Calculation of temperature stress in concrete structures based on finite elements

This paper addresses the issue of temperature stress in large-volume concrete during the pouring process caused by the heat of hydration, proposing a computational model based on the finite element method. By analyzing the experimental and empirical data on the material properties of concrete, a cubic concrete model was developed using ANSYS software to investigate the distribution of temperature and stress fields, revealing that peak temperature stress is concentrated at the top of the third layer. Furthermore, a database correlating concrete mix ratios and peak temperature stress was constructed, and ensemble learning methods were employed to train and test predictive models, identifying XGBoost as the optimal model. Additionally, the SHAP method was applied to interpret the model, revealing the influence mechanisms of factors such as cement and blast furnace slag on temperature stress, and proposing mix ratio adjustments based on SHAP values to reduce temperature stress. The findings provide theoretical and practical guidance for temperature control design and structural optimization of large-volume concrete.