Abstract

The shrinkage limit is one of the Atterberg limits and is a fundamental geotechnical parameter used to assess the settlement of engineering soils containing clays, yet is rarely tested for as part of ground investigation. This paper describes shrinkage limit test results on a variety of soils from Britain and overseas obtained using an improved laboratory testing procedure developed at the British Geological Survey (BGS). The co-relationships with the other Atterberg limits and with density are explored. In particular, the coincidence of the shrinkage limit with the water content at the peak bulk density achieved in the test is examined. The shrinkage behaviour for undisturbed and remoulded states and a 3-way relationship between water content, density and suction are demonstrated. Some tropical residual and highly smectitic soils show a very wide range of shrinkage behaviour, albeit for a small dataset, when compared with the larger dataset of temperate soils tested. Consideration is given to limitations of the new and existing test methods.