Geo Jordan 2004: Advances in Geotechnical Engineering with Emphasis on Dams, Highway Materials, and Soil Improvement

Zones of low total stresses may affect the safety of embankment dams. Earth pressure measurements are carried out in order to determine stress redistributions occurring within the dam. The reliability of these measurements is significantly reduced, however, by various influencing factors, especially due to the mechanical interaction between the earth pressure cell and the surrounding fill material. The theoretical errors of measurement based on different material behaviour may be determined with the aid of approximation formulae, which is derived from numerical computations. Experimental investigations, as well as the evaluation of measured data from dam projects, serve to indicate the influence (also numerically determinable) of transverse pressure on earth pressure cell measurements. With the aid of suitable evaluation methods, it is possible to verify earth pressure measurements in embankment dams and also to improve their reliability.

In contrast to empirical design, rigorous analytical modeling and numerical simulations provide useful predictions on time-dependent changes in filters, and the potential rate of base soil erosion from the dame core under various geo-hydraulic constraints. Moreover, simulation of filtration phenomenon by numerical schemes provides a good idea of the minimum filter thickness, based on its retention capacity. In this Keynote paper, governing equations are developed considering the principles of mass and momentum conservation combined with appropriate seepage criteria and the equations are solved by finite difference procedure to simulate the time-dependent changes in filters. The important contribution of various void models and seepage conditions assumed by various researchers are discussed and compared, identifying their scope and limitations within the framework of practical dam engineering. The theoretical developments are verified through large-scale laboratory testing for both cohesive and non-cohesive base soils.

In this paper, a thermal analysis is carried out using two and three dimensional finite element method (FEM). The analysis is carried out for a roller-compacted concrete (RCC) dam, called Al-Mujib Dam in Jordan, using two different numerical approaches. The first approach is a conventional approach in which the dam is assumed to be constructed by placing continuous RCC layers at 3m intervals in every 10 days. Whereas, the second approach, called the actual approach, considers the actual placement schedule of RCC layers that are 30 cm thick. The actual thermal properties and actual climatic conditions were considered in the analysis, and the computer FE program ANSYS was used to simulate the construction process of the Al-Mujib RCC Dam. The numerically computed temperature obtained from ANSYS is compared with the actual temperature measured in the field using Distributed Fiber Optic Temperature Measurement (DFOT) and thermocouples installed within the dam body. The study demonstrated that the two dimensional FE analysis of actual approach is adequate to predict the temperature distribution in the dam body.

The quantity of seepage under the foundation of most hydraulic structures is of great significance to dam engineers and designers since it plays an essential part in the dam design and safety. Acquiring seepage estimates under dams may not be a straightforward task especially if the underlaying stratum exhibits heterogeneous and/or anisotropic properties. Since exact or analytical solutions for this seepage problem are non-attainable, researchers have developed several techniques in order to approximate the seepage quantity. Flownets and the method of fragments are two techniques that have long been used with limited success in furnishing seepage under hydraulic structures. Several numerical models that provide seepage estimates have been developed by making use of the finite difference and the finite element methods, or other numerical modeling tools. In this research study, an efficient and accurate method for obtaining seepage estimates under hydraulic structures was developed. The finite element method was successfully implemented for providing seepage and exit gradient estimates under embedded dams with single sheet piles underlaid by heterogeneous media. The data obtained from the numerical model was compiled and plotted to form different sets of design charts. These charts could be used efficiently for obtaining seepage and exit gradient estimates under the aforementioned structures. The results obtained from the finite element model were compared and verified against existing relevant mathematical approximation and other techniques. Designers and practitioners in the field would benefit from these charts because of their simple yet accurate estimates.

Two sewage treatment systems are described, both in the State of Paraná, Southern Brazil. They involve similar hydraulic lay-outs with oxidation ponds at the end of the process, but were built on different soil profiles. One subsoil is residual from basalt (in the North part of the State) and the other, located in the South, is basically sandstone. The geotechnics of these projects is discussed, including soil parameters, cut, fill, borrow areas, soil substitution, impermeabilization, foundation and slope stability. Some particularities of both job sites are described, showing the importance of apparently simple construction details, and the authors also present comments about the performance of the plants during the past years of regular operation.

