Optimization of mixtures of soil, construction and demolition waste, and steel slag using the simplex-extreme vertices method

Abstract

In the present study, a simplex-extreme vertices methodology was applied to investigate the optimal proportions of recycled aggregates from construction and demolition waste (CDWr) and fines of electric arc furnace oxidizing slag (FS) for the stabilization of two distinct tropical soils, enabling their use in structural pavement layers. The influence of mixture components content on the unconfined compressive strength (UCS) and California Bearing Ratio (CBR) was investigated using regression analyses and response surfaces. UCS and CBR were optimized, based on the application of the Desirability function approach for determination of the best mixture compositions. The regression models yielded high coefficients of determination (R² ≥ 0.83). The optimization methodology demonstrated that the best dry mass composition was 80 % soil, 0 % CDWr, and 20 % FS for the sandy soil; and 18 % soil, 62 % CDWr, and 20 % FS for the clayey soil. X-ray diffraction (XRD) and scanning electron microscopy (SEM) analyses revealed the formation of typical products of pozzolanic reactions after a 28-day curing period. Incorporating wastes and extending the curing period caused a decrease in expansion and an enhancement in the UCS and CBR values of the soils.