sensors
Review
A Review of Recent Distributed Optical Fiber Sensors
Applications for Civil Engineering Structural
Health Monitoring
Mattia Francesco Bado 1,2, *






Citation: Bado, M.F.; Casas, J.R. A
Review of Recent Distributed Optical
Fiber Sensors Applications for Civil
Engineering Structural Health
Monitoring. Sensors 2021, 21, 1818.
https://doi.org/10.3390/s21051818
Academic Editors: Karim Benzarti
and Paulo Antunes
Received: 2 February 2021
Accepted: 23 February 2021
Published: 5 March 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/).
Sensors 2021, 21, 1818. https://doi.org/10.3390/s21051818
and Joan R. Casas 1
1 Department of Civil and Environmental Engineering, Technical University of Catalonia (UPC), Jordi Girona 1–3,
08034 Barcelona, Spain; joan.ramon.casas@upc.edu
2 Department of Reinforced Concrete Structures and Geotechnical Engineering, Vilnius Gediminas Technical
University, Saulėtekio al. 11, 10221 Vilnius, Lithuania
* Correspondence: mattia.francesco.bado1@upc.edu
Abstract: The present work is a comprehensive collection of recently published research articles
on Structural Health Monitoring (SHM) campaigns performed by means of Distributed Optical
Fiber Sensors (DOFS). The latter are cutting-edge strain, temperature and vibration monitoring tools
with a large potential pool, namely their minimal intrusiveness, accuracy, ease of deployment and
more. Its most state-of-the-art feature, though, is the ability to perform measurements with very
small spatial resolutions (as small as 0.63 mm). This review article intends to introduce, inform and
advise the readers on various DOFS deployment methodologies for the assessment of the residual
ability of a structure to continue serving its intended purpose. By collecting in a single place these
recent efforts, advancements and findings, the authors intend to contribute to the goal of collective
growth towards an efficient SHM. The current work is structured in a manner that allows for the
single consultation of any specific DOFS application field, i.e., laboratory experimentation, the built
environment (bridges, buildings, roads, etc.), geotechnical constructions, tunnels, pipelines and wind
turbines. Beforehand, a brief section was constructed around the recent progress on the study of
the strain transfer mechanisms occurring in the multi-layered sensing system inherent to any DOFS
deployment (different kinds of fiber claddings, coatings and bonding adhesives). Finally, a section
is also dedicated to ideas and concepts for those novel DOFS applications which may very well
represent the future of SHM.
Keywords: distributed optical fiber sensors; distributed optical fiber sensors; distributed sensing;
distributed monitoring; DOFS; DFOS; structural health monitoring; SHM; review
1. Preface
The present work reports numerous Distributed Optical Fiber Sensors (DOFS) moni-
toring campaigns spread over multiple Structural and Civil Engineering fields. Whilst a
sequential reading of the article’s diverse sections may provide a global point of view on
the evolution of DOFS for the Structural Health Monitoring (SHM) panorama, the authors
are intent on also providing the readers with the possibility of skipping directly to the
section of interest.
Section 2. Introduction provides a very general overview on the topic of SHM and the
importance of the implementation of DOFS. Before presenting the most recently published
DOFS-SHM integrations, Section 3. DOFS Coatings and Bonding Adhesives: The Strain
Transfer Process and the Optimal Bonding Technique tackles an issue of great relevance which
intrinsically affects any DOFS monitoring. The latter, as foreshadowed by the section’s
title, is the strain transfer process occurring in the multi-layered sensing system that
characterizes a bonded DOFS. In more detail, due to the stiffness diversity of the adhesives,
cladding and coating layers that separate a DOFS’ sensing core from a structural host
https://www.mdpi.com/journal/sensors
Sensors 2021, 21, 1818 2 of 83

surface, a certain distortion of the sampled data occurs which, in return, decreases the
reliability of the DOFS measured strain. The authors deemed mandatory reporting some
recently published findings on this crucial and intrinsic limitation before reporting any
in-situ DOFS-monitoring campaign.
Finally, the reported DOFS-SHM applications are separated into the following sections:
• Section 4. Laboratory Experimental Applications
• Section 5. The Built Environment: Buildings, Bridges and Roads
• Section 6. Geotechnical Engineering
• Section 7. Tunnel
• Section 8. Pipes
• Section 9. Wind Turbines
Finally, Section 10. An Eye towards the Future introduces the reader to those DOFS-SHM
novelties that may very well become key DOFS applications in the future.
Acknowledging the broadness of the above-listed application fields and the readers’
engineering backgrounds differences, for the proper understanding of all the DOFS appli-
cations, the authors started each section and subsection with a small introduction aimed at
providing a certain degree of contextualization. At the end of each section, instead, a brief
summary will be presented encapsulating the key findings and DOFS deployments tips.
Furthermore, the authors would like to remark that the hereafter reported figures
will all be taken from the referred articles and are, therefore, uniquely the product of their
respective writers’ efforts. In line with the concept of research and its goal of collective
growth towards the solution of society’s biggest problems, the figures were selected with
the aim of clarifying the reported DOFS deployment techniques and their outputs in
addition to sharing any particular ingenious technique that other monitoring campaigns
might draw inspiration from.
Finally, the present work situates itself as a spiritual successor to the work developed
by Barrias et al. [1] which studied the DOFS-SHM integrations until 2016, the year of
publication. Consequently, the present review article deals with the publications comprised
in the time interval 2017–2020 additionally expanding on previously overlooked topics.

2. Introduction
As clearly stated by Brownjohn [2], civil infrastructure provides the means for a society
to function. The wide range of applications that Civil Engineering offers includes buildings,
bridges, highways, tunnels, power plants, industrial facilities, geotechnical, hydraulic
structures and more. Throughout their service lifetimes, each of these is subjected to
multiple events that deteriorate and compromise their structural integrity thus exposing
their future performances to several risks. In order to avoid the social, economic and
environmental costs of their deterioration and/or failure, civil structures have different
safety and durability requirements that need to be met. In order to ensure correct structural
performance and the wellbeing of their users, a series of inspections, monitoring and
maintenance protocols have to be carried out during the structures’ service lifetimes. The
effectiveness of these monitoring methods relies on their ability to rapidly detect alterations
of the structure’s performance, allowing for their identification, characterization and control.
This idea is the basis of Structural Health Monitoring (SHM). Indeed, according to Housner
et al. [3], SHM can be understood as the continuous and regular measurement and analysis
of key structural and environmental parameters under operating conditions for the purpose
of warning of any early-stage abnormal structural behaviors.
Yet, as stated by Cawley [4], despite the recently increasing volume of research on
SHM, this enormous effort has yielded only a small number of routine industrial ap-
plications. Indeed, still at the present day, the most common manner of assessing an
infrastructure’s ability to carry on with its designed duties is through the intervention
of engineers trained in structural visual inspections [1]. A non-disregardable degree of
inaccuracy is therefore automatically incorporated in these studies due to the differences in
the inspectors’ safety condition assessment backgrounds. Wherever monitoring tools are
 tions. Indeed, still at the present day, the most common manner of assessing an infrastruc-
ture’s ability to carry on with its designed duties is through the intervention of engineers
trained in structural visual inspections [1]. A non-disregardable degree of inaccuracy is
Sensors 2021, 21, 1818 3 of 83
therefore automatically incorporated in these studies due to the differences in the inspec-
tors’ safety condition assessment backgrounds. Wherever monitoring tools are addition-
ally employed, these usually fall under the category of “traditional monitoring tech-
niques”
additionallysuchemployed,
as inclinometers, accelerometers,
these usually fall underextensometers,
the category oftotal station surveys,
“traditional monitoring load
cells and GNS-based
techniques” sensors. By themselves,
such as inclinometers, these can
accelerometers, be considered
extensometers, sufficiently
total reliable
station surveys,
sincecells
load theirand
correct deployment
GNS-based has been
sensors. extensivelythese
By themselves, investigated and they aresufficiently
can be considered nowadays
widely since
reliable acknowledged,
their correct thus ensuring reliable
deployment has beenmonitoring
extensivelyand structuraland
investigated assessments.
they are
However, widely
nowadays as stated in Baker [5], conventional
acknowledged, thus ensuring forms of inspection
reliable monitoring andand monitoring
structuralare only
assess-
as good as their ability to uncover potential issues in an accurate
ments. However, as stated in Baker [5], conventional forms of inspection and monitoring and timely manner. In-
deed,
are onlyregarding
as good the ability
as their of damage
ability detection,
to uncover traditional
potential issues tools
in anpresent
accurate several draw-
and timely
backs which
manner. Indeed, are,regarding
amongstthe others, insufficient
ability of damagedata management
detection, ability,
traditional toolsthe needseveral
present for the
drawbacks whichservice
infrastructure’s are, amongst others,during
interruption insufficient data management
their deployment, ability, the need
non-automated for
real-time
the infrastructure’s
measurements andservice interruption
interference during
risk. The above their deployment,
deficiencies, non-automated
stacked up with the real-time
lack of
measurements
reliable, scalable and andinterference
affordablerisk. The above
monitoring deficiencies,
solutions and thestacked
absence upofwith the lack
uniform execu-of
reliable, scalable and affordable monitoring solutions and the
tion methodologies [6] and some degree of standardization, are the reason why SHM is absence of uniform execution
methodologies
not yet implemented[6] andas some degree of
a standard standardization,
practice in most Civil areEngineering
the reason why SHM is not yet
structures.
implemented
Modern as a standardtechnologies
monitoring practice in most Civil Engineering
are aimed at tackling the structures.
above-mentioned limita-
tionsModern monitoring
thus boosting technologies
the sensors’ are automation
precision, aimed at tackling and data themanagement
above-mentioned speed.limita-
These
tions thus boosting the sensors’ precision, automation and data
include amongst others Optical Fiber Sensors (OFS), Global Positioning Systems (GPS), management speed. These
include
radars, amongst
Micro ElectroothersMechanical
Optical Fiber Sensors
Systems (OFS), Global
(MEMS) and Image Positioning
Processing Systems (GPS),
Techniques
radars, Micro Electro Mechanical Systems (MEMS) and Image
such as Digital Image Correlation (DIC). As a matter of fact, as properly put by Seo Processing Techniques such
[7],
as Digital Image Correlation (DIC). As a matter of fact, as properly
Civil Engineering has recently seen a convergence with various other fields, thus slowly put by Seo [7], Civil
Engineering
moving towards has recently
technologyseen convergence.
a convergenceThe with various
latter other fields,
is a theory thus the
asserting slowly moving
absence of a
towards technology convergence. The latter is a theory asserting
clear distinctive line between the products/applications of different disciplines but rather the absence of a clear
distinctive line intersection
a co-existence, between theand products/applications
reciprocal enhancement of different
betweendisciplines
the latter but rather
which, a
in re-
co-existence, intersection and reciprocal enhancement between
turn, accelerates the overall technological advancement [8]. SMART monitoring is one of the latter which, in return,
accelerates the overalland
these developments technological
the use of OFSadvancement [8]. SMARTcan
in Civil Engineering monitoring
be consideredis onea of these
quintes-
developments and the use of OFS in Civil Engineering can be considered
sential example as it joins Civil Engineering, Photonics, Materials Engineering and Com- a quintessential
example as it joins Civil Engineering, Photonics, Materials Engineering and Computer
puter Engineering.
Engineering.
Focusing in particular on OFS, these are dielectric devices used to confine and guide
Focusing in particular on OFS, these are dielectric devices used to confine and guide
light. They consist of several layers: fiber core, cladding and, occasionally, an external
light. They consist of several layers: fiber core, cladding and, occasionally, an external
jacket aimed at providing mechanical resistance to the fiber (an example is provided in
jacket aimed at providing mechanical resistance to the fiber (an example is provided in
Figure 1).
Figure 1).

Figure1.1.3D
Figure 3Dillustration
illustrationof
ofan
anoptical
opticalfiber
fibercross-section.
cross-section.

