Research Of Evaluation Methods For Rock Mass Structures In The Wudongde Hydropower Reservoir Region

Various structural types of rock masses distributed in the Wudongde hydropowerreservoir region are broken because of the frequent tectonic activity and strong exogenicforces (e.g., weathering). The deformation and destruction of the rock masses threat thestructures and the constructors, which may lead to a great loss. Therefore, the brokenrock masses in the reservoir region should be regarded.Using “controling theory of rock mass struture” as the guiding ideology, variousmeasures were used to investigate the structural characteristics of the rock masses in thereservoir region. The particular features of the rock mass parameters and themechanisms of deformation and destruction, which are influenced by rock massstructures, were researched. Furthermore, susceptibility and dynamics of debris flowsinduced by the destruct rock masses were researched. The conclusions were drawn asfollows:1. Based on the generalized conception of rock masses, the rock masses can bedivided into6categories (considering the genetic mechanism, stability degree, anddeformation/destruction mechanisms), i.e., intact and semi-intact rock masses, stronglyfractured rock masses, highly to completely weathered rock masses, semi-consolidatedstratum, ancient accumulation bodies, and recent accumulation bodies. The diameters ofthe grains (which can be defined as gravel soil) in Madianhe Group layer varysignificantly. Large numbers of vertical cracks are developed. Calcareous structures willbe destroyed upon contact with water, leading to decrescendotype collapses and slides. The grains (which can be defined as fine sand) in Longjie layer are fine. Large numbersof crossed horizontal beddings and vertical cracks are developed. Thedeformation/destruction mechanisms of the Longjie layer are the same as those ofMadianhe layer (decrescendotype collapses and slides). For the red-bed soft rocks, thecontents of clay minerals, calcites, and dolomites in the silty mudstones are high. Grainyand flaky minerals are relatively concentrated, with extremely low resistance to weather.The soft and hard rock layers show interlaced distribution; thus, the stability degree ishigh, with few collapses and landslides. Taking highly to completely weathered phylliterocks as an example, the contents of clay and mica minerals are high. Mineral grainsshow laminated and aerial structures, with low resistance to weather and low strength.For the exterior weathered metamorphic rocks, only small-scale collapses happen. Forthe large-scale loose accumulation bodies, large-scale collapse and landslides frequentlyhappen.2. Three-dimensional (3D) fracture network modeling can be applied to thefractured rock masses. The generated fractures can be described by numerical data.However, the data are nonobjective. Visualisation technique can show the spatial shapesof the fractures and the position relationships among the fractures. Then, thenonobjective fracture data can be checked.3. The fractured rock mass parameters are characterized by uncertainty, that is, sizeand spatial effects exists in the parameter determination. Only the parameter that canreflect the overall rock mass can be applied to the engineering analysis. Considering theeconomical factors, it is impossible to employ abundant samples with various sizes andzones to obtain parameter values. To solve the abovementioned problem, using3Dfracture network to generate samples can be used for uncertainty analysis.4. Considering size and spatial effects, a number of scanlines should be set for RQDdetermination. Linear representative elemental volume (LREV) can be used to reflect thesize effect of one-dimensional rock mass parameters (e.g., RQD). The analyticexpression of LREV extremely facilitates the determination of size effect. Consideringdifferent zones and parameters, the representative elemental volume (REV) values vary,leading to a poor universality. A universal REV can be obtained by comprehensively considering the spatial effects and the fracture parameters of the rock masses. Thegeneralized REV [cuboid representative elemental volume (CREV)] can fully reflect theanisotropy of rock masses. The analytic expression of CREV generated based on LREVmakes the determination of size effect convenient and economical.5. This paper proposed3methods, i.e., modified stochastic dynamics method,fracture frequency method, and Dijkstra algorithm search method, to search the criticalslip surface. These methods have fully considered the statistical features and locations ofthe fractures, making the determination of the critical slip surfaces more rational.Locations and shapes of the critical slip surfaces determined through theabovementioned methods vary. The features of the surfaces determined by modifiedstochastic dynamics and fracture frequency methods are similar, which are much smallerthan that determined by Dijkstra algorithm search method. In the field engineeringapplication, the modified stochastic dynamics and fracture frequency methods can beused in the original design stage, with larger safety factor and longer support facilities.The application of Dijkstra algorithm search method is opposite.6. The loose accumulation bodies of destruct rock masses constitute loose materialsof debris flows. Then, hazards will be educed by the rainfall. The rock mass structuresinfluence the debris flow characteristics. Susceptibility is influenced by factors of debrisflow gullies, and the influence degrees vary. The combination weighting methodconsiders both the preference of the engineers and the information among the factors;therefore, this method can educe rational weights. Then, the rational susceptibilityresults can be determined. From the point of view of conversation of energy, thedynamics equation, which can fully consider the gully topography, can be established.The dynamics equation facilitates the calculation and the results are universal.