Model tests and corresponding numerical simulations on cave model subjected to thermo-mechanical loading

2021 
Abstract Geological repository of high-level nuclear waste (HLNW), the most feasible approach for the safe and permanent treatment of HLNW without human intervention, has been investigated for years by many researchers. In some countries, real facilities for the geological repository are already under construction. In Japan, however, due to complicated geologic conditions, especially underground water and fractured rock masses, and the high risk of natural disasters, ensuring the long-term stability of the method remains a struggle. The influential factors include underground water, heat generation from radioactive waste, and thermal and chemical weathering of the surrounding rock mass as a natural barrier. It is difficult to estimate and verify the long-term stability for up to one hundred thousand years, a complicated thermal–hydraulic-mechanical-chemical coupling behavior, via any field test. The objective of the study is to develop a numerical method for predicting the long-term stability of geological repositories. As the first step toward realizing this objective, heating and loading tests on cave model made of man-made rock specimens that are composed of diatoms, gypsum, and water, which are viscoplastic materials, were conducted, based on which a newly proposed numerical method with finite element method (FEM) was used to describe the thermal, mechanical, and time-dependent behavior of the model tests for a geological repository. To ensure the accuracy of the numerical calculations, all the material parameters in the thermoelasto-viscoplastic model with consideration of overconsolidation, the structure, and the influence of intermediate stress were determined via triaxial compression/creep tests under various temperatures, confining stresses, and loading rates. Finally, the validity of the numerical method was demonstrated by model tests over a limited time span.
    • Correction
    • Source
    • Cite
    • Save
    • Machine Reading By IdeaReader
    28
    References
    0
    Citations
    NaN
    KQI
    []