An Optical Fiber Sensor for Corrosion Detection of

Reinforced Concrete Structures

Presentation at HKCI Annual Seminar 2016

Christopher K.Y. Leung

Department of Civil and Environmental Engineering

Hong Kong University of Science and Engineering

Kai Tai Wan

Department of Mechanical, Aerospace and Civil Engineering

Brunel University, UK

Introduction Steel Corrosion is a Major Problem with Reinforced

Concrete Structures

Early Detection of Potential Corrosion Enables Proper Maintenance to Extend the Life of Structures

Corroded Beam and Column Corroded Slab

Initiation of Steel Corrosion in Concrete

Iron oxide formed on steel surface is initially stable in the alkaline environment of concrete

The stable layer protects underlying material from corroding, so the corrosion rate is Negligible

The Protection is Lost when

– pH at steel surface drops below 11 (normally due to the penetration of Carbon Dioxide)

– A Critical Chloride Concentration is reached at the Steel

Steel corrosion is then Greatly Accelerated

NO Corrosion

Concrete Surface

Front of CO2

or Chloride

Steel

Significant Corrosion

Assessing the Penetration of Corrosive Agents

A Core can be taken from the Concrete Structure

To assess the penetration of CO2, pH indicator can be applied on the lateral surface to find the location of color change

To assess Chloride penetration

– The Core is cut into slices, which are ground into powder

– Tiltration is performed to find the chloride content in the powder

– Chloride concentration vs depth can hence be obtained

Disadvantages

– Time Consuming and Costly

– Results Do NOT Directly Reflect Corrosion Activities

– For CO2, pH indicator changes color at pH=9 while steel starts to corrode at pH < 11

– For Chloride Initiated Corrosion, the Critical Chloride Content decreases by an order of magnitude when pH changes from 13.5 to 12.5

In-situ Corrosion Assessment with the

Corrosion ‘Ladder’

Developed by Raupach and Schiessl at Aachan University

By measuring the corrosion current between each steel bar on the ‘ladder’ and a reference, the corrosion activity at different distances from the surface can be assessed

These bars start

to Corrode

Front of Critical

Chloride Content

The Corrosion

Ladder Steel Re-bar

The Corrosion Ladder is a very Robust ‘Instrument’ and has been used in many Real-World Projects

Limitations:

– Not Applicable to Existing Structures

– Relatively Costly

A version for Installation in Existing Structures has been developed but sealing is a problem

A Novel Corrosion Detection Concept

Optical Fiber

Thin Iron

Coating With sufficient coating thickness,

Most of the Light is reflected

Coating Thickness

REDUCED BY CORROSION

Drop in Intensity

of Reflected Light Light Loss

at Fiber End

The Coating can be Produced by the Sputtering Process, with many Fibers coated at the same time – Cost of Each Sensor can be Low

Retrofitting of Sensor into a Structure is Possible

The Sputtering Process

Physical Principle Coating of Optical Fibers

Iron Target

Bundle of Fiber (Stripped at the End)

Coated Optical Fibers

Fiber

Holder

0

100

200

300

400

500

0 50 100 150 200F

ilm

th

ick

nes

s (n

m)

Time (minutes)

Sputtering process

Series1

P=80W,2.92nm/min

Measurements on Sputtered

Substrate and Fiber

Film Thickness vs

Sputtering Time

- Measured from Glass

Substrate with the

Profilometer

Reflectivity vs

Film Thickness

- Measured from Coated

Optical Fibers in Air

with OTDR

Power Measurement with OTDR

OTDR Stands for the Optical Time Domain Reflectometer

Measurement Principle – An optical pulse is sent forward and the reflected signal monitored

as a function of time

– Reflected Power at a particular point can be obtained through comparison with the signal level at the ‘floor’

The measurement is hence insensitive to fluctuation of source powder and loss at connections – Key to reliable measurements based on Power Intensity

Time (Distance)

Ref

lect

ed P

ow

er Loss Obtained from Difference

Between Reflected Peak and ‘Floor’

A Typical OTDR

Result

Tests on Bare Coated Fibers (1)

Coated film corrodes under the presence of chlorides

and the response is dependent on chloride concentration

Fibers are put inside Salt Solution of different chloride

Concentrations

-7

-6

-5

-4

-3

-2

-1

0

1

0 5 10 15 20 25

Number of Days in the Solution

Po

wer

Lev

el

(dB

)

0.1 mol/L

1 mol/L

0.5 mol/L

Tests on Bare Coated Fibers (II)

-0.5

0

0.5

1

1.5

2

2.5

0 20 40 60 80 100 120 140

Days after Embedment

dB

Lo

ss in

Lig

ht

Po

wer

Wet

Wet

Dry

Dry

Sensor can detect corrosion and distinguish between

different corrosive environments

Fibers are embedded inside Mortar Blocks containing

salt, which are either placed in Water or left in Air

Packaging Method of the Sensor

Plastic tube

Fiber with coated end

Connector to optical system

Protecting tube

1. Coated

fiber (sensor)

and plastic

tube

2. Sensor

inserted

into the

plastic tube

3. Sensor

fusion-

spliced to

connector

4. Protecting

tube appled

over bare

fiber

Coated Fiber

Detection of Chloride Penetration

Test Set Up Specimen (Side View)

Cement Mortar

W:C:S=0.6:1:2

Fiber sensor

15mm

10mm

5mm

All Lateral Sides

sealed with Epoxy

OTDR

3% NaCl

Solution

Container Specimen

Fiber sensor

Results for Chloride Penetration Test

-25

-20

-15

-10

-5

0

5

0 20 40 60 80 100

Pow

er L

oss

(d

B)

Time (Days)

