This study aims to develop a high-generalizability machine learning framework for predicting the homogenized mechanical properties of short fiber-reinforced polymer composites. The ensemble machine learning model (EML) employs a stacking algorithm using three base models of Extra Trees (ET), eXtreme Gradient Boosting machine (XGBoost), and Light Gradient Boosting machine (LGBM). A micromechanical model of a two-step homogenization algorithm is adopted and verified as an effective approach to composite modeling with randomly distributed fibers, which is integrated with finite element simulations for providing a high-quality ground-truth dataset. The model performance is thoroughly assessed for its accuracy, efficiency, interpretability, and generalizability. The results suggest that: (1) the EML model outperforms the base members on prediction accuracy, achieving $R^2$ values of 0.988 and 0.952 on the train and test datasets, respectively; (2) the SHapley Additive exPlanations (SHAP) analysis identifies the Young’s modulus of matrix, fiber, and fiber content as the top three factors influencing the homogenized properties, whereas the anisotropy is predominantly determined by the fiber orientations; (3) the EML model showcases good generalization capability on experimental data, and it has been shown to be more effective than high-fidelity computational models by significantly lowering computational costs while maintaining high accuracy.

High porosity phenomena of U-10Mo fuel foil appear in the U–Mo/Al monolithic fuel plate under deep burnup. In this study, the fuel skeleton creep-based bubble growth model is further improved with the intrusion effect of solid fission products, multiplying the reduction factor in the bubble volume to obtain the bubble pressure. With the locally enhanced irradiation creep of the Mo-depleted region considered, a simulation of the thermo-mechanical coupling behaviors of the monolithic fuel plate L1P7A0 is carried out, based on the commercial finite element (FE) analysis code ABAQUS. A fission-induced creep rate coefficient of 250 × 10⁻²² mm³/(fission·MPa) is identified for the Mo-depleted region, with the predictions of porosity and the thickness deformation of U–Mo fuel foil agreeing well with the experimental data. The research results indicate that: (1) the locally enhanced fuel skeleton creep ability is responsible for the higher porosities near the U–Mo/Zr interface; (2) the entrance of solid fission products into the fission bubbles at high burnup is the dominant factor in inducing high porosity in the regions of the most heavily irradiated fuel foil, especially near the fuel foil edge bearing the elevated external hydrostatic pressures; (3) with the intrusion effect of solid fission products considered, the prediction of the porosity increases from ~15% to ~35% near the fuel foil edge; (4) the intrusion of solid fission products leads to extra differences between the bubble pressure and the external pressure, and simultaneously results in the strengthened fuel skeleton creep deformation contributions to the bubble growth.

The quality factor is a key metric for evaluating the performance of micro-electro-mechanical and nano-electro-mechanical system resonators. Although reducing device dimensions enhances resonator sensitivity, it also intensifies energy dissipation, thereby degrading the quality factor. Among the dominant dissipation mechanisms in micro-/nano-resonators, support loss is strongly influenced by surface effects at small scales, particularly the surface elastic modulus and initial surface stress. In this study, support loss in a double-clamped micro-/nano-beam resonator with surface effects incorporated is investigated. A dynamic model incorporating the surface elastic modulus and initial surface stress is developed based on Euler–Bernoulli beam theory and Gurtin–Murdoch surface elasticity theory. A quality-factor calculation method is then established by combining elastic wave radiation in the supports with an energy-based formulation. The theoretical predictions are validated using a three-dimensional finite element model comprising the resonator with a surface layer, supports, and a perfectly matched layer. Further, the effects of the surface elastic modulus, initial surface stress, characteristic size, and dimensionless geometric parameters on support loss are examined. The results show that the surface elastic modulus and initial surface stress increase support loss and reduce the quality factor, with the initial surface stress exhibiting the stronger influence. The influence of these parameters becomes more pronounced as the characteristic size decreases, and variations in the length-to-thickness and width-to-thickness ratios further modify their contribution to support loss. Mechanistically, surface effects alter the effective bending stiffness and axial force, thereby changing the dynamic loads transmitted to the supports and resulting elastic-wave radiation. These findings provide theoretical insights for the design of high-quality-factor micro-/nano-devices.

Fission-gas-induced porosity significantly affects the irradiation creep behavior of U-10Mo alloy, which is critical for the safe and reliable design of fuel elements. In this study, a representative volume element (RVE) model with randomly distributed pores is developed to investigate the porosity-dependent irradiation creep of U-10Mo alloy. Based on homogenization theory and macroscale dimensional variation, the effective creep strain and creep rate are obtained, and the contributions of the U-10Mo skeleton and pores to the macroscale creep are quantitatively distinguished. Systematic parametric studies examine the effects of fission rate, tensile stress, and porosity on the effective irradiation creep response. The results show that (1) the effective creep strain scales linearly with both fission rate and applied stress and increases significantly with porosity; (2) the creep deformation of the U-10Mo skeleton dominates the macroscale creep, contributing from approximately 85% at 10% porosity to approximately 63% at 25% porosity, while the pore contribution increases with increasing porosity; (3) a porosity-dependent effective irradiation creep rate model is established, with only two fitting parameters, and it exhibits good internal consistency and extrapolability. Additional simulations at porosities of 5%, 18%, and 28% agree well with the fitting curve. The model is further validated by a U-10Mo/Al monolithic fuel element irradiation-induced deformation simulation, where the predicted thickness deformation shows reasonable agreement with experimental data. This work provides a practical and physically grounded model for multi-scale thermo-mechanical coupled analyses and engineering-scale fuel performance simulations of porous U-10Mo fuels.