arXiv subjects

Sina Borgi

Publications and source records attributed to Sina Borgi.

5 recordsLinked to original sources

arXiv2026

Automated Burgers Vector Identification for Individual Dislocations in Bulk Crystals

Weak-beam imaging in dark-field X-ray microscopy (DFXM) can resolve individual dislocations in bulk crystals, but assigning Burgers vectors from the resulting contrast typically requires manual comparison with forward simulations. Here, we train a physics-informed convolutional neural network (CNN) on geometrical optics simulations of isolated dislocations in face-centred cubic (FCC) aluminium, incorporating crystallographic constraints into the learning pro- cess. The model identifies Burgers vectors from weak-beam integrated rocking-curve images. On synthetic test data, the model achieves an accuracy of approximately 93%. In an experimental cross-slip case, the constrained model as- signs 72.7% of the layer-wise predictions to the reference Burgers vector. These results show that simulation-trained, physics-informed CNNs represent a step toward automated dislocation identification in DFXM.

cond-mat.mtrl-sci↗

arXiv2026

Revealing Dislocation Interactions Controlling Mechanical Properties of Metals

During plastic deformation, metals change shape while continuously becoming stronger. The microscopic origin of these processes lies in the proliferation and movement of line defects, dislocations, and the subsequent self-organisation and pinning of dislocations on lattice imperfections, including other dislocations. The nature of these multiscale processes has remained elusive because in situ observations have not been feasible. We present 3D movies of how dislocations pile up near an obstacle, deeply within a mm-sized pure Al sample and during tensile deformation. Cross-slip is found to provide a mechanism for the dislocations to escape the pile-up, leading to pronounced intermittent behaviour. Such data support a new generation of dislocation dynamics and micro-mechanics modelling.

cond-mat.mtrl-sci↗

arXiv2026

FaCT-GS: Fast and Scalable CT Reconstruction with Gaussian Splatting

Gaussian Splatting (GS) has emerged as a dominating technique for image rendering and has quickly been adapted for the X-ray Computed Tomography (CT) reconstruction task. However, despite its growing popularity, the benefits of GS are typically not substantial enough to motivate a transition from well-established reconstruction algorithms. This paper addresses the most significant remaining limitations of the GS-based approach by introducing FaCT-GS, a framework for fast and flexible CT reconstruction. Enabled by an in-depth optimization of the voxelization and rasterization pipelines, our new method is significantly faster than its predecessors and scales well with projection and output volume size. Furthermore, the improved voxelization enables rapid fitting of Gaussians to pre-existing volumes, which can serve as a prior for warm-starting the reconstruction, or simply as an alternative, compressed representation. FaCT-GS is over 4X faster than the State of the Art GS CT reconstruction on standard 512x512 projections, and over 13X faster on 2k projections. Implementation and data available through: https://github.com/PaPieta/fact-gs.

cs.CV↗

arXiv2025

Towards Interfacing Dark-Field X-ray Microscopy to Dislocation Dynamics Modeling

Deformation gradient tensor fields are reconstructed in three dimensions (mapping all 9 tensor components) using synthetic Dark-Field X-ray Microscopy data. Owing to the unique properties of the microscope, our results imply that the evolution of deformation fields can now be imaged non-destructively, in situ, and within deeply embedded crystalline elements. The derived regression framework and sampling scheme operate under the kinematic diffraction approximation and are well-suited for studying microstructure evolution during plastic deformation. We derive the deformation conditions under which diffraction vectors extracted from DFXM images can be uniquely associated to the deformation gradient tensor field of the sample. The analysis concludes that the deformation gradient tensor field must vary linearly over line segments defined by the X-ray beam width and the diffracted ray path. The proposed algorithms are validated against numerical simulations for realistic noise levels. Reconstructions of a simulated single straight-edge dislocation show that the Burgers vector components can be recovered with an error of <2%. The mean absolute error of the reconstructed elastic distortion field was found to be <10^-6. By taking the curl of the elastic distortion field, local dislocation densities are derived, yielding a reconstructed dislocation core position with sub-pixel accuracy. The significance of directly measuring the elastic distortion and the dislocation density tensor fields is discussed in the context of continuum theory of dislocations. Such measurements can also be interfaced with continuum dislocation dynamics by providing data that can guide the development and validation, thus extending the relevant models to finite strain regimes.

cond-mat.mtrl-sci↗

arXiv2024

Observing formation and evolution of dislocation cells during plastic deformation

During plastic deformation of metals and alloys, dislocations self-organise in cells, which subsequently continuously decrease in size. How and when these processes take place has remained elusive, because observations of the structural dynamics in the bulk have not been feasible. We here present X-ray diffraction microscopy movies of the structural evolution during tensile deformation of a mm-sized aluminium (111) single crystal. The formation and subsequent development of 40,000 cells are visualised. We reveal that cells form in a stochastic and isotropic manner already at 1% strain. We show that the cell size and dislocation density distributions are log-normal and bi-modal Gaussian distributions, respectively, throughout. This insight leads to an interpretation of the formation and evolution steps in terms of universal stochastic multiplicative processes. This work will guide dislocation dynamics modelling, as it provides unique results on cell formation.

cond-mat.mtrl-sci↗