Abstract

Geopolymer concrete is an environment-friendly material and is presently accepted as an alternative to conventional concrete. It utilizes industrial by-products like fly ash and slag to reduce CO2 emissions associated with cement production. Due to the variability of fly ash composition and concrete casting conditions, significant variability is observed in the performance of the concrete produced with such materials. In this study, a non-linear regression technique is used to develop predictive models for the compressive strength of high and low calcium fly ash geopolymer concrete. The criteria such as root mean squared error (RMSE), the statistical significance of model coefficients, and the R-squared values were adopted. The developed models were validated using plots comparing the experimental and predicted response and plots showing residuals. The dependency of the compressive strength on the significant parameters were highlighted using contour and adjusted response plots. From the contour plots, it was observed that, for geopolymer concrete sourced from the high calcium fly ash, compressive strength increases with the molarity of sodium hydroxide, the activator-to-binder ratio, and the curing temperature; while it decreases with the increase in coarse aggregate content. Similar contour plots, for the case of geopolymer concrete sourced from low calcium fly ash, exhibited improvement in compressive strength increases with the molarity of sodium hydroxide, the activator-to-binder ratio, and the ratio of Na2SiO3 –NaOH solutions; at the same time, it decreases with the increase in water-to-geopolymer solids ratio. For low calcium fly ash geopolymer concrete, the proposed model predicts that compressive strength of 90 MPa could be obtained for fly ash with oxide composition comprising a silica-to-alumina ratio of 4. The regression models developed in the present study will be helpful to develop a design mix for geopolymer concrete comprising either high calcium fly ash or low calcium fly ash.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call