Module 4: lining analysis

Segmental lining design process

TBM/lining Effects of TBM Stacking and
TBM Ram loads
interface operation handling loads

Handling Check fixtures

constraints and fittings

Define Define ground Analyse effects Check structural

geometry loads on lining capacity

Manufacturing Check joint

constraints capacity

Build
Alignment Grouting loads Joint behaviour
imperfections

©Anthony Harding/Jacobs 2022

What loads are we analysing for?
 Vertical and horizontal pressures
− Ground load
− Water loads
− Ground movements (such as from Pv
adjacent tunnels)
 Other loads
− Internal loads
− Surcharge/excavation effects
Ph
− Seismic
− Thermal

©Anthony Harding/Jacobs 2022

Three analysis types
 Closed form
− Simple analyses using equations
− Provide rapid answers and help with ‘what-if’ analyses
 Bedded beam
− Analyses in structural software
− The lining is typically represented as beam elements (or plates or solid elements)
− The ground represented as supporting springs
 Continuum
− The ground is represented as solid elements
− The lining is typically represented as beam elements (or plates or solid elements)
 Procedures for analysing loads on segmental linings are broadly similar to
conventional lining design

©Anthony Harding/Jacobs 2022

Module 4 Session 1: Closed form
solutions for linings
Why use closed form solutions?
 Easy to formulate (in excel)
 Help us understand relative influences of:
− Ground load
− Differential stress
− Ground stiffness
− Lining stiffness
− Deflection
− Moment
− Axial force
 What is the problem sensitive to?
 Guides future analysis

6 ©Anthony Harding/Jacobs 2022

Inputs
 Vertical and horizontal pressures
− Ground load
− Water loads
− Ground movements (such as from Pv
adjacent tunnels)
 Other loads (effects can be added
separately)
− Internal loads
Ph
− Surcharge/excavation effects
− Seismic
− Thermal

©Anthony Harding/Jacobs 2022

General principles – Curtis/Muir Wood; Duddeck and Erdmann
Pdist
Pv Pav Pdist
Pav
Pav

Ph Pdist
= Pav
+ Pdist

Insitu pressure regime Average pressure Distortional pressure

©Anthony Harding/Jacobs 2022

Basic principles of lining ground interaction
 Lining moves in:
− Arching develops in the ground
− The ground resists some of the Arching
load
− Ground load on the lining
reduces
 For large load reductions
plasticity must be accounted for
 For small reductions in load
elastic solutions may be used

9 ©Anthony Harding/Jacobs 2022

Basic principles of lining ground interaction
 Lining moves out:
− The opposite of moving in
− Load increases
 Similar to mobilising passive
pressure in a retaining wall
 For small increases in load
elastic solutions may be used

10 ©Anthony Harding/Jacobs 2022

Effects of average pressure
 Pressure acts on lining
 Lining contracts Pav
 Load on the lining reduces
 Usually ignored in soft ground:
− Benefits typically low
− Grouting effectively ‘pre-loads’ the
lining
 In rock Pav
− Explicitly accounted for in
convergence confinement methods

11 ©Anthony Harding/Jacobs 2022

Effects of distortional pressure
 Lining distorts
− Lining resists the load in bending
− Lining deforms to an ellipse
Pdist
 At the crown Ground load
− Deforms downward (away from the ground) reduces
− Ground load reduces
 At the axis Pdist
− The lining pushes into the ground
− Ground load increases
− The distortional pressure (difference
between the pressures) reduces Ground load
increases

©Anthony Harding/Jacobs 2022

Closed form solutions
 Flexible tunnel in elastic medium
 Multiple approaches:
− Curtis/Muir Wood
− Duddeck and Erdmann
 Simple and versatile
− Can be used in spreadsheets
− Analyse ranges of inputs very quickly
Always use as a first pass to work out what your critical sections are

©Anthony Harding/Jacobs 2022

Flexibility factor 1. Relatively Bending moment
stiff lining: high Increases with
 ‘Flexibility factor’ – relative flexibility bending deflection
of lining and ground: moment
1. Rigid lining in soft ground: most of
the distortional pressure is resisted
in bending Distortional load
Solution
reduces with