The city of Curitiba Metropolitan Area, in Southern Brazil, hosts about two million people spreaded around several close counties in the State of Paraná. During the last years, a great number of hydraulic structures has been planned and constructed in this region, namely water and sewage treatment plants, reservoirs, pipelines, dams, etc. This paper reports a case of two small rivers, both being used as water supply to be treated and distributed to the local people, for which control flow hydraulic structures were designed. According to the Brazilian environmental law, whenever there is a contamination risk for a water source, there must be a plan and a set of counter measurements ready to be activated. A road crossing rivers is the example for this case. The two rivers that feed the Irai Water Treatment Plant (Piraquara and Irai), whose discharge is about 4 m³/sec, have to be blocked and control flow structures devised for an occasional emergency situation, by-passing the hazardous material that otherwise would reach the water distribution system.

This study presents the behavior analysis of a pumped power storage dam body that showed longitudinal cracks on its crest. The analysis is based on the filed monitoring results. Settlement records show that the dam crest settlement reached nearly 1% of the dam height. The seepage records show that its magnitude is about four times bigger than the design prediction. Pore pressure records show that there are essentially zero pore pressures at the down stream side of the core. However, the pore pressures at the bed rock indicate that the permeability of the bed rock is substantially larger than the expected magnitude. The color or turbidity of the seepage water does not show any sign of the erosion. The compaction record shows that the core was compacted to slightly wet side of the optimum moisture content. From the analysis, it is concluded that the cut-off effectiveness of the core is fair. A somewhat larger settlement and surface cracks are due to the wet side compaction of the clay core. A somewhat larger amount of seepage water is due to the seepage through the bed rock.

In Comparison with other massive concrete structures, thermal stresses of RCC dams have an important rule in the structural analysis, due to their construction method. Throughout this paper the Finite-Element based computer code, called STARD, along with other necessary relationships for determination of temperature and stresses are introduced. The code is then applied to the real field-scale problem of Roodbar RCC dam, to determine the impact of placement schedule on the thermal stresses of the dam. The finite element model of the dam is generated for every construction phase, by applying Birth & Death of Elements technique. The entire procedure is repeated until the whole dam is constructed. Throughout the analyses, in addition to placement schedule, the impact of some other influencing factors, such as daily variations of temperature, hydration induced heat generation, initial temperature of rock foundation, placement temperature of each layer, various boundary conditions and module of elasticity, are also investigated. Based on the results of this study, it might be concluded that for a given RCC dam, placement operation, if started on the warm season, would lead to higher thermal stresses in long-term, because of which thermal cracks may form and propagate throughout the entire body, particularly near dam foundation.

This paper evaluates the effect of volume change due to consolidation on one-dimensional transport of contaminants. The rate of consolidation and excess pore pressure dissipation result in a transient, nonlinear advective component of transport through clay. Consolidation experiments were conducted on kaolinite samples prepared at 50% water content. Bromide, a non reactive tracer, was mixed with the upper half of the kaolinite sample at concentration of 960 mg/L. Flux of bromide as a result of consolidation was monitored in the water surrounding the kaolinite sample. The results show continuous flow of bromide as a result of water drainage by consolidation. The bromide mass flux compared well with consolidation rate and ceased at the end of the consolidation process. A formulation for the simultaneous consolidation and diffusion of solutes in clays, solved by finite difference, is also presented. A hypothetical case is solved to demonstrate the effect of consolidation under single and double drainage on contaminant transport and breakthrough. The results show that consolidation in doubly drained clay impacts concentration profiles, but does not significantly impact breakthrough of the diffusive flux. Consolidation under single drainage conditions significantly impacts the diffusional flux.

