Machine learning has been widely applied to various concrete materials tasks, such as macroscopic performance prediction and mix proportion optimization. The performance of machine learning models depends strongly on the quantity and quality of the available datasets. However, constructing large-scale and high-quality concrete mix proportion datasets remains challenging because concrete experiments are time-consuming, costly, and easily affected by material sources, curing conditions, and human operations. At present, small-sample modeling problems are mainly addressed through two strategies: model optimization and data augmentation. Model optimization methods mainly improve the extraction of information from existing small-sample datasets by adjusting model structures or training strategies. However, they do not fundamentally increase the amount of data, and their improvement in prediction accuracy is still limited by the original sample size. Conventional data augmentation methods mainly rely on generative adversarial networks, which learn the real data distribution through adversarial training between a generator and a discriminator, thereby generating virtual samples to expand the original dataset. Nevertheless, when dealing with complex tabular data, generative adversarial networks may encounter unstable training, insufficient sample diversity, and distribution shifts in generated data, resulting in limited consistency between virtual and real data. To address these issues, this study proposes MixDiff, a denoising diffusion probabilistic model (DDPM) for generating virtual concrete mix proportion data under small-sample conditions. This model aims to generate high-quality virtual concrete mix proportion data.

A small-sample dataset containing 100 sets of concrete mix proportions was first established. The input features included water, cement, fly ash, mineral fines, river sand, manufactured sand, coarse aggregate, water-reducing agent, total binder content, water-to-binder ratio, total fine aggregate content, and sand ratio. The target variables included 7 d compressive strength, 28 d compressive strength, slump, expansion degree, initial setting time, and final setting time. MixDiff was developed using a fully connected neural network and a denoising diffusion probabilistic model. The model consisted of a sinusoidal timestep embedding module and four stacked residual fully connected blocks. Each residual block contained fully connected layers, batch normalization layers, and SiLU activation functions. In the forward diffusion process, Gaussian noise was progressively added to the normalized data according to a cosine variance schedule. In the reverse generation process, a neural network was used to predict the noise and progressively reconstruct concrete mix proportion samples with realistic distribution characteristics from standard Gaussian noise. Subsequently, the quality of virtual data generated by MixDiff was quantitatively evaluated from two perspectives: marginal distribution and feature correlation. Under five-fold cross-validation and grid search, several mainstream machine learning models were further used to compare the prediction performance of the baseline mode and the augmentation mode. Finally, 15 mix proportions were selected from the generated virtual dataset for experimental validation to evaluate the engineering plausibility of the generated data.

The quantitative evaluation results showed that the virtual dataset generated by MixDiff showed high similarity to the real dataset. The average Wasserstein distance, K-S statistic, and J-S divergence between the two datasets were 0.019 5, 0.095 2, and 0.120 7, respectively. The Frobenius distance between their Pearson correlation matrices was 0.332 7, indicating that MixDiff effectively preserved the marginal distribution characteristics of the real data and the correlation structure among variables. The machine learning comparison results showed that the prediction performance of all models improved significantly after the introduction of MixDiff-generated data. In the baseline mode, the average R² of the seven models was 0.817 1, and increased to 0.953 7 in the data augmentation mode. Meanwhile, the average RMSE decreased from 14.20 to 7.39, and the average MAE decreased from 8.77 to 4.66. Correspondingly, R² increased by 16.72%, while RMSE and MAE decreased by 47.96% and 46.86%, respectively. Among the machine learning models, the deep neural network achieved the best performance after data augmentation, with R², RMSE, and MAE values of 0.959 4, 7.00, and 4.52, respectively. In addition, the standard deviations of the five-fold cross-validation results were significantly reduced in the data augmentation mode. This indicates that MixDiff-generated data not only improved the prediction accuracy of machine learning models but also enhanced model stability and generalization ability. Experimental validation of 15 virtual mix proportions showed that the data generated by MixDiff had high plausibility in six key performance indicators, including 7 d compressive strength, 28 d compressive strength, slump, expansion degree, initial setting time, and final setting time. The average MAPE of each performance indicator was lower than 5%, and the overall average MAPE was 3.8%.

