It is prerequisite to estimate rock stress for designing underground rock structures constructed in various engineering projects such as disposal of high-level radioactive wastes. In general, rock mass is heterogeneous since it is composed of different rock bodies. Accordingly, stress measurements in a rock mass may give us only local stresses, which are affected by the heterogeneity of the rock mass, discontinuities and the surface topography of the region. Some researchers have tried to estimate regional stress from local stresses. However, the effects of the heterogeneity of rock mass have never been considered in previous studies.
In this study, regional stress in a heterogeneous rock mass was defined as an average strain field acting at boundaries, which are sufficiently far from a region under consideration. The average strain field is given by displacements as a function of the coordinates to satisfy the condition of continuity for the displacements at the boundaries. We proposed a new method for estimating regional stress from data of stresses measured in a heterogeneous rock mass by using a 3-D finite element method.
This method was applied to small-scale heterogeneous models composed of different rock bodies to know the effects of mechanical heterogeneity on estimation of stress within a region under consideration. The results showed that an estimation error increases as the heterogeneity of the rock mass increases. When the region under consideration is homogeneous, the location or number of measured points has little effects on the estimation error. On the other hand, when the region under consideration is heterogeneous, the estimation error in stress is greatly reduced by using data of stresses measured in different rock bodies. Thus, stress should be measured in different rock bodies to accurately estimate the regional stress. Finally, this method was applied to estimate regional stress in the Tono district, Gifu prefecture, but without consideration of Tsukiyoshi fault existing in this district. Although only sixteen stress components measured in three formations were used, the stress distribution estimated from the regional stress was in good agreement with the tendency of measured data.
View full abstract