Introduction

Thermomechanical controlled processing (TMCP) is used to maximise the properties of steel, alongside alloy and downstream process design. Optimisation of the TMCP schedule requires knowledge of microstructure development, with accurate metallurgical rules used in combination with FE, or fast mathematical, models. The predictive capability can be used to support new grade development, process optimisation, process resilience to upstream variability, fast mill models, data generation to complement mill data (for ‘scarce data’ regions at the extremes of mill operation) and for data analytics modelling providing a route to exploitation of the metallurgical science generated. Changes in steel making practices, as part of the drive to low CO2 processing, include greater scrap use in BF-BOF or increased EAF capacity. This gives rise to higher residual element contents in the steel compositions, which also needs to be considered for TMCP.

Research in Task 17 will focus on three key areas:

1. Development of full grain size distribution modelling during hot deformation incorporating how dislocation density distributions within the grain affects recrystallisation nucleation and growth behaviour and the strain partitioning behaviour across the grain size distribution. This microstructural model will be coupled to existing process models for temperature and global strain variations through thickness for different rolling profiles and schedules. The model will support predictions for austenite grain size develop in area 2.

2. Development of novel approaches to the control of austenite recrystallisation and strain induced ferrite formation to give a room temperature structure that contains a high dislocation density with recrystallisation nuclei embedded in the structure. This opens up the ability to reduce alloy content but retain, or even improve, the final properties.

3. Development of metallurgical rules for TMCP incorporating residual element effects on boundary motion. The work will extend state of the art knowledge on the effect of elements on nucleation and growth processes. Incorporating these effects into the full TMCP models will allow for potential changes in compositions to be more rapidly accommodated by modifications to process schedules with fewer mill trials, data to support the fast mill models, and potentially more resilient processing through adaptive rolling / cooling schedules.

Workshops with industrial partners are planned throughout the programme: 1) to explore industrial challenges that can form case studies in the latter part of the work using the knowledge gained; and 2) to demonstrate the capabilities (and limitations) of the predictive models and novel processing routes.

Progress to Date

WMG has focussed on developing the framework for predicting the full grain size distribution during multi-pass rolling, using industrial processing parameters and a previously generated FE model for strain and temperature distributions. A library of digital grain size distributions has been developed for input as well as experimental work on a model Fe-Ni alloy to generate the grain size dependency for recrystallisation nucleation site density, including twin boundaries, and strain distributions/partitioning. The future work will consider grain size development when partial recrystallisation occurs and effects of residual elements on boundary mobility and hence recrystallised grain sizes.

At Sheffield, a systematic approach is being taken to investigate the role that solute, including residual elements, have on the behaviour of the austenite during hot working. The aim is to optimise mechanical properties with the minimum alloy additions. The austenite grain boundary mobility is controlled by the solute, which determines grain growth, recrystallisation, strain induced ferrite formation, and transformation temperature. Plane strain compression of model alloys is determining key deformation parameters (e.g. the non-recrystallisation temperature, work hardening and recrystallisation kinetics). It is perhaps surprising how much difference in flow stress