This study proposed MixDiff, a denoising diffusion probabilistic model for concrete mix proportion data generation. The results show that the virtual data generated by MixDiff were highly consistent with the real data in terms of marginal distribution and feature correlation. The Frobenius distance between the Pearson correlation matrices was only 0.332 7. After introducing MixDiff-generated data, the prediction performance of multiple machine learning models, including RF, SVR, KNN, XGBoost, LightGBM, CatBoost, and DNN, was significantly improved, with R² values of all models increasing above 0.94. Based on the average performance of the seven machine learning models for the six output performance indicators, R² increased by 16.72%, while RMSE and MAE decreased by 47.96% and 46.86%, respectively. Among these models, DNN achieved the highest R² and the lowest RMSE in the data augmentation mode. Experimental validation of 15 virtual mix proportions showed that the generated results were highly consistent with the measured results in terms of 7 d compressive strength, 28 d compressive strength, slump, expansion degree, initial setting time, and final setting time. The MAPE of each performance indicator was below 5%, and the overall average MAPE across the six performance indicators was 3.8%. These results indicate that MixDiff can learn the underlying relationships between concrete mix proportions and macroscopic properties, and it can serve as an effective method for concrete mix proportion data augmentation under small-sample conditions.

The chemical composition of cement clinker, particularly the percentage of key oxides, is crucial for evaluating cement quality and is vital for engineering-accident liability tracing. However, traditional methods for chemical composition analysis rely heavily on unhydrated original cement samples, which are often missing, improperly stored, or ambiguously labeled in engineering management, resulting in an insufficient basis for subsequent identification. Once cement hydrates, active minerals undergo irreversible reactions, making it challenging to invert the original clinker composition directly from the hardened paste. In contrast, backscattered electron (BSE) images of the hardened cement paste contain sufficient physical information to distinguish the original chemical compositions of different cement clinkers, thus providing a potential pathway for post-event tracing. Although deep learning methods have been explored for microstructural image mapping to infer macroscopic properties, conventional regression models are typically "black-box" point predictors that cannot quantify prediction uncertainty. The conditional probability density estimation method can not only provide the most probable predicted value, but also output the complete estimated probability-density curve along with its confidence assessment, thereby enabling quantitative characterization of prediction uncertainty. Accordingly, this study proposes a conditional probability density estimation method based on a Deconvolutional Density Network (DDN). Utilizing BSE images of hardened paste together with encoded water-to-cement (W/C) ratios and curing ages as conditional inputs, this model directly predicts the full conditional probability density of the main clinker oxide contents, providing a new objective and quantitative pathway for quality liability tracing.

Cement clinker samples from 47 different plants were used to prepare cement paste specimens with W/C ratios ranging from 0.25 to 0.70 and curing ages from 1 d to 28 d. The hydration was terminated by immersion in anhydrous ethanol, and the samples were polished for subsequent scanning. BSE images were acquired via a JSM-7610F field-emission SEM at 200× magnification and an accelerating voltage of 15 kV. For each sample, two scans were performed, and nine non-overlapping regions were captured per scan, resulting in a dataset of 844 independent BSE images. The proposed model consists of a feature extraction layer based on a residual network (ResNet) with self-supervised learning and cross-layer skip connections, a conditional encoder, and deconvolutional layers. The ResNet extracts multiscale morphological features from local textures to global patterns. The conditional encoder embeds the W/C ratio and curing-age conditions into the image features. The DDN is then employed to output the conditional probability density distributions for the percentages of eight oxides (CaO, SiO₂, Al₂O₃, Fe₂O₃, SO₃, Na₂O, MgO, and K₂O). The deconvolutional layers map the fused features to a smooth probability distribution over 256 discretized intervals. To evaluate the model, accuracy (ACC) under different confidence levels (α), mean absolute error (MAE), and mean squared error (MSE) were adopted as evaluation metrics.

