Study On Optimization Of 3D Joint Network Simulation In Rock Mass And Its Application In Jijicao Block, Beishan, Gansu Province

It is important to quantitatively analyze and effectively describe the structural condition of rock mass. It is a valid way for determining the mechanical, hydraulic and engineering characteristics of rock mass. The study of the 3D joint network simulation in rock mass and its application is reviewed in detail. The paper bases on the site investigation of Jijicao block, Beishan, Gansu province, the pre-selected area of Chinese High Level Radioactive Waste. Followed the route from investigation of rock mass joints to the analysis of joint characteristics, and to 3D joint network simulation in the rock mass, the optimization of 3D joint network simulation, the characteristics of the rock mass joints and the properties of the site in Jijicao block are studied.A great deal of joint data is obtained from the joint survey and the borehole logging in Jijicao block. Attention is focused on the probability analysis of the joints geometry characteristics and the optimization of 3D joint network simulation in rock mass principally. The detail of my research is listed as following:(1) Study on the circular sampling window methodBy simulating the joint survey on outcrops, the Monte Carlo method is employed to determine the size and position of the circular window, and used to study the relationship of the width of the rectangular window and the mean joint trace length. Appropriate suggestions about the circular sampling window method and the joints survey are present.(2) Partition of the structure homogeneities in rock massIn this paper, the circular sampling window method is employed to carry out to estimate the mean joint trace length and joint trace midpoint density around the both sides of the faults. According to the interference range of two faults,Jijicao block is accurately divided into three regions where the structure of rock mass is homogenous and two regions where are interferenced by the two faults.(3) Classification of the joint orientationsImproved fuzzy clustering analysis by amending degree of membership is applied to classifying the joint orientations instead of the traditional analysis. Based on the improved fuzzy clustering algorithm, the program can deal with a large quantity of joint orientations. And degree of membership of some isolated joint orientation is modified to reduce their contribution for the results of fuzzy clustering analysis. The study shows that the results of fuzzy clustering analysis are reliable and reasonable. And the dominant orientations and classification are more precise.(4) Study on the joint sizeBased on the Monte Carlo method, the relationship of the included joint trace length and the total joint trace length is studied quantitatively by simulating the joint survey on outcrops. Furthermore, the joint size which presents the extension of the joint plane in space is involved. For the joint Poisson disc model, the joint size can be represented by the joint diameter. Warburton's method is right theoretically, but its practicability is not good for using the joint probabilistic distribution. In this paper, a new method for estimating the joint diameter distribution is suggested without using the joint probabilistic distribution.(5) Study on the joint densityWith the joint survey on outcrops, we introduce the concept of general RQD which is used with different thresholds to determine anisotropy of rock mass. By means of analyzing the variation of the joint line density in borehole, a new method for estimating the REV size of rock mass is proposed. The characteristics of the rock mass and the formation process of the joint compact district are involved through the analysis of the joint spacing. And the formation mechanism of the primary joint is described. By the weighted average method, joint trace midpoint densities of the dominant joint sets are obtained. And based on the space geometric relationship, the calculation formula of joint volume density is deduced and joint volume densities of the dominant joint sets are obtained. Some new optimistically methods in some key techniques are used to make 3D joint network simulation in the rock mass more exact, and more beneficial to its application. 3D joint network simulation in the rock mass is a valid tool, which is applied in evaluation of rock mass quality, evaluation of site property, migration of nuclide and stability of surround rock in tunnels.