Is the use of Residual Strength Justified in Excavation Support Design?

"Laboratory test data of strain softening soils exhibit the development of a maximum (peak) shear strength followed by a decrease in the strength with continued straining. The rate of strength loss decreases such that an asymptotic minimum value, known as the residual strength, is approached. This paper presents an investigation of the evolution of strength degradation in excavation supported soils that exhibit strain softening behavior. The paper presents a numerical investigation of excavation support of slicken-sided clay, and results of shear testing on materials with similar strain softening behavior. A strain softening constitutive law is presented that can be used with the finite element computer software Plaxis for analysis of excavation support. A typical design section of a deep excavation support in stiff, fissured clay is presented that illustrates the limitations and conservatism of the residual strength approach for excavation support design. Finally, a discussion is presented on the impact of stiffer walls such as diaphragm walls, compared to less stiff systems such as sheet piles, on the design parameters.INTRODUCTIONIn the Washington, D.C. metropolitan area, the Potomac Formation is generally composed of stiff, fissured clays. Geotechnical engineers often associate the presence of slicken-sided planes within the formation as evidence that the soil cannot develop its peak strength. Although it is unclear how old and continuous these slicken-sided planes are, their occurrence, persistence and orientation appear to be somewhat random. Due to this uncertainty, the design of open cut slopes in these materials is often conservatively performed assuming residual shear strength values. This design approach may be justified for long-term open cuts when there is limited amount of geotechnical information and in view of back stability analyses of past slope failures. Furthermore, the lack of confinement in open cuts allows relatively high shear strains near the surface and toe throughout their existence, which promote mobilization of post-peak strength values and the ensuing progressive failure. Indeed, such behavior has been discussed in the past (Skempton, 1970; Schnabel et. al., 1982; Derrenbacher, 1998). A thorough study by Mesri et. al. (2003) of numerous slope failures concludes that the lower bound strength for stability analysis of non-failed slopes is the fully softened strength, which is intermediate between the peak and residual strengths, but that part of the slip surface may be in a residual condition prior to failure."