Load
2. Flexible lining in stiff ground: most deflection
of the distortional pressure is
dissipated by ground deformation 2. Relatively
flexible lining:
 Important to understand the relative high deflection
flexibility

Deflection

14 ©Anthony Harding/Jacobs 2022

Flexibility factor
 The relative flexibility of the ground
Ground stiffness
Radius to centroid
Flexibility
factor Lining second moment of area
Lining elastic modulus

 At low values (high lining stiffness/low ground

stiffness):
− Lining resists most of distortional pressure
− Moments are high
 At high values (low lining stiffness/high ground
stiffness):
− Ground resists distortional displacements
− Moments are low
15
Duddeck & Erdmann (1985) ©Anthony Harding/Jacobs 2022
Flexibility factor
 Example Bending moment in lining vs stiffness of ground
− Fully bonded (no slip between 300
Unsupported
lining and ground) pipe

Bending moment (kNm/m)

250 Very soft soil
Medium clays,
Vertical pressure on
crown
381kPa 200 loose sands
Hard clays,
Horizontal pressure 150
283kPa dense sands
at axis
Sedimentary
Poisson's ratio 0.3 100 rock
Lining radius, r 3.15m
Emodulus 34,800MPa 50

Second moment of 0
0.000475m4
area, I 0.01 0.1 1 10 100 1000 10000
Stiffness of ground MPa

16 ©Anthony Harding/Jacobs 2022

Axial loads
 Also have a flexibility factor
 Typically ignored in soft ground because:
− Benefits tend to be low
− Lining is pressurised on installation by the grout
loads
 In rock:
− Benefit can be higher
− Explicitly considered in convergence
confinement methods

Duddeck & Erdmann (1985)

17 ©Anthony Harding/Jacobs 2022

Effect of accounting for shear stress
 Insitu shear stresses
 Differential distortion between ground
and lining Shear
 Friction between lining and ground can stress, T
be low
 Often examine two cases:
− Full slip (friction = 0)
− Fully bonded (no slip, friction > shear
stress)

18 ©Anthony Harding/Jacobs 2022

Effect of accounting for shear stress
 Example: Change in moment and axial force by varying
shear stress limitations
Vertical pressure on 1400 200
381kPa
crown

Bending moment (kNm/m)

1200
Horizontal pressure
283kPa
at axis 150

Axial force (kN/m)

1000
Eground 5MPa
Poisson's ratio 0.3 800
Maximum axial force 100
Lining radius 3.15m 600
Emodulus 34,800MPa Minimum axial force
400
Second moment of Limiting axial force r.Pv 50
0.000475m4
area, I 200
Maximum bending moment
0 0
0 10 20 30 40 50 60
Limiting shear stress
19 ©Anthony Harding/Jacobs 2022
Other inputs – impact on relative flexibility
 Young’s modulus
− May differ between short and long term
− Increasing concrete stiffness increases moments
− Increasing ground stiffness decreases moments and deflections
 Shear strength – the bond between the lining and the ground:
− Increasing shear strength reduces bending moments and increases maximum axial force
− However it can be uncertain
 When the above values are uncertain, a range may be used

©Anthony Harding/Jacobs 2022

Stiffness
 The joints introduce points of rotation
 This reduces the stiffness
 Can be approximated using Muir-Wood formula:

 Where is the number of segments

 Design should consider a range (typically ±25% on
Lower bound (max.
No of segments Upper bound (max. BM) Best estimate
deflection)
5 𝐼 4 0.3 × 𝐼
.𝐼
7 0.7 × 𝐼 𝑛 0.15 × 𝐼
+𝐼
9 0.5 × 𝐼 Fully flexible

©Anthony Harding/Jacobs 2022

Morgan’s equation
 Can provide an estimate of the increase
in moment resulting from a ground
deformation:

22 ©Anthony Harding/Jacobs 2022

Closed form pros and cons

Pros Cons
 Quick and easy  Elastic only
 Captures ground/lining  Circular sections only
interaction  Isotropic materials
 Very powerful for  Can’t explicitly incorporate joints
sensitivity and route wide
analysis  Can’t incorporate discrete loads
 Not suitable where plastic behaviour is expected
 Can’t capture complex ground behaviours