The estimation results of the proposed model indicate that, at a confidence level of α = 90%, the model achieves an accuracy of over 90% in estimating the percentage contents of five oxides (SiO₂, Al₂O₃, Fe₂O₃, SO₃, and Na₂O) on the test set. When the confidence level is relaxed to α = 75%, the estimation accuracy for all eight oxides exceeds 90%. This could be attributed to the DDN's capability of modeling complete probability density distributions instead of simple point estimates, effectively resolving the nonunique-solution problem stemming from the intrinsic information ambiguity of BSE imaging. In terms of MAE and MSE metrics, the model exhibits relatively higher absolute errors for components with higher contents, such as CaO. This is generally attributed to the nature of absolute error metrics, which are positively correlated with the oxide content. Conversely, low-content oxides like Na₂O show smaller estimation errors, possibly because their larger fluctuation ranges among different samples allow the model to capture clearer mapping relationships. In addition, ablation experiments further reveal that removing the W/C ratio and age inputs leads to a decrease in estimation accuracy across all confidence levels, demonstrating the effectiveness of the conditional encoder in integrating critical hydration parameters. Compared with the deterministic ResNet regression model, the proposed DDN model exhibits significantly better stability of estimation performance under higher precision requirements. This result demonstrates that the DDN model effectively mitigates the inherent uncertainty in the mapping from BSE images to chemical compositions, a limitation typical of deterministic regression models that lack reliability quantification.

The main conclusions of this study are summarized as follows. The proposed probability density estimation method successfully maps BSE images of hardened cement paste to the oxide contents of clinker. Experimental results demonstrate that the model achieves an accuracy of over 90% for five oxides at a confidence level of 90%. The adoption of the Deconvolutional Density Network enables the model not only to output the most probable value of each oxide content but also to provide a complete probability density curve with corresponding confidence assessment, thereby overcoming the limitation of traditional regression models that lack quantifiable reliability for quality tracing applications. Moreover, the ResNet-based feature extraction with cross-layer skip connections effectively captures multiscale features from BSE images, while the integration of W/C ratio and curing age via the conditional encoder significantly enhances the model predictive capability. This study demonstrates that the probabilistic retrospective estimation based on BSE images and DDN offers a robust and objective means to quantitatively infer the original clinker composition from hardened paste. As such, it holds promise as a transformative technique for cement-quality arbitration in engineering-accident liability investigations.

Thermodynamic modeling has become a useful tool for understanding hydration, phase evolution, modification mechanisms, and the durability of magnesia cements. It can predict stable phases, pore solution chemistry, saturation indices, and phase stability boundaries under defined conditions, providing theoretical support for mix proportion design and performance optimization. Existing thermodynamic modeling studies have investigated the effects of raw-material mix proportions, temperature, modifiers, carbonation, and aggressive ions on the phase assemblages of M-S-H, MPC, MOS, MOC, and RMC systems, and considerable progress has been made in understanding the formation and transformation of M-S-H/M-(A)-S-H gels, magnesium phosphate hydrates, basic magnesium sulfate and chloride hydrate phases, and magnesium carbonate hydrates. However, several limitations remain. Thermodynamic data for M-S-H/M-(A)-S-H gels, magnesium carbonate hydrates, metastable intermediate products, and the temperature-dependent parameters of some hydration products are still incomplete. In addition, high-ionic-strength pore solutions require appropriate activity coefficient models, but Pitzer parameters for multicomponent systems remain insufficient. Moreover, thermodynamic modeling mainly reflects equilibrium or quasi-equilibrium states and cannot directly describe MgO dissolution, nucleation, growth, transport, or early hydration kinetics. Future work should focus on improving thermodynamic databases, refining activity models for high-salinity systems, and coupling thermodynamic modeling with experimental characterization, kinetic modeling, and transport simulations. This will enhance its reliability in guiding material design, durability assessment, and low-carbon applications.

Fluorophosphate (FP) laser glass offers high rareearth solubility, low nonlinear refractive index, and a large Abbe number, making it a promising gain medium for high-power fiber lasers. The determination of the glass-forming region is fundamental to developing new laser glass systems. Previous studies have relied heavily on trialanderror approaches, yielding only limited data at prohibitive cost in terms of manpower, finances, and material resources. In this work, thermodynamic calculations are employed to investigate the glass-forming region of multicomponent FP glass systems. The results reveal that thermodynamic calculations can effectively predict the glass-forming region of FP glasses. Moreover, differences in cation field strength and interactions with anions are identified as critical factors governing the glassforming ability and structural stability of FP glasses. Based on these theoretical predictions, a ytterbiumdoped, high-gain FP laser glass and fiber are designed and fabricated through compositional optimization, and initial 1.0 μm laser output is successfully demonstrated.

