2nd International Workshop on Geotechnics of Soft Soils  »  Abstracts and Papers

Abstracts and Papers :


The call for papers resulted in about 80 abstracts. The list of accepted abstracts can be seen below. Selected authors will be invited to develop their submission to a keynote paper. The dead-line for papers is March 30th, 2008.


Please note when submitting a paper that it is expected that at least one of the authors is registered to the workshop to present the paper, either orally or as part of a dedicated poster session. When submitting his/her paper the main author is obliged to sign and return a Consent to Publish Form, which can be downloaded here.


By submitting a paper for publication the authors agree to transfer the copyright of his/her paper to the Publisher.


Furthermore:


- All papers should be delivered both in electronic form (MsWord or LaTeX required) as well as in identical camera-ready hard copy. The instructions can be found in http://www.balkema.nl/instructions.asp.

- The maximum number of pages per paper is 6 pages for regular contributions

- The maximum number of pages per paper is 8 pages for keynote papers


PLEASE SEND YOUR PAPERS TO THE FOLLOWING ADDRESS:


Dr Minna Karstunen
Department of Civil Engineering,
John Anderson Building,
Glasgow G4 0NG
Glasgow
UK

Tel: +441415483277
Fax: +441415532066
E-mail: amgiss@strath.ac.uk


List of accepted abstracts:




Development of Side Shear Resistance of driven pile in soft ground


WK Ng, MR Selamat, KK Choong


Bearing capacity of driven pile often increases with time after installation. This phenomenon is known as soil/pile set-up. The phenomena were reported by researchers all over the world but no consideration yet in the routine pile design works in the relevant industry. The piling mechanisms involved of large strains happened as pile driving into the soil and excess pore water pressure developed around the pile shaft. Initial side shear resistance of pile found low and slowly increased relatively as pore water pressure dissipated with time. The program consists of 3 sizes of reinforced concrete piles which installed in soft ground by con-ventional driven method. The sizes are ranged from 150mm – 250mm. Soil pressure as well as pore water pressure measured in the study. The pressures are monitored throughout the study until the pressures changes are negligible. Typical soil exploration tests such as Standard Penetration Test (SPT), dynamic cone penetra-tion test (DCP) and Cone Penetration Test (CPT) are used to investigate the soil characteristic and in-situ strength. This study is aims to investigate the development of side shear resistance on pile shaft, dissipation of pore water pressure and further explain its relationship with side shear resistance (SSS) of driven pile in soft ground.



A Constitutive model for lignosulfonate treated soft soils


Buddhima Indraratna and J. S. Vinod


A Constitutive model for lignosulfonate treated soft soils Buddhima Indraratna+ and J. S. Vinod++ +Professor, Department of Civil, Mining and Environmental Engineering, University of Wollongong, NSW, Australia, 2522; Email: indra@uow.edu.au ++ Research Fellow, Department of Civil, Mining and Environmental Engineering, University of Wollongong, NSW, Australia, 2522; Email: vinod@uow.edu.au Workshop Topic: Constitutive modeling Abstract Soft and unstable soils are conventionally stabilized using chemical admixtures such as lime, gypsum, cement, fly ash etc. These admixtures (stabilizing agents) generally alter the mineralogical structure of the clay and improves the inherent properties of the soil such as strength and stiffness. However, problems such as sulphate attack on concrete and steel structures adjacent to gypsum treated soils, problems with vegetation and groundwater contamination of chemically treated soils due to high pH levels etc, have demanded researchers to find alternative stabilizers. Recently, lignosulfonate, a by product of wood industry shows a promising prospect as stabilizing agent especially for soft grounds. Preliminary investigation on lignosulfonate as stabilizing agent showed that, amount of lignosulfonate required to stabilize soft/unstable soils is much less compared to other admixtures such as cement, lime, or fly ash. Furthermore, it has also been observed that stress strain and volume change behaviour is distinctly different from those stabilized with traditional admixtures and that the lignosulfonate soils maintain the ductile characteristics. Only a few studies are reported in the literature, hence, the mechanism of soft soil stabilization by lignosulfonate is not well understood. In view of the above, laboratory experiments were carried out on silty clay stabilized with lignosulfonate. A preliminary constitutive model for lignosulfonate treated soils has been developed. Based on the extensive laboratory experiments, the basic mechanics involved in the lignosulfonate treated soils is presented and discussed. The constitutive model was developed based on the theoretical considerations within the framework of the critical state concept. The results obtained from the model were validated with the laboratory experimental results.



IMPROVEMENT OF SUB-GRADE FOR HIGH SPEED TRACK (IN SOFT GROUND AREAS)


Dinis da Gama, C. and Moraes, R.M.


An essential prerequisite for High Speed tracks is to maintain control on the degradation of track geometry so as to keep various tolerances well within the specified limits. Degradation of track geometry is a function of track design, axle-load, speed, and sub-grade characteristics. Improvement of sub-grade in poor quality ground areas is recognized as one of the most significant factors for their stability. The dynamic soil-structure interaction of railway lines depends on many aspects like the shear modulus and damping ratio of the soil as well as the frequency and amplitude of the dynamic loading. Measurements at railway lines in Ledsgrard, Sweden (Hildebrand, 2001) and the high speed track L2 between Brussel and Paris (Degrande et al, 2001) indicate that the dynamic response of the high speed phenomenon (Holm et al, 2002). Katzenbach, R. and Ittershagen, M. (2006) presented a new measurements that clearly show that in soil conditions the mentioned phenomenon appears also in the normal speed range up to 150 km/h. One countermeasure to reduce the oscillations respectively the particle velocities of the ground and thus consequently the long term deformations of the track bed is to increase the stiffness and dynamic stability of the layer by a column shaped soil improvement. This paper describes the considerations involved and the methodologies to be adopted for geotechnical engineering applications on sub-grade improvement. The available processes to overcome the problems presented by poor ground areas are discussed along with the examination of the various available options, outlining their advantages and limitations.



Collapse and deformation behaviour of alluvial loess soils with special respect to the conditions in the Marmal-mountains, Afghanistan


Dipl.-Ing. Claas Meier, Univ. Prof. Dr.-Ing Conrad Boley, Dr.-Ing. Yazhou Zou


The soil in the Marmal mountains south of Mazar-e-Sharif, Afghanistan, is mainly alluvial loess. Existing foundation design standards are not applicable for soft ground. Due to an insufficient understanding of the properties of these soft soils extensive structural damage has occurred to buildings in this area. Similar problems have been observed in West China where high speed railway lines have been built on loess soils. Little is known about the relationship between water content and the deformation behaviour of loess soils. NATO instructed the Institute of Soil Mechanics and Foundation Engineering to carry out an extensive research programme with the aim of improving the understanding of geotechnical engineering on and with soft loess soils, in order to make recommendations about soil improvement, drainage and structural design. Particular attention was paid to the sensitivity of alluvial loess soils to changes of the water content in conjunction with the salt content. In order to learn more about loess soils, on site tests, including penetration, immersion and dynamic load plate tests were carried out around Mazar-e-Sharif and Kunduz. Undisturbed samples were taken to Munich for laboratory tests to determine relevant physical properties. In order to develop new procedures for foundation methods on loess soils, special tests, such as collapse tests in conjunction with microscopic structural analysis of the soil were carried out. Collapse tests were performed in the laboratory according to Chinese standard GBJ 123-88 in order to study the sensitivity of the soil to water. This institute’s own test procedures have been developed out of these results. A key result has been the development of tables and diagrams to enable easy estimation of loess soil load capacities. The research consists of two main parts: a. The development of a new estimation method for static loads. b. The development of a new estimation method for dynamic loads esp. due to earthquakes. In order to develop new methods for estimating the effects of dynamic loads, modified tests in a triaxial-cell were carried out. It proved possible to develop a frequency range for when loess soils are most likely to collapse. Thus, we are now able to present the new design criteria we can give for geotechnical engineering on these soft soils on the basis of our experimental and theoretical research.



Design and construction of a caisson breakwater on soft ground


A. Gens, E.E. Alonso, R. Madrid, D. Tarragó


The paper presents the design and construction of a caisson breakwater in the Barcelona Harbour. The breakwater is founded on deltaic deposits with a low bearing capacity. The soil strength is insufficient to provide adequate support to the caissons, so a key aspect of the design is the need to consider the strength improvement of the soft sots and clays due to the loading at intermediate stages. Important issues relating to the design are: - The selection of the failure mechanisms and factors of safety to be considered - The incorporation of the forces and pore pressures representing storm loading - The consideration of cyclic loading effects due to wave action The need to have improved soil strength makes it necessary to evaluate the times requires to complete intermediate consolidation stages. Given the uncertainty of this process, instrumentation has been installed to control the degree of dissipation and other phenomena. Pore pressures and horizontal and vertical displacements were measured. In the paper, the performance of the monitoring system and the observations obtained are discussed in the context of the predictions made. Finally, the safety with respect to the dynamic wave effects is estimated comparing the cyclic stress in different ground locations with the results of simple shear tests preformed on undisturbed samples using the framework of interaction diagrams.



Development of a Case-Based Reasoning System for Deep Excavation


Yuwen Yang


Because of the variability of soft soils, it is always uneasy to rationally and properly determine the values of their parameters, such as cohesive force c and internal frictional angle ƒÚ and so on. In reality, soil parameters are random variables, not given values. In this paper, an innovative approach, called case-based reasoning, is introduced and used to develop a system in the domain of deep excavation aiming to solve the problem about soil parameter determination. By collecting a large amount of the completed deep cases, the useful information involving the chosen values of soil parameters within these cases is picked up and sorted, and then input to a database we have developed where the technique of case-based reasoning is applied. From the database, for an enquired case, a user can easily find out the best matching case within the stored cases of deep excavation in the consideration of the restraint conditions along excavation boundaries, excavated depth, soil parameters and geological situations at sites. The retrieved case owns some valuable data for the reference to the enquired case, in which the given values of soil parameters are included. The procedures and processes of developing the system, retrieving a case automatically and others are detailedly described in this paper with the effort to select the rational values of soft soil parameters for the enquired case. Keywords: case-based reasoning, deep excavation, GIS, database



The Veda trial embankment – comparison between measured and calculated deformations and pore-pressure development


Rasmus Müller, Stefan Larsson


2nd International Workshop on GEOTECHNICS OF SOFT SOILS Focus on Ground Improvement 3-5 September 2008, University of Strathclyde, Glasgow, Scotland, UK The Veda trial embankment – comparison between measured and calculated deformations and pore-pressure development Rasmus Müller, Tyréns Infrakonsult AB, Sturegatan 4, SE-784 31 Borlänge, Sweden Stefan Larsson, Tyréns Infrakonsult AB, SE-448 86 Stockholm, Sweden Abstract The Veda-embankment is a part of the Ådalsbanan Railroad Line which connects the existing railway along the east coast of Sweden, from Sundsvall and further north, with the Botnia Line. The embankment is planned to be about 300 metres long and up to 16 metres high. The embankment is being built in a staged-construction mode on subsoil consisting of very loose sulphurous postglacial clay and very loose glacial clay. The staged construction and the ground improvement by pre loading are performed according to the observational method. Vertical drains are used to speed up the consolidation process. In order to obtain a reliable prediction of the deformations and the pore-pressures during the staged-construction process, a 50 metres long and 6 metres high trial embankment has been built. The trial-embankment was instrumented to measure pore-pressure development, vertical- and horizontal deformations. Furthermore, vane-tests and CPT-testing at different times has been performed in order to verify the strength increase of the soil. For the same reason, simple direct shear tests have been performed at different consolidation pressures on samples from the site. Finite-element calculations have then been performed to find a model that describes the behaviour of the soil sufficiently well. The soil-model has later been used to obtain a reliable prediction of the behaviour for the staged-construction process of the Veda-embankment. The emphasis of the paper is to present the calculated and measured performance of the trial embankment and to discuss the soil-model and how well it describes different aspects such as excess pore-pressure build-up and dissipation, horizontal displacements and settlements. The paper further discusses the benefits achieved from the knowledge about the behaviour of the subsoil of the trial embankment when preparing the final design for the large Veda-embankment. One very important information from the measurements on the trial-embankment is that the coefficient of consolidation in the horizontal direction was much lower than expected.



Smear Effect on Consolidation Behaviors of SCP-Reinforced Composite Ground


Yun-Tae Kim and Thanh Hai Do


Sand compaction pile (SCP) method is composed of compacted sand pile inserted into the soft clay deposit by displacement method. SCP-reinforced ground is composite soil which consists of the SCP and the surrounding soft soil. When a surcharge load is applied to composite ground, time-dependent behaviors are occurred in the composite soil due to consolidation according to radial flow toward the SCP. In addition, stress transfer also takes place between the SCP and the soft soil. This paper presents the numerical results of cylindrical composite ground that was conducted to investigate smear effect on consolidation behaviors of SCP-reinforced composite ground. The results showed that the smeared zone of soft clay had a significant effect on effective stress-pore water pressure response, stress transfer mechanism and stress concentration ratio of composite ground. Amount of stress transfer between the clay and the SCP was maximum in depth of z/H=0.25, and decreased with depth. Stress concentration ratio of composite ground was not constant, but depended on consolidation process. It was also found that the value of stress concentration ratio in soft clay with smeared zone was larger than that in soft clay without smeared zone. However it converged to the constant value after the end of consolidation irrespective of the smeared zone.



Countermeasures against settlement of embankment on soft ground with newly-proposed sheet pile wall method - PFS (Partial Floating Sheet-Pile) Method


Noriyoshi Harata, Jun Otani (Kumamoto University, JAPAN)


To prevent ground settlement, lateral deformation and upheaval that occur in the surrounding ground due to banking such as levees and roads on soft ground, the countermeasure with sheet pile walls driven continuously into the bearing stratum is effective. The measure with sheet pile walls forms stress discontinuity between the embankment and the surrounding ground to reduce the influence of the embankment on vertical and lateral deformation caused in the surrounding ground, and therefore differs from soil improvement and reinforced-earth methods. In the case of raising or widening of existing embankments, although the applicability of soil improvement method is limited right under the embankment, the sheet pile wall method will be highly applicable because it requires only a small space. This type of construction should be designed to secure the stabilities of the embankment and the structures around the embankment taking the property of the ground and loading condition into account. In this report, PFS (Partial Floating Sheet-pile) method in which a minimum number of sheet piles driven into the bearing stratum is used is introduced. PFS method enables to improve both cost and constructional performances based on the reasonable design method which satisfies the requirement in and around the embankment. The several field test results to observe the behavior of ground and sheet piles are reported and a simplified design method with bilinear model as soil behavior is proposed based on the test results. The effect of an estimated construction cost reduction is also described.



