Soil Overconsolidation Changes Caused by Dynamic Replacement

Abstract

In the dynamic replacement method (DR) the soil is improved by initially dropping a large weight (typically 8-20 t) pounder from a significant height up to 25 m. The created crater is filled with a stronger material (gravel, rubble, stone aggregate, debris), and the pounder is dropped once or multiple times again. The construction of dynamic replacement pillars influences the parameters of the adjacent soil. It results from the energy generated by dropping a pounder into the soil. In the current practice, these changes are not taken into the account during the design. This paper focuses on the changes of overconsolidation ratio (OCR) and in situ coefficient of lateral earth pressure (K) values estimated base on cone penetration test (CPTU) and Dilatometric test (DMT) performed at a test site. A single column was constructed and the ground around the column was examined using CPTU and DMT, performed at different distances from the column centre (2, 3, 4 and 6 m) and at different time intervals (during construction and 1, 8, 30 days later). The column was constructed in so-called transition soils (between cohesive and non-cohesive). While interpreting the results of the research, the authors addressed the matter of choosing the procedure of OCR and K indication for transition soils (in this case described as silts and/or sandy silts). Overconsolidation changes may differ depending on the chosen analysis procedure (for cohesive or non-cohesive soils). On the basis of the analysis presented in the paper and the observation of soil (acknowledged as cohesive according to macroscopic observations) during column excavation, it was decided that for more detailed analyses methods dedicated to cohesive soils should be applied. Generally, it can be stated that although the changes were complex, DR pillar formation process resulted in the increase of these parameters. The average increases of OCR and K values were 25% and 10% respectively. The post installation values are not significant from the engineering point of view, but they represent the influence of the formation process of only a single column. The described results indicate that Priebe’s column dimensioning method should be applied with caution, as it assumes the value K=1 which was not obtained in the described research. The results from the conducted tests indicate that different mechanisms occur during stone column formation with vibro-replacement and dynamic replacement. As the authors did not manage to find literature describing the results of K tests in the surrounding of a DR column, the presented results should be acknowledged as significant for designers who will apply the dynamic replacement method.