![]()
ABSTRACT: Initial attempts to include the effect of joints in analysis of unlined tunnel sections inc? first 2-d and then 3 d ubiquitous joint analysis applied to established stress formulas, e.g. the Kirsch solution and subsequently to the results of plane strain finite element analysis Subsequently, the finite element analysis was improved to handle joint elements and these were informed with opening/closing and shear non-linear constraints. Finally, we developed block theory which more naturally accommodates the three dimensional aspects of structural geological data. The original assignmere to analyze the Pile Driver test drifts at the Nevada Test Site, based on joint trace mappings, would have been better served with the block theory approach.
INTRODUCTION The organivers of this session asked for elucidation of problem areas in rock engineering in which we can now perform better than in years past thanks to the development of rock mechanics. The example I know best from personal experience has to do with the interaction of tunnels and joints. In the early 1960's, the Protective Structures Branch of the Army Corps of Engineers, Omaha District, was trying to develop design methods to create safe underground bases. After the problems they had encountered in construction of the NORAD underground complex in Cheyenne Mountain, above Colorado Springs, they were particularly motivated to find methods to include the action of faults and other discontinuities of rock. In this context, I was brought on as a consultant with the specific directive to propose and conduct studies to this effect in connection with the design and execution of a large test of alternative tunnel designs at the Nevada Test Site, code named "Pile Driver". Figure 1 shows the configuration of the Pile Driver Project, in which alternative design concepts for blast resistant tunnels were constructed and instrumented at several range distances from a nuclear detonation in granitic rock.
The Protective Structures Branch opted to include a sixteen foot diameter, unlined section supported by rock-bolts and mesh within the Pile Driver complex, to be placed in "D-drift". Other sections, in A, B, and C drifts had steel or steel and concrete liners of varying thickness. In D-drift, on the other hand, the jointed rock could be mapped both before and after the blast. It seemed a wonderful opportunity for innovation2 To start, the geological staff of the Omaha District, under Harold Jack, had used sheet metal drawing techniques to produce a complete series of unrolled joint trace maps of all the exposed rock surfaces in the Pile Driver excavations (as well as the NORAD excavation drifts). Figure 2 is the sheet prepared for the rock-bolted galleries, DL1 and DL2. Trace maps like this had been prepared for portions of the Snowy Mountain power house excavations, for some Bureau of Reclamation foundations (e.g. Grand Cotflee Dam), and perhaps elsewhere; but I had never seen complete logs of the fractures in a series of complex underground openings drawn with such care and completeness. Carl Distefano, the structural designer for this drift,. was equally impressed; but also perplexed. He suspected these data were meaningful but had no idea how to use them so he asked me. Like president Eisenhower commenting on Nixoh's achievements as vice-president, I asked for time to think about it. It is now more than thirty years later and I can tell you about the struggles to find a solution, which came after much treading water and far too late to help Mr. Distefano.
Site News