Method of measuring the stress of ferromagnetic materials based on EMAT and magnetic Barkhausen noise characteristic parameters

The magnetostrictive properties of ferromagnetic materials are related to their microstructure which varies between different stress states. This paper proposes a new method of stress detection based on EMAT (electromagnetic ultrasonic transducer) to better identify the early defect and stress state of materials. For experimental verification, an experimental platform consisting of cantilever beam, strain gauge and EMAT detection system was constructed. Compared with the traditional method of magnetostrictive effect detection, the transmitter and the receiver were deployed in 7 different ways to explore the relationship between electromagnetic ultrasonic signal and stress. According to the amplitude of EMAT signal and the intensity of magnetic field as obtained under different stress states, the relationship curves between them were drawn, with ten representative EMAT characteristic parameters selected. In addition, BP (back propagation) neural network was applied to establish the mapping relationship between the electromagnetic ultrasonic characteristic parameters and the stress state of the specimen, and the material stress was predicted with the maximum error of 8.10%. Moreover, BP neural network was adopted to combine EMAT characteristic parameters with Barkhausen characteristic parameters for establishing the mapping relationship between the characteristic parameters and stress state of specimen. The research results show that the accuracy of quantitative prediction was improved for material stress, with the maximum error reaching 6.27%.