Effect of Sample Disturbance on Compressibility Parameters

Abstract

The so-called undisturbed samples are widely used in the laboratory to obtain the engineering and index properties of the soils. Rarely does one get a perfect undisturbed sample as the method of sampling itself induces disturbances in the soil sample, such as changes in moisture content, soil structure, and stress conditions, all of which effect the compressibility and consolidation properties of the soft soils especially. In the present study, void ratio versus log effective stress results from both disturbed and undisturbed samples available in the literature are analysed, and compression and swelling indices, maximum past consolidation stress and OCR are estimated. Casagrande and log–log (Sridharan et al. in Geotechnique 41(2):263–268, 1991, [14]) methods were applied to estimate the preconsolidation pressures. Maximum past consolidation stress and OCR get reduced and the effect of past overconsolidation partly erased in the disturbed or not so perfect samples. Settlements for a given surcharge stress and layer thickness are then estimated for both disturbed and undisturbed states. Ratio of settlement based on disturbed state with respect to the undisturbed one is found to be greater than one implying settlements are overestimated in the conventional practice based on imperfect samples.