plaxis bulletin spring 2012

Title

Temporary stability of a jacket platform during installation and influence of adjacent pug marks

Piled embankments in PLAXIS 2D, 3DTunnel and PLAXIS 3D 2011

Numerical analysis of geosynthetic reinforced piled embankment scale model tests

Issue 31 / Spring 2012

Plaxis Bulletin

Pag

e 12

Table of contents

Pag

e 5

Pag

e 6

Pag

e 10

Editorial03New developments04

Recent activities18


10


06PLAXIS Expert Services update05

ColophonAny correspondence regarding the Plaxis Bulletin can be sent by e-mail to:

[email protected]

or by regular mail to:

Plaxis Bulletinc/o Annelies VogelezangPO Box 5722600 AN DelftThe Netherlands

The Plaxis Bulletin is a publication of Plaxis bv and is distributed worldwide among Plaxis subscribers

Editorial board:Wout BroereRonald BrinkgreveErwin BeerninkArny Lengkeek

Design: Jori van den Munckhof

For information about PLAXIS software contact your local agent or the Plaxis head office:

Plaxis bvP.O. Box 5722600 AN DelftThe Netherlands

[email protected]

Tel: +31 (0)15 251 7720Fax: +31 (0)15 257 3107

» The Plaxis Bulletin is the combined magazine of Plaxis and the Plaxis users

association (NL). The bulletin focuses on the use of the finite element method in geotechnical engineering practise and includes articles on the practical application of the PLAXIS programs, case studies and backgrounds on the models implemented in PLAXIS.

The bulletin offers a platform where users of PLAXIS can share ideas and experiences with each other. The editors welcome submission of papers for the Plaxis Bulletin that fall in any of these categories.

The manuscript should preferably be submitted in an electronic format, formatted as plain text without formatting. It should include the title of the paper, the name(s) of the authors and contact information (preferably e-mail) for the corresponding author(s). The main body of the article should be divided into appropriate sections and, if necessary, subsections. If any references are used, they should be listed at the end of the article. The author should ensure that the article is written clearly for ease of reading.

In case figures are used in the text, it should be indicated where they should be placed approximately in the text. The figures themselves have to be supplied separately from the text in a vector based format (eps,ai). If photographs or ‘scanned’ figures are used the author should ensure that they have a resolution of at least 300 dpi or a minimum of 3 mega pixels. The use of colour in figures and photographs is encouraged, as the Plaxis Bulletin is printed in full-colour.

Pag

e 16

Numerical analysis of geosyn-thetic reinforced piled em-bankment scale model tests

12

www.plaxis.nl l Spring issue 2012 l Plaxis Bulletin 3

Editorial

» In 2012 two completely new add-on Modules to PLAXIS software are scheduled for

release: 3D Dynamics and 3D PlaxFlow. The former one has already been released at the beginning of the year. In the New Developments column we discuss the new 3D Dynamics module in depth. The latter will be released towards the end of the year, more information on this exciting release will be given later in the year, so keep an eye on our website and stay up to date.

In this issue of the Plaxis Bulletin we have again tried to offer a collection of interesting articles and useful information for you. The first user’s article involves a study on the temporary stability of a jacket platform during installation by use of mudmats. Specific attention is paid to the pug marks formed by previous deployments of jack-up rigs and their influence on the mudmats bearing capacity.

The second user’s article involves the analysis of piled embankments using PLAXIS 2D, 3DTunnel, and PLAXIS 3D. A comparison is done of the results with the three different versions of PLAXIS software, and conclusions are drawn as to the improvements in the new PLAXIS 3D 2011 in comparison to older versions.

The third users article involves a numerical analysis of geosynthetic reinforced piled embankment scale model tests. PLAXIS simulations of the test series were performed to improve the understanding of the arching mechanism in the piled embankment. Conclusions are then drawn to validate the earlier conclusions from the analysis of the scale tests.

In addition to the contributions by PLAXIS users, the bulletin offers another article on a project where Plaxis has provided expert services to a client, where a common research program was conducted for studying in detail the effects of isotropic consolidation on soft soil.

Furthermore the bulletin gives an overview of recent activities, with a special update on our activities in Asia Pacific and North America, and we take a closer look at the recently launched Knowledge Base. Don’t forget to check the agenda on the back for an overview of our upcoming events.

We wish you an interesting reading experience and look forward to receiving your comments on this 31st Plaxis bulletin.

The Editors

Editorial

4 Plaxis Bulletin l Spring issue 2012 l www.plaxis.nl

Furthermore, new dynamic examples and movies will be provided to help users in setting up models for ground response analysis and seismic design.With these new dynamic software and support facilities we are confident to provide you with up-to-date technology to solve your dynamic problems.

References• Beaty M., Byrne P.M. (2011). UBCSAND

CONSTITUTIVE MODEL Version 904aR, documentation report.

New developments

»To highlight the release of 3D Dynamics, free-of-charge short seminars will be

organized in Seattle and San Francisco in the last week of March. Additionally, relevant governmental organizations (e.g. state DOT’s) and companies on the US West Coast will be visited. Furthermore, Plaxis will present the key features of 3D Dynamics during the software session at Geo-Congress in Oakland, California, on March 27. Plaxis contacts in North America will receive more detailed information about these events in the coming weeks. The new module will have similar capabilities as the 2D Dynamic module. Independent prescribed displacements, velocities or accelerations in the three Cartesian directions can be applied at the model bottom to realistically simulate earthquakes. Alternatively, harmonic dynamic multipliers can be assigned to external loads to simulate vibrating sources. Hysteretic damping is provided by the Hardening Soil small-strain model and corresponding damping curves are displayed in the material data set. Rayleigh can be defined to include viscous damping of soils and structures on the basis of the input of a target damping ratio and frequency range. Viscous boundaries can be applied to absorb waves at the model boundaries in order to prevent spurious reflections. Besides the new 3D Dynamic module, the Plaxis research team has worked on improved dynamic features, which will first become available in the new PLAXIS 2D 2012 version.

