Prediction method of volumetric stability and cracking resistance of concrete coupled with moisture and heat based on maturity theory and engineering application

• A coupled hygrothermal model with variable diffusion and latent heat was established. • Temperature prediction accuracy raised by 5%–52% vs. traditional models. • Strain prediction accuracy improved by up to 31%, field errors ≤±10%. • Casting temperature significantly affects thermal strain and early-age cracking. With the growing demand for large-scale infrastructure development in China—such as deep-sea, deep-underground, and urban subsurface projects—combined with the widespread use of general-purpose raw materials, there is an urgent need for more precise crack control technologies in concrete. This need stems from the imperative to reduce unnecessary material consumption and environmental impact caused by excessive safety margins. To address this, a set of governing equations that account for the mutual feedback between temperature and humidity was first proposed. A non-constant form of the diffusion coefficient was introduced, alongside latent heat terms and unsteady-state heat source terms, to establish a hygrothermal coupling model. This model was further enhanced by incorporating the effects of creep relaxation, reinforcement constraint, structural restraint, and thermal conduction characteristics of formwork, thereby forming a comprehensive multi-field coupling evaluation framework that encompasses the temperature field, moisture content field, strain field, and cracking index field. Subsequently, the proposed theoretical framework was applied to representative engineering scenarios, including large-scale concrete foundation slabs, bridge bearing platforms, large-area long-span side walls and prefabricated tunnel segments. The accuracy and reliability of the model were validated through comparisons between simulation results and field-monitored data. The results demonstrate that this method effectively overcomes the technical limitations of traditional concrete crack prediction models, particularly those relying on constant parameter assumptions and decoupled field interactions. It offers a practical and robust approach for engineering applications, providing a novel perspective for precision crack control in concrete and contributing to the broader goals of sustainability and resource efficiency