Evaluation of contributing factors on strength development of lime stabilized artificial organic soils using statistical design of experiment approach

Abstract

Lime is widely used as chemical stabilizer in soft soil stabilization. However, lime is reported to be less effective when dealing with organic soil. It is believed that the organic matter in the soil will retard the pozzolanic reaction which is responsible for strength enhancement. The heterogeneity nature of the organic matter in the soil makes the study complicated and reduced the repeatability of the test results. Hence, artificial organic soil with known organic matter and content are preferred by researchers when repeatability of the test results are required in determining the influential effect of each contribution factor. Various factors such as additive contents, effect of aging (curing periods), curing temperature, density of materials and moisture content are reported by previous researchers as the potential contributing factors towards the strength development. It is believed that the interaction of the factors also will contribute to the strength enhancement. Hence, this study is carried out to evaluate the contributing factors and its interactions on strength development of artificial organic soils with known type and contents of organic matter. Statistical design of experiment (DOE) approach was utilized to evaluate the factors and its interaction on the strength development of lime stabilized artificial organic soils by using commercial statistics package. Three main factors were investigated: effect of organic content, effect of curing periods, and effect of additive, while other factors namely curing temperature, molding water content, types of compaction and compactive effort were keep constant through controlled experiments. Processed kaolin (inorganic material) is mixed with humic acid (organic matter) to simulate the organic soil which comprised of inorganic soil and organic matter. The density of the soil specimen and its moisture content were recorded before and after the curing process. General Linear Model (GLM) was utilized to determine the significance of the main factors, two-factor interactions, and three factor interactions. The significance factors and interactions were utilized in multiple regression analysis to develop the strength prediction model which can be utilized to predict the strength of stabilized materials within the inference space defined by the experiment.