Deep neural network-enhanced prediction and carbon footprint analysis of early-age high-performance manufactured sand concrete’s stress–strain behavior

To meet the demands for rapid and environmentally sustainable construction in bridge building, the study of the mechanical properties of manufactured sand concrete (MSC) at early ages is of significant importance. This study aims to investigate the stress-strain curve relationship and construction efficiency of manufactured sand concrete (MSC) at early ages through uniaxial compression tests conducted on 216 specimens at ages of 2 d, 3 d, 4 d, 5 d, 6 d, 7 d, 14 d, and 28 d. The study examined the effects of age, water-cement ratio, and fly ash content on the peak stress and peak strain of MSC. The results indicate that the influence of fly ash content on early-age peak stress is more significant than that of the water-cement ratio. Additionally, the water-cement ratio and fly ash content significantly affect the peak strain of MSC at early ages, particularly within the 2 to 7-day age range. This study evaluates the carbon emissions of MSC as a sustainable building material over its entire lifecycle and concludes that the replacement of supplementary cementitious material (SCM) for cement is essential for emission reduction. Furthermore, a deep neural network (DNN) model with four hidden layers and 100 neurons in each layer was developed based on experimental results. The model was trained to predict the stress-strain curves of MSC under varying water cement ratios, ages, and fly ash content. DNN model was trained and validated through pre-processing and segmenting the original dataset using Pytorch deep learning (DL) libraries. DNN model accurately predicts stress-strain curves that closely align with MSC test curves.