Seminar-Advanced Geotechnical Finite Element Modeling in Analysis using PLAXIS.pdf

PLAXIS JAKARTA 2012

Plaxis Seminar, Binus University, Jakarta, Indonesia, 2012

Advanced Geotechnical Geotechnical Finite Element Modeling using PLAXIS Dr William WL Cheang Principal Geotechnical Consultant Plaxis AsiaPac Lecture notes are contributed by: Dr Lee Siew Wei Prof. Harry Tan A.Prof. Ronald Brinkgreve Dr Shen Rui Fu Ir Dennis Waterman

CONTENTS A. Section Section 1: Geotechnical Analysis using PLAXIS Programs B. Sectio Section n 2: Modelling of Deep of Deep Excavations C. Sectio Section n 3: Modelling of Piled of Piled Foundations D. Section 4: Modelling of Tunnel of Tunnel‐Soil‐Structure Interaction Problems E. Conclusions Conclusions F. Referenc References es

Plaxis Seminar, Jakarta 2012

PLAXIS JAKARTA 2012

GEOTECHNICAL ANALYSIS USING PLAXIS FINITE ELEMENT CODES

SECTION 1.0


SECTION 1 A. Versions Versions 1. Pre Pre 2010 (Version 7.x, 8.x and 9.x) 2. Post Post 2010 (Version 2010, 2011, 2012…) B. New New Developments (2011…2012) 1. On‐going software developments 2. Researc Research h projects 3. Conclusio Conclusions ns


PLAXIS JAKARTA 2012


SECTION 1.0


SECTION 1 A. Versions Versions 1. Pre Pre 2010 (Version 7.x, 8.x and 9.x) 2. Post Post 2010 (Version 2010, 2011, 2012…) B. New New Developments (2011…2012) 1. On‐going software developments 2. Researc Research h projects 3. Conclusio Conclusions ns


PLAXIS JAKARTA 2012

Plaxis 2D: Features The PLAXIS 2D (Currently at v2010 moving to v2011)

•Program including the PLAXIS Dynamics and PLAXIS PlaxFlow modules of deformation • A finite element package intended for the two dimensional analysis of deformation and stability in geotechnical engineering The PLAXIS Dynamics Module

•An extension to PLAXIS 2D •The Dynamics module offers the tools to analyse the propagation of waves of waves through the soil and their influence on structures.

•This allows for the analysis of seismic of seismic loading as well as vibrations due to construction activities. •PLAXIS Dynamics offers the possibility to perform dynamic calculations in individual calculation phases.

PlaxFlow

• An add on module to the PLAXIS 2D program. of the non‐linear, time dependent and anisotropic behaviour of soils of soils • Simulation of the and/or rock in saturated and partially saturated situations.

Plaxis VIP These special extensions are: CAD Interfaces • New Material Models • • User Defined Soil Models • Multiphase Calculations • Sensitivity Analysis


Plaxis 2D v2011

Plaxis 2D Workflow can be found at: http://www.youtube.com/watch?v=LMy895GCsBQ&lis http://www.youtube.com/watch? v=LMy895GCsBQ&list=PLF7F3CDD69090AF3A&index=1&featur t=PLF7F3CDD69090AF3A&index=1&feature=plpp_video e=plpp_video


PLAXIS JAKARTA 2012

Plaxis 3D, 3DF & 3DT

1.

PLAXIS 3D is a finite element package intended for three‐ dimensional analysis of deformation and stability in geotechnical engineering. It is equipped with features to deal with various aspects of complex geotechnical structures and construction processes

1.

3DFoundation is a finite element package intended for the three‐

2.

3DTunnel is a geotechnical finite element package which is

dimensional deformation analysis of foundation structures

specifically intended for the three‐dimensional analysis of deformation and stability in tunnel projects.


Plaxis 3D v2011


PLAXIS JAKARTA 2012

PLAXIS 3D INPUT

General toolbar Mode switches Selection explorer

Drawing area Model explorer Mode toolbar Command line


Plaxis 3D Input : Modes

Definition of soil stratigraphy

Definition of structural elements, loads and boundary conditions

SOIL

STRUCTURES

Creation of the FE mesh

Definition of pressure distribution

MESH

WATER LEVELS

Definition of construction stages

STAGED CONSTRUCTION

Let me demonstrate!

PLAXIS JAKARTA 2012


SECTION 1.1: FEM MODELS


Tunnel‐Pile‐Soil Interaction 1

Bldg. load “Plate” modelling superstructure EI & EA

Building 40m Fill 1m

CDG

Tunnel

120m

48 Franki piles (Embedded Piles)

Tunnel advance

140m 6m Ø tunnel • Analysis by Plaxis 3D ( 70,000 Tets)


PLAXIS JAKARTA 2012

Tunnel‐Pile‐Soil Interaction 1 Iso-surface of soil total displacements

Isometric view

Tunnel advance

Pile group deformations

Front view

Tunnel advance

Animation Plaxis Seminar, Jakarta 2012

Piled Foundations 1


PLAXIS JAKARTA 2012

Piled Foundations 2


Piled Foundations 2

Piled Raft Foundation for a storage platform and Stacker Reclaimer Runways Plaxis Seminar, Jakarta 2012

PLAXIS JAKARTA 2012

Piled Foundations 3


Deep Excavation

Video Plaxis Seminar, Jakarta 2012

PLAXIS JAKARTA 2012

Dam:CFRD Malaysia



PLAXIS JAKARTA 2012

Model: CFRW‐CH300‐2D (South Sumatra 2007‐09‐12)



PLAXIS JAKARTA 2012

Model: Domain Mesh


Stability Analysis: MUDMAT


PLAXIS JAKARTA 2012

Filling of Spudcan Footprints:



PLAXIS JAKARTA 2012


SECTION 1.2: NEW DEVELOPMENTS


LATEST PRODUCT RELEASES PLAXIS 2D 2011 (December 2011)

1. Design approaches 2. Anisotropic plates and geogrids 3. Direct input of bending moments 4. Sekiguchi‐Ohta model


PLAXIS JAKARTA 2012

Design Approach Facility 1. Possibility to enter a coherent set of partial factors in one location (according Eurocode 7, LRFD, etc.) 2. More structured and efficient way of modeling 3. Easy exchange of Design Approaches between different projects 4. Partial factors definition remain the entire responsibility of the user (no default values for different building codes)


Orthotropic Plates and Geogrids – Independent definitio n of stif fness and strength properties with respect to element local axis

2 1


PLAXIS JAKARTA 2012

DIRECT INPUT OF BENDING MOMENTS


LATEST PRODUCT RELEASES

PLAXIS 3D 2011 1. Shape designer 2. Steady state groundwater flow analysis 3. Section contraction (tunnels and shafts) 4. Anisotropic geotextiles 5. Parallel computing 6. Output visualization during calculation


PLAXIS JAKARTA 2012

Shape Designer – Definition of polycurve (series of curved sections) which can then be extruded


Steady State Groundwater Flow Analysis – Pore pressure distribution in a dam during full pool conservation

PLAXIS JAKARTA 2012

SECTION CONTRACTION – To model volume loss during construction of tunnels or shafts

– Applicable to plates only – Contraction definition: section

c [ %] =

Ainitial

section

- Afinal

section Ainitial


PARALLEL COMPUTING – Reduce computation time by using domain decomposition

– Two new solvers available – PICOS solver (multicore iterative)

– PARDISO solver (multicore direct)


