This study compares Ti-Zr and Ti-Ce composite deoxidation in hot-rolled low-alloy steel. Both generate dispersed submicron oxides, but Ti-Ce exhibits superior intragranular ferrite nucleation, refining grain size from -40.6 mu m to -28.7 mu m. Ti-Ce steel also shows -24% lower pole density and -84% relative more high-angle grain boundaries, achieving -194.95 J impact energy--19.4% higher than Ti-Zr at -20 degrees C. The synergy of grain refinement, more HAGBs, and optimized stress explains the enhanced toughness under continuous hot rolling with ultra-fast cooling. This work supports inclusion engineering for high-toughness low-alloy steels.
The accuracy of sheet metal forming simulations is highly dependent on the selected yield criterion. However, most existing criteria lack flexible control over yield surface curvature, thereby limiting their ability to capture the anisotropy under complex biaxial loading. To address this, a novel stress-triaxiality dependent anisotropic-asymmetric yield criterion is proposed under non-associated flow rule. The new model can be reduced to an isotropic form that allows for the broadest adjustment of yield surface curvature under near-plane strain stress state among existing isotropic criteria. The anisotropy coefficients of the new yield criterion can be determined analytically from experimental data. The model was validated across various sheet metals, including high-strength steel, aluminum alloy, magnesium alloy, and titanium alloy. Compared with the evaluated advanced yield criteria, the proposed model achieves superior prediction accuracy for yield stresses under six distinct stress states and characterizing the anisotropic hardening behavior. Finally, the new model was implemented in ABAQUS for V-bending and tube hydroforming simulations. Finite element results show that the new model accurately predicts both V-bending involving tensile and compressive stress paths and tube hydroforming spanning a wide range of stress states.
Third-element doping is an effective strategy for promoting low-temperature L 10 ordering in FePt nanoparticles, but the underlying dopant-regulated diffusion kinetics and atomic migration mechanisms remain insufficiently understood. In this work, density functional theory (DFT) calculations and experiments are combined to elucidate these processes. The calculations reveal distinct defect-mediated diffusion mechanisms. Fe preferentially migrates through nearest-neighbor vacancy jumping, whereas Pt favours a lower-energy antisite-assisted pathway. A five-step strategy integrating thermodynamic, structural, magnetic, and ordering-transition screening with DFT prediction and experimental verification is established to identify promising dopants. Large-radius dopants reduce migration barriers by more than 20% and enhance low-temperature diffusion coefficients by over three orders of magnitude. Atomic radius is identified as the dominant geometric descriptor of migration barriers, while electronegativity and valence electron number provide secondary electronic contributions. Experimental results further support the predicted trends. Large-radius elements, such as Bi and Pb, can promote L 10 ordering at low temperatures and achieve a high ordering degree and coercivity. In contrast, Mn shows limited ordering, whereas Co-, Ni-, and Cu-doped systems retain the disordered fcc phase, which is broadly consistent with the calculated trends. This work provides a framework for dopant selection and low-temperature synthesis of ordered intermetallic nanoparticles. (c) 2026 Published by Elsevier Ltd on behalf of The editorial office of Journal of Materials Science & Technology.
Multi-connected temporal subgraph querying is essential for discovering patterns in real-world systems (e.g., food delivery, financial markets), yet existing methods assume single-edge connections, leading to inaccurate or incomplete results. This work proposes TSMVI, a vertex-index-based approach that enables precise edge binding without repeated vertex or edge traversal. We prove the problem is NP-complete and design a vertex index structure with an edge-relation-code matrix encoding all 13 Allen interval relations. Evaluated on 5 datasets (3 time-series: 310 similar to 8050 nodes; 2 graphs: 451 similar to 21,303 nodes) with query graphs of 3 similar to 9 vertices, TSMVI achieves 100% query accuracy (improving by 6.4% points over baselines) and achieves 1.53 & times; average query speedup over indexed baselines.
In industrial monitoring, systems typically operate under stable conditions, leading to scarce anomaly samples and thereby limiting the effectiveness of traditional supervised detection methods. However, most existing approaches rely on a single network to process high-dimensional time series, without explicitly capturing fine-grained frequency patterns or complex interdependencies among variables, and often neglect the impact of uncertainty in anomaly detection tasks. To address these challenges, this paper proposes an Uncertainty-aware Frequency-Graph Anomaly Detection (UFGAD) framework for multivariate time series. The framework consists of two complementary branches. First, an uncertainty-aware frequency-Transformer branch is designed to capture refined frequency representations through segment-based multi-angle learning and multi-head attention, while modeling uncertainty from latent representations. Second, a graph deviation network-enhanced graph structure learning branch is introduced to extract inter-feature dependencies using a topological learning mechanism. In addition, a cloud model-based uncertainty quantification strategy is incorporated to measure data-prediction uncertainty within the graph-based framework. Extensive experiments on multiple public datasets demonstrate the strong generalization capability of the proposed method. Furthermore, the framework achieves excellent detection performance in a real-world industrial scenario.
