Geo-Congress 2019: Embankments, Dams, and Slopes

Hydraulic fracturing in dam core can cause excessive seepage, internal erosion, and potentially lead to the total failure of earth dams. Previous hydraulic fracturing studies in soils have focused on fine-grained materials (clays and silts), and little attention has been given to investigating hydraulic fracturing in materials with a significant proportion of sands and gravels. This paper outlines a laboratory testing program that investigates hydraulic fracturing in a widely-graded till material. A new large-scale hydraulic fracture cell (a modified true-triaxial apparatus) was designed and built to test maximum particle sizes in the gravel range. Dam core sampled from an operational hydropower dam in New Zealand was screened into two batches of different maximum particle size and fines content (D_max = 2.36 mm, fines = 36%, and D_max = 9.5 mm, fines 26%), and tested in the fracture cell. It was found that the shape of the fluid injection pressure curves closely matched previous research on finer materials, suggesting the failure mechanism is still hydraulic fracturing in the widely-graded materials tested. As there are many variables that influence the onset of hydraulic fracturing, the influence of the wider grading on the magnitude of fracture initiation pressure was not sufficiently defined. However, this research demonstrates that hydraulic fracturing can occur in existing dams with a high proportion of coarse material in the dam core.

The New Jersey Department of Transportation’s Office of Maritime Resources (NJDOT OMR) team developed an excavation and stockpiling plan to restore sufficient capacity at Gull Island Confined Disposal Facility (CDF) to support the required post-Superstorm Sandy dredging. Site investigation revealed a soft cohesive soil layer and relatively steep exterior slope up to 1.5 H to 1 V. A 4.5 m height stockpile with geotextile reinforcement was used to reach the desired capacity and slope stability, which was monitored by 5 inclinometers during the construction and dredging operations. Following construction protocols established by the OMR team, localized sloughing and seepage of the south dike was identified at the end of the 2016 dredging season, and an emergency slope repair program, using anchored turf reinforcement matting (TRM) and geomembrane, was conducted in early 2017. The repair successfully stabilized the south dike, with the CDF restored in time for the 2017 dredging season.

This paper presents the use of strain-based analysis of Princeville levee to assess the probability of exceeding performance limit states due to the effect of repeated rise and fall of water levels, representing severe storm cycles. To this end, the Princeville levee is selected as a case study and the strain-based analyses are conducted using the finite element program Plaxis 2D. The strain-based approach considers the uncertainty of the permeability of foundation layer and treat them as random variables. Results show a progressive development of plastic shear strain within the levee as the number of storm cycles is increased. The accumulation of such shear strain leads to increasing the probability of exceeding a given performance limit state, and therefore reflect the increasing level of failure risk. In parallel, the stability factor of safety obtained from limit equilibrium approach remains unchanged while it changed slightly for strength reduction method with increased number of water level rise and fall cycles. The conventional slope stability factor of safety approaches, therefore, do not reflect the potential probability of failure sufficiently with repeated loading which negates the ability use it to perform design on the basis of meeting a risk criterion.

Earth dam safety is important to because failure can result in devastating losses of human life and property. As Hurricane María crossed Puerto Rico, the dam spillway of Lago Guajataca in the northwest of the island suffered major damages, caused by overtopping due to extreme precipitation and other factors. After the hurricane, we initiated new surveying of the site with unmanned aerial vehicle (UAV) survey and structure-from-motion (SfM) analysis. The complete temporal analysis includes: SfM and airborne LiDAR datasets from 1930 until 2018. This analysis successfully captured the addition of volume associated with site improvements until the 1980s, approximately 31,000 m³ of volume loss due to erosion triggered by Hurricane María, and the subsequent volume gain of about 17,000 m³ related to emergency repairs in the weeks and months after the atmospheric event. The methods used are remarkably effective to measure and monitor historical and ongoing site changes at earth dams.