The
Themajority
majorityofofOFS
OFSused
usedininsensing
sensingapplications
applicationshave
havesilica
silica(SiO
(SiO2 2))glass
glasscores
coresand
and
claddings.
claddings. The refractive index of the latter is lower than that of the core in orderto
The refractive index of the latter is lower than that of the core in order tosatisfy
satisfy
the condition of Snell’s law for total internal reflection and thus confine the propagation
of the light along the fiber core only [9]. The jacket is usually made of polymeric material
or nylon and can vary in diametrical size, shape and manufacturing process. Finally, as
will be evident in this article, the external coating can be constituted by numerous kinds of
materials, layering, thicknesses and external textures.
A significant ability of OFS is the ability to measure mechanical and temperature-
variation-induced strains along the fiber length by means of light scattering and back-
Sensors 2021, 21, 1818 4 of 83

scattering occurring whenever the photons of the emitted light interact with the physical
medium through which it travels (the fiber’s core itself). When no strain or temperature is
imposed on the system, light propagates and gets reflected throughout the imperfections
with a given signature. Instead, when strain or temperature values vary, the frequency
of the back scattered light is shifted. The measured strain values are then related to the
frequency shift.
As a matter of fact, three different types of light scattering and backscattering may
occur in an OFS, namely Raman, Brillouin and Rayleigh. All hold particular optical features
that make one more suitable than the others relative to the research objectives. Brillouin
scattering, thanks to its extended measurement range capability (up to several kilometers),
is the most studied and used OFS system in Civil and Geotechnical Engineering. The
Raman back-scattering, instead, is characterized by a high dependence on temperature but
can also be used to extract physical and chemical information of a material [10] such as food
quality [11] or presence of explosive materials [12]. The third back-scattering phenomenon
is Rayleigh. Despite its 70 m sensing range limit, it provides the highest spatial resolution
of the three (even down to 0.63 mm), which is ideal for strain and damage monitoring in
the context of environmentally controlled experimental laboratory investigation.
Whilst multiple sections of the present article could be dedicated to the explanation of
the different photonic working principles of all the commercially available kinds of OFS,
this topic has been discussed innumerable times in almost every OFS application article
both new and old. Therefore, the authors believe its insertion would result in redundancy.
In its stead, Table 1 not only summarizes this article’s most commonly used acronyms
but also refers the reader to exhaustive literature on each different kind of OFS photonic
working principle (marked with the symbol * in Table 1).

Table 1. Acronyms used in the present article and literature references.

Acronym Name References

SHM Structural Health Monitoring
RC Reinforced Concrete
DIC Digital Image Correlation
SG Strain Gauges
Linear Variable Displacement
LVDT
Transformers
SRA Strain Reading Anomalies
DOFS-related
OFS Optical Fiber Sensors [1,13–15]
DOFS Distributed Optical Fiber Sensors [1,13–15]
OTDR Optical Time Domain Reflectometry [1,16,17]
Brillouin Optical Time Domain
BOTDR * [16–18]
Reflectometry
BOTDA * Brillouin Optical Time Domain Analysis [19–21]
Brillouin Optical Frequency Domain
BOFDA * [22]
Analysis
PPP-BOTDA * Pulse Pre Pump-BOTDA [23–25]
DPP-BOTDA * Differential Pulse Pair-BOTDA [26]
OFDR * Optical Frequency Domain Reflectometry [1]
OBR Optical Backscattered Reflectometer [27]
DAS * Distributed Acoustic Sensing [28,29]
DVS * Distributed Vibration Sensing [28,29]
Φ-OTDR * Phase-OTDR [29,30]
FBG * Fiber Bragg Grating [9,31,32]
* different kinds of OFS.

Table 2, instead, reports a concise summary of the performance of the various dis-
tributed sensing techniques thus providing the reader with an idea of the strengths and
weaknesses of each.
Sensors 2021, 21, 1818 5 of 83

Table 2. Performance differences between the various distributed sensing techniques [1].

Sensing Transducer Sensing Spatial

Main Object of Measurement
Technology Type Range Resolution
1 km–37
Raman OTDR Distributed 1 c–17 m Temperature
km
Brillouin BOTDR Distributed 20–50 km ≈1 m Temperature and Strains
150–200 2 cm (2 km)–2 m
Brillouin BOTDA Distributed Temperature and Strains
km (150 km)
Rayleigh
Distributed 50–70 m ≈1 mm Temperature, Strains and Vibration
OFDR/DAS
Quasi- ≈100 2 mm (Bragg Temperature and Strains and
FBG
distributed channels length) displacement

Finally, for additional reading on the subject, the authors refer the readers to the
comprehensive state-of-the-art reviews on FOS, namely Udd [33], Kersey [34] and Grattan
and Sun [35].
OFS sensors have recently improved their performance, drastically increasing their
potential pool. The latter was immediately picked up and tapped into by the worldwide
scientific community as evident from the increasingly larger number of relative published
research articles. Additionally, according to a market survey report conducted by the
Photonic Sensor Consortium and published by Information Gatekeepers [36], the increased
Sensors 2021, 21, x FOR PEER REVIEW
OFS popularity goes hand in hand with the monitored and predicted growth of the 6whole
of 87
OFS fiber sensors technological market (see Figure 2).

Figure
Figure 2.
2. Growth
Growth of
of Optical
Optical Fiber
Fiber Sensors
Sensors (OFS)
(OFS) technological
technological market
market [36].
[36].

ItIt is
is worthy
worthy ofof note
note that,
that, according
according to to Figure 2, the combined distributed and single
point
point fiber
fiber optic
optic sensor
sensor markets
markets are are projected
projected to to be over USD 1.3 billion in 2023. In In this
respect,
respect, the the present
present work
work represents
represents an effort aimed at facilitating the use of OFS for both
the
the present
present and and prospective
prospective users of this novel technology.
Now, if to focus strictly on OFS’s application to the Civil and Structural Engineering
Now,
field, it can
field, caneasily
easilybe
beasserted
asserted that it has
that slowly
it has but but
slowly steadily increased
steadily in theinpast
increased theyears
past [1,37].
years
As clearly identified by Li et al. [9], its increased popularity for SHM applications
[1,37]. As clearly identified by Li et al. [9], its increased popularity for SHM applications can be
attributed
can to the following
be attributed OFS features:
to the following OFS features:
•• They
They allow for completely
completely distributedmonitoring
distributed monitoring(in (inthe
thecase
caseofofDOFS)
DOFS) with monitor-
with moni-
ing points
toring spaced
points spacedlessless
than 1 mm;
than 1 mm;
•• Their
Their small diameter
diameterand andminimal
minimalstiffness allow
stiffness allowforfor
a very high
a very degree
high of deployment
degree of deploy-
configuration
ment complexity,
configuration complexity,no matter if these
no matter if imply circumferential
these imply surfaces,
circumferential sharp
surfaces,
sharp corners, surface irregularities and more. It is even possible to embed them in-
side structural elements with a minimal level of intrusiveness;
• Their ease of deployment as can be achieved by simply applying an adhesive over it.
• Their monitoring length is very flexible and can vary from halves of millimeter to
Sensors 2021, 21, 1818 6 of 83

corners, surface irregularities and more. It is even possible to embed them inside
structural elements with a minimal level of intrusiveness;
• Their ease of deployment as can be achieved by simply applying an adhesive over it.
• Their monitoring length is very flexible and can vary from halves of millimeter to tens
of kilometers;
• Intrinsic immunity to electro-magnetic interference;
• They are designed with a long life cycle in mind. Indeed, its main component, silica,
is highly resistant to corrosion and can withstand high tensile loading;
• Silica core OFS are highly resistant to temperature and can measure temperatures
Sensors 2021, 21, x FOR PEER REVIEW from −200 ◦ C to 800 ◦ C. 7 of 87

The great majority of photonic sensing technology applied to SHM is constituted

by discrete sensors such as Fiber Bragg Gratings (FBG) [9] as these were introduced to
damage
the marketfirst
muchoccurs andthan
earlier prevents the linking of
the undoubtedly local damage
superior DOFS. mechanisms to the global
FBG are quasi-distributed
condition
optical fiberofsensors
the structure.
in which a characteristic wavelength is used to simultaneously provide
The lack
its address of sufficient
in the monitoring
sensor network, andpoints along a fiber deployment
the measurement (temperature (spatial resolution)
and strains) [31].
is an issue entirely surpassed by DOFS (Figure 3 left). As a matter
Their applications can be dated back as early as 1993, when Intelligent Sensing for of fact, the latest model
Inno-
of OFDR
vative Optical performed
Structures Backscatterfiber Reflectometer (OBR) interrogator
optic deployments machines
on six bridges (such as[38]
in Canada the oronein
on the right side of Figure 3) is able to monitor strains with a
2001 when four FBG sensors were installed across, above and under the primary arch of spatial resolution of 0.63
mm. Cathedral in order to identify any significant structural deterioration [39]. It can be
Como’s
easily The distributed
argued that thenature
discreteof sensors’
DOFS enables
(such the mapping
as FBG) mostofcrucial
temperature,
limitationstrain
liesand vi-
in their
bration distributions in two or even three dimensions (achievable
non-distributed measurement nature. This shortcoming is quite critical in the context ofwith a DOFS mesh de-
ployment)
SHM and their
as it prevents theidentification
possibility of atprecisely
any pointpointing
along a fiber. Combined,
out the these aspects
location where al-
a potential
low for the painting of a clear picture of the global behavior of
damage first occurs and prevents the linking of local damage mechanisms to the global a structure rather than
reportingofthe
condition thetensile state of a limited number of points. Figure 3 also displays the hair-
structure.
likeThe
thinness
lack of sufficientwhich
of the fiber makespoints
monitoring it idealalong
for un-intrusive deployments.
a fiber deployment (spatialStill, as sug-
resolution)
gested by Barrias [1], this is still a recent and developing technology as
is an issue entirely surpassed by DOFS (Figure 3 left). As a matter of fact, the latest model can be perceived
ofby the relatively
OFDR scarce number
Optical Backscatter of DOFS applications
Reflectometer in SHM machines
(OBR) interrogator projects published
(such as the in re-
one
search articles.
on the right side of Figure 3) is able to monitor strains with a spatial resolution of 0.63 mm.

(a) (b)
Figure3.3.(a)
Figure (a)Distributed
DistributedOFS
OFS(DOFS)
(DOFS)fiber
fiberand
and(b)
(b)ODiSI-6000
ODiSI-6000model
modelOptical
OpticalBackscatter
BackscatterReflectome-
Reflec-
tometer (OBR) interrogator manufactured by LUNA Technologies.
ter (OBR) interrogator manufactured by LUNA Technologies.

Despite such, a discrete variety of applications have been performed including

bridges, dams, tunnels, pipelines and slopes. The goal of the present article is to shine a
spotlight on the most recent of such DOFS SHM applications and help the collective
growth towards an efficient SHM by collecting in a single place these recent efforts, ad-
vancements and findings. In particular, as mentioned in Section 1, the reported DOFS
SHM applications include laboratory experimentation, the built environment (bridges,
buildings, roads, etc.), geotechnical constructions, tunnels, pipelines and wind turbines,
all performed in the time frame 2017–2020.

3. DOFS Coatings and Bonding Adhesives: The Strain Transfer Process and the Opti-
mal Bonding Technique
Ideally, the strains measured by a DOFS silica core should be equivalent to the actual
strains present in the host structure on which the former is bonded. If this requirement
 Sensors 2021, 21, 1818 7 of 83

The distributed nature of DOFS enables the mapping of temperature, strain and
vibration distributions in two or even three dimensions (achievable with a DOFS mesh de-
ployment) and their identification at any point along a fiber. Combined, these aspects allow
for the painting of a clear picture of the global behavior of a structure rather than reporting
the tensile state of a limited number of points. Figure 3 also displays the hair-like thinness
of the fiber which makes it ideal for un-intrusive deployments. Still, as suggested by Bar-
rias [1], this is still a recent and developing technology as can be perceived by the relatively
scarce number of DOFS applications in SHM projects published in research articles.
Despite such, a discrete variety of applications have been performed including bridges,
dams, tunnels, pipelines and slopes. The goal of the present article is to shine a spotlight
on the most recent of such DOFS SHM applications and help the collective growth towards
Sensors 2021, 21, x FOR PEER REVIEW 8 of 87
an efficient SHM by collecting in a single place these recent efforts, advancements and
findings. In particular, as mentioned in Section 1, the reported DOFS SHM applications
include laboratory experimentation, the built environment (bridges, buildings, roads, etc.),
sensing fiber [19].constructions,
geotechnical Instead, sometunnels,
of the energy is converted
pipelines and windto the shear
turbines, alland normal in
performed defor-
the time
mations
frameof 2017–2020.
the intermediate layers (as well illustrated by Her and Huang [40] in Figure 4)
thus leading to strain transfer errors.
3. DOFS
The readerCoatings
should and Bonding
consider Adhesives:
that Figure 4, for The Strain
the sake of Transfer Process
simplicity, and the
only represented
Optimal
a single coatingBonding Technique
layer between the fiber core and the adhesive but in numerous applica-
tions, theIdeally, the strains
intermediate layersmeasured
are manybymore,a DOFSi.e.,silica core should
cladding, be equivalent
reinforcement layers,to the actual
diverse
coatings, jackets, etc. The whole issue is exacerbated in the presence of strain singularities was
strains present in the host structure on which the former is bonded. If this requirement
zonesfully
wheremeta then both DOFS
high-stress structuralismonitoring
concentration accumulated and laboratory
over structural
a short length of DOFS,investigation
such
would simply boil down to analyzing the OBR’s output data, no
as cracks in RC elements or structural discontinuities. Because the optical fiber bridges matter how minute
the the
object
cracks of investigation
irrespective is orthey
of whether howareaccurate the measurement
embedded requirements
or surface adhered, in orderare.
toUnfortunately,
eliminate
this is not always the case. Indeed, the packaging and the installation
this error and guarantee effective measurements, investigation on the strain transfer of the sensors involves
intermediate layers (i.e., protective coatings and adhesive layers) whose
mechanism from the substrate material (steel, concrete, etc.) to the fiber core is required presence leads to
[19,41]. Alj et al. [42], for example, indicated that the strain transfer process depended on [19].
the failure of the complete strain transfer from the host material to the sensing fiber
Instead,and
mechanical some of the energy
geometrical is converted
properties to the shear
of the different and normal
intermediate deformations
layers, of the
i.e., their elas-
intermediate layers (as well illustrated
tic modulus, their height and the initial crack opening.by Her and Huang [40] in Figure 4) thus leading to
strain transfer errors.

