ABSTRACT

In this study, extreme vertex design (EVD) was adapted for the mixture experiment involving the clayey soils’ mechanical strength properties modification for civil engineering construction purposes through utilization of geotextile materials for soil reinforcement. EVD method provides an efficient approach to mixture experiment design whereby the factor level possesses multiple dependencies expressed through components constraints formulation. Using design experts and Minitab 18 software, the design of experiments, statistical diagnostics and influences, graphical and numerical optimization were carried out. I-optimal design with Quadratic model design was utilized to explore the constrained experimental region so as to formulate mixture ingredients proportions at 10 experimental runs. I-optimality and D-optimality of 0.39093 and 1747.474 respectively was obtained with G-efficiency of 64.8%. The test soil’s general engineering properties and classification were carried out before the mechanical properties tests with respect to varying ratios of mixture components. Experimental responses were obtained through laboratory process and the generated data utilized for model development and statistical analysis. The fits-summary computation showed preferences for Quadratic and linear model source respectively for CBR and UCS respectively. The statistical influence and diagnostic test results showed that there is no significant difference between the experimental constant results and the model results. Furthermore, desirability function was utilized to achieve numerical and statistical optimization in order to arrive at the optimal solution for the mixture components combination at 0.002:0.098:0.9 for geogrid, water and soil respectively. A desirability score of 1.0 was obtained with optimal response of 19.546% and 41.270kN/m² for CBR and UCS respectively. The results indicated improvement in the problematic soil’s mechanical properties due to the incorporation of geogrid material for pavement construction. Model simulation was further carried out to test the model’s applicability with the results compared with the actual results using ANOVA and student’s t-test. The statistical analysis results indicated a p-value>0.05 which indicates there is no significant difference between the compared datasets.