Reliability-based analysis and design of geotechnical soil-structures and transparent granular soils for geotechnical testing

点击翻译按钮获取中文摘要

Reliability-based design and analysis of geotechnical soil-structures is now accepted as the way forward in geotechnical design practice for foundations, retaining walls, reinforced slopes and embankments. Load and resistance factor design (LRFD) is a variant of reliability-based design that has been used in structural engineering design codes since the 1970s. This project will provide engineers with a useful tool for LRFD calibration using spreadsheets and thus avoid Monte Carlo methods that obscure understanding of the contributions of different sources of uncertainty to probabilities of failure. The general approach includes the concept of “level of understanding” that rewards engineers with higher resistance factors for gathering more project data. Candidate applications are steel and polymeric reinforced soil walls and soil-nail reinforced systems. Calibration outcomes for resistance and load factors will be used in future additions of the National Building Code of Canada (NBCC) and the Canadian Highway Bridge Design Code (CHBDC). The random finite element method (RFEM) is a valuable research tool for analysis of reinforced slopes and embankments. It will now be extended to geosynthetic reinforced walls and foundations. The advantage of the RFEM is that it can consider the influence of soil spatial variability on soil-structure performance. The proposed work will investigate non-circular random walk techniques to find the same critical slope failure mechanisms as the RFEM, but with the possibility to avoid excessive computation times. The method will be applied to unreinforced and reinforced soil slopes and embankments, reinforced soil walls and soil nail structures. The first practical transparent surrogate granular soil was invented by the applicant and PhD student Ezzein in 2011. Since that time there has been a proliferation of interest and work investigating a large number of soil-structure interaction problems. A specially manufactured plane strain apparatus that allows the interior of a specimen of transparent soil to be viewed through a transparent membrane will be used in this work. An array of seeded painted targets will be tracked using the digital image correlation (DIC) technique. A laser will also be used to illuminate and track particles in different planes through the specimen using the DIC. These methods will allow the deformation field within a specimen to be observed directly rather than inferred from boundary measurements which is the case using natural opaque sand materials. Different transparent soils varying with respect to particle shape, size and gradation will be investigated. The testing will provide better understanding of the particle-level behaviour of granular soils during plane strain and triaxial testing. Macro-scale stress-strain properties will be available to model the same transparent soils used in centrifuge and 1g bench-scale tests.