Figure
Figure Free diagram
4. body
4. Free body diagram
for the for the symmetrical
symmetrical section
section of of thefiber
the optical optical
andfiber and the substrates
the substrates
together
together with with
their their relative
relative shearshear transfer
transfer (adapted
(adapted fromfrom
[40]).[40]).
The reader should consider that Figure 4, for the sake of simplicity, only represented a
Alj et al. continued reporting that several analytical models had been proposed in the
single coating layer between the fiber core and the adhesive but in numerous applications,
literature among which the one of Feng et al. [41] who developed a theoretical approach
the intermediate layers are many more, i.e., cladding, reinforcement layers, diverse coatings,
to describe the strain transfer process in zones containing strain singularities (cracks) and
the one of Henault et al. [43] who defined a Mechanical Transfer Function (MTF) that re-
lates the strain profile along the DOFS core to the strain profile of the host cracked con-
crete. The above analyses become especially relevant when considering the need of add-
ing additional protective layers around a DOFS core. Indeed, whilst the ideal DOFS would
Sensors 2021, 21, 1818 8 of 83

jackets, etc. The whole issue is exacerbated in the presence of strain singularities zones
where a high-stress concentration is accumulated over a short length of DOFS, such as
cracks in RC elements or structural discontinuities. Because the optical fiber bridges
the cracks irrespective of whether they are embedded or surface adhered, in order to
eliminate this error and guarantee effective measurements, investigation on the strain
transfer mechanism from the substrate material (steel, concrete, etc.) to the fiber core
is required [19,41]. Alj et al. [42], for example, indicated that the strain transfer process
depended on mechanical and geometrical properties of the different intermediate layers,
i.e., their elastic modulus, their height and the initial crack opening.
Alj et al. continued reporting that several analytical models had been proposed in the
literature among which the one of Feng et al. [41] who developed a theoretical approach to
describe the strain transfer process in zones containing strain singularities (cracks) and the
one of Henault et al. [43] who defined a Mechanical Transfer Function (MTF) that relates
the strain profile along the DOFS core to the strain profile of the host cracked concrete. The
above analyses become especially relevant when considering the need of adding additional
protective layers around a DOFS core. Indeed, whilst the ideal DOFS would be one with the
least and thinnest cladding/coating layer possible, in practice, considering the challenging
environments where the fiber could be deployed, most commercial DOFS cables include
additional coating layers that protect the core against mechanical and chemical damages.
Furthermore, the high-stress concentration transmitted to a DOFS by a crack opening may
lead to its premature failure leading to the loss of data from that instance on.
Whenever an experimental investigation requires the maximum precision possible,
the shear lag induced by any host-surface/DOFS-core intermediate layers should be
minimal, thus the preferable fiber is a simply cladded non-coated DOFS. In this case,
though, the protective function against external damages falls entirely on the adhesive
layers. Numerous experimental campaigns have been performed in the latest year to assess
which is the optimal adhesive for any specific function, i.e., bonded to a structure’s surface
or embedded inside the structure itself bonded to rebar, for instance.
As can be evinced, the optimal DOFS claddings/coatings and bonding techniques are
still elusive concepts and, therefore, still under heavy investigation. The authors believe
in the importance of reporting the recent discoveries and discussions on the present topic
first as it inherently influences every DOFS monitoring campaign henceforth.
Wang et al. [19] recently developed a state-of-the-art review on the strain transfer
theory of optical fiber-based sensors developed for Civil structures. In particular, the article
addresses the issues of strain not entirely transferred from the host material to the sensing
fiber, resulting in a portion of strain loss in the transferring path and the Strain Transfer
Theory produced to tackle this issue. Interestingly the strain transfer theory was reported
for both embedded and surface-attached sensors. The authors concluded by mentioning
that the Strain Transfer Theory could be adopted to conduct accurate interpretations of the
collected data, to analyze the influence of the installed sensors on the structural integrity
and strength and to improve the design and calibration codes of the sensors for the long-
term, effective and accurate monitoring.
Bassil et al. [44] studied the above-mentioned strain transfer mechanism and proposed
a new analytical model which assumes the absence of a perfect bond between the compo-
nents of a multilayer system. Particularly, a general expression of the crack-induced strain
transfer from fractured concrete material to an optical fiber (see Figure 5a,b) is established
and put to the test through wedge splitting tests on concrete specimens instrumented with
multiple embedded and surface-mounted fiber optic cables.
Furthermore, the authors summarized the modern-day expertise on DOFS cables:
• Different types of optical cables are available on the market;
• Some cables are conceived to be embedded inside the structure during construction,
whilst others are more suitable for surface installation;
• The cables, constituted of different materials and shapes, lead to a different strain
transfer response due to the shear lag effect in the intermediate layers;
 mentioning that the Strain Transfer Theory could be adopted to conduct accurate inter-
pretations of the collected data, to analyze the influence of the installed sensors on the
structural integrity and strength and to improve the design and calibration codes of the
sensors for the long-term, effective and accurate monitoring.
Sensors 2021, 21, 1818 9 of 83
Bassil et al. [44] studied the above-mentioned strain transfer mechanism and pro-
posed a new analytical model which assumes the absence of a perfect bond between the
components of a multilayer system. Particularly, a general expression of the crack-in-
•
duced Little
straininformation is available
transfer from fracturedon concrete
the mechanical properties
material of the cable
to an optical fiber and
(seeits con- 5a,b)
Figure
stituent elements;
is established and put to the test through wedge splitting tests on concrete specimens in-
• Little information is also available on the long-term behavior and performance of
strumented with multiple embedded and surface-mounted fiber optic cables.
these cables.

(c)
Figure 5. Crack detection using DOFS techniques: (a) Comparison of the traditional sensors, (b) strain transferring between
layers and (c) the tested cables (adapted from [44]).

In response to some of the above lackings, the authors used in their test multiple
kinds of OFDR DOFS cables (see Figure 5c). The experimental results reportedly estimated
correctly the crack openings and led to successful monitoring of the optical cable responses
through the strain lag parameter. A final key observation worth reporting here was the
Polyimide-cladded cables’ strain lag parameters λ varying around 260 m−1 whilst the
others between 20 m−1 and 60 m−1 . This meant that the Polyimide fiber reported crack-
induced strain distributions 5 to 15 times higher than the other tested cables, which allowed
for the accurate detection of tiny crack propagations.
Zhang et al. [45] proposed a mechanical model (further validating with experimenta-
tion) to study the effects of multiple closely spaced cracks on the DOFS measurements and
the effects of the cable properties (i.e., shear stiffness between cable and fiber, cable radius,
elastic modulus, interface cohesion) on the shape of the DOFS strain distributions across
cracks. The reported spring model was constituted by two layers, namely an inner one
(representing the fiber core and cladding) and an outer one (representing the outer coating,
primary layers and a jacket) with assigned relative stiffness. A noteworthy result was that,
Sensors 2021, 21, 1818 10 of 83

by independently varying the stiffnesses in the spring model, the axial stiffness of the
coating layer was reported to have a negligible influence on the shape of strain distribution.
This was the opposite for the coatings shear stiffness and shear parameter.
Alj [46] developed his doctoral thesis in the framework of the ODOBA (Observatory of
the Durability Of reinforced concrete structures) project launched by IRSN in 2016 in France
which was aimed at studying the concrete pathologies and their consequences for nuclear
power plant structures. In particular, the thesis was redacted with the objective of studying
the effectiveness of DOFS technology to detect and monitor concrete pathologies potentially
met in nuclear power plants (Delayed Ettringite Formation, Alkali-Silica Reaction or a
combination of the two). As such, the work also undertook the interpretation of long-term
strain measurements provided by OF sensors in degrading environments.
The above project spawned a research article on the topic of the durability of DOFS [47].
Here, the authors studied the durability of OFDR DOFS cables for the SHM of Reinforced
Concrete (RC) structures through an experimental campaign based on the accelerated aging
of small concrete specimens instrumented with two embedded optical fiber cables. The first
included an external coating made of a soft elastomeric material (olefin-type elastomer)
containing two single-mode OFs and two metal rods for mechanical protection. The second
one included an outer layer, made of polyethylene with six optical fibers (single-mode and
multimode) which are wound around a central rod (reinforcement part of the cable) and
embedded in an elastomeric layer. In order to simulate the effect of alkaline pore solution
of the concrete medium, the accelerated aging was achieved immerging the specimens
in an alkaline solution: c(KOH) = 0.5 mol/L + c(NaOH) = 0.1 mol/L, at three different
temperatures (20 ◦ C, 40 ◦ C and 60 ◦ C) for three months. Pull-out tests were consequently
carried out in order to evaluate the effect of the accelerated aging on the mechanical
response of the cable/concrete interfaces. The results showed that the interface between
the first cable and the concrete had not been degraded, contrariwise, the second cable’s
interface was damaged and the degradation was emphasized by temperature.
The results of the previous test are a demonstration of the large experimental effort
still required by the scientific community to fully dominate the DOFS-powered SHM
monitoring science.
Moving on to the second topic of the present section, specifically the research of the
optimal adhesive and bonding technique for the RC structures instrumentation with non-
coated OFDR DOFS. Bado et al. [48] asserted that despite DOF’s potential for the detection
of deformations and cracking in RC structures, in numerous published DOFS-monitored
experimental campaigns, researchers reported the appearance of unrealistic strain readings,
defined as Strain Reading Anomalies (SRAs), which prevented the extraction of reliable
strain data. Later, the article outlined the results obtained through an experimental test
aimed at inducing such SRAs in an isolated and controlled environment in order to isolate
and identify the physical cause of their origin. The test consisted of gradually bending
seven steel reinforcement bars with DOFS bonded to their bottom side with different
adhesives, namely cyanoacrylate, silicone, and a combination of the two. Additionally, one
of the rebars was instrumented with Strain Gauges (SG) for comparison purposes. Three of
the DOFS instrumented rebars were transversally incised (as visible in Figure 6) in order to
simulate the typical post-cracking strain transmission between concrete and steel.
 isolate and identify the physical cause of their origin. The test consisted of gradually bend-
ing seven steel reinforcement bars with DOFS bonded to their bottom side with different
adhesives, namely cyanoacrylate, silicone, and a combination of the two. Additionally,
one of the rebars was instrumented with Strain Gauges (SG) for comparison purposes.
Sensors 2021, 21, 1818
Three of the DOFS instrumented rebars were transversally incised (as visible in Figure 6)
11 of 83
in order to simulate the typical post-cracking strain transmission between concrete and
steel.

Figure 6. Test
Test setup:
setup: three-dimensional
three-dimensional specimen
specimen detailing
detailing and
and photography
photography [48].
[48].

