β-CaSiO3 and colloidal n-SiO2 based blended cement composites- their properties, regression analysis and micro-characterization studies

• Main Authors: Nair, N.A (Author), Viswanathan, T.S (Author)
• Format: Article
• Language: English
• Published: CSIC Consejo Superior de Investigaciones Cientificas 2023
• Subjects: Blended cement; Cement composite; Cements; Characterization studies; Compressive strength; Different proportions; Durability; Dynamic modulus of elasticity; E-studio; FESEM; Mechanical properties; Pozzolans; Property; Quality control; Regression analysis; Regression modelling; Silica; Silicate minerals; Sols; Strength modulus; Sulfur compounds; Water absorption
• Online Access: View Fulltext in Publisher

 This paper examines the effect on mechanical properties such as compressive strength, flexural strength, and dynamic modulus of elasticity (DYE) of different proportions of wollastonite (β-CaSiO3) and colloidal nano-silica (n-SiO2) partially replacing cement. Durability indicators (water absorption, sorptivity and sulphate treatment test) were also examined to ascertain the quality of hardened paste mixes with respect to the control mix. The regression models were found for mechanical properties using different parameters from the results obtained, and statistical relations were established and validated. Regression analysis shows the significance of every parameter considered and model for the prediction of mechanical strengths. Finally, the results were substantiated by the microstructural characterization by FESEM. β-CaSiO3 and colloidal n-SiO2 replaced cement by 15%, and 1.5%-6% with an offset of 1.5%, respectively at low (0.25), medium (0.40) and high (0.55) water/binder (w/b) ratio. FESEM micrographs showed dense Calcium-silicate-hydrate (CSH) gel and stratlingite (CASH) was formed by blended cement paste mixes containing β-CaSiO3 and n-SiO2. n-SiO2 at 3% and CaSiO3 at 15% replacements of cement (NS3 mix) was the optimum replacement level for the cement paste mix. Analysed regression model suggests that the models and parameters were found significant and can also be used for prediction (based on R2 values and p-value). ©2023 CSIC.