You are on page 1of 18

Problem Solving in

Tunnelling Projects Using


Numerical Modelling
W Y Soo1 and J Lee2
1. MSc, PE, MIEAust, CPEng, Associate Director, AECOM Brisbane,
QLD4006. Email: WaiYee.Soo@aecom.com
2. PhD, PE, FIEAust, CPEng, Associate Director, AECOM Melbourne,
VIC3008. Email: Jeawoo.Lee@aecom.com

Key Benefits of Solving Tunnelling Problems Using


Numerical Modelling

1
Fast and systematic
2
Possibility of using
3
Solution of coupled
4
Fast parametric
solution over spatial more realistic non- phenomena evaluation
variation linear material
behaviour

For example: Advanced constitutive soil Steady-state ground water Analysis output such as
Heterogenous soil layers models such as hardening flow and coupled flow settlement for impact
and soil-structure soil model, hardening deformation analysis. assessment, structure
interaction small strain model, member deformation and
generalised hardening soil forces for design.
model, cam clay model,
concrete model etc.
2D 3D
Retaining
structures of
excavation
corresponding
to 2D and 3D
structural
system
(M. Fernandes, 2012)

Case 1: Road Tunnel Construction in Kenny Hill Formation with


Deep Soil Mixing (DSM) and Cutter-Soil-Mix (CSM)
Geological Map

Soil Investigation Bore Hole Location, Survey and Tunnel Location

Geotechnical Design Parameters


Tunnel A:
Temporary Works:
Layout,
Longitudinal
Section and
Geological Profile

Geotechnical
Interpretation
Plots
Tunnel A:
Design Section

Typical Construction Sequence


Fundamental Stability Checks
Toe Stability
- The toe-in depth or embedment of retaining wall has to be determined to ensure the passive resistance can be sufficiently
mobilised.

Check for adequacy of toe-in depth Example of toe kick-out failure mechanism (Philips et al.
1993) and plastic points in PLAXIS 2D

Fundamental Stability Checks


Hydraulic Uplift
- To ensure that the base of the will not “blow in”
- In the temporary condition, the water cut-off during
excavation would be provided by DSM / CSM.
- The minimum factor of safety by hydraulic uplift in
the temporary condition shall be 1.2.

k2 >> k1

Steady-state > PlaxFlow > Phreatic

Factor of safety against hydraulic uplift

12
Fundamental Stability Checks
Floatation Assessment
- Underground structures shall be checked Considering 1m run of the structure,
for the possibility of floatation at all stages
of the construction and throughout the = unit weight of water
service life of the structure. = height of structure
= width of structure
- In the permanent condition, ground water
level is considered at the existing ground
level for design purpose.
+ +
- The minimum factor of safety for
floatation assessment in the permanent = submerged weight of backfill material
condition shall be 1.2, which was met for = average shear resistance along a - a’
Floatation Assessment the proposed structure. = self-weight of structure
= depth of backfill
- Nevertheless, the NCBP toe-in depth = width of structure
turned out governed by the toe stability = partial safety factor for weight of soil
during the excavation rather than the
= partial safety factor for weight of structure
floatation assessment in the permanent
= partial safety factor for shear resistance
condition.

Fundamental Stability Checks


FEM Safety Calculation / c-phi Reduction Simulation sequence
- Stability analysis was performed using finite element
analysis in conjunction with the ‘phi-c reduction’
approach, which is also called ‘Safety Calculation’.

- This method iteratively imposes reduction of the soil


strength parameters, namely the friction angle, and
the cohesion, c, until the model becomes unstable
instead of the usual incrementation of loads
(Zienkiewicz, 1975).

- The procedure has been made more robust in Global Stability Check Summary
PLAXIS by the addition of the arc-length technique. Stability Check Factor of Safety

- The minimum FOS is required to be not lower than Toe Stability > 1.0
1.0. Hydraulic Uplift > 1.2
Floatation Assessment > 1.2
= =
FEM Safety Calculation > 1.0
FEM Analysis
2D FE analysis with plane strain model and 3D FE Analysis to study the behaviour of the
seepage groundwater flow DSM / CSM

Temporary
Non CBP and strutting system
will be temporary retaining
wall to enable the rc box
tunnel construction.
DSM will provide water cut-off
during excavation.
Permanent
Non CBP will be deep foundation element
(piled foundation) to resist uplift force
during permanent condition.