The main new features are:• Free field boundaries: These special boundaries

adopt exactly the material behaviour of the adjacent soil elements, which provide an improved absorption of waves at the model boundaries when using advanced soil models.

• New UBCSAND model: The 904aR version of UBCSAND (Beatty & Byrne, 2011) gives a better cyclic loading behaviour and a better prediction of pore pressures and liquefaction.

In addition to PLAXIS 2D, a new Dynamic module has become available for PLAXIS 3D. With this module it is possible to

perform three-dimensional calculations of vibrations in the soil, as well as simulations of earthquakes in order to analyse the

influence of dynamic effects on buildings and other structures in and on the ground.

Ronald Brinkgreve, Plaxis bv


»The project is the construction of a container port along the Cai Mep River. The riverbank

length is approximately 500m, and container storage yard depth ranges between 600 to 750m. The total area covered by the port land is about 330,000 m² (excluding berth). The port comprises an offshore berth (reinforced concrete slab deck on piles), along which the ships will moor, and on which will be installed the unloading cranes. The berth is located 100 to 150m away from the river shore, and a “reclamation” area, where sand fill will be installed over the existing natural ground (swamp). Due to the draft of the ships, the natural riverbed will be dredged down. The reclamation area will welcome the container storage area, surrounding service roads, and ancillary buildings. The berth and the reclamation area will be connected by three approach bridges, also RC deck on piles. Due to its very poor mechanical properties, the natural soil will settle a very large amount under the load of the sand fill. Therefore, ground improvement is foreseen to address this issue. Main resultsIn this study, a finite element (FE) model has been set-up to analyze consolidation and lateral displacement of soft ground with prefabricated vertical drain (PVD) under isotropic consolidation.

A two-dimensional idealization is used for simplicity. Soft Soil model is used to idealize the soft ground. An equivalent permeability, based

on the equal discharge rate in the model and in the field, is used to characterize flow. Isotropic consolidation is simulated by specifying negative pore pressure-time history on the PVD boundaries and sand blanket. Varying the negative pressure inputs, several FE simulations are performed to predict the field behavior on soft clay.

The FE analysis was carried out in the framework of a fully coupled flow-stress analysis with unsaturated soil condition. For this purpose, a new drain element has been implemented on which negative pore pressure can be applied as a flow boundary condition.

The capability of the finite element method to properly model isotropic consolidation has been clearly highlighted during this project by comparison with the measurements performed on the Cai Mep River project

The CompanyMenard is a design-build specialty geotechnical contractor offering expertise on ground improvement for sites with poor soil. Menard combines value engineering and innovative techniques to deliver practical, sustainable solutions that can be attractive alternatives to deep (pile) foundations. From design to construction, MENARD offers a complete service for improving soil foundations.

Menard and Plaxis Expert Services have elaborated and successfully conducted a common research program for studying in

detail the effects of isotropic consolidation on soft soil. In the framework of PLAXIS Expert Services, extensive finite element

modelling work has been carried out in PLAXIS 2D and the results obtained in this context have been checked against available

in-situ measurements for a land reclamation project currently being executed and managed by Menard in Vietnam.

Jerome Racinais, Menard

PLAXIS Expert Services:Advanced knowledge on isotropic consolidation

“The extended collaboration and fruitful exchange between Menard and the Plaxis Expert Services team enabled to end-up

in a very realistic FE modeling.”



» As part of an extensive oil field development in the Deep Continental Shelf, Offshore

India, three new well platforms are to be installed in a water depth of about 85-90m. During the site survey, it was discovered that the target position for one of the platforms was located next to large pug marks, as shown on Figure 1a. These pug marks were formed earlier by a jack-up rig that had been deployed at this location. As a result, the soil has been remoulded by the penetration and extraction of 17m diameter spudcans, causing seabed depressions of about 30m diameter and 2m depth that are still present.

The jacket platform is to be temporarily supported during installation by a 40m by 40m square mudmat before piles are driven. During installation, the mudmat will be subjected to combined vertical, horizontal and moment (VHM) loading resulting from an eccentric gravity load and environmental actions. Due to the presence of the remoulded zone and seabed depression close to the mudmat, the stability during installation would be influenced and the effect of the pug marks needed to be analysed.

The position of the mudmat relative to the pug marks is shown on Figure 1b. Only one pug mark would potentially affect the stability of the mudmat. As the problem is three-dimensional and cannot be properly analysed with simplified approaches and/or two-dimensional analysis, finite element analyses were undertaken using PLAXIS 3D.

Jacket platforms used offshore for oil extraction are generally temporarily supported by mudmats during installation. These

platforms are not always installed on virgin seabed but are sometimes located close to features such as pug marks formed by

previous deployments of jack-up rigs. These seabed features potentially influence the bearing capacity of the mudmats and

need to be accounted for in the stability verification.

Jean-Christophe Ballard & Nicolas Charue, Fugro GeoConsulting, Brussels, Belgium

Figure 1a: Geophysical image of pug marks

Figure 1b: Mudmat relative to pug marks


Site ConditionsThe soil conditions at the site consist of alternate layers of clay and sand. The top soil layer consists of very soft clay up to a depth of about 9 m. This layer is underlain by a medium dense carbonate sand layer from 9 to 27 m depth, Figure 2. Below that depth, stiff clay is present.

