Stepwise regression modeling for compressive strength of alkali-activated concrete

Abstract

This paper presents the results from a parametric experimental investigation of the compressive strength of alkali-activated concrete. The effects of curing condition (moist-cured versus heat-cured), sodium oxide dosage, silica dosage, silica modulus (relative dosage of silica to sodium), solution/binder ratio, and free water/binder ratio on the compressive strength of sodium silicate-activated fly ash and slag cement concrete are evaluated. More than 5000 specimens with 676 unique combinations of mixture proportion and curing condition were tested. The marginalized effects of each parameter indicate effects similar to those identified by previous studies. Predictive models for the compressive strength of fly ash and slag cement-based concretes are developed by stepwise regression analysis. Although the specific models presented in this paper are applicable only for the materials and activators identified herein, the modeling procedures are generalized results are of doubtless utility for the design of alkali-activated concrete mixtures.