IFCEE 2018: Developments in Earth Retention, Support Systems, and Tunneling

Urban cofferdams are used to build the rigid central core of high-rise buildings and are conceived as temporary retaining excavation structures. For the case of the One Museum Park West (OMPW) building (Chicago, IL), settlement points and inclinometer data were recorded during the excavation within the cofferdam. This cofferdam stage represented 20% of the total excavated volume of this project and caused about a third of the total measured settlements. An axisymmetric numerical model using an advanced constitutive soil model, the hypoplasticity model for clays with intergranular strains, is developed to simulate an idealized lower bound solution of the ground movements resulting from the OMPW cofferdam excavation. Significant differences between computed and observed ground movements are presented when compliance effects between structural members, installation of nearby foundations, and flexibility of the cofferdam and internal bracing connections are ignored.

The performance of a circular cofferdam braced with reinforced concrete ring beams is presented in this paper. The cofferdams were built using a continuous circular wall, made of interlocked sheet pile modular pieces, internally braced with ring beams. During concrete placement of the exposed ring beams, curing conditions were affected by temperatures below the freezing point of water which resulted in lower concrete strength and stiffness values than those estimated during the design phase. A numerical model is presented to study the influence on the performance of creep, shrinkage, temperature, and aging of concrete. Lateral wall displacements are computed with the model only as a result of coupled temperature and concrete time-dependent effects and studied individually with a parametric study. Those effects contributed about 17% of the total maximum measured lateral wall deformations. Creep effects had the largest contribution with about 8%, while aging of concrete and shrinkage contributed less than 5%.

Over the past few years, we have gathered results of 150 sacrificial ground anchor tests with about 25% of them tested to pullout failure. This paper reviews the predicted capacities based on the available geotechnical data and load test results from the actual sites. This comparison develops a correlation between bond stress in different soil types to the N₆₀ value, which can be used to confirm bond stress estimation. Additional conclusions were drawn based on differences in installation methods, for example solid stem auger installation versus air rotary installation. The ultimate capacities of these anchors were compared to see how the capacities differed or if they were similar. Finally, we compared the variability in load test results performed on different anchors at the same site. We compared how deflection readings at the same load can vary as well as how the ultimate capacity of anchors can vary at a single site.

Geotechnical ground anchors have been used for more than half century. They have been further developed constantly in the course of time and have found their place in standards as a special foundation engineering system. Many significant structures could not have been accomplished or only with substantially more difficulty without ground anchors. However, there is a concern that tensioned tieback anchors installed under city streets or adjacent properties will be detrimental to future development such as utilities, tunnels or buildings. Running of TBM drives, installing sheet pile walls or placing drilled or driven piles, into tieback zones will be difficult and costly. For this reason, some municipalities have requirements that these anchors to be de-tensioned and removed once permanent structural wall is in place. In U.S., cities with these requirements are Los Angeles, Beverly Hills, San Diego in California, Phoenix, Tucson, Tempe, in Arizona, and few in Northeast. In Europe there are cities in Germany, Austria, and England. The essential removable strand anchor systems available on the market are described with emphasis of fully removable strand anchor systems in the form of compression type anchor.

Predicting the performance of tiedback soldier beam shoring walls is one of the challenges in geotechnical engineering. This paper presents the results of advanced nonlinear numerical modeling of a full-scale instrumented shoring wall. The wall was constructed in granular materials, and included one- and two-tier tiedback solider beams with wood-lagging in between them. The goals of the numerical modeling were to accurately predict the bending moments in the soldier beams, the lateral movements of the exposed height of the wall, and the ground settlements behind the wall. Staged excavation analyses were performed to more accurately include the effects of the construction sequence. The tiebacks were modeled as pre-stressed springs to account for the lock-off loads and to capture increased tieback resistance relative to wall movements. The numerical model was able to predict behavior of wall and soil settlements behind it. However, the model was unable to predict soldier beam settlements, and the data indicates that soldier beam settlement significantly impacted lateral movements and ground settlements.