The thermodynamic calculations are based on the free energy balance between liquid and solid phases in binary systems. By solving the equilibrium equation, the melting temperature is obtained as a function of composition, and the eutectic point is determined by locating the minimum of this function. For ternary and higher-component systems, the binary eutectic points are treated as “quasi-compounds”, and the calculation is performed iteratively to construct the theoretical glass-forming region. The predicted regions were validated against experimental data reported in the literature. Raman spectroscopy combined with peak-fitting analysis was used to quantify Qn species, while density and refractive index measurements were carried out for physical property evaluation. Yb³⁺-doped FP glasses and fibers were then fabricated based on the predicted glass-forming region, and their laser performance was characterized.

Quantitative comparison between predicted and experimental glass-forming regions shows overlap ratios of 74.5%, 100%, and approximately 100% for the three systems examined, confirming that the method effectively locates the main compositional range of glass formation. Raman analysis reveals a transition from Q²-dominated chain structures at high phosphate contents to Q⁰/Q¹ units at low phosphate contents, with the Q² fraction decreasing from 73.4% to 8.4% as phosphate content drops from 80%(mole fraction) to 40%. Cation field strength is identified as a critical factor: the higher field strength of Zn²⁺ (0.43) relative to Ba²⁺ (0.24) leads to stronger network bonds but greater sensitivity to composition variation, resulting in a narrower glass-forming region. Guided by the theoretical predictions, Yb³⁺-doped FP fibers were fabricated with a peak gain coefficient of 5.56 dB/cm at 1064 nm and stable dual-wavelength laser emission at 1023.3 nm and 1031.5 nm under 974 nm pumping.

This work demonstrates that thermodynamic calculations based on eutectic point prediction provide a valid pre-screening tool for locating glass-forming regions in multicomponent FP systems. Cation field strength governs the glass-forming window and structural stability. Guided by theoretical predictions, Yb³⁺-doped FP fibers were successfully fabricated with stable 1.0 μm laser output, validating the effectiveness of this combined theoretical-experimental approach for developing novel laser glass materials.

Insulating materials play a crucial role in high-voltage power transmission, protecting equipment and ensuring the reliability of the power grid by effectively isolating current and voltage. Conventional ceramic insulators (such as porcelain or aluminium oxide insulators) are widely used in high-voltage power transmission, but they have inherent limitations, including a short service life, high manufacturing costs and many surface defects, which are likely to cause electrical breakdown in the material. Furthermore, defects such as internal porosity in ceramics can lead to deterioration of their mechanical and electrical properties. In contrast, glass insulators, owing to their exceptional mechanical strength and excellent stability, have become one of the materials of choice for long-span transmission lines. Among these, sodium-calcium silicate (NCS) glass offers significant advantages, including high transparency, excellent physicochemical stability, ease of manufacture, and low cost, and thus holds great potential for application in the field of high-voltage power transmission and transformation equipment. However, the mechanical strength and electrical properties of NCS matrix glass are insufficient to meet the demands of modern power equipment for high electrical strength, high mechanical strength and low dielectric loss. Therefore, the glass composition was optimized, and the NCS matrix glass system was modified by doping with B₂O₃ and K₂O. The evolution of the network structure, optical transparency, mechanical strength and electrical properties with varying doping ratios was investigated to develop a material suitable for high-performance, long-life, high-voltage insulators.

The matrix glass composition was an NCS system. Two sets of glass samples were prepared using the high-temperature melting-quenching method. In Group B, B₂O₃ was added at varying molar fractions (0-20%(mole fraction)) while keeping the other components constant. In Group N, using the mixed alkali effect, K₂O was introduced to partially replace Na₂O, while maintaining the total content of Na₂O and K₂O constant. Subsequently, high-purity raw materials (SiO₂, CaCO₃, Na₂CO₃, H₃BO₃, K₂CO₃) were thoroughly mixed, placed in an alumina crucible and melted for 2 h at 1400 ℃. The melt was cast into preheated copper molds and annealed at 500 ℃ for 2 h to relieve internal stresses. The glass was cut into samples measuring 10 mm × 10 mm × 1 mm and 20 mm × 20 mm × 1 mm, and the surfaces were ground flat using a metallographic grinding and polishing machine. The material structure was characterized using differential scanning calorimetry (DSC), X-ray diffraction (XRD), and Raman spectroscopy. The optical properties were determined by UV-vis spectrophotometry. Mechanical strength was measured using a three-point bending test. The electrical properties included volume resistivity (measured at room temperature using a resistivity meter), dielectric loss, and dielectric breakdown strength (under DC voltage).

