Brinthan Kanesalingam

Spatially resolved strain measurements are crucial to understanding the properties of engineering materials. Although strain measurements utilizing techniques such as transmission electron microscopy and electron backscatter diffraction offer high spatial resolution, they are limited to surface or thin samples. X-ray diffraction methods, including Bragg coherent diffraction imaging and X-ray topography, enable strain measurements deep inside bulk materials but face challenges in simultaneously achieving both high spatial resolution and large fields of view. Dark-field X-ray microscopy (DFXM) offers a promising solution, with its ability to image bulk crystals at the nanoscale while offering a field of view approaching a few hundred micrometres. However, an inverse modelling framework to relate the angular shifts in DFXM explicitly to the strain and lattice rotation tensors is lacking. In this paper, we develop such an inverse modelling formalism. Using the oblique diffraction geometry for DFXM, enabling access to non-coplanar symmetry-equivalent reflections, we demonstrate that the reconstruction of the full deformation gradient tensor F ^(g) , and consequently the strain and lattice rotation matrices, is possible. In addition to the formalism, we develop a computational framework that both forward calculates the anticipated angular shifts and reconstructs the average F ^(g) for an individual pixel from DFXM experiments. Finally, utilizing the established formalism and computational framework, we present methods for sensitivity analysis to relate individual components of the rotation or strain tensor to specific angles of DFXM. The developed sensitivity analysis also enables explicit computation of the errors associated with the reconstruction of each component. The formalism, the computational framework and the sensitivity analysis established in this paper should assist both the interpretation of past DFXM experiments and the design of future DFXM experiments.

About 1.9 gigatonnes of steel is produced every year emitting 7\% (2.7 gigatonnes) of global CO$_2$ in the process. More than 50\% of the CO$_2$ emissions come from a single step of steelmaking, known as ironmaking. Hydrogen based direct reduction (HyDR) of iron oxide to iron has emerged as an emissions free ironmaking alternative. However multi-scale phenomena ranging from nanometers to meters inside HyDR reactors exhibit detrimental microstructure evolution which resists gaseous transport of H$_2$/H$_2$O, slows reaction rates and disrupts continuous reactor operation. To resolve the conundrum between atomic and reactor scales, we devise a percolation-theory model to reconcile nanoscale porosity with macroscopic properties relevant to reactor design models. Using synchrotron nano X-ray computed-tomography, we quantify the evolution of pores in iron oxide pellets, and demonstrate how nano-scale pore networks influence micro and macro-scale flow properties such as permeability, diffusivity and tortuosity. Our new modeling framework bridges the gap between scales and offers the criteria to accelerate HyDR by at least 5x via feedstock-reactor synergies based on percolation.