micromachines
Article
A Spiral Distributed Monitoring Method for Steel
Rebar Corrosion
Jianzhi Li 1, * , Yiyao Zhao 2 and Junjie Wang 2






Citation: Li, J.; Zhao, Y.; Wang, J. A
Spiral Distributed Monitoring
Method for Steel Rebar Corrosion.
Micromachines 2021, 12, 1451.
https://doi.org/10.3390/mi12121451
Academic Editor: Limin Xiao
Received: 1 November 2021
Accepted: 21 November 2021
Published: 26 November 2021
Publisher’s Note: MDPI stays neutral
with regard to jurisdictional claims in
published maps and institutional affil-
iations.
Copyright: © 2021 by the authors.
Licensee MDPI, Basel, Switzerland.
This article is an open access article
distributed under the terms and
conditions of the Creative Commons
Attribution (CC BY) license (https://
creativecommons.org/licenses/by/
4.0/).
Micromachines 2021, 12, 1451. https://doi.org/10.3390/mi12121451
1 Structure Health Monitoring and Control Institute, Shijiazhuang Tiedao University,
Shijiazhuang 050043, China
2 School of Materials, Shijiazhuang Tiedao University, Shijiazhuang 050043, China;
zhaoyiyao1995@163.com (Y.Z.); junjiewang315@163.com (J.W.)
* Correspondence: lijianzhigang@163.com
Abstract: This paper proposes a novel spiral-wound, optic-fiber sensor to monitor the corrosion
of steel bars. At the same time, the winding parameters, such as winding angle and pitch, were
first theoretically deduced. Then, to decrease light loss, a practically distributed sensor wound onto
the protective mortar layer was developed by increasing the winding curvature radius. The spiral
distributed sensors were experimentally verified for their feasibility. Experimental results showed
that the spiral fiber strain depended on the thickness of the protective mortar layer. Furthermore,
the spiral distributed strain well reflected the cracking process of concrete. In addition, the concrete
cracking time depended on the thickness of the protective concrete layer. Accordingly, this method is
feasible for evaluating the initial and final cracking behaviors of concrete structures and provides a
sight for steel bar corrosion.
Keywords: spiral wound fiber; corrosion monitoring; Brillouin distributed sensor; reinforced concrete
1. Introduction
Reinforced concrete structures are widely used in civil engineering applications for
their durability and fire resistance. However, most reinforced concrete structures fail to
meet their engineering requirements owing to insufficient durability [1]. This is mainly
attributable to the damage resulting from the corrosion of steel bars, further affecting the
concrete structures [2,3]. Over the past few decades, theoretical analysis of the reinforced
concrete structure damage induced by corrosion [4–10] has shown that the corrosion process
is divided into three stages: the corrosion-expansion stage, the protective-layer-cracking
stage, and the crack-extension stage. Meanwhile, many techniques have been proposed
and developed to monitor steel corrosion in concrete structures [11–13]. These are mainly
divided into two categories: the first category includes some traditional electrochemical
monitoring methods, such as the half-cell potential method, the constant-current method,
the AC-impedance method, the electrochemical-noise method, and the concrete-resistance
method [14]. The indispensable reference electrode [15,16] is vulnerable, owing to its
time-varying characteristics. Therefore, steel corrosion cannot be accurately monitored at
any time during long-term service. A second category comprises corrosion-monitoring
methods based on optical fiber sensors. Optical-fiber sensing technology has attracted
extensive attention, owing to its advantages, such as the use of lightweight fibers, high
measurement accuracy, strong anti-electromagnetic interference capability, and corrosion
resistance of the fibers. Consequently, it has been gradually applied to monitoring steel
corrosion [17–24]. However, the corrosion sensors used in the optical fibers can only
monitor corrosion at certain points and fail to measure corrosion at any location along a
steel bar.
Hence, this study developed a spiral distributed fiber sensor to monitor the corrosion
of steel bars in concrete structures. Compared to the existing distributed optical-fiber
https://www.mdpi.com/journal/micromachines
Micromachines 2021, 12, x FOR PEER REVIEW 2 of 11

Micromachines 2021, 12, 1451 2 of 11

Hence, this study developed a spiral distributed fiber sensor to monitor the corrosion
of steel bars in concrete structures. Compared to the existing distributed optical-fiber
sensing methods, the novelty of the spiral distributed fiber lies in increasing the corrosion-
sensing methods, the novelty of the spiral distributed fiber lies in increasing the corrosion-
mass-loss tested range and lengthening the tested distributed fiber sensor, which is
mass-loss tested range and lengthening the tested distributed fiber sensor, which is attribute
attribute to the reduced light loss of the distributed sensor. Furthermore, the desired
to the reduced light loss of the distributed sensor. Furthermore, the desired winding
winding parameters were first achieved by theoretical analysis and experimentally
parameters were first achieved by theoretical analysis and experimentally verified. Thus, a
verified. Thus, a spiral distributed method was utilized to measure steel corrosion in the
spiral distributed method was utilized to measure steel corrosion in the long-term service
long-term service of rebar.
of rebar.