Performance of Highway Embankment on Low Improvement Ratio Deep Mixing Method


Jutha Sunitsakul, Attasit Sawatparnich, and Satipong Apimeteetamrong


Department of Highways, Thailand, reconstructed the highway route number 3117 (KlongDan-Bangbor), which connects two major highways to the eastern part of the country. Soft Bangkok clay layer in the reconstruction area is about 12-15 meters thick. Excessive settlement of highway embankment will be expected if the embankment is constructed on unimproved Bangkok clay. In this reconstruction project, the highway embankment constructed on improved soft Bangkok clay, deep mixing cement column, shallow cement stabilization, and embankment preload. Total pressure cells, piezometers, vertical inclinometers, settlement plates, and rod extensometers, have been installed in order to monitor the highway embankment performance during and after the reconstruction. Horizontal displacement is expected to occur in the deep mixing cement column section even though the safety factor of slope stability is high. Stress distributions from highway embankment thru cement columns are around 55 to 66 percent for the heads of cement columns at level -1 and 0 m, respectively. Pore water pressure measurement from piezometer demonstrated the load of the highway embankment thru soil layer beneath. Field measurements from settlement plates showed the differential settlement of the highway embankment on between the cement columns. With some adjustments on the engineering properties of the soft Bangkok, finite element analyses provide good estimated settlement of highway embankment.



Structured Soils: Implications for the Practicing Engineer


Sergei Terzaghi, Minna Karstunnen


Whilst structured and anisotropic behaviour of soils have been recognised in academic circles for many years, the implications of such behaviour has not been recognised by the practicising geotechnical engineer to their detriment. Three structured soils will be modelled using a constitutive model (s-clay1s) that models both anisotropy and structuration and demonstrate the impact on real problems that commonly face practicioners. It is intended to that, by modelling the soils to show the match (or otherwise) to the real exhibited behaviour from these soils (two residual soils from New Zealand, and an alluvial soil from Australia), the application of the concepts to real problems. Two of the problems involve slope stability realted problems, and the third embankment behaviour. Each of the problems, in the absence of the concepts of structure and anisotropy represent insoluble problems, yet are readily explained by the concepts embodied in the concept of structured soil. It should be noted that the problems involved were solved in the absence of using an advanced model such as the s-clay1s. The purpose is to show the implications and consequences of such behaviour on real problems,along with the advantages of using the concepts embodied in structured soils for the practioner. These advantages are real, in that the offer both the opportunity to save money, and also to avoid embarassement when behaviours are not as expected according to conventional theory. Discrepancies between what is observed and that predicted will be noted and discussed, as that provided the opportunity to advance the state of art.



Research on creep and consolidation characteristics of soft soil


CHEN Xiao-ping ZHOU Qiu-juan


Based on the mass of contrast tests, creep consolidation characteristics and calculating models of typical soft soil from Guangdong are systematically studied. The tests results show that the deformation proportion of consolidation and second consolidation is related to the factors, including initial load and loading ration in test, consolidation degree, water content and void ratio of soil samples. The complications of affecting secondary consolidation coefficient are loading history, loading ration and current consolidation state of soft soil. Preloading can reduce secondary consolidation coefficient. The scale between second consolidation coefficient and compression index is constant, about 0.025¡«0.10. Direct shear-creep characteristic is relation to vertical consolidation pressure, and the creep rule presents strain hardening characteristics. Consolidation action can play down the creep effect, so the creep characteristic in triaxial test is connected with drainage condition. In coordination loading addition, the relationship between deviator stress and axial strain is nearly straight-line in CD test, but in CU test, it assumes non-linear and there is visible yield characteristics. When the continuous loading manner is adopted in UU test, the failure strength will be improved. Bases of above results, the models of creep and consolidation are analyzed form considering respectively nonlinear constitutive relationship and shear drainage condition.



Development of artificial structure in Leda clay


S. Sasanian, T. Newson


The formation and mechanical behaviour of the sensitive clays of eastern Canada has been a subject of research investigations for a long time. There is evidence of natural cementitious bonding between particles in these clays. The high sensitivity of these soils has been ascribed to a high undisturbed strength due to the presence of cementing agents, such as carbonates, sesquioxides (amorphous or crystalline), organics, etc., and a very low remoulded strength due to the mineralogy and composition of particles. This paper will discuss the artificial cementation of Leda clay from eastern Canada. The experimental method used to achieve cementation in reconstituted samples of Leda clay will be described. The bonding mechanism has been investigated using scanning electron microscopy images and the effect of artificial cementation on the pore size distribution was studied by conducting mercury porosimetry. A series of oedometer and triaxial tests have been performed on three sample types: (1) artificially cemented, (2) reconstituted and (3) intact, natural samples to investigate the impact on the mechanical behaviour of the cementitious bonding. In general, the results show that formation of the bonds due to calcite cementation reduces compressibility and permeability, and increases the strength of these clay soils. The effect of the precipitation of calcium carbonate on the fabric and the resulting mechanical behaviour will be discussed in light of the geotechnical performance of sensitive cemented clays.



Soil improvement with vibrated stone columns - influence of pressure level and relative density on friction angle


Ivo Herle, Jimmy Wehr, Michael Arnold


The application of gravel for soil improvement with pile-like vibrated stone columns has attracted a considerable attention during the last decade. For the design of such a soil improvement, friction angle of the fill material (rockfill, angular or subrounded gravel) plays the crucial role in the calculation of bearing capacity of the stone columns and depends markedly on pressure level and relative density. Additionally, size, distribution and shape of grains, together with their mineralogical composition also influence the mechanical behaviour of the soil skeleton in the stone columns. In the contribution, a review of published experimental data on gravel and rockfill materials is supplemented by own test results. Shear tests with densely prepared specimens yield in most cases very high friction angles which lie above 50 degree at common normal stresses. Thus, conventional design values of 40 degrees are usually too conservative. Additionally, it is reported on direct and indirect measurements of density inside stone columns installed in situ and on small-scale model tests of stone columns in the laboratory. By exposing vibrated stone columns produced by the Keller method it could be shown that their in situ void ratio corresponds to the maximum density represented by e_min according to the German standard DIN 18126. Implications of the obtained experimental results for the design practice are discussed.



Numerical modelling of stone columns in soft clay under an embankment


T.M. Weber, M. Gäb, U. Vogler, S.M. Springman


A series of numerical calculations have been performed in order to investigate the behaviour of floating stone columns in soft clay under embankment at working load level. Three different approaches are studied more in detail using finite element calculations: idealisations in 2D and 3D and a homogenisation technique. Challenges in numerical modelling of stone columns include the estimation of the effect of smear around stone columns, which strongly influence the consolidation behaviour of the clay, as well as the consideration of construction effects due to stone column installation. Employing a 2D numerical model also needs the conversion of the spatial stone column grid into a 2D stone trench structure, which demands that an adjustment will be made of the soil parameters within the column grid. The homogenisation method also bears problems as it does not allow for considering all effects of consolidation analysis in an implicit way. One objective of this study is to establish how well the behaviour of an embankment on stone columns can be modelled in 2D or using the homogenisation method with less computational effort in comparison to a full 3D analysis. The results from geotechnical centrifuge tests of an embankment on soft clay improved with floating stone columns provide the basis for the numerical comparison described in this paper. The main characteristics of the embankment deformation, and the development of pore water pressures within the column grid, give an indication about the quality of the numerical models investigated.



Settlement predictions of embankments on organic soils in engineering practice


Dykstra, C.J., Mathijssen, F.A.J.M., Molenkamp, F.


Settlement predictions of embankments on organic soils in engineering practice Dykstra, C.J., Mathijssen, F.A.J.M.1,2), Molenkamp, F.2) 1) Royal Boskalis Westminster nv 2) Delft University of Technology CITG, section Geo-Engineering Key words: settlement predictions, embankments, organic soils, preloading, consolidation methods, isotache modelling, observational method, sample disturbance Abstract: The increasing number of Design Construct and Maintenance contracts in the past 5-10 years for large infrastructural projects in the Netherlands has resulted in a knowledge innovation in the field of applicability and reliability of settlement models in engineering practice. This paper focuses on the experience gained with two isotache type settlement models during the construction phase of some recent projects in the Netherlands. The a,b,c model based largely on the work of Den Haan (1994) using a natural strain formulation and a Bjerrum model mainly based on the work of Bjerrum (1967, 1973) and Yin and Graham (1989-2002) with both “linear strain” and “void ratio as state parameter” formulations, are compared. The theoretical background of both models and the main differences are briefly discussed, followed by a parameter determination from laboratory data and back analysis of parameters and reliability determination using field measurements. Application of these models during embankment construction is discussed for several projects. These embankments are typically built using a staged construction method with application of vertical drains for increasing the rate of consolidation, and preloading for reduction of the post-construction residual settlements. The back calculated values for consolidation and compressibility parameters clearly indicate that good approximations can be obtained while using the observational method in combination with relatively simple consolidation and compressibility models. Finally, various recommendations are presented for the improvement of the predictability of consolidation and settlement behaviour of embankments in engineering practice, after discussing various aspects like sample disturbance, finite strain consolidation and residual secondary compression after preloading.



Modeling piled foundation by means of embedded piles


E.G. Septanika, P.G. Bonnier and R.B.J. Brinkgreve


ABSTRACT: Recently, the embedded pile model - consisting of slender beams, skin and foot interfaces – has been successfully implemented in the Plaxis 3D Foundation Program. By utilizing the embedded pile model, the piles are being generated without affecting the existing mesh structure. The piles are added afterwards into the existing 3D finite element mesh, in which the pile can arbitrarily cross the soil element interior and not necessarily along the element edges. During the mesh re-generation, the embedded interfaces are also generated to model the skin interaction and the foot resistance. The interfaces represent both the stiffness and the strength of such pile-soil interaction system. Rigid and flexible connection or inelastic interaction can be modeled by choosing appropriate characteristic of interfaces. The skin interaction can be described by means of linear, multi-linear of soil-dependent traction models. The foot resistance is decribed by an elastoplastic-like spring. The results of the embedded pile model on single pile behavior, in both compression tests and pull-out tests, have been compared to the field test data. Further, the performance of embedded pile in modeling the pile group behavior is also evaluated. This paper considers a number of cases to demonstrate the 3D modeling and numerical capability of the developed embedded pile model.



AN EXPERIMENTALLY-BASED CONSTITUTIVE MODEL FOR DEEP-MIX STABILIZED QUICK CLAY


Bujulu, P. M., Grimstad, G., Sandven, R. B., Grande, L. O.


Quick clay changes, when remoulded, from a relatively stiff and brittle material to a viscous liquid mass, thereby losing its shear strength to less than 0.5 kPa. In the deep mixing process, quick clay is usually completely remoulded into a viscous fluid and then mixed with binder materials. The eventual engineering properties of the stabilized material depend on the fabric and particle cementation resulting from the chemical reaction of the binders. These factors are dependent on the nature of the original soil, type and proportions of the binder materials, mixing and curing conditions and the length of the curing period. Cement and lime are the common binders used for deep stabilization. Depending on the diameter of the columns and lime-to-cement ratio, a total of more than 75 to 150 thousand tons of cement are needed for deep stabilization in the Scandinavian countries per year. Thus, a very significant amount of money is used for this binder material annually. In order to check this cost, some industrial wastes in form of industrial slags and ashes have recently been introduced to replace a portion of cement, where this has proved to be technically, environmentally and economically feasible. Fly ash, silica fume and ground granulated blast furnace slag are the commonly used industrial wastes so far. A study is presently carried out at NTNU to investigate the potential of utilizing wastepaper sludge ash (WSA) for deep-mix stabilization of quick clays. WSA, which is a waste from paper recycling factories, is expected to provide an economical and environmentally advantageous substitute for cement in construction of lime-cement columns. Different mixing ratios of lime, cement and WSA have been studied at different curing periods. Uniaxial, triaxial and oedometer tests have been performed at NTNU at 7, 28, 56, 90 and 180 curing days. WSA has proved the potential to replace cement in deep-mix stabilization of quick clay. Based on these results a constitutive model for stabilized quick clay has been proposed. The model takes into account the actual mechanical properties of the stabilized material. Key words: Quick clay, deep-mix stabilization, wastepaper sludge ash, constitutive model



Comparison of vibrocompaction methods by numerical simulations


Michael Arnold, Ivo Herle, Jimmy Wehr


Densification of loose granular soils with vibrocompaction belongs to the state of the practice today. Nevertheless, the role of various technological factors, like direction, amplitude and frequency of vibration, is still not well understood. Numerical simulations represent a cost-effective tool for testing different vibrocompaction setups provided that realistic numerical and constitutive models are applied. The paper describes the results of 3-D simulations using an approach combining a nonlinear elastic model within a dynamic FEM with hypoplastic numerical element tests. The shear stiffness degradation in the elastic model follows from the hypoplastic constitutive model for sand with the small-strain stiffness (intergranular strain) extension. With respect to the compaction efficiency, three zones can be distinguished. In the first zone closest to the vibrator, high shear strains produce dilatancy which prevents a maximum possible density. In the intermediate zone, an optimum densification is achieved. An extent of this zone depends on the way of shearing, i.e. on the type of vibrator. If leaving other parameters unchanged, a deep vibrator generates a larger compacted region than a top vibrator. In the far zone, the shear deformation does not noticebly surpass the quasi-elastic range of the soil behaviour and, thus, does not produce sufficient volumetric changes.



3D NUMERICAL MODELLING OF THE BEHAVIOUR OF LARGE BUILDINGS FOUNDED ON MEXICO CITY SOFT CLAYS


J.F. Rodriguez, N.P. Lopez and G. Auvinet


Analytical or two-dimensional numerical models used to represent soil-structure interaction tend to oversimplify the real conditions of the problem and cannot be expected to provide really reliable results. Recently, powerful and user-friendly three dimensional (3D) numerical tools have been developed. In this paper, two cases of large buildings founded on Mexico City soft clays that were analysed with 3D models are presented. The first case refers to a group of nine large buildings arranged in a 32,000m2 rectangular area, located in the urban area of Mexico City in the so-called “lake zone”. The area is affected by regional subsidence induced by intense pumping from deep sand strata for water supply. Due to spatial variations of the thickness of the compressible layers, differences in buildings geometry, dates of construction and service loads of each structure, the group has presented differential settlements which exceed the acceptable limits set in the local building code. The complexity of the problem and the importance of the buildings, led to the implementation of a 3D finite element model which allows taking into account the above factors and simulate the past behaviour of the buildings group as well as the expected effects of a rigid inclusions underpinning system. The second case is about an 86m high tower constructed over a 7m deep partially compensated box-type foundation combined with 156 friction piles. Since construction in 1964, important differential settlements have been observed. Tilting of the tower is about one meter. As part of a rehabilitation project, it was consider necessary to check the stability of the structure in static and seismic conditions performing a 3D finite element analysis simulating the past behaviour of the tower and surrounding structures. The contribution to stability of a group of special “control” point-bearing piles that will be added to the present foundation system was also evaluated.