Rapid drawdown (RDD) controls the design of the upstream face of many earth dams and can also be important for the waterside slopes of levees. However, unlike dams, levees retain high water for relatively short flood durations and may not reach steady state seepage. Little guidance is available for selecting an appropriate start-of-drawdown phreatic surface that can be used with multi-stage undrained analysis procedures. Rapid drawdown analysis of the Village Creek wastewater treatment plant levee near Fort Worth, TX, is explored in this case study. The levee consists of a blend of clayey sand to fat clay, underlain by layers of mixed alluvium. Three methods are used to select an approximate start-of-drawdown phreatic surface and are compared to a single piezometer measurement from a past flood. Method 1 uses steady state results for water level at about 80% of the peak flood stage. Method 2 uses results directly from transient seepage analyses. Method 3 utilizes a recently developed response surface. The phreatic surfaces predicted by each method are used as the starting point for undrained rapid drawdown analyses. Methods 1 and 3 give similar factors of safety, despite differences in the start-of-drawdown seepage patterns. Method 2 results in a higher factor of safety because it predicts significantly less seepage development during the flood. The results illustrate the importance of the final piezometric surface used in the third stage of the rapid drawdown analysis.

Partially-penetrating wells have been used since the late 1700s for agricultural and construction dewatering and since the 1930s for flood protection structure seepage control. Systems design requires consideration of boundary conditions, foundation characteristics, and well losses to determine the needed number, size, spacing, and penetration of wells to achieve project objectives. Design often involves the estimation of flow to and head reduction from an infinite line of partially-penetrating artesian wells with a line source of seepage. Electrical analogy and physical models with theoretical studies led to the development of a design chart (nomogram) for this purpose by the 1950s. This nomogram has been widely employed and yields uplift factors which can be used to estimate flows and foundation head reductions for a given system geometry. While not as broadly employed for systems design, analytical solutions developed in the 1970s–1980s, and recent three-dimensional computer modeling can also be used to compute uplift factors. This paper offers a review of partial penetration artesian well systems design and provides finite element modeling-derived uplift factors. While isolated errors were identified in a few historical publications, this study further validates prior approaches, and shows that the various methods provide practically similar uplift factor solutions for well systems design.

The U.S. International Boundary and Water Commission (IBWC) discovered cracks and a partial slope failure on a newly refurbished levee section and adjacent floodplain along the Rio Grande River in Brownsville, TX. The partial failure followed a significant drop in water level in April 2014. A geotechnical investigation was performed by the U.S. Army Engineer Research and Development Center (ERDC) to determine what caused the partial levee failure. The site investigation was conducted using cone penetration tests (CPT) and soil borings. Inclinometers and piezometers were installed to monitor the site and establish groundwater conditions. It was found that a series of events, combined with the local geologic conditions, led to the partial slope failure. Events included the 2012 levee construction, fluctuation and rapid drawdown conditions in the Rio Grande River, and a higher elevation of Lake Brown (an oxbow of the Rio Grande) relative to the river. A progressive-type failure mode was identified as the probable mechanism to explain the deformation observed in the field and was confirmed by both seepage and stability analyses.

In recent years, face slab rupture has occurred on several high CFRDs, resulting in large leakage and even potential safety problems. In this paper, a case study on Buxi high CFRD (H=135.8 m) in China is conducted to illustrate mechanism analysis of the face slab rupture by using numerical analysis. It is found that large deformation and differential deformation of rockfill, cavity zones under the face slab, and structural defects of joints are the most important causes of the face slab rupture. Using the monitoring data, dam deformation and leakage were analyzed. This analysis predicts that leakage of the dam after rehabilitation will decrease by at least 80%. Rehabilitation measures are proposed and adopted for repairing the ruptured slabs and joints as well as the damaged waterstops. The leakage of the dam has decreased to less than 20% compared with the value for the same water level before rehabilitation, which is consistent with the pre-remediation prediction. Finally, conclusions are drawn, including suggestions for dealing with the face slab rupture issue of high CFRDs.

Complex deformation processes observed on slow moving landslides are attributed to non-uniform shear surfaces, hydrogeology, rock mass heterogeneity, and uncertainties associated in the determination of the predominant slope failure mechanisms. Effective early warning systems must have a good monitoring system and be designed on sound scientific and mathematical principles. Social, environmental, and economic implications of impending disaster should motivate and determine the accuracy of these systems. Analysis was made of points to consider in designing a monitoring and early warning system for the active landslide on Kariba Dam south bank slope. These considerations included geological conditions, landslide kinematics, and risk scenarios. Displacement and deformation threshold definitions that allow enough time for evacuation of wild life and people that live downstream Kariba Dam are paramount. Potential landslide triggers, that are useful in threshold definition, were identified. The existing monitoring system was modified to take into account strain rates in critical sections of the slope. It will take both a well-designed system and easy to implement evacuation procedure to protect the populace that is susceptible to the imminent disaster.