2. Materials and Methods

2. Materials and Methods
2.1. Effect of Winding
2.1. Effect Radius
of Winding on theon
Radius Bending Curvature
the Bending of Optical
Curvature Fiber Fiber
of Optical
The spiral distributed
The spiral winding
distributed fiber was
winding fiberunrolled, as shown
was unrolled, in Figure
as shown 1. Here,
in Figure α is α is
1. Here,
the helix angle, which represents the rising angle of the winding fiber, β
the helix angle, which represents the rising angle of the winding fiber, while , called the the
while β, called
winding
winding angleangle of distributed
of distributed fiber,fiber,
is theiscomplementary
the complementaryangleangle
of αof
. α.

FigureFigure
1. Spiral expansion.
1. Spiral expansion.

The equations of theofspiral

The equations curvecurve
the spiral are expressed as follows:
are expressed as follows:
x = r cos θ
x = r cos θ
y = r sin θy = r sin θ
L
(1) (1)
z =L± 2π θ = ±rθ cot α
z =  θ =  rθ cot α
2π
where r is the radius of the steel bar (cm), L is the spiral distribution pitch (cm), α is the
where r is the
winding radius
angle, andofθ the steel
is the bar angle,
center (cm), Lθ is ωt. spiral distribution pitch (cm), α is
= the
the winding
Theangle, and θ of
dependence is the
the center
bending angle, θ = ωt ρ. of the spiral distributed fiber upon the
curvature
The dependence
spiral of theand
winding radius bending
winding pitch is ρillustrated
curvature of the spiral distributed
as follows [25]: fiber upon the
spiral winding radius and winding pitch is illustrated as follows [25]:
2π
ρ = q2 π (2)
ρ= (2πr ) 2
+ L 2 (2)
(2 πr )2 + L2
Simultaneously, the reciprocal of the bending curvature ρ is the bending radius of a
Simultaneously, the reciprocal of the bending curvature ρ is the bending radius of a
spiral distributed sensor. Moreover, the bend losses of fiber depend on the bending radius.
spiral distributed sensor. Moreover, the bend losses of fiber depend on the bending radius.
The bend losses of fiber exponentially fall with an increasing radius and ultimately keep
The bend losses of fiber exponentially fall with an increasing radius and ultimately keep
stability [26]. Accordingly, to decrease the light loss induced by fiber bend, the minimal
stability [26]. Accordingly, to decrease the light loss induced by fiber bend, the minimal
bending radius was achieved in this study.
bending radius was achieved in this study.
Figure 2 depicts the relationship between the spiral winding radius, the spiral dis-
Figure 2 depicts the relationship between the spiral winding radius, the spiral
tribution pitch, and the curvature of the fiber. The curvature of spiral winding decreases
distribution pitch, and the curvature of the fiber. The curvature of spiral winding
with the increasing winding radius when the winding pitch is constant. In addition, the
decreases with the increasing winding radius when the winding pitch is constant. In
curvature also depends on the winding pitch of the distributed fiber. When the winding
radius is unchanged, the fiber curvature gradually decreases with the increasing fiber
Micromachines 2021, 12, x FOR PEER REVIEW 3 of 11
Micromachines 2021, 12, x FOR PEER REVIEW 3 of 11
Micromachines 2021, 12, 1451 3 of 11

addition, the curvature also depends on the winding pitch of the distributed fiber. When
addition,
the windingthe curvature
radius is also depends the
unchanged, on the winding
fiber pitchgradually
curvature of the distributed
decreases fiber.
withWhenthe
the
pitch.winding radius is
The above-mentioned
increasing unchanged, the fiber curvature
phenomena are clearly
fiber pitch. The above-mentioned gradually
shownare
phenomena in Figure decreases
clearly 3. with
Accordingly,
shown in Figure the
the3.
increasing
increasing
Accordingly, fiber
fiber pitch.
thewindingThe
increasingabove-mentioned
pitch canwinding
fiber phenomena
not onlypitch
decrease are decrease
the only
can not light clearly
loss, butshown
also in Figure
lengthen
the light loss, the3.
but
Accordingly,
tested steel the
bar. increasing fiber
also lengthen the tested steel bar. winding pitch can not only decrease the light loss, but
also lengthen the tested steel bar.

Figure 2.
Figure 2. Curvature
Curvature versus
versus radius
radius of
of rebar.
rebar.
Figure 2. Curvature versus radius of rebar.

Figure 3. Curvature versus fiber pitch.

