A Multifrequency Approach to Interpret Defect Signals Superimposed by Disturbing Signals According to the Causing Defect Type and Size

Abstract

A multifrequency algorithm is applied to suppress the effects caused by disturbing signals. The remaining defect signals then exist in a scalar (one-dimensional) form. The defect indication can be assessed by using artificial reference defects of known depths giving equivalent indication peaks. It is not possible, however, to distinguish which type of defect has really caused the read-out value. Before the defect size can be determined it is necessary to know the defect type. A method of distinguishing the different types of defects has been developed. The multifrequency algorithm is applied twice by adding at least one additional measuring value (this can be the real part or the imaginary part of an additional testing frequency) or replacing the test frequencies by entirely new ones. In both parallel channels the indications of the disturbing parameters are suppressed. Provided that the two channels react in different ways to the different types of defects, one obtains a two-dimensional read-out value. Displaying the two channels on an oscilloscope achieves the following results: (1) the disturbing signals are suppressed and concentrated at the zero point of the screen, (2) the defects give indications with different phase angles corresponding to the different defect types, and (3) the heights of the defect indications are correlated with the defect size. Using a computer program based on the permutation principle, the angles between the phase directions of the defect types can be optimized. Results are presented for the testing of built-in heat-exchanger tubes and welded joints and claddings.