Stability of steep slopes in cemented sands

Brian D. Collins; Nicholas Sitar

doi:10.1061/(asce)gt.1943-5606.0000396

Home/Archive/USGS

USGS · 70236351

Stability of steep slopes in cemented sands

Brian D. Collins·Nicholas Sitar

View original source ↗Export metadata ↓

Abstract

The analysis of steep slope and cliff stability in variably cemented sands poses a significant practical challenge as routine analyses tend to underestimate the actually observed stability of existing slopes. The presented research evaluates how the degree of cementation controls the evolution of steep sand slopes and shows that the detailed slope geometry is important in determining the characteristics of the failure mode, which in turn, guide the selection of an appropriate stability analysis method. Detailed slope-profile cross sections derived from terrestrial lidar surveying of otherwise inaccessible cemented sand cliffs are used to investigate failure modes in weakly cemented [unconfined compressive strength ( UCS ) < 30   kPa "> ( UCS ) < 30 kPa (UCS)<30 kPa ] and moderately cemented ( 30 < UCS < 400   kPa ) "> ( 30 < UCS < 400 kPa ) (30<UCS<400 kPa) sands and their role in the evolution of the geometry of the slopes. The results show that high-resolution slope topography, such as can be obtained with terrestrial lidar, is essential for identifying altogether different failure modes in weakly cemented (shear-mode) and moderately cemented (tensile-mode) sand slopes. Analyses show that the standard Culmann method for steep slopes is inappropriate for modeling the stability of cemented sand slopes since it tends to overpredict expected crest retreat and underestimate failure plane angle. Instead, a simplified analysis using infinite slope assumptions, but applied to a slope with finite dimensions subject to changing geometric conditions, such as toe erosion and slope steepening, is suggested for analysis of weakly cemented sand slopes. For moderately cemented sand slopes, a limit equilibrium analysis directly comparing the cliff tensile stress and cemented sand tensile strength is shown to reasonably predict failure conditions and timing as a result of either slope steepening or tensile strength loss, presumably from wetting in most cases.

Explore related subjects

Journal of Geotechnical and Geoenvironmental Engineering Article

Keep this discovery

Explore connections, maps & timelines

BibTeX RIS

Brian D. Collins, Nicholas Sitar. 2011. Stability of steep slopes in cemented sands. https://doi.org/10.1061/(asce)gt.1943-5606.0000396

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related USGS reports

USGS2026

Discrete element investigation of the influence of shallow soil density on the manifestations of strike-slip surface fault rupture

This study investigates the influence of soil relative density on strike-slip surface fault rupture manifestations using three-dimensional numerical simulations performed with the discrete element method (DEM). The simulations capture the formation of distinctive fault strands within complex flower structures using tens of millions of grains. The tendency for dense soils to localize shear manifests as multiple localized shear bands within wide zones of deformation in strike-slip fault rupture, whereas diffuse shear deformation develops within narrow zones in loose soils. The spatial extents of soil deformation are consistent between simulations having similar relative densities but with different quantities of grains. However, individual shears are more distinguishable in assemblages of finer grains than in assemblages of coarser grains. The simulations show the progressive development of new shears within the bounds of previously developed shears. Shear activity transitions inward as fault activity diminishes along the outermost shears and continues along newly developed shears until a vertically dipping throughgoing shear structure develops that accommodates most of the fault displacement thereon. The throughgoing fault develops at smaller fault displacements in looser soils because the first shear rupture propagates closer to the vertical direction and does not undergo as much inward translation of shear activity, as is observed in denser soils. In all simulations, ground surface uplift develops between nonintersecting active shears, and ground surface subsidence tends to develop where new shears intersect previous shears. The surface traces in these simulations are shown to be consistent with analog models and case histories of surface fault rupture occurring in different shallow subsurface materials. Although computationally costly, these modeling results are valuable for providing a strong numerical supplement to traditional analog models used to represent the mechanics of strike-slip zones in soil, and they provide quantifiable stresses and large-strain deformations throughout the model domain.

