Assessing the Risk of Thermal Cracking in Hardening Concrete – a Comprehensive Parametrical Study

Paper i proceeding, 2025

Modelling the temperature evolution and the subsequent stress development in hardening concrete is a complex problem that involves numerous parameters. In turn, determining those parameters often requires extensive laboratory testing which must be repeated when new concrete recipes are developed. Simplifying testing procedures by focusing only on the most relevant parameters could accelerate the use of concretes with novel supplementary cementitious materials and a lower carbon footprint. This work investigated the key parameters influencing early-age thermal cracking through a comprehensive parametric analysis. The study focused on modelling the early-age behaviour of a restrained wall including 20 different parameters describing the rate of heat evolved, mechanical and thermal properties as well as creep and shrinkage among others. In addition to one-factor-at-a-time (OFaT) analysis, Monte Carlo simulations were conducted to assess the influence of statistical variations in the input parameters on the risk of cracking. Moreover, different methods were used to evaluate the Monte Carlo simulation results including correlation analysis, linear regression analysis and Random Forest feature importance. Results showed that the cement content and final hydration heat had the largest impact in absolute terms, but the tensile strength and modulus of elasticity were most relevant when the variability of the parameters was accounted for. The results are in line with earlier studies and highlight the importance of accurate input data for mechanical properties and the need for standardized testing methods to predict cracking behaviour reliably.