Contraction

PLAXIS JAKARTA 2012

PARALLEL COMPUTING EXAMPLE • Multi‐layer ground with tunnel : – 100 000 elements – 148 000 nodes – 414 000 d.o.f’s

Tot

Dec

Back

Iter

Cores

Speedup

175

57.5

90.1

62

1

1.000

133

21.5

84.2

74

2

1.396

84

18.5

40.5

39

4

2.501

69

13.7

29.3

45

8

3.432


OUTPUT VISUALIZATION DURING CALCULATION

Will open the Output program when the calculation is still running


PLAXIS JAKARTA 2012

ON-GOING SOFTWARE DEVELOPMENTS

– New modelling workflow PLAXIS 2D – Soil Constitutive Models – Model parameters definition from laboratory test results by inverse analysis in Soil Lab test

– Free‐field boundary elements – Reinforcement element for pile modelling in PLAXIS 2D – Structural forces in solid element in PLAXIS 2D – Thermo‐hydro‐mechanical coupling – New PLAXIS 3D add‐on modules: Dynamics and Transient GWF


NEW MODELLING WORKFLOW PLAXIS 2D


PLAXIS JAKARTA 2012

SOIL CONSTITUTIVE MODELS User‐defined soil models: 1. Anisotropic S‐Clay1(S) model

15 10 ] a P k [ y x



2. Anisotropic Creep Model

5 0 -5

0

50

100

150

200

-10 -15

3. Barcelona Basic model (unsaturated soil)

p'[kPa]

4. Hypoplastic model with intergranular strain 5. UBCSAND model (liquefaction)


PARA METER OPTIMIZATION BY INVERSE ANALYSIS

1. Based on Soil Test facility 2. Import of real lab test data (triaxial, oedometer) 3. Optimisation of selected model parameters based on particle swarm algorithm 4. Different curves can be considered simultaneously

 Best

match between curves from real tests and model simulation


PLAXIS JAKARTA 2012

PARA METER OPTIMIZATION BY INVERSE ANALYSIS - EXAMPLE Experimental parameters:  = 24º c = 5.5 kN/m2 Eoed= 9700 kN/m2 E50 = 9700 kN/m2 Calculated HS parameters:  = 24.30º c = 4.68 kN/m2 Eoed= 9627 kN/m2 E50 = 9509 kN/m2

Hardening Soil model fit 180 160 140 | 3 120 a m g 100 i S 1 80 a m g i S 60 |

40 Experimental

20

Calculated 0 0

0.02

0.04

0.06

0.08

0.1

0.12

Strain 1


FREE FIELD BOUNDARY ELEMENTS

– Free field condition definition – 1D soil column – Tied horizontal displacement on left and





right boundaries (Ux2=Ux1)

Y

X


PLAXIS JAKARTA 2012

FREE FIELD BOUNDARY ELEMENTS – Practical application of free field elements in PLAXIS Structure y r a d n u o b s u o c s i V

Soil

Structure y r a d n u o b s u o c s i V

Dynamic input (acc or vel)

n o i t i d n o c d l e i f e e r F

y r a d n u o b s u o c s i V

Soil

y r a d n u o b s u o c s i V

n o i t i d n o c d l e i f e e r F

Dynamic input (acc or vel)

Free Field elements


REINFORCEMENT ELEMENT IN PLAXIS 2D – Offer pile modelling capabilities in 2D • Development of line interface elements inserted between soil and the pile (Same modelling strategy as 3D embedded pile)

• The beam representing the pile slides over the 2D geometry and not through the 2D geometry

2D model


3D Equivalent representation

PLAXIS JAKARTA 2012

REINFORCEMENT ELEMENT IN PLAXIS 2D – Different than combining plate with surrounding interfaces • Soil cannot flow freely (as it should in between the piles) • Interfaces introduce unrealistic failure surfaces

2D model

3D Equivalent representation


Structural Forces in Solid Element in 2D – Beam modelled as solid elements under pure flexion

– View of integrated stresses along drawn neutral axis


PLAXIS JAKARTA 2012

THERMAL FLOW, THERMAL EXPANSION, THM COUPLING 1. Taking temperature effects into account: A. B. C. D.

Expansion Soil freezing Phase transition (ice, water, vapour) Change of properties

2. Geo‐energy applications A. Heat / cold storage B. Geothermal energy


Research Projects Participation in Research projects:

– Piles (inst. effects, embedded piles) i.c.w. TUD, TUGraz – Liquefaction of underwater slopes i.c.w. TUD – Geo‐Install (soil modelling, MPM) EU project (# partners) – Notes (dynamics) EU project i.c.w. TCD – Cyclic liquefaction, geotech EQ.eng. i.c.w. UC Berkeley, UIUC – Stochastic FEA i.c.w. TUD

PLAXIS JAKARTA 2012

MODELLING OF DEEP EXCAVATIONS

SECTION 2.0


GEOMETRY‐ MODEL DISCRETIZATION 2-D Plane Strain


3-D MODEL

PLAXIS JAKARTA 2012

GEOMETRY‐ MODEL DISCRETIZATION Axi-symmentry


PLAXIS JAKARTA 2012


3‐D MODELS

Piled building

Tower crane

N

Strut layout Piled building Plaxis Seminar, Jakarta 2012

PLAXIS JAKARTA 2012

3‐D MODEL OF AN EXCAVATION

Top of PW (70/90)

Top of Grade III or Better

N

Complex Soil-Structure Interaction Plaxis Seminar,Problem Jakarta 2012

CONSTITUTIVE MODELS

1. 2. 3.

Linear elastic, perfectly plastic Hyperbolic stress-strain curve (stiffness degradation for  > 1E-4) Non-linear stiffness from very small strains (1E-6)

1: Mohr Coulomb

1e-6

2: Hardening Soil Plaxis Seminar, Jakarta 2012

1e-5

1e-4

1e-3

1e-2

1e-1

3:Hardening Soil + Small Strain Overlay

PLAXIS JAKARTA 2012

SURFACE HEAVE IN INITIAL EXC./CANTILEVER WALL 3 m deep excavation with cantilever wall 20kPa 5m

3m 7m Dry sandy material

FSP III sheetpile

• 3 analyses with Mohr Coulomb, Hardening Soil & Hardening Soil-Small models using equivalent soil input parameters • Compare ground movements, wall displacements & wall stability Plaxis Seminar, Jakarta 2012

SOIL INPUT PARAMETERS FOR 3 ANALYSES Parameters for soil strength & initial stress state Analyses Material Model 1 2 3



(kN/m3) MC 20 HS 20 HSsmall 20

c'

'

(kPa) 5 5 5

(Deg) 35 35 35

 (or ur ) [-] 0.3 0.2 0.2

 

R inter

[-] 0.426 0.426 0.426

0.67 0.67 0.67

Parameters for soil stiffness prior to failure Analyses Material Model 1 2 3

MC HS HSsmall

Eref (or E50ref or Eoed ref ) (MPa) 30 30 30

Eur ref

pref

m

G0

0.7

(MPa) 90 90

(kPa) 100 100

[-] 0.5 0.5

(MPa) 150

[-] -5 2×10

• For derivation of soil stiffness parameters, a. HS model from standard drained triaxial compression tests b. HSsmall model from small-strain triaxial tests or field tests (e.g. downhole / crosshole seismic survey)


PLAXIS JAKARTA 2012

PREDICTED SURFACE SETTLEMENT BEHIND WALL Distance behind wall (m) 0

5

10

15

20

25

30

0.006

Heave

0.004 0.002 ) m 0.000 ( t n e m -0.002 e l t t e S -0.004

Settlement

-0.006

MC HS

-0.008

HSsmall -0.010

• MC predicts unrealistic surface heave 4 mm • HS & HSsmall predict max. surface settlement 9 mm