Journal of Geotechnical and Geoenvironmental Engin↗

USGS2024

Framework for mapping liquefaction hazard–Targeted design ground motions

Liquefaction-induced ground failure poses substantial challenges to geotechnical earthquake engineering design. Current approaches for designing against liquefaction hazards, as specified in most seismic provisions, focus on estimating a liquefaction factor of safety ( 𝐹⁢𝑆𝐿 ) and typically characterize earthquake loading using design parameters based on probabilistic or deterministic ground motion levels. Because 𝐹⁢𝑆𝐿 is estimated deterministically, this basis of design neglects considerable uncertainties for estimating liquefaction triggering and its consequences and results in a lack of liquefaction-specific design criteria, particularly as structural design has advanced toward risk-targeted performance objectives. This study presents a framework for developing liquefaction-targeted design criteria based on a minimum acceptable return period of liquefaction, informed by probabilistic liquefaction hazard analysis (PLHA). PLHA quantifies annualized rates of liquefaction by considering contributions from (1) the full ground-motion probability space, and (2) uncertainties in liquefaction triggering using probabilistic models. PLHA is used in this study to characterize the current, effective return periods of 𝐹⁢𝑆𝐿 ( 𝑇𝑅,𝐹⁢𝑆 ) obtained from conventional liquefaction hazard analysis (CLHA) using uniform-hazard ground motions. 𝑇𝑅,𝐹⁢𝑆 is evaluated in a parametric study of nearly 100 sites throughout the conterminous United States. The results indicate large geographic variations in acceptable liquefaction hazard levels, with implied 𝑇𝑅,𝐹⁢𝑆 ranging between approximately 1,000 to 3,000 years. To address these inconsistencies without the computational demands of full PLHA, a framework is proposed for developing a liquefaction-targeted design peak ground acceleration, 𝑃⁢𝐺⁢𝐴𝐿 , for use in liquefaction models that result in consistent liquefaction design levels across all geographic locations. The mapped 𝑃⁢𝐺⁢𝐴𝐿 is shown to be somewhat sensitive to site-specific properties, and adjustment factors are developed and presented. The proposed 𝑃⁢𝐺⁢𝐴𝐿 mapping procedure produces 𝐹⁢𝑆𝐿 estimates that are consistent with those obtained from full PLHA at a target 𝑇𝑅,𝐹⁢𝑆 , providing a promising roadmap to incorporating PLHA concepts into current liquefaction design methods.

Journal of Geotechnical and Geoenvironmental Engin↗

USGS2019

Effects of infiltration characteristics on the spatial-temporal evolution of stability of an interstate highway embankment

Infiltration-induced landslides are among the most common natural disasters threatening modern civilization, but conventional methods for studying the triggering mechanisms and predicting the occurrence of these slides are limited by incomplete consideration of underlying physical processes and the lack of precision inherent in limit-equilibrium analyses. To address this problem the spatial-temporal evolution of failure is investigated in a seasonally unstable section of interstate highway embankment, known as the Straight Creek landslide, Colorado. The study includes multi-year site investigation, monitoring, and numerical simulation using a rigorous hydromechanical framework along with a field of local factor of safety method. The sensitivity of episodic landslide reactivation to infiltration characteristics is evaluated. Results indicate that annual cumulative snowmelt infiltration, which typically accounts for approximately 75% of total annual cumulative infiltration and occurs over a short period in the spring, has the most substantial impact on slide activation. The rate of snowmelt infiltration varies independently of annual cumulative snowmelt infiltration and cumulative infiltration in the previous year, but still affects antecedent soil moisture conditions at the onset of snowmelt infiltration and therefore also the level of slide activation. These findings are used to establish specific thresholds for exacerbated slide movement using annual snowpack accumulation, forecasted snowmelt rate, and the previous year’s snowmelt, an approach which may be applied for predicting movement at this and other recurring or potential slide sites.

Colorado↗