Uncertainty analysis in the discrete fracture network generation In order to reduce the risk of simulation based on probability distribution functions: A case study of Emamzadeh Hashem tunnel

Rock mass consists of intact rocks and discontinuities such as fractures, which have a significant impact on the mechanical and hydraulic properties of the rock mass. For example, in sensitive analyzes such as tunnel stability simulation. Therefore, it is very important to determine the exact parameters of stone jointing, such as orientation and length. Using Discrete Fracture Network is one of the common methods to simulate jointed rock. Since the construction of the fracture network is done using statistical distribution functions, the uncertainty in the construction of the fracture network has always been one of the great challenges of engineers. In this study, the fracture network of of Emamzadeh Hashem tunnel was constructed by a code developed in MATLAB and 3DEC software, and the influence of statistical distribution functions on the uncertainty in the construction of the fracture network was investigated. The results indicate that a negative exponential distribution can lead to relatively large errors in constructing the fracture network, especially when used to generate the fracture dip direction.. Also, the results of the parametric study showed that the use of statistical distribution functions that have the variance of the data in their PDF (Probability Distribution Function) can increase the accuracy in the production of fracture parameters such as slope, slope direction and effect length, and To reduce the uncertainty in the production of the fracture network.