Dynamic REcrystallization in Anisotropic Materials – DREAM

This project will associate “robust” experimental techniques to characterize local heterogeneities on anisotropic model materials (Electron BackScattering Diffraction, and Digital Image Correlation) with two complementary modeling approaches to characterize and simulate the basic DRX mechanisms.
This will be achieved through three inter-connected steps:
- The experimental study on polycrystalline ice and magnesium that are model materials whose DRX mechanisms have the appropriate kinetics for ex-situ and in-situ characterization. Indeed, the first steps of nucleation are much too fast in most metals, and much too slow in rocks to be correctly observed. In addition, the high viscoplastic anisotropy of ice and magnesium provides highly heterogeneous strain and stress fields at both
intra- and intercrystalline scales.
- The development of the first model of dynamic recrystallization using a dynamic continuous theory of dislocation fields (FDM), which accounts for plasticity mediated by dislocation transport, lattice incompatibility, internal stress fields and elastic/plastic compatibility conditions at grain boundaries in a polycrystal.
- The combination of the virtues of FDM and of an original “3D FE-level set” approach, which associates CPFEM with microstructural evolution, through the implementation of the FDM-based elastic/plastic compatibility conditions at grain boundaries in the FE-level set approach, and physical substantiation of the level set treatment of grain boundary motion.
This combination of two models at the forefront of the international state of the art in this domain should enable the adequate enrichment of the CPFEM formulation in order to introduce physically-based nucleation criteria and kinetic laws for accurate modeling of DRX at the polycrystalline scale. This step will thus produce an original « intermediate-scale » model joining the advantages of a mesoscopic (RVE) modeling (CPFEM - level set) and the crystal-scale FDM approaches.

The main results obtained so far are the following:
- A high resolution (relatively for ice) characterization of the interaction between strain field and nucleation mechanisms during dynamic recrystallization
- A detailled characterization of nucleation mechanisms in ice, showing the complexity of various processes taking place simulataneously
- EBSD characterization of microstructure evolution in ice under annealing of initialy deformed samples
- Development of techniques to characterize grain boundary and their misorientations relation adapted to the collaborative toolbox MTEX
- Preliminary results on strain fields during deformation of Zinc (instead of Mg)
- Development of a Field Dislocations Modeling approach that consideres different evolutions for every slip systems, and its development in a FE code, with validation in comparison with experimental observations on single and poly-crystals of ice.

Three main objectives will be followed for the last year and half:
- End the development of the deformation and heating stage to perform in-situ tests under the CrystalProbe SEM of Géosciences Montpellier, in order to observe the evolution of dislocation arrangements during dynamic recrystallization of Zinc (instead of Mg).
- Perform some large scale (polycrystal) simulations of DRX in ice and Zinc with the CPFEM approach of Cemef, and investigate the best procedure to take into account the Field Dislocation Modeling prediction of nucleation mechanisms. Toward a progressive coupling between the two approaches
- Extend the observations performed on ice to the interaction between fracturing and dynamic recrystallization close to the ductile to brittle transition, at high temperature

- T. Chauve, M. Montagnat and P. Vacher. Strain field evolution
during dynamic recrystallization nucleation; a case study on ice.
Acta Materialia, 2015, 101, 116-124 .
- M. Montagnat, T. Chauve, F. Barou, A. Tommasi, B. Beausir and C. Fressengeas, New insights into dynamic recrystallization of ice from EBSD
orientation mappings, Frontiers in Earth Sciences, 3:81 .
- T. Chauve, M. Montagnat, F. Barou, K. Hidas, A. Tommasi, D.
Mainprice. Investigation of nucleation processes during dynamic
recrystallization of ice using cryo-EBSD. 2016. In Press
- K. Hidas, A. Tommasi, D. Mainprice, T. Chauve, F. Barou and M.
Montagnat. Microstructural evolution 1 during thermal annealing
of ice-Ih, 2016. Submitted
- T. Richeton, L.T. Le, T. Chauve, M. Bernacki, S. Berbenni and M.
Montagnat. Modelling the transport of geometrically necessary
dislocations on slip systems: application to single- and multicrystals
of ice, 2016. Submitted

Dynamic recrystallization (DRX) strongly affects the evolution of microstructure (grain size and shape) and texture (crystal preferred orientation) in materials during deformation at high temperature. Since texturing leads to anisotropic physical properties, predicting the effect of DRX in metals is essential for industrial applications, in rocks for interpreting geophysical data and modeling geodynamic flows, or in ice for predicting ice sheet flow and climate evolution. DRX reduces the energy stored during plastic deformation of polycrystals via nucleation of new grains and grain boundary migration. The stored energy is linked with heterogeneous microstructures of geometrically necessary dislocations (GNDs), and is strongly affected by grain interactions. Yet, despite a large effort at characterizing DRX, the link between the strain heterogeneity, the internal stress field, the dislocation arrangements and nucleation is still missing. The role of the internal stress field and dislocation structures on DRX, in particular on the nucleation, and the resulting texture development will be central in the “DREAM” project.
DREAM will use strongly anisotropic viscoplastic materials, i.e. ice and magnesium, as model materials to analyze the relations between DRX mechanisms and dislocation arrangements. Thanks to the competencies coordinated in the project, DREAM will make use of complementary experimental tools to probe the dislocation microstructures and long-range correlations built during material loading, and their evolution during DRX: Digital Image Correlation for strain field measurements, Electron Back Scattered Diffraction (EBSD) for lattice misorientation characterization, Acoustic Emission for dynamic nucleation analysis, and neutron-based Laue diffraction for accessing nucleation in 3D. Innovative in-situ EBSD analyses during static heating and deformation will be possible thanks to the unique configuration of the CrystalProbe SEM of Géosciences Montpellier (inclined column).
Experimental data will be used to validate, further develop and couple two modeling approaches: a field dislocation mechanics (FDM) model and a coupled Crystal Plasticity Finite Element (CPFEM) – level set model. FDM is a continuum approach of elasto-plasticity that deals with long-range correlations and internal stresses due to GND arrangements and with plasticity mediated by dislocation transport. It is therefore well adapted to simulate nucleation, but is limited to small polycrystals (a few grains). The description of the interfaces (grains, sub-grain boundaries) accounts for compatibility conditions of the plastic distortion rate. These grain interaction conditions, which are key for handling features like dislocation pile-ups, are not present in the Crystal Plasticity Finite Element Method (CPFEM) or in phase field models of DRX. The FDM model will provide maps of GNDs, internal stress and elastic strain fields, to be compared with experimental observations, allowing for establishment of a nucleation criterion. The CPFEM approach will be appended with the tangential continuity condition at interfaces, and combined with a level set method, to allow modeling of DRX in large polycrystals. The driving force inducing grain boundary motion in the level set approach will be based on the FDM outputs. By progressively linking the inputs/outputs of the two models, we will estimate the level of complexity needed to build a predictive model of DRX in realistic polycrystalline configurations. Validation of this model will be based on direct comparisons with the FDM maps of GNDs, and the experimental observations (local lattice misorientations and the GND distribution they reflect).
Breakthroughs are expected from both experimental and modeling works, as the chosen model materials (ice and Mg), the experimental methods and modeling tools all enable to focus on elucidating the fundamental mechanisms at work in grain nucleation and interface motion during DRX.