Investigation of factors affecting the performance of deep mixing columns by model experiments

In the scope of this thesis, parameters affecting the performance of fly ash additive deep mixing columns (DMC) manufactured in a clay soil were investigated. Experimental studies were carried out on columns manufactured with a deep mixing machine (large scale DMC) and soil-slurry mixtures prepared with mixer (small scale DMC). Designs of experiments were organized using the Taguchi method. The results obtained by determining the DMC performance values at each research area and at the end of the relevant curing periods were evaluated with S/N and multivariate variance analysis (ANOVA), and mathematical models were established with regression analysis and then optimization studies were carried out. In order to construct large scale DMC, the manufacturing parameters required for homogeneous and continuous DMC formation were investigated first. Optimum construction parameters were determined as cement dosage is 325 kg/m3, the rotation speed of the mixing blades is 80 rpm, the diameter of the grout nozzles is 3 mm and the liquidity index of the soil is 1. The unconfined compression (qu) and triaxial compressive strengths (quu), soil improvement ratio (RI), consolidation index (Cc) and permeability coefficient (k) values of fly ash additive DMC designs were determined at 7, 28 and 56 days of curing times. Optimum construction parameters for the best performance were determined as the binder dosage is 425 kg/m3, the fly ash ratio is 40%, the superplasticizer ratio is 3%, the water/binder ratio is 0.8 and the liquidity index of the soil is 1. The use of cement and slurry cost will be reduced by 40% by using 40% fly ash in DMC applications. While the strength performance of small scale DMC is 40% and 30% higher in qu and quu values, respectively, compared to the performance of large scale DMC, it is obtained 20% and 45% less in Cc and k values representing consolidation and permeability properties, respectively. In small and large scale DMC designs; when curing time (t), binder factor (α) and content (aw) increase, qu, quu and RI increase, Cc and k decrease. As the total water/binder ratio (WT:Wb) and/or the liquidity index (IL, mix) of the soil-slurry mixture increases, qu and quu decrease, while Cc and k increase. The best condition for DMC performance was obtained when IL,mix was (1.2-1.3)IL. With the SEM images, the condition of the voids and the formation of C-S-H, C-A-H gels, CH plates and etrenjit needles in the microstructure of DMC samples have been investigated and associated with DMC performance. The results obtained from experimental and statistical studies have been compatible with the literature.