Abstract
A practical mix design and performance metrics for fly ash-phosphogypsum blended geopolymer paste and mortar is lacking. This study focused on investigating the physico-chemo-mechanical performance and mix design of ambient-cured fly ash-phosphogypsum geopolymer paste (FPGP) and mortar (FPGM). The geopolymer was formed by reacting binary powder precursor (fly ash+phosphogypsum) and alkaline activator (NaOH+Na2SiO3). The fly ash was substituted with varying weight proportions of phosphogypsum (i.e., 10 %, 20 %, 30 %, 40 %, 50 %). Experimental testing and analysis were done based on ASTM standards, XRF, XRD, SEM-EDS, ANOVA, and Python. The workability, setting time, flexural strength, compressive strength, morphology, and mineralogy were studied. The multivariate regression approach was used to model the relationship between compressive strength and alkaline liquid/precursor. The optimum mixture conditions for FPGP and FPGM were 30 wt% phosphogypsum, alkaline liquid/precursor of 0.4, 10 M NaOH, Na2SiO3/NaOH of 1.5, and binder/aggregate of 1.0. The results showed that the workability of FPGP ranged from 185 mm to 112 mm while that of FPGM ranged from 145 mm to 100 mm. The FPGP and FPGM had considerably lower initial and final setting times of (18 - 37 min, 81 - 155 min) and (14 - 29 min, 67 - 142 min), respectively. The compressive strength of FPGP ranged from 7.3 MPa to 27.24 MPa while that of FPGM ranged from 9.5 MPa to 43.27 MPa. There was a strong connection between compressive and flexural strength giving R-2 values > 0.8. Based on the adjusted R-2 values and ANOVA, the proposed mix design models were statistically significant since the p-value (0.000) was less than the 0.05 significance level and the value of F (197.953) was greater than the critical value of 1. At high curing duration, the concentration of NaOH and alkaline liquid/precursor facilitated the dissolution of Ca2+ in phosphogypsum and Si(4+)and Al3+ in fly ash forming a C-(N)-A-S-H matrix crucial to the strength development of geopolymers. The statistical metrics implied a good performance accuracy and reliability of the proposed model equations and can find practical applications in construction materials design. For optimal results, it is recommended to use the prediction models within the specific input parameters employed in this study.