Hierarchical crack identification from electric potential measurements using pattern recognition approach and optimization technique*

Abstract

A hierarchical crack identification scheme is proposed based on electric potential measurements. It consists of three stages for progressive estimation of the size and the location of a crack embedded in an electric current carrying body. The scheme includes two different identification procedures, applied on the different stages, which in the proposed method are internally connected as one whole operation, since each stage uses the results from the previous one. The identification procedure is organized in such a way that it can be stopped after each stage, for each stage provides information for the presence of the crack and its dimensions. The completion of the first stage gives a comparatively rough estimate of the size of the crack, classifying it as a negligible, small, medium or large one. The second stage obtains intervals to which the parameters characterizing the crack belong. In the last stage a precise enough estimate of the crack parameters is obtained by applying optimization technique. The method is applied on a case study for the identification of a surface opening semi-elliptic crack in a prismatic body with a known carrying plane. The identification procedure is tested with simulated results for the considered case, and a very good agreement is obtained between the assumed and the obtained crack parameters which is demonstrated by the low relative errors.