Figure 3.
Figure 3. Curvature
Curvature versus
versus fiber
fiber pitch.
pitch.
2.2. Effect of Spiral Angle on a Spiral Distributed Sensor
2.2.
2.2. Effect
Effect of
of Spiral
Spiral Angle
Angle on
on aa Spiral
Spiral Distributed
Distributed Sensor
Sensor
The dependence of the spiral stress σ upon the principal stress σ x and the winding
The dependence of the spiral stress σ upon
σ upon
the
the principal stress σσx and the winding
angleThe
angle
dependence
ααof of
thethe
spirally
of the
spirally spiral stress
distributed
distributed fiber
fiber areare expressed
expressed
principal
as:as:
stress x and the winding
angle α of the spirally distributed fiber are expressed as:
  π 
σ =σσ=x cos 2 2 β = σ cos
σ x cos
β2= σx cos
x
2 2π
 π−−αα  (3)
(3)
σ = σ x cos β = σ x cos 2 2 2 − α  (3)
2 
Equation
Equation (3) (3) indicates
indicates that
that the
the fiber
fiber stress
stress can
can be
be decreased
decreased due
due toto the
the helix
helix angle.
angle.
Thus, the distributed fiber withstands the greater corrosion expansion stress,helix
Thus,Equation
the (3) indicates
distributed fiber that the
withstands fiber
thestress
greatercan be decreased
corrosion due
expansion to the
stress, angle.
compared
compared
Thus,
with the distributed fiber withstands the greater corrosion expansion
with the ring distributed sensor. Therefore, the spiral method improves the corrosiontest
the ring distributed sensor. Therefore, the spiral method improves stress,
the compared
corrosion test
range
range of
with theof the
ringsteel
the bar.
distributed
steel bar. sensor. Therefore, the spiral method improves the corrosion test
rangeThe
Theof sensor
the specimen
steel
sensor bar.
specimen was
was fabricated
fabricated using
using aa carbon
carbon steel
steel bar
bar with
with aa diameter
diameter of of 16
16 mm.
mm.
Based on
The Equation
sensor (2),
specimen the bending
was curvature
fabricated using of
a the
carbonhelical
steelfiber
bar was
with calculated
a
Based on Equation (2), the bending curvature of the helical fiber was calculated based diameter based
of on
16 mm.
on
the different
Based winding diameters, helix angles, and winding pitches, as shown
the different winding diameters, helix angles, and winding pitches, as shown in Table 1. on
on Equation (2), the bending curvature of the helical fiber was calculatedin Table
based 1.
the different winding diameters, helix angles, and winding pitches, as shown in Table 1.
Micromachines 2021, 12, 1451 4 of 11

Table 1. Structural parameters of spiral distributed sensor.

Bending Curvature
Diameter (mm) Helix Angle (◦ ) Winding Pitch (cm)
(cm−1 )
16 60 8.7 0.625
16 55 7.1 0.719
16 45 5.0 0.885
16 30 2.9 1.087
26 30 4.7 0.667
32 30 5.8 0.541
36 30 6.5 0.463

Using the Brillouin optical time-domain analysis (BOTDA), the following results were
concluded, based on the above theoretical calculation and analysis:
(1) The curvature of the winding fiber decreased with the increase in the helix angle and
the diameter of the steel bar. At the same time, we concluded that the lower the helix
angle and the spiral pitch were, the greater the curvature of the winding fiber was.
(2) (When the winding diameter was 16 mm, the optical signals of the distributed sensing
fiber with a 60◦ helix angle were obviously better than those with helix angles of 30◦ ,
45◦ , and 55◦ ; then, the light loss of the distributed sensor with a 60◦ helix angle was
less than that of the distributed sensor with the helix angles of 30◦ , 45◦ , and 55◦ . The
bending curvature of the fiber should preferably be less than 0.625 cm−1 . Then, the
corresponding bending radius of a spiral distributed sensor is approximately 1.6 cm.
The corresponding macro-bending loss is in a lower range [26], which has basically
no effect on a spiral distributed sensor.
(3) The mortar layers of the steel bar were casted. Their thicknesses were 5, 8, and
10 mm, respectively. Then, the corresponding winding radii of the same helix angle
distributed fiber (30◦ ) were 26, 32, and 36 mm, respectively. Moreover, the bending
curvature of the optical fiber can be calculated, as shown in Table 1. The bending
curvature of less than 0.667 cm−1 is desirable.
In summary, the spiral method improves the corrosion rate test range of steel bar. The
desired curvature of the winding fiber (0.667 cm−1 ) is a critical parameter for the distributed
sensor, which contributes to not only decreasing light loss, but also to improving strain
accuracy of the distributed sensor. At the same time, the fiber curvature is difficult to
intuitively measure. We transformed the fiber curvature into the helix angle and spiral
pitch was experimentally measured. Hence, to decrease the fiber light loss, the practically
distributed sensor wound on the protective mortar layer was proposed.

3. Experiment
A grouting experiment was carried out with a cement grouting gun. The mixing ratio
of water, cement, fine aggregate, and water-reducing agent was 210:500:1000:3 in weight.
The specimens with the above-mentioned thicknesses of the mortar layers were named 5 #,
8 #, and 10 # specimens, respectively. The schematic diagram of the distributed sensor is
shown in detail in Figure 4. The three steel bar corrosion sensors were made and embedded
in the concrete specimens.
Finally, the electric accelerated corrosion experiments were conducted. Before the steel
bar was energized, the outer optical fiber with a spiral distribution was fixed on the mortar,
as shown in Figure 5. Simultaneously, to protect the bare 2 cm steel bars at the bottom of
the specimen from corrosion, the bottom surface of the mortar specimen was also sealed
with AB epoxy glue. Subsequently, the mortar specimen was completely immersed in
a water tank containing 5% NaCl solution and started to be corroded. There were three
identical HY3005MT digital DC power suppliers used in the accelerated corrosion testing.
The power suppliers have a steady current of 0.1 A. It was assumed that the corrosion of
rebar is uniform. Meanwhile, the distributed fiber strains of 5 #, 8 #, and 10 # specimens
Micromachines 2021, 12, 1451 5 of 11

Figure 4. Schematic of the specimen.