In this study, TDR detection of the studied inorganic (NaCl) and organic pollutants (acetone, benzene, heptane, tricholoethylene, and xylene) was evaluated experimentally and theoretically. For inorganic pollutants, samples were prepared with different water contents and NaCl concentrations, which were tested using TDR. For organic pollutants, samples were prepared with pore fluid content of 14% and different ratios of organic fluids to deionized water ranging from 0 to 100%. The TDR waveforms were analyzed using eigendecomposition technique. Eigenvalues were calculated by using autoregressive modeling and the singular value decomposition technique. Then, the most significant eigenvalues were identified, based on their power, and used in the multivariate statistical analysis. In addition, for organic pollutants, the relative eigenvalues of the first mode were calculated and correlated to organic chemical type and content, as well as octanol water partition coefficient.

In this study, mineral transformation, due to application of the newly developed ALFA solidification/stabilization process to sulphide bearing soils, and changes in the hydro-mechanical properties of the stabilized/solidified matrix were discussed. Samples were treated with lime, fly ash, lime and fly ash, and Aluminium, Lime, and Fly Ash (ALFA). The success of the treatment techniques was evaluated from the viewpoint of their interaction, chemical stability, and hydro-mechanical properties such as shear strength and hydraulic conductivity.

The flow of water through rigid non-submerged vegetation in wide floodplains was modeled. Rigid rods were used to model the vegetation. Different configurations of these rods represented different vegetation densities. Each configuration was tested at three flow discharges. Values of friction factor Manning's n for these different configurations were calculated. Results of this study were presented as functional relations that relate the friction factor to the flow and vegetation density parameters. The validity and the uncertainty of the friction factor values were tested. Tests were conducted to determine if the results of two trials for each experiment were consistent. It was found that 95% of the differences between depth readings from the two trials were less than 3.7%. This value was considered within the range of experimental errors. The Manning's n values across one, two, three, and four rows of rods were measured and used to determine whether the Manning's n could be calculated using the upstream condition. It was found that there were no significant differences between the n values computed for different numbers of rows. This confirms that the value of n can be calculated using the upstream conditions. Thirty six different configurations of rod densities and flow discharge values were tested. Twenty four values of n were calculated for each of these configurations. These values included two trials for each case, three values at the lateral dimension of the flume and four values of n measured for 1, 2, 3, and 4 rows. It was shown that all the data computed for a given configuration at a given discharge arise from the same population. The uncertainties of the data were found to be within an acceptable range given the large variation of reported Manning's n.

When rehabilitating a concrete-faced open water canal with a geomembrane-based surface sealing system, the fixation must be designed properly to prevent the tearing out of the membrane in any loading case. Appropriate fixations have not yet been investigated for materials such as PVC, HDPE and FPP which are used for this type of application. For the purpose of testing fixations a special experimenting device has been built at the Laboratory of Hydraulic and Water Resources Engineering of the Technische Universität München. The device allows the testing of any fixation of geomembranes onto any kind of surface. Loads of up to 30kN/m can be applied to the fixation at variable angles. In this paper results of `pull-out' tests carried out in the Laboratory will be presented.

Economically and environmentally feasible alternatives are being explored for re-use of scrap tires. One of the alternatives has been shredding of scrap tires to produce tire shreds ranging in size from 50 to 300 mm for use as alternative to crushed stones (gravel) as drainage media in landfill leachate collection systems. For leachate drainage layers constructed with gravel, the hydraulic conductivity of drainage layer, slope of the landfill bottom layer and drainage distance from the upstream boundary to the drainage outlet are constant for a specific landfill; the only variable being the rate of inflow into the landfill leachate drainage layer. With leachate drainage layers constructed using tire shreds, the hydraulic conductivity of the leachate drainage layer becomes an additional variable parameter. This is because tire shreds are highly compressible in nature and the resultant compression affects the porosity and in turn the hydraulic conductivity of the tire shred leachate drainage layer. This study performs laboratory experiments to evaluate the permeability characteristics of a tire shred sample under various vertical stresses to simulate the permeability behavior of a tire shred drainage layer subjected to vertical stresses by municipal waste overburden. The laboratory results are used to estimate the maximum leachate head on the primary liner using established formulas. The performance of the tire shred drainage layer in controlling the maximum leachate head on the primary liner is compared with those of gravel drainage layers.