Due to the presence of pug marks close to the target position for one of the platforms, a geotechnical investigation was carried out to characterize the ground conditions in the vicinity of the pug marks. The purpose was to measure the remoulded shear strength parameters and map the extent of the remoulded areas. The site investigation programme consisted of cone penetration tests as well as drilling with sampling.

The data from the geotechnical investigation enabled three zones around the pug marks to be defined: a fully disturbed zone, a partially disturbed zone and an intact zone. The corner of the mudmat is in contact with the partially disturbed zone of one of the pug marks, as shown on Figure 2. The undrained shear strength profile for each zone is also defined on Figure 2 (where z is the depth below the initial (undisturbed) seabed in meters). The extent of the zones of soil disturbance was checked by simulating the penetration and extraction of the spudcans and good agreement was found. (These analyses are not presented in this article).

The sand properties were found to be of secondary importance as the failure mechanism for the mudmat develops in the top soft clay layer, as described below.

Stability Assessment of Jacket MudmatThe presence of the pug mark in the vicinity of the jacket mudmat is expected to reduce the safety factor against bearing capacity failure as well as cause tilt during settlement of the structure supported by the mudmat. The purpose of the

Figure 2: Site conditions and soft clay layer properties



analysis was twofold: first to check the safety factor against bearing capacity failure allowing for the full 3D geometry of the problem, and second to assess any tilt during mudmat settlement due to the presence of the pug mark. The most critical load case was where the moment loading acts in the direction of the pug mark.

Problem geometryThe lateral dimensions of the 3D model are 95 x 110 m. The model thickness is 30 m assuming the soil stratigraphy described before. These dimensions were selected such that the model boundaries have negligible effects on the results. As shown on Figure 3, a cylindrical zone of 18 m radius in the top clay layer is considered to be fully remoulded by the prior penetration of the spudcan at the pug mark. In this area, a seabed depression of 2 m is considered. A further 4 m wide zone with partially remoulded conditions is considered around the fully remoulded zone. The partially remoulded zone extends below the mudmat corner.

Load casesSimplified VHM load cases where the moment and horizontal loads act along the diagonal of the mudmat in the direction of the pug mark were first analysed. VHM yield surfaces were developed to identify the most critical load paths and make sure the minimum safety factors were met for the different scenarios. Specific load cases were then verified with the same model. Eccentric vertical loads were used to introduce moment loads.

Details of the finite element modelThe finite element mesh is shown on Figure 4. The mesh comprises a number of 10-noded tetrahedral elements. The mesh global coarseness was first set to “Medium”. Then the local refinement factor was decreased to 0.35 in the upper clay layer where the failure mechanism develops. In total, the model comprises 70375 elements and the average element size is about 2 m.

Preliminary analyses were first performed for the base case without a pug mark and for which analytical solutions exist. The aim was to check for any effects due to mesh size on the accuracy of the

Figure 3: PLAXIS 3D model geometry

solution. A compromise was found between the accuracy of the solution and computational time. It was estimated that the over-estimation of the true solution due to discretization errors was maximum 5% for the selected mesh, which was judged to be reasonable.

The soil was modelled as an isotropic elasto-perfectly plastic continuum, with failure described by the Mohr-Coulomb yield criterion. The clay layers are assumed to behave “undrained” and are characterized by a cohesion equal to the undrained shear strength su with u=0. The elastic behaviour was defined by a Poisson’s ratio =0.49, and a constant ratio of Young’s modulus to undrained shear strength E/su. The sand layer is assumed to behave “drained” and is characterized by effective stress shear strength parameters c’ and ’.

The strength of the mudmat/clay interface was modelled using an interface factor a, where the

Table 1: Selected parameters for 3D FE analyses

Soil Type CLAYIntact CLAYRemoulded CLAYPartially Remoulded SAND

Submerged unit weight ' [kN/m³] 5 4 5 9.5

Young’s modulus at mudlineE [kPa] 2400 0 1200 10000

Poisson’s ration [-] 0.49 0.49 0.49 0.2

Undrained shear strength at mudlinesu [kPa] 4 0 2 -

Cohesionc’ [kPa] - - - 0.1

Angle of friction ’ [°] - - - 30

Angle of dilation [°] 0 0 0 0

Rate of increase of E with depthEinc [kPa/m] 397 397 397 0

Rate of increase of su with depthsu,inc [kPa/m] 0.667 0.667 0.667 -

Interface factor [-] 0.5 - 0.5 -

maximum shear stress at the interface max= su. The “rough” and “smooth extremes of interface strength correspond to =1 and =0 respectively. An intermediate roughness was assumed with =0.5, which is a typical assumption for steel/soft clay interface. A no-tension condition allowing separation of the mudmat from the seabed was permitted at the mudmat/clay interface.

The jacket mudmat is modelled as a 40m by 40m rigid plain square plate. The seabed is assumed to be perfectly flat below the mudmat.

Design parametersThe selected parameters for the 3D FE analyses are summarized in Table 2. The elastic behaviour of the clay layers was defined by a ratio of Young’s modulus to undrained shear strength E/su = 600. This is a typical value for clays with a Plasticity Index around 30%. The selected parameters for the sand layer are also presented.

