01 Schweiger PDF

FINITE ELEMENT ANALYSIS OF DEEP FOUNDATIONS AND TUNNELS PRACTICAL APPLICATIONS
Helmut F. Schweiger Acknowledgement: Franz Tschuchnigg, Thomas Marcher (ILF Innsbruck)

Computational Geotechnics Group Institute for Soil Mechanics and Foundation Engineering Graz University of Technology
Alert Workshop 2010 - Session Engineering Geostructures, Aussois, 4 October, 2010
Finite element analysis of deep foundations and tunnels

DEEP FOUNDATION (Bucharest) DEEP FOUNDATION (Vienna) STATION TUNNEL (Bratislava) JET GROUT FOUNDATION (Vienna)
CONTENT
DEEP FOUNDATION OF TOWER FOR SHOPPING MALL (Bucharest) DEEP FOUNDATION OF TWO ADJACENT HIGHRISE BUILDINGS (Vienna) ANALYSIS OF STATION TUNNEL (Bratislava) JET GROUT FOUNDATION FOR RAILWAY STATION (Vienna)

CONTENT
DEEP FOUNDATION OF TOWER FOR SHOPPING MALL (Bucharest) Project overview Soil conditions Finite element models Optimisation of the foundation concept Results Validation of numerical model Conclusion

SHOPPING MALL WITH SKY TOWER
Sky Tower

PROJECT OVERVIEW - SKY TOWER
Facts: - Dimensions 93.4 x 61.7 m - Height more than 130 m - Excavation depth 20.4 m - Slab thickness 2.6 / 1.5 m - Point loads up to 14 900 kN
Results of interest: 1. Maximum differential settlements (of 2.6 m thick slab) 2. Arrangement of diaphragm wall elements
Alert Workshop 2010 - Session Engineering Geostructures, Aussois, 4 October, 2010 5

SOIL CONDITIONS
Silty Clay
~20.5 m 60.0 m ~25.0 m
Soil profile: - Core drillings down to -60 m - Altering layers of sands and silty clays - Drained conditions are assumed
Assumption: Alternate layers continue
Sand
Constitutive models: Diaphragm wall panels => Mohr-Coulomb Floors => Linear elastic Soil layers => Hardening Soil Model

CONSTITUTIVE MODEL
Hardening Soil Model 1. 2. 3. 4. Hyperbolic relation between strains and deviatoric stress for drained triaxial stress paths Distinction between deviatoric primary loading and unloading - reloading Mohr-Coulomb failure criterion Stress dependent stiffness
1 =
1 q 2 E 50 1 q / q a
2 sin 1 sin

m
' qf = (c cot 3 )
E 50 = Eref 50
c cot + 3 c cot + pref

c cot + 3 c cot + pref
Eur = E
ref ur

FINITE ELEMENT MODELS
Core walls
Floors (Top-down method)
=> All models consist of about 52 000 15-noded wedge elements



Layout 1 Top view Deep foundation elements in 3D FE model
Cross section
~ 62 m
2D "Plane strain" model
~ 94 m
=> Only silty clays are modelled below -60 m

=> 2D (plane strain) model overestimates differential settlements by approx. 100% => 2D representation of geometry / load situations not possible
uy,max~ 165 mm uy,diff.max ~ 100 mm
=> 2D representation of deep foundation elements not possible
=> For optimisation of foundation concept full 3D modelling is required
uy,max~ 95 mm uy,diff.max ~ 50 mm

OPTIMISATION OF FOUNDATION CONCEPT
Discontinuous barrettes
Different lengths of barrettes
Circle of panels beneath concentrated loads
=> Maximum settlements between 85 - 100 mm => Differential settlements between 42 - 60 mm => Full connection of barrettes difficult to achieve in practice

OPTIMISATION OF FOUNDATION CONCEPT Final foundation concept

Top view FE discretization of panels
25 m
15 m
12

OPTIMISATION OF FOUNDATION CONCEPT Final foundation concept

Contour lines of settlements Vertical displacements of slab
=> Maximum vertical displacements of about 105 mm => Maximum differential settlements of about 65 mm => Between point A and C (2.6 m thick slab) differential settlements of about 47 mm

VALIDATION OF NUMERICAL MODEL (IN SITU LOAD TEST)

FE model Measurements vs calculation
=> FE analysis predicts slightly higher settlements > conservative estimate of parameters => Effect of small strain stiffness is not taken into account