FEM Analysis
Analysis Output from 2D FE Analysis and the Design
Settlement Contour Plot at the Railway Protection Zone 1st and 2nd Reserve
Tunnel A Tunnel B

CSM

CSM Wall as Retaining Wall CSM Wall as Cut-Off Wall

Comparison Among Different Typical Cut-Off Retaining Wall


CSM Wall

FEM
Input

Advanced Constitutive Model: Shotcrete / Concrete Model

Ground improvement is modelled with Hardening Soil Model Ground improvement is modelled with Concrete Model
and there are tension cut-off points and no tension cut-off points here
Case 2: Railway NATM Tunnel Construction:
Ground Reaction Curve and Settlement Analysis

NATM Tunnel
Three-dimensional (3D) FE-Model Deformed Mesh, 2 Steps Excavation, and soil/rock layers
NATM Tunnel

Three-dimensional (3D) FE-Model Deformed and d=1.0m

NATM Tunnel
Longitudinal Section of the 2-Steps Excavation
NATM Tunnel
Ground Reaction Curve

(a) Crown deformation (+: downward) b) Display of load reduction method adopting ground response curve
Deformation profile and value at tunnel crown from 3D analysis

Display of load reduction method adopting


ground response curve (Vermeer et al, 2005)

Ground reaction and (1-MStage) for PLAXIS 2D excavation staging

NATM Tunnel
Comparison of Two- and Three
Dimensional FE-Analysis

FIG 17 presents the structural member


forces in the tunnel lining calculated from
the 3D and 2D analyses when the tunnel
excavated at 20 m advance ( value =
0.46). Despite the difference in the
magnitude of member forces, the maximum
member forces are reasonably comparable
to one another.
FIG 17 Comparison of three- and two-dimensional structural forces
NATM Tunnel
Detail Impact Assessment

FIG 18 3D FE-Analysis: NATM tunnel excavation induced groundwater drawdown and the existing power pylon.

NATM Tunnel
Detail Impact Assessment

FIG 19 3D FE-Analysis: NATM tunnel excavation induced groundwater drawdown and the settlement contour.

FIG 20 3D FE-Analysis: Canopy tube axial force (N) and bending moment (M).
Case 3: Possible Ground Settlement due to Tunnel Excavation
and Groundwater Drawdown during 4 Hours CHI

3D PlaxFlow: deformed mesh and permeabilities of each soil / rock layers Snapshots during CHI

FIG 21 CHI and 3D PlaxFlow deformed Mesh

TBM Tunnel: CHI


Flow Analysis Results

Possible groundwater ingress during CHI Groundwater head contour and groundwater head at CHI location

Groundwater head contour at 150m from CHI location Groundwater head at 150m from CHI location
TBM Tunnel: CHI
Ground Settlement
Stage At CHI Location At 150mm from CHI
Cumulative Maximum Cumulative Maximum
Maximum Settlement Maximum Settlement
Settlement At Phase Settlement At Phase
uz uz uz uz
mm mm mm mm

0 Initial phase 0 - 0 -

1 XX tunnel excavation
15.5 15.5 20.4 20.4
(1% VL)

2 CHI for 4 hrs 20.1 4.6 20.7 0.3

Ground settlement at CHI location during Stage 02 and Stage 03

Ground settlement at 150m from CHI location during Stage 02 and Stage 03

TBM Tunnel: CHI


Results and Assessment
Stage At CHI Location At 150mm from CHI Total discharge for 4 hours CHI = 0.165 m3/s
Cumulative Maximum Cumulative Maximum = 990 l/min
Maximum Settlement Maximum Settlement (compared to site record 500 – 800 l/min)
Settlement At Phase Settlement At Phase Consolidation parameters for F2 Clay: mv = 0.3m2/MN, Cv=2 m2/yr
uz uz uz uz
mm mm mm mm Consolidation Settlement
0 Initial phase 0 - 0 - The groundwater head is estimated to drop = 103.5 – 99 = 4.5m
1 XX tunnel F2 clay (compressible layer) thickness, H = 3m
15.5 15.5 20.4 20.4 = 22.5 kN/m2
excavation (1% VL)
Possible consolidation settlement = 14mm
2 CHI for 4 hrs 20.1 4.6 20.7 0.3