23 ©Anthony Harding/Jacobs 2022

Module 4 Session 2: Bedded
beam models
Bedded beam

 Lining is represented as
Pv
beams Ph
 Ground is represented as
springs
– Normal springs with
tension release
– Shear springs (sometimes
omitted or limited)
 Joints can be modelled:
– Pins
– Rotational springs
 Can be 2D or 3D

25 ©Anthony Harding/Jacobs 2022

Bedding springs

 Ground springs represent

Pv
the ground action Ph
 Normal Springs
– Usually set to tension only
– Sometimes omitted in the
crown
– Often set to “Zero tension”
 Tangential springs
– Usually less stiff than
normal springs
– Can be omitted
– Shear can be limited
(cohesion or friction)

26 ©Anthony Harding/Jacobs 2022

Loading

 Loads are specified directly Pv

– Don’t account for arching Ph
 Variable ground loads
– Patch loads (wedge,
grouting)
– Varying loads
 Internal loads
 Self weight

27 ©Anthony Harding/Jacobs 2022

Load factors and table
 Load can be described SLS1 SLS2 SLS3 ULS1 ULS2 ULS3 ULS4 ULS5
individually and then Self weight 1.0 1.0 1.0 1.35 1.0 1.35 1.35 0.9
combined Ground load 1.0 1.35 1.0

 Some codes require the Water 1.0 1.0 1.35 1.0 0.9
pressure
loads be factored in tis
Flood water 1.0
way. pressure
Rock wedge 1.0 1.35
Grout loads 1.2
Internal 1.0 1.1
water
pressure

28 ©Anthony Harding/Jacobs 2022

Lining model

 Lining is represented as
Pv
beams Ph
– Usually straight between
springs
– EI from lining (usually per
m out of plane)
– Can use reduced I
 Can model joints
– Pin (overestimates
rotations, underestimates
moments)
– Rotational spring between
two elements

29 ©Anthony Harding/Jacobs 2022

Modelling springs – Janssen joint

tcz

tcz
tcz

True joint Janssen joint Janssen joint - open

30 ©Anthony Harding/Jacobs 2022

Beams and springs
 2D representation
 All loads can be directly controlled
− Ground
− Water
− Internal
− Grout

31
©Anthony Harding/Jacobs 2022
Coupled 2D models

 Multiple bedded beam

models:
– Two whole rings
– Two half rings
– One ring with two half rings
on either side
 Elements between rings:
– Radial, circumferential, and
normal stiffness
– Friction between rings
– Connectors
 Captures force transfer
between rings

32 ©Anthony Harding/Jacobs 2022

Bedded plate models
 Beams -> plates
 Can use reduced stiffness
(Muir-wood) or model joints
explicitly
 Captures longitudinal variation
in forces
 Very useful for openings!

33 ©Anthony Harding/Jacobs 2022

Example: designing for internal water pressure
 Worst case internal pressure 25 kPa
 3D bedded beam model
− One whole ring plus two half rings
 Ground restraint included
 Bolts modelled
 Dowels modelled
 Maximum bolt force: 75kN
 No-bolt scenario checked
Bedded beam pros and cons

Cons Pros
 Elastic only  Can model discrete loads
 Don’t capture ground  Can incorporate joints
relaxation/arching  Can capture stiffness changes
(following load)
 Can model connections between segments
 Can’t apply ground
movements  Can model connections between rings
 Can quickly generate multiple load cases with multiple
load factors

35
©Anthony Harding/Jacobs 2022
Timing of loading
 Linings are affected by creep
− Usually modelled as a change in Young’s Modulus: ,

− Immediate ground and water loads
− Immediate loads on structure

− Long term ground loads
− Loads that develop slowly (e.g. pile loads)
 Combinations:
− for loads applied a significant time before
− for loads at time of application

36 ©Anthony Harding/Jacobs 2022

Example of using different Young’s modulus
 Road tunnel
 Buildings at the surface
 Internal structures are built shortly before tunnel opening
 Load cases:
− Collision/impact
− Seismic