PREDICTED HEAVE AT EXC. LEVEL IN COFFERDAM Distance in front of wall (m) -5

-4

-3

-2

-1

0

1

2

3

0.025 MC

0.020

Wall

0.015 ) m ( e v a e H

0.010

0.005

0.000

-0.005

• MC predicts 20 mm heave at cofferdam centreline • HS & HSsmall predict 11 mm & 8 mm respectively


HS HSsmall

PLAXIS JAKARTA 2012

PREDICTED WALL RESULTANT DISPLACEMENT MC Ux=6mm

HS Ux=11mm

HSsmall Ux=10mm Ux: wall horizontal displacement


PREDICTED STABILITY OF WALL 3 Sum-Msf = FOS

FOS=2.8

2.5

MC

2

Rotation mechanism with FOS 2.8

1.5

3

Sum-Msf = FOS

FOS=2.8

2.5 2

HS

1.5

3 Sum-Msf = FOS

FOS=2.8

2.5 2

HSsmall

1.5


• “Phi-c' reduction” for predicting FOS • FSP III sheetpile properties: EI=34440 kNm2/m; EA=3.92×106kN/m Mp=369 kNm/m; Np=3575 kN/m

PLAXIS JAKARTA 2012

SUMMARY OF PREDICTIONS Analyses MC HS HSsmall

Surface settlement behind wall Heave 4 mm (not OK) Settle 9 mm Settle 9 mm

Heave at excavation level Heave 20 mm

Wall horizontal displacement 6 mm

FOS for wall stability 2.8

Heave 11 mm Heave 8 mm

11 mm 10 mm

2.8 2.8

1. MC predicts incorrect surface heave behind wall a. related to soil stiffness (E) prior to failure b. different ways of modelling E in 3 constitutive models 2. Stability of wall has FOS = 2.8 for 3 analyses a. related to soil shear strength b. all 3 constitutive models use Mohr Coulomb failure criterion with c'=5 kPa & '=35°


VARIATION OF SOIL STIFFNESS IN EXCAVATION 1. Soil stiffness is not constant and varies with a. stress-level. Higher stress, higher stiffness b. strain-level. Higher strain (or displacement), lower stiffness c.

stress-path (recent soil stress history).

d. Rotation of stress path, higher soil stiffness 2. During excavation, soil elements at different locations experience different changes in stress, strain & stress-path direction


PLAXIS JAKARTA 2012

SOIL STRESS PATHS NEAR EXCAVATION GCO No.1/90

• A: unloading compression; B: unloading extension • Rotation of stress paths at A & B Plaxis Seminar, Jakarta 2012

SOIL STRESS PATHS NEAR EXCAVATION 20kPa

25

20kPa Failure line

20

3m

A

A

15

B

K0 Exc.

A B

10

7m

) a P k ( t 5

K0

20kPa

Exc.

B

0

5m

-5 -10

A: unloading compression B: unloading extension

Failure line

-15 0

10

20

30 s' (kPa)

Rotation of stress path at A,  A 90° w.r.t. K0 direction Rotation of stress path at B, B 160° w.r.t. K0 direction ≈

≈


40

50

60

PLAXIS JAKARTA 2012

STRESS PATH DEPENDENT SOIL STIFFNESS Stress path rotation, 

Shear modulus, 3G’ (MPa)

t

°

=0° =180° K 0

° =90°

s'

Atkinson et al. (1990)

°

°

Triaxial tests on London Clay

Shear strain (%) -1 -0.1 -0.01 0.01 0.1 1 =0°, no change in stress path direction =180 °, full reversal of stress path direction Plaxis Seminar, Jakarta 2012

STRESS PATH DEPENDENT CDG STIFFNESS Stress-level

Test series

Extension Compress

Compression Extension

=90°

Wang & Ng (2005) • At s 0.01%, shear stiffness in extension 60% higher than in compression Plaxis Seminar, Jakarta 2012

PLAXIS JAKARTA 2012

WHY MC PREDICTS INCORRECT SURFACE HEAVE? 1.

MC models a constant soil stiffness prior to failure – not realistic

2.

In reality, stiffness of soil elements near excavation varies according to a. stress-level b. strain-level c. direction of stress-path

3.

Realistic prediction of wall deflections & ground settlements in all excavation stages requires a constitutive model that considers above factors, e.g. HS & HSsmall models

4.

HS & HSsmall consider factors (1), (2) & (3) in determining the operational soil stiffness (E), i.e. E is changing during excavation


INFLUENCE OF SMALL STRAINS AT FAR FIELD AREAS


PLAXIS JAKARTA 2012


SECTION 2.1:EXAMPLES



PLAXIS JAKARTA 2012



PLAXIS JAKARTA 2012



PLAXIS JAKARTA 2012



PLAXIS JAKARTA 2012



PLAXIS JAKARTA 2012



SECTION 2.2: VALIDATIONS


PLAXIS JAKARTA 2012

Plaxis vs. SAP2000 • Model a non-symmetrical deep deep exc. • DWall, 6 strut layers, 24m deep exc. • Compare Compare structural structural behavi behaviour our - DWall DWall deflections/bending moments/shear forces, strut forces

20m

• Recommenda Recommendation tion on design of reinforcement based on 3D results • Plaxis 3D 3D Foundation Foundation V2.2 - analyses analyses by GCG (Asia)

28m

• SAP2000 V12.0.2 V12.0.2 (BD (BD No. S0749) analyses by AECOM

25m

85 Plaxis Seminar, Jakarta 2012

Plaxis 3D Foundation

SAP2000

Element size ~1.3m Plaxis Seminar, Jakarta 2012

Element size ~1m

PLAXIS JAKARTA 2012

Plaxis 3D Foundation

SAP2000


Validation 3 – Deformed Mesh Plaxis 3D Foundation

SAP2000


PLAXIS JAKARTA 2012

Validation 3 – DWall Deflection


Validation 3 – Strut Axial Force


PLAXIS JAKARTA 2012

Validation 3 – DWall Bending Moment


MODELLING OF PILED FOUNDATIONS

SECTION 3


PLAXIS JAKARTA 2012

OUTLINE A. Volume piles B. Embedded piles 1.

Concept

2.

Model

3.

Properties

4.

Deformation behaviour

5.

Elastic region

6.

