VALIDATION OF SWELLING ESTIMATION METHODS USING A LABORATORY-SCALED INSTRUMENTED SOIL COLUMN

Abstract

Damages to lightweight structures due to expansive soils present a significant global challenge in arid and semi-arid regions. These areas are highly susceptible to climatic changes that alter moisture conditions in expansive soils, leading to swell-shrink behavior. The complexity of ground movements induced by moisture (suction) variations is exacerbated by many factors influencing the process, making accurate prediction a formidable task. Various swelling prediction methods are currently employed worldwide, each with its own limitations and varying degrees of success. Consequently, validating any heave prediction method before its application to specific soil types is crucial. This validation can be accomplished through either a field instrumentation program or a laboratory-scaled model to simulate actual field conditions, with the last being preferable due to its cost-effectiveness and ability to control boundary conditions. However, there is a significant lack of studies investigating the applicability of current swelling prediction methods to Queensland soils. This study proposed the development of a long-term operable instrumented soil column (ISC) subjected to four phases (wet-dry-wet-dry) to examine the climatic-induced hydro-mechanical responses of unsaturated expansive grey Vertosol, which is widespread in South-East Queensland. The observed ISC data validated the widely applied current heave prediction methods. Aitchison and Peter (1973) and Oedometer-based methods demonstrated consistency in estimating movements in all four phases compared to other selected methods. The findings of this study provide confidence for accurately predicting expansive soil movements in Queensland's expansive soils, thereby minimizing associated damages and maintenance costs.