Ground Improvement by Dry Deep Mixing Lime-Cement Column Panels as Excavation Support

Abstract: Many urban areas near the coastal regions of Sweden are characterized by post-glacial clay deposits with very low undrained shear strength and high compressibility. Column type ground improvement by the Deep Mixing, DM, method using a binder mixture of lime and cement is commonly used in areas with poor soil conditions due to its cost-effectiveness, predominantly for settlement reduction and to improve the stability of embankments. With increasing urbanization and infrastructural development in these areas there is great interest in the industry in extending the practice of the Deep Mixing method to include other applications such as deep excavation and temporary and permanent improvement of natural slopes. Swedish experience related to use of the DM method for excavation support is limited and the current design recommendations for DM columns installed in the passive zone are very restrictive regarding the allowable mobilized column strength, resulting in a design which is often not cost-effective.In order to increase the use of the method to include applications where DM columns are subjected to unloading and lateral loading conditions, the mobilized strength and stiffness properties of the columns as well as the soil-column interaction need to be reliably predicted. The main objective of this study is to present a consistent method to adequately predict the behavior of lime-cement columns installed as excavation support in the passive zone of the structure and to investigate the strength and stiffness properties of lime-cement improved clay under different unloading and laterally loading conditions together with the soil-column interaction under these conditions.In order to investigate the field behavior of lime-cement column panels as excavation support, two experimental full-scale tests were performed. In each of these tests, a braced steel sheet pile wall supported by panels of overlapping lime-cement columns was first excavated to a pre-determined depth and thereafter loaded to failure by stepwise increasing a load applied behind the sheet pile wall. The tests provided a case record of deformations, stresses, and pore pressure responses, and failure mechanisms of the structures focusing on the improved soil. These tests showed that column-type ground improvement installed as panels of overlapping columns in the passive zone of a sheet pile wall significantly increases stability and reduces both excavation- and loading-induced structural forces and vertical and horizontal displacements in the soil.This thesis also presents the results of a laboratory study involving undrained and drained isotropic consolidated triaxial compression, extension and tension tests on laboratory improved clay with a binder of lime-cement similar to that used in the experimental field tests. Based on undrained triaxial test results, a relationship between the undrained strength, effective consolidation stress, and overconsolidation ratio is presented for different stress paths to failure. From the drained triaxial tests it was found that a failure surface comprising of two failure functions, one for tension failure and one for shear failure, similar to that observed for cemented sand, is consistent with the experimental data. Finally, a 3D FE-study of the experimental field tests considering the laboratory observed stress-strain behavior and mobilized strength of lime-cement improved clay was conducted. The results of these analyses are promising and failure load, deformations and structural forces in the retaining structure were predicted reasonably well.Summarizing the most important findings and conclusions from this study:-          Lime-cement columns panels installed in the passive zone acting as excavation support for a sheet pile wall will significantly increase the stability of the structure.-          Lime-cement column panels installed as excavation support are effective in reducing excavation induced displacements that can be of major concern for deep excavations conducted in areas with soft clay layers.-          The undrained strength of lime-cement improved clay at low consolidation stresses, corresponding to approximately 10 m of depth in field conditions, is dependent of the stress path to failure and it was found to be significantly lower for unloading stress paths compared to lateral loading stress paths, i.e. stress induced anisotropy. -          The Young’s Modulus of lime-cement improved clay evaluated from undrained triaxial extension tests was significantly higher, 2.7 to 4.1 times, compared to the corresponding Young’s Modulus evaluated from the undrained triaxial compression tests. Also, significantly more brittle stress-strain behaviour was observed for triaxial extension tests compared to triaxial compression tests, regardless of applied stress path to failure and type of test, i.e. undrained/drained. -          Results of the Finite Element analysis of the conducted experimental tests show that the current Swedish Design Guide for lime-cement columns installed in the passive zone overestimates the material undrained strength when based on results from Unconfined Compression tests, but also significantly underestimates the material drained strength. Since the Swedish Design Guide specifies that the lowest of the undrained/drained column strength should be chosen in the design, the consequence is often a too conservative design as the strength increase in the improved clay is not properly considered.