Case Long term analysis Short term analysis

Traffic impacts Ground and water loads Internal structures
Surface surcharges from buildings Traffic loads
Traffic impacts
Seismic Ground and water loads Internal structures
Surface surcharges from buildings Traffic loads
Seismic loads/deformations

37 ©Anthony Harding/Jacobs 2022

Cracking and hinges
 In tunnel linings deformation of the lining
typically reduces moments
 Cracking reduces moments
− In zones of maximum moment Cracked length
− As a first pass might adopt 50% stiffness
− Some structural software will calculate it for
you
 Hinges can reduce moments
− Rotational limits specified in codes (RC
sections) Hinge
− Rotation usually limited to 4 mrad or 0.23°
(plain and FRC)

38 ©Anthony Harding/Jacobs 2022

Module 4 Session 3: Finite
element methods
Finite element/discrete element/finite difference

 Ground is represented as
solid elements
 Lining can be beam or
shell elements, or solid
elements
 Soil behaviour can be
captured with
constitutive models

40 ©Anthony Harding/Jacobs 2022

Constitutive elements

Soil models Structures Interfaces

• Elastic • Beams • Bonded

• Mohr Coulumb • Plates/shells • Pins
• Hardening Soil • Rotational • Springs
• Cam Clay springs • Friction
• Hoek-Brown • Hinges • Gaps
• …

41 ©Anthony Harding/Jacobs 2020

What do I want to know?
 Modelling is easy but validation can be hard
 What is my question?
− What is the effect of…
− How much does…
− Can I justify a reduction in…
 Try to do just as much modelling as you need to answer the question

42 ©Anthony Harding/Jacobs 2020

Example 1: Shallow launch

 Upper sediments much

softer than lower
 Risk of flotation
 Protection slab above
surface
– Protects services
– Reduces flotation risk
 Oucomes
– Checked displacements in
flotation case
– Worst case moments
– Allowed for changes in
stiffness

43 ©Anthony Harding/Jacobs 2022

Example 2: constitutive model for complex behaviour

 Captured:
– Relaxation
– Short term stress changes
in clay (undrained)
– Pore water pressures in
clays
 Long term consolidation
effects
– Changes in ground stress
– Movements
– Changes in axial force and
bending

44 ©Anthony Harding/Jacobs 2022

Example 3: effects of second tunnel
Reduced hoop Reduced horizontal
 Tunnel in rock loads stresses
 High horizontal stresses
 Effects of second tunnel
– Reduced load on lining
– Deformation (squatting)
– Increased moments
First tunnel Second tunnel

Increased moments

45 ©Anthony Harding/Jacobs 2022

Example 4: 2D axisymmetric for minimum pressure

 The problem:  Net water pressure is outwards

– Design surge pressure in  If soil pressure < 90kPa then the lining can be in
tunnel tension
– 90kPa higher than external
water pressure  What is the minimum soil pressure on the lining?
– If ground loads are low
lining goes into tension!
Net pressure = -90kPa
 Answer: convergence
confinement in soil!
Pnet  uexternal  uint ernal  P ' soil

46 ©Anthony Harding/Jacobs 2022

Internal pressure
 Soil loads are very difficult to Factored
predict accurately load Ground loads
 Normal design procedures yield
upperbound values
 Maximum tension occurs with
minimum ground load

Net water
pressure
Design
value

Net pressure

Simple analysis Refined analysis

©Anthony Harding/Jacobs 2020
Axisymmetrical modelling
 TBM operation effects the
relaxation that occurs
Grout pressure
 The influence of the Annulus pressure
Annulus pressure Grout pressure
moved by length L
moved by length L
operational factors was Area of grout pressure from
required Face pressure Fixed radially
previous step fixed radially
 An axisymmetrical model was
used to evaluate the relative Axis of Symmetry:
influence of the different Face excavated Tunnel Centre Line
factors by length L Step
Stepn+1
n
 Lining was checked for
resulting pressure

©Anthony Harding/Jacobs 2020

Example 5: 3D convergence confinement in rock

 Sequentially excavate ring

lengths in rock
 TBM length is unsupported
 Lining loads are calculated
explicitly within the model
 Second tunnel can also be
captured

49 ©Anthony Harding/Jacobs 2022

Example 6: 3D models for minimum ground load

 Lining is permanently in
tension
– Conservative assessment of
minimum ground load
– Tension is large: around
300kN/ring
 Can we increase the
minimum design load and
decrease the required bolt
load?

