Answers are grounded in this paper and cite their sources.
Modeling predicts sulfate ion transport in fiber-reinforced concrete, suggesting improved durability through polypropylene fibers.
• Developed a machine learning–enhanced 2D chemo-mechanical transport model to predict sulfate ion transport in fiber-reinforced concrete under drying–wetting cycles. • Two types of fiber (polypropylene and polyethylene terephthalate) with different lengths were exposed to Na₂SO₄ solution under multiple temperatures for 360 days. • Develop an optimized diffusion coefficient equation that incorporates moisture, porosity, temperature, damage degree, and calcium hydroxide consumption. • Demonstrated that fiber reinforcement substantially improves durability, with PET fibers showing superior resistance to sulfate-induced expansion and cracking. • Employed model explainability tools (Taylor diagram, SHAP, and PDP) to clearly identify dominant deterioration mechanisms and quantify feature influence on sulfate diffusion. Rising global temperatures are accelerating the deterioration of concrete structures in coastal and marine environments due to aggressive sulfate attack, highlighting the need for reliable prediction of sulfate-ion transport under complex exposure conditions. This study integrates numerical modeling with machine learning (ML) to enhance the prediction of sulfate diffusion behavior in fiber-reinforced concrete subjected to cyclic drying-wetting. A simplified two-dimensional (2D) chemo-mechanical transport model was developed to analyze sulfate-induced deterioration of concrete prisms. Plain concrete and concrete reinforced with polypropylene and polyethylene terephthalate fibers of different lengths were exposed to Na₂SO₄ solution under drying-wetting cycles at 25, 30, and 35°C for 360 days. Key macroscopic properties, including sulfate penetration depth, porosity, damage degree, and diffusion coefficient, were measured, while microstructural evolution was examined using SEM, XRD, and TGA. A total of 586 experimental data points were used for model development and validation. The base 2D transport model was extended to incorporate moisture variation, temperature effects, fiber length-to-diameter ratio, porosity, damage degree, and calcium hydroxide consumption. Five ML algorithms with optimized hyperparameters were employed to calibrate model coefficients, and interpretability analyses were conducted using SHAP values, Taylor diagrams, and Partial Dependence Plots. Results indicate that damage degree is the dominant factor governing sulfate diffusion, whereas calcium hydroxide consumption has a minor influence. PP and PET fiber reinforcement reduced concrete porosity to approximately 71% and 68% of that of OPC concrete, respectively. The proposed ML-integrated framework accurately predicted sulfate diffusion behavior, achieving a coefficient of determination ( R² ) of up to 98.14%.
No takes yet. Share an insight, caveat, or question.
Shayanfar et al. (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: