A suite of shear tests was performed on a set of scale model monoliths to investigate the influence of macroscopic, three-dimensional, foundation roughness on shear strength. Two of the monoliths were designated “critical” blocks, and sheared over the model foundation under a variety of lateral boundary conditions, with the remaining monoliths providing lateral confinement. The lateral boundary conditions included: 1. No lateral constraint; 2. One-sided lateral constraint; 3. Loose two-sided constraint and; 4. Rigid two-sided constraint. The resulting differences in configurational shear strengths, closely related to the unconstrained sliding trajectories and parameterized by the “effective” friction angle of each block, are used to highlight the reinforcing action of lateral confinement over asymmetric sliding topography. It is shown that the opposing unconstrained trajectories of the critical monoliths promote the formation of a kinematic wedge, whose effect on the overall stability of the model dam is evaluated numerically.
The observed tendency of a strongly asymmetric foundation surface to produce lateral, in addition to vertical, movement in a sliding body ([1-3]) is of great importance to dam engineering. A shearing, vertically dilating monolith has only free space to resist its motion in the vertical direction. On the other hand, a laterally dilating monolith is forced to interact with its neighbors while sliding forward; the altered trajectory and sideforces resulting from this interaction lead to additional frictional resistance and serve to increase the effective friction angle1, much as two-dimensional roughness increases the apparent friction angle of a symmetric block [4]. Traditional limit equilibrium-based design did not consider this interdependent mechanism, and made no positive assumptions about the potentially strengthening consequences of lateral confinement [5].
As a result, inherited expectations about the behavior of an existing structure are often based on the laterally unconstrained shear properties of its constituent monoliths.
In 2004, the proposed retrofit of a federally operated gravity dam in California (originally constructed in the mid-twentieth century) necessitated a thorough evaluation of that structure’s as-built shear behavior [5]. Among the issues that arose during facility review was the presence of an incompletely excavated channel fault footwall underneath several monoliths on the left side of the dam (Figure 1). The potentially adverse impact of this feature was balanced by the discovery that other factors not considered in the original friction-angle calculations could actually prove beneficial to overall structural stability. Among these factors was the strong topographic asymmetry exhibited by the rock-concrete interface.
The shear tests discussed in this paper were performed as part of a wider effort to determine whether the in situ effective friction angles of the channel fault monoliths were indeed higher than their design values. While the alteration-plated surface is thought to underlie several monoliths, this investigation focuses specifically on the behavior of monoliths M15 and M16, whose positions and foundation profiles identify them as the critical structural elements. The laterally unconstrained, one-side constrained, and two-side constrained shearing of model monoliths M15 and M16 was performed using a modular testing apparatus originally designed for investigating end-effects [1].
