Science and Education Publishing American Journal of Mining and Metallurgy 2 3 2014 09 17 Empirical Evaluation of Slag Cement Minimum Setting Time (SCMST) by Optimization of Gypsum Addition to Foundry Slag during Production 51 56 EN C. I. Nwoye Department of Metallurgical and Materials Engineering, Nnamdi Azikiwe University, Awka, Nigeria I. Obuekwe C. N. Mbah S. E. Ede C. C. Nwangwu D. D. Abubakar AJMM2014233 10.12691/ajmm-2-3-3 2014 08 16 2014 09 13 2014 09 17 An empirical evaluation of slag cement minimum setting time (SCMST) has been successfully carried out through optimization of gypsum addition to foundry slag in the course of the cement production. A model was derived and used as a tool for predictive analysis of the cement setting time based on gypsum input. The model aided optimization of gypsum addition indicates a minimum setting time of 14.1054 minutes at an optimum gypsum input concentration of 6.4847%. Beyond 6.4847% gypsum addition, the slag cement setting time increases drastically; a situation typifying immiscibility and lack of homogeneity between the cement slurry and the extra gypsum addition. This is because increased gypsum addition (above a specified quantity) is unlikely to forms a coherent mass with a specified and fixed liquid volume, resulting to delayed and differential setting. The derived model expressed as; is quadratic and single factorial in nature. The slag cement setting time per unit gypsum addition are as obtained from experiment and derived model are 4.75 and 4.996 mins./% respectively. Statistical analysis of the experimental and derived model-predicted results for each value of the gypsum input concentration considered shows standard errors of 1.8974 and 1.5485% respectively. Deviational analysis indicates 10.46% as the maximum deviation of the model-predicted slag cement setting time from the corresponding experimental value. The validity of the model was rooted on the expression 0.0226ɤ + 0.2101 α = 0.0162 α2 + 1.0001 where both sides of the expression are correspondingly approximately equal. The validity of the derived model-predicted results also was ascertained using SPSS 17.0. The results indicate check variance = 0.001, standard deviation = 0, and model operational confidence of 95.0% at a significant level: 0.05.