Journal of Geotechnical and Geoenvironmental Engineering

Forum papers are thought-provoking opinion pieces or essays founded in fact, sometimes containing speculation, on a civil engineering topic of general interest and relevance to the readership of the journal. The views expressed in this Forum article do not necessarily reflect the views of ASCE or the Editorial Board of the journal.

Based upon large-scale model tests conducted at Nanjing Hydraulic Research Institute, China, the surface erosion and intermittent mass failure along a dam’s axis and the formation of headcut and its migration in the longitudinal section were determined as the key breaching mechanisms for a cohesive dam due to overtopping. In this work, based on the breach mechanism, a numerical model has been developed to simulate the overtopping process of cohesive dams. A comparison among the performances of three large-scale model tests shows that the proposed model exhibits the best overall performance among the three selected physically based dam breach models. The sensitivity studies show that the three models are all sensitive to soil erodibility, although overall the proposed model and Windows Dam Analysis Modules (WinDAM) B are more sensitive than the National Weather Service (NWS) BREACH model. In addition, it is demonstrated that the proposed model performs better and provides more detailed results than the three selected parametric models.

A reinforced soil foundation (RSF) consists of layers of geosynthetic reinforcement and compacted granular fill material. The RSF approach is a fast, sustainable, and economical alternative to shallow foundation design. This paper presents the development of a prediction equation for estimating the settlement of footings placed on reinforced soil. The parameters that are considered in the prediction equation include footing geometry (width and length), soil friction angle and cohesion, reinforcement characteristics (stiffness, spacing, length, and number of reinforcement layers), and applied static loads from 50 to 600 kPa. For the prediction equation development, a parametric study was first conducted using a validated finite difference numerical model. The results of the parametric study were then used to conduct a regression analysis to develop the prediction equation for estimating the maximum settlement of RSF. The equation was validated using three case studies. The developed prediction equation will be useful for practitioners in preliminary RSF design.

Shear-wave velocity, $V_s$, is a mechanical geotechnical parameter required for the dynamic response of geomaterials. This property can be advantageously assessed both in the field and in the laboratory. Although field tests may offer the most precise methods to obtain $V_s$ of a certain soil, they, however, do not permit conditions different from those encountered in the field to be readily investigated. Several laboratory techniques of $V_s$ measurement have been developed, most notably the resonant column and the piezoelectric bender elements. Although the latter is widely used, it leads to a number of difficulties that cannot be denied, such as uncertainties in first arrival’s detection, near-field effects, and mixed radiation of primary and shear waves. In an ongoing effort to minimize/eliminate these difficulties, the geotechnical group at the Université de Sherbrooke (UdeS) developed the piezoelectric ring-actuator technique (P-RAT), which can be easily incorporated into traditional geotechnical apparatus, an advantage that facilitates its utilization in other geotechnical laboratories as a powerful tool in $V_s$ measurement. This paper summarizes the results of parallel tests of P-RAT installed in typical oedometer cells to measure the $V_s$ of soft sensitive clay samples extracted from two different sites in Québec, Canada. These tests were carried out at two different institutes, the UdeS and the École de technologie supérieure (ÉTS). The main purpose was to examine the reliability of the P-RAT test results by applying the same test procedures to similar test materials. The results show that the change in the used sensors, input signals, data acquisition system, and technical operators between the two laboratories has no practical effect on the measured $V_s$ values of all tested samples, which confirms the robustness of the technique and promotes its incorporation in other geotechnical apparatus and laboratories. Results obtained at both institutions were also used to present consolidation curves of the tested soft clays in terms of their shear-wave velocities and establish unique correlations between the stress-normalized shear-wave velocities, $V_{s1}$, with the overconsolidation (OCR) and void ($e$) ratios for each site investigated.

Concentrated leak erosion occurs in transverse cracks in dams when the reservoir rises above the base of the cracks. As water flows through the crack, the sides of the crack will erode if the hydraulic shear stress exceeds the critical shear stress of the soil, and the crack will widen. For cracks above the phreatic surface, the sides of the partially saturated soil in the sides of the cracks will swell with increasing moisture content, narrowing the crack. These two phenomena occur concurrently. Laboratory erosion and swell tests have been carried out on five soils which are representative of the soils used in the core of embankment dams. Using this data and the equations controlling the rate of erosion, crack width versus time has been modelled. This shows that for the soils tested the crack will close for cracks up to about 5 mm wide for a flow gradient of 0.1, and 1 mm to 2 mm wide for a flow gradient of 0.4. For wider cracks the erosion will progress.

Design for the tensile strength limit state of steel strip reinforcement in mechanically stabilized earth (MSE) walls must provide for loss of steel thickness due to corrosion. The paper combined key features of zinc and steel corrosion models found in the literature with corrosion rate statistics for buried galvanized steel strips. Examples of loss of strength of reinforcement strips of different thickness for up to 120 years after construction were presented in probabilistic terms. Results of probabilistic analysis for the tensile strength limit state at a design life of 75 years were demonstrated for a typical wall using three different models to compute the maximum tensile load in a strip under operational conditions. The same three models were used to demonstrate a methodology for load and resistance factor design (LRFD) calibration for the tensile strength limit state for steel reinforcement strips. The calculations showed that the accuracy of the load model with respect to measured loads has a strong influence on the magnitude of computed resistance factor for a prescribed load factor and maximum target level of reliability.

