The field monitoring of the climatic-induced behaviour of the expansive soil has always been difficult, expensive and time consuming. The uncontrollability of the field boundary conditions and the difficulty in accurately measuring them have worsened the problem. As an alternative, the instrumented model set-ups are ideal for long-term monitoring of expansive soils since the laboratory compacted expansive soils become environmentally stabilised after few wet–dry cycles. There had been a very limited laboratory-based column set-ups for the observation of expansive soils under unsaturated conditions with an appropriate set of sensors embedded at known depths. The major difficulties associated with model tests are considerable boundary effect and sensor-to-soil area ratio due to the insufficient physical model dimensions. In this study, the research need for a laboratory model set-up with minimised boundary effects has been addressed by a large instrumented soil column, which could more closely represent environmentally stabilised soil. The current results depict the expected pattern for the variations of soil suction, volumetric water content and soil displacement under wetting and drying phenomenon, which accentuates the applicability of instrumented soil column for the investigation of climatic-induced expansive soil behaviour.

Expansive grey Vertosols are predominant in Australia, especially in Queensland. These soils induce a significant amount of swell strain on the lightweight structures founded on/in these soils when soil moisture changes due to climatic influence. According to AS2870: Residential slab and footing design code (AS2870, 2011) and available literature, active zone depth which contributes to the climate induced-soil displacement in Queensland can be up to 5m. To design lightweight structures founded within the active zone depth, it is imperative to understand swelling behavior of expansive grey Vertosols. The resultant swell strain of these expansive soils is primarily a function of initial water content or initial soil suction and overburden pressure (surcharge). However, there is an extremely limited amount of data available for unsaturated swelling characteristics of expansive soils in Queensland. In this study, a series of conventional oedometer tests were conducted to investigate the effect of the initial water content, initial soil suction and the surcharge on the swelling behavior of expansive grey Vertosol that is predominant in South-East Queensland. The swell strain of expansive grey Vertosol displayed greater sensitivity to surcharge when compared to initial water content or suction. Developed relationship between swell strain per unit change in volumetric water content and surcharge depicted statistically strong (R 2 = 0.89) agreement. The application of the knowledge gathered from this study will benefit the decision making in semi-arid regions when building lightweight structures on grey Vertosol.

The moisture variations in expansive soils cause shrink-swell behaviour, resulting in distress to the structures founded in/on problematic soils. The oedometer based tests can be used to determine swell behaviour of soil; however, limited research has been conducted for vertical shrinkage estimations. In this study, a series of conventional oedometer tests were conducted to investigate the vertical shrinkage of grey Vertosol due to soil moisture variations under different surcharges. A statistically strong relationship (R 2 = 0.99) was observed for shrinkage per unit change in volumetric water content under shallow overburden pressures (surcharges). The validation of the shrinkage was conducted by simulating field conditions under induced drying cycle. Derived shrinkage prediction equation and Aitchison's method showed underestimations of 10.1% and 44.0% of the actual shrinkage respectively. Briaud's and Dhowian's models overestimated the value by 59.0% and 44.5% respectively. This study emphasizes the applicability of the conventional oedometer based shrinkage test for a reasonable estimation of vertical shrinkage for a given expansive soil. Thereby, proposing a simple and practical approach to obtain shrinkage characteristics for geotechnical engineering applications.

No abstract

Expansive soils exhibit swell-shrink behaviour in wet-dry periods resulting in distresses on light-weight structures founded on/in them. Therefore, it is essential to investigate the climate-ground interaction when designing structures on expansive soils. Laboratory-based models are preferred to investigate the climatic-ground interaction of expansive soils due to the uncontrollability of the boundary conditions and expenses associated with field monitoring. More flexibility in analysing the climatic-induced hydraulic responses in expansive soils can be achieved by finite element modelling of data from physical model tests. However, these laboratory-based models regularly encounter the effects of boundary flaw, preferential flow paths and entrapped air that needs to be accounted for when numerically simulated. In this study, the authors aim to numerically model the hydraulic responses in an instrumented Vertosol soil column (ISC) under controlled laboratory conditions. The effects of the preferential flow paths and boundary flaws were incorporated into a modified hydraulic conductivity as a practical approach to model the hydraulic responses in ISC. Influence of the entrapped air was rectified by a suitable correction factor. These findings present a practical method for geotechnical practitioners to accurately estimate the suction and volumetric water content profiles in laboratory-based expansive soil model tests.