Optimizing Construction Parameters of Dmc in High Plasticity Soils

Abstract

In this study, the parameters affecting the construction of deep mixing columns (DMC) in high plasticity clayey soil in the laboratory environment were investigated. For this purpose, laboratory type deep mixing machine, mixing blade, slurry pump are designed and manufactured. The characteristics of the soil to be improved (water content or consistency of the soil), slurry characteristics (cement dosage and water/cement ratio) and the selected system parameters in the column manufacturing (the drilling speed of the drill body, the rotation speed of the mixing blades, the nozzle diameter) affect the homogenous and continuous column formation. In this study, the clayey soil was placed into the 60 cm diameter and 110 cm high tank with the targeted water content (35-45-52%) and consolidated under a certain pre-consolidation pressure. Thus, a soil environment has been formed close to the field conditions. Then 30 cm diameter and 60-70 cm high DMC was manufactured with the laboratory type deep mixing machine. For the column production, the cement dosage was 250-325-400 kg/m3 and the water/cement ratio was 1. During column construction, the penetration rate of the drill body is adjusted to transfer the neccessary cement grout to the soil. The rotational speed of the mixing blade is 60-80-100 rpm. There are 4 injection nozzles on the mixing blade. The diameter of these nozzles varies between 2-3-4 mm. In order to understand the relation between the above mentioned parameters and their levels, a test program was created according to Taguchi's 4-parameters and 3-levels L9 design table. Column was produced according to the experimental program and core samples were taken from the columns after 28 days. A term of improvement ratio (ratio of DMC strength to untreated soil strength) was defined to evaulate experimental findings by statistical methods as S/N analysis, ANOVA, and regression analysis, and then optimum construction parameters were determined.