Modelling of floating stone columns beneath an embankment using different constitutive models


Martin Gaeb, Daniela Kamrat-Pietraszewska, Helmut Schweiger, Minna Karstunen


Constructions on soft soils have become a common necessity within the last decades. Stone columns are often an effective solution to keep (differential) settlements within acceptable limits and to guarantee the bearing capacity of foundations. Still, there is a lack of knowledge of the interaction between soil and the inclusions. Further complication is that due to the installation of the columns, there will be a significant change in soil properties of the improved soil, e.g. change in permeability and change in stiffness. Presented are various simulations that deal with a field trial in Klagenfurt, Austria. Different measuring devices have been used to investigate the performance of a floating stone column foundation. These mainly include measurements of pore water pressure and settlements at different depths. Therefore, they provide a basis for back calculations that can lead to a better understanding of the processes discussed above. The simulations are carried out using advanced constitutive soil models which allow taking into account anisotropy, destructuration and creep. These models were developed during the EC-funded SCMEP and AMGISS projects. It is of interest how these models behave in comparison to a standard linear-elastic perfectly plastic Mohr-Coulomb model. Furthermore, the results will show if the models are able to perform well with respect to field conditions.



Numerical modelling of deep mixed columns below embankments on soft soils.


Harald Krenn, Minna Karstunen


Construction of railway or motorway embankments on soft soils is always difficult due to large settlements and stability problems. In many cases, the required construction and design criterions cannot be achieved without additional ground improvement. The advent of deep soil stabilisation or often referred to as deep mixing method in the 70’s has given tremendous impetus to the construction projects in areas with soft soils. Deep mixing is a soil improvement technique in which lime and/or cement are mixed in-situ to create a columnar reinforcement in the soft soil layers. The installation of the columns will reduce settlements and increase the bearing capacity of the ground and as consequence reduce stability problems for embankment constructions. Numerical analysis in combination with advanced models soil models for soft soils and the deep mixed material gives an insight in the complex interaction between soil and column. Current practice in numerical modelling is to use a unit cell, which allow just to model a single column and the surrounding soil or plane strain analysis where the columns are modelled as wall elements. In addition only simple constitutive models are used which don not represent the observed stress-strain-strength behaviour of the deep mixed materials and the soft soil. Inherent and plastic strain induced anisotropy is an important characteristic of soft soil which will influence the predicted settlement behaviour of the systems. In addition to anisotropy , natural soils exhibit bonding between the particle. The bonding can be progressively destroyed during plastic straining and induce additional settlements to the soil-column system. Additionally to the realistic soil and column behaviour also the true geometry of the problem has to be accounted for. This can be done by using three dimensional numerical models.



Some experiences from full-scale test embankments founded on lime-cement columns not reaching firm bottom


Olsson, M. Edstam, T. Alén, C.


In the Göta River valley in the western part of Sweden 80 km of a new motorway and a new high speed railway will be constructed during the period 2007-2012, at a cost of approximately 1.1 milliard €. The ground conditions mainly consists of deep layers of soft, high-plastic clay which was deposited in marine conditions some 10.000 years ago. Embankments have to be constructed for the main part of both the motorway and the railway. In order to reduce the settlements and improve the stability conditions of the embankments, lime-cement columns will be used extensively, i.e. some 9 million linear metres of columns will be installed. Already during the feasibility studies, in year 2000, a large amount of test-columns were installed in order to study their geo mechanical properties (strength, stiffness, permeability). Furthermore 3 full-scale test embankments, founded on lime-cement columns not reaching firm bottom (i.e. “floating columns”) were constructed in year 2001. The test embankments were considered necessary due to lack of reliable design methods for estimating consolidation settlements of embankments founded on floating columns. The embankments were heavily instrumented in order to study the settlements of the embankments. Measurements were made of the variation of the settlements with depth in the columns, in the clay between the columns and in the clay below the columns respectively. Additional measurements included pore pressure changes in the clay below the embankments as well as horizontal displacements in the clay just outside the embankments. In order to evaluate the results of the measurements, engineering judgement and 1D analytical analyses were combined with 1D, 2D and 3D FEM analyses. Based on such analyses, simplified design methods for calculating consolidation settlements were developed to be used within the project. Furthermore more general, yet relatively simple, design methods were developed in order to be used for more general conditions of lime-cement columns not reaching firm bottom.



EXPERIMENTAL AND NUMERICAL ANALYSES OF PULL-OUT RESISTENCE OF NAILS DRIVEN IN PYROCLASTIC SILTY SAND


G. Di Fonzo, A. Flora, G. Manfredi, M.V. Nicotera, A. Prota


Soil nailing is a ground improvement technique that, in recent years, has been increasingly used all over the world for the support of excavations and slopes. Nailing is done by closely spaced passive inclusions (nails) having tensile, shear and bending resistance. These passive elements are set up directly driven into the soil or grouted in pre-drilled holes and then possibly connected to a facing. Even though the technique has proven to be safe and cost effective and a number of papers has been devoted in the last thirty years or so to the comprehension of soil-nail-facing interaction, there is still a lack of information on some of its basic features. As a consequence, design practice is often oversimplified and essentially based on limit equilibrium methods for the definition of a global safety factor (Gassler & Gudehus, 1981; Shen et al.,1981; Schlosser, 1982; Elias & Juran, 1991). For the assessment of nails behaviour, the calculation of a local safety factor with reference to a possible pullout mechanism must be added. This is not an easy task too, as the estimate of a nail working load and of the true behaviour at the soil nail interface is far from being a routine calculation. In design practice, a rather frequent alternative to this simplified analytical approach is the use of FEM or DEM codes for the assessment of both global and local stability. However, the use of a numerical (typically 2D) approach in design is rather ambiguous in this case, as it basically fails to take into account some inherently 3D basic features of the technique, like soil-nail, nail-nail, and nail-facing interactions. Recently, at the University of Napoli Federico II a comprehensive research programme was started to study soil nailed structures in unsaturated pyroclastic silty sands, which can be used as an effective mean to tackle some of the typical local geotechnical hazards. The programme includes tests on instrumented nails in a large pullout apparatus with the local measurement of suction, site testing to failure on a large field trial, 2D and 3D numerical analyses. This paper will first show the results of laboratory pull out test under different testing conditions, showing the influence of water content and confining stresses. Then, 3D numerical analyses carried out to simulate the pull out tests will be presented, showing the relevant influence of nail-soil interface properties on the overall results.



Cambuslang Investment Park - Construction on Clyde Alluvium


Alan Swanson & Iain Clow


Phase 2 of Cambuslang Investment Park involved the development of an area in the south east of Glasgow, approximately 6.5km from the City centre and 0.5km to the south of junction 1 of the M74 motorway. Development proposals included commercial and industrial premises comprising warehousing, production, office, utility buildings and associated roads hardstandings and car parks. A large part of the area was inundated during the major flood of 1994 and in order to reduce flood risk to an acceptabe level it was proposed to upfill the site by up to 6m in order to create raised building platforms. High floor loadings and significant depths of upfill raised the possibility of excessive settlements due to the consolidation of the compressible alluvial soils underlying the site. In order to reduce risk and satisfy funders requirements for low cost foundation solutions it was proposed to surchage the upfilled platforms to simulate floor loadings and improve the properties of the underlying superficial deposits. Extensive site investigation works were carried out to confirm the nature and extent of the Clyde Alluvium and allow determination of the engineering properties of the main soil groups. Examination of U100 and piston samples revealed excessive sample disturbance casting doubt on the reliability of the geotechnical parameters provided and as a result, analysis concentrated on the cone penetration test (CPT) profiles and back analysis of previous preload trials. As part of the earthworks, monitoring was installed, including over 40 vibrating wire piezometers and 100 rod settlement gauges, at original formation and platform level. Through analysis of the monitoring data, settlement periods of 5 to 7 months were predicted for the anticipated total settlements, ranging from 30mm to 500mm. URS, through the Earthworks Contract ensured that the works were carried out in accordance with the drawings and Specification, controlled the quality of the earthworks import and monitored ongoing settlements and pore water pressures therefore allowing the rolling reuse of surcharging materials and the rapid release of buildiing platforms for subsequent development



A review of field trials investigating the performance of partial depth vibro stone columns in a deep soft clay deposit


Colin J. Serridge


There is increasing pressure to develop marginal, deep soft clay sites within the UK, using more economical techniques than the traditional approach of piling, particularly for low rise development/s and where some settlement can be tolerated. Ground improvement techniques provide potential and practical alternatives to piling. Of the various techniques attempted to date, the most commonly used technique is the technique of “reinforcing” soft clay soils with (vibro) stone columns. Historically, the wet-top feed (vibro-replacement) technique of stone column installation has been used in soft clay soils, particularly where these are below ground water level, but environmental concerns regarding disposal of effluent generated, advances in vibro technology and economic constraints have resulted in the wet-top feed technique being largely superseded by the dry bottom-feed (vibro-displacement) technique. The limited published field data relating to vibro stone columns in soft clay soils have concentrated on the wet top-feed technique. This paper therefore provides an update of an on-going programme of research using the large database of information obtained during an instrumented vibro stone column field trial, (utilising the dry bottom-feed method), carried out at the former soft clay research site at Bothkennar in the Forth valley, Scotland. Whilst much useful laboratory and numerical based research has been undertaken on stone/granular columns in soft clay soils over the last 10 years or so, there exists the inherent issue of scale effects and of being able to compare the results with actual field data. The main aims of the field trials at Bothkennar were to investigate ground response to installation of partial depth (partially penetrating), vibro stone columns in a deep soft clay soil deposit using the dry bottom-feed method together with the effects of stone column length (depth), spacing and a thin surface “crust” (founding depth), on the behaviour of strip/pad foundations constructed over the installed trial stone columns. The data has highlighted important implications with regard to the way we approach both stone column installation and design, for partial depth vibro stone columns in deep soft clay soil deposits.



INFLUENCE OF STONE COLUMN DEFORMATION ON SURROUNDING SOIL CONSOLIDATION


Jorge Castro and César Sagaseta


An analytical closed-form solution for the radial consolidation around stone columns is presented. The solution considers the influence of vertical and radial deformation, either with elastic or elasto-plastic behaviour. A multi-step load is applied. As other previous solutions, initially developed for rigid column, or including only vertical deformation, the solution is obtained for a “unit cell” in axial-symmetry using an average excess pore pressure in the soil along the radius. The elastic and plastic deformations of the column lead to equivalent coefficients of consolidation for the radial flow, which enables the application of the available methods of integration of the consolidation equation. For elastic column, the solution gives the variation of strains and stresses between the undrained and final states, for which it coincides with the existing elastic solutions. Plastic strains in the column are modelled with Mohr-Coulomb yielding criterion and a non-associated flow rule, with a constant dilatancy angle. A parametric study shows the influence of the main problem features. Because of the radial deformation and the yielding of the column, the stress concentration and the settlement reduction factors are similar to empirical values. Consolidation is in good agreement with numerical solutions. A design example is used to illustrate the application to practical cases.



Experimental analysis in laboratory of radial consolidation around a stone column


Cimentada,A. and Da Costa,A.


Stone columns are a method of soil reinforcement commonly used in soft soils for foundation of road embankments. These columns increase the bearing capacity of the foundation as well as reduce and accelerate the settlement. This paper deals with an experimental study in laboratory in which samples of kaolin (diameter 254 mm, H¡Ö70mm) are reinforced with a single totally penetrating stone column. Two different diameters of column are chosen for testing in order to know the influence of this parameter in the behaviour of the treatment. The diameter ratios are 3 and 4 corresponding to a column diameters of 84.67mm and 63.5mm respectively. The scale factor is about 1/10 in the designed model. Columns are installed by freezing them before inserting into a drilled hole in the center of the consolidated soft soil. The specimens are tested in a Rowe-Barden cell applying several steps of load under known boundary strain and drainage conditions. Tests with samples of kaolin without column have been also carried out under the same conditions in order to compare the behaviour of reinforced and non-reinforced soil. This research is conducted to investigate the radial consolidation process around the column by means of obtaining the equivalent consolidation coefficient, the reduction and development throughout the time of the settlement and the load distribution between column and soil. A suitable instrumentation has been designed for measuring of pore pressures and total stress during the steps of load in order to achieve the proposed goals. The results have been compared with existing analytical solutions.



Evaluation of the effect of face bolting on tunnel construction using X-ray CT


Daiki Takano and Jun Otani


In recent years, mountain tunneling method such as NATM (New Austrian Tunneling Method) has been used for tunnel construction even in the urban area. In Japan where land area is restricted, it is required that tunnel structure is constructed even in unconsolidated ground or low overburden ground. When excavating a tunnel, tunnel face becomes very unstable due to stress release caused by excavation. It is a main issue to maintain the stability at tunnel face during construction. To create the condition for safe construction, it is necessary that tunnel face is reinforced by any auxiliary methods (i.e. face bolting method or fore poling method). In order to evaluate the mechanism of tunnel face failure, authors have developed a system of tunnel pull-out model test that could be carried out in the system of X-ray CT scanner. And using this system, the failure patterns due to tunneling without reinforcement and with face bolting method were visualized using X-ray CT and it was observed that face bolting method prevents the progress of the failure zone due to tunneling. The purpose of this paper is, firstly, to simulate four cases of the failure patterns which are without auxiliary method and face bolting method with three different length of the bolts due to pulling out tunnel model from the ground; and secondly, to investigate the behavior of model ground using X-ray computed tomography (CT) scanner in order to visualize the failure zone in three-dimensions. Finally, the effect of auxiliary methods is evaluated. Conclusions of this study are summarized as follows; (1) The effect of face bolting methods can be investigated visually in three-dimensions using X-ray CT scanner; (2) For the case of face bolting method, the failure can be prevented fairly well in not only front of the tunnel face but also up to the ground surface, and (3) The increment of the volume of failure zone depends on the length of face bolts.



Automated ground improvement process


Leena Korkiala-Tanttu, Markku Juvankoski and Petri Valasti


In the last years the Finnish infra structure research has focused on the automated construction systems. POHVA project concentrates on the automation of the ground improvement process. POHVA project is divided into two subprojects. The first ongoing phase started in spring 2005 and will end in December 2007. The objective of the first phase is to develop an interactive 3D model for deep mixing and piling including site investigation methods, data processing and analysing methods which meet the needs of automated production management system. Another objective is to develop resistivity sounding method for deep mixing purposes. The first phase created readiness for the more economic, ecoefficient and better performance deep mixing method. By refining the 3D water content space, which has been created from the resistivity sounding results, to the strength estimation of the deep mixing the amount of binder agent can be optimized along the columns. So the homogeneity of the deep mixing can be improved. This together with the more exact placement of the columns achieved by machine automation will improve the quality and performance of the product. The second phase has just started and it will concentrate on the applying information and automation technology into piling and deep mixing methods.