Output

C. Verification & validation 1. Axial loading, pile groups, lateral loading D. Further research


Volume piles Volume piles: Piles composed of volume elements or wall elements with pile properties • Use Cylinder command to create pile geometry Cylinder 0.6 20 24 (creates a cylinder with 0.6m radius, 20m length and 24 sections) • Alternative: Import cylinder • Pile can be inclined in PLAXIS 3D! (not in 3D Foundation)


PLAXIS JAKARTA 2012

Volume piles

Volume piles:

• Import cylinder


Volume piles Volume piles: After creating pile geometry: • Create soil material set with concrete properties for pile • Tubes: Apply plate around pile volume; create plate material set • Apply interface around pile geometry • To activate pile in calculation phase: - Assign pile properties - Tubes: activate plate - Activate interface


PLAXIS JAKARTA 2012

Volume piles Volume piles: Limitations of volume piles: • Takes many elements • Limited number of piles feasible • Installation effects not considered • Possibly bad element shapes (check mesh quality)


Embedded piles – Concept Sadek & Shahrour (2004): A three dimensional embedded beam element for reinforced geomaterials  Beam arbitrarily through volume elements  Shear interaction between beam element and surrounding soil. Septanika (2005) A finite element description of embedded pile model  Shaft interaction similar to Sadek & Shahrour (2004) - Tip interface  NEW: - Shaft interface


PLAXIS JAKARTA 2012

Embedded piles –

Model

kt

kn pile

kt kn

ks

kt

tskin

Ffoot

ks

kn

soil

t

Skin stiffness: tmax ks : axial stiffness Kn ,kt : lateral stiffness

k 1

Skin tractions: ts = qs/length = ks (uspile‐ussoil) ≤ tmax tn = qn/length = kn (unpile‐unsoil) tt = qt/length = kt (utpile‐utsoil)

urel

ks Base stiffness: kb : base/foot stiffness

s

Base/Foot force: Fb = k b (ubpile ‐ ubsoil)

t

≤ Fmax

kb n


Embedded piles – Model Embedded piles: • Beam nodes: Real nodes; 6 d.o.f.’s per node (u x u y u z r x r y r z ) • Interface nodes: Virtual nodes, 3 d.o.f.’s per node (u x u y u z ), expressed in volume element shape functions


PLAXIS JAKARTA 2012


Embedded piles – Properties

Properties (in explorer): Connection: • Rigid (only at beams / plates) • Hinged • Free


PLAXIS JAKARTA 2012

Embedded piles – Properties

Material set with embedded pile properties: • Pile type and material - Type: Massive circular pile, Circular tube, Massive square pile • Interaction properties (defines pile bearing capacity)


Embedded piles Bearing Capacity= ½ (Ttop+Tbot)×Lpile + Fmax Ttop

Lpile

Tbot

Fmax


PLAXIS JAKARTA 2012

Embedded piles – Deformation behaviour • Pile bearing capacity is input and not result of FEM calculation t

F

tmax

Specified bearing capacity

k 1

urel

Global pile response from soil modelling and pile‐soil interaction

F Fmax k 1

u

urel


Embedded piles – Without elastic region .

Load-Displacement Curves - Vertical Pile EB+CS 1250

Defined Capacity

Capacity Reached (Premature Failure)

Defined  Pile Capacity

1193.2 kN

1000

) 750 N k ( d a o L 500

VERYFINEMESH FINEMESH MEDIUMMESH

250

COARSEMESH VERYCOARSEMESH PileCapacityDefined 0 0

50

100

150

200

250

Displacement (mm)

Without elastic region: Early (soil)Jakarta failure Plaxis Seminar, 2012 for fine meshes

300

350

400

PLAXIS JAKARTA 2012

Embedded piles – Elastic Region .

• Around shaft • Around foot

Soil stress points inside elastic region are forced to remain elastic Plaxis Seminar, Jakarta 2012

Embedded piles – Output Displacements, bending moments, axial forces, shaft friction, foot force

N

u

C B A Plaxis Seminar, Jakarta 2012

Ts

PLAXIS JAKARTA 2012

Verification & validation

Verification & validation by Plaxis, METU, TUGraz, TUDelft * - Shaft friction, end bearing, total capacity - Axial loading (compression, extension) - Lateral loading (external loading, soil movement) * 1. 2. 3. 4.

5. 6.

Related reportsandpublications: Engin H.K.(2006).Vali dation ofembeddedpiles,Plaxis InternalReport. Engin H.K., Septanika E.G.and Brinkgreve R.B.J.(2007). Improvedembeddedbeamelements for the modelling ofpiles.In: G.N.Pande & S. Pietruszczak (eds.), Int.Symp. on NumericalModels in Geomechanics – NUMOG X, 475-480. London:Taylor& Francis group. EnginH.K.(2007).A Reportontension piles testing using embeddedpiles,Plaxis InternalReport. Engin H.K., SeptanikaE.G.,Brinkgreve R.B.J., Bonnier P.G. (2008). Modelin g pile d foundation by means of embedded pile s.2nd Internatio nal Workshop on Geotechnic s of Soft Soils - Focus on Ground Improvement.3-5 September2008, Univ ersity of Strathclyde,Glasgow,Scotland. (Acceptedfor publication) S ep ta ni ka E .G ., B ri nk gr ev e R .B .J ., E ng in H .K . ( 20 08 ). E st im at io n o f p il e g ro up b eh av io r u si ng e mb ed de d p il es , t he 1 2t h I nt er na ti on al Conferenceof InternationalAssociationfor Computer Methods and Advances in Geomechanics (IACMAG),1-6 October, 2008,Goa, India. T sc hu ch ni gg F . ( 2 00 9) . E mb ed de d p il es – 1. Report. CGG_IR021_2009.TechnischeUniversitätGraz.

7.

Tschuchnigg F. (2009). Embedded piles – 2. Report. Improvements. Technische Universität Graz.

8.

DaoT.P.T.(2011). Vali dation ofPLAXIS embeddedpilesfor lateral loading.MSc thesisGeo-engineerin g.Delft University ofTechnology.


Verification & validation – Axial loading (Plaxis)

Single Layer :  = 0 , Cohesive Soil (Case 1): c = 50 kPa  = 0


PLAXIS JAKARTA 2012

Verification & validation – Axial loading (Plaxis)

Single Layer :  = 0 , Cohesive Soil (Case 1): c = 50 kPa  = 0


Verification & validation – Axial loading (METU) Pile load test Alzey Bridge near Frankfurt (Bored Pile)

Hardening Soil model Pre Overburden Pressure = 50 kPa

El-Mossallamy, Y (1999)


PLAXIS JAKARTA 2012

Verification & validation – Axial loading (METU) Alzey Brig de Sing le Pile Lo ad Test 3500

PILE CAPACITY

3000

2500

) N k ( d a o L

2000

1500

TotalLoad

SkinFriction

Base Resistance

1000

PILE CAPACITY

HS-CS

500 HS-CS-BaseRes.

HS-CS-Ave. SkinFriction

0 0

5

10

15

20

25

30

35

40

45

50

Plaxis Seminar, Jakarta Settlement (mm) 2012

Verification & validation – Pile groups (TUDelft)

Pile group example by Poulos:


PLAXIS JAKARTA 2012

Verification & validation – Pile groups (TUDelft)

(a) Poulos & Davis (1980) (b) Randolph (1994) (c) Strip on springs analysis, using the program GASP (Poulos,1991) (d) Plate on springs approach, using the program GARP(Poulos, 1994a) (e) Finite element and boundary element method of Ta & Small(1996) Seminar, Jakarta 2012 (f) Finite el ement and boundary elementPlaxis method of Sinha(1996).

