Innovation in geotechnical instrumentation to realize performance based design

Kenichi SogaUniversity of California-Berkeley

Acknowledgement

• GeoVirgina Organizing Committee and Qamar Kazmi

• University of California, Berkeley – PhD students and post-docs

• University of Cambridge, Cambridge Centre for Smart Infrastructure and Construction – PhD students and Researchers

Examples of distributed strain/displacement sensors

Distributed fiber optics - Embedded sensor for life-long monitoring• Fibre optics – 30-100• Fibre optics – 0.03-0.1 mm resolution (for 1 m gauge length)

Computer Vision and LIDAR• Fixed system – 0.1 mm resolution• Not Fixed system – 3-5 mm resolution

WSN – Continuous monitoring at difficult-to-access sites• Wisen• Utterberry – sub millimeter resolution

“Value” as a tool

Value = Product(what it claims it will

do)

Product(what it claims it will

do)+

Confidence in delivery

(do I believe it will work?)

Confidence in delivery

(do I believe it will work?)

“Innovations have more chance of adoption if their benefits are mapped” Dr Keith Bowers, London Underground Limited

118 km from east to west37 stations9 new stations (8 sub-surface)Increase London's rail-network capacity by 10%

Crossrail – New London Underground Line in London

Monitored Section of the Royal Mail TunnelCrossrail – Liverpool Street Station – C510

Monitored SectionCrossrail Platform Tunnel Excavation

Bloomfield Box Excavation

After Crossrail

How Big Is It?

≈ 2.74 m

≈11 m

≈6 m

Constructed: 1917 - 1923

Suspended: 2003Made of cast iron

Fixed support

Tunnel lining

Monitoring area

Camera

Site Overview Camera setup

Low frequency motion

Bending Mode

Shearing Mode

Matthew Wilcock

Arup – M. Devriendt

Bending or Shearing?

(After Attewell, 1986)

Smax

Longitudinal Settlement

Radius of Curvature

Radius of Curvature

2009

2012

2015

Shear LPDT Orientation

Expected settlement profile

New Tunnel

Compressive (-ve) when ring nearest tunnelling approach shifts downward with respect to neighbouring ring!

Data for these sensors above sign reversed in

data processing.

Layout of Sensors & Location of Sensors 11, 34-40

Data presented

2881@ Node 11

2940@ Node 40

29.5m

New Tunne

l

120 Wireless Sensors to monitor movements of 60 joints

Wilcock

Pilot tunnel

Main tunnel

Bond Street – LUL Station

Royal Mail Tunnel

Passenger Adit

Instrumented Section (35m)

Longitudinal crossing Transverse crossingBending > Shearing Shearing > Bending

The two case studies show that the cast-iron tunnels are more tolerable to bending (i.e. can accommodate smaller Radius of Curveture (ROC))

• Distance range ≈10-30km• Readout resolution = 0.05m• Gauge length resolution = 0.2-1m• Strain Resolution = 10-30me

Distributed Sensing providing “Continuous Strain/temperature/vibration Profile” along the fibre optic cable

Frequencyshift 0

RayleighBrillouin Brillouin

RamanRaman

B B

,T,T

T

0

Backscattered spectrum

1. Send light through the fibre and detect the back scattered light2. Identify the location of back scattering from the time interval3. Measure Frequency shift due to strain (ΔF)4. Translate ΔF to strain: ε = f(ΔF)5. Obtain the strain profile: ε(x) over the whole length of the cable

Sensing cables

(a-d) Strain/acoustic sensing(e-f) Temperature sensing

Fujikura ReinforcedFibre Optic CableRobustness

27

CH5 CP2CP1

Sprayed concrete liningPerformance during cross passage opening

Tunnel liningBending and axial performance during construction and in long term

-1000.0

0.0

1000.0

2000.0

3000.0

4000.0

5000.0

6000.0

7000.0

8000.0

9000.0

0.0 100.0 200.0 300.0

Axia

l For

ce (k

N/m

)

Bending Moment (kNm/m)

Capacity limit

Muir Wood Full Slip

Muir Wood FullBondDuddeck Full Slip

Duddeck Full Bond

Joint check - MaxloadJoint check - MaxdeflMuir Wood Full Slip+point loadsMuir Wood FullBond + point loadDuddeck Full Slip +point loadDuddeck Full Bond+ point loadSegment 1

Segment 2

In January this year, New York Governor Cuomo announced a plan to prevent the 15-month-long L-train shutdown set to begin in April.