50 ©Anthony Harding/Jacobs 2022

Riachuelo scheme – river catchment
 Sewer outfall
 Pumped from Diffuser with
31 Risers
treatment plant
 Over 10km
Buenos Aires
City

Treatment
Plant
(Not part of
Contract)
Outfall
Outfall long section
 Maximum
net internal
Net internal pressure pressure is
120kPa Risers 120kPa
Max River level = 15.90 m
(1500m)

0+000 5+000 10+400 12+000

Modelling

 Modelled:
– Face pressure
– Shield restraint
– Grouting
– Hardening soil model
 Sensitivity analyses:
– Face pressure
– Overcut
– Grout pressure
– Mohr-coulomb soil model
 Result
– Increased ground loads
– Bolt forces 220kN/ring
53 ©Anthony Harding/Jacobs 2022
Bolts
 T28, stainless steel bolts
 160 mm long plastic
sockets
 Long term creep and
strength loss of plastic
 Extensive testing
 Just met capacity!

54 ©Anthony Harding/Jacobs 2020

Numerical modelling – pros and cons

Pros Cons
 Can analyse complex 3D shapes (such as cross  Requires the most effort to:
passage excavations and openings) – Create models
 Can capture relaxation/convergence explicitly – Validate
 Can model the effects of nearby excavation – Check

 Can model the joints  There is no guarantee you will get

any benefit
 Can model complex construction sequences
 Can capture plastic soil behaviour
 Can capture long-term effects, such as
consolidation

55 ©Anthony Harding/Jacobs 2022

Session 4 Modelling connections
between elements
Impact of the joints on lining flexibility
Joint geometry and rotation
 Joint can be in full contact or partial contact
 Joints can carry moments, but are more flexible

Rotation with full Rotation with partial

No rotation
contact contact (joint opening)

©Anthony Harding/Jacobs 2022

Modelling springs: Janssen springs
 Joint is represented as a
short column
 Height of column is the
same as the contact width
 Width of column is reduced
when the joint is open Column

Actual Assumed
geometry geometry

58 ©Anthony Harding/Jacobs 2022

Modelling springs: Janssen springs

 Joint is represented as a Joint is represented as a

short column, length L short column as shaded
 The width of the column is:
– L, if the joint is closed
– x, if the joint is open
 FE models usually report L L
higher stiffnesses
 The behaviour typically
matches experimental L x
behaviour quite well
Rotation with full Rotation with partial
contact contact (joint opening)

59 ©Anthony Harding/Jacobs 2022

Janssen springs - application
Moment vs rotation
 Provides a way of explicitly 900
modelling the joint 800
 Provides a reasonable 700

Moment (kNm)
match to experimental
data 600

 Can be complex to 500

implement 400
– Non-linear 300
– Dependent on axial force N=3500kN
200
 Other models exist but N=5000kN
offer little improvement 100 N=7000kN
(presenter’s view) 0
0 5 10 15 20 25 30
Rotation (mrad)
60 ©Anthony Harding/Jacobs 2022
Reduced stiffness approach
 Applied to the whole lining stiffness
 Can be approximated using Muir-Wood formula:

 Where is the number of segments

 Design should consider a range (typically ±25% on
 Can often bound the problem as well as more complex analyses
Lower bound (max.
No of segments Upper bound (max. BM) Best estimate
deflection)
5 𝐼 4 0.3 × 𝐼
.𝐼
7 0.7 × 𝐼 𝑛 0.15 × 𝐼
+𝐼
9 0.5 × 𝐼 Fully flexible
Extract from company internal design guide, circa 2000

©Anthony Harding/Jacobs 2022

Connections between rings
 Elements between rings:
− Radial
− Circumferential
− Longitudinal (normal stiffness, Radial
usually compression only) Circumferential
 Friction: how much can I rely on? Longitudinal
 Connectors (dowels)