Verification & validation – Pile groups (TUDelft) Moment (MNm/m)

Average Settlement (mm) 1,2

50,0 45,0

1,0

40,0 35,0

0,8

30,0 0,6

25,0 20,0 15,0 10,0 5,0 0,0

d n F D 3 s i x a l P

s i v a D & s o l u o P

h p l o d n a R

) P S A G ( p i r t S

) P S A G ( e t a l P

l l a m S & a T E F

a h n i S E B + E F

0,4 0,2 0,0


DifferentialSettlement(mm)



l l a m S & a T E F

a h n i S E B + E F

% Load on Piles

10,0

100,0

9,0

90,0

8,0

80,0

7,0

70,0

6,0

60,0

5,0 4,0 3,0 2,0 1,0 0,0



) P S

A G ( e t a l P

l l a m S & a T E F

50,0 a h n i S E B + E F

40,0 30,0 20,0 10,0


0,0


h p l o d n a R



l l a m S & a T E F

a h n i S E B + E F

PLAXIS JAKARTA 2012

Verification & validation – Axial loading (TUGraz)


Verification & validation 3D model - volume piles: 70 mm

2D model: 72 mm

3D model - embedded piles: 74 mm Plaxis Seminar, Jakarta 2012

PLAXIS JAKARTA 2012

Verification & validation – Axial loading (TUGraz) Conclusions from research at TUGraz (based on 3D Foundation): • Embedded pile gives good results in serviceability states • Layer-dependent option preferred to obtain realistic shaft friction • Increased interface stiffness needed at pile tip * • Pile should end at corner node *

* Implemented in PLAXIS 3D


Verification & validation – Lateral loading (TUDelft) Validation for lateral loading: • Comparison with volume pile • Lateral movement of pile in horizontal soil slice • Lateral loading of pile top • Lateral loading by soil movement (embankment construction) • Comparison with measurements from centrifuge test • Lateral loading by soil movement (embankment construction)


PLAXIS JAKARTA 2012

Verification & validation – Lateral loading (TUDelft) Lateral movement of pile in horizontal soil slice: > Embedded pile almost behaves as volume pile due to elastic region


Verification & validation – Lateral loading (TUDelft) Lateral loading by soil movement due to embankment construction > Bending moments in reasonable agreement with measurements


PLAXIS JAKARTA 2012

Verification & validation – Lateral loading (TUDelft) 1. Conclusions from research at TUDelft: 2. Embedded piles have capabilities for lateral loading behaviour in case of rough pile-soil contact (full bonding) and small soil displacements 3. When using ‘standard’ mesh around embedded piles (no local refinement), stiffness and lateral capacity are over-estimated (~30%)


Further research 1. Research at TUDelft on pile installation effects: 2. Press-replace technique to simulate pile installation with the purpose to generate data for different situations 3. Results are used in generalized model, where (embedded) piles are ‘wished-in-place’ and installation effects are ‘superimposed’


PLAXIS JAKARTA 2012

CONCLUSIONS A. Volume pile 1. Pile composed of volume elements or wall elements with pile prop’s 2. Massive piles or tubes (wall elements) 3. Not feasible for many piles B. Embedded piles 1. Efficient way to model different types of piles 2. Validated for axial loading, pile groups and lateral loading C. Limitations of embedded piles: 1. Primarily for bored piles (no installation effects) 2. Primarily for serviceability states 3. Mesh-dependency of results 4. Full bonding considered in lateral movement


TUNNELS AND TUNNELLING

SECTION 4.0


PLAXIS JAKARTA 2012

CONTENTS

A. 4.1 Introduction to Plaxis Approach a. Input and construction of FE model b. Conclusions B. 4.2 Some Validations C. 4.3 Case Histories


Modelling of Tunnelling in Plaxis 3D

•

To – – – –

be able to: Model tunnel geometries in different ways Model construction stages for tunnels Model volume loss due to tunnel construction Analyse deformations, stability, lining forces

PLAXIS JAKARTA 2012

Geometric modelling issues Circular tunnel shapes (TBM tunnels) • Create cylinder using Cylinder command or using Import facility cylinder 4 100 48

• Decompose cylinder volume into surfaces • Apply plate and negative interface features to cylinder contour

Geometric modelling issues Circular tunnel shapes (TBM tunnels) – Example


PLAXIS JAKARTA 2012

Geometric modelling issues Cross passages and entrance shafts – Example Hint: Draw cross section surface and use Extrude command to create shafts PLAXIS 3D will automatically create intersections


Geometric modelling issues Non-circular tunnel shapes • • • • •

Using shape designer* to create tunnel contour Create surface from tunnel contour using right-hand mouse button Extrude surface Decompose tunnel volume into surfaces Assign Plate and Negative interface features to tunnel surface

* new in 3D 2011


PLAXIS JAKARTA 2012


Geometric modelling issues

Non-circular tunnel shapes – Example


PLAXIS JAKARTA 2012

Geometric modelling issues

Importing tunnel geometry using CAD model • DXF triangulated surface model - Model should be ‘cleaned’ before importing in PLAXIS 3D • 3DS model • Use Import command or corresponding tool in Structures mode


Construction stages

Creating geometry for construction stages • • • •

Divide tunnel in excavation sections (top heading, bench, invert) Divide tunnel in longitudinal steps by defining cross section planes Intersect tunnel with excavation sections and cross section pl anes Remove unnecessary sub-surfaces around tunnel


PLAXIS JAKARTA 2012

Construction stages

Creating geometry for construction stages – Example (exploded view)


Modelling volume loss

Volume loss can be modelled by: • Defining Contraction * (TBM tunnels) in Structures mode, e.g: Contraction Fase_Volume_1_1

or use contraction tool or right-hand mouse menu • Activate contraction in Staged construction mode

* New in 3D 2011 Plaxis Seminar, Jakarta 2012

PLAXIS JAKARTA 2012

Contraction


Modelling volume loss

Alternatively, volume loss can be modelled by: • Applying Volumetric strain to volume (Staged construction mode) - Distinction can be made between xx, yy, zz


PLAXIS JAKARTA 2012


SECTION 4.1: VALIDATIONS


Validation 1 – Plaxis 3D Tunnel vs. Plaxis 2D Plaxis 2D

• Model a plane strain tunnelling • Layered ground Fill, Alluvium, CDG • GWL 2 mbgl • 6m dia. tunnel, tunnel axis 23 mbgl • Stress relief by 30% due to tunnel exc. • Linings take 70% initial soil stress

Plaxis 3D Tunnel

• Plaxis 2D V8.2 (BD No. G0133) - 456 nos 6-noded triangular elements • Plaxis 3D Tunnel V2.4 - 4,560 nos 15noded wedge elements • Fineness of 2D & 3D meshes identical in-plane


PLAXIS JAKARTA 2012

Validation 1 – Input Parameters


Validation 1 – Ground Surface Settlement


PLAXIS JAKARTA 2012

Validation 1 – Lining Hoop Force & Bending Moment Plaxis 2D

Plaxis 3D Tunnel

Hoop force

Bending moment


Validation 2 – Plaxis 3D Tunnel vs. Centrifuge Test in Sand Centrifuge model

• Stability of shallow tunnel in sand • Minimum tunnel support pressure ( T) before tunnel collapse • Centrifuge tests by Atkinson & Potts (1977) in Leighton Buzzard Sand • Acceleration 75g, 60mm dia. model tunnel is 4.5m dia. prototype tunnel • Centrifuge tests at C/2R ratios of 0.34, 0.63, 1.0, 1.37 & 2.0 • Plaxis 3D Tunnel replicates centrifuge tests in prototype scale • Predicted T compared to measured T

Atkinson, J. H. & Potts, D. M. (1977). Stability of a shallow circular tunnel in cohesionless soil. Geotechnique, 27(2), 203-215. 146 Plaxis Seminar, Jakarta 2012

PLAXIS JAKARTA 2012

Validation 2 – Input Parameters quoted by Atkinson & Potts (1977)


Validation 2 – Collapse Mechanism


PLAXIS JAKARTA 2012

Validation 2 – Comparison

Atkinson & Potts (1977)


Validation 3 – Plaxis 3D Tunnel vs. Centrifuge Test in Clay • Stability of tunnel heading in clay • Minimum tunnel support pressure ( T) in unlined section P before collapse • Centrifuge tests by Kimura & Mair (1981) in soft kaolin clay • Acceleration 125g, 60mm dia. model tunnel is 7.5m dia. prototype tunnel Centrifuge model

