Optimizing installation parameters of DM columns in clay using Taguchi method

In this study, the installation parameters of the deep mixing (DM) column were optimized to manufacture a homogeneously mixed column and to reduce variability in the strength of the column. DM columns, 30 cm in diameter and 70 cm in length, were manufactured on clayey soils of different consistencies using a DM machine and slurry batch plant in the laboratory. The strength of DM columns and soil improvement ratio were investigated using installation parameters such as the cement dosage, the rotational speed of the mixing blades, the nozzle diameter, and the liquidity index of the soil. The design of experiments was conducted using the Taguchi method with a 4-parameter and 3-level L9 orthogonal array table. According to the observations during the installation of columns, the mixing time required to obtain a homogeneous soil-slurry mixture depended on the liquidity index of the fresh soilcrete. In addition, an interface layer, which is stiffer than the unimproved clay, was observed between the column and the unimproved soil. The thickness of the interface layer is 3–8% of the column diameter. Statistical evaluation was performed with S/N (signal-to-noise) and variance (ANOVA) analyses using the 28-day core strength of the columns. Optimum installation parameters were determined as cement dosage of 325 kg/m3, the rotation speed of the mixing blade is 80 rpm, nozzle diameter is 3 mm, and liquidity index of the soil is 1. The coefficient of variance for strength values decreased from 0.29 to 0.1 when the column was constructed with optimum installation parameters.