Soil nailing is a popular reinforcement method for stabilizing slopes, excavations, and retaining walls in many countries. The pullout capacity of the soil-nail interface is a critical parameter for the design and safety assessment of soil-nailed systems. In this paper, a three-dimensional (3D) finite element (FE) model is developed using the commercial program ABAQUS to simulate the pullout test of a soil nail in a soil-nail pullout box. The stress-strain behavior of the soil and soil-nail interface is described by a mohr-coulomb model and coulomb friction model respectively. A novel temperature-based reduction of friction coefficient has been applied in order to reduce the roughness of soil-nail interface as the slip occurs. Validation of the FE model is carried out by comparing the results of numerical simulations with experimental data from a previous work. The comparisons between the modelling results and experimental data have demonstrated that the established FE model can simulate the pullout behavior of a soil nail in a soil mass properly. Based on this, the verified FE model has been used to investigate the effect of different nail diameters, soil elasticity moduli and pullout rates on the pullout behavior of the soil nail.

Excessive ground movements are a major concern in urban excavations because they can lead to significant displacements and rotations in adjacent structures. One of the key elements to add stiffness to an excavation support system is the retaining wall. However, it is conventionally designed based on limit equilibrium. The internal wall shear forces and bending moments are estimated employing apparent earth pressure diagrams and the internal hinge or tributary area methods. Although these approaches will prevent structural failure of the support wall, they may result in excessive wall deformations and ground movements. This paper presents an empirical method to estimate lateral wall deformation profiles and internal bending moment along a retaining wall in clays. The method is based on inclinometer data collected from 30 case histories worldwide and allows the designer, having the maximum lateral wall displacement and the height of the retaining wall, to predict the shape of lateral wall deformations for deep excavations based on soil type (i.e., based on the undrained shear strength parameter). The internal bending moment along the wall is derived from the fundamental relations of mechanics of materials and the classical theory of beams that relate the bending moment to the components of translation of the member. It assumes that the wall material is linear elastic.

This paper presents the results of a finite element (FE) numerical analysis that was developed to simulate the fully-instrumented geosynthetic reinforced soil integrated bridge system (GRS-IBS) at the Maree Michel Bridge in Louisiana. Four different loading conditions were considered in this paper to evaluate the performance of the GRS-IBS abutment due to dead loading, tandem axle truck loading, service loading, and abnormal loading. The two-dimensional FE computer program PLAXIS 2D 2016 was selected to model the GRS-IBS abutment. The hardening soil model proposed by Schanz et al. (1999) was used to simulate the granular backfill materials; a linear-elastic model with Mohr-Coulomb frictional criterion was used to simulate the interface between the geosynthetic and backfill material. Both the geosynthetic and the facing block were modeled using linear elastic model. The Mohr-Coulomb constitutive model was used to simulate the foundation soil. The FE numerical results were compared with the field measurements of monitoring program, in which a good agreement was obtained between the FE numerical results and the field measurements. The range of maximum reinforcement strain was between 0.4 and 1.5%, depending on the location of the reinforcement layer and the loading condition. The maximum lateral deformation at the face was between 2 and 9 mm (0.08–0.4% lateral strain), depending on the loading condition. The axial reinforcement force predicted by FHWA (Adams et al. 2011) design methods were 1.5–2.5 times higher than those predicted by the FE analysis and the field measurements, depending on the loading condition and reinforcement location. However, the interface shear strength between the reinforcement and the backfill materials predicted by Mohr-Coulomb method was very close to those predicted by the FE.

