ANN modeling of mechanical properties in high-volume fly ash concrete: multi-objective cost optimization using NSGA-II for sustainable construction

Abstract

The present study utilizes an Artificial Neural Network (ANN) to develop a comprehensive model for predicting the mechanical properties of High-Volume Fly Ash (HVFA) concrete, encompassing compressive strength, modulus of elasticity, flexural, and splitting tensile strength. Eight concrete mixes with different fly ash content (0–60%) were taken for the present investigation. Experimental data encompassing compressive strength, modulus of elasticity, flexural strength, and splitting tensile strength, recorded over 180 days, were utilized to train the ANN model. The dataset underwent thorough analysis to yield a statistically robust model capable of estimating diverse mechanical properties of concrete containing any proportion of fly ash at any given concrete age. The strengths projected by the unified ANN model were compared with actual values, revealing remarkable proximity between the anticipated and experimental compressive strength values. Thus, the ANN-based model presents a dependable approach for assessing the strength of fly ash concrete. Moreover, a comprehensive multi-objective cost optimization model was developed using NSGA-II to ascertain the most cost-effective and economical mix of HVFA concrete with maximum compressive strength. This research offers valuable insights for designers and practicing engineers involved in sustainable construction endeavors, particularly when considering the incorporation of fly ash within concrete structures.