Journal of Engineering Research

As a novel green cementitious material, the setting time of geopolymer grouting materials directly impacts construction safety and efficiency. Traditional empirical methods struggle to achieve accurate predictions due to complex factors including mix proportions, temperature, humidity, and active mineral characteristics. To address this challenge, this study proposes an intelligent modeling-based setting time prediction method. Through experimental data collection and machine learning algorithms for feature extraction and correlation analysis of key influencing factors, a setting time prediction model was developed and validated through multiple experiments. Results demonstrate that this method exhibits high prediction accuracy and strong adaptability, providing effective references for grouting process parameter design and construction control. This enhances the application reliability and engineering controllability of geopolymer grouting materials.

Geopolymer grouting materials; Setting time; Intelligent prediction; Machine learning