Figure 4: PLAXIS 3D mesh

ν

φ

α

α α

α

γ

ν

ψ

φ

α

φ

τ



Figure 7: Typical failure mechanism (total displacements at failure)

Figure 5: VM envelopes with H constant and with/without pug mark

Figure 6: Load-tilt curve with H constant with/without pug mark, for design vertical load

Figure 4: PLAXIS 3D mesh

ResultsThe obtained VM yield surfaces for different scenarios are shown on Figure 5. These results show how the presence of the pug mark and a horizontal force degrades the VM capacity of the mudmat. This reduction can potentially lead to safety factors against bearing capacity failure that become unacceptable. Note that for small vertical loads, the moment resistance decreases. This is because of the no tension assumption at the mudmat/clay interface. In reality, for rapid loading and if the the mudmat is not perforated, some tension may develop at the interface and yield higher moment capacity. The no tension assumption is therefore a cautious approach in this case. Safety factors have to be applied to the ultimate yield surfaces to obtain the allowable surfaces. Then, it must be verified that the design load cases are within the allowable surfaces. This guarantees that the safety factor is met for every load path. Safety factors in the range 1.5 to 2.0 are generally used depending on the nature of the loading (i.e. short-term or permanent). The allowable yield surface can either be obtained by dividing the ultimate values by the appropriate safety factor or by using directly reduced undrained shear strength values in the analyses. The two approaches give the same result in this case.

The presence of the pug mark and the applied horizontal force also has an impact on the induced tilt, as shown on Figure 6. The induced tilt in degrees along the diagonal in the direction of the pug mark is plotted versus the applied moment for the cases with / without pug mark and with / without horizontal force. As an example, a typical failure mechanism is illustrated on Figure 7 when the moment loads are acting along the mudmat diagonal in the direction of the pug mark.

ConclusionsA shallow foundation subjected to a combined VHM loading and located next to a pug mark is a 3D problem for which simplified approaches for analysis do not exist. This type of problem needs to be analysed by means of 3D FE analyses. The software package PLAXIS 3D has been used successfully on this project. The analysis allowed confidence to be established for the selected location of the mudmat with respect to the pug mark. In contrast, a simplified 2D analysis would suggest that the proximity of the mud mat to the pug mark was unacceptable.


» The previous slip road in Woerden needed to be reconstructed. Without advanced

construction methods, the soft soil on the location would have caused severe differential settlements. Usually, such soft soils are preloaded during a long time. However, the option of preloading was rejected in this case, because of required time span. Another option, the application of Styrofoam was rejected, because the road should remain available in case of flooding.

The option of a piled embankment was considered. The advantages of the piled embankment are the small settlement, the independence of heterogeneity of the subsoil, long lifetime, and short time between the start of the construction and time to take the slip road into service.

The construction contains about 900 concrete prefabricated piles, a geosynthetic reinforcement and a layer of construction aggregate. The piles have substantial differences in length, because the required bearing capacity can only be acquired in the deeper layer of sand that varies in height.

The representative cross section is located on a canting part of the road. The section contains an

old ditch that is filled with water. For the purpose of drainage a new ditch was required. The ditches are shown in the figures.The original design of the reinforcement in the piles below the slip road was made in PLAXIS 2D 9.0. The embankment contains several sensors to monitor displacement, forces en bending moments in the piles in a test section of the road. In order to make an accurate prediction of these measurements, a model was made in 3DTunnel 2.4. Recently, the model was also rebuilt in PLAXIS 3D 2011.

The cross sections of all three different models are almost the same. The layers containing peat or soft clay are modelled with the Soft Soil Creep model, and the other layers are modelled with the Hardening Soil model. The model in 3DTunnel has the smallest depth that is allowed by symmetry rules; only half a pile and half of the space between the piles. This choice had to be made because of the limited calculation possibilities with 3DTunnel. The model in PLAXIS 3D 2011 contains two rows of piles and half of the space between the piles on both sides of the model.

The most interesting parts of the differences between the 3 models are the geogrid and the

In recent years, there is a growing interest in the use of piled embankments. This interest also initiated several investigations, the publication of for example the Dutch design guideline for piled embankments (CUR226, 2010) and the German EBGEO (2010). The slip road, containing a piled embankment, of the highway A12 in Woerden (25 km south of Amsterdam) was designed using CUR226. The reinforcement of the piles was, as prescribed in CUR226, designed using PLAXIS 2D (9.0). During the construction of the road several types of sensors were placed on the piles, pile caps and geogrids. In order to predict especially the moments in the piles, measured with optical fibres attached to the steel piles, a 3DTunnel model was build. Recently, this model was adapted and rebuild in PLAXIS 3D 2011. This paper compares the results of the models with PLAXIS 2D (9.0), 3DTunnel (2.4) and PLAXIS 3D 2011. The measurements are not yet available.


Eelco Oskam, Movares, Utrecht, The Netherlands

piles. Except the piles, all material models are equal in the 3 PLAXIS versions. In the PLAXIS 2D model, plates are used with a hinge to model the piles. In the 3DTunnel model, a pile is constructed out of three parts. Two clusters (per layer) of concrete piles are used in combination with a plate. This plate has a bending stiffness of 1/1000 of the real pile. On top of the piles, there are two node-to-node anchors connected to the pile cap, in order to create a hinge. For the PLAXIS 3D 2011 model, an embedded pile with hinge was used.

The 2D-model only determines the tensile forces in the geogrids in the direction perpendicular to the road. The direction along the road is obviously not modelled. In the 3DTunnel and PLAXIS 3D 2011 models, the shape and density of the meshes are different. The PLAXIS 3D model has a finer mesh and the shapes of the elements in the geogrids are triangular.

For the situation without traffic load, all three PLAXIS models predict similar bending moments in the piles, (with no correction for the different phases) after 10 years of settlement (see figure 1). In this figure the bending moments of the fourth pile from the left is shown. Note that the model of PLAXIS 3D 2011 contains two rows of piles.