Micromachines 2021, 12, x FOR PEER REVIEW 5 of 11
were recorded by the BOTDA system. The BOTDA analyzer was a Japanese-made Neubrex
6040A with a 10 cm spatial resolution. The schematic diagram is shown in Figures 6 and 7.
Finally, the electric accelerated corrosion experiments were conducted. Before the
steel bar was energized, the outer optical fiber with a spiral distribution was fixed on the
mortar, as shown in Figure 5. Simultaneously, to protect the bare 2 cm steel bars at the
bottom of the specimen from corrosion, the bottom surface of the mortar specimen was
also sealed with AB epoxy glue. Subsequently, the mortar specimen was completely
immersed in a water tank containing 5% NaCl solution and started to be corroded. There
were three identical HY3005MT digital DC power suppliers used in the accelerated
corrosion testing. The power suppliers have a steady current of 0.1 A. It was assumed that
the corrosion of rebar is uniform. Meanwhile, the distributed fiber strains of 5 #, 8 #, and
10 # specimens were recorded by the BOTDA system. The BOTDA analyzer was a
Japanese-made Neubrex 6040A with a 10 cm spatial resolution. The schematic diagram is
shown4.
Figure
Figure 4.in Figures of
Schematic
Schematic 6 and
of the 7.
the specimen.
specimen.

Finally, the electric accelerated corrosion experiments were conducted. Before the
steel bar was energized, the outer optical fiber with a spiral distribution was fixed on the
mortar, as shown in Figure 5. Simultaneously, to protect the bare 2 cm steel bars at the
bottom of the specimen from corrosion, the bottom surface of the mortar specimen was
also sealed with AB epoxy glue. Subsequently, the mortar specimen was completely
immersed in a water tank containing 5% NaCl solution and started to be corroded. There
were three identical HY3005MT digital DC power suppliers used in the accelerated
corrosion testing. The power suppliers have a steady current of 0.1 A. It was assumed that
the corrosion of rebar is uniform. Meanwhile, the distributed fiber strains of 5 #, 8 #, and
10 # specimens were recorded by the BOTDA system. The BOTDA analyzer was a
Japanese-made Neubrex 6040A with a 10 cm spatial resolution. The schematic diagram is
shown in Figures 6 and 7.

Micromachines 2021, 12, x FOR PEER REVIEW 6 of 11

Figure5.5.Schematic
Figure Schematicillustration
illustrationofofthe
thespiral
spiraldistribution
distributionofofthe
theoptical
opticalfiber.
fiber.

Figure 5. Schematic illustration of the spiral distribution of the optical fiber.

Figure 6. Diagram of accelerated rebar corrosion.
Figure 6. Diagram of accelerated rebar corrosion.
Micromachines 2021, 12, x FOR PEER REVIEW 6 of 11

Micromachines 2021, 12, 1451 6 of 11

Figure 6. Diagram of accelerated rebar corrosion.

Figure 6. Diagram of accelerated rebar corrosion.

Figure7.7.Experimental
Figure Experimentalsetup
setupof
ofaccelerated
acceleratedcorrosion.
corrosion.

4.4.Results
ResultsandandDiscussion
Discussion
The
The relationship
relationship between
between fiber
fiber strain
strain and
and corrosion
corrosion rate
rate isis shown
shown in in Figures
Figures 8–10.
8–10.
Obviously,
Obviously, the spiral fiber strain depends on the corrosion mass loss rate of the steel bar.
the spiral fiber strain depends on the corrosion mass loss rate of the steel bar.
The
Thecorrosion
corrosionmass
massloss
lossrate
rateisisthe
theratio
ratioofofthe
thecorrosion
corrosionmass
massloss
lossto tothe
theinitial
initialrebar
rebarmass,
mass,
named
namedthe thecorrosion
corrosionrate. The
rate. The spiral fiber
spiral strain
fiber increases
strain generally
increases generally withwith
the corrosion rate.
the corrosion
Moreover, the sensing fiber strain with the corrosion rate is divided into three
rate. Moreover, the sensing fiber strain with the corrosion rate is divided into three stages, stages, i.e.,
Stage I, Stage
i.e., Stage I, II, and II,
Stage Stage
andIII. The findings
Stage Щ. The are consistent
findings with the references.
are consistent with the Therefore,
references.
the spiral distributed sensor reflects well the degradation of reinforced
Therefore, the spiral distributed sensor reflects well the degradation of reinforced concrete.
concrete.
(1) Stage I: In this duration, the steel bar was slowly and gradually corroded. The volume
Figureof7.the
Experimental
corrosion setup
productsof accelerated
accumulated,corrosion.
leading to a subsequent increased expansion
2000 Reinforced
force, which layer fiber
correspondingly contributed to the increased fiber strain. The helix fiber
4. Results
strainand
1800
Discussion
Mortar
increased layer fiber linearly with the corrosion rate.
almost
(2) TheStagerelationship
II: The fiberbetween
strain increased rapidly,
fiber strain andwhich was rate
corrosion attributed to concrete
is shown cracking.
in Figures 8–10.
1600 Ⅲ
In addition, the strain of the depends
distributed onfiber sensor located
massin therate
steelofbar
the(the
steelinner
Ⅱ
Obviously, Ⅰthe spiral fiber strain the corrosion loss bar.
1400
fiber) wasmass
The corrosion greaterlossthan
ratethose
is theof the of
ratio distributed fibermass
the corrosion sensor located
loss to theininitial
the mortar layer
rebar mass,
Strain/µɛ

