Modeling and Optimization of the Compressive Strength of an Alkali-Activated Material Using a Mixed Full Factorial Design

Abstract

The aim of the present work is to carry out a technical study on blast furnace slag-based alkali-activated binders, using the design of experiments methodology. A four-factor, mixed-level (two and three) full factorial design was employed to evaluate and model the significance and interactions of four independent factors on the compressive strength: curing environment, activator type (alkali silicate (AS) and alkali hydroxide (AH)), activator content, and curing time. The experimental compressive strength data were adequately fitted by empirical models with determination coefficients (R2) of 0.89 and 0.93 for AS and AH activators, respectively. The most significant effects of factors on the compressive strength are classed according to this order: curing time > activator-to-precursor mass ratio > curing temperature for the AS case and curing temperature > activating solution concentration > curing time for the AH case. The most favorable situation corresponding to maximum compressive strength is obtained with a minimum curing temperature of 20 °C, an activator-to-precursor mass ratio of 0.42, and a maximum curing duration of 28 days, for a desirability value of 0.95, in the case of the AS activator. When using the AH activator: a desirability function-based optimization targeting a compressive strength above 50 MPa showed that this could be achieved beyond 16 days of curing, with a minimum concentration of the alkaline solution worth 5 M (considering its corrosive nature), and a medium temperature representative of North African countries (≈30 °C), for a desirability value of 0.35.