A SUSTAINABLE DESIGN FRAMEWORK INCORPORATING THE STATISTICAL AND SPATIAL VARIABILITY OF CEMENT-TREATED SOILS FOR USE IN EXCAVATION SUPPORT

Abstract

In sites with weak soils, underground construction activities such as excavation and tunnelling can result in large soil movements. While the use of cement-treated soils can help reduce ground movements, it often increases the construction costs significantly. This thesis focuses on the proposal of a new design framework which allows the systematic selection of cement-treated soil parameters. This is achieved by first studying the statistical and spatial variability of this material, the information from which is then used to formulate the statistical framework. Deterministic and random finite element analyses of an existing monitored deep excavation where cement-treated soil was used are then performed to examine the system performance in a probabilistic manner. The use of this proposed framework can aid practicing engineers in the selection of characteristic material values that is expected to result in a more rational and sustainable design of deep excavation support systems.