named1200
(the
theouter fiber). The
corrosion rate.sharp increase
The spiral in fiber
fiber strain
strain is obviously
increases shown
generally with in the
Figures 8–10.
corrosion
Cracking
In this duration, AB epoxy glue on the bottom surface of the mortar
rate. Moreover, the sensing fiber strain with the corrosion rate is divided into three stages,
1000 specimen came
off, soI,the
i.e., 800
Stage AB glue
Stage was reused
II, and Stage Щ. to protect the bare are
The findings 2 cmconsistent
steel bars, as shown
with the in Figure 8.
references.
(3) Stage
Therefore, III: The fiber strain remained unchanged or was less than
the spiral distributed sensor reflects well the degradation of reinforced the peak strain, owing
600
to the rust loss.
concrete.
400
Glue shedding
200 Reinforced layer fiber
2000 0.0 0.2 0.4 0.6 0.8 1.0 1.2
Mortar layer fiber
1800 Corrosion rate/%
1600 Ⅱ
Ⅲ
Ⅰ
1400
Strain/µɛ

1200
Cracking
1000

800

600

400
Glue shedding
200
0.0 0.2 0.4 0.6 0.8 1.0 1.2
Corrosion rate/%

Figure 8. Corrosion curves of sample 5 #.

 Micromachines 2021, 12, x FOR PEER REVIEW 7 of 11
Micromachines 2021, 12, x FOR PEER REVIEW 7 of 11
Micromachines 2021, 12, 1451 7 of 11

Figure 8. Corrosion curves of sample 5 #.

Figure 8. Corrosion curves of sample 5 #.
3500
3500 Reinforced layer fiber
3000
Reinforced
Mortar layerlayer
fiberfiber
3000 Mortar layer fiber
2500
2500

Strain/µɛ
2000
Strain/µɛ 2000
1500
1500
1000
1000
500
500
0
0 0.0 0.2 0.4 0.6 0.8 1.0 1.2
0.0
0.4
Corrosion 0.2
0.6
rate/%0.8 1.0 1.2
Corrosion rate/%
Figure
Figure 9.
9. Corrosion
Corrosioncurves
curvesofofsample
sample 8 #.
8 #.
Figure 9. Corrosion curves of sample 8 #.
4000
4000 Reinforced layer fiber
3500 Reinforced
Mortar layerlayer
fiberfiber
3500 Mortar layer fiber
3000
3000
2500
Strain/µɛ

2500
Strain/µɛ

2000
2000
1500
1500
1000
1000
500
500
0
0
0.0 0.2 0.4 0.6 0.8 1.0 1.2
0.0 0.2 0.4 0.6
Corrosion rate/%0.8 1.0 1.2
Corrosion rate/%
Figure 10. Corrosion curves of sample 10 #.
Figure 10. Corrosion
Figure Corrosioncurves
curvesofofsample
sample1010
#. #.
(1) Stage I: In this duration, the steel bar was slowly and gradually corroded. The volume
(1) of Stage
In I: In thisthere
addition,
the corrosion duration, the
was stochastic
products steel bar
accumulated, wasleading
slowlycharacteristic
fluctuation and
to graduallyofcorroded.
a subsequent spiral fiber
increased The volume
strain
expansion in the
of
thirdforce,the
stages corrosion products
of 5 correspondingly
which #, 8 #, and 10 # in accumulated,
Figures 8–10.
contributed leading
to theTheto a subsequent
test data
increased increased
fiberwere checked
strain. expansion
The helix afterwards.
fiber
force, increased
In general,
strain which
a single correspondingly
peak
almost oflinearly contributed
BGS spectrumwith theshown to the in
corrosion increased
Figure
rate. 11 fiber
is strain.
desired, The helix fiber
which failed to
(2) strain
causeStage increased
the measurement
II: The fiber almost linearly
error.
strain On with
increased the
the contrary, corrosion
rapidly, the rate. was
double
which peaks in thisto
attributed BGS spectrum
concrete
(2) chaotic
and Stage II:BGS
cracking. The
In fiber strain
spectrum,
addition, theas increased
shown
strain of the rapidly,
in distributed
Figure which
12a,b,fiber was the
caused
sensor attributed to steel
concrete
straininfluctuation.
located the bar The
REVIEWcracking.
Micromachines 2021, 12, x FOR PEERabove-mentioned In addition,
phenomena the strain
were of
thethe distributed
origins of the fiber
tested
(the inner fiber) was greater than those of the distributed fiber sensor located in thesensor located
frequency in
shift the steel
error, 8 bar
which
of 11 in
(the
turn mortar
led to inner
the fiber)
inaccuracywas greater
of the than
spiral those
fiber of the
strain. distributed
layer (the outer fiber). The sharp increase in fiber strain is obviously shown fiber sensor located in the
mortar
in Figureslayer8–10.(theInouter fiber). The
this duration, ABsharp
epoxy increase
glue oninthe fiber strainsurface
bottom is obviously shown
of the mortar
in Figurescame
83 specimen 8–10.off,In this duration, AB epoxy glue on the bottom
so the AB glue was reused to protect the bare 2 cm steel bars, as surface of the mortar
specimen came off, so the AB glue was reused to protect the bare 2 cm steel bars, as
82 shown in Figure 8.
shown in Figure 8.
(3) 81 Stage Щ: The fiber strain remained unchanged or was less than the peak strain, owing
(3) to Stage
the Щ:
rustThe fiber strain remained unchanged or was less than the peak strain, owing
loss.
80
to the rust loss.
In addition, there was stochastic fluctuation characteristic of spiral fiber strain in the
BGS(dB)