To evaluate the effect of plasticity characteristics and mineralogical composition on postcyclic shear strength degradation, 18 soils, prepared as different proportions of kaolinite, montmorillonite, and ground quartz, were characterized using static simple shear and cyclic simple shear tests with postcyclic monotonic loading. The postcyclic undrained strength ratio ($s_{u,pc}/{\sigma }_c^{\prime }$) was found to depend on the clay mineral, plasticity index (PI), and cyclic stress ratio. In soils with kaolinite as the clay mineral, a greater reduction in shear strength due to cyclic loading was noted at lower PIs, whereas soils with montmorillonite as the clay mineral typically experienced reductions between 0% and 30% of the static undrained shear strength. A relationship was established between degradation in undrained shear strength and postcyclic effective stress ratio (PC-ESR), which is the ratio of the consolidation stress to the effective vertical stress after cyclic loading and is equivalent to the reciprocal of the difference between the pore pressure ratio and one. A linear relationship between the ratio of $s_{u,pc}/{\sigma }_{pc}^{\prime }$ (where ${\sigma }_{pc}^{\prime }$ is the effective vertical stress immediately after cyclic loading) and the undrained strength ratio ($s_u/{\sigma }_c^{\prime }$) of a normally consolidated soil and PC-ESR was also developed. Results from nine natural soils showed good agreement with the relationships developed for mineral mixtures.

Microbial-induced calcium carbonate precipitation (MICP) was experimentally implemented on two coal ash materials (i.e., CA1 and CA3) to investigate the efficacy of the treatment on the hydraulic and compressibility properties of the material. Oedometers used in testing were modified to enable the injection of treatment solutions and to provide the ability to monitor shear wave velocity. The coal ash specimens were treated to reach different levels of shear wave velocity, which corresponded to varying amounts of ${\mathrm{CaCO}}_3$ content. The hydraulic conductivity and shear wave velocity values were assessed throughout the loading process. Results indicate that the MICP treatment decreased the compressibility (compression index decreased by a factor as high as 2), and the hydraulic conductivity (decreased by a factor as much as 0.05). The slope of $\log ({\sigma }_{\mathrm{V}}^{\prime })\text{-}{\mathrm{V}}_{\mathrm{S}}$ relationship (${\beta }^{\prime }$) decreased from 0.31 for the untreated material to 0.01 with treatment. In parallel, the increase in the shear wave velocity after treatment increased the intercept of the relationship between $\log ({\sigma }_{\mathrm{V}}^{\prime })\text{-}{\mathrm{V}}_{\mathrm{S}}$ (i.e., ${\alpha }^{\prime }$) from 38 to $383\mathrm{m}/\mathrm{s}$. The correlation between small-strain and large-strain parameters are explored and discussed.

Because of its great length, one subway terminal excavation in Shanghai, China, was divided into two pits for sequential construction. Although it featured an adequate earth-supporting system and the field construction followed proper procedure, wall and ground displacements of the first excavated pit were up to 1.8–2.7 times local specifications. Then, basal strata of the second pit were jet grouted. Unfortunately, worse excavation performance was tracked. Technical review of design plans, construction logs, and field data, and postconstruction investigations disclosed that for a long and narrow deep excavation bottoming out in soft clays, reinforcing basal strata prior to excavation is essential and attention should be paid to the adverse effect arising from great wall exposure length, $L_e$, during excavation. Although both basal treatment and $L_e$ are of decisive importance, lack of suitable basal treatment imposed more adverse effect on excavation performance when $L_e>4H_e$, where $H_e$ = final excavation depth.

This case study presents an investigation on building settlement induced by dewatering for a deep excavation constructed in a deposit in Tianjin, China. The excavation site is located in the Tianjin urban center, around which there are many existing buildings. Geological conditions at the site involve a multiaquifer system with one phreatic aquifer over three confined aquifers. To maintain the stability of the bottom of the excavation, a cutoff wall was used to block groundwater seepage with large diameter wells to lower the groundwater level. Field instrumentation was installed during construction to monitor (1) the quantity of withdrawn groundwater, (2) the groundwater head, and (3) the settlement of surrounding buildings. The field observations yielded the following results and implications: (1) pumping of groundwater could lead to large building settlement and a large area of ground settlement; (2) most of the settlement caused by dewatering did not rebound; (3) the cutoff wall played a significant role in reducing the building settlement during dewatering; and (4) the dewatering-on-demand procedure should be implemented during excavation.

The influence of spatial variation in undrained shear strength is an important practical consideration for assessing the stability of excavations or natural slopes. Prior research has focused on slope stability in normally consolidated or lightly overconsolidated clays, commonly assuming that undrained shear strength ($c_u$) increases linearly with depth but remains constant in the horizontal direction. However, depositional history, geologic settings, and variations in groundwater conditions may result in a strength profile that varies vertically and horizontally. A parametric study was performed to determine slope stability considering undrained shear strength that increases linearly with depth but exhibits an initial undrained shear strength at the ground surface that varies over horizontal distance—a relationship defined geometrically by an angle called the $c_u$ profile inclination ($\theta$). The analyzed solutions are presented through charts of normalized stability numbers ($N_s$). The results demonstrate that increasing the shear strength gradient increases slope stability, but considerable differences in stability were found between $c_u$ profile inclinations. The importance of the $c_u$ profile inclination is significantly influenced by the observed failure and slope geometry.