Traditional silicate cement materials produce a large amount of CO₂ during production, making it urgent to seek alternatives. Alkali-activated slag cement is a good substitute, as its production process has low carbon emissions and energy consumption, and it can comprehensively utilize various types of industrial waste residue while possessing superior physical and chemical properties. However, the shrinkage of alkali-activated concrete can be larger than that of traditional silicate concrete. To address this issue, the present study utilized slag powder as the raw material, sodium silicate (water glass) as the alkaline activator, and incorporated fly ash and fine sand to study the dry shrinkage and autogenous shrinkage values of alkali cementitious material under different content. Furthermore, combined with the change trend of pore structure, the impact of their content on the drying shrinkage and autogenous shrinkage of alkali-activated slag cement was discussed. Based on the author’s previous research, it was found that by sacrificing a certain mechanical strength, adding fly ash and fine sand can effectively reduce the drying shrinkage and autogenous shrinkage values of alkali-activated slag cement. The higher the content, the greater the strength loss of the material and the lower the shrinkage value. When the fly ash content was 60%, the drying shrinkage and autogenous shrinkage of the alkali-activated slag cement mortar specimens decreased by about 30% and 24%, respectively. When the fine sand content was 40%, the drying shrinkage and autogenous shrinkage of the alkali-activated slag cement mortar specimens decreased by about 14% and 4%, respectively.

Modifying the admixture of alkali-activated cementitious materials using components such as fly ash and fine sand may reduce CO₂ emissions and conserve natural resources and energy. This study adopted strength testing, scanning electron microscopy, and mercury intrusion porosimetry to investigate the influence of different admixtures on the compressive strength and flexural strength of alkali slag cementing materials and the microstructure characteristics of hardened slurry under the action of load. The flexural strength of alkali slag cement slurry and mortar was reduced by replacing slag powder with fly ash. Content of fine sand less than 20% had little effect on the strength of alkali slag cement mortar; however, when the content of fine sand exceeded 30%, the strength decreased significantly. The hydration degree at 3 d was large, and the density of slurry increased with the extension of age. Increased fly ash or fine sand content decreased the density of the slurry, and increased fly ash resulted in a large number of unhydrated fly ash particles in the cementitious materials. Addition of fine sand resulted in a large number of microcracks in the slurry, which gradually decreased with the extension of hydration age.

The moisture diffusion behaviors of 3D woven composites exhibit non-Fickian properties when they are exposed to a hydrothermal environment. Although some experimental works have been undertaken to investigate this phenomenon, very few mathematical works on non-Fickian moisture diffusion predictions of 3D woven composites are available in the literature. To capture the non-Fickian behavior of moisture diffusion in 3D woven composites, this study first utilized a time fractional diffusion equation to derive the percentage of moisture content of a homogeneous material under hydrothermal conditions. A two-stage moisture diffusion model was subsequently developed based on the moisture diffusion mechanics of both neat resin and 3D woven composites, which describes the initial fast diffusion and the long-term slow diffusion stages. Notably, the model incorporated fractional order parameters to account for the nonlinear property of moisture diffusion in composites. Finally, the weight gain curves of neat resin and the 3D woven composite were calculated to verify the fractional diffusion model, and the predicted moisture uptake curves were all in good agreement with the experimental results. It is important to note that when the fractional order parameter α < 1, the initial moisture uptake will become larger with a later slow down process. This phenomenon can better describe non-Fickian behavior caused by initial voids or complicated structures.

With the continuous promotion of urbanization, the generation of construction and demolition waste (CDW) is increasing. The environmental problems and safety hazards caused as a result need to be resolved. In this paper, based on the system dynamics (SD) theory, the modeling, the cost, and the environmental benefit of CDW resource management under the life cycle assessment (LCA) are proposed. Specifically, we propose a combined policy derived through three variables, namely, fines, subsidies, and charges. The target is to reduce illegal dumping behavior and landfill volume and to improve the recycling volume and environmental benefits. The model is constructed with the help of the software VENSIM, and the validity and feasibility of the model are demonstrated with data from Nantong City. The results show that a single policy cannot simultaneously improve environmental benefits, illegal dumping, recycling, and landfill behavior. A combined policy combines the advantages of three single policies, fines, subsidies, and charges, which not only can effectively curb illegal dumping and landfill disposal under the premise of prioritizing environmental benefits, but can also promote the recycling of CDW. The reasonable range for the fine is 300–350 CNY/ton; the rational range for subsidies is 30–40 CNY/ton; and the flexible range for treatment charge is 40–80 CNY/ton. The model can be used for the quantitative assessment of urban CDW management costs and environmental benefits and can also provide a theoretical basis for the government.

This paper investigates the potential of deep learning in predicting the compressive strength of ultra-high-performance concrete (UHPC) by developing a convolutional neural network (CNN) model with two convolutional layers. The proposed CNN architecture demonstrates capability in accurately predicting UHPC compressive strength from tabular data encompassing various material compositions. Ten input variables were selected, including the cement content, water content, silica fume content, silica powder content, silica sand content, superplasticizer content, and curing parameters. The model was trained and tested on a dataset comprising 219 samples. Experimental results indicated excellent predictive performance, with the CNN achieving a coefficient of determination (R²) of 0.959 on the test set and a mean absolute percentage error (MAPE) of 5.55%, demonstrating both accuracy and stability. Comparative analysis revealed that the CNN’s performance was comparable to established machine learning methods like XGBoost (R² = 0.961), which are typically more suited for tabular data. Furthermore, SHAP (SHapley Additive exPlanations) analysis confirmed the model’s interpretability. These findings collectively suggest that the CNN-based approach shows considerable promise for predicting compressive strength across diverse UHPC formulations.