While empirical design methods typically produce general preliminary recommendations to cover a wide variety of rockmass behaviour (within a given rock quality range), a mechanistic approach considers specific failure mechanisms and adjusts design accordingly. Hoek and Brown (1980), Hoek et al. (1995), Brady and Brown (1993), and others give detailed treatment to many of these mechanisms and to the appropriate support strategies. A modest selection is covered here.
Stress Induced Boundary Crushing
Hoek et al. (1995) describe a methodology for analysis of excavation induced stresses for the purpose of support design. Two-dimensional or three-dimensional elastic analysis may be used to evaluate the induced stresses around complex excavation shapes. These stresses can be compared to an appropriate strength criteria to determine the extent of rockmass failure (Sections 2.13 and 2.15).
In hard and brittle rockmasses, the maximum compressive stress can be contoured around an excavation boundary (where the minor or least compressive stress approaches zero) and compared with the uniaxial compressive strength of the rock. Where the calculated compressive stress exceeds one-half (Section 2.13.3) of the strength determined from testing of laboratory samples, it can be assumed that, in the long term, the rockmass will become significantly damaged and may require support. This method of analysis is most appropriate in highly stressed, hard, brittle rockmasses with moderate to low initial fracturing or structure. Failed rock in these environments is unlikely to possess much residual strength and will require full support after the creation of the failed zone. For regions within one excavation radius, compare the induced stress difference (- 1) to 0.5 times the laboratory UCS. Areas where the stress difference exceeds this 3 value may be prone to damage and eventual weakening and rupture.
In softer and more fractured rockmasses, the use of a confinement dependent criteria such as Hoek-Brown (Hoek et al., 1995) is warranted. Plastic analysis may be used to assess the potential for progressive failure and to investigate the role of stress redistribution and self-stabilization. Once a zone of potential failure has been established from such models, it may prudent to simply design cablebolt support to sustain the deadload of this failed zone (Factor of Safety or Strength Factor <1). Often this will prove economically impractical and may not be necessary, since failed rock often retains limited ability to support itself. If plastic analysis has been employed, the support must extend into the zone of confinement or the zone above the back (e.g. in a roof support problem) where the minor (least compressive) stress begins to increase. This indicates the development of self-stability. The location of this boundary will, however, be highly dependent on the strength and deformation parameters used in the analysis.
It should be noted in any case that cablebolts are unlikely to arrest the onset of rock failure under high stress and may do little to alter the progression of such failure into the rockmass. The objective here is to hold the failed material in place so that the broken rock itself can generate the necessary onfinement to reduce the extent of progressive damage and instability. In highly plastic (deformable) rockmasses under high stress, it is also unlikely that cables will be effective in arresting the progression of failure. In addition, in these environments, the induced displacements may be too great for the system to handle and cable strand rupture may be inevitable in pre-installed systems.