CONCLUSION FOR THIS PROJECT
Due to the geometrical layout 2D analysis proved to be too conservative. A number of 3D calculations have been performed investigating different arrangements of diaphragm wall panels. The final concept of the deep foundation consists of two discontinuous circles beneath the highly loaded areas. The computed maximum settlements are 105 mm. The expected maximum differential displacements are about 65 mm. The numerical back analysis of an in situ load test shows reasonable agreement The soil parameters have been estimated prior the test and no adjustments have been made. The analyses provide a good, albeit somewhat conservative estimate of settlements for the Sky Tower.
15

CONTENT
DEEP FOUNDATION OF TWO ADJACENT HIGHRISE BUILDINGS (Vienna) Project overview Soil conditions and its numerical modelling Finite element models Optimisation of the foundation concept Results Conclusion
16

TWIN TOWERS
~220m ~168m
Concept for foundation: deep foundation with diaphragm wall panels


PROJECT OVERVIEW - TWIN TOWERS

Underground car parks
56.3 m
TOWER II
24 m
TOWER I
58.9 m
64.1 m
Facts:
TOWER I: 220 m high (around 60 stories) => constructed first TOWER II: 165 m high Foundation system: diaphragm wall panels
18

SOIL CONDITIONS AND ITS NUMERICAL MODELLING
Constitutive model Hardening Soil Model (HS): Slab
E 50 = E E ur = E
ref 50
c' cot '+ 3 c' cot '+ p ref c' cot '+ 3 c' cot '+ p ref
ref ur
E oed
Barrette length 30m
ref c cot + 1 = E oed c cot + p ref
Hardening Soil Small Model (HSS):

ref c' cot '+ 3 G0 = G0 c' cot '+ p ref
19

G0
Defintion of additional parameters for the HSS model: G0: Correlation between small strain stiffness and stiffness at larger strains after Alpan (1970) => Young's modulus interpreted as Eur
0.7
0.7: Stiffness reduction curves after Vucetic and Dobrey (1991)
Parameter E50ref (kPa) Eoedref (kPa) Eurref (kPa) pref (kPa) m (-) G0 (kPa) 0.7 (-)
Gravel 40 000 40 000 120 000 100 0.00 150 000 0.0001
Sandy silt 20 000 20 000 50 000 100 0.80 62 500 0.0002
Fine sand 25 000 25 000 62 500 100 0.65 78 125 0.0002
HS model
HSS model
20

21

FINITE ELEMENT MODELS - OVERVIEW
=> Both towers are modelled => Models consist of around 50 000 15-noded wedge elements

FINITE ELEMENT MODELS - DETAILED MODELS
Detailed model for Tower I
Barrettes with unit dimensions of 3.6 m x 0.6 m
Detailed model for Tower II
23

FINITE ELEMENT MODELS - FIRST RESULTS TOWER II Detailed model Tower I:

- Barrette length 25 m - uy,max of 80 mm - Interaction of towers - Differential settlements => Optimisation of foundation system
Eccentric loads lead to large differential settlements
24

OPTIMISATION OF FOUNDATION CONCEPT Final design of deep foundation system
Arrangement of diaphragm wall panels for Tower I
Arrangement of diaphragm wall panels for Tower II
25

RESULTS
26

RESULTS
=> Barrette lengths between 20 m and 30 m => Maximum vertical displacements of both towers ~80 mm => Differential settlements reduced to acceptable value

RESULTS WITH 64 BIT CALCULATION KERNEL
Model details: => No geometrical simplifications => ~ 140 000 15-noded wedge elements => Mesh refinement => 300 diaphragm wall panels are modelled explicitly
Expectations: => Validate first approach => Detailed information about settlement trough between towers
_________________________________________________________ Model Nr. El. Nr.Nodes Nr.El in 2D plane _________________________________________________________ 32 bit model 49 096 131 993 2 888 64 bit model 136 710 361 243 6 510 _________________________________________________________
28

RESULTS WITH 64 BIT CALCULATION KERNEL
=> Results are similar to first approach => Maximum vertical displacements ~80 mm => More accurate with respect to interaction and settlement trough between tower

RESULTS
Settlements / differential settlements => Road up to 40 mm
Tolerable => Highway - Point A 14 mm Inclination 1/900 => Train - Point B 18 mm Most critical area Inclination 1/600
=> Measurements during construction are necessary
30