The addition of B₂O₃ significantly alters the glass network structure. When a small amount of B₂O₃ is introduced, its [BO₃] triangular units tend to replace some of the [SiO₄] tetrahedra of SiO₂, forming an irregular boron-oxygen network structure. This structural transformation reduces the proportion of bridging oxygen atoms in the glass network, leading to a decrease in network connectivity. As the B₂O₃ content increases further, [BO₄] tetrahedra gradually form, the number of bridging oxygen atoms increases, the glass network structure becomes denser, and the structural stability of the glass is enhanced. This structural change enhances the rigidity of the network. When 5%(mole fraction) B₂O₃ was added, the breakdown strength of the sample reached a maximum of approximately 31.93 kV/mm, with a flexural strength of 108.37 MPa. However, when the B₂O₃ content was excessive, a large number of [BO₃]/[BO₄] units appeared in the glass, leading to the breaking of the Si-O-Si chains and a decrease in network polymerization, which severely affected glass performance.

Compared to Na⁺ ions, the larger radius of K⁺ ions means that introducing K₂O increases steric hindrance within the glass network, resulting in a denser glass structure and thereby improving flexural strength. When 4% K₂O was added, the sample exhibited a flexural strength of 110.55 MPa, and the dielectric loss reached its lowest value (0.0027). The results indicate that an appropriate amount of K₂O doping for partial replacement of Na₂O can improve the glass's flexural strength and stiffness; however, the introduction of excessive K₂O may lead to an increase in defects within the glass network, thereby reducing the material's mechanical properties.

In summary, the incorporation of an appropriate amount of B₂O₃ into NCS glass, combined with the mixed alkali effect resulting from the substitution of Na₂O by K₂O, is an effective strategy for enhancing the mechanical strength and electrical properties of glass insulating materials. The results indicate that as the B₂O₃ content increases, the material's flexural strength exhibits a marked nonlinear change; when 5% B₂O₃ is added, both the flexural strength and electrical breakdown strength of the material are enhanced. The introduction of an appropriate amount of K₂O (partially replacing Na₂O) alters the polarizability of the glass via the mixed alkali effect, thereby enhancing the material's electrical breakdown strength while maintaining low dielectric loss. When the B₂O₃ and K₂O contents were 5% and 4%, respectively, the prepared sample exhibited high mechanical strength (flexural strength > 110 MPa), excellent insulation properties (room-temperature volume resistivity > 10¹² Ω·m, dielectric loss < 0.003 at 50 Hz) and high electrical strength (breakdown strength > 32 kV/mm). These findings provide a theoretical basis and material-design guidance for glass insulating materials used in high-performance electrical equipment.

Although significant progress has been achieved in understanding oyster-concrete interactions, the dualistic corrosion-protection mechanisms of oyster attachment remain complex and incompletely understood, with the evolutionary dynamics of oysters as dominant sessile organisms on concrete substrates and their lifecycle impacts on concrete integrity subject to ongoing debate owing to the intricate interplay of diverse marine environmental factors, seasonal variations, and developmental phases that lead to differential growth patterns and distinct effects on concrete durability. Future research should prioritize conducting systematic in situ field trials within authentic marine environments to enable precise mechanistic analysis of adhesion and corrosion processes, facilitate quantitative assessment of oyster life-cycle effects on concrete performance, and establish comprehensive models capable of predicting the transition between corrosive and protective outcomes based on critical thresholds of coverage density, biofilm maturity, and local hydrodynamic regimes. Furthermore, although oyster attachment can be detrimental under conditions of low-density colonization, the formation of stable, dense biofilms may confer significant protective benefits to concrete substrates by impeding aggressive aggressive-agent ingress, refining surface pore structure, and mitigating physical erosion, warranting in-depth investigations into this protective mechanism through integration of biological principles such as selective breeding and behavioral conditioning of target oyster species to promote rapid colonization of intertidal concrete surfaces and engineer cohesive fouling communities capable of providing effective long-term material preservation, ultimately contributing to the development of controllable, large-scale bioprotective strategies for sustainable marine infrastructure management.