• Centrifuge tests at C/D of 1.5 to 3.0, P/D of 0 to 3 • Plaxis 3D Tunnel replicates centrifuge tests in prototype scale with C/D = 3, P/D = 0, 0.5, 1, 2 & 3 • Predicted T compared to measured T

Kimura, T. & Mair, R. J. (1981). Centrifugal testing of model tunnels in soft clay. Proc. 10 th Int. Conf. Soil Mech. & Found. Eng., Stockholm, Vol. 1,2012 319-322. Plaxis Seminar, Jakarta

PLAXIS JAKARTA 2012

Validation 3 – Plaxis 3D Tunnel Model & Stability Ratio N

Stability Ratio, N

Prototype scale




PLAXIS JAKARTA 2012

Validation 3 – Collapse Mechanism

P/D=0

P/D=2




PLAXIS JAKARTA 2012

Validation 4 – Plaxis 3D Tunnel vs. SAP2000 Plaxis 3D Tunnel

• Model 6m dia. circular lining subjected to 100 kPa external radial pressure • Lining 0.25 m thick, E=20 GPa, =0.2 • Compare lining radial displacement, hoop force, axial force & bending moment • Plaxis 3D Tunnel V2.4 uses “Plate” element • SAP2000 Nonlinear V7.40 (BD No. S0476) uses “Shell” element

SAP2000

• Both predictions compare to known theoretical solutions


Validation 4 – Theoretical Solution Cylinder Under External Radial Pressure

Watkins, R. K. & Anderson, L. R. (2000). Structural mechanics of buried pipes. CRC Press. Young, W. C. & Budynas, R. G. (2002). Roark’s formulas for stress and strain. McGraw-Hill, 7th edition. Plaxis Seminar, Jakarta 2012

156

PLAXIS JAKARTA 2012



Validation 5 – Plaxis 3D Tunnel vs. Closed Form Solution + Boundary Element Method

• Model a 2x2 pile group near a 6m tunnel in clay 5.7m

• Hypothetical example by Loganathan et al. (2001) analysed using closed form solution + boundary element method GEPAN • Not an exact solution, not measurement • Volume loss ratio modelled 1%

20m

• Plaxis 3D Tunnel analysed the example 1.1m

25m

• Compare pile settlement, horizontal displacement, axial force, bending moment

6m dia. Front pile

Rear pile

0.8m

Loganathan, N., Poulos, H. G. & Xu, K. J. (2001). Ground and pile-group responses due to tunnelling. Soils and Foundations, JGS, 41(1), 57-67. 158 Plaxis Seminar, Jakarta 2012

PLAXIS JAKARTA 2012

Validation 5 – Plaxis 3D Tunnel Model Deformed mesh

Exaggeration scale 150


Validation 5 – Comparison Pile Settlement

Horizontal disp.

Axial force

Bending moment


PLAXIS JAKARTA 2012

Validation 6 – Tunnelling below Hua Tai Building, Sheung Wan

Sheung Wan Crossover Box

30-32 New Market Street • Hua Tai Bldg. built in 1964, 10-storey R.C. frame structure, founded on 73 nos. of 0.457m dia. Franki piles • 5.8m dia. overrun tunnel built in 1980s, trimmed 17 nos pile toes, Fill grouted, 161 increase size of central raft Plaxis Seminar, Jakarta 2012

Tunnelling Beneath/Near Building Piles • Advantage of 3D over 2D analysis 1. progressive advance of tunnel face

Proposed U/T tunnel

2. assess stability of tunnel face/heading Existing overrun tunnel

3. model individual piles

(Proposed D/T tunnel)

4. model plan area of buildings 5. model varying support pressure on tunnel face & along/around TBM 6. soils vary in tunnel axis direction 0 -10 -20 -30 -40


PLAXIS JAKARTA 2012

Removal of Existing Tunnel Linings

• 3D analysis required because 1. soil arching in x, y, z (tunnel axis) directions

2. stability of localised unlined section 3. unlined & lined sections exist 4. shotcrete properties change with time in z direction Plaxis Seminar, Jakarta 2012 5. soils

vary in z direction

Validation 6 – Tunnelling below Hua Tai Building, Sheung Wan

WIL

Overrun

• Open-face shield tunnel 2.6 bar air pressure Overrun

• Bldg settled 6-9 mm, ground settled 4-6 mm

WIL

GCO (1985). Technical Note TN 4/85 – MTR Island Line: Effects of Construction on Adjacent Property. GEO, Eng. Development Dept., HK.

• Valuable case history for benchmarking


PLAXIS JAKARTA 2012

Validation 6 – Plaxis 3D Tunnel Model

40,388 15-noded wedge elements


Validation 6 – Modelling of Tunnelling

• Progressive advance of tunnel face • Varying support pressure on tunnel face & along/around shield (average ~2.6 bar) • Building load, piles & cap modelled • Bldg. stiffness considered – Parallel Axis Theorem or sum of EI for individual storeys • Circular piles idealised as square piles 166 Plaxis Seminar, Jakarta 2012

PLAXIS JAKARTA 2012


SGI


Validation 6 – Ground & Pile Displacement Front

Rear


PLAXIS JAKARTA 2012

Validation 6 – Comparison of Settlement Building

Ground surface


Validation 7 – Interface Behaviour

100kN


PLAXIS JAKARTA 2012

Validation 7 – Interface Behaviour

100kPa


Validation 7 – Straight Interface Input Shear Strength


PLAXIS JAKARTA 2012

Validation 7 – Curved Interface Input Shear Strength


Validation 7 – Comparison 50kPa x 1m2 = 50 kN

Straight interface Input shear strength 50 kN

Curved interface 160kPa x

2.9688m2

= 475 kN

Input shear strength 474 kN


PLAXIS JAKARTA 2012


SECTION 4.2: APPLICATION 1


Scenario 1: Impact of tunnelling on existing piles

PLAXIS JAKARTA 2012

Scenario 2: Impact of piling loading on existing tunnels

(1) A proposed development was located adjacent to the future development MRT twin tunnels; (2) The piling system within MRT Protection Zone adopts bored piles so as to minimize the dynamic impact during construction. RC piles outside MRT Protection Zone Bored piles within MRT Protection Zone

MRT Protection Zone

Future MRT twin tunnels


PLAXIS JAKARTA 2012

HOW to simulate the problem using Plaxis 3D?