Pile foundations for bridges must often resist lateral loads produced by earthquakes and thermal variation. Because of the lack of large scale tests, there is almost no guidance for engineers in assessing the lateral resistance of piles located close to these MSE walls. The objective of this study was to assess lateral pile resistance as a function of distance from the wall face and predict force induced in the reinforcements. Full-scale lateral load tests were conducted on 24 piles located at 2, 3, 4, and 5 pile diameters behind 4.6 and 6-m high MSE walls. P-multipliers were developed to account for the decrease in lateral soil resistance. Little reduction occurred for piles located more than about 4 pile diameters from the wall. Measured tensile forces in the MSE reinforcements increased as the load on the pile increased and decreased rapidly with transverse distance from the direction of loading. Multiple linear regression equations were developed to predict maximum tensile force induced in reinforcing elements by lateral loading.

In this study, a numerical investigation was conducted to study the settlement of GRS piers under applied axial loads. A finite-difference program was used to model full-scale GRS piers. The backfill soil is simulated using the plastic hardening model combined with strain-softening behavior. The developed model is validated against the results of full-scale GRS pier performance tests. The numerical results were compared with the observed pier settlement data, and it was found that the model adequately captures the behavior of GRS piers under axial loading. After validation, the influences of reinforcement strength, reinforcement vertical spacing, pier height, and pier cross-sectional dimensions were investigated through a parametric study. Results indicated that increasing the reinforcement strength and decreasing its spacing have a significant benefit in reducing the settlement of GRS piers.

A reinforced soil foundation (RSF) consists of closely-spaced layers of geosynthetic reinforcement and compacted granular fill material used to support bridge piers and abutments. The RSF approach is a fast and economical alternative to traditional shallow foundations. This paper presents a prediction model for estimating settlement of a footing placed on an RSF under service loads. The parameters that are considered in the prediction model include footing geometry (width and length), soil friction angle and cohesion, reinforcement characteristics (stiffness, spacing, and length), and applied static loads from 50 to 600 kPa. In order to develop this prediction model, a parametric study was conducted using a validated finite difference numerical model. The results of the parametric study were used to conduct a regression analysis to derive a prediction model for estimating the maximum settlement of foundation placed on reinforced soil. Such a prediction model will be useful for practitioners in preliminary RSF design.

Fly ash is widely accepted as a construction fill material. Utilization of fly ash as a backfill material in mechanically stabilized earth (MSE) walls is explored in the present study. The common modes of failure in MSE walls include (1) pullout of reinforcement, and (2) sliding of backfill material along the surface of the reinforcement. Interfacial shear tests are typically performed in a modified shear box to study the sliding mode of failure. In this study, ash from two different thermal power plants located in India- Neyveli Thermal Power Plant and Ramagundam Thermal Power Plant- are used to study the interfacial shear properties between the fly ash material and the geogrid type of reinforcement. The interface properties of fly ash-geogrid are compared with that of conventional fill material-geogrid. The post peak loss of strengths is found to be less in the case of geogrid reinforced fly ash indicating the change in behavior of the fill material from brittle to ductile. The interfacial shear parameters, adhesion and interface friction angle, are found to range from 13.6-to-25.6 kPa and 33.8-to-45.5°, respectively. The interaction coefficient of fly ash and geogrid ranges between 0.87 and 0.92. The interface properties provided in the study may be adopted in the design of geogrid reinforced fly ash structures.

Application of lightweight cellular concrete (LCC) materials are increasingly popular in major projects due to their multiple benefits including significantly low surcharge loads. While implementing agencies have shown their interest in using such materials in backfills of mechanically stabilized earth (MSE) retaining walls, the investigation is limited to the determination of the static properties. However, with the increased use of LCC retaining walls in seismically active regions, it is necessary to understand the seismic performance of these structures. In this study, a 1.8 × 1.2 × 1.2 m block of LCC retaining wall with unit weight of approximately 4 kN/m³, reinforced at the mid-height with a geo-grid layer, was subjected to seismic loading of different amplitudes and frequencies on a shake table to evaluate the response of the wall and geo-grid to seismic loading. Numerous accelerometers and strain gauges were installed at different heights to evaluate the deformation of the LCC block at different depths and relative to the geo-grid. To evaluate the effect of surcharge load and simulate different layers, the blocks were separately loaded seismically at the surcharge stresses equivalent to the surcharge load of 3 and 3.7 m, i.e., a total retaining wall height of 4.2 and 4.9 m, respectively. Specifically, the LCC retaining wall was subjected to a series of sinusoidal cyclic loads with amplitudes ranging from 0.1 to 0.2 g and frequencies of 2 and 3 Hz. Additionally, the ground motions recorded from the 1994 Northridge Earthquake were also applied to the model. Observations of the model during the shake table testing did not indicate the formation of any fractures within the LCC retaining wall. The results obtained from accelerometers and strain gauges indicated that the LCC retaining wall moved monolithically, meaning no observable differences were noted in the displacements of the LCC material relative to the geo-grid.