62 ©Anthony Harding/Jacobs 2022

Normal stiffness and friction
 Longitudinal force:
− Can be present from TBM ram thrust
− Can be present from segment movements
− Need to account for creep and shrinkage
 Friction

Friction
− Code supplied values may be used with
caution Longitudinal force
− Steel moulds + release oil can lead to low
friction values

63 ©Anthony Harding/Jacobs 2022

Dowels
 Cover most movements
− Radial (shear)
− Circumferential (shear)
− Longitudinal (usually
tension only)
 Stiffness/movement
defined by manufacturers
 Need to check capacity

64 ©Anthony Harding/Jacobs 2022

Bolts
 Circumferential bolts have limited
shear stiffness
 Usually consider tension only

65 ©Anthony Harding/Jacobs 2022

Module 4 Session 5: Checking
capacity (moment-axial)
Checking capacity
 The purpose of the analysis is to
prove you have sufficient capacity
∅𝑀 ≥ 𝑀 ∗
 Varies between codes of practice
𝑉 ≤ ∅𝑉
− Moment-axial
− Shear in section 𝑀 ≤𝑀
− Shear at joint (friction)
−
−
Bearing
Bursting
Capacity ≥ Effect
− Punching shear 𝑉 ≤𝑉
− Cone pull-out
∅𝑉 ≥ 𝑉 ∗

𝑀 ≤ ∅𝑀

©Anthony Harding/Jacobs 2022

Moment-axial Moment-Axial

 All points within curve Point at Capacity

Points within
capacity

Axial Force

Moment
Envelope Load case 1 Load case 2

©Anthony Harding/Jacobs 2022

Moment-axial force: basic concepts
 In ULS:
− Consider a number of neutral axis locations in the section
− Assume a compression block up to a maximum strain (usually 0.3 to 0.35%)
− Assume a tension block in concrete
 Tension = 0 for plain concrete
 Tension per defined tensile stress block for fibre reinforced concrete
− Detemine the resulting axial force in tension

69 ©Anthony Harding/Jacobs 2022

Moment-axial force: basic concepts

Centroid
Neutral
 For a given neutral

axis
axis: Compression Tension
Ls
− Compressive strain = Max strain
maximum Fct Fs
− Develop stress blocks
 Axial force Lct

+
 Bending moment

+ Fc

Lc

70 ©Anthony Harding/Jacobs 2022

Compression blocks
 Vary from code to code:
− Rectangular
− Bilinear
− Parabola/rectangular Fcd
− Parabola
 Some codes allow
multiple methods

0.8x εc2
εcu2
x

71 ©Anthony Harding/Jacobs 2022

Tension blocks
 Rectangular
model
− Can be applied
over the whole
tension area or
over 0.9x
 RILEM model
− Needs values
from EN14651:
 fctm,fl 0.37fR4
 fR1 0.45fR1
 fR4
0.9x to 1.0x
 Other models per fctm,fl
codes of practice
72 ©Anthony Harding/Jacobs 2022
The moment-axial force diagram

9000

8000

7000

6000

Axial froce (kN)

5000

4000

3000

2000

1000

0
Maximum tension 0 50 100 150 200 250 300 350
-1000
Bending moment (kNm)

73 ©Anthony Harding/Jacobs 2022

The moment-axial force diagram

9000

8000

7000

6000

Axial froce (kN)

5000

4000

3000

2000

1000

0
0 50 100 150 200 250 300 350
-1000
Bending moment (kNm)

74 ©Anthony Harding/Jacobs 2022

The moment-axial force diagram

9000

8000

7000

6000

Axial froce (kN)

5000

4000

3000

2000

1000

0
0 50 100 150 200 250 300 350
-1000
Bending moment (kNm)

75 ©Anthony Harding/Jacobs 2022

The moment-axial force diagram

9000

8000

7000

6000

Axial froce (kN)

5000

4000

3000

2000

1000

0
0 50 100 150 200 250 300 350
-1000
Bending moment (kNm)

76 ©Anthony Harding/Jacobs 2022

The moment-axial force diagram

9000

8000

7000

6000

Axial froce (kN)

5000

4000

3000

2000

1000

0
0 50 100 150 200 250 300 350
-1000
Bending moment (kNm)