Ultra-high-performance concrete (UHPC) is widely applied in bridges, prefabricated structures, and advanced construction owing to its high strength, dense microstructure, and excellent durability performance. Its performance is strongly governed by the type and dosage of supplementary cementitious materials (SCMs), including silica fume (SF), ground granulated blast-furnace slag (GGBS), superfine GGBS (sGGBS), fly ash (FA), and fly ash microspheres (sFA), as well as polycarboxylate superplasticizer (PCE) dosage and curing age (CA). However, the relationships among mixture parameters, water distribution, hydration, and strength development are highly nonlinear, and conventional empirical models often fail to provide accurate predictions.

Water film thickness (WFT) is a key microscopic parameter reflecting the free water covering the surfaces of solid particles, directly influencing particle lubrication, dispersion, and fresh-state behaviour in high-solid-volume UHPC systems. Compressive strength (CS) is another essential index for mechanical evaluation. Accurate prediction of both WFT and CS is crucial for mixture design and performance optimization. Nevertheless, UHPC experiments are costly and time-consuming, leading to limited datasets. To overcome this small-sample challenge, this study proposes an integrated framework combining conditional tabular generative adversarial network (CTGAN) for data augmentation, Bayesian optimization (BO)-based eXtreme Gradient Boosting (XGBoost) for predictive modelling, and interpretable machine-learning methods (SHAP and PDPs) for mechanism analysis.

A total of 160 experimental data points were obtained from binary, ternary, and quaternary cementitious systems. The input variables included cement content, PCE dosage, SF, GGBS, sGGBS, FA, sFA, and CA, while the output variables were WFT and CS. P·I 42.5 cement was used as the reference cement, and quartz sand with particle sizes of 0.84-2.00 mm was used as the aggregate. In the binary systems, one mineral admixture was used to replace cement by volume at replacement levels of 0%, 5%, 10%, 15%, and 20%. In the ternary systems, the SF content was fixed at 10%. In the quaternary systems, the GGBS content was fixed at 20% based on the ternary system. WFT was calculated using the wet packing method, and CS was measured after curing for 3, 7, 28 d, and 90 d.

CTGAN was used for data augmentation to expand the original small dataset. The model learned the joint distribution of the real experimental data through adversarial training between the generator and discriminator. After optimization, 1000 synthetic samples were generated for WFT prediction, and 900 synthetic samples were generated for CS prediction. The quality of the synthetic data was evaluated by comparing histograms, cumulative distribution curves, and correlation matrices between real and synthetic datasets.

The synthetic datasets were used to train XGBoost regression models. BO was applied to optimize hyperparameters, including tree depth, learning rate, number of estimators, subsample ratio, column-sampling ratio, minimum child weight, and regularization. Model performance was evaluated by R², MAPE, RMSE, and MAE. SHAP and PDPs were employed to interpret feature contributions and nonlinear effects.

The comparison between real and synthetic datasets showed that CTGAN effectively preserved the main distributional characteristics of the original data. The generated samples had value ranges and distribution patterns similar to those of the real samples. The cumulative distribution curves also showed good consistency between real and synthetic datasets. In addition, the correlation matrices indicated that CTGAN captured not only the marginal distribution of individual variables but also the correlation structure among key variables. Therefore, the generated data could provide a reliable data basis for BO-XGBoost model training under small-sample conditions.

The BO-XGBoost model exhibited excellent prediction performance. For WFT, test R² reached 0.986, with training and test RMSE values of 0.909 3 nm and 0.387 6 nm, respectively. For CS, test R² was 0.976, with RMSE values of 5.244 8 MPa and 3.837 9 MPa. Predicted values clustered around the 45° ideal line, with most samples within ±10% error. Residuals were concentrated and symmetric, indicating no systematic bias, and loss curves showed stable convergence without overfitting. These results demonstrate that CTGAN augmentation effectively improved the ability of XGBoost to capture complex nonlinear relationships under small-sample conditions.

The SHAP and PDP analyses further revealed the influence mechanisms of different variables. For WFT, sFA showed a strong positive contribution, mainly because its fine and spherical particles improved particle dispersion and increased effective water-film formation. SF had a positive effect at low-to-medium dosages, but excessive SF could reduce WFT because of its high specific surface area. GGBS and sGGBS tended to show negative effects at high dosages, which may be related to denser particle packing and compression of effective water-film space. PCE showed a nonmonotonic effect: moderate PCE improved WFT, while excessive PCE could weaken the stability of the water-film structure.

