Building, Updating and Verifying Fracture Models in Real Time for Hard Rock Tunneling

Type of Document Dissertation
Author Decker, Jeramy Bruyn
Author's Email Address jbdecker@vt.edu
URN etd-04242007-160555
Title Building, Updating and Verifying Fracture Models in Real Time for Hard Rock Tunneling
Degree PhD
Department Civil Engineering
Advisory Committee
Advisor Name Title
Matthew Mauldon Committee Chair
Erik Westman Committee Member
Joseph E. Dove Committee Member
Marte Gutierrez Committee Member
William Dershowitz Committee Member
Keywords

* tunnels
* 3d visualization
* database
* mean fracture size
* fracture intensity measures
* Monte Carlo method
* stereology
* fractures
* trace length
* trace density

Date of Defense 2007-04-20
Availability unrestricted
Abstract

Fractures and fracture networks govern the mechanical and fluid flow behavior of rock masses. Tunneling and other rock mechanics applications therefore require the characterization of rock fractures based on geological data. Field investigations produce only a limited amount of data from boreholes, outcrops, cut slopes, and geophysical surveys. In tunneling, the process of excavation creates a priceless opportunity to gather more data during construction. Typically, however, these data are not utilized due to the impedance of sampling and analysis on the flow of construction, and safety concerns with sampling within unlined tunnel sections. However, the use of this additional data would increase the overall safety, quality, and cost savings of tunneling.

This study deals with several aspects of the above, with the goal of creating methods and tools to allow engineers and geologists to gather and analysis fracture data in tunnels without interrupting the excavation and without compromising safety. Distribution-independent trace density and mean trace length estimators are developed using principles of stereology. An optimization technique is developed utilizing Differential Evolution to infer fracture size and shape from trace data obtained on two or more nonparallel sampling planes. A method of producing nearly bias free empirical trace length CDF’s is also introduced. These new methods and tools were validated using Monte Carlo simulations. A field study was conducted in an existing tunnel allowing the above methods and tools to be further validated and tested. A relational database was developed to aid in storage, retrieval, and analysis of field data. Fracture models were built and updated using fracture data from within the tunnel. Utilization of state of the art imaging techniques allowed for remote sampling and analysis, which were enhanced by the use of 3d visualization techniques.

free literature