· · · Conference Proceedings hosted by TU Delft Library

The paper details the role modelling (analytical, physical and numerical) has played in the development of design methods for basal reinforced piled embankments; specifically the level of embankment arching and the determination of the magnitude of the tensile loads in the basal reinforcement. The paper also reviews the major national design approaches describing how the different methods operate. Comparison of the different methods is discussed.

The London Geotechnical Centrifuge Centre located at City University, London is one of four currently active centrifuges in the UK. The centrifuge is well used and much of the geotechnical research at City University employs physical modelling but supported by a large element testing facility and numerical modelling capability. The centre was established in 1990 and the facility was extensively upgraded to provide more space for sample preparation and model making in 2004. In 2012 the centre is supporting 4 doctoral research projects in addition to visitors from China, Italy and UK. The focus of research is urban construction processes with an underlying theme of sustainability. Over the last 10 years the group has established a capability for tackling complex modelling problems and is currently investigating the application of smart instrumentation and control at high g.

The current paper presents a novel experimental method which is able to capture the effects of suffusion by substitution of the fines in a sample by salt of a similar grain size. The setup is tailored to optically capture the change in soil structure behind a glass window in a plane strain strongbox using a digital camera. Subsequently, digital image correlation techniques have been used to quantify the structural change. The first model test shows promosing insight in the pseudo suffusion mechanism. The test setup therefore offers a valuable addition to permeameter tests.

Correct scaling of breach analysis of river levees is a challenging task that is not easily accomplished by physical modelling. Several small-scale physical model tests have been conducted at 1-g level, which cannot truly represent the stress-dependency of soils, whereas the scaling issues arising from centrifuge modelling have not been fully explored.Two key features have to be considered when modelling the prototype behaviour. On the one hand, the whole embankment should be included in the model to ensure that flow nets are valid. This is not always easy to achieve due to space limitations within the strongboxes used. On the other hand, full control of water levels, prior and during breaching, is of principal interest.This contribution shows how both of these features can be modelled for levee breaching by taking advantage of the availability of space within a drum centrifuge and its versatile toolplate.

Structural analysis of new devices to be used in centrifuge facilities can be a challenging task. Initial calculations, based on simple basic static principles, are usually used. However, as devices placed in a geotechnical centrifuge are often complex structures, advanced analysis methods to assess their performance under enhanced acceleration fields are desirable. This paper presents the analysis of the performance of a new strongbox designed for a centrifuge facility at ETH Zurich by using a commercial Product Lifecycle Management (PLM) software.

Animal burrows have been reported to cause extensive damage to existing levees and other earthen structures in different parts of the world. In this study a centrifuge model that has been designed to investigate the impact of animal burrows on the performance of an existing earthen structure is described. The proposed methodology simulates animal burrows in an earthen structure by introducing cylindricalshaped openings inside a pre-designed model. A homogenous levee with 1:1 side slopes and a horizontal toe drain was chosen for this preliminary investigation. The steps taken to construct the model and the effect of the adopted burrow simulation technique on the model performance are discussed. Crest settlement was measured before and after the burrow introduction and during the increase in water level to failure. Conclusions regarding the adequacy of the adopted technique are made.

This paper demonstrates the use of a geotechnical drum centrifuge in the calibration of a model to predict the peak punch-through penetration (qpeak) resistance of a spudcan on sand overlying clay. A series of loose sand overlying clay tests was performed and combined with an existing database of tests performed on dense sand overlying clay. The performance of the failure stress dependent model proposed by Lee (2009) and Lee et al. (2009) has then been assessed using this combined dataset which encompasses a wider range of soil properties and problem geometries than was used in the original calibration of the model. The single empirical stress distribution factor (DF) that is employed in the model was then optimised using a back calculation procedure for all tests. The scatter of the optimised DF values was then compared for the original bi-linear calibration proposed by Lee et al. (2009) and a new nonlinear power law calibration. The new non-linear relationship enables the model to better predict qpeak over a wider range of problem geometries and for conditions of both loose and dense sands overlying clay. The work demonstrates the importance of the geotechnical centrifuge in calibrating such models given that at present numerical methods are unable to reliably capture such punch-through behaviour and good quality field data of punch-through failure of spudcan foundations is unavailable

A 1/18-scale model of a 4-story reinforced concrete structure was tested at 18 g in the US Army Centrifuge to support the development of computational modeling of progressive collapse of multi-story buildings. The model structure was 4 bays long and 3 bays wide with column spacing about 33 cm in each directions and story height of 20.3 cm. Three columns around one corner at the 1st story were explosively removed while the model was spinning in the centrifuge at 18 g. The column removal resulted in complete collapse the bays above those columns but collapse did not progress to the rest of the structure.