Construction of a Lightweight-Dam on Organic Soils – FE-Analyses and Practical Experiences


Hartmut Schuller, Peter Krameter


A former bog next to Lake Ossiach in Carinthia (Austria) with an area of approximately 45 ha, which has been drained for some decades for agricultural reasons, will be flooded again. A dam with a length of approximately 400 m and a height of up to 1.5 m is required to protect the adjacent agricultural areas from flooding. Subsoil investigations revealed a 3 to 8 m thick layer of peat below the ground surface. Underneath, sandy to clayey silts in weak to stiff consistency, partially with organic contents, have been found. The groundwater table is located close to the ground surface. Deeper soil layers contain confined water, partially with artesian groundwater tables. In the stage of dam design, finite-element analyses have been performed for a prognosis of settlements, porewater pressures and time dependent behaviour of the organic and fine-grained soil layers. These calculations using PLAXIS with different soil models resulted in relatively large deformations, predominantly within the peat layer. During construction of the dam, unexpected large settlements occurred in one section of the dam, while settlements in the other areas remained more or less within prognosis. The reason for this seems to be local in-homogeneities of the peat layer which could not be recognised by ground investigation. Therefore, it was not possible to continue the adopted construction concept. Extensive finite-element calculations have been performed to find a solution for dam construction in a technically and economically feasible manner. As a consequence of these studies, a lightweight fill material (blocks of expanded polystyrene EPS) was used for the body of the dam. A layer of well-graded granular material on top of the EPS guarantees the required safety against uplift. Construction of the dam according to the new concept has been finished in July 2007. Since then, settlements are measured at regular intervals and compared to the prognosis. This enables not only a review of results of the FE-analyses but also an evaluation of long term settlements in peat and organic soils.



CONSTRUCTION OF AN EMBANKMENT ON AND WITH AN IMPROVED SOFT INDUSTRIAL SLUDGE


P.H. YONATAN, J. TENNEKOON, D. VAN GEMERT, J. MAERTENS, J. HOUTMEYERS & M. GOORDEN


The behaviour of an industrial sludge, arising from fertiliser and feed phosphate production, has been found to be similar to that of clayey soil. Prior to its use as a construction material for an embankment, the soft sludge has been improved by means of a filter press chamber to reduce the water content and, thereby, to obtain a stiffer material. Filter pressing has yielded relatively stiffer sludge filter cakes with a water content of less than 80%. However, the behaviour of the dewatered sludge (or filter cakes) is relatively unknown which is of importance when considering it as a construction material. Extensive laboratory tests have been conducted in order to characterise the original sludge and the filter pressed cakes. From the results of laboratory tests, similarities as well as differences have been found in correlation with naturally occuring clayey soil. One of the most important findings from the laboratory tests, is that laboratory shear tests on fresh dewatered sludge have shown the material to be cohesive but under drying and re-wetting cycles, its behaviour changes dramatically. The cohesion reduces to practically zero and the angle of internal friction increases to in excess of 30°. These properties are especially important when considering the construction of embankments. For that reasons, the construction of an embankment with a size of 70x190m and 10m height, can cause significant deformation and result in stability issues both in the embankment and in the foundation. This paper deals with the assessment of the sludge and the dewatered sludge filter cakes as well as the behaviour of an embankment constructed with and on the improved industrial sludge. Deformations and settlements of the embankment are being monitored since 2001 by means of hydrostatic settlement profile, and compared with numerical models, showing reasonable agreement.



Design and research activities related to the enlargement works in two major harbours in Italy


J. Persigilli, P. Ruggeri, E. Sakellariadi and G. Scarpelli


Construction works in poor ground conditions are often performed without the support of reliable theory and models, but rather confiding in empirical rules and the contractors\' experience. On the other hand, the resources necessary for developing adequate theory and models are usually scarce. Typically, access to information and the necessary data to perform useful research is provided through the possibility to participate in construction projects requiring the formal support and backing of some research institution, while the main interest of collaborating contractors and authorities is the finished work. In such a context the Authors are involved in two major construction projects regarding the works for enlargement and development of the harbours of Ancona and Ravenna, both situated on the Adriatic coast of Italy. In this paper the progress of the resulting research experience will be presented. In the case of the Ancona harbour, the project regards an extension of the breakwaters of about 1 km to sea. The seabed, at about 16 m depth, consists in a series of soft soil layers, becoming a muddy soil of extremely poor quality for a length of about 200 m. The main difficulties encountered in the original project were the estimation of the quantity of material required for the construction, and the determination of material properties for settlement prediction. It was therefore decided to construct an adequately instrumented test embankment. Data acquired will include settlements, pore water pressures, and sub bottom profiling for evaluation of penetration in the mud sea bed; this data will be interpreted and compared in numerical analysis and predictions, with the aim to arrive at an adequate model for performing design analysis. In the case of the Ravenna harbour, the original project regards an improvement of the mechanical properties of the subsoil in order to allow the deepening at the foot of existing or new quays. Moreover, since the structural design involves anchoring of sheetpile walls with high levels of anchor forces, anchorages with a grout body will be used. The details of the preliminary testing campaign are at this moment being discussed. To this end a test site is being designed for a 500 m length of quay. Within this context, some test loading has already been executed and measurements have been obtained, which present some interpretation problems. The Authors are taking part in the analysis and interpretation of this data, the outcome of which shall be reported in the paper.



Numerical analysis of embankments on soft soils incorporating vertical drains


José Leitão Borges


In spite of all experience obtained over the last decades, undertaking embankments on soft soils still collocates diverse and delicate problems that are determined by the poor geotechnical characteristics of the foundation soils: (i) low shear strength significantly limits the load (embankment height) that is possible to apply with adequate safety for short term stability and (ii) high deformability and low permeability determine large settlements that develop slowly as pore water flows and excess pore pressure dissipates (consolidation). The use of vertical drains in embankments on soft soils is one of the most adequate techniques when the main purpose is to accelerate the consolidation. In the paper, the three-dimensional behaviour of an embankment on soft soils incorporating vertical drains (geosynthetic band drains) is analysed by a numerical model based on the finite element method. The computer program incorporates the Biot consolidation theory (coupled formulation of the flow and equilibrium equations) with constitutive relations simulated by the p-q-theta critical state model. Three dimensional coupled analyses are performed. Special emphasis is given to the analysis, during and after the construction period, of the excess pore pressures, settlements, horizontal displacements and stress levels. In order to analyse the effect of the use of vertical drains, the same problem without drains (two dimensional analysis) is also performed. In spite of its three dimensional behaviour in terms of stress and water flow, the results show that settlements at the embankment base with vertical drains are approximately uniform along longitudinal direction, as shown by field measurements. This is justified by the existence of “arching effect” inside the fill, which is expressed by the application of a non uniform vertical load on the foundation surface, as shown by the numerical results.



On the difficulty of characterizing the properties of mixed materials obtained from soil-cement columns


Alain LE KOUBY and Iliya PETKOVSKI


Within a railtrack project, soil-cement columns were built in a clayey silt soil. Different diameters (400 and 600mm) , w/c (1 and 1.2) ratios and density of cement (between 200 and 300 kg/m3) were chosen. Load testing have been carried out on the columns in order to get their bearing capacity. Subsequently, excavations of the loaded columns and of test columns have been done and lumps samples were taken in order to investigate the mechanical behaviour as well as the physical and chemical properties of the soil-cement material. The first difficulty to cope with was the heterogeneity of the soil-cement material resulting from the mixing method. Only one top down step was achieved. Laboratory tests were carried out. The sampling of the lumps of mixed material was very complicated as some pieces were too soft due to the lack of mixing. Mechanical tests aimed at determining an average behaviour of the soil-cement material. Unconfined compression tests have been done to give some ideas of the resistance of the mixed material and to compare it with the cement material and a homogeneous material which has been mixed in the laboratory. Static and dynamic Young modulus were also evaluated. Oedometer tests were also made to show the evolution of the settlement coefficient with the amount of cement. In addition, some investigations on the Scanning Electron Microscope (SEM) analyses were carried out on treated and untreated soil to investigate the shape of the mixed material and of the pores. Permeability tests also showed interesting results with the varying amount of cement..



Influence of curing temperature on the strength of cement stabilised artificial clays


I. P. Marzano, A. Al-Tabbaa and M. Grisolia


As the mechanical properties of most cement-stabilised soils changes over time (Porbaha et al, 2000), developing an understanding of the time-related performance of such treated soils is essential in understanding their durability and long-term effectiveness. Accelerated ageing techniques are commonly applied for this purpose. On such technique is the use of elevated temperatures to speed up the hydration processes of the cement (Fuessle and Taylor, 1999). The paper will present results of a laboratory investigation on the influence of curing temperature and time on the mechanical characterisation of cement stabilised clays as used in deep soil mix applications. Numerical modelling is also presented, which was applied to experimental data in order to develop a prediction of the longer-term performance of the cement stabilised clays. A number of models have been developed to predict such time-related performance, one such model being the maturity concept (Carino, 1984) which involves use of the ‘global’ activation energy, which derives from the Arrhenius equation, and is applied using simple graphical techniques as reported by Chitambira et al (2007). Experimentally, this model was applied to a series of tests performed at different temperatures, namely 10, 20 and 40°C. For each temperature, unconfined compressive strength tests were performed at specific time intervals in order to follow the curing and the increasing of strength with time. The clay used was an artificial clay made up of 40% Kaolin and 60% of silt (using silica flour). This clay had a Plastic Limit of 20.0% and a Liquid Limit of 35.6%, classified as CL according to the Casagrande Plasticity Chart. Three different water contents of 50, 40 and 33% were used for this clay in order to compare the performance of different soft clays. Portland cement was used as the wet binder at a 10% addition by weight and was applied at a water:cement ratio of 1. References Carino N. J, (1984). The maturity method: Theory and Application, Journal of Cement, concrete and Aggregates, ASTM, 6(2), pp. 61-73. Chitambira, B. Al-Tabbaa, A., Perera, A.S.R. and Yu, X.D. (2007). “The activation energy of stabilised/solidified contaminated soils”, Journal of Hazardous Materials. Vol. 141(2), pp 422-429. Fuessle R., and Taylor M., (1999). Accelerated Aging of stabilized hazardous wastes, Report No. RR-82, Waste Management Research Centre. Porbaha, A. Shibuya, S. And Kishida, T. (2000). State of the art in deep soil mixing technology. Part III: geomaterial charaterization. Journal of Ground Improvement, Vol 3, pp 91-110.



Numerical modelling of small-scale geogrid encased sand column tests


Joel Gniel, Abdelmalek Bouazza


The use of stone columns as a ground improvement technique in soft cohesive soils is increasingly being extended to sites with poorer conditions. This is being achieved with the use of geosynthetic reinforcement which acts to provide additional lateral support to columns, preventing excessive bulging and column failure. Although the use of geotextile encasement has been investigated and implemented on numerous projects throughout Europe, research into the complex reinforced behaviour is ongoing. In addition, the concept of using other geosynthetic materials such as geogrid for column encasement is relatively new. In order to investigate the effect of geogrid encasement on stone columns, small-scale laboratory testing of model sand columns was undertaken. In conjunction with this, a numerical modelling study was undertaken to further understand the interaction between the geogrid, column material and surrounding soil. Particular emphasis was placed on comparing the behaviour of partially encased columns to fully encased columns. The behaviour of single columns was also compared to column groups. This paper describes the finite element modelling that was undertaken to replicate model sand column behaviour and predict the performance of further laboratory testing. Results of the numerical study indicate good correlation with the observed behaviour in laboratory tests.



Evaluation of Ground Improvement by Groups of Stone Columns using Field Measurements and Numerical Analysis


Fabian Kirsch


Vibro Stone Columns are widely used to improve the load-settlement characteristics of soft soils either as an infinite pattern under wide spread loading or as a column group beneath shallow foundations. The design is usually based on analytical and seni-empirical procedures such as the methods of Goughnour-Bayuk or Priebe. The paper to be presented deals with groups of vibro stone columns under vertical loading. Since the overall behaviour is depending on the interaction of the stone column material with the surrounding soil it is important to correctly address the influence of the column installation. The results of multiple field measurements of the stress and stiffness modification due to the vibro stone column installation are presented. Additionally the results of extensively instrumented load tests on groups of five and four columns respectively are presented. These results together with the results of model tests are used to calibrate a true 3-D numerical model, which is further on used to examine amongst others the following aspects: Modelling of installation effects, Influence of the constitutive model, Area ratio and improvement factor, Influence of column length and end bearing situation, Load distribution amongst the columns and influence of plate stiffness, Vault formation in distributing layers. Additionally some considerations on single column load testing are given. The paper concludes with an outlook on current research activities and future lines of development.



Reliability-Based-Design Application to Slopes Stability Analysis


Selmi Mbarka, Ltifi Mounir, Hassis Hedi


The landslides are generally spectacular; they occur in very varied circumstances and affect natural as well as artificial slopes. Since decades, the slope safety estimate against rupture was done in safety factor terms which is, in fact, a precaution against our real ground properties ignorance. Currently, we attend a new approaches development allowing to treat qualitatively the natural soil variability. Such approaches aim to take into account all uncertainties that affect the input parameters used in the stability mechanical design. In this context, the probabilities application to the slope stability analysis has acquired for a few years a growing interest. It seems to be the precious tool permitting to describe the spatial variability of soils properties and, therefore, to have a more realistic soil modelling with a high objectivity degree. Several applications of a reliability-based-design to slopes stability conclude that the critical circle can be determined precisely using the reliability index concept. We will present in this paper a reliability approach application to the slope stability analysis. The spatial variability of soil properties is taken into consideration by using Gaussian random fields. The reliability measure is the Cornell reliability index which is calculated by applying the Monte Carlo simulation. First results showed that this can be an alternative approach which can provide a better performance in the slope stability analysis. Key Words: slopes stability, reliability-based-design, spatial variability, random fields, reliability index



ANALYSIS OF THE AGEING EFFECTS ON THE PROPERPTIES OF A SILTY SOIL


Mounir Ltifi 1, Jean Paul Tisot 2


The aging of soils is a widely known and studied phenomenon which practical importance not unnoticed. However, compared to other soil properties, this is far from recognized in technical literature, and in practice, it is often forgotten because when doing so, it remains on the safety side. There are numerous examples in the literature of time-dependent changes in the properties of soils, or aging effects. Most of these aging effects are of increases in the cone penetration resistance. Time-dependent increases in penetration resistance have been measured in hydraulically placed fills and freshly densified deposits, with the largest increases following the use of ground modification techniques such as vibrocompaction, dynamic compaction, and blast densification. It is not known what causes these increases in penetration resistance to occur. The objective of this research was to gain an understanding of the possible mechanisms responsible for aging effects in silty soil. A laboratory testing program was developed to study the influence of different variables on the presence and magnitude of aging effects. In this fact, we have carried out classical shear tests and permeability tests on Xeuilley silt samples. Xeuilley silt is located southwestern Nancy, France. The time effect was considered at the level of the aging time after compaction with constant water content. Undrained shear tests and triaxial permeabilty tests were carried out on samples prepared at different water contents. Results of these tests showed an increase in the undrained shear strengh and deformation modulus related to a decrease in the failure strain as a function of the time. This is probably due to the mechanisms witch condition undrained behavior of material during aging. Finaly, permeability tests show that the intrinsec permeability tends to decrease according to the duration of storage.