RESULTS - HS VERSUS HSS (TOWER I)
=> Up to depth of -36.6 m Distribution is similar Difference is 25% => Beneath foundation elements Difference increases => At depth of -75 m HSS predicts 51% less vertical displacements
=> HSS predicts less settlements => Reduced influence of bottom boundary condition => right depth of influence is taken into account implicitly by the constitutive model

RESULTS - HS VERSUS HSS (CROSS SECTION A-A)
Difference of settlement troughs HS: - Settlements at the surface are higher - Spread of relevant settlement is wider - Conservative results for both maximum settlements and differential settlements
In other applications it is possible that the HSS model yields steeper settlements troughs => more critical scenario for adjacent buildings

Same length of deep foundation elements => high differential settlements. Final foundation concept with lengths between 20 - 30 m improves the settlement behaviour of the towers. In most calculations the geometry is simplified in that way that only one tower is modelled in detail and the foundation of the second tower is modelled with a homogenized block. The computed maximum vertical displacements are about 80 mm. The settlements of the highway and the railway lines are in the order of 15-20 mm. With a 64 bit calculation kernel both towers are modelled in detail. The settlements are similar to the simplified models but the settlement trough between the towers is more accurate. When using the HSS model the influence of the model boundary condition (in particular bottom boundary) is reduced and a more realistic settlement behaviour can be obtained.
33

CONTENT
ANALYSIS OF STATION TUNNEL (Bratislava) Design of tunnel lining according to EC7 TEN-T Project Analysis of Underground Station Tunnel Summary and Conclusion
34

EC 7 - INTRODUCTION
EC7 Design Approaches (DA1 to DA3)
Partial Factors for Action Design Approach Permanent Unfavourable Variable
Partial Factors for Soil Strength and Resistance tan c Undrained Shear Strength Passive Resistanc e
DA1/1 DA1/2 DA2 DA3
1.35 1.00 1.35 1.00
1.50 1.30 1.50 1.30
1.00 1.25 1.00 1.25
1.00 1.25 1.00 1.25
1.00 1.40 1.00 1.40
1.00 1.00 1.40 1.00
DA2* Partial factors are applied to effects of action
35

PROJECT TEN - T
Interconnection of Corridors Track expansion Connection of Airport
36

PROJECT TEN - T
Plan View of Underground Stations
37

PROJECT TEN - T
Cross Section of Platform Tunnel
38

GEOMETRY AND PARAMETERS OF NIVY STATION
Q1 Q5 N1
185 m 24m
67 m 13m
N5
Material parameters used for HS-Small

Soil Unit Q1 Q5 N1 N5 E50ref kN/m 7000 100000 18300 18200 Eoedref kN/m 7000 100000 18300 18200 Eurref kN/m 20160 200000 49000 48500 c kN/m 2 16 0.1 21 5 20 34 25.5 27 G0 MN/m2 22500 240000 65000 60000 0.7 1.5e-4 1.5e-4 1.5e-4 1.5e-4
Material parameters used for MC model

Soil Unit Q1 Q5 N1 N5 E kN/m 9500 140000 21000 21000 0.2 0.2 0.2 0.2 c kN/m 2 16 0.1 21 5 20 34 25.5 27
Stiffness parameters are reference values at tunnel level Water table: 5.5 m below surface Drained analysis Alert Workshop 2010 - Session Engineering Geostructures, Aussois, 4 October, 2010 39

CALCULATION (CONSTRUCTIONS) STAGES

Phase No. 1 2 3 4 5 6 7 8 9 Description Initial Phase Pilot Tunnel - Stress Release 20% Pilot Tunnel Excavation, Load 80% - Lining 5 GPa Pilot Tunnel - Lining Hardening 15 GPa Top Heading - Stress Release 20% Top Heading - Excavation & Removal of part of Pilot Tunnel - Load 85% - Lining 5GPa Bench & Invert - Stress Release 20% - Lining Hardening in TH 15GPa Bench & Invert - Excavation & Remove of Temp. Invert - Load 80% - Lining (B&I) 5 GPa Bench & Invert - Lining Hardening 15 GPa
40

RESULTS - NORMAL FORCES (CALCULATED VALUES)

DA3 DA2*
Mohr Coulomb
max. Value 1800 kN/m
HS-Small
Difference between DA2* and DA3 are not significant in case of calculated normal forces