Diffusion Test

5mm

10mm

15mm

Corrosion starts at sensor closest to the surface exposed

to chlorides

Performance of Sensor in a Drilled Hole

Experiment Set Up Specimen preparation

OTDR

Sodium Chloride

Solution

Fiber Sensor

Cement Mortar

W:C:S=0.6:1:2

Fiber sensor

New cement mortar

W:C:S=0.6:1:2

Relative chloride

concentration

Results for Sensors installed in the

Holes of Mortar Blocks

-45

-40

-35

-30

-25

-20

-15

-10

-5

0

0 10 20 30 40 50 60 70 80 90 100

Time (Day)

Re

flectiv

ity (

dB

)

0.4% NaCl

1% NaCl

2% NaCl

Uncoated fiber end in air

Uncoated fiber end in water

Feasibility for Post-installation of Sensor is Demonstrated

Sensor Output vs corrosion current

Experimental Set-Up

(modified ASTM G109)

Fiber Sensor

Ground clamp

Steel bars

3% NaCl Solution

V R

R=100Ω

Adhesive tape

10mm

Corrosion Current =V/R

-30

-20

-10

0

10

20

30

0 5 10 15 20 25 30 35 40 45 50 55 60 65 70Time (Day)

Re

fle

ctivity (

dB

)

Reflectivity of Optical Sensor

Macrocell Current

Ma

cro

ce

ll C

urr

en

t (µ

A)

Reflectivity of uncoated fiber in air

Reflectivity of uncoated fiber in water

Threshold of

macrocell current

Results (1)

Corrosion Initiation detected by current measurements

6 days before the optical sensor

Results (2)

-50

-40

-30

-20

-10

0

10

20

30

0 10 20 30 40 50 60 70 80

Time (Day)

Refle

ctiv

ity (

dB

)

Reflectivity of Optical Sensor

Macrocell Current

Macro

cell

Cu

rre

nt (μA)

Threshold of macrocell current

Reflectivity of uncoated fiber in air

Reflectivity of uncoated fiber in water

- Corrosion Initiation detected by optical sensors 16 days

before the current measurement

- Different is not significant for practical long-term monitoring

Sensor Installation in Concrete Structures

New Structures (angled sensor)

Concrete Surface Bottom of Sensor

Glued to Surface

of Steel Bar

Steel Reinforcement

Fiber

Sensors

Fiber corrosion sensor

Sensor head

Flat part glued to steel

Fixture

Inclined part

Sensor head

Fixture end

Sensor Outlet

Sensor Encased in GFRP

Existing Structures

Steel Reinforcements

Steel

Reinforcements

Optical

Fiber

Sensor Hole Drilled

from inside

of Box Section

Optical Fiber Sensor

in the Drilled Hole

Hole Fully

Sealed on

the side

Chloride Penetration from the

Lower Surface to be Monitored

Box Section Solid Section

Field Trial in an Existing Structure

10 sensors installed on two piers of a

footbridge right next to the sea

Sensors are installed

in holes drilled on

these surfaces

Locations of the Sensors

Sensor #4

60mm from back surface

40mm depth

SEA Sensor #5

300mm from surface

100mm depth

Sensor #2

80mm from surface

40mm depth

Sensor #1

60mm from surface

40mm depth

Sensor #3

100mm from surface

40mm depth

5 sensors installed in each pier

Installation Procedure (1)

Locating of Steel Reinforcements

Cleaning of Hole with

Compressed Air

Drilling of Hole for the Sensor

Installation Procedure (2)

Sealing of Hole

Surface with

Epoxy Coating

Sensors Placed inside the Holes Filling of hole with Cement Grout

The sensors have survived THREE big typhoons since their installation

OTDR Readings

0

5

10

15

20

25

30

35

40

45

50

80 130 180 230

Distance from the Bulkhead of OTDR (m)

Inte

nsity o

f th

e R

efle

cte

d S

ign

al (d

B)

13-Aug-0815-Aug-0819-Aug-0817-Sep-0825-Sep-0829-Oct-08

Sensor 1Sensor 5

Sensor 2 Sensor 4

Sensor 3Reference

The loss for each sensor is measured at different times

The reference is a fiber embedded inside an epoxy block

Signal Loss vs Time

Results appear to correlate well with location of sensors

-30

-25

-20

-15

-10

-5

0

0 50 100 150 200 250 300 350Number of Days

Reflectivity (

dB

)

Sensor 06

Sensor 07

Sensor 08

Sensor 09

Sensor 10Reflectivity of uncoated fiber in air

Reflectivity of uncoated fiber in water Sensor 1

Sensor 2

Sensor 3

Sensor 4

Sensor 5

1 2 5 4

3

Correlation with Chloride Content in Pier

Powder taken around the 3 sensors on the seaward side (at 25-40mm depth) and also around the sensor in the middle of the pier (at 85-100mm depth)

Chloride content obtained through tiltration with silver nitrate solution

Location Chloride Content (%wt of cement)

Near the Sea 0.314

Center of Pier 0.214

Results seem to be reasonable as chloride threshold is

normally between 0.2 to 0.4% cement weight

sensors

Powder taken

Within these regions

SEA

Conclusion

A New Corrosion Sensor based on OTDR Measurement of Reflected Light Intensity is developed together with a simple packaging technique

The Sensor can be mass produced at low cost and is suitable for both new and existing concrete structures

Laboratory Results show that

– The sensor can monitor corrosion penetration

– The sensor can be post-installed into an existing structure

– There is reasonable agreement between sensor measurement and macro-cell current

Site Trial indicates robustness of sensor and reasonable correlation between measured corrosion and sensor location

Through this investigation, the potential of a novel corrosion sensor is demonstrated