For CS, CA was the most important factor and showed a strong positive contribution, reflecting continuous hydration and microstructure densification with curing time. sGGBS, SF, and sFA generally promoted strength development. sGGBS contributed to strength through microfilling and potential pozzolanic activity, while SF improved strength through pore refinement and matrix densification. In contrast, GGBS and FA showed negative effects at high dosages because excessive cement replacement reduced clinker content and limited the availability of calcium hydroxide (Ca(OH)₂) for secondary reactions. These results demonstrate that different mineral admixtures have dosage-dependent and material-specific effects on UHPC performance.

This study establishes a multi-response prediction and interpretation framework integrating CTGAN, BO-XGBoost, SHAP, and PDPs for UHPC under small-sample conditions. CTGAN effectively generated high-quality synthetic data while preserving original distribution and correlation characteristics. The optimized BO-XGBoost model achieved outstanding predictive accuracy and generalization, with test R² of 0.986 for WFT and 0.976 for CS. The interpretability analysis demonstrated that the effects of mineral admixtures on WFT and CS are highly dependent on material type and dosage: sFA and SF are beneficial for WFT within suitable ranges, whereas sGGBS, SF, and sFA promote CS; excessive GGBS and FA impair CS due to limited reaction activity. The proposed framework offers an effective data-driven approach for UHPC mixture design, performance prediction, and mechanism interpretation when only limited experimental data are available.

This review systematically summarizes recent advances in enhancing the temperature resistance of alumina aerogels through three key strategies: structural-unit engineering, extrinsic doping, and isolation phase construction. These approaches have effectively delayed alumina phase transformation and suppressed high-temperature sintering, increasing the short-term service temperature from approximately 1000 ℃ to as high as 1600 ℃ while maintaining low thermal conductivity and structural integrity at elevated temperatures. However, for applications requiring higher temperatures and longer exposure times, future research should prioritize multiscale synergistic design, infrared opacification, mechanical reinforcement, and deeper investigation of the microscopic mechanisms of high-temperature phase transformation and sintering. Integrating the integration of emerging technologies such as multiscale simulation and artificial intelligence is expected to accelerate the development of alumina aerogels that simultaneously achieve ultrahigh-temperature structural stability and durable thermal insulation functionality.

Future developments in the field are likely to follow three main trajectories. First, standardized definitions and transfer rules for transferable variables are required to reduce inconsistencies among models and to support reproducible coupled workflows. Second, the integration of uncertainty-aware calibration and sensitivity analysis into process-window design is warranted as fundamental constitutive relations and kinetic parameters are constrained by measurement limitations and condition dependence. Third, integrated workflows combining high-fidelity simulation, reduced-order modeling, and data-driven surrogates are essential for fast component-scale screening and optimization. Driven by these advances, full-process modeling can evolve from retrospective interpretation toward prospective control of process-structure-property linkages in ceramic BJ, thereby improving reliability, dimensional fidelity, and performance consistency across diverse material categories.

This review summarizes recent progress on the synthesis, mechanisms, and applications of persistent luminescent and mechanoluminescent nanomaterials. We emphasize that advances across both top-down and bottom-up routes substantially improve control over particle size, morphology, and-critically-the defect/trap structures governing carrier trapping and detrapping. We highlight a trap-centric perspective that links delayed emission in persistent luminescence with stress-assisted carrier release in mechanoluminescence via comparing the physical bases of the two phenomena. Translating these materials to the nanoscale can broad their utility, particularly for autofluorescence-free bioimaging and diagnostics, as well as self-powered, visually intuitive stress sensing and flexible optoelectronic devices. Collectively, these studies illustrate a shift from viewing luminescence as an isolated metric toward engineering integrated, stimulus-responsive nano-systems tailored to application demands.

Some key challenges remain despite a rapid development. A central limitation is the trade-off between achieving precise nanoscale structural control and preserving the deep, stable trap populations needed for strong afterglow or mechanoluminescent output. In addition, a predictive and quantitative understanding of how interfaces, surface states, and localized mechanical fields regulate carrier dynamics in nanomaterials is still lacking. For practical deployment, a long-term stability under environmental and cyclic loading, biocompatibility, and scalable yet reproducible fabrication must be addressed. Future work should involve: 1) advancing in-situ/operando characterization and computation to map trap distributions and carrier pathways, 2) developing hybrid or composite architectures that balance structural integrity with optical functionality, and 3) validating application-driven prototypes in realistic settings from deep-tissue theranostics to structural health monitoring.