<p>This paper investigates the need to explicitly consider strain localisation in constitutive model calibration and numerical tunnel analysis from a practical engineering point of view. Motivated by the conditions expected in the planned Swiss deep underground repository for radioactive waste, which will be located in Opalinus Clay – a material that exhibits softening behaviour – the simplest possible tunnelling problems satisfying rotational symmetry in the absence of localisation are considered: (a), the ground response curve (GRC) and, (b), the ground-support interaction for a circular tunnel. Two modelling approaches are comparatively evaluated: (i) a rigorous approach, in which strain localisation is considered in model calibration and regularisation is applied in finite element (FE) analyses to account for mesh dependence and to correctly consider the softening rate; and (ii) a simplified approach, in which model calibration assumes uniform specimen deformation and no regularisation is performed. Calibration by these two methods is based on the results of a fictitious biaxial test, representative of Opalinus Clay samples. The relatively simple biaxial test problem also serves to illustrate some of the peculiarities and limitations of modelling problems involving strain softening. The rigorous GRC calculations are based on an extremely fine spatial discretisation and reveal the effect of bifurcation on the deformation field around a tunnel and the resulting variation in radial displacements. The comparative simplified analyses confirm the expected mesh sensitivity of the solution but show that this effect is small and that the simplified approach is adequate for the considered problem layout and material parameters. This is also true for the ground-support interaction problem, provided that the ground-lining interface is modelled realistically. These conclusions are of significant value for the design computations concerning the short-term behaviour of the ground during the construction of the underground openings of the proposed Swiss radioactive waste repository. Nevertheless, despite the focus on a specific project and the limitation of these conclusions to the conditions of this project, the approach of this paper is methodologically valuable, not least for other projects with different conditions.</p>

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

On the need to Consider Strain Localisation in Undrained Analyses of Tunnels Crossing Opalinus Clay

  • Athanasios Agalianos,
  • Ioannis Anastasopoulos,
  • Georgios Anagnostou

摘要

This paper investigates the need to explicitly consider strain localisation in constitutive model calibration and numerical tunnel analysis from a practical engineering point of view. Motivated by the conditions expected in the planned Swiss deep underground repository for radioactive waste, which will be located in Opalinus Clay – a material that exhibits softening behaviour – the simplest possible tunnelling problems satisfying rotational symmetry in the absence of localisation are considered: (a), the ground response curve (GRC) and, (b), the ground-support interaction for a circular tunnel. Two modelling approaches are comparatively evaluated: (i) a rigorous approach, in which strain localisation is considered in model calibration and regularisation is applied in finite element (FE) analyses to account for mesh dependence and to correctly consider the softening rate; and (ii) a simplified approach, in which model calibration assumes uniform specimen deformation and no regularisation is performed. Calibration by these two methods is based on the results of a fictitious biaxial test, representative of Opalinus Clay samples. The relatively simple biaxial test problem also serves to illustrate some of the peculiarities and limitations of modelling problems involving strain softening. The rigorous GRC calculations are based on an extremely fine spatial discretisation and reveal the effect of bifurcation on the deformation field around a tunnel and the resulting variation in radial displacements. The comparative simplified analyses confirm the expected mesh sensitivity of the solution but show that this effect is small and that the simplified approach is adequate for the considered problem layout and material parameters. This is also true for the ground-support interaction problem, provided that the ground-lining interface is modelled realistically. These conclusions are of significant value for the design computations concerning the short-term behaviour of the ground during the construction of the underground openings of the proposed Swiss radioactive waste repository. Nevertheless, despite the focus on a specific project and the limitation of these conclusions to the conditions of this project, the approach of this paper is methodologically valuable, not least for other projects with different conditions.