79
third In addition,
stages of 5 #,there was10
8 #, and stochastic
# in Figures fluctuation
8–10. The characteristic
test data were of spiral
checked fiber strain in the
afterwards. In
third
78
stages of 5 #, 8 #, and 10 # in Figures 8–10. The test
general, a single peak of BGS spectrum shown in Figure 11 is desired, which failed to data were checked afterwards. In
general,
77 a single peak of BGS spectrum shown in Figure
cause the measurement error. On the contrary, the double peaks in this BGS spectrum and 11 is desired, which failed to
cause
chaotic
76 the
BGSmeasurement
spectrum, aserror. shown Oninthe contrary,
Figure 12a,b, the double
caused the peaks
straininfluctuation.
this BGS spectrum
The above- and
chaotic BGSphenomena
mentioned spectrum, as shown
were the in Figure
origins of12a,b,
the caused
tested the strain
frequency fluctuation.
shift error, whichThe in
above-
turn
75
mentioned
led phenomena
to the inaccuracy were
of the the origins
spiral of the tested frequency shift error, which in turn
fiber strain.
led74to 10.6
the inaccuracy
10.7 10.8 of the
10.9 spiral
11.0 fiber
11.1 strain.
11.2 11.3
Frequency(GHz)

Figure11.
Figure 11. BGS
BGS spectrum
spectrum with
withaasingle
singlepeak.
peak.
 76

75

74
10.6 10.7 10.8 10.9 11.0 11.1 11.2 11.3
Frequency(GHz)
Micromachines 2021, 12, 1451 8 of 11
Figure 11. BGS spectrum with a single peak.

(a) (b)
Figure
Figure 12.12.
BGSBGS spectrum
spectrum of of
thethe spiral
spiral distributed
distributed sensor.
sensor. (a)(a)
BGSBGS spectrum
spectrum with
with double
double peaks;
peaks; (b)(b)
BGSBGS spectrum
spectrum without
without
Brillouin peak.
Brillouin peak.
These
These above-mentioned
above-mentionedfiber fiberstrain
strainvibrations
vibrations showed
showed good
goodagreement
agreement with
with thethe
changes in the concrete structure. There were three stage of distributed fiber strain
changes in the concrete structure. There were three stage of distributed fiber strain in good in good
agreement
agreement with
withthethe
three stages
three of of
stages thethe
concrete
concrete structure,
structure, as as
shown
shownin in
Figure 13.13.
Figure In In
thisthis
figure,
figure, it can be seen clearly that the corrosion products flowed from the specimenthe
it can be seen clearly that the corrosion products flowed from the specimen into into
water due todue
the water the to
concrete cracks.cracks.
the concrete Mechanical analysis
Mechanical was carried
analysis out for out
was carried the concrete with
for the concrete
cover thickness c and reinforcement diameter d. In the figure, δ represents
with cover thickness c and reinforcement diameter d. In the figure, δ represents the the deformation
of deformation
concrete dueoftoconcrete
rust expansion force.
due to rust At first, force.
expansion the corrosion
At first, products filledproducts
the corrosion the porous filled
area and the elastic deformation was generated in the corrosion duration.
the porous area and the elastic deformation was generated in the corrosion duration. As the corrosionAs
rate increased, the expansion force generated by the corrosion products exceeded the tensile
the corrosion rate increased, the expansion force generated by the corrosion products
strength of the concrete. Subsequently, concrete cracking was generated. Hence, a sharp
exceeded the tensile strength of the concrete. Subsequently, concrete cracking was
increase in the fiber strain occurred, owing to the generation of concrete crack, until the
generated. Hence, a sharp increase in the fiber strain occurred, owing to the generation of
cracks propagated to the surface of the specimens. Therefore, the concrete cracks were
concrete crack, until the cracks propagated to the surface of the specimens. Therefore, the
generated, propagated, and penetrated the specimen, which directly led to the corrosion
concrete cracks were generated, propagated, and penetrated the specimen, which directly
products’ outflow into the water through the concrete cracks. The generation and outflow
led to the corrosion products’ outflow into the water through the concrete cracks. The
of the corrosion products of steel reached a state of dynamic equilibrium or the more the
Micromachines 2021, 12, x FOR PEER generation and outflow of the corrosion products of steel reached a state of dynamic
rustREVIEW
flowed out. Hence, the spiral fiber strain may be unchanged or less than the peak 9 of 11
equilibrium or the more the rust flowed out. Hence, the spiral fiber strain may be
strain, owing to the rust loss in the last duration.
unchanged or less than the peak strain, owing to the rust loss in the last duration.

