Quality inspection of electroplated materials using planar type micro-magnetic sensors with post-processing from neural network model

Abstract

The possibility of employing planar type micro-magnetic sensors for the evaluation of the quality of electroplated materials as well as to inspect the presence of defects in its near-surface is investigated. The impedance of a planar type micro-magnetic sensor in the proximity of any metal surface is a complex function of many parameters including near-surface material properties. A two-dimensional model of two types of planar micro-magnetic sensors having meander and mesh type configurations has been developed for the analytical calculation of magnetic vector potential, flux-linkage and impedance. The impedance of the sensor is used for the evaluation of the quality of the electroplated materials. Usually an off-line generated grid system is used for the on-line evaluation of near-surface material properties. Use of a simple neural network model is proposed for the post-processing of output parameters from the measured impedance data as an alternative to the grid system.