These incisions further had varying steepness degrees in order to simulate simulate different
different
transmission lengths. Finally, different
different loading speeds were applied (rapid/impact and
loading speeds were applied (rapid/impact and
slow monotonic
monotonicone).
one).The
Theresults
resultsshowed
showed the bonding
the bonding adhesives
adhesiveshaving an influence
having on the
an influence on
DOFS performance
the DOFS but not
performance buton
notthe
onrise
theofrise
anomalies. Expectedly,
of anomalies. considering
Expectedly, the difference
considering the dif-
between the modulus
ference between the of elasticityof
modulus of elasticity
silicone andof cyanoacrylate, the former seemed
silicone and cyanoacrylate, thetoformer
delay
the appearance of SRAs to the expense of the strain measurement precision in addition
to redistributing the strains along the fiber. Oppositely, the rigid but fragile nature of the
cyanoacrylate reportedly provided the results with superior precision but at the expense
of the adhesive resistance and endurance. The combination of the two provided a hybrid
behavior. Finally, the reasons triggering the SRAs were narrowed down to a strain threshold
and to the friction with concrete present whenever the DOFS is embedded inside an RC
structure.
This last conclusion was later demonstrated to be the main reason for the rise in
SRAs in Bado et al. [49]. The article was aimed at providing a contribution to the effort of
establishing the best thin non-coated OFDR DOFS/rebar bonding technique whenever the
instrumented bar is to be embedded inside of RC structures. The investigated bonding
techniques were:
• DOFS positioned inside a groove and bonded with cyanoacrylate;
• DOFS glued to the rebar’s surface with cyanoacrylate with a protective layer of silicone
on top;
• DOFS positioned inside a groove, bonded with cyanoacrylate and further covered
with silicone.
The authors also indicated as a crucial requirement the absence of any influence of the
bonding technique on the DOFS sampled strains, thus, the guarantee that what is actually
reported by the fiber are the actual strains occurring in the monitored surface rather than
 strumented bar is to be embedded inside of RC structures. The investigated bonding tech-
niques were:
• DOFS positioned inside a groove and bonded with cyanoacrylate;
• DOFS glued to the rebar’s surface with cyanoacrylate with a protective layer of sili-
cone on top;
Sensors 2021, 21, 1818 • DOFS positioned inside a groove, bonded with cyanoacrylate and 12 of 83 covered
further
with silicone.
The authors also indicated as a crucial requirement the absence of any influence of
the bonding technique on the DOFS sampled strains, thus, the guarantee that what is ac-
a strain present solely in the fiber.
tually reported by theThe first
fiber arephase of the
the actual testoccurring
strains (performed onmonitored
in the a bare rebar)
surface rather
assessed that the presence
than a strain present solely in the fiber. The first phase of the testno
of a layer of silicone and/or the one of a groove had influenceon a bare
(performed
on the extrapolated DOFS
rebar) when
assessed compared
that the presence to readings performed
of a layer of withthe
silicone and/or a DOFS
one of simply
a groove had no
bonded to the rebar with cyanoacrylate. This was proved to be valid
influence on the extrapolated DOFS when compared to readings performed both in tension andwith a DOFS
in bending. The second phase of
simply bonded to the
the test
rebarsawwiththe embedmentThis
cyanoacrylate. of the
wasinstrumented rebars
proved to be valid both in ten-
sion and in bending. The second phase of the test saw the embedment
inside a prism of concrete to form multiple RC tensile elements (also named RC ties). These of the instrumented
RC specimens were rebars inside
of two a prism
kinds, of concrete
cracking and to form multiple
non-cracking RC designed
ones tensile elements (also to
expressly named RC
study these two different scenarios. Interestingly, during this second test phase, the authorsdesigned
ties). These RC specimens were of two kinds, cracking and non-cracking ones
expressly
were able to monitor to study
the steel rebars’ these two well
strains different scenarios.
beyond the yieldInterestingly, during by
limit as testified thisthe
second test
phase, the authors were able to monitor the steel rebars’ strains well beyond the yield limit
peaks of Figure 7 which are the physical representation of the rebar’s different sections
as testified by the peaks of Figure 7 which are the physical representation of the rebar’s
slowly plasticizing one after the other.
different sections slowly plasticizing one after the other.

FigureFigure
7. Strain7.profiles
Strain of a 15 × 15
profiles of ×a 210
15 ×mm
15Ø20 Reinforced
× 210 mm Ø20Concrete (RC) Concrete
Reinforced tie test with the tie
(RC) DOFStestbonded with
with the a combi-
DOFS
nation of groove, cyanoacrylate and silicone [49].
bonded with a combination of groove, cyanoacrylate and silicone [49].
Additionally, as also visible in Figure 7, the oscillations of the profiles, easily confus-
Additionally, as also visible in Figure 7, the oscillations of the profiles, easily confusable
able with signal noise, were stated to be the influence of the transversal rib on the tensile
with signal noise, were stated to be the influence of the transversal rib on the tensile
performance of the rebar. Additionally, it was observed on all the specimens that the
magnitude of the defined rib-induced noise became increasingly larger (see Figure 8) with
the load thus increasingly disturbing the sought-after signal, that is the strain profile.
The authors continued reporting an unexpected phenomenon, possibly connected
to the under-performance of the bonding techniques, which caused an alteration of the
DOFS strain readings whenever the fiber was in the presence of extra bending-induced
tension (smoothing out strain profiles) and/or monitoring strains in cracked cross-sections
(magnifying their amount). Additional work on the involved transfer mechanisms could
be the key to such phenomena. Finally, it was asserted that, whilst none of the studied
bonding techniques could completely guarantee the removal of all SRAs, the combination
of a longitudinally incised groove, cyanoacrylate and silicone was the bonding technique
that led to the smallest amount of SRA thus preserving the readability and reliability of the
strain data (see Figure 9).
 tension (smoothing out strain profiles) and/or monitoring strains in cracked cross-sections
(magnifying their amount). Additional work on the involved transfer mechanisms could
be the key to such phenomena. Finally, it was asserted that, whilst none of the studied
bonding techniques could completely guarantee the removal of all SRAs, the combination
Sensors 2021, 21, 1818 of a longitudinally incised groove, cyanoacrylate and silicone was the bonding technique
13 of 83
that led to the smallest amount of SRA thus preserving the readability and reliability of
the strain data (see Figure 9).

Face A_Bare rebar: w/o groove Face A_Concrete: w/o groove

40 40

20 20
]

]
Noise [

Noise [
0 0

-20 -20

-40 -40
40 60 80 100 120 140 160 180 240 260 280 300 320
DOFS coordinates [mm] DOFS coordinates [mm]

Load 1 = 17.0 kN Load 2 = 64.4 kN Load 3 = 116.0 kN

Face B_Bare rebar: w/o groove Face B_Bare rebar: w/ groove Face B_Concrete: w/ groove
40 40 40

20 20 20
]

]
Noise [

Noise [

Noise [
0 0 0

-20 -20 -20

-40 -40 -40

20 40 60 80 100 120 130 140 150 160 170 180 190 200 220 240 260 280 300
DOFS coordinates [mm] DOFS coordinates [mm] DOFS coordinates [mm]

80 Face A_Bare rebar: w/o groove

Load 2 Load 3 Yielding load
Average strain reading

Face A_Concrete: w/o groove

Face B_Bare rebar: w/o groove
60 Face B_Bare rebar: w/ groove
noise [ ]

Face B_Concrete: w/ groove

40
Load 1
20

0
0 20 40 60 80 100 120 140 160
Load [kN]
Sensors 2021, 21, x FOR PEER REVIEW 14 of 87
Figure8.8.Influence
Figure Influenceofofthe
therib
ribpattern
patternon
onthe
thesteel
steelstrain
strainprofiles
profilesduring
duringthe
thedouble
double pull-out
pull-out test
test ofof the
the RCRC
tietie
1515
× ×1515××240
240
mm Ø20 (top five figures). Relative average reading noise (bottom)
mm Ø20 (top five figures). Relative average reading noise (bottom) [49]. [49].

Figure
Figure9.9.Total
Totalamount
amount of
of anomalistic
anomalistic readings
readings grouped
grouped per
per bonding
bonding technique
technique [49].
[49].

Barrias et al. [50] reported the results of a laboratory test campaign on two 150 × 150
× 600 mm RC beams whose bottom face was instrumented with four differently bonded
longitudinal OFDR DOFS segments, namely with Silicone, Polyester, Epoxy and Cy-
anoacrylate. The beams were subjected to three-point load tests carried out under an elas-
tic regime (uncracked condition) before being, in a second cycle, cracked and loaded to
failure. The influence of the spatial resolution on the data acquisition was additionally
Sensors 2021, 21, 1818 14 of 83

Barrias et al. [50] reported the results of a laboratory test campaign on two 150 × 150 ×
600 mm RC beams whose bottom face was instrumented with four differently bonded longi-
tudinal OFDR DOFS segments, namely with Silicone, Polyester, Epoxy and Cyanoacrylate.
The beams were subjected to three-point load tests carried out under an elastic regime
(uncracked condition) before being, in a second cycle, cracked and loaded to failure. The
influence of the spatial resolution on the data acquisition was additionally analyzed. On
the latter, the spatial resolution of 1 mm presented an undesired high spatial variability
(reportedly due to the concrete’s heterogeneity of the aggregates) which did not occur with
a spatial resolution of 1 cm. On the adhesive performance, in uncracked conditions, all
adhesives compared fairly well with the data measured by parallelly installed SG with a
slight edge from the silicone’s part whose plots were smoother and more uniform. When
cracks arose, whilst all the segments were able to detect and locate the crack formation,
the measurements at this location became erratic in all segments except for the silicone.
Nevertheless, for the latter, the length influenced by the crack formation is relatively and
suspiciously wider.
The last observation was very probably due to the aforesaid [48] silicone-induced
stress redistribution.
Alj et al. [42] investigated the strain response of thick-coated DOFS sealed in a groove
on the surface of a concrete structure using a polymer adhesive and aimed to identify
optimal conditions for crack monitoring. Initially, a Finite Element Model was developed
(see Figure 10) to understand the strain transfer process between the DOFS core and the
host structure which showed that the softer the adhesive is, the lower the strain transferred
Sensors 2021, 21, x FOR PEER REVIEW 15 of 87
to the DOFS glass core due to redistribution over a broader length of the sensor (similarly
to what was observed in [48–50]).

Figure
Figure10. Finite element
10. Finite elementmesh
meshofof(a)(a) the
the 3D3D geometry
geometry andand
(b) (b) the cross-sectional
the cross-sectional viewview
of theofsys-
the
tem [42].
system [42].

Later,
Later,mechanical
mechanicalteststestsononcylindrical
cylindricalconcrete
concretespecimens
specimensinstrumented
instrumentedwith withaapoly-
poly-
imide
imide coated (external diameter 160 μm) OFDR DOFS bonded/sealed using severaladhe-
coated (external diameter 160 µm) OFDR DOFS bonded/sealed using several adhe-
sives
sivesproved
provedthat strain
that measurements
strain measurements provided by DOFS
provided by DOFSwerewere
consistent with those
consistent with from
those
conventional
from conventionalsensors. Furthermore,
sensors. they confirmed
Furthermore, that bonding
they confirmed DOFSDOFS
that bonding to thetoconcrete
the con-
structure using using
crete structure soft adhesives allowed
soft adhesives to mitigate
allowed thethe
to mitigate amplitude
amplitude of of
local strain
local strainpeaks
peaks
induced by crack openings, which may prevent the sensor from early
induced by crack openings, which may prevent the sensor from early breakage. Unfortu-breakage. Unfortu-
nately,
nately,this
thisparticular
particularresult,
result,also
alsocorroborated
corroboratedby bythe
theprevious
previousarticles,
articles,cuts
cutsboth
bothways
waysasas
the strain alleviation reduces the sensitivity of bonded DOFS and may
the strain alleviation reduces the sensitivity of bonded DOFS and may alter specificallyalter specifically
sought
soughtafter
afterstrain
strainreadings.
readings.This
Thiswas
wasevident
evidentininthe
thestrain
strainprofiles
profilesprovided
providedby bythe
theDOFS
DOFS
(see
(see Figure 11) which was bonded to the test specimen with three polymer adhesivesused
Figure 11) which was bonded to the test specimen with three polymer adhesives used
in this experiment (Bi-component epoxy, Mastic silicone and Silicone rubber) and sub-
jected to an actuator displacement reaching 0.6 mm.
 sives proved that strain measurements provided by DOFS were consistent with th
from conventional sensors. Furthermore, they confirmed that bonding DOFS to the
crete structure using soft adhesives allowed to mitigate the amplitude of local strain pe
induced by crack openings, which may prevent the sensor from early breakage. Unfo
Sensors 2021, 21, 1818 nately, this particular result, also corroborated by the previous articles,15cuts
of 83both way
the strain alleviation reduces the sensitivity of bonded DOFS and may alter specific
sought after strain readings. This was evident in the strain profiles provided by the D
(see Figure 11) which was bonded to the test specimen with three polymer adhesives u
in this experiment
in(Bi-component
this experimentepoxy, Mastic silicone
(Bi-component and
epoxy, Silicone
Mastic rubber)
silicone andand subjected
Silicone rubber) and
to an actuator displacement reachingdisplacement
jected to an actuator 0.6 mm. reaching 0.6 mm.

Figure 11. Strain11.

Figure profiles
Strainrecorded
profilesbyrecorded
DOFS bonded to the
by DOFS concrete
bonded tospecimen with specimen
the concrete three soft adhesives
with three(Bi-component
soft
epoxy, Mastic silicone and Silicone rubber) [42].
adhesives (Bi-component epoxy, Mastic silicone and Silicone rubber) [42].

