Structure, Properties and Modeling of Transported Residual Soil

Massachusetts Institute Of Technology, Cambridge MA

Investigators

Abstract

Studies of soil behavior have historically focused on properties of sedimented clays and clean sands in which the material micro-structure and engineering properties are defined by the deposition process and subsequent stress history. In contrast, the structure and properties of residual soils are controlled by in situ weathering processes that involve chemical dissolution and reprecipitation. Existing methods of classifying residual soils and predicting their behavior are far from adequate and can be linked directly to problems in construction. The PI's have recently been involved in a project in San Juan de Puerto Rico, where unexpectedly large ground settlements have occurred during tunneling through deposits of old alluvium. Further research has established that this material can be classified as a transported residual soil (i.e., weathered after deposition), a class of soil barely documented in the literature, but of apparently quite widespread occurrence. Extensive block samples of the soil already obtained by the PI's provide an excellent opportunity to investigate the structure, engineering properties and spatial variability of the material. To date, these preliminary investigations have identified iron oxides as the cementing agent, and have found an extraordinary reduction in bulk hydraulic conductivity due to breakdown in the structure in 1-D compression tests. This project will undertake a comprehensive experimental study of the material micro-structure and investigate how the breakdown in this structure can be linked to changes in the shear strength, deformation and flow properties. The work will benefit from the prior studies and availability of high quality block samples. Data from the experimental program will be used to formulate a constitutive model, based on a conceptual interpretation of the micro-structure. Model validation will then focus on predictions of material response associated with destructuring. This project will provide a valuable and unique database of properties for a transported, residual soil comparing the behavior of intact (in situ), destructured and intrinsic properties. By understanding the material structure, the project will also contribute towards standardizing the classification of residual soils. Future application of a well validated constitutive model will lead to more reliable predictions in geotechnical analyses of projects in the old alluvium and other similar deposits.