In this investigation, fragility analysis of a cantilever retaining wall with a height of 6-m and cohesionless and cohesive backfills is conducted to evaluate the probability of failure due to earthquake peak ground acceleration (PGA). Calibrated seismic analyses of retaining walls are conducted using the finite difference method. Seven different earthquake ground motions are used to assess the seismic behavior of retaining walls under different earthquake conditions. The maximum relative displacement of retaining walls was determined for models with various earthquake loads as well as different backfill materials. Consequently, fragility analyses are conducted to determine the effect of earthquake intensity in terms of PGA and backfill cohesion on the failure risk of the retaining wall. Fragility results are provided based on probability curves demonstrating the likelihood of different levels of damage in a given earthquake PGA for various backfill cohesions. According to the fragility analysis, with consideration of 50% as the probability of failure, the threshold PGA free-field of the failure zone is 0.47 and 0.75 g, respectively.

Although geosynthetic-reinforced soil (GRS) bridge abutments have been used widely in highway infrastructure projects, understanding their response during earthquake loading is a remaining issue that is affecting their implementation throughout the U.S. In particular, the three-dimensional (3D) seismic response of GRS bridge abutments is critical to consider as the direction of shaking in the field may be uncertain and because experimental work has shown that it is possible to have horizontal deformations in directions other than the primary direction of shaking. Further, any horizontal deformations of the GRS bridge abutment may be linked with settlements of the bridge seat, so 3D deformations are important to understand. This study involves 3D numerical simulations of the seismic response of a hypothetical GRS bridge abutment experiencing one-dimensional horizontal shaking in the longitudinal direction of the bridge beam, focusing on the lateral deformations of the GRS bridge abutment in the longitudinal and the transverse directions. Time histories of the horizontal deformations in both directions are presented and synthesized with settlements of the bridge seat to evaluate the applicability of using the Federal Highway Administration (FHWA) design method linking vertical and lateral static deformations by assuming zero-volume change in the GRS mass to seismic loading.

Numerous geotechnical structures may be subjected to low-frequency cyclic movements over their design lives. A prime example is the case of integral bridge abutments, which have been shown to expand and contract into and away from the retained soil due to seasonal temperature changes. The stresses generated within the retained soil mass for these structures have been reported to increase with increasing number of cycles, a process that has been referred to as earth pressure ratcheting. To further investigate the extent of this increase in lateral earth pressure over time, a laboratory experiment was conducted involving a moving wall. The wall was repetitively displaced at pre-defined amplitudes into and away from the retained soil mass in a rotational mode. One sidewall of the container holding the soil was constructed to be transparent to allow for visualization of the shear bands that develop during passive and active stress states. The experimental results were used to calibrate a finite element model suitable to extend the performance evaluation considering a wider range of parameters. The comparison between measured and numerically predicted results indicates that lateral earth pressure and total thrust acting upon the wall can be accurately estimated at specified wall displacements.

