Robust deconvolution by modified Wiener filtering , 1988 .

[2]  Jerry M. Mendel,et al.  Optimal seismic deconvolution , 1983 .

[3]  D. Middleton.  An Introduction to Statistical Communication Theory , 1960 .

[4]  I. Barrodale,et al.  An Improved Algorithm for Discrete $l_1 $ Linear Approximation , 1973 .

[5]  N. Wiener.  The Wiener RMS (Root Mean Square) Error Criterion in Filter Design and Prediction , 1949 .

[6]  S. Zohar,et al.  Fortran subroutines for the solution of Toepltiz sets of linear equations , 1979 .

[7]  C. H. Chen.  Signal Processing and Pattern Recognition in Nondestructive Evaluation of Materials , 1988, NATO ASI Series.

[8]  C. H. Chen,et al.  High resolution spectral analysis NDE techniques for flaw characterization, prediction and discrimination , 1988 .

[9]  S. Zohar,et al.  Correction to "Fortran subroutines for the solution of Toeplitz sets of linear equations" , 1980 .

[10]  Alan V. Oppenheim,et al.  Discrete-Time Signal Pro-cessing , 1989 .

[11]  Ian Barrodale,et al.  Algorithm 478: Solution of an Overdetermined System of Equations in the l1 Norm [F4] , 1974, Commun. ACM.

[12]  J. P. Burg,et al.  Maximum entropy spectral analysis. , 1967 .

[13]  O. R. Gericke,et al.  Determination of the Geometry of Hidden Defects by Ultrasonic Pulse Analysis Testing , 1963 .

[14]  K. Cook,et al.  ULTRASONIC FREQUENCY ANALYSIS. , 1969 .

[15]  John Kormylo,et al.  New Fast Optimal White-Noise Estimators for Deconvolution , 1977, IEEE Transactions on Geoscience Electronics.

[16]  N. Levinson.  The Wiener (Root Mean Square) Error Criterion in Filter Design and Prediction , 1946 .

[17]  J. L. Hock,et al.  An exact recursion for the composite nearest‐neighbor degeneracy for a 2×N lattice space , 1984 .

[18]  N. R. Chapman,et al.  Comparison of the L1, And L2 Norms Applied to One-at-a-Time Spike Extraction From Seismic Traces , 1983 .