Artificial Neural Network (ANN) Based Microstructure Modelling of 22MnB5 Boron Steel During Tailored Quenching in Hot Stamping Process

Because of demand for lower emissions and better crashworthiness, the use of boron Steel has greatly increased in manufacturing of automobile components. For many applications in the automotive industry it is required that only certain regions in a hot stamped part are fully hardened whereas other regions are required to have softer microstructure. For example in an automobile structural component like B-pillar, which may undergo impact loading, it is desirable that there are certain regions in it which are softer and more ductile so that the component's overall energy absorption is improved. The innovative process of tailored quenching allows this by controlling the local cooling rates, through actively dividing the tooling into heated and cooled zones. A barrier to optimal application of the technique is that a reliable microstructural model is required to quantify the fraction of softer ferrite phase which results in the different regions of a tailored quenched component. Currently most of the existing models for phase distribution prediction in boron steel, only take into account the thermal history of the region while not considering the effect of deformation on the final microstructure. In this paper an Artificial Neural Network (ANN) based model is developed, which predicts the final volume of softer ferrite phase in different regions of a tailor quenched component by taking into account both the thermal history and the effect of deformation. Gleeble testing was done to physically simulate the thermal & mechanical conditions which the different regions of a tailor quenched component undergo during hot stamping. The ferrite phase fraction in microstructure from each of the Gleeble sample was quantified using standard 2% Nital etching procedure and image analysis. The data values obtained from these Gleeble & metallography experiments were used to develop and validate an ANN-based model for final ferrite volume prediction in the microstructures of tailor hot stamped parts. This ANN-based model can be easily coupled with thermo-mechanical Finite Element simulation.