Abstract The morphological characteristics of Nd–Fe–B powders critically determine the packing behavior, stress transmission, and final magnetic properties during the orientation compaction process. However, direct numerical simulation of the coupled magnetic-mechanical alignment remains computationally prohibitive and physically challenging. This study decouples the problem by employing discrete element method (DEM) simulations to specifically investigate the mechanical-induced de-orientation mechanisms during the pressing stage, while treating the DEM particles as scale-up agglomerates with rigorous dimensionless number similarity analysis to maintain physical consistency with real powder systems. The Hysteretic Spring contact model was adopted to capture the elastoplastic deformation behavior of metallic agglomerates, with model parameters calibrated via angle-of-repose experiments (optimal rolling friction coefficient: 0.1226, simulation-experiment deviation <2%). Dimensionless analysis confirms that the simulated agglomerates match real powders in terms of packing fraction and effective internal friction angle, validating the scale-up approach. Simulation results reveal that reducing the agglomerate size from 2.8 mm to 2.0 mm (with correspondingly narrower size distribution) decreases green compact porosity by 12.13%, average compressive stress by 31.44%, and average torque by 50.20%. Advanced micromechanical analyses demonstrate that finer, uniformly distributed powders exhibit more gradual coordination number evolution during compaction, avoiding abrupt stress concentration and associated shear-induced rotation. The simulation validity was confirmed through Huang Peiyun’s double-logarithmic pressing equation with correlation coefficients R 2 approaching unity. Experimental validation using five powder groups (sauter mean diameter (SMD) ranging from 2.530–3.607 μ m) demonstrates that the favorable powder morphology range occurs at SMD of 2.500–2.700 μ m with D99/D10 < 7 under the processing conditions investigated in this study, arising from the competing effects of magnetic alignment efficiency (dominated by agglomeration at fine end) and mechanical retention efficiency (dominated by stress concentration at coarse end). Microstructural analysis qualitatively corroborates the simulation predictions. This work establishes a validated DEM framework specifically addressing the de-orientation aspect of compaction and provides quantitative guidance for powder morphology optimization in high-performance sintered Nd–Fe–B magnets.
Abstract This study combines detailed experimental characterization with image-based finite element modelling to investigate the compressive micromechanical response of hypereutectic Al–Mg 2 Si composites containing 15–30 wt.% Mg 2 Si. Optical microscopy and scanning electron microscope (SEM)–energy-dispersive x-ray spectroscopy (EDS) observations reveal a three-phase microstructure consisting of an α -Al matrix, eutectic (Al + Mg 2 Si), and primary faceted Mg 2 Si particles. The size, morphology, and distribution of these phases strongly influence the mechanical response and damage evolution. Hardness, compressive yield strength (CYS), and ultimate compressive strength increase with increasing Mg 2 Si content owing to enhanced load transfer, increased matrix constraint, and the higher volume fraction of hard Mg 2 Si reinforcement. However, increasing Mg 2 Si content also promotes the formation of coarse, dendritic, clustered, and hollow primary Mg 2 Si particles, resulting in increased stress concentration and reduced deformation capability. Fractographic analysis indicates that particle fragmentation and interface-related damage are associated with sharp particle corners, internal hollows, eutectic bridges, and particle–matrix interfaces, where deformation incompatibility is most pronounced. Image-based representative volume element models employing a Ramberg–Osgood constitutive formulation reproduce the experimental compressive stress–strain response and CYS with good agreement. The simulations reveal progressively increasing stress concentration, equivalent plastic strain localization, stress-triaxiality gradients, and localized plastic-dissipation zones with increasing Mg 2 Si content, providing a micromechanical interpretation of the experimentally observed deformation behaviour. The numerical model identifies regions susceptible to damage initiation but does not explicitly simulate particle fracture, interfacial debonding, or crack propagation. Localized plastic dissipation is interpreted as an indicator of intense irreversible deformation and damage susceptibility rather than the driving mechanism for fracture. Overall, increasing Mg 2 Si content enhances the load-bearing capability and compressive strength of the composites while promoting deformation localization that reduces strain-to-failure. The combined experimental and numerical approach provides a robust framework for correlating microstructural morphology with localized micromechanical response in hypereutectic Al—Mg 2 Si composites.
Abstract Acoustic strain waves play a key role in enabling high-resolution, non-destructive characterization of thin film multilayer structures. This leverages picosecond sonar, where a pump laser pulse generates the strain wave that travels through the sample. A thorough understanding of strain wave propagation, including wave shape evolution, remains essential for precise data interpretation. We numerically studied strain wave propagation in various samples composed of Ti, Si 3 N 4 , SiO 2 , and Si layers. Two physical models were applied and solved using different numerical methods: a hydrodynamic model implemented with a Lagrangian finite difference code, and a thermoelastic model solved via the finite element method. The models yield consistent or equivalent results across the three cases studied: interaction of a strain wave with a single discrete interface, interaction with a thin interlayer, and interaction with a gradient interface.
Abstract The coupled diffusion and deformation characterizing diffusive processes on the nanoscale provide significant challenges to many simulation methods due to the combination of fast atomic vibrations and slow diffusion. Diffusive molecular dynamics (DMD) models such processes by a scale separation modeling the thermal vibrations as a spatial position probability distribution and diffusion as atom sites with variable concentrations, which evolve based on kinetic equations. This study introduces a novel type of fully anisotropic spatial probability distribution to improve the approximations for atom sites with anisotropic potential energy wells, e.g. close to surfaces or lattice defects and atoms in lattices with a relatively high degree of intrinsic anisotropy. A comparative study between the classic isotropic and newly introduced anisotropic probability distributions is conducted for the case of magnesium hydride with molecular dynamics simulations using the same interatomic potential serving as reference solution. Tests range from lattice and elastic constants of perfect crystals to more complex cases like the bending of nanobeams and reveal previously unreported shortcomings of isotropic DMD, leading to a loss of lattice symmetry and nonphysical behavior. The results obtained using anisotropic shape functions avoid this behavior and follow the reference solution significantly better while maintaining comparable computational effort. This confirms the requirement for DMD to take anisotropic vibrations into account when simulating magnesium hydride and other systems with anisotropic atom environments.
Abstract Elucidating the mechanism of electroplasticity is essential for the widespread application of electrically assisted forming technology to hard-to-deform metallic components. In this study, a two-dimensional discrete dislocation dynamics (DDD) model incorporating the electroplastic effect is developed by introducing non-uniform temperature fields induced by Joule heating. The effects of Joule heating and the electron wind on dislocation nucleation, glide, and thermal activation escape are systematically quantified from both thermal and energetic perspectives. Dislocation evolution in titanium alloy single crystals subjected to electric current during uniaxial tensile deformation is investigated across both high and low strain rate regimes using this electroplastic-DDD framework. The results show that the non-uniform temperature fields modify the dislocation source strength and drag coefficient, thereby influencing dislocation nucleation and glide. Furthermore, the additional Joule heating energy E Joule and electron wind energy E wind supplied to dislocations reduce the energy barrier for thermal activation escape, promoting dislocation depinning. It is demonstrated that the contribution of E wind is approximately seven orders of magnitude smaller than that of E Joule , with Joule heating identified as the dominant mechanism governing the electroplastic response in titanium alloys.