The DOFS bonded with Bi-component epoxy revealed the development of a strain
peak in the central part of the profile from the actuator displacement of 0.2 mm. DOFS
bonded with the two silicone adhesives, instead, did not show clear peaks development
but rather small triangular shaped profiles visible only from the actuator displacement
of 0.3 mm. Indeed, as already observed in [48,50], the very low stiffness of these silicone
adhesives (about 5 times lower than that of the Bi-component epoxy) favored the diffusion
of the originally concentrated stress over the entire length of the bonded DOFS. Conse-
quently, the authors suggested that the adhesive stiffness should be optimized to ensure
a compromise between protective function and crack monitoring performance. Alj et al.
finally generalized the above FEM to describe the strain response of bonded DOFS in the
presence of crack and an analytical expression relating a DOFS peak strain to the crack
opening was proposed.
In Barrias et al. [51], the authors aimed at assessing OFDR DOFS’ ability and per-
formance to monitor bridges on a long-term basis thus subjected to a large number of
load cycles. The study also studied the magnitude of DOFS debonding and fatigue failure
risk. To this end, two similar specimens were submitted to a fatigue test up to 2 million
load cycles, the amplitude of which was representative of what is expected on a standard
highway bridge under vehicular traffic. According to the authors, the first beam was
fatigue loaded in an un-cracked state and its results showed a reasonably good agreement
between the DOFS measurements, the theoretical results and the ones measured by the
SG as well and for all adhesives tested. The second beam, instead, was intentionally
pre-cracked and its results showed how DOFS continued to provide strain measurements
coherent with the applied load despite some differences related to the used bonding agents.
For example, the best performance was obtained for the epoxy and cyanoacrylate, whose
measurements matched perfectly with those from the SG. Larger differences were, instead,
observed against the other two bonding agents. Overall, though, neither fatigue failures
nor fiber/adhesive debonding were detected.
Conclusively, the present section introduced several recently published results and
observations on the matters of strain transfer in the multi-layered DOFS bonding systems
Sensors 2021, 21, 1818 16 of 83

and of the ideal bonding technology for laboratory-targeted coating-less DOFS. The main
takeaways on the former were:
• Feng et al. [41] and Henault et al. [43] developed theoretical approaches to describe
the strain transfer mechanism (also defined Mechanical Transfer Function) that relates
the strain profile of the host cracked concrete with the DOFS core;
• Bassil et al. [44] proposed a new analytical model formulating a general expression of
the crack-induced strain transfer from fractured concrete material to optical fibers;
• Alj et al. [42] indicated that the strain transfer process depended on mechanical and
geometrical properties of the different intermediate layers, i.e., their modulus of
elasticity, their height and the initial crack opening;
• Articles [44,52] presented particularly varied selections of DOFS cables with equally
deep studies of their performances;
• Whenever an experimental investigation requires the maximum precision possible, the
strain shear lag induced by any host–surface/DOFS-core intermediate layers should
be minimal, thus the preferable fibers are simply cladded non-coated DOFS;
• It was experimentally assessed that simple Polyimide-cladded cables had a strain
lag parameter inferior to other coating-including ones, thus allowing for accurate
detection of tiny crack propagations;
• On the durability of DOFS cables embedded inside RC structures, an experimental
campaign with accelerated aging of the RC saw a larger DOFS/concrete interface
degradation for cables with an external coating made of a polyethylene than the ones
made out of soft elastomeric material (olefin-type elastomer).
On the best bonding/protection techniques of thin coating-less fibers on RC structures:
• Numerous researchers reported the appearance of unrealistic strain readings, defined
as Strain Reading Anomalies (SRAs), which prevented the extraction of reliable strain
data;
• Whenever DOFS are embedded inside an RC structure, SRAs are mainly caused by
the friction with concrete or (in the case of externally bonded fibers too) by steep strain
gradients like the ones occurring in the proximity of surface non-linearities, such as
cracks;
• Whilst it could not completely guarantee the absence of SRAs, the combination of a
longitudinally incised groove, CYN and SI was reportedly the bonding technique that
led to the smallest amount of SRAs;
• It was observed on DOFS instrumented RC ties that load increments increased the
magnitude of the influence of the ribs on the strain profile of steel rebars;
• Several DOFS-powered experimental campaigns detected a behavioral feature of sili-
cone whenever it was used for the protection of DOFS. That is a certain redistribution
of the strains present in a specific point/section of the DOFS along a larger segment,
consequently sampling lower strains than the ones originally present;
• This is in line with the results of a performed FEM analysis stating that the softer the
adhesive is (in this case silicone can be characterized as a “soft” adhesive), the lower
the strain transferred to the DOFS glass core due to redistribution over a broader
length of the sensor;
• Following a test on DOFS-instrumented RC beams subjected to fatigue testing (2
million load cycles) the most resilient bonding adhesives were reportedly epoxy and
cyanoacrylate.

4. Laboratory Experimental Applications

The application of DOFS for structural monitoring is still rather young, especially
when compared to classic tools such as SG, LVDT and inclinometers. It can be stated
that an “instrumental maturity” is still lacking when it comes to optical fiber sensing,
a shortcoming that numerous research groups and projects are attempting to overcome
through laboratory experimentation. Indeed, most research articles published in the latest
year still motivate their experimental efforts with the study of the applicability, suitability
Sensors 2021, 21, 1818 17 of 83

and performance of DOFS deployments on any particular structure, under any particular
testing conditions on the lookout for any particular mechanical response. Furthermore, it
is common to see a comparison between DOFS’ outputs and parallelly sampled ones by
means of common and well-established monitoring tools. The most common variables
under investigation are the following:
• DOFS cladding, coating and reinforcements
• Adhesives or technologies used to achieve the DOFS/structure bonding and protection
• The layout of the deployment
• Positioning of the DOFS: bonded to a structure’s surface versus embedded inside it
Considering all the above variables, the solution for the best DOFS deployment is
branching and permuting in nature. Consequently, over the years, multiple techniques
have been proposed which will eventually “converge” in a single one in the close future, in
accordance with the objective at hand. Only under such conditions, it would be possible to
develop standards and guidelines for DOFS monitoring.
The present section is aimed at presenting and putting side to side the laboratory
experimental efforts performed in the latest years in an attempt to help to hasten the
above-mentioned “instrumental converge”.
It is worth mentioning, as will be noticeable further on, the large number of laboratory
experimental campaigns that integrate DOFS monitoring with a novel technology under
the name of Digital Image Correlation (DIC). Briefly, DIC is an innovative, cheap, non-
contact optical technique for measuring strains and displacements on any primed surface.
It works by comparing digital photographs of a component or test piece at different
deformation stages and, by tracking blocks of pixels of a complex stochastic pattern
painted on the monitored surface, it measures its surface displacement, building 2D and 3D
deformation vector fields and strain maps [53]. As it is easily discernible, the integration
of this technique inside a DOFS monitoring campaign not only does not lead to any
interference but also allows to integrate, verify and validate the DOFS internally measured
strains with externally measured ones. Two recent studies, Mata-Falcón et al. [54] and Bado
et al. [55], explicitly analyzed this DOFS-DIC integration.
The former explored the combined use of DIC and OFDR DOFS for the strain mea-
surements of two 600 × 600 × 125 cm reinforced panels subjected to diagonal tension. The
OFDR DOFS were glued with epoxy inside 1 × 1 mm grooves carved along both legs of
every stirrup, summing up to 24 measuring lines across the whole specimen. As visible in
Figure 12, the DOFS-sampled reinforcement strains correlated well with the position and
opening of the cracks (monitored by the DIC) proving the compatibility and potential of
DIC and DOFS to measure phenomena like bond, crack behavior and shear transfer.
 Sensors
Sensors 21,21,
2021,
2021, 1818PEER REVIEW
x FOR 18 87
19 of of 83

Figure
Figure 12.12. Crack
Crack locations
locations andand openings,
openings, steel
steel stresses
stresses andand forces
forces inin frontlayers
front layersofofstirrups
stirrupsfor
for Panel
Panel
1 at 1(a)
at 80
(a)kN
80 kN
andand (b) 110
(b) 110 kNwell
kN as as well as Panel
as for for Panel
2 at 2(c)at90
(c)kN
90 and
kN and (d) 150
(d) 150 kN [54].
kN [54].

The same
4.1. Reinforced conclusions
Concrete were also
Beams: Cracking andreached in Bado et al.
Other Serviceability [55] who monitored with
Analyses
OFDR DOFS the strains present inside three steel rebars embedded in differently
The present section collects the large research effort performed throughout theshaped
last
and differently behaving (cracking and non-cracking) RC ties during testing. As already
years on DOFS distributed strain sensing for RC beams experimental campaigns.
evident, this dual monitoring methodology is often employed for experimental works on
Berrocal et al. [59], for example, aimed at assessing the suitability and accuracy of
cracking RC structures.
thin polyamide -ladded OFDR DOFS for the crack monitoring of RC beams subjected to
Therefore, with this frame of mind, the following sub-sections will explore recent
three-point bending tests (monotonic and cyclic loading) in addition to determining the
research articles on laboratory experimental campaigns deploying DOFS to RC structures
crack location and width evolution over time. The latter was achieved by means of com-
and more. They will be grouped in different sections exploring DOFS deployments on
parison between DOFS reinforcement strain profiles and externally sampled DIC meas-
RC beams, RC ties, RC slabs, corrosion tests and finally on miscellaneous experimental
urements. Low-bend loss polyamide-coated DOFS (155 μm diameter) were bonded di-
campaigns such as fire resistance, concrete–textile hybrids and more.
rectly to the rebar’s surface, along one of the longitudinal ridges, with cyanoacrylate ad-
Finally, additional reading on laboratory applications of DOFS can be found in Huo
hesive before being protected by a silicone layer. The comparison of the crack location
et al. [56], Bassil [57], and Zhang [58].
obtained from DOFS and DIC revealed that most of them were located within an accepta-
ble4.1.
range of ±3 cm
Reinforced from each
Concrete other
Beams: (as visible
Cracking in Figure
and Other 13).
Serviceability Analyses
The present section collects the large research effort performed throughout the last
years on DOFS distributed strain sensing for RC beams experimental campaigns.
Sensors 2021, 21, 1818 19 of 83

Berrocal et al. [59], for example, aimed at assessing the suitability and accuracy of thin
polyamide -ladded OFDR DOFS for the crack monitoring of RC beams subjected to three-
point bending tests (monotonic and cyclic loading) in addition to determining the crack
location and width evolution over time. The latter was achieved by means of comparison
between DOFS reinforcement strain profiles and externally sampled DIC measurements.
Low-bend loss polyamide-coated DOFS (155 µm diameter) were bonded directly to the
rebar’s surface, along one of the longitudinal ridges, with cyanoacrylate adhesive before
being protected by a silicone layer. The comparison of the crack location obtained from
DOFS
Sensors 2021, 21, x FOR and DIC revealed that most of them were located within an acceptable range of
PEER REVIEW 20 of 87
±3 cm from each other (as visible in Figure 13).

Figure Figure
13. Multiple
13. strain profiles
Multiple from
strain the DOFS
profiles where
from thethe red triangular
DOFS where the makers indicate themakers
red triangular determined crack position
indicate the
based on the strain profile at maximum load and the grey shaded area correspond to position determined from the Digital
determined crack position based on the strain profile at maximum load and the grey shaded area
Image Correlation (DIC) [59].
correspond to position determined from the Digital Image Correlation (DIC) [59].
This range is understandable considering the different depths at which the transver-
This range issal
understandable considering
cracks were measured the different
and their probable depths
lack of at which the transversal
perpendicularity compared to the
cracks were measured
beams’and axis.their probable lack
An additional of perpendicularity
takeaway was the need forcompared to the beams’
a certain amount of strain noise-
axis. An additional takeaway
removing data was the need for
post-processing in aorder
certain amount of
to determine thestrain
crack noise-removing
position. The crack width
data post-processing in order were
and evolution to determine the crack
further calculated byposition. The
integration of crack width
the strain andover
profile evo-a certain
lution were furtherlength adjacent to
calculated bythe crack and through
integration the removal
of the strain profileofover
the contribution of the tension
a certain length
stiffening
adjacent to the crack between the
and through cracks. A good
removal ofagreement (±20 μm)ofwas
the contribution thegenerally observed between
tension stiffening
between cracks. A DOFS
good and DIC crack (widths
agreement ±20 µm)monitoring, as can be
was generally seen in Figure
observed 14. DOFS and
between
DIC crack widths monitoring, as can be seen in Figure 14.
Berrocal et al. [60] also investigated the suitability of embedded robust OFDR DOFS
(featuring a protective external rugged polyamide coating) to accurately assess the perfor-
mance requirements, in terms of vertical deflection and crack width, of three 20 × 25 ×
300 cm RC beams subjected to four-point bending. The strain measurements performed
by means of the thicker fiber were additionally integrated with the ones extracted from
thin polyamide-cladded ones and from an external DIC system. The thin DOFS were
bonded directly onto the steel bars, at the concavity created by the longitudinal ridge,
by means of cyanoacrylate adhesive (after removing the mill-scale and degreasing the
area with acetone) with the addition of a protective layer of a one-component water-proof
oxygen-free silicone rubber (Figure 15c). The robust DOFS were installed longitudinally
along the rebars (Figure 15f) either supported along a longitudinal rebar, bridging the
stirrups or resting on the formwork using electric tape to fix them in place (Figure 15d,e).
The results revealed that a certain strain attenuation occurred in the embedded robust
DOFS compared to the commonly used thin polyamide coated DOFS. However, the pres-
ence of the protective sheath prevented the appearance of SRAs. Performance wise, the
robust DOFS were able to provide a good estimate of the beam deflections with errors of
between 12.3 and 6.5%. Similarly, crack widths computed on the base of the DOFS strain
measurements differed by as little as ±20 µm with the DIC results.

