Multispectral impedance quality testing of coil-coating system using principal component analysis

Abstract

The task of quality checking of the coil-coated system was undertaken using impedance spectroscopy. In order to properly characterize the system the new impedance spectrum was calculated as an averaged result of 12 experimental spectra obtained from 12 different places of the sample at the same time of immersion in 3% of NaCl. The experimental spectra were recorded for a 24-h period at 1 h intervals between starting points of each measurement. Averaged impedance at each frequency point of the calculated spectrum was characterized by standard deviation. Impedance analysis using two-time constant equivalent electrical circuit was conducted showing the permissible parameter scatter for the checked system. A new classification method of obtained parameters using principal component analysis was presented. This approach enables to identify samples of low quality. © 2010 Elsevier B.V.