jrc

Accurate simulation of geometrical properties of fractures is an important goal in rock engineering. One of the most capable methods for simulating the random nature of geometrical properties of fractures is Discrete Fracture Network (DFN) random modelling, which presents the heterogeneous nature of fractured rock mass with statistically defined geometrical properties. Up to now, all properties of fractures such as location, shape, orientation, size (persistence), spacing, and opening of joints have been simulated and applied in 3D DFN modelling. In this research, a statistical solution based on Kernel’s non-parametric distribution is used for simulating roughness. Through this method, even those geometric properties of fractures which do not have their own specific distribution functions can be simulated. After simulating the roughness value, the roughness geometry should also be simulated in a way that evokes the roughness value. Therefore, in order to simulate the surface of fractures in this research, the DRS method is applied in 2D and then, developed into 3D. At the end, simulation of discontinuity’s roughness is added as a separate package to DFN-FRAC3D computer program. DFN-FRAC3D computer program, as one of the most capable tools in this field, is able to develop a 3D fracture network block model by using the surveyed data and then simulating geometrical properties of the fracture; thus, by applying the results of this research in this compute software, all geometrical properties of fractures can be simulated. Finally, in order to explain the results of this research, outcomes of DFN-FRAC3D computer program for both with and without applying the roughness property on DFN are compared.

Summary:

One of the important geometric features of rock mass discontinuities is its surface roughness. The discontinuity level has different heights that differentiate each level from the other. In practice, it is not possible to assign a roughness value to each of these levels, so to levels that are similar in height; The same amount of roughness is assigned. In analyzing the problems of stone mechanics related to coercion, it is necessary to choose one of the many levels of discontinuities that have the same amount of roughness. Discontinuity Roughness Simulation (DRS) is a random selection of different levels of surface roughness for a certain amount of surface roughness. In this method, the level of discontinuity is simplified by connecting several pieces together; So that each piece has an unevenness. By simplifying and using different layouts of components that make up the discontinuity surface, the DRS method can produce different levels, the JRC value of which is almost the same.

Several methods have been proposed to measure the roughness feature of rock mass, all of them work based on the elevation of its fracture's measurement. Precise measurements and illustration of roughness values were the subjects of the previous studies.

Methodology and Approaches:

Barton's field method and determination of the maximum distance, a, between fracture and the field survey profile, L, is the foundation of the proposed method here. So fracture's length would be divided into n pieces in which its length is less than L and there exist an asperity with the height of a. By considering a number between 0 and an as the flange's height and a random number in the [0,l] interval as the distance between two adjacent flanges, locus of the fracture's flange points would be determined. The locus of the fracture's atrio points would be determined by measuring the distance between two adjacent flange and atrio points of a fracture which is a number between 0 and two adjacent flanges.

In order to validate and ensure the accuracy of the results obtained from the application of the DRS method, the validity, and reliability of the method have been checked for all possible JRCs. The results of the validity test for different JRCs indicated that the value of the mean percentage of absolute error (MAPE) for different parts of a simulated fracture is always less than 10 percent and this means that if a fracture needs to be simulated using the DRS method in order to produce a roughness value of JRC = 10, the JRC value for all parts of the simulated fractures would a number be between 9 and 11. The reliability of the DRS method for different JRCs has also been investigated. Outcomes showed that the value of MAPE in 100 of repetition of DRS method implementation for a part of the simulated fracture that is randomly determined is less than 6 percent, which is an acceptable value and a confirmation of the accuracy of the DRS method to simulate and produce fracture’s roughness.

In studies of mechanical and hydro-mechanical behavior of rock joints, it is important to consider the joint surface geometry. Statistic methods are used for quantification of rock joint surface geometry. In these methods the geometry of joint surface are quantified by Centre Line Average height (CLA), Root Mean-Square of height (RMS), Auto Correlation Function (ACF), RMS of the first derivative of profile (Z2) and Relative Length (RL). In most of previous studies, one parameter (such as Z2 or RL) was used for quantification of surface roughness. In this study, the efficiency of statistical parameters for rock joint quantification is evaluated. With this regard, Barton’s profiles and some profiles with regular geometry are used. Each profile is digitized and statistical parameters are computed. It is revealed that in order to study the rock joints behavior, it isn’t appropriate to use one or some limited number of statistical parameters.