Statistical Methods for Modeling the Compressive Strength of Geopolymer Mortar

Area of Science:

• Materials Science
• Civil Engineering
• Sustainable Construction Materials

Background:

• Geopolymer concrete is an eco-friendly alternative to Portland cement (PC) due to PC's high carbon dioxide emissions.
• Fly ash (FA) is a preferred binder for geopolymer concrete owing to its cost-effectiveness, availability, and suitability.
• Accurate prediction of geopolymer mortar compressive strength (CS) is crucial for its widespread application.

Purpose of the Study:

• To develop and evaluate multiscale models for predicting the compressive strength (CS) of fly-ash-based geopolymer mortar.
• To compare the predictive accuracy of linear regression (LR), multinomial logistic regression (MLR), and nonlinear regression (NLR) models.
• To identify the most influential parameters affecting the CS of geopolymer mortar.

Main Methods:

• Compilation of 247 experimental datasets from existing literature on fly-ash-based geopolymer mortar.
• Development of LR, MLR, and NLR models using thirteen input parameters, including mix proportions, curing age, and temperature.
• Statistical evaluation of models using R², RMSE, SI, OBJ, and MAE metrics.

Main Results:

• The nonlinear regression (NLR) model demonstrated superior performance compared to LR and MLR models.
• NLR model achieved a coefficient of determination (R²) of 0.933 and a scatter index (SI) of 0.138.
• Sensitivity analysis indicated that the SiO₂ percentage of fly ash and the alkaline liquid-to-binder ratio are the most critical parameters influencing CS.

Conclusions:

• NLR is a highly effective method for predicting the compressive strength of fly-ash-based geopolymer mortar.
• The developed models provide a reliable tool for forecasting geopolymer mortar CS, aiding in sustainable construction.
• Understanding the impact of SiO₂ content and alkaline liquid ratio is key for optimizing geopolymer mortar performance.