Understanding the magnitude and distribution of earthquake induced soil pressures is an important concern in designing retaining walls and underground structures. Seismic soil pressures on flexible and yielding structures are mainly analyzed through limit-state methods. To develop better understanding of these methods and their accuracy, a series of large-scale shake table tests was performed. A unique system of vertical underground structures was placed in an 8 m-diameter laminar soil box and embedded in a two-layered soil strata. The system was subjected to sinusoidal motions and scaled ground motion records of the Kobe 1995 earthquake. Apparent soil pressure records for different acceleration levels were analyzed and compared with classical and recently developed limit-state methods for both cohesionless and cohesive (c-φ) soil layers. Results indicate that all limit-state methods under-predicted seismic pressures at small to moderate accelerations (<0.4 g). This paper will provide a brief review of limit state methods in literature followed by an overview of the experimental program, the data analysis, and a discussion of the suitability of existing methods in predicting pressures on flexible underground structures.

Most traditional research on buried culverts has looked at live load distribution through soil onto buried culverts without due attention to the dynamic amplification of moving loads. The few studies that looked at the dynamic amplification of buried culverts idealized the load-soil-culvert problem using a plane-strain assumption. Under such conditions, the finite area axle loads are erroneously modeled as strip loads acting over the entire width of the culvert. In this study, the plane-strain assumption is removed and the load-soil-culvert system is modeled and analyzed as a three-dimensional problem. Dynamic load allowance (DAF) was determined from field measurements, two-dimensional finite element method (2-D FEM), and three-dimensional finite element method (3-D FEM). The 2-D and 3-D FEM models resulted in average DAFs of 1.10 and 1.03, respectively. The average DAF calculated from field measurements and the American Association of State Highway and Transportation Officials (AASHTO) recommended formula are 0.973 and 1.30, respectively. Overall, the AASHTO DAFs are the highest and the field DAFs are the lowest. The plain strain assumption adopted in the 2-D FEM models resulted in DAF values that are higher than the 3-D FEM and field evaluated DAFs. The DAF calculated from three-dimensional models is the closest to the field measured DAFs.

Construction of the Anacostia River Tunnel in Washington, DC required the tunnel boring machine to pass beneath several utilities that were sensitive to movement as well as critical to the regional infrastructure. The utilities included a 1.2-m (48-in.) diameter water line, a 2.7-m (108-in.) diameter force main, a 0.8-m (30-in.) diameter water line and a 1.1-m (42-in.) diameter water line, all with varying degrees of redundancy. A compensation grouting program was designed, tested and implemented to reduce the risk of settlement of these utilities during tunneling. Challenges with this scope of work included site locations with access constraints, complex subsurface conditions, and very sensitive utilities with varying foundation systems. Compensation grouting parameters were established through an effective trial program and the grouting was then implemented in conjunction with an extensive monitoring program. Potential settlements of the utilities during tunneling of up to 13 mm (0.5 in.) were successfully mitigated by the compensation grouting program. This paper will discuss details of the trial and production grouting programs implemented during this critical portion of the Anacostia River Tunnel construction.

This paper presents methodology and tools for use in evaluation of potential hydraulic jacking in the concrete-lined section of a pressurized headrace tunnel immediately adjacent to a steel-lined tunnel section. The locations of potential in situ stress deficiency in the rock mass are uncertain but could be located in the transition section. Therefore, the performance of reinforced concrete-lining in the transition zone and elsewhere are evaluated for hydraulic jacking. The evaluation of hydraulic jacking was performed by using finite element method (FEM) and discrete element method (DEM). The computer programs ABAQUS and UDEC were used for FEM and DEM analyses, respectively. In addition to the analytical approach, a review of the available approaches focusing on the performance of concrete-lined pressure tunnels was performed. The response of the tunnel system including the lining and surrounding rock mass was evaluated for various scenarios. The evaluation includes: (1) extent of hydrojacking; (2) exfiltration from the tunnels; and (3) structural stability of the tunnel system. A series of sensitivity analyses were performed using numerical modeling to parametrically evaluate the influence of rock mass joint variations on the results.