Centrifuge modelling to compare ground improvement techniques on double porosity clay landfills


Emma Pooley, Sarah Springman, Jan Laue


Open cast coal mining in Northern Bohemia produces overburden waste of overconsolidated clay lumps that are placed in spoil heaps or used as backfill in exploited mines. This forms landfills with a double porosity structure consisting of voids between lumps (inter-granular) and voids within lumps (intra-granular). The total porosity can be up to 70 %, making the soil highly compressible even after self-weight consolidation. Soil behaviour is characterised by large absolute and differential settlements, necessitating ground improvement before construction. Methods of soil improvement used in practice on lumpy clay landfills typically involve either the installation of sand compaction piles, or in-situ densification through preloading or heavy tamping. Model tests on scaled double porosity clay landfills were carried out in the ETH Zurich drum centrifuge. A period of self-weight consolidation, during which pore pressures and surface deformation were monitored, was followed by the application of ground improvement. Three ground improvement techniques were tested: sand compaction piles, heavy tamping, and the use of a geotextile. Sand piles were installed in-flight and a variety of geometries were modelled. Heavy tamping was also carried out in-flight. In order to assess and compare the strength of the clay fill, in-flight displacement-controlled rigid foundation tests were carried out on improved and unimproved soil models. The total load on the foundation, and pressure distribution under it, were measured. Load-displacement and pressure distribution data from the centrifuge foundation tests were analysed. This was used to compare the effectiveness of sand compaction piles, heavy tamping, and geotextiles as ground improvement techniques for double porosity clay.



Mechanical properties of clay stabilised with conventional and novel binders


S.Jegandan and A. Al-Tabbaa


This paper presents part of study which focuses on the applicability of conventional and novel binders in Deep Soil Mixing (DSM) applications. The objective of the study was to investigate partial or complete replacement of Portland cement (PC), which has been used as a conventional binder in most applications regardless of the design criteria or its environmental effects. In this work, industrial by-products such as ground granulated blast furnace slag, magnesia, both reactive and dead burned,, pulverised fuel ash, incinerator bottom ash and cement kiln dust were considered as possible replacements for PC either partially or completely in various combinations. Laboratory experiments were carried out to assess the performance of these binders in terms of the mechanical properties of stabilised clay. Artificial clay consisting of 40% kaolin and 60% silt (silica flour) with 50% water content was used, which has a liquid limit of 35% and plastic limit of 30%. The grout dosage and water: binder ratio in the grout were also considered and varied in the experiments. Performance of the binders was assessed in terms of the unconfined compressive strength (UCS), static stiffness (E50) and strain at failure (€f). Furthermore, the microstructure, using scanning electron microscopy (SEM) and x-ray diffraction (XRD) tests, was used in order to understand the chemical reactions taking place between binders and soil and to link changes in the microstructure to the mechanical properties measured. The figure below show a set of the mix samples prepared.



The Behavior of Pile-Net Composite Foundation


Jie-jin Chen ,Jian-dong Niu


Pile-net composite ground is an important method for ground improvement, which comprises both vertical piles and horizontal ¡°net¡±¡ªgeosynthetic. In this kind of ground, net, piles and soil work together and perform compatibly. Many advantages can be found in it: small settlement deformation, good stability, short construction period and convenient construction, etc. However, because of the complicated action mechanism, there is not a proper design method for it except recourse to experience. Bearing mechanics of pile-net composite foundation involved in complex interaction of embankment fills, reinforced cushion, pile and soil. With a in-situ test of certain high-speed railway trial embankment in Shanghai City, China, pile and soil settlement deformation, the stress and strain of geogrid and contact pressure were monitored. The change process in settlement and the bearing properties of geogrid and pile-soil stress ratio have been analyzed.The pile-soil stress ratio is an important parameter for design and calculation of pile-net composite ground, so it is very significant to research on it. Based on previous research, a calculation formula for pile-soil stress ratio of pile-net composite ground composed of sand piles and sand cushions beneath which one-layer geosynthetic is embedded is proposed. The comparison with observed data in field shows that the calculated results are relatively accordant with practical results. This formula can be the reference in design and the basis of farther research in the future.



ARS Pile, Proposal of Procedures


Arie Sriyono


Micro piles have already known as one of solution of ground improvement to build upper structure above soft soil. Bamboo micro piles in many years have proven it is effectiveness in many traditional to modern building in Indonesia. One of micro pile design is found by Dr.Ir. A.R. Soehoed. The design has been proven in several development sites such as Pantai Mutiara Project. Being in very soft soil with high liquidity, the A.R. Soehoed design of micro pile (ARS Pile) could bare heavy rubble mound sea dykes together with bamboo raft layers. The consolidation process is the key to the ARS Pile successful ground improvement. Using vertical drain that attach to the pile reduce the pore pressure. Nevertheless the mathematical process of calculating the ARS Pile yet not been agreed. This paper will show a proposal of calculating procedures and conducting in situ measurement of ARS Pile. The mathematical procedure in this paper is using 2D Plaxis 8.5 from Plaxis BV , the Netherland and the in situ measurement is using Dutch CPT. This paper will show that soil structure interaction is one of key point. Another key point is Pore Pressure behavior by loading and it’s release using ARS Pile.



A unified method to describe the influences of intermediate principal stress and stress history in constitutive modeling


M. Kikumoto, T. Nakai, H. Kyokawa, F. Zhang and H. M. Shahin


It is a commonly accepted view that soil behaves anisotropically due to the past stress histories. This characteristic of soil is usually called stress-induced anisotropy, which is often modeled by applying kinematic / rotational hardening rule. On the other hand, though ordinary isotropic hardening models are rather simple, they cannot describe the stress-induced anisotropy. Satake, one of the authors and some other researchers already indicated a possibility that even an isotropic hardening model is capable of describing the stress-induced anisotropy by applying fabric tensor which suitably considers the influence of stress histories and reflecting it on the stress variables. It is, however, difficult to quantitatively estimate the variation of the fabric. Therefore, a constitutive model formulated with the fabric tensor, which can be applied to the analysis of boundary / initial value problems, have not been developed. Our research group proposed the modified stress tensor tij, by which the influences of intermediate principal stress on the deformation and the strength of soils are properly considered. It has been noticed by experimental and numerical evidences that tij suitably reflects the fabric tensor under monotonic loading paths. Therefore, tij is extended to a new stress tensor tij* to be able to consider simultaneously the influences of the intermediate principal stress and the past stress histories, i.e., stress-induced anisotropy. The evolution rule of tij* is provided referring to the past experiments and numerical analyses. In this paper, the modified stress tij* is applied to an isotropic hardening elastoplastic model as an example. The validity of the model is checked by comparing the results calculated by the model with results of true triaxial tests and hollow cylinder tests on medium dense sand, in which cyclic loading, independent variations of three principal stresses and rotation of principal stress axes are considered.



Modeling on influences of the density and the bonding in geomaterials


T. Nakai, F. Zhang, H. Kyokawa, M. Kikumoto and H.M. Shahin


An elastoplastic model for soils, which is named subloading tij model (Nakai and Hinokio, 2004), has been developed and has applied to the analyses of some boundary value problems. This model can describe typical features of soil behaviors such as the influence of the intermediate principal stress, the influence of the stress path on the direction of plastic flow, the influence of density and/or confining pressure and others, regardless of small numbers of the material constants. However, since this model mainly deals with the behavior of the remolded and reconsolidated soils, it hardly describes the structured soils such as natural deposited clays. A simple method to describe stress-strain behavior of structured soils under normally and over consolidated states in general stress systems is presented in this research. Based on an idea that structured soils can be made due to a kind of bonding effects between soil particles, a simple elastoplastic model is formulated by extending subloading tij model. Firstly, a new interpretation of subloading surface concept for describing the influence of density and/or confining pressure is illustrated. Secondly, a method to take into consideration the effect of bonding on the soil behavior in constitutive modeling is developed. By introducing this method for considering bonding to the subloading tij model, the stress-strain behavior of structured soils can be described. The validity of the proposed model is checked through numerical simulation of oedometer tests and drained and undrained shear tests on structured clays.



The dynamic analysis of soil reinforced by vertical stiff inclusions


A. HATEM(1), I.SHAHROUR(2), S. LAMBERT (3)


Soft soil improvement by vertical inclusions is an interesting alternative for construction on compressible soils from a technical, ecological and economical point of view. Technique concept is to transfer load applied on surface to a substratum through a system composed of stiff piles and gravel earth platform intercalated between improved soil and structure. Reinforcement could be completed by a geotextile, introduced at the platform base. Different elements composing this innovative system of reinforcement are in complex interaction. In seismically active environments, the application of such solution is not yet evaluated. Though, it is believed that the concept (inclusion plus granular soil) has the ability to enhance the seismic capacity of the foundation. According to A.Pecker, plastic deformation and energy dissipation in the horizontal direction would occur in the gravel platform that it will act as a (plastic hinge), while stiff inclusions would increase the strength capacity of the foundation soil as well as dissipating an important amount of seismic energy. In this work, the purpose is to study the seismic behaviour of piled supported earth platform. The analysis will take into account the complex interaction soil-pile-platform-foundation-structure, the heterogeneity of such system and the non-linear soil comportment. Calculations would be performed using three dimensional difference finite modeling. The study will compose of two parts. The first part concerns analysis related to the influence of the gravel platform on the load transfer mechanisms and efforts induced in the different elements of the system under seismic load. The second part is devoted to study the influence of the main factors in dynamic analysis such as the relative rigidity, the loading frequency, number and escapement among the inclusions, etc.



SOIL FRACTURING OF SOFT SILTY CLAYS FOR THE REINFORCEMENT OF A BELL TOWER FOUNDATION


Guido Gottardi, Luca Cavallari, Michela Marchi


Soil-fracturing is a rather recent technique for soft soils reinforcement. For many years, grouting for stabilization and improvement of not permeable fine-grained soils has been a recurrent challenge for geotechnical engineers. More recently, the development of new methodologies as well as an increasing number of cases in which settlements must be compensated have driven current practice to inject soft soils too. Very different injection pressures can be reached depending upon relevant over consolidation ratio, injection rate and grout viscosity. In general, two opposite soil responses can be observed: cavity expansion or soil-fracturing. Many grout injection tests have already been performed at a laboratory scale; however, a thorough understanding of such soil behaviour has not been achieved yet. In the paper, data from a well investigated field study are reported: a case history, in which extensive soil fracturing was carried out in order to improve the mechanical characteristics of the very soft silty clay beneath the second tallest bell tower in Venice and thus to prevent further settlements, without directly underpinning the existing foundation. To this purpose, a lay-out of about 90 injection pipes was introduced into the subsoil all around the bell tower foundation, until 12 m depth, and by using a double packer system each valve (50 cm spaced) was injected 3 or 4 times with a suitably calibrated cement-bentonite mixture. A rather detailed experimental programme of in situ and laboratory soil testing was carried out before, during and after the ground improvement intervention. Data analysis, together with carefully recorded injection parameters, enables to better understand the mechanisms of soil fracturing and the soil response at a real scale.



3D numerical modelling of the stone columns in soft clay


NGUYEN Ngoc-Thanh, Pierre FORAY, Etienne FLAVIGNY


The reinforcement of soil by stone columns is one of the methods being increasingly used in civil engineering in the design of structures in soft soil, particularly with regard to embankments, foundations reservoirs and works submitted to intermediate loads. The design methods and justifications of this type of foundation have been proposed by several authors Greenwood (1970) Priebe (1995), Dhouib et al., (2004, 2005), Debats (2006) among others. Numerical modelling is also increasingly used to evaluate the three dimensional behaviour of stone columns. We present in this paper a numerical modelling of the behaviour of reinforcements by stone columns in very soft clay. The simulation is using the software finite elements Plaxis 3D Foundation. Emphasis is given in taking into account the installation of the columns in the simulation. In practice, under the effect of the compaction of the column, the surrounding soil undergoes a lateral expansion inducing an increase in the horizontal stresses and the Young’s modulis. This lateral expansion of the stone column can be simulated numerically by applying to the boundaries of the columns values of radial expansion corresponding to those observed in practice. This expansion is also related to the ratio of column and soil modulis. Three configurations of the columns were considered in the calculation: (i) an elementary cell containing one column, (ii) a group of 4x4 columns and (iii) a group of 8x8 columns. In each case, the influence of parameters of the soil and the columns were analyzed. The calculations were carried out following a first phase of installation and a second phase of loading. The results show that the use of lateral expansion in the simulation increases significantly the factor of reduction of settlements, leading to values close to those observed in practice. From the numerical results taking into account the lateral expansion of the column, a new chart for the design of the columns is proposed. It is depending on the spacing between stone columns, the ratio of columns and soil modulis, and the value of lateral expansion of the column. The numerical results underline the importance of taking into account the installation in the simulation: the reduction of settlements in calculated subsidence due to the columns is only 10 to 30% without expansion, while the introduction of an expansion stage gives a reduction of the settlements of 30 to 70%, very close to that observed in the field practice.



Characterization and modeling of consolidation and seepage behavior of soft sediment at low stress levels


Jeffrey S. Melton and Rafael Prieto


Geotextile based reactive capping is a promising technology for the in situ treatment of contaminated sediment because it has the potential to reduce costs and environmental impact when compared to dredging. The cap consists of a reactive amendment (e.g. activated carbon, phosphate minerals, zeolites) held between two geotextiles. Typically, the mat placed on the contaminated sediment and 25 cm of clean sand or sediment is placed over the mat for protection. When the contaminants move out of the sediment and into the cap, they are captured by the reactive amendments. While this technology has promise, it must be optimized to make it cost effective, which requires a better understanding of how the underlying soft sediment behaves when a cap is placed. It is clear that the underlying sediment undergoes consolidation, but the sediment consolidation process cannot be accurately measured using an oedometer consolidation test because the weight of the porous stones and the loading plate is comparable to the weight of the protective layer of sand. A modified seepage consolidation test was used to characterize the consolidation process at low stress levels (0 to 5 kPa), which is consistent with the weight of the reactive capping mat. This paper discusses the results of the seepage consolidation tests, and how they were used to evaluate various constitutive models for soft sediments. In addition, model results are compared with laboratory data to show which models work well for finite element analysis of soft sediment deformation at low stresses. Finally, the limitations and advantages of these models for soft sediments are presented.



