Plastic anisotropy and destructuration of soft Finnish clays

Aalto University, P.O. Box 11000, FI-00076 Aalto www.aalto.fi Author Mirva Koskinen Name of the doctoral dissertation Plastic anisotropy and destructuration of soft Finnish clays Publisher School of Engineering Unit Department of Civil and Environmental Engineering Series Aalto University publication series DOCTORAL DISSERTATIONS 169/2014 Field of research Soil Mechanics and Foundation Engineering Manuscript submitted 16 June 2014 Date of the defence 12 December 2014 Permission to publish granted (date) 20 October 2014 Language English Monograph Article dissertation (summary + original articles) Abstract Due to growth of population, there is a need to construct on areas on soft soils. Therefore, there is also a need for tools for modelling deformations of soft soils. Soil models called SCLAY1, accounting for fabric anisotropy, and S-CLAY1S, accounting additionally for bonding and destructuration, were developed. Aim of this work was to design and perform a comprehensive series of laboratory tests in order to validate the S-CLAY1 and S-CLAY1S models. Four typical soft Finnish clays were chosen for test materials. The clays were Otaniemi clay from Espoo, POKO clay from Porvoo-Koskenkylä motorway site, Murro clay from Seinäjoki and Vanttila clay from Espoo. These clays represent a typical variety of moderately to highly sensitive clays. The sensitivity is a demonstration of interparticle bonding of natural clays. The series of tests included tens of triaxial tests on both reconstituted and natural samples. In tests on reconstituted samples, anisotropy can be studied in the absence of bonding. Natural clays, however, exhibit anisotropy and bonding. The triaxial tests were consolidation tests with different stress ratios, and drained and undrained shear tests. Additionally, oedometer tests and index tests were carried out. The behaviour of Murro test embankment was also calculated with the two models. Model parameters for S-CLAY1 and S-CLAY1S models were determined from the test data, and the test results were simulated as single element simulations. The simulations showed that accounting for anisotropy enhanced the predictions of tests on reconstituted samples significantly compared to predictions with an anisotropic model. Using certain parameter values, the anisotropic S-CLAY model could also be used for satisfactory predictions of tests on natural samples, but, however, using the destructuration law incorporated in the S-CLAY1S model, the predictions could be improved further. The behaviour of Murro test embankment was also modelled rather realistically using the S-CLAY1S model compared to observations of settlements and horizontal displacements. S-CLAY1 and S-CLAY1S do not account for rate effects or elastic anisotropy, but despite of that, the models are a promising tool for geotechnical design and a basis for model development in the future.