Three-dimensional probabilistic site characterization relies on geotechnical measurements obtained from indoor laboratory and in-situ tests. Previous probabilistic site characterization studies assumed that these geotechnical measurements are lattice data. In engineering practice, however, geotechnical data are often nonlattice because of inconsistent sampling intervals, measurement errors, and missing or incomplete records at certain depths. Recently, the authors proposed a three-dimensional probabilistic site characterization method based on multi-outputs sparse Bayesian learning (3D-MSBL), which accounts for both spatial variability across locations and cross-correlations among geotechnical properties. However, the original 3D-MSBL method requires lattice input data to enable Kronecker product-based computation, thereby limiting its applicability to the nonlattice data commonly encountered in geotechnical engineering practice. To address this limitation, this paper extends the original 3D-MSBL framework to accommodate nonlattice geotechnical data. Based on the stochastic expectation–maximization algorithm, tailored imputation strategies are developed to treat the missing data, which are categorized into two types. The proposed method is examined on three case studies to demonstrate its effectiveness. The results show that the proposed method can provide reasonable imputation results for missing data and enables the imputed lattice data to support reliable predictive performance of the 3D-MSBL model. The proposed method offers a practical and robust framework for probabilistic site characterization using nonlattice geotechnical data.
Abstract This study proposes a novel energy-dissipation isolation bearing, which is composed of Z-shaped dampers and a lead-core rubber bearing. Under cyclic loading, the damper provides additional stiffness and energy dissipation for the support through hysteresis behavior. The seismic performance of the novel energy dissipation isolation bearing was tested through compression-shear tests, taking into account the influence of the material type of the damper and the loading direction. In addition, a finite-element model was established to verify the test results. The results show that the novel energy-dissipation isolation bearing has a stable hysteresis curve and good energy dissipation performance. Under shear strains of 50, 100, and 150%, compared with LRB, the energy dissipation improvement range of the novel energy-dissipation isolation bearing remains stably within the range of 16–35%. The in-plane damper dissipates energy through bending-tensile deformation. The damper in the out-of-plane direction dissipates energy through bending and torsional deformation. Compared with Q235 steel dampers, iron-based shape-memory alloy dampers have a more significant improvement in energy dissipation performance, horizontal bearing capacity, and equivalent horizontal stiffness, and have smaller residual displacement. The numerical model accurately represents the hysteretic behavior of the novel energy-dissipating isolation bearing and effectively captures the stress concentration regions in the Z-shaped dampers. The predicted results show good agreement with the experimental data. In addition, the hysteresis characteristics of the novel energy-dissipation isolation bearing are almost isotropic, enabling it to resist seismic forces in any horizontal direction.
Pulmonary embolism (PE) detection in Computed Tomography Pulmonary Angiography (CTPA) remains challenging due to the presence of small embolic lesions, low contrast, and complex three-dimensional vascular structures. To address these challenges, we propose IVT-PENet, an Information-Gated and Variance-Enhanced U-Net with Direction-Aware Positional Encoding and a Channel-Spatial Transformer. First, a Direction-Aware Positional Encoding (DAPE) module is embedded throughout the encoder–decoder pathway to preserve directional anatomical priors and maintain spatial consistency. Second, an Information-Gated and Variance-Enhanced Atrous Spatial Pyramid Pooling (IG-VEASPP) module is developed to enhance multi-scale contextual representation and improve the discrimination of subtle embolic candidates. Third, a Channel-Spatial Attention Transformer (CSAT) is introduced at the bottleneck to capture long-range three-dimensional contextual dependencies through sequential channel–spatial attention. In addition, a dynamic probability-annealing strategy is employed to adaptively regulate random noise and intensity perturbations during training, thereby improving model robustness. The proposed framework adopts a detection-oriented design, where vessel-aware anatomical guidance and embolic candidate generation jointly facilitate the subsequent detection process. Experimental results on the CAD-PE dataset demonstrate that IVT-PENet achieves a sensitivity of 0.895, a FROC score of 0.812, and an AUC of 0.851. Furthermore, without any retraining or fine-tuning, IVT-PENet attains a sensitivity of 0.942, a FROC score of 0.764, and an AUC of 0.881 on the external FUMPE dataset, demonstrating strong cross-dataset generalization capability. These results indicate that IVT-PENet effectively captures multi-scale embolic characteristics and provides reliable pulmonary embolism detection in complex vascular environments. The source code is available at https://github.com/GuYuIMUST/IVT-PENet.
Lithium-rich layered oxides (LLOs) have emerged as promising cathode materials for lithium-ion batteries due to their high capacity and low cost. However, their practical application is hindered by inherent structural instability and poor cycling performance. Addressing the cost and environmental concerns associated with Co reliance in conventional cathode materials, this work investigates a series of Co-free Li28FelNimMn20-l-mO48 (LF2NMO for l = 2, m = 5; LF4NMO for l = 4, m = 4; LF6NMO for l = 6, m = 3) systems. The influences of Fe-doping concentration and the penetrated-Fe number into the Li2MnO3 domain on delithiated structures, lattice stress, oxidation processes of transition metal (TM) ions and oxygen ions, oxygen Bader charge evolution, oxygen release enthalpy and structural stability are systematically examined using DFT calculations. Results demonstrate that the LF4NMO-P(1Fe)(the number 1 in 1Fe is the penetrated-Fe number into the Li2MnO3 domain) composition, exhibits the optimal structural stability. Upon the full delithiation it shows the smallest volume change of 0.23
Ferroelectric materials offer a transformative solution for developing high-responsivity photodetectors by leveraging their unique photovoltaic effect. However, the low output photocurrent that is associated with carrier separation efficiency is the main issue that restricts the applications of ferroelectric photodetectors. Here, constructing a MoS2/YMO/LSMO heterostructure, a high photocurrent density of 11.9 mA/cm2 was achieved under standard solar illumination, representing an approximate 3.4-fold increase compared to the 3.50 mA/cm2 observed in the YMO/LSMO structure. The significantly enhanced photocurrent output further improved the heterojunction’s photodetection performance, exhibiting a responsivity (R) of 0.72 A/W and a specific detectivity (D*) of 2.28 × 1012 Jones. Furthermore, the device demonstrated excellent broad-spectrum detection capability, with a photoresponse range spanning from the ultraviolet (UV) to near-infrared (NIR) spectral region. The microscopic mechanism underlying the performance enhancement originates from the built-in electric field at the MoS2/YMO interface strengthening the driving force for carrier separation, while MoS2 acting as an electron transport layer significantly improves the transport efficiency of charge carriers. This work provides novel insights into developing high output photocurrent photodetectors.