Ritschard Dam was constructed from August 1993 to September 1995 by the Colorado River Water Conservation District. The dam is 125 feet high zoned embankment with a relatively thin clay core, downstream filters and drains, and rockfill shells composed of terrace gravels, shale, and clay. Changes to the design occurred as a result of conditions encountered during construction resulted in localized variability and reduced stiffness in the rockfill shell material. Since construction, the dam has experienced settlement of approximately 2.3 feet (approximately 1.8% of the dam height) and translation in the downstream direction of approximately 1.0 feet as measured at the dam crest. Deformation modeling was performed using the FLAC finite difference program to simulate the historic deformations of the dam and provide a tool that could be used to evaluate the mechanism for the movement and assess the performance implications of the deformations.

A new segment of U.S. Highway 14 was constructed over highly organic clay near Nicollet, Minnesota, by using prefabricated vertical drains and surcharging in 2016. The highly compressible soils were predicted to settle 3 feet at the maximum fill area over the duration of 2 to 3 years. During the design phase, coefficient of consolidation (c_v) values obtained from 1-D consolidation tests were used to estimate the time-rate of settlement curve. The estimated primary settlement (S_p) of the embankment was about 37 inches after 6 months of surcharging. Nonetheless, a back-analysis using permeability change index (C_k), which has the advantage of representing decrease in permeability due to decrease in void ratio, was performed to model the primary settlement response of the highly organic soils. The computed settlement and pore water pressure show a good agreement with the field data. After the removal of surcharge, the post-surcharge settlement behavior, including rebound and compression, were continuously measured and recorded for more than a year. By using laboratory secondary compression index (C_α) and post-surcharge secondary compression indices (C'_α and C"_α) derived from the field data, the projection of embankment performance over the pavement design life was made.

This case study summarizes an extensive surcharge monitoring program for a project in Juneau, Alaska. The project required working in the inter-tidal zone to facilitate the expansion of the existing uplands area for future development. The geotechnical exploration encountered a deposit of soft, low plasticity clayey silt of variable thickness over a sloping bedrock profile. The preliminary stability analysis performed as part of the geotechnical report determined that the low undrained shear strength was inadequate to support the total new embankment height. The geotechnical recommendations consisted of a temporary surcharge embankment constructed in stages to induce consolidation and increase the strength of the clayey silt prior to construction of the permanent embankment. Due to the thickness of the clayey silt layer, prefabricated vertical drains (PVDs) were installed across the site to expedite consolidation and strength gain, as well as to minimize and dissipate excess pore water pressures. An extensive instrumentation monitoring program was implemented which included inclinometers, piezometers, and settlement plates to monitor and evaluate horizontal and vertical deformations of the embankment and development of excess pore pressure during and after surcharge placement. The monitoring program was developed to ensure that excess pore pressure was minimized in the clayey silt. The results from the monitoring determined that consideration of construction sequencing and fill placement, coupled with drainage from the PVD would have yielded an improved design over the “wish-in-place” alternative. This paper summarizes the lessons learned from geotechnical site conditions, embankment stability analyses, PVD design, instrumentation installation and data reduction, prediction and observation methods for settlement, and general comments regarding the staged construction procedure.

Excessive erosion was observed within the slopes of a 1.3 km long earth embankment supporting an active roadway. The embankment extends up to 15 m above the bottom of an adjacent river. The slopes exceed 1 horizontal to 1 vertical (1 Hz: 1Vl) in some locations and are mostly covered with vegetation. Airborne LiDAR technologies were used to capture the existing topography and generate a digital terrain map which was used to determine changes in elevation and calculate the degree of steepness. The subsurface soils consist of high quantities of silts indicating a high susceptibility to erosion. The current embankment health was assessed by evaluating the stability of the slopes using surface and subsurface information. Several mitigation schemes were developed to increase the health of the embankment and protect the existing roadway. The selected scheme considered multiple aspects including surface drainage, constructability, and cost.