77 ©Anthony Harding/Jacobs 2022

The moment-axial force diagram

9000

8000

7000

6000

Axial froce (kN)

5000

4000

3000

2000

1000

0
0 50 100 150 200 250 300 350
-1000
Bending moment (kNm)

78 ©Anthony Harding/Jacobs 2022

The moment-axial force diagram

9000

8000

7000

6000

Axial froce (kN)

5000

4000

3000

2000

1000

0
0 50 100 150 200 250 300 350
-1000
Bending moment (kNm)

79 ©Anthony Harding/Jacobs 2022

The moment-axial force diagram

9000

8000

7000

6000

Axial froce (kN)

5000

4000

3000

2000

1000

0
0 50 100 150 200 250 300 350
-1000
Bending moment (kNm)

80 ©Anthony Harding/Jacobs 2022

The moment-axial force diagram

9000

8000

7000

6000

Axial froce (kN)

5000

4000

3000

2000

1000

0
0 50 100 150 200 250 300 350
-1000
Bending moment (kNm)

81 ©Anthony Harding/Jacobs 2022

The moment-axial force diagram
Maximum compression
9000

8000

7000

6000

Axial froce (kN)

5000

4000

3000

2000

1000

0
0 50 100 150 200 250 300 350
-1000
Bending moment (kNm)

82 ©Anthony Harding/Jacobs 2022

Moment-axial Moment-Axial

 All points within curve Point at Capacity

Points within
capacity

Axial Force

Moment
Envelope Load case 1 Load case 2

©Anthony Harding/Jacobs 2022

Serviceability
 Process is as per ULS
 Usually set a maximum tensile strain
− Based on crack width strain (tension)
− Based on maximum steel stress (tension)
− Based on maximum compressive stress (usually 0.4-0.6f’c)
 Strain gradient must meet all these criteria
 Compressive stress uses a triangular stress block
 Tensile stresses are based on actual strains
− Triangular concrete stress to fctm,fl
− 0 beyond fctm,fl for plain concrete
− FRC strength beyond fctm,fl for fibre reinforced concrete

84 ©Anthony Harding/Jacobs 2022

Shear
 Shear in section
− Check in accordance with local codes
− Use shear enhancement if required
− Seldom an issue

Shear (friction)
 Shear at joint (friction)

Axial force
− Concrete is very smooth
− Sometimes simpler to reverse-check

85 ©Anthony Harding/Jacobs 2022

Bearing, punching shear

 Bearing, punching shear

− Local concentrated loads
− Check per local codes
− Consider beneficial effects of outside
pressure

86 ©Anthony Harding/Jacobs 2022

Cone pull-out
 Cone pull-out
− Drill and fix anchors
− Bolts and dowels

87 ©Anthony Harding/Jacobs 2022

Deformation checks
 Sources of deformation
− Alignment tolerance
− Build inaccuracy (shape
tolerance)
− Ground loads
− Construction of second
tunnel Spaceproofing
− Construction of cross line/tolerane line
passage
 Need to consider all
sources

88 ©Anthony Harding/Jacobs 2022

 Module 4 Session 6: Critical
sections

89
How do we refine a design? Moment-Axial

 What do we mean by Point at Capacity

‘refine’?
− Make the segment thinner
− Reduce the concrete
strength

Axial Force
− Reduce reinforcement
− Use smaller bolts, dowels,
gaskets
 Where do we focus our Point beyond capacity
effort?

Moment
Envelope Load case 1 Load case 2

©Anthony Harding/Jacobs 2022

Refinement 1: reducing conservative assumptions
Potential impact of
 Is it possible that the design is conservative asumptions
too conservative?
− Inaccuracies in inputs/analysis Capacity
overestimate loads
Predicted load
 Types of conservatism:

Load
−Use of full overburden Actual load
−High ring stiffness

Critical case

Non-critical case
−2D vs 3D effects
−Conservative surcharges
−Low shear resistance at lining
ground interface
−…

91 ©Anthony Harding/Jacobs 2022

Reducing conservatism: example
 TBM tunnel in clay
 6m diameter
 Nearly100m deep
 Originally designed for full overburden
96m
− Maximum load at crown: ~1.9 Mpa
− Maximum TBM face pressure < 1.2MPa
 Solution:
− Allow for 30% relaxation
− Evaluate consolidation load recovery
− Analyse long term loads with long term stiffness
− Check a full overburden case with load factors = 1.0