Soil-bentonite (SB) slurry trench cutoff walls are widely used for subsurface containment of contaminants and groundwater control. In many cases, particularly for environmental containment applications, the walls are intended to serve as a permanent barrier with a life expectancy measured in decades. The compatibility of these barriers with the environment is considered during most design studies but a comprehensive review of the longevity of these walls is absent from the literature. This paper identifies key longevity issues, summarizes published literature to date and describes the current practice related to the longevity of soil-bentonite slurry trench cutoff walls. Principle issues potentially affecting long-term performance include compatibility between the backfill and the groundwater, impact of wetting/drying in the zone of the fluctuating water table, long-term property changes due to secondary compression of the backfill, potential for desiccation of the near-surface zone of the barrier wall and the potential for hydraulic fracturing due to the relatively low state of effective stress. The paper concludes with the authors’ summary opinions and suggestions for future research.

The hydraulic press-in piling method, which only generates extremely low vibration, works very efficiently in low headroom and hard soil conditions because of the way the system is designed. This paper will present two case studies on emergency foundation repair for railway bridge piers; one in Ohio in the U.S. and the other in Fukuoka, Japan; both of which successfully applied the method.

Two alternative methods were evaluated to measure bentonite content in soil-bentonite mixtures: (1) the water vapor sorption (WVS) method; and (2) the electrical conductivity (EC) method. The WVS method is a new technique developed for this study; the EC method was developed previously and modified in this study. Soil-bentonite mixtures blended with a pug-mill to have 5–6% bentonite content were sampled from three different locations of an engineered barrier. Samples of the base soil used for the barrier and the bentonite were also obtained. Calibration curves were created in the laboratory using base soil and bentonite collected from the field to determine the bentonite contents of samples of the field mixture. Bentonite contents determined from the field-mixed samples were compared to bentonite contents computed from scale records maintained during construction of the engineered barrier. Mixtures prepared in the laboratory and mixtures sampled in the field were treated chemically and physically to eliminate the influence of the initial distribution of bound cations on the test outcome and to amplify the sensitivity of the testing methods to bentonite content. The WVS and EC methods have linear correlations with bentonite content and can be used to determine the bentonite content of field mixtures accurately. However, physical and chemical treatments must be applied for accurate determination of bentonite content.

Earthen levees are critical civil engineering infrastructures in the U.S. Many of them are beyond their design life and it is important to evaluate their safety and reliability for mitigation of any potential hazards. Traditionally, the stability of the earthen levees has been investigated using the deterministic analysis in terms of factor of safety. Recently there is a trend to utilize probabilistic methods for assessing the stability of earthen levee problems since significant uncertainties exist in the earthen levee analysis. Typically in a geotechnical engineering project only limited field investigation is conducted. The input parameters for stability analysis of earthen levees can be highly uncertain. To explicitly consider the uncertainties, this paper develops a framework for probabilistic assessment for the safety of earthen levees using the fuzzy finite element method. The strength reduction method is adopted for evaluating the safety of the levee and then the probability of failure for the levee is derived based on the fuzzy set based reliability method comprehensively considering the uncertainty in the input parameters. The derived results can provide valuable information about risk assessment and risk informed decision making for the stakeholders of earthen levees. A case study is presented to demonstrate the significance of the proposed framework for probabilistic assessment of earthen levee stability.

Field deformation of geosynthetic reinforced soil (GRS) is commonly inferred using strain gauges that are bonded to the layers of geosynthetic reinforcement. Attaching the gauges to the geosynthetic is a challenging process, as the use of overly rigid bonding techniques can affect the accuracy of measured strains. This paper describes two commonly used methods for foil strain gauge attachment, which were utilized in parallel to instrument the geotextile layers in a GRS bridge abutment. For this project, GRS construction was utilized, with closely spaced geosynthetic layers and a well-compacted, high-quality granular soil being used to construct two reinforced soil bridge abutments; these abutments were then used to provide direct bearing support for structural bridge members. The effect of the gauge type and the attachment technique on the strain readings will be presented. Finally, practical issues related to gauge survivability will be discussed.