Figure 14. Comparison between the crack width measured by the DIC and calculated from the
DOFS measurements [59].

Berrocal et al. [60] also investigated the suitability of embedded robust OFDR DOFS
(featuring a protective external rugged polyamide coating) to accurately assess the per-
 sal cracks were measured and their probable lack of perpendicularity compared to the
beams’ axis. An additional takeaway was the need for a certain amount of strain noise-
removing data post-processing in order to determine the crack position. The crack width
and evolution were further calculated by integration of the strain profile over a certain
Sensors 2021, 21, 1818 length adjacent to the crack and through the removal of the contribution of the tension
20 of 83
stiffening between cracks. A good agreement (±20 μm) was generally observed between
DOFS and DIC crack widths monitoring, as can be seen in Figure 14.

14. Comparison
Figure 14. Comparison between
betweenthe
thecrack
crackwidth
widthmeasured
measuredbybythe
theDIC
DICand
andcalculated from
calculated thethe
from DOFS
DOFS measurements
measurements [59]. [59].

Brault and
Berrocal et Hoult
al. [60][61]
alsomonitored simultaneously
investigated the suitabilitythe
of deflections
embedded and crack
robust widths
OFDR of
DOFS
thirteen different
(featuring RC beams
a protective tested
external underpolyamide
rugged three-point bending.
coating) The goal was
to accurately to evaluate
assess the per-
the performance
formance of nylon-coated
requirements, in terms ofOFDR DOFS
vertical cables and
deflection for serviceability performance
crack width, of as-×
three 20 × 25
sessments and compare the latter with other monitoring technologies’, namely DIC and
linear potentiometers. The DIC cameras were aimed such that their field of view encom-
passed the middle section of each beam’s span. The results showed that, if to use a FOS
layout that avoided cracks larger than 0.15 mm and accounting for flexural deformations
only, the distributed beam deflections could be accurately measured up until load levels
approaching failure. Furthermore, comparing the DOFS-based crack width calculation
outputs with DIC’s measurements, it was found that the FOS could measure crack widths
with an average measurement difference of 0.007 mm for crack widths up to 0.15 mm,
0.016 mm for crack widths between 0.15 and 0.18 mm, and 0.031 mm for crack widths
between 0.18 and 0.30 mm. The authors concluded suggesting that DOFS could be used
for the monitoring of an RC structure’s compliance with the crack width limits prescribed
by the codes.
Similar RC beams were also used in Brault and Hoult [62] with the objective of
reporting the development of a practical technique to measure distributed reinforcement
strains in RC members beyond cracking. For validation purposes, OFDR DOFS outputs
were further compared against those of conventional SG. Some key takeaways were the
good agreement between the DOFS and SG sampled reinforcement strains at crack locations
and between the latter. The authors observed that exaggeratedly high reinforcement strains
and/or noisy strain results were measured at crack locations (SRAs), likely due to concrete
shearing or pinching of the DOFS. Furthermore, it was noticed that mainly unreliable
strains were measured once reinforcement yielding commenced. Finally, when comparing
the progression of external crack widths with the corresponding reinforcement strains
at crack locations, it was detected that certain cracks initiated gradually whilst others
more abruptly.
 thin polyamide-cladded ones and from an external DIC system. The thin DOFS
bonded directly onto the steel bars, at the concavity created by the longitudinal ridg
means of cyanoacrylate adhesive (after removing the mill-scale and degreasing the
with acetone) with the addition of a protective layer of a one-component water-proo
Sensors 2021, 21, 1818 ygen-free silicone rubber (Figure 15c). The robust DOFS were installed 21 of 83longitud
along the rebars (Figure 15f) either supported along a longitudinal rebar, bridgin
stirrups or resting on the formwork using electric tape to fix them in place (Figure 1

(f)
Figure 15. Installation of the DOFS:
Figure 15. Installation of (a)
the installation of robust DOFS
DOFS: (a) installation cable
of robust in acable
DOFS reinforcement bar by inserting
in a reinforcement bar by it into a
previously milled groove; (b) comparison of thin and robust DOFS; (c) installation of thin DOFS on the surface of a rein
inserting it into a previously milled groove; (b) comparison of thin and robust DOFS; (c) installation
forcement bar by bonding it with cyanoacrylate adhesive and protecting it with silicone; (d) installation of robust DOFS
of thin DOFS on the surface of a reinforcement bar by bonding it with cyanoacrylate adhesive and
on the surface of a reinforcement bar by mechanically anchoring the cable to the reinforcement with electric tape; (e) multi
protecting it with silicone;
layer configuration of embedded DOFS(d) installation
in the of robustand
beam specimens DOFS on the surface
(f) loading of aDOFS
setup and reinforcement barconfiguration
installation by
mechanically anchoring the cable
for the RC beam specimens (adapted from [60]). to the reinforcement with electric tape; (e) multi-layer configuration
of embedded DOFS in the beam specimens and (f) loading setup and DOFS installation configuration
for the RC beam specimens (adapted from [60]).

An additional DOFS-DIC integration was reported in Poldon et al. [63]. Here, a large
RC beam was designed to fail in shear under three-point bending conditions with its
flexural and shear reinforcement strains monitored with OFDR DOFS and its superficial
strains/displacements with DIC. DOFS was bonded to the longitudinal reinforcement
ribs and, notably, on 16 stirrups with a cyanoacrylate adhesive before being covered with
consumer-grade silicone. According to the authors, the presented measurement technique
using FOS and DIC allowed for a detailed study of shear failures and the quantification of
the contribution of steel and concrete to the beam’s stiffness at every stage.
Sensors 2021, 21, 1818 22 of 83

Shear failure was also analyzed in Rodriguez et al. [64,65] who investigated an experi-
mental methodology to obtain the shear cracking pattern of RC beams. In particular, the
latter suggested the formation of a 2D grid composed of one or two OFDR DOFS bonded
to the shear span of a beam. The methodology was tested on three partially pre-stressed
8 m long I-shaped RC beams subjected to shear failure with a DOFS mesh bonded with
epoxy adhesive. During such test the peaks appearing in the sampled DOFS strain profiles
were associated with the presence of cracks (detectable even before their actual visual ap-
pearance). As the latter evolved the DOFS grid’s horizontal and vertical segments allowed
Sensors 2021, 21, x FOR PEER REVIEW
to successfully define the crack pattern despite the different inclinations of the cracks23along
of 87

the web height (see Figure 16).

Figure 16. Real cracking pattern and obtained with DOFS in two different beams (adapted from [64]).
Figure 16. Real cracking pattern and obtained with DOFS in two different beams (adapted from [64]).
Apart from the definition of the crack pattern, an algorithm was also proposed to
calculate
Apartthe crack
from thewidths based
definition of on
the the
crackstrain profiles
pattern, in both horizontal
an algorithm was also and vertical
proposed to
directions.
calculate theOncrack
the extraction
widths basedof theonabove shearprofiles
the strain crack widths,
in boththe proposedand
horizontal DOFS method
vertical di-
was compared
rections. On thewith the more
extraction traditional
of the above shearone based on the deployment
crack widths, the proposed of DOFS
displacement
method
rosettes in [65]. The results showed a relative similarity between the
was compared with the more traditional one based on the deployment of displacement two analyses’ results
despite some mismatches attributed to the high level of measurement error
rosettes in [65]. The results showed a relative similarity between the two analyses’ results involved in the
rosette method. Additionally, the authors observed the presence of high-strain
despite some mismatches attributed to the high level of measurement error involved in DOFS points
unrelated
the rosetteto the cracked
method. visible pattern.
Additionally, the authors These could be
observed thepotentially
presence ofattributed
high-strainto DOFS
small
premature
points unrelated to the cracked visible pattern. These could be potentially attributedthe
cracks, damages in the bonding of the fiber, presence of small cavities in to
concrete surface or
small premature simple
cracks, failures in
damages of the
the bonding
fibers’ backscattering interpretation
of the fiber, presence performance.
of small cavities in
Shear failure
the concrete surfacewas
or also
simplestudied in Fischer
failures et al. [66]
of the fibers’ who performed
backscattering a complex
interpretation load
perfor-
test on
mance. a prestressed concrete beam subjected to four-point bending in order to investigate
the various configurations
Shear failure was also of the deployed
studied in Fischer measuring
et al. [66]device, OFDR DOFS/cables
who performed a complex load(see
Figure 17), and application methods.
test on a prestressed concrete beam subjected to four-point bending in order to investigate
The authors
the various suggestedof
configurations that
theindeployed
structuralmeasuring
concrete construction,
device, OFDR the robustness and
DOFS/cables (see
durability of the monitoring DOFS (in addition to their basic suitability and sensitivity)
Figure 17), and application methods.
should reflect their application (glued to the concrete surface, applied to reinforcement bars
or embedded in concrete). Thus, the authors investigated both conventional polyimide and
nylon coated DOFS but also more robust measuring cables (referred to as V0, V1, and V9
and N3) which include various plastic and metal layers (thus comprising comparatively
thicker cross-sections). The formers were bonded to the surface of the test specimen with a
cyanoacrylate-based adhesive whilst the latter were embedded inside it. During the test,
all the observed flexural cracks formed peaks in the strain curve of the measuring cable.
 was compared with the more traditional one based on the deployment of displacement
rosettes in [65]. The results showed a relative similarity between the two analyses’ results
despite some mismatches attributed to the high level of measurement error involved in
the rosette method. Additionally, the authors observed the presence of high-strain DOFS
points unrelated to the cracked visible pattern. These could be potentially attributed to
Sensors 2021, 21, 1818 23 of 83
small premature cracks, damages in the bonding of the fiber, presence of small cavities in
the concrete surface or simple failures of the fibers’ backscattering interpretation perfor-
mance.
Shearso,
Similarly failure was also studied
the formation in Fischer
and opening et al.cracks
of shear [66] who performed
close a complex
to the supports couldload
be
test on a prestressed
identified concreteguided
by the upwardly beam subjected
sensors V9,to four-point
V1, and N3bending
in thoseinspecific
order to investigate
points where
the
the various
DOFS and configurations of theThe
shear cracks cross. deployed
robust measuring cables
device,V9OFDR
and V1DOFS/cables (see
were reportedly
Figure 17), and
more adapt application
to reporting themethods.
steel strains beyond the cracking stage.

Figure 17.
Figure DOFS cables
17. DOFS cables used
used in
in [66].
[66].

Mehdi
The Mirzazadeh
authors suggestedandthat
Green [67] studied
in structural an application
concrete of OFDR
construction, DOFS and DIC
the robustness and
for the measurement of the crack widths, deflections, and strains in eight
durability of the monitoring DOFS (in addition to their basic suitability and large-scale RC
sensitivity)
beams tested under static and fatigue loading at room temperature (+15 ◦ C) and at low
should reflect their application (glued to the concrete surface, applied to reinforcement
temperature (−20 to −25 ◦ C). Half of the 200 × 400 × 4200 mm beams had no shear
reinforcements whilst the remaining ones did. Single-mode nylon coated DOFS were
bonded both at the top and the bottom of the beams’ side surfaces which had previously
been ground and cleaned with alcohol. Additionally, in both positions two parallel rows of
DOFS were deployed, one fully bonded and the other partially bonded. DOFS reportedly
measured strains with an accuracy close to the one of parallelly installed SG under both
temperature conditions. A noteworthy observation was that the sum of the strain-optic
error (thermal stress-induced error when a coating is present) and the thermo-optic error
(error caused by the alteration of the fiber refractive index due to temperature variations)
was close for the two different tested fibers. The first of these was a single-mode fiber
(900 µm coating diameter) whilst the second a polarization-maintaining fiber (230–260 µm
coating diameter). Finally, the article introduced a methodology to determine the thermal
expansion coefficient of the optical fibers with thin coatings.
Sawicki et al. [68] proceeded to validate the performance of DOFS (integrated with DIC
and extensometers) for the SHM of Ultra-High-Performance Fiber-Reinforced Cementitious
composites (UHPFRC). In particular, a full-scale UHPFRC beam was instrumented with
three lines of OFDR DOFS positioned inside 2 × 2 mm grooves present on each face
of the beam (the fibers were glued in place with a bi-component epoxy adhesive). The
displacement-controlled four-point bending test was characterized by different loading–
unloading steps. Overall, the experimental campaign yielded a successful detection of the
UHPFRC beam structural discontinuities from the DOFS’ part. Yet, during the elastic stage
of the beam’s structural response, large measurement noises caused the failed observation
of the DIC strain variations leading the authors to state that the latter still remained a
highly complex and sensitive monitoring technique. However, in the cracking stage, DIC
allowed for the tracking and characterizing of the localized fictitious cracks, i.e., localized
discontinuities bridged by multiple fibers (see Figure 18).
The authors further observed that polyimide-coated fibers allowed for the early and
accurate detection and localization of micro-cracks. Conclusively, Sawicki et al. highlighted
the importance of a deep understanding of the sensor properties (like crack and temperature
sensitivity), durability and aging (tackled by [46,47]), long-term fatigue (tackled by [51])
for holistic global and local strain measurements.
Sensors 2021,
Sensors 21, x1818
2021, 21, FOR PEER REVIEW 2524of
of 87
83

Figure
Figure 18.
18. DOFS-integrating
DOFS-integrating DIC
DIC strain
strain distributions
distributions at
at different
different load steps [68].

