Microscopic pore characteristics governing macroscopic freeze thaw durability and service life prediction of ternary blended concrete
- 1School of Transportation and Logistics Engineering, Xinjiang Agricultural University
Abstract
To optimize the mix design of frost-resistant concrete for cold-region infrastructures and establish a probabilistic reliability evaluation method for its long-term durability, this study takes the ternary composite system consisting of fly ash, ground granulated blast-furnace slag (GGBS) and defoamer as the research subject. A total of 360 rapid freeze–thaw cycle tests were performed on five batches of concrete labelled H0-H4 to systematically investigate the evolution laws of mass loss rate, relative dynamic elastic modulus, compressive strength degradation and micro-pore structure throughout the freeze–thaw process. Multi-scale characterization techniques including X-ray diffraction (XRD) and scanning electron microscopy (SEM) were utilized to quantitatively analyze the influence mechanisms of micro parameters such as porosity, pore size distribution and air void spacing factor on the macroscopic frost resistance of concrete. By integrating backpropagation (BP) neural network, XGBoost machine learning model, Wiener stochastic process and Weibull distribution theory, an integrated framework coupling freeze–thaw damage evolution and service life prediction of concrete was established, which quantitatively revealed the individual and coupled regulatory effects of each component on concrete frost durability. Test results demonstrate that the ternary composite mixture H0 delivers the optimal frost durability. After 360 freeze–thaw cycles, all its macroscopic performance indicators outperform other batches significantly, owing to its dense hydration gel network, abundant closed micro-fine air voids and uniformly distributed pore structure. Model comparison reveals that the BP neural network achieves higher accuracy in single-point performance prediction, yet it cannot predict long-term damage evolution trends or conduct probabilistic service life assessment. The two-stage Wiener–Weibull framework proposed in this work effectively addresses this limitation and identifies defoamer as the core regulator mitigating freeze–thaw damage propagation. The prediction results show that the average freeze–thaw service life of mixture H0 reaches 812 cycles, 21.0% longer than that of mixture H4 with the worst anti-freezing performance. This research provides theoretical support and analytical references for the mix proportion optimization and long-term durability assessment of frost-resistant concrete applied in cold regions with large temperature differences.
Related
- Microscopic characteristics of adult dyshormonogenetic goiter: attempt to histologically estimate subtypes
- Damage evolution and microscopic failure characteristics of dense sandstone at different water contents
- Differential Enteric Glial and Mast Cell Characteristics Distinguish Microscopic Colitis Subtypes Relative to Healthy and Disease Controls, with Treatment-Associated Reductions in Glial Reactivity and Mast Cell Degranulation in Collagenous Colitis.
- Prediction of Spinal Pain Intensity from Perception-Dependent Stimulation Thresholds and Patient Characteristics: A Regression and Machine Learning Approach
- Can maternal characteristics improve due date prediction without ultrasound? Development and internal validation of a machine-learning prediction model in 600 000 pregnancies in Sweden
BibTeX
@article{zhong2026microscopic,
title = {{Microscopic pore characteristics governing macroscopic freeze thaw durability and service life prediction of ternary blended concrete}},
author = {Xiaoling Zhong and Shiqiang Yin and Xueshuai Liu and Chen Zhang and Shibo Dong},
journal = {Scientific Reports},
year = {2026},
month = jul,
doi = {10.1038/s41598-026-61504-8},
url = {https://doi.org/10.1038/s41598-026-61504-8},
note = {CC BY},
}
RIS
TY - JOUR
TI - Microscopic pore characteristics governing macroscopic freeze thaw durability and service life prediction of ternary blended concrete
AU - Xiaoling Zhong
AU - Shiqiang Yin
AU - Xueshuai Liu
AU - Chen Zhang
AU - Shibo Dong
PY - 2026
DA - 2026/07/01
JO - Scientific Reports
DO - 10.1038/s41598-026-61504-8
UR - https://doi.org/10.1038/s41598-026-61504-8
AB - To optimize the mix design of frost-resistant concrete for cold-region infrastructures and establish a probabilistic reliability evaluation method for its long-term durability, this study takes the ternary composite system consisting of fly ash, ground granulated blast-furnace slag (GGBS) and defoamer as the research subject. A total of 360 rapid freeze–thaw cycle tests were performed on five batches of concrete labelled H0-H4 to systematically investigate the evolution laws of mass loss rate, relative dynamic elastic modulus, compressive strength degradation and micro-pore structure throughout the freeze–thaw process. Multi-scale characterization techniques including X-ray diffraction (XRD) and scanning electron microscopy (SEM) were utilized to quantitatively analyze the influence mechanisms of micro parameters such as porosity, pore size distribution and air void spacing factor on the macroscopic frost resistance of concrete. By integrating backpropagation (BP) neural network, XGBoost machine learning model, Wiener stochastic process and Weibull distribution theory, an integrated framework coupling freeze–thaw damage evolution and service life prediction of concrete was established, which quantitatively revealed the individual and coupled regulatory effects of each component on concrete frost durability. Test results demonstrate that the ternary composite mixture H0 delivers the optimal frost durability. After 360 freeze–thaw cycles, all its macroscopic performance indicators outperform other batches significantly, owing to its dense hydration gel network, abundant closed micro-fine air voids and uniformly distributed pore structure. Model comparison reveals that the BP neural network achieves higher accuracy in single-point performance prediction, yet it cannot predict long-term damage evolution trends or conduct probabilistic service life assessment. The two-stage Wiener–Weibull framework proposed in this work effectively addresses this limitation and identifies defoamer as the core regulator mitigating freeze–thaw damage propagation. The prediction results show that the average freeze–thaw service life of mixture H0 reaches 812 cycles, 21.0% longer than that of mixture H4 with the worst anti-freezing performance. This research provides theoretical support and analytical references for the mix proportion optimization and long-term durability assessment of frost-resistant concrete applied in cold regions with large temperature differences.
SN - 2045-2322
ER -