Application of Geosynthetics Reinforcement in Freezing-Thawing Roads: A case study from Norway


Even Øiseth, Arnstein Watn, Vikas Thakur


Low bearing capacity during the thawing period gives a lot of problem related to rutting and edge deformation on roads in Nordic countries. Geosynthetic reinforcement has been used in some road rehabilitation projects to improve the bearing capacity and to reduce the need for load restrictions on the traffic during the thawing period. This paper will present a case from the island Hitra in Sør-Trondelag, Norway, where different types of geosynthetic reinforcement has been used in road rehabilitation project. The sub soil in this area consists to a large extent of soft peat. During the thawing period the bearing capacity of the road was very low and large deformation required restrictions on the traffic load. SINTEF rock and soil mechanics group have evaluated the results from the reinforced section on this road. This paper will present different types of field measurements and long-term observations of the deformation to evaluate the effect of the geosynthetic reinforcement. The results indicate that the rehabilitation project has been successful as the rutting and deformation are within acceptable range on all sections. The plate load tests indicate that the resistance against plastic deformation significantly high, and so the amount of rutting will be low as well.



Model Studies of Circular Foundations on Soft Soils


Ahmet Demir, Murat Ornek, Mustafa Laman, Abdulazim Yildiz


Soft clay deposits cover most of the residential areas of the world. Shallow foundations, when built on soft clay deposit, have low load-bearing capacity and undergo large settlements. Construction in loose or soft subsurface soil conditions may not sometimes be possible without using a suitable soil improvement technique. The soft clay behavior can be improved by totally or partially replacing inadequate soft soils with granular fill being compacted in layers. It is also possible to further improve the bearing capacity by reinforcing the base course with geogrid. In the present study, laboratory model tests were carried out to determine the improvement of bearing capacity and settlement behavior of circular shallow foundations supported by a compacted granular fill with and without geogrid reinforcement on soft clay soil. For this purpose, laboratory model tests were conducted in a circular steel box. In these tests, soft clays prepared in constant water content were placed into the test box in layers and compacted in constant compaction energies. And then, granular fill materials were placed on this soft clay bed using geotextile for separating purpose. Circular rigid foundations with different diameters were used for transmitting loads to soils. In addition to this, geogrids with one or more layers were used on soft clay for reinforcing the granular fill. Loads were applied automatically with a constant speed using load cell and deformations were recorded using data acquisition system. After laboratory tests, numerical analyses were conducted using geotechnical computer software PLAXIS (Finite Element Code for Soil and Rock Analysis) and the results were compared with test results. The results of the experimental and numerical studies indicate that the depth and width of the granular fill and the geogrid reinforcement have considerable effects on the bearing capacity and settlement behavior of the circular shallow foundations. It is proved that the improvement depends also on the reinforcement geogrid arrangement. The use of geogrid reinforcement in a compacted granular fill on soft clay soil helps much more improvement on their bearing capacity performance.



Prediction and performance of stone columns to reduce settlements under a breakwater dam in Trieste bay (Italy)


Cola S., Gabrieli F. & Simonini P.


To protect the port of Trieste (Italy) from southwestern storms, the construction of a new 400 m-long breakwater wall was very recently undertaken. The dam consists of several precast concrete caissons loading a large rubble embankment posed on to the 20 m deep sea bottom. The ground below the sea bottom consists of a 20 m thick layer of normally consolidated very soft silty clay, resting of a flysch formation. Assessment of bearing capacity at all stages of construction together with settlement calculation using finite element computation indicated as necessary the use of an improvement system to increase soft soil strength and stiffness. Among the different types of possible solutions, a stone column system was selected. Stone columns were installed by driving a casing up to 14 m under the sea bottom, and then filling the hole with sub-angular cobblestones during the casing extraction. In order to reduce costs of improvement works, the stone column system was designed to reinforce the soils especially in the zone under the core of the dam. In addition, to prevent possible lateral spreading of soil under the bank slopes, a higher number of columns were driven into the upper 3 m layer of the extremely soft soil, above which a stiff geogrid-reinforced granular layer was installed, before starting with the bank construction. By this system, the maximum settlement at the centre of the dam, estimated with a staged finite element (FE) analysis, was reduced at around 1.8 m. The occurrence of such relatively large settlement imposed a careful site control of settlement evolution during the construction. Moreover, before the caisson installation, the bank elevation has to be risen by adding a rubble layer on its top, in order to achieve the correct wall elevation at the end of soil consolidation. Settlement measurements during each stage of bank construction, still being carried out by a simple and low-cost but reliable monitoring system, are used to back-analyze site soil response and recalibrate FE computations to predict settlements of the dam at each stage before completion, leading to some eventual variation of construction schedule. In this paper, the results of site monitoring are presented and compared with the results of FE analysis performed before and during the embankment construction, showing the relevance of the observational method to update the predictions of ground movements and to carefully control the whole construction process.



Creep behaviour of the Venice lagoon heterogeneous clayey silty soils


Berengo, Leoni, Simonini, Vermeer


To protect the city of Venice against recurrent flooding, a huge project has been undertaken involving the design and construction of movable gates located at the three lagoon inlets for controlling tidal flow. The design of such an important geotechnical structure requires the knowledge of the mechanical behaviour of the heterogeneous subsoil of Venice lagoon. Several in situ and laboratory tests have been carried out in the past, thus allowing a proper understanding of the long term mechanical behaviour of the subsoils. In order to assess the predictive capabilities of a well established constitutive model for creep in soft soils, that is the Soft Soil Creep model, a numerical study has been carried out. The creep behaviour was first simulated by reproducing one-dimensional tests to extend the study to a real boundary value problem. A well-instrumented cylindrical trial embankment was built in 2002-03 and displacements have been measured for several years, thus allowing anexcellent database for comparisons with the numerical results. In order to model the anisotropic behaviour of natural Venice soils, a further study has been carried out by using the new Anisotropic Creep Model developed at Stuttgart University and the obtained results are discussed thus showing the influence of the assumption of anisotropy on the predicted displacements.



Deep stabilized test embankment at the Suurpelto area in Espoo, Southern Finland


Timo Stapelfeldt, Matti Lojander, Helsinki University of Technology, Harri Tanska, City of Espoo, Antti Ojala, Geological Survey of Finland, Juha Forsman, Ramboll Finland OY


The city of Espoo has started construction of infrastructures of a new area about 10 km west of Helsinki City Centre. The area consists of clay and silt deposits up to 25m depth. In the upper part of the deposit organic rich clay and silt can be found. Also a thin layer of sulfide rich clay has been observed. The city of Espoo constructed a test embankment in the year 2006. One part of embankment is founded on cement and lime cement columns (d = 700 mm, h ≈ 18 m) which reached the bottom of the clay layer. Another part was constructed on columns being about 5 m shorter. The purpose of the test embankments is to evaluate the best possible binding material and the amount of binding material for soils in Suurpelto area, and also to investigate the stabilization of the stiff clay layer under soft soil layers. The quality of the columns was tested with column penetrometer tests and vane tests on the columns. Settlements of the embankment as well as pore pressures under the embankment have been measured. The settlement of the embankment on long columns almost stopped after some weeks. The embankment on short columns continues settling (>20 cm) after 10 months. This article deals with litostratigraphy, physical and geotechnical properties on Suurpelto clay and silt. Classification tests, triaxial tests, oedometer tests and permeability tests have been carried out.



Numerical modelling of a test embankment on soft clay improved with vertical drains


Timo Stapelfeldt, Pauli Vepsäläinen, Helsinki University of Technology


To improve foundations on soft clays many different ground improvement techniques exist on the market, among these the vertical drain technique. This simple and cost efficient ground improvement method is used to accelerate radial drainage and consolidation by reducing the drainage path in soft soils. Thus, a gain in strength is achieved allowing larger loads to be placed on the soil. The test embankment under consideration is constructed both on improved and unimproved soft clay. Field measurements have been performed over several years and settlements can still be measured. Also, it can be observed that the unimproved part of the subsoil is also influenced by the improved part. Due to their sedimentation processes the here considered normally consolidated soft clay exhibits a high degree of plastic anisotropy, which is typical for glacial deposits in Northern Europe. Thus, constitutive modelling needs to consider initial anisotropy as well as the change of anisotropy during subsequent loading. The constitutive models used are critical state type models called S-CLAY1 and S-CLAY1S. The former model accounts for plastic anisotropy and the latter additionally for inter-particle bonding and degradation of the bonds. In the proposed paper the performance of vertical drains will be demonstrated by numerical analyses of the test embankment. Numerical simulations will be compared with new measurements recently performed.



Modelling of triaxial behaviour of stabilized clay and mud


Asko Aalto


Until these days the design of stabilized structures has been made with analytical methods in Finland. Basic unconfined compression tests on stabilized laboratory samples had been enough for giving sufficient information to designer. Nowadays effective FEM-applications were become more and more common. The new material model of stabilized material is needed. This paper describes the first phase of developing the constitutive material model for stabilized soft soil. Drained and undrained triaxial compression tests were performed on lime-cement stabilized clay and mud. For simulating the stress-strain behaviour different normalization methods were used. The effect of stress applied during curing has also been studied. The final reached preconsolidation pressure is combination of the curing time (cementation) and the stress applied during curing. This combination before have a great effect on strength and deformation properties of stabilized material. The mobilization of strength parameters of elastoplastic (deviatoric) hardening and softening stabilized material were studied. At early stage of the deformation the frictional component is relatively small. However, the strength due to cohesion of the stabilized material decreases with the increase in deformation and finally the frictional strength mainly controls the strength of the material. The strength and deformation behaviour (deviatoric hardening / softening, critical state) of stabilized material has compared with unstabilized normal consolidated and highly overconsolidated soil.



Improvement of a soft clay by preloading and vertical drains: field measurements and numerical predictions.


Bandini V., Biondi G., Cascone E., Galletta A.


The construction of two cylindrical oil tanks on a deposit of soft clays required to accelerate the consolidation process prior the construction of the tanks in order to limit soil settlements at service conditions. The traditional method of preloading associated with vertical wick drains was used for the purpose and monitoring of soil settlements was carried out during the preloading period and during the hydraulic leakage loading test of the tanks. A detailed geotechnical characterization of the foundation soil was available, based on in situ and laboratory tests. A back analysis of the soil behaviour at the location of one of the tanks was performed and the results of the analysis are compared with the actual performance of the tank foundation soil resulting from field measurements. Numerical analyses have been performed using a finite element code. The mechanical behaviour of the soil has been described using an elastic-plastic rate independent constitutive model with isotropic hardening, capable to reproduce the soil non-linearity. The model has been calibrated using the soil stress-strain curves obtained in consolidated undrained triaxial compression tests. The results of a coupled consolidation analysis provide a satisfactory agreement of the observed time-settlement curve at the centre of the tank and of the overall field of displacements monitored during the preloading period and the hydraulic leakage loading test.



Finite Element Simulation of Mandrel Penetration in a Normally Consolidated Soil


Ali Ghandeharioon, Buddhima Indraratna, Cholachat Rujikiatkamjorn


Good quality geologic materials for construction are also becoming scarce. Due to these reasons and because of the environmental restrictions on certain public works, ground improvement is becoming an essential part of infrastructure development. As a result, Civil Engineers are forced to utilise even the poorest soft clay foundations for buildings, highways and railway tracks. The application of preloading with prefabricated vertical drains is regarded as one of the classical and popular methods in soil improvement practice. However, the installation of vertical drains using steel mandrel causes significant remoulding of the subsoil with pore water pressures trend to build up during mandrel driving, especially in the immediate vicinity of the mandrel. The resulting smear zone will result in reducing lateral permeability, which adversely affects soil consolidation. In this paper, finite element simulation of mandrel penetration in normally consolidated soils is presented using a commercial finite element code (ABAQUS). A coupled analysis with large-strain frictional contact is used to simulate soil-mandrel interaction behaviour during the mandrel penetration. A modified Cam clay model is used to represent a soft soil. The numerical predictions are then compared with the results obtained from the large-scale laboratory to predict the extent of the smear zone. A large-scale consolidation test was conducted with the speed controlled mandrel installation machine. Associated excess pore pressures at various locations were recorded and compared with the numerical predication.



The Influence of Smear Zone on the Behaviour of Haarajoki Test Embankment


Abdulazim Yildiz and Minna Karstunen


Natural soft soils are highly anisotropic and micro-structured (bonded) materials. It is well known that the fabric and the initial microstructure of clay play an important role in its compressibility and peak shear strength. In recent years there has been considerable development in understanding the behaviour of soft clays and a number of elasto-plastic constitutive models incorporating such as anisotropy and/or destructuration have been published in the literature. Two recently proposed constitutive models, namely S-CLAY1 that accounts for initial and plastic strain induced anisotropy and its extension, called the S-CLAY1S that accounts, additionally, for interparticle bonding and degradation of bonds Soft clay deposits have low strength and high compressibility. Prefabricated vertical drains (PVDs), in combination with preloading have been widely used to accelerate the settlement and increase the strength of soft clays. In the analysis of PVDs, many factors should be considered. One of the most significant aspects is smear effect. In the field, PVDs are installed using a steel mandrel. The installation process usually causes significant disturbances of the soil around the mandrel. The term smear is generally referred to the disturbance that occurs during the installation of the vertical drain. Consequently, the compressibility and consolidation characteristics of the smear zone are significantly changed because the structure of the soil is destroyed by mechanical disturbance. Hence, the behaviour of soil stabilised with vertical drains can not be predicted accurately if the effect of smear is ignored. Previous numerical studies in literature have only taken into account a reduced permeability due to remolding in the smear zone. However, disturbance in the smear zone causes a decrease in the preconsolidation pressure and, more generally, an increase in the estimated settlements. These parameters must also be into account in the analysis. This paper studies the influence of smear zone on the behaviour of Haarajoki test embankment, considering the section that was improved with vertical drains. Finite-element calculations are performed with Plaxis 3D Foundation. Haarajoki deposits can be characterised as very sensitive anisotropic soft clay. The water content is often higher than the liquid limit. Hence, considerable disturbance is expected in the subsoil during the installation of vertical drains. The soft clay behaviour is represented with S-CLAY1 and S-CLAY1S models. It has been shown that not only reduced permeability but also compressibility and consolidation characteristics of disturbed zone have significant influence on the PVDs behaviour.



