Data-driven modeling and optimization of nano-belite cementitious composites incorporating internal NaCl and Na2SO4 additives
- 1Department of Civil Engineering, University of Sistan and Baluchestan
Abstract
This study evaluates the interactive effects of nano-belite cement replacement ( 0-−20% wt. ) and internal salt additions ( 5% wt. of either NaCl or Na2SO4 ) on the compressive strength and bulk electrical resistivity of concrete composites up to 180 days of curing. A dataset of 225 experimental records was used to train a multilayer perceptron (MLP) neural network, which was subsequently coupled with the non-dominated sorting genetic algorithm II (NSGA-II) to optimize the mixture design. Experimental results indicate that under salt-free conditions, a 10% nano-belite replacement optimizes microstructural development, yielding a compressive strength of 50.26 MPa and an electrical resistivity of 174.40Ω · m at 180 days. The internal addition of Na2SO4 acts synergistically with nano-belite by promoting pore refinement through secondary hydration products, elevating the 180-day compressive strength to 52.49 MPa . In contrast, internal NaCl incorporation exhibits an antagonistic relationship at nano-belite dosages exceeding 5% , causing a drop in electrical resistivity due to enhanced ionic mobility in the pore network. The developed MLP model predicted the experimental outcomes with high accuracy ( R2>0.98 for both properties). Multi-objective optimization via NSGA-II identified a balanced optimum replacement level at 9.82% nano-belite. Independent validation testing of ten physical specimens fabricated at this optimal dosage demonstrated close agreement with model predictions, yielding relative errors of 2.94% for bulk electrical resistivity and 13.13% for compressive strength. Additionally, explicit algebraic equations derived from the network weights provide a direct mapping tool for mix design without the need for active machine learning execution.
Related
- Mechanical performance of basalt fiber reinforced cementitious mortars incorporating calcined red mud
- Pathomic Texture Composites from H&E, and Immunohistochemistry of Prostatectomy Specimens for Exploratory Modeling of Time to Biochemical Recurrence
- Population pharmacokinetic modeling of voriconazole based on real-world data for dosing regimen optimization
- A whole-brain modeling framework for tDCS montage optimization in drug-resistant epilepsy
- Water and salt migration model of solidified sulfate-contaminated soil by full solid waste cementitious materials
BibTeX
@article{jamshidzehi2026datadriven,
title = {{Data-driven modeling and optimization of nano-belite cementitious composites incorporating internal NaCl and Na2SO4 additives}},
author = {Habib Alrahman Jamshidzehi and Mohammad Reza Sohrabi and Ahmad Reza Abbasian},
journal = {Scientific Reports},
year = {2026},
month = sep,
doi = {10.1038/s41598-026-67355-7},
url = {https://doi.org/10.1038/s41598-026-67355-7},
note = {CC BY},
}
RIS
TY - JOUR
TI - Data-driven modeling and optimization of nano-belite cementitious composites incorporating internal NaCl and Na2SO4 additives
AU - Habib Alrahman Jamshidzehi
AU - Mohammad Reza Sohrabi
AU - Ahmad Reza Abbasian
PY - 2026
DA - 2026/09/01
JO - Scientific Reports
DO - 10.1038/s41598-026-67355-7
UR - https://doi.org/10.1038/s41598-026-67355-7
AB - This study evaluates the interactive effects of nano-belite cement replacement ( 0-−20% wt. ) and internal salt additions ( 5% wt. of either NaCl or Na_2SO_4 ) on the compressive strength and bulk electrical resistivity of concrete composites up to 180 days of curing. A dataset of 225 experimental records was used to train a multilayer perceptron (MLP) neural network, which was subsequently coupled with the non-dominated sorting genetic algorithm II (NSGA-II) to optimize the mixture design. Experimental results indicate that under salt-free conditions, a 10% nano-belite replacement optimizes microstructural development, yielding a compressive strength of 50.26 MPa and an electrical resistivity of 174.40Ω · m at 180 days. The internal addition of Na_2SO_4 acts synergistically with nano-belite by promoting pore refinement through secondary hydration products, elevating the 180-day compressive strength to 52.49 MPa . In contrast, internal NaCl incorporation exhibits an antagonistic relationship at nano-belite dosages exceeding 5% , causing a drop in electrical resistivity due to enhanced ionic mobility in the pore network. The developed MLP model predicted the experimental outcomes with high accuracy ( R^2>0.98 for both properties). Multi-objective optimization via NSGA-II identified a balanced optimum replacement level at 9.82% nano-belite. Independent validation testing of ten physical specimens fabricated at this optimal dosage demonstrated close agreement with model predictions, yielding relative errors of 2.94% for bulk electrical resistivity and 13.13% for compressive strength. Additionally, explicit algebraic equations derived from the network weights provide a direct mapping tool for mix design without the need for active machine learning execution.
SN - 2045-2322
ER -