This paper investigates temperature changes in the abutment of a geosynthetic reinforced soil integrated bridge system (GRS-IBS). A GRS-IBS structure is a type of bridge construction that uses geosynthetic reinforced soil (GRS) abutments to support prefabricated bridge structural members. GRS abutments are composed of high quality granular soil and closely spaced geosynthetic layers; the close spacing of the reinforcement increases the confinement and stiffness of the compacted aggregate, resulting in a strong and internally-supported abutment system for the bridge superstructure. Thermal expansion and contraction behavior of the bridge and abutment plays a role in the long-term behavior of GRS-IBS structures over the course of many seasons and years of service. This paper examines seasonal heating and cooling behavior within a typical GRS bridge abutment, using temperature data collected from an array of thermistors installed in a GRS-IBS constructed in the state of Delaware in the U.S. In order to monitor the temperature changes in the abutment of the constructed GRS-IBS, thermistors were installed on the fabric in the GRS abutments and changes in the abutment temperature were continuously measured using these embedded thermistors over 3 years of in-service operation. Measured results showed that the changes in the abutment temperature follow the same trend as the air temperature and are proportional to the elevation and the distance from the facing wall. To illustrate this behavior, temperature distribution contours in the abutment on different days and months over the course of a year are presented. Finally the effect of temperature change on measured strain readings for the GRS abutment is explained and some recommendations are made for minimizing its effect or correcting for its effect in future field studies of this type.

This paper examines a significant bridge project and four factors that contributed to the design, construction, and sustainability of the project. The four factors include the environment, society, site conditions, and economics. Engineers and contractors are well-versed in designing and building transportation projects within the constraints of federal environmental guidelines and strict budgets while keeping the needs of the local community and environment in mind. This is especially true for projects with locally “normal” site conditions. A project overview of Oklahoma State Highway 99 over the Washita River looks at what happened when all four factors were extreme. The project endured many delays, alternative design studies, public opposition, and difficulties during construction. The project was eventually completed, but not before Mother Nature contributed her input.

Infiltration of rainwater into a slope causes an increase in pore water pressures, changes in the degree of saturation and a decrease in the shear strength of the soil. Each of these can cause a reduction in the factor of safety of the slope and may trigger a landslide. In this study, model slopes were prepared using a clayey soil to have different densities and instrumented with tensiometers and inclinometers. These slopes were subjected to different intensities of rainfall until complete saturation was obtained. Based on the experimental results, relationships were developed between the seepage velocities observed at various locations in the slope to evaluate the spatial distribution of seepage velocity within the slope at different slope inclinations. Moreover, the study result shows that when the relative compaction is less than 75%, volumetric compression of the slope was observed with the introduction of rainfall, whereas volumetric expansion of the slope was observed with the introduction of rainfall when the slope was compacted at relative compactions higher than 75%.

In this research, a sensitivity analysis was performed to study the effect of variation in maximum shear modulus (G_max) values of different layers of a levee structure on the seismic response and stability of the levee slopes. The analysis was implemented by systematically varying the G_max values of the different layers to simulate the effect of possible variation in estimated G_max values obtained from in-situ/laboratory tests and correlation equations. The G_max values of the layers were altered in two ways: variation in (1) individual layers; and (2) group of layers; thus, simulating the effect of incorrect estimation in G_max values of a particular layer and a group of layers, respectively. The levee was subjected to the acceleration time-history data of two earthquakes with considerably different predominant frequencies. It was observed that the inaccurate estimation of G_max values, especially for the deeper layers, significantly affects the natural frequency and peak crest acceleration. Moreover, a prominent change in the factor of safety of slopes under seismic loading conditions was observed with variation of G_max values for the near resonance condition. The effects were pronounced when the G_max values were varied for a group of layers, rather than individual layers.