Figure 13. Schematic

Figure diagram
13. Schematic of the
diagram influence
of the of reinforcement
influence corrosion
of reinforcement on concrete.
corrosion (a) The
on concrete. firstfirst
(a) The stage; (b) the
stage; second
(b) the second
stage; (c) the
stage; (c) third stage.
the third stage.
Moreover, the cracking times of specimens 5 #, 8 #, and 10 # specimens were ap-
Moreover,
proximately the 16
6, 10, and cracking times ofasspecimens
h, respectively, 5 Figures
observed in #, 8 #, and
8–10,10 # in
and specimens were
Table 2. At
approximately 6, 10, and 16 h, respectively, as observed in Figures 8–10, and in Table 2.
At these durations, their corresponding corrosion rates were 0.16, 0.2, and 0.35%,
respectively. Therefore, it was concluded that the cracking times and corresponding
corrosion rates were highly related to the thickness of the protective layer. The concrete
crack was well recorded, as shown in Figure 14. Obviously, the cracks depend on the
Micromachines 2021, 12, 1451 9 of 11

these durations, their corresponding corrosion rates were 0.16, 0.2, and 0.35%, respectively.
Figure 13. Schematic diagramTherefore,
of the influence
it wasofconcluded
reinforcement
that corrosion on concrete.
the cracking times and(a)corresponding
The first stage; corrosion
(b) the second
rates were
stage; (c) the third stage.
highly related to the thickness of the protective layer. The concrete crack was well recorded,
as shown in Figure 14. Obviously, the cracks depend on the corrosion rate.
Moreover, the cracking times of specimens 5 #, 8 #, and 10 # specimens were
approximately 6, 10,
Table 2. Corrosion andand
rates 16corresponding
h, respectively, as observed
cracking times. in Figures 8–10, and in Table 2.
At these durations, their corresponding corrosion rates were 0.16, 0.2, and 0.35%,
respectively.Specimen Corrosion
Therefore, it was concluded thatRate
the(%)
cracking timesCracking Time (h)
and corresponding
corrosion rates were
5# highly related to the thickness
0.16 of the protective layer. 6The concrete
crack was well 8# recorded, as shown in Figure 0.2014. Obviously, the cracks depend
10 on the
corrosion rate. 10# 0.35 16

Figure
Figure 14.14. Generated
Generated concrete
concrete crack.
crack.
To sum up, the corrosion expansion of steel bar and cracking of the outer mortar
Table 2. Corrosion rates and corresponding cracking times.
contributed to the fiber strain variations. Thus, it is evident that significant strain variations
in the spiral winding fiber reflected
Specimen concrete
Corrosion cracking.
Rate (%) The spiral distributed
Cracking Time (h)strain can
precisely reflect
5# the corrosion of reinforcement
0.16 concrete. 6
8# 0.20 10
5. Conclusions
10# 0.35 16
This study developed a new method of monitoring the corrosion of steel bars using
spiral wound fibers. Its feasibility was theoretically analyzed and experimentally verified.
To sum up, the corrosion expansion of steel bar and cracking of the outer mortar
The main findings of this study are as follows:
contributed to the fiber strain variations. Thus, it is evident that significant strain
(1) Thein
variations bending curvature
the spiral winding of fiber
the spiral distributed
reflected fiber
concrete sensor is
cracking. Therelated
spiraltodistributed
the winding
radius and winding pitch of the sensing fiber. The desired
strain can precisely reflect the corrosion of reinforcement concrete. curvature of the winding
fiber (0.667 cm−1 ) is a critical parameter for the distributed sensor.
(2) Fiber strain variations can precisely reflect the corrosion of reinforcement concrete.
Therefore, using spiral distributed fiber sensors to monitor corrosion of steel bars and
concrete cracking is a feasible technique.
(3) The cracking times and corresponding corrosion rates are closely related to the thick-
ness of the protective layer. Accordingly, the proposed method with the distributed
sensor wound on the protective mortar layer is verified to be practical.

Author Contributions: Conceptualization and formal analysis, J.L.; data curation, Y.Z.; writing—
original draft preparation, J.L. and Y.Z.; investigation and resources, Y.Z. and J.W. All authors have
read and agreed to the published version of the manuscript.
Funding: This research was funded by National Natural Science Foundation of China (51778379)
and the Natural Science Foundation of Hebei Province (E2019210201).
Micromachines 2021, 12, 1451 10 of 11

Institutional Review Board Statement: Not applicable.

Informed Consent Statement: Not applicable.
Data Availability Statement: The data presented in this study are available from the corresponding
author, (J.L.), upon reasonable request.
Conflicts of Interest: The authors declare no conflict of interest.