Most critical section adopted for the present 3D FEM analysis


Typical cross section Road surface

Bored pile dia. 1000mm with 40m length with 28m into underlying OA soils


PLAXIS JAKARTA 2012

Boreholes at this local area are adopted for the interpretation of subsurface soil profile


GIBR soil parameters are adopted for the analysis. Effective drained parameters are adopted due to the long‐term nature of the project

Bored pile dia. 1000mm with 40m length to rest on the underlying hard OA with SPT N>100


PLAXIS JAKARTA 2012

Illustration of effective drained soil parameters following GIBR adopted in 3D FEM analysis


Illustration of effective drained soil parameters following GIBR adopted in 3D FEM analysis


PLAXIS JAKARTA 2012

Constructing the 3D FEM mesh…

3D FEM mesh with subsurface soil profiles, pile groups, tunnels

Pile groups

25kPa surcharge

Working load on pile cap

Top fill

OA (E) OA (D)

F1 F2

Upper and closer tunnel

OA (C) OA (B)

Underlying hard OA (N>100)


Lower and farther tunnel

PLAXIS JAKARTA 2012

Hiding of some soil elements to reveal the tunnels and piles Pile groups 25kPa surcharge Working load on pile cap

Bored piles dia. 1m with 40m length

tunnels


Scenario 1: Pile groups assumed to be constructed first; Effect of 2 tunnelling (with 2% volume loss each) on the adjacent pile groups

PLAXIS JAKARTA 2012

Simulation sequence: 25kPa surcharge

Pile groups with loadings applied first

Tunnels NOT constructed yet Plaxis Seminar, Jakarta 2012

Lower tunnel activated with 2% volume loss

Lower tunnel activated with 2% volume loss Plaxis Seminar, Jakarta 2012

PLAXIS JAKARTA 2012

The invert of of the the tunnel was restrained from heaving up, so as to induce maximum tunnel shrinking inward with maximum impact to surrounding ground

Cross‐section of model of model tunnel

3D view


A surprise: tunnel has an overall shrinking in, the restraint at the invert has NOT effect…


PLAXIS JAKARTA 2012

A surprise: tunnel has an overall shrinking in, the restraint at the invert has NOT effect…

“hexagon” “hexagon” tunnel tunnel composed of of 30 30 composed of of 24 24 sides, each 12 sides, each 15 Plaxis Seminar, Jakarta 2012

A relief to remove the unpleasant surprise

Correct restraint of invert of of tunnel tunnel


PLAXIS JAKARTA 2012

Followed by the upper tunnel activated with 2% volume loss

Followed by the upper tunnel activated with 2% volume loss


Final tunnel volume loss shapes (scaled up by 25 times)


PLAXIS JAKARTA 2012

Final ground movement contour plot Max ground movement around tunnel crown, and dissipates away from the tunnels Immediately above the tunnel, the induced ground surface settlement is about 25mm; while the ground movement at the adjacent site is about 10mm 10mm 25mm


The induced max pile deflection is only about 6mm due to the 2 tunnelling with 2% volume loss each


PLAXIS JAKARTA 2012

The induced max pile settlement is less than 5mm


Max pile axial force of 5386kN before tunnelling; and 5766kN after 2 tunnelling, an increment of 380kN, or about 7% increment only.

Axial force BEFORE 2 tunnelling


Axial force AFTER 2 tunnelling

PLAXIS JAKARTA 2012

Max pile BM towards tunnels (M2‐2) of 90kNm before tunnelling; and 104kNm after 2 tunnelling, an increment of 14kNm which is negligible for a bored pile of 1m diameter.

Bending moment towards tunnels Plaxis Seminar, Jakarta 2012 M2 ‐2 BEFORE two tunnelling

Bending moment towards tunnels M2‐2 AFTER two tunnelling

Max pile BM parallel to tunnels (M3‐3) of 60kNm before tunnelling; and 63kNm after 2 tunnelling, indicating negligible increment of BM parallel to the two tunnelling.

Bending moment towards tunnels M3‐3 BEFORE two tunnelling Plaxis Seminar, Jakarta 2012

Bending moment towards tunnels M3‐3 AFTER two tunnelling

PLAXIS JAKARTA 2012

Final pile max loading condition: Final Max working axial force = 5766kN; FOS=1.4; Factored axial force =5766*1.4 = 8072kN Max working BM: M2‐2 = 104kNm; M3‐3 = 63kNm; So Composite BM = 122kNm; FOS=1.4; Factored BM = 170*1.4 = 170kNm

The final loading state is located well within the M‐N plot envelope



SECTION 4.2: APPLICATION 2


Stop

PLAXIS JAKARTA 2012

Zones of Influence

Zone C

Zone A

Zone B

Zone B

Zone C Pile settlement C B A

Selementas et al. (2005)

45º

45º

h t p e D

For pile toe located in Zone A: pile head settlement > soil surface settlement; decrease in pile axial force Zone B: pile head settlement soil surface settlement Zone C: pile head settlement < soil surface settlement; increase in pile axial force ≈


ANALYSIS OF TUNNEL‐PILE INTERACTION A. Typically use the combination of 1. empirical relationships/closed‐form solutions to estimate greenfield ground movements; and 2. boundary element methods to compute pile deformations and stresses A. Suitable for preliminary assessment, with some limitations

B. Alternatively, use 3D numerical analysis Pros: model tunnelling, tunnel‐pile ‐building interaction &

geotechnical

entities in one single analysis Cons: complicated, relatively long analysis time & require constitutive model for soil non‐linear behaviour


advanced

PLAXIS JAKARTA 2012

EXAMPLE OF TUNNELLING BELOW PILED BUILDING 25m

0 mbgl 2m

25m P5 Rear P6

P4

Pile cap Fill

5 mbgl

9m 4m

MD

10 mbgl

CDG

P1 10m

1m

10m P2 Front P3 6m Ø tunnel

1m

4m

Tunnel advance direction

20 mbgl 2m Ø pile

Tunnel 6m Ø 30 mbgl 31.5 mbgl Rock P1/P4

Pile design load 15MN (~5MPa) 3m Ø bell-out P2/P5

P3/P6


INFORMATION FOR TUNNEL, PILES & GROUND

A. 6 m diameter tunnel excavated by TBM, tunnel axis depth at 20 mbgl in Completely Decomposed Granite B. 15‐storey building supported by 6 nos of 2 m diameter bored piles with 3 m diameter bell‐outs in rock at 32 mbgl C. Each pile takes 15 MN design load (~5 MPa). D. Building plan size is 25 m by 9 m, pile cap 2 m thick E. Stratigraphy is 5 m Fill, 5 m Marine Deposits, 20 m CDG and rock. Groundwater table at 2 mbgl F. Tunnel constructed in between piles, tunnel edge to pile edge distances are 1 m, 4 m and 10 m


PLAXIS JAKARTA 2012

SOIL SMALL STRAIN NON‐LINEAR STIFFNESS

0.01%

0.1%

1%

Atkinson & Sallfors (1991)


CDG Small Strain Non‐linear Stiffness Laboratory small strain stiffness results for CDG samples Ng et al. (1998)

• Hardening Soil + Small Strain Overlay (HSsmall) constitutive model to consider CDG small strain non-linear stiffness

1600 Triaxial_Upper

1400

Adopted line

1200

Triaxial_Lower HSsmall_Upper

' p / c

e s

G

1000

HSsmall_Lower

800

HSsmall_Baseline

600 400 200 0 0.0001

0.001

0.01 0.1 Shear strain (%)


1

10

PLAXIS JAKARTA 2012

3D Finite Element Model (Plaxis‐GiD) Load 15 MN

Rear

“Plate” Pile cap

Building 40m Bored pile Front Fill MD CDG

Tunnel

Tunnel face 149m TBM length

Rock

120m

Bell-out 43,000 elements

Linings

Refined mesh around tunnel & building piles Plaxis Seminar, Jakarta 2012

TUNNEL CONFINEMENT (FACE SUPPORT) PRESSURE A PIII

PIV

PI

Rear 6m Ø

Front

TBM shield 9m PII

PVI PV A PIII

PV Section A-A Plaxis Seminar, Jakarta 2012

Pressure increases with depth

• Confinement (face support) pressure (PI to PII) = hydrostatic pore pressure + overpressure • Higher confinement pressure, lower ground loss • Along TBM shield, tunnel support pressures vary to consider 1. conical shape of TBM shield / over-cutting 2. ground loss into tail void in rear • Any combination of support pressure profiles can be modelled

