Rheological, mechanical and durability evaluation of alkali-activated pastes and concretes designed based on fly ash and steel slag

Abstract

The development of societies presents a high impact on the environment, which includes the consumption of natural resources, environmental pollution with greenhouse gases, and a high generation of solid waste. Portland cement production is one of the main causes of such impacts, and the search for more sustainable materials to replace it is a challenge to overcome. In this scenario, there are the Alkali-Activated Binders (AAB), frequently called "geopolymers" in the literature related to construction. This is a new generation of non-hydraulic binders, activated by the reaction of precursors, composed of aluminosilicate sources, associated with alkaline activators. It can represent a reduction in CO2 emissions between 55% and 75% compared to Portland cement concrete. Thus, this MSc work focuses on the evaluation of alkali-activated pastes and concretes produced based on fly ash (FA) and steel slag (SS) precursors with alkaline activators composed of different proportions of sodium hydroxide and sodium silicate. The experimental approach was distributed in three stages. The first stage focused on the development the design procedure to formulate and mechanically evaluate AAB pastes and concrete. In this stage, 36 binder’s formulations were evaluated by compressive strength, aiming to achieve 60 MPa under thermal curing. They were composed by silica modulus (S/N) of the activator of 0.75, 1.25 and 1.75, alkali dosage (N%) of 6%, 8% and 10%, proportions of FA to basic oxygen furnace (BOF) SS from 50% to 50% and 75% to 25%, and ambient and thermal curing (65 oC). At this stage, it was found that lower silica modulus and lower alkali dosage were beneficial to the increase of compressive strength. An AAB paste based on N% of 10%, S/N of 0.75 and 75% to 25% of FA to SS proportion presented a suitable combination to produce concrete with high slump flow (575 mm) and compressive strength (64 MPa). The second stage focused on the multipoint rheological characterization of pastes and concretes. The tests were performed using flow tests and oscillatory tests in wither a bench dynamic shear rheometer for pastes or using a planetary rheometer for concretes. The pastes presented a plastic viscosity of 1.63 Pa.s to 10.1 Pa.s and no yield stress and the AAB concrete exhibited 3 times higher mixing energy than Portland cement concrete. Both paste and concrete had rheological behaviors that were well-adjusted by pseudoplastic models. The third stage focused on assessing the impact of replacing conventional aggregates by Baosteel's Slag Short Flow (BSSF) SS aggregates on the fresh and hardened state properties, besides the characterization of the concretes regarding durability. The concretes presented good passing ability, but slow flow related to the high viscosity of the activator. In the hardened state, the concretes with conventional aggregate showed better mechanical performance, with compressive strength of 64 MPa and Young's modulus up to 26.2 GPa, under thermal curing. The utilization of steel slag aggregates produced a decrease in compressive strength, to values of 49.6 MPa under thermal curing, but an improvement in self-compacting properties, reaching slump flow values of 737 mm. Compared to conventional concrete, AAB concrete presented controversial indications of durability, with low water penetration under pressure but high indications of chloride ion penetration, based on electrical measurements. In other words, conflicting conclusions depend on the chosen tests, which suggests that further research is necessary for developing proper methods for durability analysis. At the end of the study, knowledge about AAB pastes and concretes was produced, as well as an efficient mix design methodology.