References
1. Fahim, A.; Ghods, P.; Isgor, O.B.; Thomas, M.D.A. A critical examination of corrosion rate measurement techniques applied to
reinforcing steel in concrete. Mater. Corros. 2018, 69, 1784–1799. [CrossRef]
2. Hong, N. Corrosion and protection technology of steel bars in concrete (6)—corrosion inhibitor and cathodic protection of steel
bars. Ind. Constr. 2000, 1, 57–60. [CrossRef]
3. Shen, J.; Wu, S. Study on the performance and corrosion model of corroded steel bars in concrete under atmospheric environment.
China Concr. Cem. Prod. 2004, 3, 46–50. [CrossRef]
4. Zang, J.; Fang, C. Research progress on service life prediction of reinforced concrete structures under chloride attack. J. Mater. Sci.
Eng. 2019, 37, 848–854+859.
5. Wang, X.; Jin, W.; Yan, Y. Path probability model of corrosion-crack assessment for existing reinforced concerte structures. J.
Zhejiang Univ. (Eng. Sci.) 2010, 44, 1191–1196.
6. Liu, Y.; Weyers, R.E. Modeling the time-to-corrosion cracking in chloride contaminated reinforced concrete structures. Mater. J.
1998, 95, 675–680.
7. Zhao, Y.; Jin, W. Corrosion ratio of reinforcement bar in reinforced concrete construction at the moment of cracking due to
corrosion expansion. J. Hydraul. Eng. 2004, 11, 97–101.
8. Jin, W.; Zhao, Y.; Yan, F. The mechanism of corroded expansion force of reinforced concrete members. J. Hydraul. Eng. 2001, 7,
57–62.
9. Lu, C.; Jin, W.; Liu, R. Reinforcement corrosion-induced cover cracking and its time prediction for reinforced concrete structures.
Corros. Sci. 2011, 53, 1337–1347. [CrossRef]
10. Chernin, L.; Val, D.V.; Volokh, K.Y. Analytical modelling of concrete cover cracking caused by corrosion of reinforcement. Mater.
Struct. 2010, 43, 543–556. [CrossRef]
11. Ahmad, S. An experimental study on correlation between concrete resistivity and reinforcement corrosion rate. Anti-Corros.
Methods Mater. 2014, 61, 158–165. [CrossRef]
12. Femenias, Y.S.; Angst, U.; Elsener, B. Monitoring pH in corrosion engineering by means of thermally produced iridium oxide
electrodes. Mater. Corros. 2018, 69, 76–88. [CrossRef]
13. Loto, R.T.; Busari, A. Influence of white aluminum dross on the corrosion resistance of reinforcement carbon steel in simulated
concrete pore solution. J. Bio- Tribo-Corros. 2019, 5, 19. [CrossRef]
14. Zhang, M.; Geng, O. Research status and development trend of steel corrosion monitoring technology in concrete. Concrete 2007,
2, 26–28.
15. Abbas, Y.; Pargar, F.; Olthuis, W.; Berg, A.V.D. Activated carbon as a pseudo-reference electrode for potentiometric sensing inside
concrete. Procedia Eng. 2014, 87, 1437–1440. [CrossRef]
16. Muralidharan, S.; Saraswathy, V.; Berchmans, L.J.; Thangavel, K.; Ann, K.Y. Nickel ferrite (NiFe2O4): A possible candidate
material as reference electrode for corrosion monitoring of steel in concrete environments. Sens. Actuators B Chem. 2010, 145,
225–231. [CrossRef]
17. Chen, X.; Fang, L. Performance analysis and experimental verification for FBG sensors applied for smart structure. Key Eng.
Mater. 2007, 336–338, 1357–1360. [CrossRef]
18. Lv, H.; Zhao, X.; Zhan, Y.; Gong, P. Damage evaluation of concrete based on Brillouin corrosion expansion sensor. Constr. Build.
Mater. 2017, 143, 387–394. [CrossRef]
19. Liang, R.; Tao, J.; Jia, Z.; Li, D.; Yuan, C.; Li, H. Pipeline corrosion and leakage monitoring based on the distributed optical fiber
sensing technology. Measurement 2018, 122, 57–65. [CrossRef]
20. Suresh, R.; Tjin, S.W.; Ngo, N.Q. Application of a new fiber Bragg grating based shear force sensor for monitoring civil structural
components. Smart Mater. Struct. 2005, 14, 982–988. [CrossRef]
21. Tan, C.H.; Shee, Y.G.; Yap, B.K.; Adikan, F.R.M. Fiber Bragg grating based sensing system: Early corrosion detection for structural
health monitoring. Sens. Actuators A Phys. 2016, 246, 123–128. [CrossRef]
22. Tang, F.; Chen, Y.; Li, Z.; Tang, Y.; Chen, G. Application of Fe-C coated LPFG sensor for early stage corrosion monitoring of steel
bar in RC structures. Constr. Build. Mater. 2018, 175, 14–25. [CrossRef]
23. Chu, T.; Lu, C.; Zhang, Z.; Lu, J. The durability of natural aggressive environmental conditions on the tensile performance
of FRP bars. In Proceedings of the 5th Advanced Materials, Mechanics and Structural Engineering Conference, Seoul, Korea,
7 October 2018.
24. Wang, X.; Wang, W. The influence and research of the metal-coated FBG on temperature sensitivity. Adv. Mater. Res. 2009, 146–147,
1894–1897. [CrossRef]
Micromachines 2021, 12, 1451 11 of 11

25. Wang, Y.; Chen, J.; Rao, Y. Long period fiber grating sensors measuring bend-curvature and determining bend-direction
simultaneously. J. Optoelectron. Laser 2005, 16, 1139–1143. [CrossRef]
26. You, S.; Hao, S.; Yin, Z.; Li, X. Analysis and Test of Bend Loss in Single-mode Fiber. Acta Photonica Sin. 2003, 32, 409–412.
[CrossRef]