RESULTS - BENDING MOMENTS (CALCULATED VALUES) - MC MODEL

DA*2 & DA2*-1 DA2*-2 DA3-3
Max Value 153
Max Value 103
Max Value 69
DA3 & DA3-1
DA3-2
Max Value 151
Max Value 103
DA3-1-1
Max Value 162
No Mpl
Mpl
Mpl/1.5
42

RESULTS - BENDING MOMENTS (CALCULATED VALUES) - HS-SMALL

DA2* & DA2*-1 DA2*-2 DA3-3
Max Value 98 153 DA3 & DA3-1
Max Value 102 103 DA3-2
Max Value 69
Max Value 101 151 DA3-1-1
Max Value 103 103
Max Value 103 162
Mohr Coulomb
No Mpl
Mpl
Mpl/1.5
43

SUMMARY OF RESULTS - BENDING MOMENTS

Max Calculated Moment
DA3-3 DA3-2 DA3-1-1 DA3-1 DA3 DA2*-2 DA2*-1 DA2* 0 50 100 150 200 250 MC material model HS Small
In presented case the calculated bending moments are almost independent of applied partial factors to soil (DA3 )
Max Design Moment

DA3-3 DA3-2 DA3-1-1 DA3-1 DA3 DA2*-2 DA2*-1 DA2* 0 50 100 150 200 250 MC material model HSS material model
=> DA2* yields higher design values (calculated bending moment x 1.35) than DA3
Using advanced material model yields in lower bending moments in this case
44

Small differences in calculated normal forces => design forces largest for DA2*. Modelling behaviour of a shotcrete lining as linear-elastic results in very conservative lining design. Relative stiffness between ground and tunnel lining has some influence on design forces => should there be a partial factor on stiffness of ground? Both design approaches (DA2* and DA3) have advantages and disadvantages => recommendation is difficult => more studies are required before definite conclusions can be drawn (if possible at all ) => compromise DA1?
45

CONTENT
JET GROUTING FOUNDATION FOR RAILWAY STATION (Vienna) Project layout Soil profile Finite element models Validation of numerical model Results
46

PROJECT OVERVIEW - PLAN VIEW
~ 500m
~ 400m
PLAN VIEW

PROJECT OVERVIEW - ARRANGEMENT OF JET GROUT COLUMNS
100m
175m
48

TYPICAL CROSS SECTION - SOIL PROFILE
8 m - gravel 3 m - sandy silt 14 m - sand
11 m - sandy silt 2 m - sand 10 m - stiff base layer
49

FINITE ELEMENT MODEL
> Model is too big - in phase activating the jet grout columns failure occurred
> 70000 elements > 1200 embedded piles
50

FINITE ELEMENT MODEL Three different models for the sensitive zones
Zone A: => 130 embedded piles => 51376 elements
Zone C: => 114 embedded piles => 47022 elements
Zone B: => 615 embedded piles => 47464 elements


VERIFICATION EXAMPLE
2D model
52

3D model - volume piles
3D model - embedded piles
53

VERIFICATION EXAMPLE 2D model: 72 mm
3D model - volume piles: 70 mm
3D model - embedded piles: 74 mm
ALL PILES VERTICAL


2D model: 68 mm 3D model - embedded piles: 70 mm
OUTER PILES INCLINED


EMBEDDED PILE MODEL FOR ZONE B
Zone B: => 615 embedded piles => 47464 elements
Model without soil - top view Model without soil - bottom view

EMBEDDED PILE MODEL FOR ZONE B
> 615 Piles

Different pile lengths Different pile inclinations > Rest is modelled as blocks

EMBEDDED PILE MODEL FOR ZONE B - RESULTS
inclined embedded piles
vertical displacements - uy,max = 63 mm
58

EMBEDDED PILE MODEL FOR ZONE B - RESULTS
Global settlement behaviour of construction is obtained because complete structure is modelled including areas with significant different load intensities. Detailed results in the zone where the embedded piles are modelled. Benefit of zone B (embedded piles): spacing, pile length and pile diameter can be modified with reasonable effort.

