The use of moisture probes to infer changes in suction due to controlled inundation behind a full scale trial retaining wall

Abstract

To better understand the behaviour of unsaturated clays for retaining wall design in Adelaide, South Australia, a trial soldier pile retaining wall supporting an 8 m deep excavation was constructed for a proposed 3 km long underpass. The full scale trial retaining wall was constructed prior to the start of summer to allow the soil to shrink over an extended dry period, typical of an Adelaide summer. At the end of summer, the soil was wetted up via a controlled inundation program that replicated a pipe burst or leaking water main, critical considerations for retaining wall design. As part of an extensive testing and monitoring program, in situ moisture probes were used to quantify the change in soil moisture content before, during and after inundation. This paper presents a practical means to estimate soil suction based on the calibration of moisture probe data and the use of a site specific soil-water characteristic curve. Limitations of using moisture probes to infer suction are discussed, with suggestions provided for the future use of similar moisture probes in very stiff to hard clay soils as encountered on this site.