In the Netherlands, several field measurements were carried out in piled embankments with a geosynthetic basal reinforcement (GR). This paper presents a series of nineteen 3D model experiments on piled embankments. Purpose of the tests was to find an explanation why the calculated GR strains exceed the GR strains measured in the field. This paper focuses on the starting points of the test series, the test set-up and the scaling rules and gives a summary of the results. Van Eekelen et al., (2011b and 2011c) describe the results of the tests extensively. Five starting points were leading to the development of the test set-up. (1) Possibility to evaluate the two calculation steps separately, (2) Possibility to evaluate the influence of consolidation of the subsoil, (3) Inclusion of GR, (4), Modelling the fill realistically, (5) a realistic stress level and scale. For the test conditions (static load, laboratory scale), it was found that consolidation of the subsoil results in an increase of arching. This is not in agreement with the current calculation models. Loading on the GR is concentrated on the strips lying above and between adjacent piles (the “GR strips”) which is in agreement with the current calculation models. The measured load on a GR strip has the distribution of an inverse triangle, although the load may be even more concentrated around the pile caps than this indicates. This is not in agreement with the current calculation models. Implementing this in the CUR/EBGEO calculation model results in 19-26% less GR strain.

Stability assessment of existing dikes build on soft soil generates questions on the available stability calculation techniques and material models. Especially peat behaviour is not easily captured by standard design methods. This has led to disapproval, based on calculations, of dikes that have been in operation for several hundreds of years. To establish the weak points in the stability assessment prescribed by the available handbooks on dike engineering, a series of 5 field tests is started. These field tests give the option to compare the different design methods and different parameter assessment techniques to the test results. The series includes single stage and multi stage loading. This paper focuses on the single stage loading tests. The tests result in a failure mode that differs from the circular type failure planes that are used in engineering practice. Horizontal and vertical fractures dominate the active side of the failure while at the passive side the peat is compressed. The tests indicate that tensile strength of peat, or peat fibres, is important in understanding stability problems in which thick peat deposits are involved. Furthermore, the tests show that over-consolidated behaviour is important to understand the measurements, while in engineering practice peat is usually modelled as a normally consolidated material.

A series of small scale physical modelling tests are performed in a geotechnical drum centrifuge in order to investigate the triggering mechanisms of landslides due to rainfall. They are conducted under controlled conditions of rainfall intensity and duration, ambient relative humidity, wind, and temperature. These tests have been designed to study the possible failure mechanisms proposed for a full scale landslide experiment. Accordingly, different shapes and hydraulic properties of the bedrock, in terms of drainage and exfiltration, are provided for the model. A three dimensional close range photogrammetric technique is used to track the movements and monitor the volumetric changes of the ground during the cycles of wetting and drying. The slope elevation is filmed during and following the rainfall events using a high speed camera and the deformation vectors and strains are elaborated using the PIV method. Details of the design of the climate chamber are discussed in this paper.

Piping is one of the possible failure mechanism for dams and levees with a sandy foundation. Water flowing through the foundation causes the onset of grain transport, due to which shallow pipes are formed at the interface of the sandy layer and an impermeable blanket layer. In the past, the mechanism has been investigated predominantly in densely packed sands, in which the process was observed to start at the downstream side (backward erosion). Recently performed experiments in loose sand (van Beek et al. 2009) showed a different failure mechanism (forward erosion). In this article additional experiments of piping in loose sands are described for investigating the relevance of the forward process for practice. In these experiments the type of process was found to be dependent on the presence of shear resistance between sand box cover and top sand grains, that causes grains to be fixed geometrically. Without this shear resistance the process was found to be forward, whereas with this shear resistance the process was found to be backward oriented. The change in degree of fixity and relative density as a result of loading is investigated with electrical density measurements. The experiments show that the forward process is not relevant for levees in practice, in which the cohesive blanket layer causes the sand grains to be fixed properly.

