Journal Article

Optimal concrete mixture proportions for environmental, economical, and mechanical objectives in concretes exposed to high temperature and re-curing regime based on NSGA-II and MOPSO

Toufigh, Vahab, Kharrazi, Hamed

2025

Overview

Nowadays, developing economical and environmentally friendly concretes along with considering mechanical and durability objectives is of great significance. This study developed optimal concrete mixture proportions for different heating and re-curing regimes. The objective functions were 28-day, residual, and re-cured compressive strengths, cost, and embodied CO 2 emissions. A data set, including approximately 500 mix designs, was collected from the literature to develop an optimal ANN model to predict the first three objectives. The genetic algorithm and particle swarm optimization were used to tune the ANN model’s hyperparameters. NSGA-II and MOPSO algorithms, besides this optimal ANN model, were then used to develop the optimal mixture proportions for different heating and re-curing regimes. TOPSIS was utilized to sort the non-dominated solutions and select the optimal one based on different weight vectors. The results showed that the R 2 values for the verification set in predicting the 28-day, residual, and re-cured compressive strengths were 0.999, 0.987, and 0.977, respectively. The sensitivity analysis revealed that the heating regime (especially peak temperature) is the most fundamental input feature in predicting the post-fire behavior of concretes. The optimal concrete mixture proportion depends mainly on the case study, which considers economic, environmental, and post-fire mechanical objectives. Most of the mixtures were high-strength concretes when all objectives were equally weighted. Increasing the cost and embodied CO 2 emissions weights resulted in cost-efficient and environmentally friendly concretes in exchange for inferior mechanical properties. The developed mixtures by NSGA-II were mostly blended with metakaolin when all objectives equally weighted. Silica fume appeared as SCM in mixtures when the cost and embodied CO 2 emissions weighted five times that of the mechanical objectives. The proposed framework facilitates the appropriate selection of the mixture proportion.

• Publisher: Springer Berlin Heidelberg,Springer Nature B.V
• Subject: Algorithms / Aquatic Pollution / Carbon Dioxide / Carbon dioxide emissions / case studies / Concrete mixes / Construction Materials / cost effectiveness / Curing / data collection / Datasets / Design optimization / durability / Earth and Environmental Science / Ecotoxicology / Emissions / Environment / Environmental Chemistry / Environmental Health / Genetic algorithms / heat / Heating / High strength concretes / High temperature / Hot Temperature / Machine learning / Mechanical properties / Metakaolin / Mixtures / Neural networks / Objectives / Optimization techniques / Particle swarm optimization / Research Article / Sensitivity analysis / Silica / Silica fume / temperature / Waste Water Technology / Water Management / Water Pollution Control