The Canarsie Tunnel, which opened in 1924, has shown deterioration after flooding from Hurricane Sandy in 2012.

A pile loading test…

A building construction at the Isle of Dog, London

Loizos Pelecanos Duncan Nicholson

River Terrace DepositsLambeth Group

Thanet Sand

Chalk

Made Ground / Alluvium

Large diameter piles - 2.4m dia.Very deep - avoid tunnels ( 25m into Chalk)

LU tunnel

Thick raft span over tunnels

LU tunnel

Musa Chunge Cedric Kechavarzi Vivien KwanEcho Ouyang

www.loadtest.com

1.5m

Made Ground

Thanet Sand

River Terrace DepositsLambeth Group

Alluvium

Chalk

Sister Bars

Osterberg Cells

2 fibre optic cables

51m

6m

7m

Level 1

Level 2

Level 3

Level 4

Level 5

Level 6

• Diameter = 1.5m• Length = 51m• Osterberg‐cell• Load up to 31MN

No disturbance to actual construction operations

   

Conventional Strain Gauge System

Distributed FO system

Layer 1

Layer 2

O-Cell

PileLoad distribution

CompressionFriction

DownUp

Mechanism of O-cell testing

• Strain gauges

-40

-35

-30

-25

-20

-15

-10

-5

0

50 200 400 600 800 1000

LEVE

L (m

OD

)

STRAIN ()

Strain gauges

• Strain gauges

• Extensometers

• DFOS

-1000-800-600-400-2000-40

-35

-30

-25

-20

-15

-10

-5

0

50 200 400 600 800 1000

CHANGE IN CABLE STRAIN ()

LEVE

L (m

OD

)

STRAIN ()

Strain gauges

DFOS data

DFOS trend

Axial strain Instrument comparison

Axial strain Different FO cable comparison

0

5

10

15

20

25

30

35

40

45

50

-800 -600 -400 -200 0

Dep

th, z

[m]

Axial strain, ε_a [με]

P = 25.66MN

Cable S-3-1

Axial strain Different FO cable comparison

0

5

10

15

20

25

30

35

40

45

50

-800 -600 -400 -200 0

Dep

th, z

[m]

Axial strain, ε_a [με]

P = 25.66MN

Cable S-3-1

Cable S-3-2

Axial strain Different FO cable comparison

0

5

10

15

20

25

30

35

40

45

50

-800 -600 -400 -200 0

Dep

th, z

[m]

Axial strain, ε_a [με]

P = 25.66MN

Cable S-3-1

Cable S-3-2

Cable S-4-1

Axial strain Different FO cable comparison

0

5

10

15

20

25

30

35

40

45

50

-800 -600 -400 -200 0

Dep

th, z

[m]

Axial strain, ε_a [με]

P = 25.66MN

Cable S-3-1Cable S-3-2Cable S-4-1Cable S-4-2

-40

-35

-30

-25

-20

-15

-10

-5

0

5

10

0 0.01 0.02 0.03 0.04

Elev

atio

n [m

]

Vertical displacement, u [m]

Vertical Displacement Profiles

Load (MN)

Dis

plac

emen

t (m

m)

Load (MN)

Load (MN)

Dis

plac

emen

t (m

m)

Load (MN)

Load (MN)

Dis

plac

emen

t (m

m)

Load (MN)

Top of O-cells

Level 2

Level 3

Level 4

Level 5

Level 6(highest)

Fibre optic integrated strain

Extensometers

Displacements are relative to the top of the pile

Small displacement

Very small displacement

Top of O-Cell

Level 2

Level 3

Level 4

Level 5

Level 6

Extensometer

Construction can be challenging alignment concrete quality and

placement soil collapse

Visible inspection not possible

Repair and rework is very difficult

Not all anomalies are defects/detrimental

FHWA‐NHI‐10‐0161.