A software programme for designing columnar reinforced soils


M. Bouassida, L. Hazzar & P. De Buhan


The design of foundations built on columnar reinforced soils basically requires a first verification related to bearing capacity aspect and a second verification which aims the settlement aspect. For this purpose a comprehensive methodology has been proposed recently by focusing on the design enabling the determination of optimized substitution factor or the so called improvement area ratio (Bouassida, 2007). This factor is defined as the ratio of total columns cross section dived by total area of foundation. In this framework the elaboration of a software programme started since 2005 at National Engineering School of Tunis (ENIT) in collaboration with National School of Bridges and Pavements (ENPC, Paris). The first version of this software named “Columns” is yet achieved. All columns, 3D modelled, are located under the foundation and are assumed of end-bearing type. All columnar reinforcement techniques (stone columns, sand compacted columns, deep mixing, etc.) can be handled by the use of “Columns” software. Ultimate bearing capacity determination is carried out by lower bound direct approach of limit state design [1]. While settlement prediction is based on variational method in linear elasticity by which columns are also 3D modelled, [2]. Adding to this, based on “poro-elastic” approach, “Columns” software includes the prediction of the evolution of consolidation’s settlement as a result of a loading’s history process (i.e. staged construction). For practical and illustration purposes, bearing capacity and settlement estimations are included in “Columns” software. The computation results given by “Columns” software have been assessed after analysis of about fifteen (15) case histories involving full scale loading tests as well as model tests in laboratory. The elaboration of “Columns” software has been supported by a financial funding provided by the Tunisian’s scientific research and technology secretary of state through a project aimed valorisation of research results.



Design and construction of a sludge lock trench to prevent extrusion beneath an embankment


Christopher Currie & Morag Mackay


Jacobs were appointed by a Design and Build Contractor to undertake the detailed design for the proposed N6 Kinnegad to Kilbeggan project in Eire. The scheme comprised 28km of dual carriageway, 3 grade separated junctions and 18 major structures. Design of 8m high approach embankments was required for one of the side roads, the L50062. Detailed design of the L50062 was undertaken based on limited ground investigation information available. The ground conditions were anticipated to comprise 0.5m to 1.5m of peat underlain by 3 to 5m of weathered glacial till in turn underlain by stiff cohesive till. All peat was to be removed prior to embankment construction and the design incorporated removal of any local soft spots encountered during construction. During construction of the N6 mainline in the vicinity of the L50062, the ground conditions encountered were much poorer than anticipated. Further investigation was undertaken which confirmed the presence of between 1.1m->3.2m of very soft gravelly sandy clay below the northern approach embankment. As a consequence of this very soft clay, the original embankment design had to be revised. The design required to incorporate basal reinforcement to provide an acceptable factor of safety against failure of the embankment side slopes. Foundation extrusion was identified as an issue due to the depth of the soft clay, and the potential for this was assessed in accordance with BS8006. No additional land was available adjacent to the embankment, therefore the recommendation given in BS8006 to increase the side slope length of the embankment could not be considered. In order to prevent lateral extrusion from occurring a design solution was required that would provide adequate lateral confinement. This was provided in the form of a “sludge lock” trench. The ‘sludge lock trench’ was constructed beneath both sides of the approach embankments and also under the embankment toe parallel with the N6 mainline. The trench was excavated to 2.4m below the base of the peat replacement fill material and was backfilled with Class 6A fill material. The bottom layer of geogrid, required for stability of the approach embankment, was continued down into the trench along with a geotextile, to aid stability and allow separation of the fill materials and the very soft cohesive materials to be maintained. This paper will describe the design process, issues raised by the limited and contradicting ground information, option constraints and constructability considerations.



Influence of gouting on the long-term behaviour of tunnel in soft clay


Dong-Mei Zhang, Zhen-Yu Yin, Hong-Wei Huang, and Minna Karstunen


The purpose of this paper is to investigate the influence of the grouting during the construction of tunnel on the long-term behaviour of tunnel in soft clay. A simple but robust elasto-viscoplastic model EVP-MCC accounting for soil viscosity was employed, coupled with consolidation theory. The field-based measurement from Shanghai metro tunnel line-2 shows that the effect of grouting on the excess pore pressure is quite remarkable. Its effect on the long-term settlements over tunnels can be rather significant, particularly when tunnels are embedded in soft and compressible soils. The metro tunnel constructed by shield tunneling method in a soft compressible clayey deposit at Shanghai in China is presented, with the soils parameters of Shanghai clay for our EVP model including hydraulic property. A multi-step numerical simulation was carried out to simulate the construction of tunnel and its subsequent consolidation stage. The influence of grouting on the evolution of long-term surface settlement, as well as ground loss, is simulated by imposing prescribed volumetric strain on the disturbed area around the tunnel due to shield tunneling. The volumetric strain could be detected from the grouting ratio of grouting volume and the volume of the physical gap of shield tunnel. The axial forces and bending moments of the tunnel lining were also examined. Finally, some conclusions are drawn from the numerical simulations.



Encapsulation of contaminated soil in Lempola Finland


A. Leivo, Md. Hassan, O. Ravaska, M. Lojander, A. Palolahti, T. Toivanen


The article deals with encapsulation of lead shots from an old shooting range in connection with the construction of the motorway E18. Contaminated soil (peat) was encapsulated beside the slip road of the motorway. The related study is a part of a project funded by the Academy of Finland aiming to figure out the possibilities to utilize Finnish surplus clay by means of adding small amounts of cement in it. The project was started by testing several clay samples taken from the construction area. After gradation tests and permeability tests in oedometers equipped with facilities to measure also permeability, suitable clays were selected for encapsulation material. Mass stabilization was chosen to increase the shear strength of the clay. Several samples were stabilized in the laboratory with different amounts of cement and with different setting times. After that unconfined compression tests and permeability tests were carried out. Compaction of the unstabilized lower clay layer and mass stabilization of the upper clay layer were carried out in August, 2007. Stabilization was carried out on the site using the mass stabilization equipment and on next day the clay was spread and compacted. Stabilization was carried out on the site using the mass stabilization equipment and on next day the clay was spread and compacted.



Modelling the staged construction of embankment on soft clay


Zhen-Yu YIN, Minna Karstunen


The purpose of this paper is to model the staged construction of embankment on soft clay by using a new developed constitutive model accounting for the anisotropy, destructuration and viscosity of soft clay, coupled with consolidation theory. The model is based on the framework of Perzyna¡¯s overstress theory and the elasto-plastic model SCLAY-1S. The Murro test embankment constructed on a soft compressible clayey deposit in Finland is briefly presented, with the soils parameters of Murro clay for our EVP model including hydraulic property. Finite element analysis of the embankment was then carried out to compare the field measurement within 14 years. After this validation of the constitutive model, the determined parameters were then used to study the staged construction of the embankment, especially for the temporary surchange. The influence of the height and the slope of this temporary embankment (surchange) on the long-term behaviour of embankment were studied as well as the duration of the surchange. The comparisons between the measured and predicted results (settlements, horizontal displacements and excess pore pressure) demonstrate that the proposed model coupled with consolidation theory can well describe the long-term behaviour of embankment on soft clay. The results of the numerical simulations provide a database as a guideline for the construction of embankment.



Undrained uplift capacity of plate anchors buried in soft saturated clay


Singh, S. P. & Ramaswamy, S. V.


This paper examines the undrained uplift capacity of plate anchors and the factors affecting the same, through a series of laboratory model tests. The effect of shape of anchors, embedment depth, pullout rate and consistency of soil on soil breakout capacity as well as on suction breakout capacity has been investigated. All the tests were conducted using high plastic commercial clay having liquid limit of 75 % and plasticity index of 44 %. The embedment ratio of anchors was varied as 1, 2, 3, 4, 6 and 8 whereas the consistency of soil was varied as 0.1, 0.25, 0.40, 0.50 and 0.60 and the rate of pull was varied as 1.4, 3.5, 5.0, 14.0 and 21.0 mm/min. Based on the results of the present investigation it is concluded that suction force forms a major part of the undrained breakout capacity of anchors located at shallow depth. The contribution of suction force (Qs) to breakout capacity (Q0) decreases linearly with the increase in embedment depth upto critical depth after which it remains almost constant. The rate of increase of anchor capacity with pull out velocity is found to be same as the rate of increase of undrained strength of soil with applied strain rate. Irrespective of the pull out rate, the ratio of the suction force to the breakout capacity of the anchors is found to be same at a given embedment depth with minor variations. The magnitude of suction force increases with the consistency of the soil, but the non-dimensional suction breakout factor (Fsc) is found to be independent of soil consistency. In brief the breakout capacity of plate anchor is a function of the shape of anchor, the embedment ratio of the anchors, the rate of loading as well as consistency of the soil. Based on the present experimental results and other published works, empirical relationships have been suggested to take into account the rate of pull and shape of anchor on their breakout capacity.



Ground improvement with cement-rubberchips stabilisation


Chee-Ming Chan


For development on naturally soft and weak soil deposits, pre-construction treatment of the ground is necessary to ensure safety and stability of the structures erected. One of the common methods used is the stabilisation technique, which essentially introduces chemical agents, such as cement and/or lime, to the soil and forming stabilised columns or platforms through mixing. This study focused on using a novel mixture of cement and rubberchips as stabilising agent for a soft Malaysian clay. The rubberchips added were waste from a rubber sheet processing plant therefore en-hancing the ecomonical and green values of the new stabilising agent. In the study, cylindrical stabilised clay specimens were prepared with various rubberchips content and 5 or 10 % cement, then aged for 14 days before being tested in an unconfined compressive strangth apparatus. The strength was found to im-prove significantly with small quantities of the cement-rubberchips stabiliser. Also, comparison of the stiffness moduli with cement-stabilised clay specimens indicated that the mixed stabiliser produced a soil matrix which bahaved in a more elastic manner. In short, the cement-rubberchips proved to be an effective soft soil stabilising agent, while having the added-value of being environmentally-friendly and sustainable. Therfore, it is a promising technique for engineering practice.



Laboratory investigation of permeability characteristics of cement stabilised soft Finnish clay with dry mix method


M.M. Hassan & O. Ravaska


There are large and thick deposits of soft postglacial clay in Finland, which are possible to be utilized instead of dumping it as waste, provided that its state can be changed from liquid into plastic form at low cost for workability and it will meet the requirements of permeability. Generally, horizontal permeability of clay is higher than the vertical one in its natural state and vertical permeability is higher than that in the remolded state. For stabilised soil, different permeability characteristics are reported in various research records. Depending on the mixing method (wet/dry), admixture type and amount, sample preparation in laboratory, test type, soil type, compaction etc. permeability characteristics may be different. This study mainly examines the permeability characteristics of cement stabilised soft Finnish clay related to the dry mixing method. Three different soft clays were chosen for this study one of which was organic clay. Different cement contents were used to evaluate engineering parameters. Plastic limit, pH, one-dimensional oedometer tests with permeability measurement and flexible wall permeability tests are conducted to study the material behaviour and evaluate the permeability properties. General geotechnical index properties of treated clays like water content, unit weight, plastic limit and pH of stabilised clay are presented. Compression index and recompression index of cement stabilised clay are also evaluated and found that the compression index increases and recompression index decreases with increasing cement content. Permeability properties of three clays with different cement amounts are presented.



Dynamic properties of two Canadian soft clays


Muhsin Elie Rahhal


It has been observed, especially after the 1985 Mexico City earthquake, that the dynamic behavior of soft clays throughout the world is a very important geotechnical problem. To assess the level of ground motion, dynamic properties of soil are required, mainly the shear modulus and damping ratio. Only low strain testing will be considered in the present work. At small strains, the stress-strain response is still relatively linear. The small strain shear modulus (Gmax) and damping ratio D play a key role in dynamic response analyses. Due to difficulties in measuring small strains, Gmax are generally evaluated indirectly using either laboratory or field tests. In the laboratory resonant column test, Gmax and D are determined from the resonance frequency of a soil cylinder under vibration. Although many studies have considered the dynamic properties of soils, the aim of this paper is to supplement the soil test data and to contribute to the understanding of the behaviour of two eastern Canadian soft clays. Two eastern Canadian clays were thoroughly analyzed. They come respectively from the Saint Alban area and from LG1 (La Grande 1) dam in Quebec. Both clays are soft, silty, of low plasticity and slightly overconsolidated. The LG1 clay is more sensitive than the Saint Alban clay. The Hardin type resonant column apparatus was used. The specimens were vibrated for 30-60 seconds at shearing amplitudes lower than 0.002% to determine the resonant frequency. Readings of acceleration, torque, frequency, height and volume changes, and resonant column calibration parameters allowed the determination of the shear modulus, damping, void ratio and the shearing strain. All specimens were reconsolidated at vertical stress varying between 45 and 120 kPa with both isotropic (K0=1) and anisotropic (K0=0.65) loading conditions. Under the same confining stress, the vibration tests were repeated at increasing time intervals over a period of 2 to 4 days to study the influence of the confinement ti! me. It is shown that the increase of Gmax over time at each pressure level can be divided into two phases: the first resulting from void ratio changes during primary consolidation, and the second resulting from strengthening of bondage in clays. On the other hand, in high void ratio clays, it appears difficult to predict Gmax using existing models based on void ratio, OCR, and mean effective stress. Ratios such as Gmax/Cu might allow a better use of models to represent the behavior of soft clays.



Three Dimensional Analysis of Pvd-Improved Soft Soils


Abdulazim Yildiz and Minna Karstunen


Natural clays are highly anisotropic because of the mode of their sedimentation and the preferred horizontal orientation during deposition of plate-shaped clay particles. In addition, most natural clays are structured that is compose of “fabric” (a geometry arrangement of soil particles in micro-scale) and inter-particle bonding that reflect the soil composition, history, present state and environment. Neglecting the effects of anisotropy and/or destructuration may lead to highly inaccurate predictions of soft clay response in the analysis. Two recently proposed constitutive models, namely S-CLAY1 that accounts for initial and plastic strain induced anisotropy and its extension, called the S-CLAY1S that accounts, additionally, for interparticle bonding and degradation of bonds. FEM has been recently used for modelling of PVD improved soft soils under embankment loading. In FEM, the behavior of soft soils improved by PVDs has usually been analysed by two-dimensional (2D) plane strain models while actual problem of water flow into a vertical drain is axisymmetric. In addition, the consolidation around vertical drains is truly three-dimensional (3D), hence, in the field, a large number of vertical drains will each have their own independent influence zone. Therefore, a proper design of an embankment involving discrete vertical drains is required to be conducted with a fully three-dimensional analysis. This paper presents 3-D numerical analysis of a benchmark embankment on Poko clay incorporating PVDs. The soft clay behaviour under the embankment is represented with Modified Cam Clay, S-CLAY1 and S-CLAY1S models. A more complex interpretation of 3D behaviour of a vertical drain system was initially investigated and it seen that 3D finite element modelling is often not practical for engineers. Hence, a simple approach proposed by Chai et al. (2001) is used in the analysis. It represents the effect of vertical permeability of natural subsoil and the effect of radial drainage due to PVD, by using an equivalent vertical permeability (kve). With the proposed method, the analysis of PVD-improved subsoil becomes the same as that for the unimproved case. The method is simple and can be directly applied in the 3-D finite element modelling of PVD-improved subsoil. First, verification of the simple method proposed by Chai et al. (1999) is made based on unit cell analysis when used elasto-plastic models. Then, the whole embankment is analysed during and after the construction period by using the simple approach. It is recommended that the proposed simple approximate method is a useful tool for engineering practice.



