Investigations of threshold effects in soil deformations
Some of the methods used in deformation analyses are briefly reviewed. It is observed that often there is a large difference between data quoted from laboratory tests and that obtained from the back analysis of either field tests or back analysed structures. Some of the factors which affect soil stiffness are reviewed. The major factors affecting soil stiffness are identified as the stress history of samples, the current stress state and the current stress path. The effect of sample disturbance is to reduce measured stiffnesses, opinion has in the past attributed the differences in stiffness to this factor. However this is not supported by the results of high quality laboratory tests in which field and laboratory data agree well. The importance of care when defining stiffness moduli and the need for the definition to be given at appropriate points on the stress-strain curve are emphasised. Some indication of the significance of recent stress history in the form of threshold effects is discussed, although this is based on limited evidence in the case of stress path effects. The soils to be tested and the equipment to be used for the soil test program are discussed. Particular emphasis is placed on the calibration of the apparatus and system compliance. Details of the tests conducted are briefly given with details of the method of preparation of soil and soil samples for testing. The results of these tests indicate that as the deviation of the current stress path from that immediately preceding increases so the stiffness on a given stress path increases. In addition it is observed that the strain increment ratio is non-unique. These effects are observed to be most important for high plasticity soils. Further tests indicate the independence of time and stress path threshold effects. The variation of these effects with stress history and overconsolidation ratio are investigated. It is found that these effects are most significant for soils which have been isotropically compressed and swelled to have an overconsolidation ratio of two. The results of these tests indicate some of the reasons for difficulties in tests to examine path dependence in earlier research. Finally some conclusions from this experimental work are drawn together and topics for further research suggested. The importance of considering the complete stress history of samples including recent stress paths and periods of rest are emphasised for inclusion in the stress path method.