Adverse soft ground conditions are encountered in a recent project in South America, in which a large embankment up to 25 m high is being constructed over very soft forest floor consisting of peaty soil and soft clay extending as much as 10m below the ground surface. To tackle the immense challenges, a battery of geotechnical monitoring instruments was deployed during the ground treatment and the staged embankment construction to provide continuous monitoring of the response of the foundation soils under the fill load increments as well as the deformations of the embankment. This paper presents some selected findings and lessons learned from the analysis of the instrumentation monitoring data leading up to a slide incident in which part of the embankment under construction was massively cracked. This case demonstrates that, with a well-planned instrumentation scheme, dedicated construction observation effort, and careful assessment of the monitoring data, the observational method can be successfully applied to assess the deformation and stability of embankment on soft ground, help identifying potential construction risks, and guide the construction activities accordingly.

Excessive erosion and bank failure within river corridors pose a significant risk to infrastructure and natural systems and reliable characterization of bank erosion is a critical component when designing mitigation strategies and conducting post-restoration monitoring. Advances in unmanned aircraft system (UAS) and digital photogrammetry techniques provide a powerful new tool for capturing high resolution topographic data at less cost. This case study focuses on a 1-km section of the New Haven River in central Vermont, located in the northeastern United States. The New Haven River reach has exhibited instability due to its sensitive landscape position and history of watershed and channel disturbances. A fixed-wing UAS was deployed repeatedly over a two-year period (2016–2018). Digital elevation models (DEMs) and DEMs of difference quantified the volumetric change along the survey area where notable erosion and deposition occurred. Results showed the UAS to be capable of producing high quality topographic data at fine resolutions and an effective method for monitoring continued bank erosion.

Fiber-reinforcement has been proven to be an effective technique for stabilizing new or existing slopes against shallow failures. For sandy soil, the main benefit of fiber reinforcement is the increase in the friction angle. Such an increase will theoretically allow for designing slopes that are steeper or higher for any given design factor of safety. The friction angle of fiber-reinforced sand can be predicted from the properties of the unreinforced sand and the properties of the individual fibers using available discrete or energy-based empirical models. It has been proven that these model predictions suffer from a given level of model uncertainty. As a result, the predicted friction angle of fiber-reinforced sand may be more uncertain than that of unreinforced sand. In this study, the uncertainty in the friction angle of fiber-reinforced sand is quantified by combining traditional spatial variability with model uncertainty. The combined uncertainty is then incorporated in a reliability-based design methodology that is tailored for assessing the risk of failure of fiber-reinforced infinite slopes. The results indicate that the design factor of safety against slope stability failure may have to be increased in fiber-reinforced infinite slopes to maintain the same level of reliability as traditional non-reinforced slopes. The required factor of safety is affected by the choice of the model used to predict the fiber-reinforced friction angle and by the anticipated degree of spatial variability in the friction angle of the unreinforced soil.

Methods of stability analyses of slopes, such as limit analysis or the limit equilibrium method, require postulation of the pattern of failure and the mode of material yielding. They typically do so by assuming the shear-type material failure (mode) along kinematic discontinuities (pattern). Observations of slope failures, however, lead to the conclusion that both soils and rocks often display a rupture-type mode of failure with large separation components of displacements, in addition to the shear-type failure. Such modes are observed, for example, as opening cracks at the crest in early stages of slope failure and as blocks separating from one another, giving rise to toppling failures. The shear strength of rocks is typically represented by a non-linear strength envelope, whereas soils are more commonly described with a linear function of normal stress. The role of the strength envelope and the plastic flow rule will be discussed in the context of rupture modes in slope failures. Examples of slope analyses with geomaterial rupture will be shown with a focus on the safety assessment of slopes. It will be demonstrated that neglecting rupture modes in stability analyses may contribute to overestimating the slope safety.