SUMMARY OF PROBLEMATIC ISSUES
DEEP FOUNDATION OF TOWER FOR SHOPPING MALL soil profile + parameters geotechnical report largely in Romanian language accuracy of stresses in foundation elements DEEP FOUNDATION OF TWO ADJACENT HIGHRISE BUILDINGS geotechnical expert based settlement estimate on very simple model interaction with structural engineers ANALYSIS OF STATION TUNNEL EC7 compatible design? JET GROUT FOUNDATION FOR RAILWAY STATION design of jet grout columns (code for concrete?) accuracy of stresses in columns
FINITE ELEMENT ANALYSIS OF DEEP FOUNDATIONS AND TUNNELS PRACTICAL APPLICATIONS
Helmut F. Schweiger Acknowledgement: Franz Tschuchnigg, Thomas Marcher (ILF Innsbruck)

Computational Geotechnics Group Institute for Soil Mechanics and Foundation Engineering Graz University of Technology

01 Schweiger PDF

Uploaded by

Copyright:

Available Formats

01 Schweiger PDF

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

01 Schweiger PDF

Uploaded by

Copyright:

Available Formats

FINITE ELEMENT ANALYSIS OF DEEP FOUNDATIONS AND TUNNELS PRACTICAL APPLICATIONS

Helmut F. Schweiger Acknowledgement: Franz Tschuchnigg, Thomas Marcher (ILF Innsbruck)

Alert Workshop 2010 - Session Engineering Geostructures, Aussois, 4 October, 2010

Finite element analysis of deep foundations and tunnels

Alert Workshop 2010 - Session Engineering Geostructures, Aussois, 4 October, 2010

Finite element analysis of deep foundations and tunnels

Alert Workshop 2010 - Session Engineering Geostructures, Aussois, 4 October, 2010

Finite element analysis of deep foundations and tunnels

SHOPPING MALL WITH SKY TOWER

Alert Workshop 2010 - Session Engineering Geostructures, Aussois, 4 October, 2010

Finite element analysis of deep foundations and tunnels

PROJECT OVERVIEW - SKY TOWER

Finite element analysis of deep foundations and tunnels

~20.5 m 60.0 m ~25.0 m

Assumption: Alternate layers continue

Finite element analysis of deep foundations and tunnels

c cot + 3 c cot + pref

Finite element analysis of deep foundations and tunnels

FINITE ELEMENT MODELS

Floors (Top-down method)

=> All models consist of about 52 000 15-noded wedge elements

Finite element analysis of deep foundations and tunnels

FINITE ELEMENT MODELS

2D "Plane strain" model

=> Only silty clays are modelled below -60 m

Alert Workshop 2010 - Session Engineering Geostructures, Aussois, 4 October, 2010

Finite element analysis of deep foundations and tunnels

FINITE ELEMENT MODELS

=> 2D representation of deep foundation elements not possible

=> For optimisation of foundation concept full 3D modelling is required

Finite element analysis of deep foundations and tunnels

OPTIMISATION OF FOUNDATION CONCEPT

Different lengths of barrettes

Circle of panels beneath concentrated loads

Finite element analysis of deep foundations and tunnels

OPTIMISATION OF FOUNDATION CONCEPT Final foundation concept

Alert Workshop 2010 - Session Engineering Geostructures, Aussois, 4 October, 2010

Finite element analysis of deep foundations and tunnels

OPTIMISATION OF FOUNDATION CONCEPT Final foundation concept

Finite element analysis of deep foundations and tunnels

VALIDATION OF NUMERICAL MODEL (IN SITU LOAD TEST)

Finite element analysis of deep foundations and tunnels

CONCLUSION FOR THIS PROJECT

Alert Workshop 2010 - Session Engineering Geostructures, Aussois, 4 October, 2010

Finite element analysis of deep foundations and tunnels

Alert Workshop 2010 - Session Engineering Geostructures, Aussois, 4 October, 2010

Finite element analysis of deep foundations and tunnels

Concept for foundation: deep foundation with diaphragm wall panels

Finite element analysis of deep foundations and tunnels

PROJECT OVERVIEW - TWIN TOWERS

Alert Workshop 2010 - Session Engineering Geostructures, Aussois, 4 October, 2010

Finite element analysis of deep foundations and tunnels

SOIL CONDITIONS AND ITS NUMERICAL MODELLING

Constitutive model Hardening Soil Model (HS): Slab

ref c cot + 1 = E oed c cot + p ref

Hardening Soil Small Model (HSS):

Alert Workshop 2010 - Session Engineering Geostructures, Aussois, 4 October, 2010

Finite element analysis of deep foundations and tunnels

SOIL CONDITIONS AND ITS NUMERICAL MODELLING