Sensor Fusion for Electromagnetic Stress Measurement and Material Characterisation

Abstract

Detrimental residual stresses and microstructure changes are the two major precursors for future sites of failure in ferrous steel engineering components and structures. Although numerous Non-Destructive Evaluation (NDE) techniques can be used for microstructure and stress assessment, currently there is no single technique which would have the capability to provide a comprehensive picture of these material changes. Therefore the fusion of data from a number of different sensors is required for early failure prediction Electromagnetic (EM) NDE is a prime candidate for this type of inspection, since the response to Electromagnetic excitation can be quantified in several different ways: e.g. eddy currents, Barkhausen emission, flux leakage, and a few others. This chapter reviews the strengths of different electromagnetic NDE methods, provides an analysis of the different sensor fusion techniques such as sensor physical system fusion through different principles and detecting devices, and/or feature selection and fusion, and/or information fusion. Two sensor fusion case studies are presented: pulsed eddy current thermography at sensor level and integrative electromagnetic methods for stress and material characterisation at feature (parameters) level.