Using Terzaghi 1D consolidation,


t90 = 0.848 x 32/ 2 = 3.816 yr (approx. 4 yrs)
21mm settlement happens in 3.816 yr
Conclusion Possible settlement for 4 hrs CHI for F2 clay = 14 * 4 hr / 3.816 yr
= 0.00168mm
From PLAXIS 3D FEM and PlaxFlow: < 1mm
• Estimated discharge flow rate during the CHI is approximately 990 l/min.
• With highly fractured rock beneath (GIV and GII /GI), the predicted ground settlement during 4 hours CHI is approximately 5mm
(compared with site record is approximately 3 to 4mm)
• Predicted long-term consolidation settlement for Kallang Formation F2 Clay is 14mm. This will take about 4 years to complete
the consolidation settlement.
• Therefore, under the 4 hours of CHI, the anticipated consolidation settlement is less than 1mm.
Concluding Remarks

1
Successful Numerical
2
Cut-&-Cover Tunnel:
3
NATM Tunnel:
4
TBM Tunnel:
Prediction, Cost Saving and FE Modelling for Basic Stability Ground Reaction Curve & Detailed Ground settlement due to
Safe Construction Check Impact Assessment groundwater drawdown
during CHI

• A successful numerical • The 1st case demonstrates FE • The 2nd case demonstrates a • The 3rd case provides a
prediction of the modelling can ensure the stability reasonable approximation of a 3D practical solution to a
performance of a against conventional failure arching effect in 2D FE modelling tunnelling project where a
supported excavation can modes. of NATM tunnelling by method. drawdown of groundwater
bring a significant benefit • The mechanism of a failure mode • The value, which represents during an unavoidable event
to a tunnelling project by such as toe-in stability can be support pressure in a ground such as CHI in TBM
achieving cost saving and identified in numerical modelling reaction curve, is estimated when tunnelling poses a risk to
safe construction. by tracking yielding/failure points, an upper heading round is existing structures in a busy
• This paper presents case which can help optimise the excavated before the installation urban area.
studies of tunnelling extent of ground improvement of shotcrete lining in the case • The 3D transient seepage
projects in which technical where necessary. presented. modelling and subsequent
difficulties were resolved • The hydraulic uplift of an • The 3D modelling facilitates the mechanical analysis
employing numerical underground structure can be estimation of ground subsidence presented manage to provide
modelling in design and observed via steady-state induced by groundwater for a reasonable prediction of
construction stages. groundwater flow analysis and fluctuation caused by tunnelling water ingress and resulting
subsequent mechanical analysis. so that the impact of tunnelling on ground deformation, which
• Safety calculation based on c-phi the adjacent existing structure aid in controlling ground
reduction provides a global can be quantitatively assessed, subsidence with I&M at the
stability check for a deep and mitigation of construction site.
excavation scheme. induced hazards.

Recommendations

Finite element analysis is a useful tool for the design and


construction of tunnels throughout preliminary and
construction phases; however, the modelling results should
be implemented and interpreted with a good judgement of
the engineer such as ground models, constitutive models,
parameters, implementation, failure criteria and safety
metrics.

Groundwater control is essential in urban transportation tunnelling


where tunnelling induced harm on existing structures are likely to
GWT be an issue. Dewatering and/or seepage-mechanical coupled
analysis will assist in mitigating construction induced hazards. To
this end, advanced numerical modelling needs to be allowed for
even in early phases of the tunnel design.
Thank you

You might also like