Broth and Hoult [69] investigated the potential of using OFDR DOFS for the monitor-
ing of a series of slender and deep RC beams subjected to three and four-point bending
cyclic loading. In particular, the embedded distributed sensors were used to assess the
deflection growth induced by the cyclic loading (3600 loading cycles at 1 Hz). DOFS were
bonded to sanded and degreased rebars by means of a cyanoacrylate adhesive and later
protected with an additional layer of silicone. The second of the three testing phases (initial
loading, cycling loading and loading to failure) saw a deflection growth (see Figure 19)
Sensors 2021, 21, 1818 25 of 83

attributable
Sensors 2021, 21, x FOR PEER REVIEW to the combined result of crack formation, crack growth, creep in the concrete
26 of 87
and the degradation of the bond between the steel reinforcement and concrete.

19. Strain along

Figure 19. along the
the bottom
bottom reinforcement
reinforcementduring
duringcycling
cyclingfor
forfour
fourdeep
deepbeam
beamspecimens
specimens[69].
[69].
Finally, the authors highlighted the fact that DOFS was able to withstand cyclic
loadings all the
Finally, theway to failure.
authors highlighted the fact that DOFS was able to withstand cyclic load-
ings all the way to failure. day, still few tests are available regarding the issues of DOFS-
However, at the present
monitored fatigue
However, at and deterioration.
the present day, still few tests are available regarding the issues of
DOFS-monitored fatigue and out
Barrias et al. [70] carried an experimental campaign on two 100 × 180 × 800 mm
deterioration.
beams Barrias et al. [70] carried out an DOFS
with polyimide-coated OFDR installedcampaign
experimental both on theonembedded
two 100 × rebars andmm
180 × 800 on
the
beamsspecimen surfaces. The beams
with polyimide-coated OFDR were tested
DOFS under three-point
installed both on the bending
embedded starting
rebarswith
and an
on
initial monotonic loading cycle, followed by an unloading stage (triggered
the specimen surfaces. The beams were tested under three-point bending starting with an well beyond the
cracking stage) and
initial monotonic concluded
loading cycle,byfollowed
another byloading cycle. In stage
an unloading order (triggered
to verify the
wellreliability
beyond
and
the cracking stage) and concluded by another loading cycle. In order to verify the rebar
accuracy of the DOFS measurements, additional SG were installed along the in
reliabil-
three locations. Overall, the optic fiber bonded to the rebar delivered
ity and accuracy of the DOFS measurements, additional SG were installed along the rebargood results despite
disturbances in theOverall,
in three locations. surroundings of the
the optic crack.
fiber The latter
bonded to thewas
rebaralso reportedly
delivered gooddetected
resultsand
de-
located by the superficial DOFS albeit only until a certain stiffness variation in the beam,
spite disturbances in the surroundings of the crack. The latter was also reportedly detected
beyond which large disruptive SRAs appeared in the proximity of the damage location.
and located by the superficial DOFS albeit only until a certain stiffness variation in the
Bassil et al. [71] proceeded to the performance assessment of Coda Wave Interferome-
beam, beyond which large disruptive SRAs appeared in the proximity of the damage lo-
try and OFDR DOFS (deployment visible in Figure 20) for the damage detection of a 20 ×
cation.
20 × 100 cm RC beam tested under three-point bending.
Bassil et al. [71] proceeded to the performance assessment of Coda Wave Interferom-
The beam was also instrumented with SG and Linear Variable Displacement Trans-
etry and OFDR DOFS (deployment visible in Figure 20) for the damage detection of a 20
formers (LVDT). The DOFS was glued to the beam’s front and back surfaces, to the front
× 20 × 100 cm RC beam tested under three-point bending.
bottom rebar, mid top rebar, mid bottom rebar and back bottom rebar. The results showed
The beam was also instrumented with SG and Linear Variable Displacement Trans-
that the implementation of Coda Wave Interferometry and DOFS allowed for the detection
formers (LVDT). The DOFS was glued to the beam’s front and back surfaces, to the front
of early subtle changes in RC structures until crack formation. On DOFS, the authors con-
bottom rebar, mid top rebar, mid bottom rebar and back bottom rebar. The results showed
cluded that, beyond the latter stage, their ability to quantify the crack openings permitted
that the implementation
effective monitoring of RC ofstructures.
Coda Wave Interferometry and DOFS allowed for the detec-
tion of early subtle changes in RC structures until crack formation. On DOFS, the authors
concluded that, beyond the latter stage, their ability to quantify the crack openings per-
mitted effective monitoring of RC structures.
Sensors 2021, 21, 1818 26 of 83
Sensors 2021, 21, x FOR PEER REVIEW 27 of 87

Figure 20. Ultra Sonics (US) and DOFS sensors attached to the rebars before casting of concrete
Figure 20. Ultra Sonics (US) and DOFS sensors attached to the rebars before casting of concrete [71].
[71].
Wu et al. [72] tested an RC beam under four-point loading with OFDR DOFS both
glued toWu itsetsurface
al. [72]and tested an RC beam
embedded insideunder four-point
it on three separateloading
layers.with OFDR DOFS
Experimental both
results
glued tothat
indicated its surface
cracks andcould embedded
be detected inside it on peaks
as sharp three separate layers. Experimental
of the measured results
strain distribution
indicated that cracks could be detected as sharp peaks of the
profiles, despite also noting that strains pertaining to neighboring cracks could interfere measured strain distribution
profiles,
with each despite also noting
other. Indeed, the that
authors strains pertaining
pointed out thatto neighboring
only neighboring cracks cracks
could interfere
with a
with each other. Indeed, the authors pointed out that
separation of more than 6 cm could be discerned with a 1 cm spatial resolution. only neighboring cracks with a sep-
Finally,
aration
both of more than
the widening and 6thecmpropagation
could be discernedof crackswith werea successfully
1 cm spatialdetected
resolution. andFinally,
describedboth
tothe widening
have and the
a slow initial propagation
growth followed ofbycracks were
a linear successfully
growth stage. detected and described to
have a slow
Chen initial
et al. [73]growth
brought followed
up the by a linear under
condition growthwhichstage.an RC structure might be
Chen et al. [73] brought up the condition
subjected to bi-directional loads. They then proceeded to use under which an RC structure Pump)
a (Pulse-Pre might be sub-
PPP-
jected to bi-directional loads. They then proceeded to use a (Pulse-Pre
BOTDA DOFS (with a 10 cm spatial resolution) to assess the superficial structural strains Pump) PPP-BOTDA
ofDOFS
a 10 ×(with
15 ×a150 10 cm
cm spatial
RC beam resolution)
subjectedtotoassess the superficial
bi-directional loading. structural strains of a 10 ×
The strain–deflection
15 × 150 cmwas
relationship RCestablished
beam subjected usingtothe bi-directional
explicit function loading. The method
analysis strain–deflection relation-
and the implicit
ship was
function established
mapping methodusing the explicit
based function
on the DOFS analysis method and the implicit function
readings.
mapping
Malek method
et al. [74]based
utilized on OFDR
the DOFS DOFS readings.
for the study of the steel/concrete bond degrada-
tion dueMalek et al. [74] utilized
to reinforcement yielding OFDR in RC DOFS
beams forsubjected
the studytooflateral
the steel/concrete
loading. The bond degra-
performed
dation due to reinforcement yielding in RC beams subjected to lateral loading. The per-
experimental program, representing a drastic departure from the common bond/slip
determination
formed experimental tests (i.e.,program,
the singlerepresenting
and double apull-out drastictensile
departuretests),from
was the
based on a
common
pseudo-static pushover testtests
bond/slip determination of four
(i.e.,25the× single
35 × 285 andcm cantilever
double RC beams
pull-out tensiletested
tests),vertically.
was based
Instrumentation-wise,
on a pseudo-static pushover two different
test oftypes
four 25 of ×optical
35 × 285fibers
cm were used,RC
cantilever a PVC-coated
beams tested tight
ver-
buffer OFS
tically. cable (for both internal
Instrumentation-wise, twoand external
different sensing)
types and afibers
of optical polyimide-coated
were used, afiber PVC-
(bonded on the
coated tight longitudinal
buffer OFS cable rebars).
(for both Thus, during
internal andthe test, the
external strainsand
sensing) were captured
a polyimide-
continuously at a spacing resolution of 2.6 mm and with
coated fiber (bonded on the longitudinal rebars). Thus, during the test, the strains an accuracy of 25 µε both were
on
the outside and on the inside of the specimen. This allowed
captured continuously at a spacing resolution of 2.6 mm and with an accuracy of 25 με for the determination of the
concrete/steel
both on the outsideslip, steel
andstress,
on thebond inside stress,
of the bond deterioration
specimen. length for
This allowed andthethedetermination
locations of
cracks.
of the Finally, the authors
concrete/steel introduced
slip, steel a novel
stress, bond model
stress, bondto predict pre- and
deterioration post-yield
length and the bond
loca-
behavior.
tions of cracks. Finally, the authors introduced a novel model to predict pre- and post-
Sounthararajah
yield bond behavior. et al. [75] presented an experimental campaign aimed at evaluating the
OFDRSounthararajah
DOFS reliabilityetand accuracy
al. [75] presentedwhenan measuring
experimental the flexural
campaign strains
aimed of aatCemented
evaluating
Pavement
the OFDR DOFS reliability and accuracy when measuring the flexural strainsbending
Material (CPM) beam. In particular, both static and dynamic four-point of a Ce-
tests werePavement
mented performed on DOFS
Material and beam.
(CPM) LVDT-instrumented
In particular, both 50 ×static
50 ×and 200dynamic
mm CPM beams.
four-point
The latter were
bending tests manufactured
were performed withonHolcim
DOFS and RoadLVDT-instrumented
Base pavement material 50 ×from the Gosnells
50 × 200 mm CPM
Quarry, Perth, Western Australia combined with a
beams. The latter were manufactured with Holcim Road Base pavement materialGeneral Purpose cement as a from
binder the
substance. An acrylate-coated single-mode fiber (SMF-28e)
Gosnells Quarry, Perth, Western Australia combined with a General Purpose cement was bonded to the bottom andas a
Sensors 2021, 21, x FOR PEER REVIEW 28 of 87

Sensors 2021, 21, 1818 27 of 83

binder substance. An acrylate-coated single-mode fiber (SMF-28e) was bonded to the bot-
tom and side surfaces of the beam specimens (with a two-part epoxy adhesive) after hav-
ing been properly cleaned with acetone. The experimental output showed a fair agree-
side surfaces of the beam specimens (with a two-part epoxy adhesive) after having been
ment between DOFS measured and calculated flexural strains suggesting their perfor-
properly cleaned with acetone. The experimental output showed a fair agreement between
mance for this scope despite its taxing surface preparation. According to the authors, the
DOFS measured and calculated flexural strains suggesting their performance for this scope
test substantiated
despite DOFS’ capability
its taxing surface of detecting
preparation. Accordingthe
tocrack initiation
the authors, theand
testpropagation in
substantiated
CPM beams.
DOFS’ capability of detecting the crack initiation and propagation in CPM beams.

4.2.
4.2. Reinforced
Reinforced Concrete
Concrete Tensile Elements: Steel/Concrete
Tensile Elements: Steel/Concrete Bond
Bond Stresses,
Stresses, Cracking
Cracking and
and Concrete
Concrete
Shrinkage
RC ties are often used to illustrate cracking, deformation and bond behavior of RC
structures thanks
structures thankstototheir
theirsimplicity
simplicityandanda areasonably
reasonably good
good representation
representation of of
thethe distri-
distribu-
bution
tion of internal
of internal forces
forces andand strains
strains in the
in the tensile
tensile zonezone of RC
of RC structures
structures [76].[76]. As such,
As such, a
a cer-
certain
tain number
number of of research
research teams
teams haveattempted
have attemptedtotoupgrade
upgradethe thecommon
commonRC RC tie
tie test with
the embedment of DOFS (bonded to the reinforcement bars).
Bado et al. [77], for
for example,
example, performed a double pullout test on five differently sized
RC ties (with OFDR DOFS bonded
RC ties (with OFDR DOFS bonded toto their
their rebars
rebars in
in different
different manners
mannersas as visible
visiblein inFigure
Figure
21) with the goal of studying the steel/concrete interaction, i.e., bond
21) with the goal of studying the steel/concrete interaction, i.e., bond and slip.and slip.