For the situation with traffic load, it appears that the bending moments in the 3DTunnel model are smaller than the two other predictions, and probably not realistic. This difference is initiated in the start of the phases with updated mesh. The updated mesh is required to properly model the membrane effect in the geogrid. During the excavation, the bottom of the excavation shows a heave. When the piles are activated, a pre-stress occurs. This is the main cause for the smaller bending moments. For the same reason, a large bending moment in the model of PLAXIS 2D is present at -19m, but for this model correction is possible. Since the piles consist of one type of linear elements.

Figures 1 and 2 show the tensile forces in the geogrids, in the directions perpendicular to the road. It should be noted that CUR226 predicts a maximum tensile force along the road of 96 kN/m, and across the road of 112 kN/m. This agrees very well with the 3D predictions.

The differences between the PLAXIS models do not increase when the load is applied. The plots of the forces in the geogrids show that all PLAXIS models, and especially PLAXIS 2D, give peaks at the sides of the pile caps (these peaks decrease when the element sizes are reduced further), but generally, the maximum forces are similar to the PLAXIS 3D model. The forces in the geogrids in the 3D models are similar, but the PLAXIS 3D 2011 version gives higher values, especially in the direction parallel to the road.

It can be concluded that PLAXIS 3D 2011 is better suitable for modelling piles than 3DTunnel and gives very promising results. The most important improvement in PLAXIS 3D 2011 is the embedded pile, since the bending moment and initial strain of the piles in the installation phase of the piles (with updated mesh) are less disturbed. The modelling in PLAXIS 3D 2011 is also easier and faster than the modelling in 3DTunnel.

In the near future, the calculations results of the calculations will be compared to the measured data.

References• CUR 226, 2010. Ontwerprichtlijn paalmatrassys-

temen (Design guideline piled embankments), ISBN 978-90-376-0518-1 (in Dutch).

• EBGEO, 2010. Empfehlungen für den Ent-wurf und die Berechnung von Erdkörpern mit Bewehrungen aus Geokunststoffen – EBGEO, 2. Auflage, German Geotechnical Society, ISBN 978-3-433-02950-3 (in German). Also available in English: Recommendations for Design and Analysis of Earth Structures using Geosyn-thetic Reinforcements – EBGEO, 2011. ISBN 978-3-433-02983-1 and digital in English ISBN 978-3-433-60093-1).

Figure 2: Force in geogrid, with traffic loadingFigure 1: Bending moment in pile, without traffic loading

PLAXIS 2D 3DTunnel PLAXIS 3D


Piled embankments with geosynthetic reinforcement are applied on soft soils and have several advantages. For example, the piled embankment can be constructed rather fast and has a small settlement after construction or is even settlement-free. Another advantage is that a piled embankment can be built next to sensitive constructions. A piled embankment consists of a field of piles with pile caps. On top of that, one or more layers of geosynthetic reinforcement (GR) are applied. On top of the GR the embankment can be constructed.


Ir. Theresa den Boogert, TU Delft (now Mobilis), Ing. Piet van Duijnen, Mobilis and Ir. Suzanne van Eekelen, Deltares/TU-DelftFoto: Project N210, Huesker

Figure 1: Load distribution in reinforced piled embankment

» In 2010 the Dutch design guideline, CUR226 (2010), for the design of piled embankmenst

was published. To validate the guideline, several field tests have been performed. From the field measurements it has been concluded that the design method is rather conservative. Improving the design guideline would reduce the construction costs of piled embankments. To understand the physical behaviour of the piled embankment and to validate design models, experimental scale tests have been performed by Deltares in partnership with Huesker, Naue, TenCate and Tensar. The results of the scale tests were analysed and published by Van Eekelen, et al. (2011a, 2011b and 2011c).

Plaxis simulations of the test series were performed to improve the understanding of the arching mechanism in the piled embankment, and where possible, to confirm the conclusions from the analysis of the scale tests. The simulations are part of the master thesis performed by Den Boogert (2011). First the definition of the load distribution in the embankment will be presented. The scale tests are described in the second paragraph. Then the content of the finite element model is

discussed. The results of the finite element model are analysed and compared to the results of the scale test. The paper ends with conclusions and recommendations. Definition of load distributionThe vertical load on the piled reinforced embankment is distributed to the soft subsoil in three load parts: A, B and C (shown in figure 1). The load parts are defined by: part A is transferred directly to the piles by arching, part B is transported via the GR to the piles, and load


part C is carried by the soft subsoil. The load parts are vertical loads and are given in kN/pile. Scale testsA section of an embankment is modelled in a metal box of 1.1 x 1.1 x 1 m3. Four piles are situated on the bottom of the box. The soft subsoil between the piles is modelled with a watertight foam cushion filled with water. A tap allows drainage from the foam cushion during the test, which models the consolidation process of the soft subsoil. The GR is attached to a steel frame and situated on top of the foam cushion with a sand layer of ca. 2 cm in between. On top of the GR, an embankment of 0.42 m is constructed of granular

material (crushed rubble). The top load on the embankment is applied with a water cushion. This provides an equally distributed top load. The metal box is closed by a cover and tie rods. A side and top view of the scale test set-up is given in figure 2. The scale tests are performed in several steps of consolidation by draining the foam cushion and increasing top load. The load steps and consolidation steps alternate: each top load step is followed by ca. 3 consolidation steps. At the end of the scale test, vacuum pressure is applied to the foam cushion. This reduces the subsoil support to zero. After every drainage or top load step, the

system is allowed to stabilise for several hours. The load distribution is measured with pressure cells. Pressure cells are placed on top of piles, one above and one underneath the GR. The pressure cell above the GR measures load part A and the pressure cell underneath the GR measures load parts A+B. Load part B is calculated by subtracting load part A from load parts A+B. Additionally, the pressure in the foam cushion is measured, which gives load part C. The top load is measured with a water pressure meter in the water cushion. The vertical deformation of the GR is measured on three locations with a liquid levelling system. The locations of the measurements are given in figure 2.

