Real-time in-line steel microstructure control through magnetic properties using an EM sensor

•Macro FE modelling on electromagnetic responses to steel microstructures/mechanical properties.•Non-destructively real-time in-line steel microstructure inspection using magnetic properties.•Dynamic control via varying the cooling strategies or heat treatment using EM sensor feedback.•High temperature tests using lab-based furnace and run-out table (ROT) with cooling system. Magnetic and electric properties (such as low field relative permeability and resistivity) are sensitive to changes in both steel microstructure and temperature. Recently an electromagnetic (EM) sensor system (EMspec™) has been installed to non-destructively on-line monitor the phase (microstructure) transformation in strip steels during the cooling process after hot rolling. To use an EM system to provide dynamic control via varying the cooling strategies or heat treatment using sensor feedback, which can give higher quality steel products with excellent mechanical properties at reduced cost, requires accurate interpretation of the EM sensor signals and predictive capability of the signals from desired microstructures at the relevant temperatures. A 3D FE model is reported here that allows the EMspec™ sensor output (Zero Crossing Frequency, ZCF) to be related to the steel microstructure (phase fraction) using the relationships between permeability and resistivity with microstructure and temperature. The model has been verified by room temperature measurements on various steel grades samples (varying microstructure and strip thickness). High temperature experimental tests have been carried out using a lab-based furnace and run-out table (ROT) with cooling system, mimicking the real-time monitoring of phase transformation of steel strip products. The experimental results have been compared to predicted sensor signals for the known transformation behaviour, determined independently using dilatometry. In this paper the process by which the model can be used to predict the ZCF values for different transformation behaviour, for example different ferrite fractions prior to bainite/martensite formation in a two phase steel, which in turn can be used to control the cooling strategy to achieve a desired microstructure and mechanical properties is discussed.