PROBLEMS WITH PILE CONSTRUCTION

SoilPile

Find the pile radius which match the temperature profile (20 x 4 x 50 = 4000 data sets)

Source of concrete heating

?

Rui Yi Cedric Kechvarzi

1

2

3

4

0

10

20

30

40

50

1 1.5

Dep

th(m

)

Diameter(m)

Fibre optics

Potential for Whole-life Management?Construction Quality Control Real Loading Performance

Future Proofing (EQs, nearby constructions..)

Deployment of Post Grouting Technique to improve Drilled Shaft End-Bearing Resistance

Tina Schwamb

Thames Tunnel

Abbey Mills Pumping station(Captures 39 million tonnes of sewage a year)

Beckton sewage treatment plant

(Schwamb et al., 2014)

Made Ground

River Terrace DepositsAlluvium

London Clay

Lambeth Group

Thanet Sand

Chalk

70m

30m• Abbey Mills shaft:

– 30m ext. diameter– 70m deep

• Diaphragm walls:– 20 panels– 1.2m thick– 84m deep

Shaft F Details

Schwamb et al. 2013

Clough and O’Rourke (1990)Peck (1969)

Circular shafts“plane strain” shafts

0

5

10

15

20

25

30

35

40

45

0 20 40 60 80

Settl

emen

t [m

m]

Distance from shaft wall [m]

New & Bowers (1994)Study at Heathrow shaft ofD = 11 m, H = 26 m

H = shaft depthd = distance from shaftS = settlement = 6 x 10‐4 (empirical factor)

0

2

4

0 5 10

Measured settlement:dwall constructionexcavation

New & BowersH = 70 m = 6 x 10‐4  (0.06%) S/H < 0.005%!

• Reference Design PLAXIS analysis – 7mm

• Reference Design – modified New & Bowers – 13mm

• Hard to justify approach taken without any empirical data

• Potential cost and risk implications Tideway Tunnel

Original design considerations

Bendingstrain

5.2mm

1.2mm

6.1mm

Strain cable

Temperature cable

Diaphragm walls Joints between panels Reduced cirumferential

stiffness???

Monitoring Scheme

Bendingstrain

Made Ground

River Terrace Dep.Alluvium

London Clay

Lambeth Group

Thanet Sand

Chalk

-1 0 1 -1 0 1 -1 0 1 -1 0 1

-2 0 2

0

10

20

30

40

50

60

70

80

-2 0 2

0

10

20

30

40

50

60

70

80

-2 0 2

0

10

20

30

40

50

60

70

80

-2 0 2

0

10

20

30

40

50

60

70

80

Dep

th [m

bgl]

Inc. Curvature [10-4 m-1]

Inc. BM [MNm]

Fibre OpticPanel P15Panel P20PLAXIS

150Diaphragm wall

6mbgl

24mbgl

Fully constrained / No drainage

Not constrained / Drainage

14.45

2D axisymmetric mode

Mohr‐Coulomb soil model

Wall: concrete C50/60 E = 37 GPa(short‐term) thickness = 1.2 m

Wall installation effects are considered with reduced K0values (WIP)

For comparison use serviceability limit state (SLS) results

]

-1 0 1-90

-80

-70

-60

-50

-40

-30

-20

-10

0

-1 0 1-90

-80

-70

-60

-50

-40

-30

-20

-10

0

-1 0 1-90

-80

-70

-60

-50

-40

-30

-20

-10

0

-1 0 1-90

-80

-70

-60

-50

-40

-30

-20

-10

0-0.5 0 0.5

Dep

th [m

bgl]

