This paper describes the hydration products and microstructural formation processes that yield excellent mechanical properties in “zero-cement concretes” (ZCC) produced by chemical activation of a blend of silica fume, blast furnace slag, and fly ash, using a CaO-rich additive (commercially supplied as an expansive agent but taking a chemical activation role here), a high superplasticizer dose, and a very low water content. These concretes reach 70 MPa at 28 days and then continue to gain strength beyond 150 MPa after 5 years, either under sealed conditions or exposed on a rooftop in the climate of Tokyo, Japan. The reaction products of ZCC are dominated by C-A-S-H gel, accompanied by aluminate hydrates of different layered double hydroxide forms; this unconventional cementitious blend yields reaction products that are familiar from Portland cement and blended binder systems. The ferronickel slag used as fine aggregate in these mixes makes an important contribution to the balance of fresh-state and hardened-state properties by modifying hydration chemistry.
The elastic stiffness of bulk concrete materials results from the complex interaction of aggregates, voids, and hydrated cement (which can have multiple hardened phases at multiple length scales). Given the complexities associated with understanding the arrangement of these particles within bulk concrete volumes, estimations for elastic modulus often rely on empirical correlations with unit weight and compressive strength. Such estimations are inherently scale-dependent and fail to capture variability in mix designs, particularly the variability found in specialty concrete mixes. To develop a scale-independent method for estimating elastic modulus from mix-design volume fraction information, this study explores a novel bottom-up approach using cement paste phase stiffness values determined through micro-mechanical experimentation and randomized Monte-Carlo spring arrangement simulations. Statistical representations of cement paste phase stiffness distributions and bulk volume fraction data are combined to provide estimations for elastic stiffness in both the composite cement paste and bulk concrete containing fine aggregate and fibers. Resulting a priori estimations of UHPC cement paste stiffness from the micro-mechanical upscaling simulations were within 4% of measured values (based on mix-design and void volume fraction information alone) for a selected sample of mix proportions. When applied to the two UHPC mixes containing fibers and fine aggregate, upscaling simulations consistently overpredicted the measured elastic modulus, likely due to the aggregate-cement interfacial transition zone (ITZ) properties that were not captured in the micro-mechanical testing.