Analyses For Orientation Errors Of Fractures Using 3D Discrete Fracture Network

Accuracy and reliability of original data are frequently affected due to errorsoccur inevitably whether in field measurement or indoor statistics and analyses fororientation of discontinuities. There are many factors that can lead to those errorsincluding those created by men and nature. It is necessary for geological engineers toconsider sampling errors so as to get data with more reliability.There are 3 sources of errors: measurement error, sampling error and conditionerror. Errors from measurement and condition could be reduced through measuringrepeatedly, improving equipment and changing research method. But for samplingerror, it is chiefly caused by geometrical relation among fractures and the samplingwindow and should be lowered by computing. Terzaghi and Kulatilade studied theerror in 1965 and 1984 respectively using geometry and probability theory. But themethods they provided are not directly perceived through the senses at the sametime, formula derivation process is very complex. So another simple visual method,3D discrete fracture network modeling, is introduced to evaluate the error in thispaper. On the basis of probability and random theory, 3D fracture network modelingtechnology generates 3D discrete network of rock mass with Monte-Carlo method.Shape of each discontinuity is presented by a thin disc, spatial coordinates of its centerare x, y and z. the diameter, dip direction, dip are D, αandβ respectively. Visual 3Dfigures can be generated using the technology after treated by Open GL and sectionsalong any direction can be cut for showing structural feature.A check face is added into the fracture network. Fractures are intersected whenthe check face rotates in the model. Through comparing the mean orientationsgathered from check face with the true, a research result is presented, then, influencefrom the number of fractures intersected with the check face to the statistics result isstudied in the paper.