The use of a geotechnical centrifuge for properly capturing the mechanisms in the sand adjacent to a foundation pile is well established. Larger centrifuge facilities were previously preferred over small centrifuges, because of the difficulties of the required miniature instrumentation. The technology since then has moved on and recent technical developments, such as MEMS sensors, facilitate the use of small models including all the necessary instrumentation. However, correct scaling and modelling of piles installed in sand in the geotechnical centrifuge remains essential. Especially, stress waves in the sample, created during pile driving, are impractical to scale. The latter, however, is not only limited to small centrifuge facilities. Therefore, the modelling should not include stress waves. A feasible approach for that is to limit pile installation effects in coarse material, which ensures drained conditions, to the application of cyclic load reversals.

Current procedures relate potential liquefaction induced settlements to foundation size and liquefiable depth. However, the analysis of field case histories suggests that the influence of foundation bearing pressure may also be a significant factor influencing such phenomena. Data from 24 buildings that suffered settlement and tilting as a consequences of soil liquefaction during the February 27th 2010 Maule earthquake in Chile, are herein analyzed and compared with data from other earthquakes. Although case history data play a crucial role in geotechnical earthquake engineering, in many cases their analysis is limited to speculation therefore experimental verification is often required. Thanks to the significant development in dynamic geotechnical centrifuge modelling in the last 30 years, we are today able to carefully reproduce field motions enhancing the reliability of experimental results. This is the case for the centrifuge tests discussed in this paper, which have been the first tests performed on the University of Dundee’s new Actidyn Q67-2 servo-hydraulic earthquake shaker.

Three on-going dike safety studies (on: macro stability, piping and flow slides) in the Netherlands make use of geotechnical physical models. A short outline of these projects is presented; the physical models chosen are described and discussed. The three studies use different physical models, depending on the research questions at the beginning of the model test series, the heterogeneity that is anticipated in the field, the scaling laws and the knowledge level. The paper describes why a certain model was chosen

Reinforcing compressible soils by rigid inclusions is a method to reduce and homogenize settlements under many types of structures. A granular mattress, located between the structure and the group of piles, transfers part of the loads on the surface to the head of the piles anchored in rigid substrate. An experimental device, a mobile tray, has been especially designed in order to allow a better understanding of this reinforcement technique. This mobile tray simulates the settlement of the soft ground (not present here) located between the piles. With this device, a parametric study of the loadtransfer mechanism in the mattress is conducted in centrifuge at 20g. Loads at the top of the piles and settlements at different places above the granular mattress are measured during the the mobile tray going down. A possible way to improve this reinforcement technique is to insert a geosynthetic layer between the head of the piles and the granular mattress. In centrifuge, due to scaling laws, the choice of the geosynthetic has to be taken very carefully. Tests with and without geosynthetic are performed for different thicknesses of granular mattress. Then the improvement of the load-transfer mechanism by the addition of a geosynthetic is studied in centrifuge.

Unidirectional irregular waves of varying significant wave-height and peak period but with a constant spectral shape (JONSWAP spectrum, J=3.3) were generated over a sandy 3:20 sloped plane beach to investigate scour fronting a dune erosion control system constructed from geotextiles, slope inclined at 45º. Both passive (i.e., three dune erosion control systems with two configurations) and active (i.e., one nearshore submerged structure with four configurations) structures are investigated. A twodimensional physical movable-bed model simulating the prototype dune-beach systems of Estela, located along the NW Portuguese coast, is employed in this study. The paper presents a brief characterization of the prototype conditions and discusses requirements and limitations on the choice of model scale for the waves, the sediments, and the geotextile materials.

Massive failure of submerged slopes form a major threat for many dikes along estuaries in the Netherlands, the subsoil of which consists of alternating layers of loosely packed and more densely packed sand. Static liquefaction in the loosely packed sand plays an important role. Insufficient knowledge is available to predict the flow of sand and the retrogression of the instability after initial liquefaction. Liquefaction flow slides have been studied during an extensive experimental research program in the period 1973 - 1977 on behalf of the design of the storm surge barrier in the Oosterschelde estuary. Results of this program are revisited to learn about the response of a loosely packed sand layer to a local instability. The program included more than a hundred of tests in a large and medium sized flume that were filled with sand. Each sand body had a horizontal surface and a slope as boundaries. A remarkable difference was observed between the tests in which retrogressive liquefaction flow slides did and those where such slides did not occur. This may be quite relevant for the understanding of flow slides in natural slopes with pockets of loosely packed sand. The test set-up, measurements and main results of the large scale tests are described in this paper.