Inc. Curvature [10-4 m-1]

Fibre OpticPanel P15Panel P20PLAXIS

-0.5 0 0.5 -0.5 0 0.5 -0.5 0 0.5 Comparison:FLAC (MC)FLAC (Adv)FO P15FO P 20

Other soils above Chalk

Chalk

FLAC modelSoil models

***************************

Elements yield in shear

Mohr Coulomb for CHALK:K = 1000 MPa | G =  600 MPa | E = 1500 MPa ’ = 35ᵒ | c’ = 20 kPa

Other soils above Chalk

Chalk

Hoek‐Brown model for CHALK:K = 1000 MPa | G =  600 MPa | E = 1500 MPaAdditional Hoek‐Brown constants

Elements elastic

FLAC modelSoil models

-5 0 5-90

-80

-70

-60

-50

-40

-30

-20

-10

0

-5 0 5-90

-80

-70

-60

-50

-40

-30

-20

-10

0

-5 0 5-90

-80

-70

-60

-50

-40

-30

-20

-10

0

-5 0 5-90

-80

-70

-60

-50

-40

-30

-20

-10

0

-5 0 5-90

-80

-70

-60

-50

-40

-30

-20

-10

0

Dep

th [m

bgl]

Lateral wall movement [mm]

Comparison:FLAC (MC)FLAC (Adv)Inc P10Inc P 20

o Low cohesion = 20 kPa in chalk to be on ‘safe side’

o For the wall this had the opposite effect:o Stiffer soil causes sharper bulge o Shaper bulge means more 

localised bending

FLAC modelWall stiffness

-2 0 2-90

-80

-70

-60

-50

-40

-30

-20

-10

0

-2 0 2-90

-80

-70

-60

-50

-40

-30

-20

-10

0

-2 0 2-90

-80

-70

-60

-50

-40

-30

-20

-10

0

-2 0 2-90

-80

-70

-60

-50

-40

-30

-20

-10

0-1 0 1

Dep

th [m

bgl]

Inc. Curvature [10-4 m-1]

Inc. BM [MNm]-1 0 1 -1 0 1 -1 0 1

Comparison:FLAC (37 | 37)FLAC (26 | 26)FLAC (37 | 18.5)FLAC (37 |  3.7)FO P15FO P 20

Stiffness in vertical and radial direction [GPa]Stiffness in circumferential direction [GPa]

FLAC (37 | 3.7) overpredicts here

FLAC (37 | 3.7) is approx. right here

Northern Line Extension Project

Thermal Integrity FO monitoring of 100 Panels and 74 Piles, replacing other methods such as sonic logging.

Prof Tom O’RourkeDr Brad Wham

Elizabeth tunnelLos Angeles Department of Water and Power

American River Levee Upgrade Project

• Sacramento Metropolitan area remains one of the most at risk areas for flooding in the United States.

• Levees constructed in the previous flood control project (1850-1950), Sacramento River Flood Control Project, were constructed of poor materials

• Flows in either the American or Sacramento Rivers will probably stress the network of levees to the point of failure.

FO Monitoring of cement bentonite cut-off wall, currently upgraded.

US Army Corps of Engineers

51 km Smart TunnelSingapore’s Deep Tunnel Sewerage System (DTSS)

Smart Road Corridors by Meso-Scale In-PavementDistributed Infrastructure Sensing

Alicia

Best Practice Guides for Monitoring Civil Infrastructure

Distributed Fibre Optics Sensing Wireless Sensor Networks

Summary• Innovation in sensors as part of Internet of Things

– Exciting opportunities for Geotechnical Engineering to understand the real performance of infrastructure and construction.

• For example, distributed fibre optics (especially embedded) can give useful strain data that no other sensors can give.

• Monitoring system should be an integral part of the construction package– Quality Control– Maintenance– Reuse

• This leads to Performance-design, construction and maintenance.

Thank you