Figure 2: Side view and top view of scale test set-up



Finite element modelThe scale tests are simulated with 3DTunnel version 2.4. 3DTunnel was used, because updated mesh could be applied and arching in the embankment is a 3D problem. Updated mesh is necessary to use, because the function of the GR depends on the deformation and the tension force cannot be modelled if the deformation is not captured in the calculation. The updated mesh function captures the tensile strains in the geosynthetic elements and the geosynthetic is no longer horizontal. Next to that the new PLAXIS 3D version was not available at the time. The geometry of the model is based on the geometry of the scale test. Because the geometry of the scale test is symmetric, one quarter of the scale test, one pile with surrounding soil, is modelled. The boundary conditions are horizontally fixed. The side and top view of the model are presented in figure 3. The material properties are summarized in table 1 and 2 and will be described in the next section. In the test series, circular piles are applied. For the Plaxis simulations, the geometry of the circular pile is converted to a square pile. The properties of the pile are based on the parameters of PVC. PVC is modelled linear elastic and non-porous material. Next to the pile the foam cushion is modelled. The watertight and soaked foam cushion behaves linear elastic in the scale test. The scale tests were controlled by both top load and draining the foam cushion and therefore decreasing the water

Figure 3: Top view and side view of finite element model

g[kN/m3] Eref [kN/m2] u[-] EA [kN/m]

Pile 13.6 2.9E6 0.0 -

Subsoil 10.2 10 0.2 -

GR - - - 2269

Frame 70.5 2.1E8 0 -

g[kN/m3] c [kN/m2] f [°] y [°] E50 [kN/m2] Eoed [kN/m2] m [-] Eur [kN/m2] uur [-] rref [kN/m2] R1 [-]

Sand above pile 20.1 1 40.9 10.9 51470 51470 0.5 154410 0.2 100 0.9

Sand next to pile 18.7 1 32.5 2.5 19660 19660 0.5 58980 0.2 100 0.9

Granular material 16.7 1 47.0 11.0 58870 58870 0.7 176610 0.2 100 0.9

Table 1: material properties of pile, subsoil, GR and frame (linear elastic)

Table 2: material properties of sand and granular material (Hardening Soil model)

pressure in the foam cushion. To simulate the drainage of the scale test, the measured water pressure is prescribed in the model by applying a phreatic level to the clusters of the foam cushion. Therefore the measured water pressure is converted into a pressure head. The axial stiffness of the GR is determined from five tensile tests. The tensile tests are performed according to DIN EN ISO 10319. The GR is attached to a steel frame. The steel frame is modelled, the weight of the frame disturbs the load distribution. The parameters of the steel frame are based on the properties of steel. The sand layer on the pile and foam cushion and the granular material are modelled with the Hardening Soil model. The parameters of the sand and granular material are determined with triaxial

tests. The sand layer is split up in two parts, a part above the pile and a part directly on the subsoil. The parameters are different for both parts. The sand on top of the pile is expected to behave very stiff, because the sand on the pile will be clamped between the GR and the pile. Therefore, the sand on the pile will be compressed more and will have higher stiffness and strength properties. The sand on the subsoil will follow the settlements of the subsoil and geosynthetics. In figure 4 the 3D finite element mesh created by Plaxis is shown. During the execution of the scale tests, part of the load is dissipated due to friction. The friction between the wall and the granular material is between 10% and 20%. Normally an interface is applied to model the friction. This interface should be applied along the box walls, which means at the left and back side of the model. In Plaxis 3D

ref ref ref



Tunnel an interface cannot be applied at the back side. To keep the amount of load distribution comparable to the scale test, the top load is reduced by the amount of friction, and no interface is applied. The disturbance of the friction on the load distribution in the embankment is therefore neglected in the model. The friction between the piles and the foam cushion is assumed to be small and its influence on arching within the fill is limited. Therefore the friction between the pile and foam cushion is also neglected. The calculation phases of the model are based on the scale test procedure. The top load and water pressure measured during each step of the scale test procedure is an input value in the calculation phases. During the initial phase, the water pressure and SMweight are set to zero, to avoid an asymmetric situation. In the following phases, the scale test is build up and the soil weight is activated. Then the measured load with corresponding water pressure is applied. During the last phase, where there is no subsoil support against the GR, the subsoil and water pressure are deactivated. ResultsDuring the vacuum phase there is a constant high top load applied and there is no subsoil support. Therefore, the vacuum phase has the largest deformation. This is the most representative situation and will be presented in the figures below. The calculated principal stresses in the vacuum phase are shown in figure 5. From the figure, soil arching can be observed. The calculated vertical displacements are shown in figure 6. The differential displacements on top of the embankment are very small. The tensile forces in the GR are presented in figure 7. The tensile forces in the GR are concentrated in ‘tensile strips’. The tensile strips are the areas of that GR that lie on top of and between adjacent piles. The maximum tensile forces are found in the GR at the edge of the piles. The exact location of the peak values cannot be determined, because the mesh is too coarse.