Most analyses on translational landslides use the traditional infinite slope theory. However, the simplicity of the infinite slope model hinders accounting for essential field observations. In this regard, catastrophic landslide model, which was developed originally for a submarine condition, was applied to see if it can be properly adjusted for a subaerial case as well. The model employs dynamic fracture mechanics to enumerate the dynamic propagation of shear band, which eventually comes down to landslide failure with an estimate of the landslide failure size and velocity. In order to apply the model for unsaturated condition, suction stress of the vadose zone is calculated for the case of downward seepage. The model estimates the critical length, which indicates the stability condition of a slope and also assesses the landslide failure lengths (size) with reasonable assumptions (e.g., mild slope angle). This analysis exhibits that the proposed model has a potential to simulate subaerial cases properly as well as submarine ones.

One of the major concerns in the behavior of a dam are the changes to exit gradient and the impact on the slope stability under sudden drawdown conditions. Drawdown increases the gradient difference between the internal areas (i.e., dam body) and the external areas (i.e., the reservoir). It can also cause increased seepage forces on the upstream slope and on the foundation surface which may result in movement of soil particles in the flow direction causing erosion problems. In this paper, a numerical approach, based on the finite element method (FEM) is used to analyze the seepage through the dam and its foundation to determine the upstream exit gradients and slope stability. GeoStudio programs using the FEM are used to conduct both steady-state and transient analyses. Three typical designs of earth dams, with and without the core, are used to study the seepage and the upstream slope stability before and after the upstream drawdown event. Water flux, flow direction, exit gradient, and the slope factor of safety are calculated in each case. The results show that the core is important in reducing the flux through the dam. Constructing a cutoff under the core increases the efficiency of the core and lowers the phreatic line. However, the results also showed that the submerged weight increases when the earth dam is constructed with a core or with a complete low permeability cutoff. Increasing the submerged weight, in turn, causes higher water flux to flow out of the dam under the drawdown condition increasing the exit gradient. The exit gradient at the upstream slope may reach critical levels and cause failure of the dam. The factor of safety of the slope is also reduced because of increasing the buoyancy of the soil at the upstream side of the dam. Some preliminary results are presented in this paper.

Quantifying the impact force from a sliding mass is very important in risk assessment and engineering design against landslides. It gives insights into the behavior of granular flow and helps understanding the behavior of landslides. There are various parameters such as inclination angle, initial density, friction between granular flow and flow bed, and flow sliding distance, which affect impact force. A series of 2D sand flume experiments were performed to systematically investigate the effects of these parameters on impact force. In this paper the effects of sliding distance were studied by changing the location of a load measurement device along the sand flume. It was observed that depending on the inclination angle, the impact force may decrease or increase by increasing the distance between the flow origin and the obstruction (i.e., load measurement device). At inclination angles smaller than the friction angle of the sand, increasing the sliding distance will result in a reduction of the maximum impact force, while at inclination angles higher than the friction angle of the sand the impact force will increase as the sliding distance increases. In the experiments performed in this study, the maximum impact force decreased by 31% and increased by 92% at 30 and 60°, respectively, when the sliding distance was increased from 0.65 to 2.49 m.

Failures of I-wall levee systems were found to be the major cause of the extensive flooding after the Hurricane Katrina and the events thereafter. These events suggest that the current design procedures must be revised to reduce the overall probability of failure of the system by incorporating the uncertainties in the input parameters into the system. In this paper, a framework to perform design optimization of a typical I-wall levee made of clay resting on a sand foundation is proposed. The uncertainties in the system such as undrained shear strength of levee fill, friction angle of sand foundation, and flood water level behind I-wall were considered as random variables in the design optimization procedure. The design variables used were the penetration depth of I-wall from the levee crown, width of levee crown, and landside slope of levee. Several design cases were simulated and finite element analysis was performed for selected cases to estimate the overall stability of the system considering the possible variations in the random variables. A response surface was developed based on the FEM simulation results to represent the factor of safety as a function of random and design variables and implemented in optimization procedure. In the design optimization, total cost of I-wall levee system was minimized to reach cost-efficient designs and at the same time the probability of failure was evaluated and minimized. Finally, the optimization yielded a set of preferred designs known as Pareto front from which the optimal design can be selected.