Fluid flow in three-dimensional (3-D) microstructures is simulated using a finite difference numerical scheme. This numerical scheme uses the non-staggered grid system that uses one computational cell only to solve the governing equations of steady incompressible flow. Fluid flow is simulated in idealized 3-D microstructures to verify the numerical model. These idealized microstructures are a medium of small fissures and a medium of packed cubes. In these idealized microstructures, the numerical solution of the governing equations is controlled using periodic boundary conditions. The numerical results compared well with the closed-form solutions for permeability of these media. Simulation of fluid flow in a real granular material is also presented in this study. The X-ray computed tomography (CT) system was used to capture the 3-D microstructure of this granular material.

The behavior of granular materials has been thoroughly investigated in the last two decades by many researchers using simple to highly advanced experimental and numerical models. The failure concept in general describes the progress of deformations that begin at a certain point and propagates to the rest of the soil mass. This phenomenon will eventually cause the strains to localize into narrow zones known as shear bands depending on many factors such as; stress path, boundary conditions, local material properties and the degree of heterogeneity. Hence, it is very important to characterize such phenomena and carefully evaluate the behavior of the full scale structure. To be able to achieve this, a non-local theory needs to be implemented to relate the local variables to the neighboring ones and thus come up with some smoothen gradients. The Cosserat theory and the gradients approach (approximation of non-local theory) were found to be quite successful in capturing localization through some internal length scales which are usually taken as the mean particle size in granular materials. In this paper, micro-structural properties and their effect on the constitutive and global behavior of the granular materials will be presented with emphasis on the importance of the micro-length scale.

This paper presents a higher order gradient multi-slip formulation to model the effect of inhomogeneous deformation in granular materials. In most current formulations, the flow stress in the plastic constitutive equation is modified with a higher order gradient term of the effective plastic strain as a means of modeling the effect of inhomogeneous deformation. Granular material consists of discrete particles and associated voids. Therefore, a higher order gradient constitutive model must explicitly account for its gradients in terms of the heterogeneous granular microstructure. The effect microstructure is accounted for by considering the mobilized friction coefficient in the model to be a function of a higher gradient porosity distribution. The double-slip version of the model has been implemented in a standard finite element code and used to simulate biaxial shear tests on dry sand. The effect of the two new formulations and the current formulations using effective plastic strain on the prediction of strain localization are examined.

The site selection study for a municipal landfill can be a costly and critical affair in the wake of limited availability of areas suitable for landfilling. Siting a landfill typically requires processing and evaluating a significant amount of spatial data with respect to various siting rules, regulations, factors and constraints. Implementing such a complex spatial analysis with drawing tools is certainly expensive and tedious. A modern geographical information system (GIS) is capable of processing a large amount of spatial data, thereby potentially saving time and money. The objective of this paper is to enable site planners and engineers who work with GIS to have a tool for preliminary site-selection studies. ArcMap a component of ArcGIS 8.2 was customized to generate a specifically designed toolbar for landfill site selection (LSS toolbar). Such a tailored GIS will enable engineers to investigate and compare results of applying different criteria, constraints and scoring schemes on the final suitability map for landfill site in an area. Customization is performed using the Microsoft Visual Basic for Application. (VBA).

Travel-time tomographic imaging of foundation soils and earth dams presents several difficulties. These difficulties are related to the data interpretation due to the non-linearity caused by the coupling of stress heterogeneity and anisotropy in soils, limited data sets, and diffraction around inclusions. In this paper, preliminary numerical and experimental studies have been implemented to address and simulate problems related to the interpretation of a simulated stress field. The numerical simulation of ray tracing evaluates the effect of ray curving in the data interpretation, while the experimental study simulates the quality of the data for its application into inversion techniques. The experimental laboratory data is gathered using a discrete two-dimensional model for wave propagation. The experimental data capture the effects of curved rays, small data sets, and restricted source-receiver angles on the evaluation of the near surface properties and the monitoring of the state of stress. Results will help in the development of inversion models for the evaluation of effective stresses during the construction of embankments and dams.