During construction of a 40 to 65 foot high, 3,000 foot long, 1H:2V reinforced steep slope it exhibited signs of global instability. At full height, additional movement of several feet occurred over a larger area and scarps formed. Installation of toe berms and shear bars by the RSS designer were only partially effective in stabilizing the slope because of underlying colluvial materials. The primary entrance roadway for the shopping center and a fire force main located at the top of slope were at risk. Analysis of the as-built RSS found inadequate factors of safety. In situ soil parameters in unstable areas were back-calculated and a concept for repair was developed. To arrest additional movements, installation of tie-back anchors extending from the face of the slope terminating 30 feet or more into sound rock was proposed. Actual anchor lengths varied from 103 to 220 feet and difficult drilling conditions were encountered in several areas. During construction of the tie-back anchors portions of the toe berm became unstable, threatening a major sign for the shopping center. Some anchors found poor rock conditions, requiring changes to install longer tendons. Re-evaluation of overall stability of the RSS was needed in one area with especially low tie-back proof load test results. Challenges of designing and constructing a tie-back repair on the face of an actively moving steep slope will be shared.

Rainfall-induced landslides constitute a major risk globally to the built environment. Practically all traditional slope stability analyses that include rainfall influences do so by varying the groundwater flow patterns with increasing pressure heads and often a rising groundwater table. Rainfall infiltration inside soil slopes result in changes to the hydrologic behavior of the soil. Critical conditions within the hydrologic behavior correspond to failure of slopes. A better understanding of the influences of the hydrologic behavior on landslide mechanics is therefore of significant geotechnical focus. This study investigated hydrologic behavior by performing an unsaturated slope stability analysis for several published case studies. This study used a fully coupled hydro-mechanical analysis with time-dependent boundary conditions to evaluate the case data. Although the hydrologic behavior was described by the wetting and drying soil-water characteristic curve (SWCC) behavior, the wetting SWCC was estimated using only the drying data. A factor of safety analysis for the case histories was performed at their final stages. The findings from this research show the wetting SWCC parameters have greater influence on the hydrologic behavior of a slope during rainfall, when compared to drying SWCC parameters.

Understanding the behavior and quantifying the impact force of a granular sliding mass are crucial to landslide hazard assessment and designing protective structures against landslide debris. The study of a granular sliding mass provides useful insight into the characteristics of sudden onset, rapid granular flow. There are several factors, such as inclination angle, basal friction between the granular flow and the flow bed, particle size and shape, and distance to a rigid obstruction, which influence the impact force and behavior of granular flow. In order to systematically investigate the effects of these factors on the impact force from granular flow, a two-dimensional sand flume was designed and constructed. This sand flume included a load measurement device consisting of six dynamic load cells, which served as the rigid obstruction. In this study, the effects of particle characteristics on the impact force were studied by using two different granular materials with mean particle sizes of 0.25 and 1.98 mm. Impact forces and flow front velocities were measured and compared. The results indicate that the different basal friction angles for the two materials result in different flow velocities and final static forces. Moreover, the different flow velocities, densities, depths, and stiffness result in different dynamic force components.

Infiltration-induced landslides are common geological hazards in the world. When they occur along highways they threaten public safety and private property, block highway traffic, and damage transportation infrastructure. Since the hydrology of the site affects the effective stress distribution and consequently the stability of the slope, monitoring the dynamic changes in pore water due to climatic changes is important. This work presents the study of the hydrological behavior of an active landslide on an embankment on Interstate-70 west of the Eisenhower/Johnson Memorial Tunnels in Summit County, Colorado; where seasonal movement of more than 1.5 m has been recorded during the past decades. At this site, the failure surface was identified to be within the range of groundwater table locations. Information from previous reports on the site was combined with data obtained from three additional boreholes to generate a soil profile. Three piezometers were installed near the crest and toe of the landslide and monitored hourly for over four years. A unique hydrological behavior was observed; the groundwater level beneath the westbound shoulder of I-70 changes twice as much as the groundwater level beneath the eastbound shoulder, with the two locations being only 30 m apart. A conceptual model that captures the principal mechanisms at the site is proposed. Observations of the seasonal embankment movement are related to the location of groundwater table, and confirmed with slope stability analysis using an extended modified Bishop’s method that accounts for suction stress in unsaturated soils.

