Drift wall boundary conditions

In the case of a boundary of one of the two remaining types the particle can arrive at the boundary so that a boundary condition is needed. One may impose the absorbing condition. The reflecting boundary condition, however, makes sense only if the particle is able to drift back away from the wall into the interval. This is the case in type (iv), but not in (iii). 

The boundary conditions shown in Figure 1 were determined from FEBEX in situ stress data (Pahl et al., 1989). An additional boundary condition was used to model the time-dependent stress induced by the alpine miner during excavation of the drift. A pressure value of 25 MPa was applied to the wall in successive sections of the drift when that material was "excavated" numerically in the calculation. 

Abstract Results from the four-year long heating phase of the Drift-Scale Heater Test at the Exploratory Studies Facility at Yucca Mountain, Nevada, provide a basis to evaluate conceptual and numerical models used to simulate thermal-hydrological coupled processes expected to occur at the proposed repository. A three-dimensional numerical model was built to perform the analyses. All model simulations were predicated on a dual (fracture and matrix) continuum conceptualization. A 20-percent reduction in the canister heat load to account for conduction and radiation heat loss through the bulkhead, a constant pressure boundary condition at the drift wall, and inclusion of the active fracture model to account for a reduction in the number of fractures that were hydraulically active provided the best agreement between model results and observed temperatures. The views expressed herein are preliminary and do not constitute a final judgment of the matter addressed or of the acceptability of its use in a license application... 

The loading of the structure is mechanically very simple since it consists in an initial isotropic stress field related to the dead weight. Concerning the mechanical boundary conditions, a zero normal stress is prescribed on the free boundaries of the pattern (wall of the drifts and well, and ground surface), and the symmetry planes are characterized by a zero normal displacement. Before excavation, the rock mass is supposed to be in a compressive stress state, and the principal minor stress 03, indicator of maximum compression, is equal to CTh (and 0i=O2=o =ayy). This stress increases (in absolute value) with depth, from -1.1 MPa at the top of the wells to -1.6 MPa at its base, and to -3.1 MPa in lower limit of the model. The excavation of drifts and wells causes a disturbance of this initial stress field (see fig.3). It is noticed that, apart... 

Lastly, the walls of the drifts are subjected to a convection condition with air circulating at a temperature T2 of 20°C in the lower drift and Tj of 60°C in the upper. From the geometry of each drift and air flow circulating in each well, the exchange coefficients hi and h2 are respectively about 11 W/mVK for the upper drift and 8 WW/K for the lower. Using these boundary conditions and a diffusion model taking of account the conduction of heat in the rock mass, the temperature field has been calculated by CAST3M (see fig.4). 

Masomy acceptance criteria (in Annex C) are quantified in terms of wall drift - in common with ASCE 41 for flexure-controlled walls, but in contrast to ASCE 41 for shear-controlled walls for which strength criteria apply. For flexure-controlled walls, drift hmits are given in terms of the slenderness ratio of width to the height of the point of contraflexure, which gives limits similar to those given in ASCE 41 for fix-fix boundary conditions of wall piers but doubles the limits for cantilever walls (mapping the limit states as discussed for steel). 

---