Figure 4: 3D finite element model Figure 5: Effective principal stresses of the vacuum phase

Figure 6: Vertical displacement of the vacuum phase

The load distribution for the FEM model and the scale test are plotted in figure 8. The horizontal axis presents the net load. The net load is the top load minus subsoil support and friction. Load parts A and B are presented on the vertical axis in kN per pile and as percentage of the total load (A+B+C). The figures show two types of loading: top load increase and drainage of subsoil (consolidation). The load transferred directly to the piles due to arching is load A and the load transferred through the GR to the piles is load B. During the first part of the test, until the net load is ca.11 kN/pile, the calculated results of load parts A and B agree quite well with the measured results. Then the calculated results diverge from the measured results. Load part A is overestimated and load part B is underestimated. The calculated

load parts A and B show a smooth relationship with the net load. This agrees with the conclusion of the measurements. During the first drainage step with zero top load, the percentage of load part A (A %) increases significantly. This means that arching occurs immediately. Not only during the first drainage step, but also in the following drainage steps load part A % increases. This shows that subsoil settlement is needed for the development of arching. This conclusion can be drawn for both the measured and calculated results. During the steps with increasing top load, load part A % on the embankment decreases. From this it follows that during increasing top load the arching effect decreases, as long as consolidation does not occur.



Figure 9: Displacement

Figure 8: Load part A and B in kN/pile and in % of total load A+B+C

Figure 7: Tensile forces in GR of vacuum phase in x-direction (left) and in y- direction (right)



Displacements have been measured at three places: in the middle of four piles (z1), in the middle of two piles (z2) and close to a pile (z3). Displacements z1 and z3 are shown in figure 9. Displacement z2 is not shown, because this displacement does not differ from displacement z3. The displacement is presented as a function of the net load. From the comparison of results between Plaxis and the measurements it can be concluded that the displacement is underestimated significantly by Plaxis. Farag (2008) also found much lower settlements in his Plaxis calculations. In CUR 226 (2010) this is solved by modelling a gap underneath the GR in the Plaxis calculations. Several possible causes of the underestimated displacements have been investigated: among them the behaviour of the subsoil, of the GR and of the granular material. Each individual aspect gives a very limited improvement conform the measured displacements. Therefore, the cause of the underestimated displacements should be investigated in more detail. Conclusions and recommendationsFrom the FEM model is concluded that arching occurs in the granular material. The effective vertical stresses are concentrated on top and the area next to the piles. The tensile forces in the GR are concentrated in ‘tensile strips’ between the piles. The exact location of the greatest tensile forces could not be determined because of the coarseness of the mesh. In accordance with the measurements, Plaxis calculations give a smooth relationship between the net load and load parts A and B, and the GR settlements. During the first part of the test, the load distribution of the model agrees quite well with the measured load distribution. During the second part of the test, the load transferred through arching is overestimated and the load transferred through the GR is underestimated. In general, Plaxis finds an increasing arch during drainage of the subsoil (consolidation), this is in agreement with the measurements. The displacements calculated with Plaxis are underestimated compared to the scale test results. However, the largest displacement of the GR is found at the middle of four piles. To increase the accuracy of the Plaxis results, the model should be calculated with a more refined mesh. The loss of load (due to friction) during the scale test is an important part of the scale test. It produces disturbance in the load distribution of the granular material. The friction should be included in the Plaxis model by an interface, therefore it is advised to perform numerical analysis of the scale model test with the full 3D version of Plaxis. The cause of the differences in calculated and measured displacement of the GR has to be investigated in more detail.

References• CUR 226, 2010, Ontwerprichtlijn paalmatrassys-

temen ISBN 978-90-376-0518-1 (in Dutch)• Den Boogert, T.J.M., 2011. Piled embankments

with geosynthetic reinforcement, Numerical analysis of scale model tests, Master of Science thesis, Delft University of Technology.

• Van Eekelen, S.J.M., Bezuijen, A., Lodder, H.-J. & Van Tol, A.F., 2011a. Model experiments on piled embankments Part I, Geotextiles and Geomembranes, 2011, http://dx.doi.org/10.1016/j.geotexmem.2011.11.002

• Van Eekelen, S.J.M., Lodder, H.J., Bezuijen, A., 2011b, Paalmatrasproeven I, Vervormingen van geokunststoffen in een paalmatras en de daa-ruit volgende belastingsverdeling, GeoKunst 42, april 2011, 42-44

• Van Eekelen, S.J.M., Van der Vegt, J.W.G., Lodder, H.J., Bezuijen, A., 2011c, Paalmatraspro-even II, belangrijkste conclusies, GeoKunst 43, juli 2011, pp 46-50


Recent activities

Plaxis Asia-Pacific updateUnder the PLAXIS Expert Services, PLAXIS AsiaPac office has conducted two sessions of PLAXIS trainings since commencing in October 2011. The 2 days trainings offered an introduction to PLAXIS 2D targeted mainly to first-time PLAXIS users. Practicing engineers including a handful of professional engineers, make-up the trainings’ attendees. However, the class size was limited to 6 for each session partly due to the available space in the Singapore office but more objectively, to

allow close interaction between the course trainer and the participants.

The content of the course covers the basic functionality of the PLAXIS 2D program and also the fundamental principles of the computational geomechanics. This includes the learning of; applicability of the structural elements, finite element meshing and boundary conditions with respect to different geotechnical problems, initial stresses set-up, safety analyses, concept

of plasticity, modeling of undrained problems, introduction of an advanced soil model and many more.

The recent course also covered briefly on the newly built feature - design approaches. Four hands-on examples were also provided in the trainings for familiarization with the PLAXIS program.

The overall structure of this tailored basic training enables the participants to attend the PLAXIS standard and advanced courses with learning ease and better understanding.

Following the success of the October 2011 and January 2012 intake, PLAXIS AsiaPac is considering offering this well-recognized course to its region, on a more regular basis.