Many slope failures could have been predicted well in advance by carefully observing warning signs and subtle site conditions. Summaries of five case histories of major slope failures and missed predictions are presented. The failed slopes consisted of: (1) slope failure of a levee around a major wastewater treatment plant during construction of a slurry wall. The significance of a thin organic-silt seam sloping adversely into the slurry wall and a very high water-head in the adjacent sludge landfill were not recognized during design; (2) Failure of a slope impacting a chemical storage structure at a water treatment plant. Obvious signs of an ancient landslide at the site and an adversely sloping weak seam were not detected during the original geotechnical exploration; (3) An extensive landslide in a large residential subdivision impacting more than 300 residents and damaging 27 homes. Historically known anomalies in the geologic formation, slickensides, and the potential for fully-softened and residual strengths were not recognized during design of significant fill and ground supported retaining walls; (4) A massive landslide on colluvial soils at a lakeside development. The very apparent presence of colluvial soils on which embankments were constructed and natural seepage paths that were blocked off by embankment fill were not considered during design of the development; and (5) A deep slope failure along a bentonite seam in the shale formation beneath a major roadway embankment adjacent to a lake. The 25-year history of this location included many preliminary signs of a deep slope failure, but these signs went unheeded until major movement was initiated following a period of heavy rainfall.

In many parts of Texas, particularly North Texas, pavements are constructed on highly plastic expansive clay subgrades with fill embankment slopes. Desiccation cracks often appear in these pavements due to cyclic swelling and shrinkage of the expansive clayey subgrade. These cracks offer a suitable gateway for rainwater infiltration which ultimately leads to shallow slope failure. The objective of the current paper is to propose a two-step mechanism to combat rainfall-induced shallow slope failure at U.S. Highway 287 located in Midlothian, Texas. The two-step mechanism includes the installation of recycled plastic pins for lateral stabilization of the slope followed by the installation of a modified moisture barrier for preventing the infiltration of rainwater through desiccation cracks. To evaluate and compare the effectiveness of the proposed two-step mechanism, three sections were installed in July 2017: the control section, the section with recycled plastic pins only, and the section with modified moisture barrier and recycled plastic pins. The performance of these sections was monitored using moisture sensors installed at different depths and an inclinometer installed at each section. Moreover, to evaluate the moisture variation at each of these sections, resistivity imaging (RI) was conducted on a monthly basis. Based on the analysis of preliminary results, the slope section with the modified moisture barrier showed insignificant moisture variation in subgrade soil while, for the section without the modified moisture barrier, the moisture variation ranged between 10% and 35%. In addition, a lateral deformation of 8.23 mm was observed in the slope section with recycled plastic pins and modified moisture barrier while a lateral deformation of 25.25 mm was observed in the section without recycled plastic pins and modified moisture barrier.

Active slope movement was identified during the site investigation phase for the construction of a new fly ash transfer facility at a power plant in West Virginia. The site is located near the base of a large slope and possesses a complex geology. The soils at the site consist of colluvium overlying alluvium deposited adjacent to the Ohio River. The slope above the project site exceeds 75 meters. Instrumentation installed during the site investigation phase identified active slope movement. Construction of the facility required re-grading the site and constructing a 9-meter cut at the base of the hillside. This was made possible through the construction of an anchored soldier pile and lagging retaining wall. The wall was designed to withstand the active landslide. Its performance was assessed with instrumentation that was monitored before, during, and after construction over a period of two years. The data collected suggests that the wall is supporting the landslide load with minimal deflection.

The M_w 7.8 2015 Gorkha earthquake that occurred on April 25, 2015, caused a death toll of over 9,000 people and a property damage of over U.S.$10B, over 40% of the Nepalese GDP. Majority of the damages include structural failures of buildings, damages to roads, landslides, and a few liquefaction cases. One among the well discussed infrastructural damages that occurred after the earthquake sequence was a cumulative of 19 cm uniform settlement of a hydropower dam structure which was under construction during the earthquake shaking. This case study is pertinent to that hydropower dam settlement. The case study project is the largest hydropower project in the country and the settlement of dam not only delayed the construction progress, but also brought up various questions pertinent to the standard practices used for the design of dam foundations in the country. The dam was constructed on over 100 m thick sediment deposited about a kilometer upstream of a landslide dam that was formed over 500 years ago. Due to the unavailability of any ground motion stations near the dam site, ground motion data published by the USGS was used to calculate seismically induced settlement of the foundation material during the earthquake using different available methods. Although different methods used in this study resulted in different amounts of settlement calculations, analyses results by three methods showed that the calculated settlement close to the observed values in the field. This study includes the systematic evaluation of foundation settlement under the seismic conditions during the earthquake and a discussion on the lessons learned from the dam settlement case history after the 2015 Gorkha earthquake. The case study paper will be helpful in future for evaluating potential settlement of floating dams during earthquakes.