In order to investigate the influence of grain size and mineralogy on the strength of weak rocks, eleven samples of weak rocks were collected from different areas of Muscat, Sultanate of Oman. Grain size properties included effective particle size, diameter, and the clay content. The behavior of clay and other minerals was studied with the help of Atterberg limits and XRD. Point load index, slake durability, cohesion and friction angles were taken as a measure of strength. The results indicate that grain-size can be correlated with the strength. A decrease in clay content and D₁₀ resulted in higher strength values. The presence of quartz was not found to influence the angle of friction. However, slake durability of the weak rocks was significantly increased by the presence of quartz.

Pile foundations have been distinguished as one of the favorable solutions for construction in difficult soil and structure conditions. The design of pile foundations primarily depends upon the function of pile whether to improve ground conditions, prevent slope movements, provide adequate bearing capacity, provide resistance to lateral loads, or resist uplift loads. The design of pile foundations should assure that the piles will effectively distribute or/and transfer the applied vertical load deep enough into to a higher effective stress level to provide the necessary resistance. The majority of current design formulas have been proposed for dynamic and pseudo-static loading conditions, and they suffer the absence of the physical soil-structure interaction. In this paper, a complementary design methodology is proposed for predicting the compressive load capacity of driven piles in sand. This method accounts for the post-driving residual stresses, level of effective stress, pile characteristics and incorporates the friction fatigue process as the pile penetrates into the soil. The method features statistical correlations that were formulated based on field testing data collected from previously published works. These correlations are verified using an independent database and by the use of other researchers. The proposed methodology was validated using the measured bearing capacities of 18 pile load tests, and was shown for the reasonable predictability and accuracy.

For the first time, exact secant stiffness and loading matrices for Timoshenko beam resting on variable one-parameter Winkler foundation are developed, using the power series approach. Effect of shear on deflection is taken into account in constructing the uncoupled differential equations. The proposed exact element is analyzed thoroughly to include the effect of shear on both deflection and internal forces; the results of such rigorous analysis show that the proposed exact secant stiffness matrix gives a big saving in the computer time with better accuracy of results. The proposed problem solution is compared with finite element analysis performed using the multi-purposes package MSC/NASTRAN. The deflection due to shear is found to be considerable in beams having small span to depth ratios with large values of modulus of foundation. It is also concluded that effect of shear must be taken into account in settlement calculations.

Shale is the most abundant rock type in a group called Wianamatta. The group is the major geological sequence in the Sydney Basin. Bringelly shale, the major formation of the group, is comprised of claystones and siltstones with occasional sandstone layers. Weak cementation and presence of swelling clay minerals have caused a number of problems with structures founded on Bringelly shale. The results of triaxial tests on reconstituted specimens are compared with the behaviour of a natural rock specimen. In order to study the de-structural behaviour of the shale, the material was subjected to high stress so that low porosity structure can be produced. Same materials that have experienced higher stress show significantly reduced frictional strength and pattern of behaviour inconsistent with critical state concepts.

This paper evaluates the potential use of the Geogauge test device to reliably measure the in-situ stiffness characteristics of highway materials and embankments. To assess this, both laboratory and field testing programs were conducted. The laboratory tests were conducted on different materials prepared inside two test boxes (5ft x 3ft x 3ft) located in GERL lab at Louisiana Transportation Research Center (LTRC). While the field tests were conducted on several highway sections selected from different highway projects. In addition, nine sections were constructed and tested at the LA-DOTD Pavement Research Facility (PRF) site. The testing program included conducting Geogauge tests, in addition to the static Plate Load Test (PLT) that was used as a reference test. The Geogauge stiffness modulus obtained from laboratory and field tests was correlated with PLT moduli, and good correlations were obtained. The results of this study suggest that the Geogauge can be a promising tool for in-situ measuring the stiffness of highway materials and embankments.