Analysis of Stiffness Regime Within a Soil Block Stabilised Using DSM Columns


A O\'Sullivan S Terzaghi


The traditional design approach for stone and DSM columns is based on a 2-D or 3-D replacement ratio, where aggregate parameters of the soil and columns are used for the treated soil block. This approach does not take into account any additional increase in soil stiffness between the columns as a result of the confinement which the columns provide. The paper will describe field testing carried out in New Zealand on two similar soil blocks which were stabilised using groups of DSM columns. At each location, a series of Seismic CPTs were carried out at close spacing prior to and following the installation of DSM columns. The results appear to show a significant increase in soil stiffness between the columns and highlight additional benefits which the installation of the columns has on the overall stability of the soil block. The paper will outline the results of the investigation and will describe the correlation between the results obtained on site with those predicted by 2D and 3D Plaxis models. Finally the paper will seek to provide additional assistance to designers and modellers of Stone/DSM columns which will allow them to carry out some simple checks on the benefits of the group effect of the columns



Applications of volume averaging technique in numerical modelling of deep mixing


Urs Vogler


The properties of very soft clays, silts and organic soils can be improved by column-type soil improvement. Common improvement technology in Northern Europe and the UK is the deep mixing method: By mixing binder in the in-situ soil the bearing capacity and deformation behaviour of the soil is improved to allow for the construction of embankments or foundations. The paper is focusing on the use of an enhanced numerical method, the volume averaging technique for analysing the elastoplastic behaviour of improved soft soil deposit. The volume averaging technique is a homogenisation technique, which allows mapping back the full three-dimensional problem of circular columns under an embankment or a strip foundation to a plane-strain homogenised material. The aim is to safe computational effort but still describing the full constitutive behaviour of soil and columns under the embankment by considering individual advanced constitutive models for each material. The volume averaging technique and the individual constitutive models used are described in the paper. Soft clay is modelled with the recently proposed S-CLAY1 model, accounting for anisotropic strength of the soil and change of anisotropy due to plastic straining. The columns are modelled with a constitutive model considering a hyperbolic stress strain relationship, stress dependent stiffness, different elastic stiffness and a shear hardening law. To show the performance of the proposed method for different applications two benchmark calculations have been performed: A strip foundation and an embankment over a column improved soil have been analysed with the volume averaging technique and with a conventional full three-dimensional finite element analysis. The quality of the results is discussed and recommendations for the use of volume averaging in finite element analysis are given.



Predicting Horizontal Deformations under an Embankment Quickly and Accurately using a Neural Network


R. van der Meij


Horizontal deformations can cause damage to objects adjacent to a newly constructed embankment. A tool has been developed to be able to determine in an early state of a project of such a risk is at hand. Artificial Neural Networks are particularly suited to find a relationship between an input and an output variable, even if there is a complex relationship between the two. In this perspective, it’s useful to investigate whether there is a relationship between the maximum vertical and maximum horizontal displacement under an embankment, especially because it is possible to predict the maximum vertical displacement with sufficient accuracy. Only the total vertical displacement is not enough information to predict the horizontal deformations, though. To make a good prediction of the horizontal deformations one needs, for the drained case, at least three parameters. These are the thickness of the soft soil layers, the height of the embankment and an estimation of the total vertical settlements. Given only these three parameters, an artificial neural network can reproduce the finite element result with an error less than 1 percent. To make a good prediction of the undrained case (in combination with consolidation) one needs more parameters. A scientifically proper way to achieve this, is by having two extra input parameters: the initial height of the embankment and the initial excess pore pressure due to this previous load. These two parameters are of vital information to know how near to failure the slope is, and therefore how much plastic deformations will take place. Another, more convenient parameter to achieve this is by using the safetyfactor. It can be concluded that, for a given geometry, the horizontal deformations can be predicted using only three to five parameters. This quick prediction has the same quality as a complete finite element calculation. The input is very simple and can therefore be used in an early stage of an engineering project.



ON THE NUMERICAL SIMULATION OF A TUBE-INSTALLED DISPLACEMENT PILE


Syawal Satibi; Martino Leoni; Pieter A. Vermeer


ABSTRACT: Considering effects of pile installation is a challenging task for numerical simulations of displacement piles. Displacement piles include full and partial displacement piles. For example, driven, tube-installed, screw and continuous flight auger piles. These type of pile create stress fields due to their installation process. This paper presents 2-D axysimmetric finite element analysis of a tube-installed displacement pile. This type of piles (or columns) are installed by jacking or vibrating a closed-bottom tube into the ground. Upon withdrawal of the tube, the cavity is filled with concrete, so there is an increase of the radial stress around the pile but no skin friction due to the installation process. In the 2D axisymmetric finite element analysis, for initialising the effects of installation, stress-controlled cavity expansion (K-pressure) is considered. An advanced soil constitutive models, which consider soil stiffness depending on the stress level and strain level, has been applied. It is shown that realistic stress fields can be obtained by stress controlled cylindrical cavity expansion up to the appropriate value of K. The method is considered as a feasible method for engineering practice to simulate the effects of pile installation using standard small strain finite element analyses. The applicability of this method is not limited to displacement piles but can be also extended to column type foundations in general.



Enhenced viscoplastic modelling of soft soils


Lyesse Laloui and Suzanne Chalindar


Strain rate has important effects on the behaviour of soft soils. To take into account such effects, various viscoplastic constitutive models have been proposed in the literature over the past decades. A widely used formulation is the well-known Perzyna’s model, in which the yield function used for describing the viscoplastic strain is rate-independent and becomes larger than zero when viscoplastic straining occurs. What we call the consistency model on the other hand, uses a rate-dependent yield surface the derivative of which governs the viscoplastic strain. It can hence be seen as an extension of the classical elastoplastic approach to account for rate dependency. This paper presents an innovative approach to model viscoplastic behaviour of soft soils, making use of comprehensive experimental observations of strain rate effects on the evolution of the apparent preconsolidation pressure. In particular, this allows the modelling of the evolution of the vertical yield stress at any void ratio as well as the use of the unique effective stress-strain-strain rate concept. The presented modeling approach is qualitatively shown to be capable of simulating the three main types of displays of rate sensitivity, which are: strain rate effect (on strength and volume variation), creep and relaxation. Finally, a validation exercise on a series of experimental data is presented. The results display a very good performance of the proposed viscoplastic approach, with easily obtained model parameters.



Analyse of the temperature effect on the efficiency of Prefabricated Vertical Drains for in-situ consolidation of soils


Lyesse LALOUI, Simon SALAGER, Nina MATTSSON


Preloading or surcharge loading is a frequently used technique in construction to obtain adequate bearing capacity and acceptable post-construction settlement. To accelerate the settlement process it is common to combine the preloading with installation of vertical drains. The primary purpose of the vertical drains is to shorten the drainage path. The preloading creates a hydraulic gradient which causes the pore water to be squeezed out during the consolidation and the vertical drains enable the pore water to flow horizontally towards the drains and then freely along the drain. For fine-grained soils however, the time required to allow the dissipation of the excess pore-water pressure is often unacceptably long due to the very low permeability of these soils. This paper proposes to evaluate and to improve a new innovative technique allowing to reduce the time required to return to pore-water pressure equilibrium and, consequently, to increase the rate of settlement. This technique consists of using Prefabricated Vertical Drains fitted with a heat source (T-PVD). The heat source leads to an increase of the soil temperature around the drains and therefore also to an increase of the permeability due to the effect of heat on the water viscosity. Firstly, complete characterisations of the thermo-mechanical behaviour of the tested soils has been carry out. These characterisations have given a good understanding of the temperature response of soils and determined parameters needed by the thermo-mechanical model used for simulations. Secondly, the effect of temperature on the settlement process has been experimentally analysed by means of a large oedometer apparatus designed for this study. The soil was compacted in the chamber and a T-PVD installed in the center. This apparatus allows to measure, all along the thermo-mechanical process, the pore pressure and the temperature evolution at different points of the sample, the over-all consolidation and the water volume release. Finally, numerical simulations which take into account thermo-hydro-mechanical couplings are presented. They enable back-analysis of the results of the experimental part and prediction of the consolidation rate in real-sized road embankements installed with T-PVDs.



An experience of consolidation of extremely soft mud for one of Europe’s largest projects “ The AIRBUS A-380” assembly factory in Hamburg – Germany


Jean-Luc Chaumeny, Johannes Kirstein, Serge Varaksin


Contiguous to the existing Airbus plant in Hamburg (Germany) is an old sand quarry along the Elbe river. Alluvium has filled up the pit over the last 50 years and covered the old sand alluvium with a thickness of locally over 12 meters and remaining under tidal influence. The soft soil is characterised by mud on the top and different layers of clay, soft clay, gyt-tja and peat below. The project consists of two steps: first the closing of 160 hectares by a dyke and second the fill with optimized consolidation processes of the soft soil. An alternative to sheet piles was adopted by the dyke founded on geotextile confined sand columns (Möbius Process) to provide bearing capacity on clay and peat with less than 0.5 kN/m² shear strength. 20 millions tons of sand were placed by a sprinkling method and 30 millions meters of vertical drains were placed combined with Vacuum Consolidation (Menard Proc-ess) in the structural areas. The article describes the concept, the parameters and the methods used for the planning. The techniques of vertical drainage and especially Vacuum Consolidation are explained in the context of the field works. The predicted settlements are com-pared with the monitoring results of four year performance.



Experimental Study on Behaviors of Piled Composite Ground with Different Connections between Pile Head and Raft


Gang Zheng, Yu Diao


Rigid piles or micropiles are commonly used in ground improvement to form piled composite ground recently in China. This paper presents a series of experimental studies, including laboratory small scale model tests, field large scale model tests and filed loading tests, in order to investigate the effect of different connections between pile head and raft on the interaction among pile, soil and raft of piled composite ground. The tests mainly involved three types of connections, of which the first one is that pile head is embedded into raft, and the second one is that pile head and raft is separated from raft by a layer of compressible sand cushion, and the third one, which is suggested by the author, is that a gap is introduced between the pile head and raft. In this paper, the behaviors of the piled composite ground with different connections, including bearing capacity, settlement, load sharing between soil and pile and pile axial force and skin friction were presented and discussed. Besides, the mechanism of load transfer in piled composite ground with gap was analyzed. The results of tests indicate that the piled composite ground with gap has an advantage over conventional piled composite ground and pile composite ground in aspects of settlement controlling and bearing capacity. Based on the analysis of results obtained from tests mentioned above several conclusions can be drawn as follows: The piled composite ground with gap has similar properties of settlement to that of natural ground before pile head and raft are in contact. It also resembles the conventional piled composite ground in aspect of settlement after pile head makes contact with raft. The piled composite ground with gap or compressible layer has a superior integrity to conventional piled composite ground in aspect of settlement due to that the soil beneath the raft, especially the soil surrounding the upper part of piles, is compacted in advance before piles and soil carry load together. The preloading and compaction of soil among piles can increase the normal stress acting on pile body and thus can increase the shaft resistance of pile. Further more, the slip at the pile and surrounding soil interface can be reduced. This can result in the reduction of pile tip penetration.



IN-SITU AND LABORATORY INVESTIGATIONS IN SOFT GLACIAL SOILS AS BASE FOR DESIGN AND DECISIONS


Steiner, Walter


Abstract: Glacial soils deposited during the ice age by advancing and retreating glaciers and associated rivers in the Alps have left heterogeneous, locally variable and often deep deposits. With the denser urbanization it becomes necessary to use these areas for infrastructure con-struction and buildings above and below ground. The geotechnical characterization of the ground requires an understanding of the geologic history. Based on several case histories from the past two decades experience with in-situ tests, mainly Flat Dilatometer, DMT, Cone Penetration tests CPT with measurement of pore pressures CPT-U, and laboratory tests and performance monitoring will be shown in perspective. Beneficial effects as well as require-ments for improvement will be discussed. The specific case histories are: · An embankment adjacent to a cut on slightly pre-consolidated glacial clay · A cut-and cover tunnel in ice-marginal glacial clays and sands and postglacial clays · A trench for an underpass in soft clay beneath a viaduct founded on piles using DMT and CPT-U and triaxial tests · A slab foundation for a multiple storey building used instead of piles based on the application of DMT and CPT-U tests. · An avalanche gallery with an embankment that suffered large settlements on loose silt. The characteristics could be more reliably evaluated with DMT tests. For each case histories in-situ tests mostly combined DMT (Flat Dilatometer) and CPT-U (Cone Penetration Test) were applied, together with cored borings and results from mechani-cal (triaxial) tests on samples in the laboratory were used. Effects of testing procedures may have a large effect on properties. Pieces of information obtained from different in-situ and laboratory tests will give a more complete picture of the ground properties (strength, deform-ability, maximum past pressure, consolidation properties).



MODELLING DESTRUCTURATION AND ANISOTROPY OF BOTHKENNAR CLAY


K. McGinty, M. Karstunen, S.J. Wheeler


An elasto-plastic model for soft, normally or lightly overconsolidated clays is presented. This model, named S-CLAY1S (Koskinen et al., 2002), accounts for the effects of both destructuration and changes in anisotropy. The model includes an inclined yield curve with a rotational hardening law to allow rotation of this yield curve, representing the evolution of anisotropy during plastic straining. An additional hardening law is included in order to represent the loss of interparticle bonding during plastic straining (destructuration). Experimental data from several suites of multistage drained triaxial stress path tests on samples of Bothkennar clay are presented alongside model simulations from S-CLAY1S and this has allowed calibration of model parameters. A total of seven model parameters are required. The procedure for calibration of model parameters is presented and it is shown that a small number of stress path tests can be used in order to obtain suitable parameter values. Further comparisons between experimental data, S-CLAY1S model simulations and simulations from a preceding model which does not account for destructuration are presented. These comparisons show clearly show that S-CLAY1S model predictions provide a good match to the experimentally observed behaviour and show a significant improvement over the predictions of the preceding model.