PLAXIS JAKARTA 2012

MODELLING OF TUNNEL FACE ADVANCE

g g n TBM shield n i i n n i i (elements nulled) L L

1.5 1.5m

g g n i n TBM shield i n i n i (elements nulled) L L

1.5 1.5m g g n i n i n i n i

TBM shield (elements nulled) L L

• Soil elements inside TBM shield are deactivated • Apply tunnel support pressure profiles • Shield is not modelled • For each face advance, shift tunnel support pressures forward & correspondingly erect new lining behind TBM • The process is repeated as tunnelling progresses

1.5 1.5m


MODELLING OF STRUCTURES • Piles & pile cap modelled by solid elements

• Interface elements along pile shafts & on pile cap vertical faces • Consider flexural stiffness (EI) & axial stiffness (EA) of superstructure by incorporating a “Plate” structural elements on top of pile cap • Superstructure EI estimated by (Potts & Addenbrooke, 1997) 1. Parallel Axis Theorem (bending about building neutral axis); or 2. Summation of EI for individual building storeys

• Tunnel linings modelled by “Plate” elements


PLAXIS JAKARTA 2012

PREDICTION ON GROUND SURFACE SETTLEMENT Overpressure 20 kPa

Overpressure 20 kPa, G/F VL 1.6%

Distance from tunnel centreline (m) -60

Fill

0

MD

-40

-20

0

VL 0.31%

20

40

60

-4 ) m -8 m ( t n-12 e m e l t t -16 e S

CDG

Tunnel

-20

VL 1.61% Mid-building Greenfield Gaussian

-24

• Gaussian curve with K = 0.45 • Close to K 0.5 from HK tunnelling experience

• Lateral spreading of displacements in MD layer • Settlement trough becomes wider

≈


Prediction on Pile Transverse Displacement Overpressure 20 kPa

Transverse horizontal disp. (mm) -5

-4

-3

-2

-1

+10D 0 0

-2D Front Rear +2D +10D

5

+2D

10 ) l g 15 b m ( h t 20 p e D

25 30 35


Rear 1m

P2 Front -2D

Tunnel advance

PLAXIS JAKARTA 2012

Prediction on Pile Longitudinal Displacement Overpressure 20 kPa

+10D

Longitudinal horizontal disp. (mm) -3

-2

-1

0

1

2 0 5 +2D

10

) l g b 15 m ( h t p 20 e D

-2D

Rear 1m

Tunnel advance

Front

P2 Front

25

Rear -2D

30

+2D +10D

35 Tunnel advance


Prediction on Pile Settlement & Axial Force Overpressure 20 kPa 0 0

Settlement (mm) -1

0

-2 0

P2

-2D Front

5

5

Rear 10

+2D

) l g b15 m ( h t p20 e D

+10D 7

25 30 35

B C

P2

-2D Front Rear

10 ) l g b15 m ( h t p e20 D

25

A

Increase in axial force (MN) 1 2 3

30 35

Pile toe Plaxis Seminar, Jakarta 2012

+2D +10D

4

PLAXIS JAKARTA 2012

Prediction on Pile Bending Moment Overpressure 20 kPa Transverse moment (kNm) 1500 0

500

-500

Longitudinal moment (kNm) -1500

1500

500

-500

-1500

0

P2

P2

5

5

10

-2D

) l g b 15 m ( h t p 20 e D

Front Rear +2D +10D

10

-2D

) l g b15 m ( h t p20 e D Tunnel advance

25

25

30

30

35

35

Front Rear +2D +10D


Check on Potential Structure Damage OP 10kPa

45

P2 35

)_ N M ( 25 N , e 15 c r o F l 5 a i x A

Distance from tunnel centre (m) -10 - 5 0 5 10 15

OP 20kPa OP 30kPa OP 40kPa

-5

-15 -15 -10 -5 0 5 10 Moment, M (MNm)

0.3

0.0 )_ m -0.4 m ( t n e m e-0.8 l t t e s . g d l-1.2 B

OP 20kPa OP 30kPa OP 40kPa

)0.2 % ( L / 

Cat. 3

0.1

=0.14 mm

2 0.0

15

Pile N-M Interaction Diagram

OP 10kPa

Cat. 4 & 5

0 0

-1.6

Building deflection

1 0.1

 h (%)

0.2

0.3

Burland’s chart

PLAXIS JAKARTA 2012

Comparison with Closed Form Solution Greenfield subsurface settle. (mm)

-35

-25

-15

-5

0

5

Greenfield subsurface horiz. disp. (mm)

15

-15

-10

-5

0

0

Fill

0

Fill 5

MD

5

MD 10

CDG

) l g b m 15 ( h t p e 20 D

10 ) l g

15 b

CDG

m ( h t p 20 e D

25 Loganathan et al. (2001) 3D analysis

30

Rock

25

Loganathan et al. (2001) 3D analysis

Rock

35

30 35

Greenfield subsurface section corresponds to P2 location 3D analysis: Overpressure 20 kPa, G/F V L 1.61% Plaxis Seminar, Jakarta 2012

3D FEA vs. Analytical Solution Issues

3D FEA

Analytical Solution

Ease of use

• Complicated • Long analysis time

• Relatively easy • Less analysis time

Ground conditions

• Layered soil

• Homogeneous soil

• Need realistic constitutive model

• Estimated greenfield deformation less good for layered soil

Tunnelling progress

• Model face advance

• Only pile response in transverse direction

• Pile response in transverse & longitudinal directions


PLAXIS JAKARTA 2012

3D FEA vs. Analytical Solution Issues Ground loss, VL

3D FEA

Analytical Solution

• Model confinement pressure & predict V L

• Assume a certain VL

Effect on • Model tunnel, piles, piles/building building & their interaction in one single analysis

• Different boundary element programs for pile axial and lateral responses

• Results from piles & building used directly in structural check

• Specific analysis for pile group effect • Dedicated modification factors account for building rigidity


REFERENCES (1) 1.

Atkinson, J. H. & Sallfors G. (1991). Experimental determination of soil properties. Proc. 10

th

ECSMFE, Florence, Vol.3, 915-956 2.

Burland, J. B. (1995). Assessment of risk of damage to buildings due to tunnelling and excavation. 1

st

Int. Conf. on Earthquake Geotech. Engrg., IS Tokyo. 3.

Geotechnical Control Office (GCO) (1985). Technical Note T4/85 - MTR Island Line: Effects of Construction on Adjacent Property. Civil Engrg. Services Dept., Hong Kong.

4.

Hake, D. R. & Chau, I. P. W. (2008). Twin stacked tunnels - KDB200, Kowloon Southern Link, Hong Kong. Proc. 13 rd Australian Tunnelling Conference, 445-452.

5.

Loganathan, N., Poulos, H. G. & Xu, K. J. (2001). Ground and pile-group responses due to tunnelling. Soils and Foundations, 41(1), 57-67.

6.

Moller, S. (2006). Tunnel induced settlements and structural forces in linings. PhD thesis, University of Stuttgart.

7.

Moller, S. & Vermeer, P. A. (2008). On numerical simulation of tunnel installation. Tunnelling & Underground Space Technology, 23, 461-475.

8.

Ng, C. W. W., Sun, Y. F. & Lee, K. M. (1998). Laboratory measurements of small strain stiffness of granitic saprolites. Geotechnical Engineering, SEAGS, 29(2), 233-248.


Seminar-Advanced Geotechnical Finite Element Modeling in Analysis using PLAXIS.pdf

Recommend Documents