Soil Improvement method by Mass Stabilization : Settlement analysis in stabilized sulfide soil

Mass stabilization is not a common soil improvement method in Sweden, and even less so in sulfide soils. Previous research indicates that a linear elastic material model is suitable for settlements in stabilized sulfide soil. A construction project in the northern part of Sweden applied mass stabilization in sulfide soil. This was an opportunity to study the mass stabilized sulfide soil in the field. This master thesis was executed in parallel to that construction project. Ground preparations in the project meant that 5-8 m fill would be placed on fine-grained sulfide soil. Soil improvement and specific foundation designs were determined to be required. Mass stabilization of the sulfide soil was one of the soil improvement methods used. The binder recipe for the mass stabilization was determined to be 60 kg cement/m3. Which also was the binder recipe analyzed in this thesis. The goal of the thesis was to investigate if settlements in stabilized sulfide soil could be predicted with a linear elastic model. This was done by using soil parameters derived from: o Back calculations of measured settlements caused by a test embankment. o UCS tests. o Oedometer tests. Deriving soil parameters by back calculation from the test embankment was done by: Assuming Terzaghi’s one dimensional consolidation theory. Defining the settlement curve by Casagrande’s logarithm of time fitting method. And then iterating the elastic modulus so that calculated settlements would match the defined settlements. The same method used for back calculations of the test embankment was performed in micro-scale on the Oedometer tests. Deriving soil parameters from the UCS tests was done according to previous recommended methods presented by (Al-Jabban, 2019) and (Åhnberg, 2006). The derived soil parameters were then used in a numerical model in Plaxis 2D to calculate settlements. The material model Mohr-Coulomb was used in Plaxis 2D. Three different settlement calculations were performed, one for each source of soil parameters. The calculated settlement results were compared to measured settlements caused by the test embankment. The calculated settlements gave good predictions of measured settlements in stabilized sulfide soil during immediate- and primary consolidation. However, depending on the binder recipe, secondary consolidation might occur in stabilized sulfide soil. Some binder recipes underneath the test embankment seemed show no secondary consolidation. But the analyzed binder recipe, 60 kg cement/m3 experienced secondary consolidation. Secondary consolidation cannot be calculated with the linear elastic material model used in Plaxis 2D. Nor can it be evaluated from the UCS test. However, if secondary consolidation occurs, it can be added with hand calculations by parameters derived from “Casagrande logarithm of time fitting method.” The calculated settlements from back calculation of the test embankment and from the oedometer tests gave good predictions of the settlements during secondary consolidation. But this should be confirmed by measuring settlements over a longer period of time. One advantage with soil parameters derived from laboratory investigations is that it is a faster method than soil parameters derived from back calculations of a full-scale field test.