Time-dependent deformation is a major concern for the design and operation of rockfill dams. Meanwhile, the scaling between the stress-strain behavior of the reduced specimens for laboratory tests and the prototype rockfills for dam constructions has not been appropriately emphasized nor yet to be addressed. The recently developed rate-dependent breakage mechanics model is integrated here with the fracture-mechanics based scaling law to predict the long-term deformation of rockfill dams. The predictive capacity of this new method is evaluated through a comprehensive case study on a 233 m concrete face rockfill dam (CFRD), the Shuibuya CFRD. We show that the proposed model can predict reasonably well the settlement response across a long-time span from construction to operation stages. The horizontal displacement, stress contours, and breakage distributions are also presented for analyzing the global response of the Shuibuya CFRD and the evolution of its internal gradations.

A highway bridge spanning an at-grade railroad right-of-way was designed using two MSE wall supported abutments. The earthwork included the soil approach embankment and the MSE wall sand backfill. Steel H-piles were inserted through the backfill in corrugated metal pipes and then driven to the desired depth to support the bridge abutments. Before construction of the MSE walls, rammed aggregate piers (RAPs) were installed beneath each MSE wall and backfill to reduce total and differential settlements and increase bearing capacity of the weak foundation soils. The west MSE wall and approach embankment did not perform well and had to be demolished and re-constructed but the east wall performed well. Field settlement and slope inclinometer data show that the west MSE wall moved laterally due to the weak foundation soils. This paper discusses the cause of the lateral wall movements and presents recommendations for the design of future MSE walls founded on RAPs.

In the past two decades major landslides devastated the town of La Conchita, California. The January 2005 landslide was a debris flow that killed 10 people, and damaged 36 residential homes. The 2005 landslide mobilized the south flank of a larger landslide that occurred in March 1995, and destroyed 9 homes. La Conchita is an important case history because the 2005 landslide was largely a consequence of decisions taken after the 1995 failure. It shows the importance of good engineering recommendations (including slope winterization) and the central role of owners and government agencies. During the last three decades, the author has made detailed observations of the slopes at La Conchita, performed subsurface and laboratory investigations, and conducted numerical modeling and conventional slope stability analyses whose principal aim was to understand the conditions and failure mechanisms of the 1995 and 2005 slope instabilities. This article discusses the history of the site, and remedial schemes (e.g., winterization, grading, and dewatering) to reduce the landslide hazards to the town of La Conchita.

In this study, the mechanical behavior of Shirin-Dare Dam was evaluated at the end of construction, initial impounding, and operation time through using the results of instruments and numerical analysis conducted by FLAC software. The results showed that the settlements calculated from the numerical analysis were in reasonable agreement with the data recorded by the instruments in the field. This confirms the validity of the material properties selected based on field instrumentation. The numerical results pertaining to the calculations of the pore water pressure distributions, embankments stresses, and the settlement behaviour of the dam showed acceptable performance of the studied dam. The sensitivity analysis of hydraulic fracturing of the dam using the arch ratio resulted in comparable performance of similar geostructures across the world.

The design and construction of an 800-m segment of an embankment over nonhomogeneous soils, including a layer of organic clay, is presented in this paper. The embankment was constructed over 13,670 controlled modulus columns that were designed to transfer loads to deeper, more competent soil layers. The combined length of the columns was about 210 km. The length of individual columns varied from 8.8 m to 22.8 m. A geosynthetic-reinforced load transfer platform was designed to aid in transferring the embankment loads to the columns. Settlements of 30 cm were detected during embankment construction. A correlation was observed between the extreme settlements and the locations of culverts. Efficacy of the column supporting system was estimated through back calculations using the settlement measured data. Analysis and observations of the settlements’ distribution led to the conclusion that the major cause of low efficacy was suppressed arching in the fill layer above the columns in the vicinity of culverts. Remedial action included reducing the loads through replacement of natural fill with lightweight expanded polystyrene blocks, and drilled shafts in one location. After the project conclusion, pavement settlements between 8 and 14 mm were observed within eight months, and no distress was reported in the following two years.