Understanding the interaction mechanisms between soil and geosynthetics is a key factor in the design and analysis of reinforced soil structures. The soil-geosynthetic interaction can be determined by conducting direct shear or pullout tests. Most of the research efforts have been focused on investigating the pullout behavior of reinforcement in granular fills. Recently, marginal fills with poor engineering characteristics have gained increasing interests due to the cost effectiveness of utilizing the available cohesive soils in the construction of reinforced retaining walls and embankments. This raises the need for the evaluation of the interaction parameters between geosynthetics and cohesive soils. To do so, series of reduced-scale and full-scale (field) pullout tests were carried out on instrumented reinforcement intrusions in cohesive backfill by the Louisiana Transportation Research Center. This paper presents the results of laboratory and field pullout tests programs conducted on different geosynthetic reinforcements in marginal cohesive backfill. The results presented in this paper include the evaluation of coefficient of interaction (C_i) between different geosynthetics and the cohesive backfill, the variation of the pullout resistance with overburden depth, and a comparison between the laboratory and field pullout results.

A test programme has been carried out in the geotechnical centrifuge of the University of Delft to investigate how different design parameters influence the failure mechanisms and displacements of dikes stabilized by three innovative methods. The three methods under investigation are known as: mixed in place, soil nailing and expanding columns. The tests included models without reinforcements (9 tests) as well as models partly simulating mixed in place (6 tests), soil nailing (6 tests) and expanding columns (5 tests). Some of these tests are discussed in this paper. The parameters tested were expected to have a significant influence on the three innovative methods. Centrifuge research indicates that, in comparison with non-reinforced models, all three innovative methods can improve stability. The extent of the improvement depends heavily on the design parameters, such as the smoothness, the number, the inclination and the stiffness of the reinforcing structural devices. It was possible to adjust the design of all three methods so that the model did not fail.

Modern engineering is continually being developed to achieve a more economic and at the same time safe structure. The monitoring systems nowadays play a very important role as a main part of the design procedure and as a tool to prove the safety and serviceability requirements. The rapid development of the infrastructure makes the application of highly developed monitoring systems, particularly in geotechnical engineering, a challenging task for civil engineers. This paper presents the main aspects of how to design and perform a geotechnical monitoring program and how to evaluate the measurements. Three case histories will be presented to demonstrate the practical applications of monitoring programs in conjunction with special numerical analyses to achieve a safe structure regarding the serviceability requirements as well as an economical design regarding the total cost. These case histories include: embankment on compressible soil, piled raft foundation and deep pit excavation. The possibilities, the advantages and the need of highly developed geotechnical monitoring systems will be discussed.

This paper presents the results of a study performed to evaluate the resilient modulus of typical Wisconsin soils. A laboratory-testing program was conducted on soils that commonly occur in Wisconsin to determine their physical properties, compaction characteristics, and resilient modulus. The resilient modulus test was determined by the repeated load triaxial test. Tests were conducted on soil samples prepared under variety of moisture contents and unit weights. Test results showed that the resilient modulus of clay and silty sand soils decreased with the increase of the deviator stress and decrease of the confining stress. Soils prepared under 95% of γ_dmax on the dry side of the compaction curve exhibited increase of the resilient modulus with the increase of the deviator stress. This paper provided preliminary results on the behavior of various Wisconsin soils under repeated load triaxial test.

Design of reinforced soil structures is coming out from very conservative presumptions that lead to highly over designed and uneconomical constructions. The main reason is low value of the creep tensile strength that is measured on the geosynthetic reinforcement elements in the laboratory without interaction with the soil. SG Geotechnika has been monitoring, for several years, behaviour of many reinforced soil structures. Comparison of measured values of extension (and corresponding stress) of the geosynthetics reinforcement with predicted values shows that the difference can be hundreds of percent on the safe side. Results of monitoring will serve for updating of Czech design code for reinforced soil structures.

This paper presents the results from a preliminary laboratory evaluation of recycled blended calcium sulfate (BCS) material, with an emphasis on its water resistance. The results indicate that the strength development or deterioration of BCS material heavily depends on the initial water content and relative humidity of curing conditions. This phenomenon can be elucidated by the interfacial free energy concept. BCS was then stabilized by adding portland I cement, class C fly ash, and ground granulated blast-furnace slag (GGBS) in an attempt to enhance the strength and water resistance of BCS material. The effectiveness of these cementitious materials was subsequently examined in terms of strength enhancement, water resistance, and microstructures of the stabilized BCS material. Of all the cementitious materials used in this study, GGBS was found to be the most effective agent to stabilize BCS.