92 ©Anthony Harding/Jacobs 2022

Refinement 2: potential unsafe designs
Potential impact of
 Is it possible that the design is unconservative assumptions
unsafe?
− Inaccuracies in inputs/analysis Actual load Capacity
underestimate loads
 When we have started with
Predicted load

Load
simple assumptions
 Now we need more accurate

Critical case

Non-critical case
assumptions to be sure the Impact of
design works unconservative
assumptions is of the
same magnitude

93 ©Anthony Harding/Jacobs 2022

Potential unsafe designs: example
 Shallow launch
− Surface excavation
− High degree of variation of ground
loads from crown to invert
− Multiple ground types
− Potential to underestimate bending
moments
 Plaxis model
− Simulate excavation
− Check for multiple ground loss
possibilities
− All results within capacity

94 ©Anthony Harding/Jacobs 2022

Why the focus on critical sections/cases?
 Simple models:  
Case 1 Case 2 Case 3 Case 4
− Simple inputs
− Run quickly

Accuracy/effort
− Results are easily
summarised
 Quickly validated
 Quickly checked
 Easily reported
 Result: lower
design effort

95 ©Anthony Harding/Jacobs 2022

Why the focus on critical sections/cases?
 More refined  
Case 1 Case 2 Case 3 Case 4
models
− More inputs

Accuracy/effort
− More validation 
− More reporting Case 2 Case 4

 More design effort

 Only spend it where 
you have to: Case 4
− Critical sections!

Very high design

effort Case 1 Case 2 Case 3 Case 4

96 ©Anthony Harding/Jacobs 2022

Critical sections
 Typical critical elements: Moment-Axial
− Maximum axial load
− Maximum displacement Max moment
− Maximum moment
− Maximum shear

Axial Force
− Adverse combination
 Typical locations
− Deepest Critical case
− Shallowest
− Softest ground
Moment
− Max water pressure
Envelope
− Mixed face
Load case 1
− Local loads/conditions Load case 2

©Anthony Harding/Jacobs 2022

 Module 4 Exercise

98
Module 4 exercise: Closed form
 Undertake closed form
calculation using either:
− Curtis/Muir Wood
− Duddeck and Erdmann
 Recommend using excel, and
creating columns of data
 Try not to paste data: you may
want to play with the inputs
later
 PLEASE DON’T TRY TO
REPLICATE MY
CALCULATIONS!

99 ©Anthony Harding/Jacobs 2022

Module 4 exercise: Closed form
 Plot moments and axial Moments and axial forces
1800 120.0
forces by chainage
− Axial forces on primary 1600 100.0

− Moments on secondary 1400 80.0

1200 60.0

Axial Froce (kNm/m)

Moment (kNm/m)
1000 40.0

800 20.0

600 0.0

400 -20.0

200 -40.0

0 -60.0
0 200 400 600 800 1000 1200 1400
Chainage

Hoop thrust in lining at axis, Hoop thrust in lining at crown, Maximum bending moment in lining,

100 ©Anthony Harding/Jacobs 2022

Module 4 exercise: Moment-axial
 Pre-populated moment-axial
inputs

101 ©Anthony Harding/Jacobs 2022

Module 4 exercise: Moment-axial
 Enter results into the sheet
 They will automatically plot on
the chart
 Try to link data rather than
paste values
− Allows you to change inputs
and see the changes in output

102 ©Anthony Harding/Jacobs 2022

Module 4 exercise: Moment-axial
 Plot all the results on your Moment-Axial force diagram. Section width: 1000 mm; Section depth: 300 mm; Concrete strength: 50
MPa; Residual flexural strength: 3.0 MPa
moment-axial 9000
Capacity
8000 Min axial
Max Axial
7000
Stacking and handling

6000

5000
Axial froce (kN)
4000

3000

2000

1000

0
0 50 100 150 200 250 300 350
-1000
Bending moment (kNm)

103 ©Anthony Harding/Jacobs 2022