Plaxis USAA special activity organized by Plaxis USA was a week of visits to three state DOTs (Department of Transport). While some state DOTs are well aware of Plaxis and also use Plaxis, others are not yet familiar with Plaxis. Plaxis USA recognizes the important role DOTs play in civil, structural and geotechnical engineering in their respective states. Therefore one-day seminars were provided at the Louisiana Department of Transport and Development (LADOTD), Virginia Department of Transport (VDOT), and North Carolina Department of Transport (NCDOT). Topics included FE, slope stability analysis, foundation design and the use of PLAXIS. Each seminar had a turn-up of 20 to 25. These seminars took place in August 2011 and were organized by Plaxis USA representative Jasper Van Der Bruggen.

Heavy traffic at PLAXIS’ exhibitor booth could be seen at the Pan-Am CGS Geotechnical Conference (Toronto, October 2011) and the Annual DFI conference (Boston, October 2011). Attendees at Pan-Am CGS came from across the Americas,


and many PLAXIS users at this conference expressed interest in the latest developments of the PLAXIS Dynamics module, while many others took an introductory DVD with them. At the DFI conference many were interested in the latest developments regarding PLAXIS 3D.

Training courses were organized in Portland OR (August 2011) and Atlanta (February 2012), both cities hosting a Plaxis course for the first time.

The course in Portland included a special day on dynamic analyses and earthquake engineering. Professor Andrew Whittle (MIT) and Professor Juan Pestana (UC Berkeley) gave lectures in Portland.

The course in Atlanta concluded with a special day on 3D modeling and foundations, guest instructors were Professor Richard Finno (Northwestern) and Dominic Assimaki (Georgia Tech).

Knowledge BaseLast year we launched the new Plaxis Knowledge Base. This Knowledge Base is the platform to share our knowledge with the Plaxis community. Within the Knowledge Base you can find an array of support information from instructional movies to interesting articles, tips and tricks, known issues, and much more. Here you can also find our technical download center, with drivers, Codemeter files, PLAXIS Connect and more.

All previously published Plaxis Bulletin articles will also be added to the knowledge base, as well as interesting (scientific) articles and publications. Thanks to the advanced search engine that can even search in PDF files, it will be much easier to find, for example, an old Plaxis Bulletin article by searching just for keywords or the author.

In the past, some of our users responded that our communication with users about Tips and Tricks, but also known issues, was not clear or transparent. By giving these two important items a place in the Plaxis Knowledge base, we hope to connect with the Plaxis community in a more transparent and open manner.

To make sure that this information reaches the Plaxis community, we made a coupling with the PLAXIS Connect program: Plaxis users will stay up to date with the latest news, known issues with tips how to circumvent the problem, will be offered useful tips and tricks, and even instructional movies, to make geotechnical modelling with Plaxis even easier. This is also a good reason to have PLAXIS Connect installed on your computer.

We will continue to expand the information in the Plaxis Knowledge Base over time. So take a look every now and then, and stay up to date with your Plaxis Knowledge. You will find the Knowledge Base under the support menu on the Plaxis website or go directly to kb.plaxis.nl

Title

16 Jalan Kilang Timor#05-08 Redhill Forum

159308 Singapore

P.O. Box 572 2600 AN Delft

The Netherlands

www.plaxis.nlTel +31 (0)15 2517 720Fax +31 (0)15 2573 107

Plaxis AsiaPacific Pte LtdSingapore

Tel +65 6325 4191

Plaxis bvDelftechpark 53

2628 XJ Delft

March 11 - 13, 2012Standard Course on Computational GeotecnicsDubai, U.A.E.

March 14 - 15, 2012DFI Middle East ConferenceDubai, U.A.E.

March 19 – 22, 2012Advanced Course on Computational GeotechnicsSchiphol, The Netherlands

March 20 - 22, 20123rd Brazilian Congress on Tunnels and Underground StructuresSão Paulo, Brazil

March 25 – 29, 2012Geo-Congress 2012Oakland CA, U.S.A.

April 12, 2012Singaporean Plaxis Users MeetingSingapore

April 17 – 19, 2012Advanced Course on Computational GeotechnicsIstanbul, Turkey

April 19, 2012Malaysian Plaxis Users MeetingKuala Lumpur, Malaysia

April 19, 2012DFI Liquefaction State-of-the-Art ForumSt. Louis MO, U.S.A.

April 26, 2012PLAXIS SeminarLondon, United Kingdom

April 27, 2012Masterclass “Risk Related to Geotechnical Finite Element Analysis”Delft, The Netherlands

May 28 – 30, 20122nd International Conference on Performance-based Design in Earthquake Geotechnical EngineeringTaormina, Italy

May 31 – June 2, 201212th Baltic Sea Geotechnical ConferenceRostock, Germany

June 5 - 7, 2012Advanced Course on Computational GeotechnicsSingapore

June 11 – 15, 2012Curso Avanzado de Geotecnia ComputacionalQuerétaro, Mexico

June 19 - 21, 2012Standard Course on Computational GeotechnicsKuala Lumpur, Malaysia

June 20 – 22, 20122nd European Conference on Unsaturated SoilsNapoli, Italy

June 27 – 29, 2012Advanced Course on Computational GeotechnicsNew York NY, U.S.A.

June 26 - 27, 2012Russian Plaxis Users MeetingSt. Petersburg, Russia

July 15 – 18, 2012ANZ 2012Melbourne, Australia

September 24 - 27, 2012Advanced Course on Computational GeotechnicsTrondheim, Norway

September 26, 2012Norwegian Plaxis Users Meeting 2012Trondheim, Norway

November 14 – 16, 2012European Plaxis Users Meeting 2012 Karlsruhe, Germany

Activities 2012

plaxis bulletin spring 2012

Documents