Figure
Figure 21.
21. (a)
(a) Tested RC ties’
Tested RC ties’ with
withtheir
theirrespective
respectivebonding
bondingtechnologies,
technologies,with
with(c)(c) and
and without
without (b)(b)
the
the addition of a protective silicone layer. The DOFS-instrumented rebars were monitored by
addition of a protective silicone layer. The DOFS-instrumented rebars were monitored by means of
means of an (d) ODiSI-A OBR interrogator [77].
an (d) ODiSI-A OBR interrogator [77].
According
According to to the
the authors,
authors, before
before the
the advent
advent of DOFS, due
of DOFS, to the
due to the lack
lack of
of tools
tools able
able to
to
assess the rebar strains in an accurate, completely-distributed and un-intrusive
assess the rebar strains in an accurate, completely-distributed and un-intrusive fashion, fashion,
most
most structural
structural analyses
analyses relying
relying on such knowledge
on such knowledge (such as tension
(such as tension stiffening)
stiffening) resorted
resorted
to
to theoretical,
theoretical, empirical
empiricalor ornumerical
numericalsolutions.
solutions.They
Theycontinued
continued suggesting
suggestingthat thethe
that intro-
in-
duction of DOFS allowed to finally acquire that insight. In the reported experiment,
troduction of DOFS allowed to finally acquire that insight. In the reported experiment, the
RC
the ties’ crack
RC ties’ positions
crack detected
positions by DOFS
detected werewere
by DOFS compared against
compared the ones
against detected
the ones by a
detected
by a simultaneous DIC monitoring. Briefly, thanks to DOFS, the authors detected some
specimen’s hybrid bond/slip behavior falling between a crack formation stage’s and stabi-
lized cracking stage’s whenever they were characterized by a length dangerously close to
 Sensors 2021, 21, x FOR PEER REVIEW 29 of 87
Sensors 2021, 21, 1818 28 of 83

simultaneous DIC monitoring. Briefly, thanks to DOFS, the authors detected some speci-
men’s hybrid bond/slip behavior falling between a crack formation stage’s and stabilized
surpassing the maximum theoretical
cracking stage’s whenevercrack spacing
they length. The by
were characterized latter was dangerously
a length translated intoclose to sur-
different bond-slip profiles.
passing The authors
the maximum concluded
theoretical byspacing
crack notinglength.
that for future
The latterDOFS applica- into dif-
was translated
tions a combination of bond-slip
ferent gluing adhesive
profiles. (cyanoacrylate for example)
The authors concluded by notingwith the
that foraddition of a applica-
future DOFS
tions a combination
protective layer (silicone) of gluing adhesive
was the preferable (cyanoacrylate
DOFS/rebar for example)
bonding technique aswith the addition of a
it removed
protective
undesired SRA-related layer (silicone) was the preferable DOFS/rebar bonding technique as it removed
uncertainties.
undesired
Cantone et al. [78] alsoSRA-related
provided uncertainties.
a comprehensive work on the determination of the
rebar–concrete interaction byetmeans
Cantone al. [78] of
also provided
OFDR DOFSa comprehensive work on the
and DIC monitored RC determination
structures of the
rebar–concrete interaction by means of OFDR DOFS and DIC monitored RC structures
tested under pullout, double pullout and bending conditions. Similar to the previous
tested under pullout, double pullout and bending conditions. Similar to the previous ar-
article, the authors also positioned the DOFS inside 1 mm-depth grooves performed along
ticle, the authors also positioned the DOFS inside 1 mm-depth grooves performed along
two or four rebartwosides (see rebar
or four Figure 22).(see Figure 22).
sides

Figure Figure
22. (a) Back-scattering mechanismmechanism
22. (a) Back-scattering due to fiber due
core to
impurities, (b)impurities,
fiber core view of the fiber-optic sensor
(b) view of glued to a steel
the fiber-optic
rebar, (c) sketch
sensor of thetofiber
glued optical
a steel cable
rebar, (c) placed
sketchinofthe
thegroove and adhering
fiber optical to the steel
cable placed due
in the to two-component
groove and adheringglue
to and
(d) description of the fiber optic cable structure with angle polished connector on one side and termination (which avoids
the steel due to two-component glue and (d) description of the fiber optic cable structure with angle
light reflection) on the opposite [78].
polished connector on one side and termination (which avoids light reflection) on the opposite [78].
Amongst the experimental campaign conclusions, the followings are particularly rel-
Amongst the experimental
evant. The observed campaign conclusions,
strain gradients and high the followings
strain are particularly
values sampled during the pullout
relevant. The observed strain gradients
test confirmed and
that this test high straincould
methodology values sampled
prove during thetopullout
to be unsuitable characterize the
test confirmed that this test
response methodology
of embedded could prove
reinforcement. to be
During theunsuitable
multi-cycle to characterize
double pullout tests, by
the response of embedded reinforcement.
means of DOFS, the authors During the multi-cycle
were successfully double
able to track thepullout tests,
increasing bond bydegrada-
means of DOFS, thetionauthors
(see Figure
were23)successfully
providing results
able toin track
soundthe agreement
increasingwith the literature.
bond degradationFinally, dur-
ing the testing
(see Figure 23) providing of in
results DOFS
soundinstrumented
agreementRC beams,
with the lower bending-induced
literature. Finally, duringstress
thevariations
testing of DOFS were detected inRC
instrumented thebeams,
longitudinal
lowerreinforcement
bending-induced with respect
stresstovariations
the theoretical
were values cal-
culated withreinforcement
detected in the longitudinal elastic-crackedwithanalyses.
respectThis
towas reportedly due
the theoretical to imperfect
values calculated closure of
cracks and negative tension-stiffening effects.
with elastic-cracked analyses. This was reportedly due to imperfect closure of cracks and
Zhang et al. [79] investigated the ability of DOFS to detect concrete cracking and large
negative tension-stiffening effects.
strain steel deformation of six 120 × 120 × 1000 mm RC ties with different OFDR DOFS
Zhang et al.cable
[79]structures,
investigated the ability of DOFS to detect concrete cracking and
coating materials and diameters (see Figure 24) which were both embed-
large strain steelded
deformation
in the concrete and120
of six × 120
bonded to ×the1000 mm
rebar. TheRC tiesreported
article with different
how OFDR OFDR successfully
DOFS cable structures, coating materials and diameters (see Figure 24) which
detected continuous rebar deformations and the evolution of the concrete–steel were both bond
embedded in the concrete
stress up untiland
andbonded
beyond to thethe rebar.range
yielding Theasarticle
well asreported
identifyinghow OFDR
the crack-indicative
successfully detected continuous
strains. rebar deformations and the evolution of the concrete–
steel bond stress up until and beyond the yielding range as well as identifying the crack-
indicative strains.
Sensors 2021,
Sensors 2021, 21,
21, x
x FOR
FOR PEER
1818 PEER REVIEW
REVIEW 30 of
29of
30 87
of 87
83

Figure 23.
Figure 23. Tensile
Tensiletests—specimen
tests—specimenTC04:
tests—specimen TC04:(a)
TC04: (a)cracking
(a) crackingpattern
cracking patterninin
pattern inthe
the
the cyclic
cyclic
cyclic loading
loading
loading phase,
phase,
phase, (b)
(b)(b) stress-strain
stress-strain
stress-strain relationship
relationship
relationship for
forfor
all
all
all cycles
cycles (in
(in grey)
grey) and
and cycle
cycle #6
#6 (in
(in green)
green) indicating
indicating load
load levels,
levels, (c)
(c) profiles
profiles of
of stresses,
stresses, calculated
calculated on
on the
the
cycles (in grey) and cycle #6 (in green) indicating load levels, (c) profiles of stresses, calculated on the basis of the measured basis
basis of
of the
the
measured DOFS
measured DOFS (here
(here defined as
as FOM)
FOM) strains
strains andand (d)
(d) bond
bond stresses
stresses ττ profiles
profiles along
along the
the steel
steel rebar
rebar (adapted
(adapted from [78]).
DOFS (here defined as defined
FOM) strains and (d) bond stresses τ profiles along the steel rebar (adapted from [78]). from [78]).

Figure 24.
Figure 24. DOFS
DOFS cables used in
cables used in [79].
[79].

Similarly to
Similarly to [77],
[77], the
the authors
authors highlighted
highlighted the the potential
potential of of DOFS
DOFS combined
combined withwith dou-dou-
ble pullout
ble pullout tests
tests for
for the
the evaluation
evaluation of existing RC
of existing RC structure
structure behavioral
behavioral predictive
predictive models
models
through the
through the bond
bond stress
stress analysis.
analysis. Worth
Worth noting
Worth noting areare the
the conclusions
conclusions on on the
the performance
performance of of
each DOFS
each DOFS cable,
cable, whose
whose different
different composition
composition was was shown
shown to to have
have aa substantial
substantial influence
influence
on the
on the measured
measuredstrain
measured strainresponse
strain response
response especially
especially
especially beyond
beyond
beyond thethe onset
the of nonlinear
onset
onset of nonlinear
of behavior
nonlinear and and
behavior
behavior into
and
large
into deformation
into large
large deformation regimes.
deformation regimes. Generally
regimes. Generally speaking,
Generally speaking, the
speaking, the measurement
the measurement sensitivity
measurement sensitivity parameter
sensitivity param-
param-
decreased
eter decreased
eter decreased withwithan increase
with an increase
an in the
increase incable
in coating
the cable
the cable thickness
coating
coating due due
thickness
thickness to
dueboth elastic
to both
to both and and
elastic
elastic inelastic
and ine-
ine-
coating
lastic deformations.
lastic coating
coating deformations.
deformations. The sensitivity
The also tended
The sensitivity
sensitivity to decrease
also tended
also tended with increased
to decrease
to decrease loading
with increased
with increased likely
loading
loading
due
likelytodue
likely debonding
due to betweenbetween
to debonding
debonding the cablethe
between and
the concrete
cable
cable (with
and concrete
and concretethe (with
exception
(with the of cable NZ_Y_09).
the exception
exception of cable
of cable
NZ_Y_09). Finally, a certain influence, particularly on the bond with concrete, wasexer-
Finally,
NZ_Y_09). a certain
Finally,influence,
a certainparticularly
influence, on the bond
particularly with
on concrete,
the bond was
with reported
concrete,to be
was re-
re-
cised
ported by
ported to the
to be coating
be exercised
exercised bystiffness
by the and
the coatingthe cables’
coating stiffness surface
stiffness and
and the properties.
the cables’ Another
cables’ surface key
surface properties. requirement
properties. Another
Another
was
key reportedly
key requirement
requirement thewasfiber
was survivability,
reportedly
reportedly the both
the fiberin
fiber terms of measurement
survivability,
survivability, both in
both survivability
in terms
terms (i.e., the
of measurement
of measurement
deformation limit
survivability (i.e.,
survivability before
(i.e., the which
the deformation data
deformation limit can still
limit before be obtained)
before which
which data and
data can the
can stillcable
still be survivability
be obtained)
obtained) and and(i.e.,
the
the
limit
cable at which
survivabilitythe fiber
(i.e., is permanently
limit at which damaged).
the fiber is On the
permanently latter,
cable survivability (i.e., limit at which the fiber is permanently damaged). On the latter, the authors
damaged). pointed
On the out
latter,
that a high sensitivity did not necessarily correlate with low survivability. For example,
Sensors 2021, 21, 1818 30 of 83

cable NZ_Y_09 reportedly boasted twice the sensitivity compared with NZ_K_20 with 50%
data loss occurring at comparable crack widths.
Sienko et al. [80] presented the outcome of a laboratory test designed to verify the
suitability of a standard “tight jacket” OFDR DOFS for the strain analysis of RC members.
In particular, the authors deployed the optical fiber cable inside a 77 × 110 × 1050 mm RC
tie along several longitudinally drilled holes before bonding it with epoxy resin. The same
adhesive was used to bond the fiber to the degreased reinforcement bar parallelly to its
longitudinal rib. According to the authors, the results showed good measurement accuracy
for micro-cracks and narrow crack widths (range up to 0.05 mm) and only moderate
accuracy in the case of wider cracks (up to 0.30 mm). This was attributed to the slip between
the DOFS core and its external jacket leading to fiber strains not corresponding fully with
the actual deformation state of the concrete member. However, it was also reported that
the measurement methodology allowed for precise crack widths determination as to the
sum of fiber elongations within the sections between the cracks.
Davis et al. [81] instrumented multiple reinforcement bars with OFDR DOFS before
embedding them inside RC prisms, thus forming RC ties. Their goal, measuring the steel
rebars’ distributed strain profiles during the concrete shrinkage period of time (3, 7, 21 and
28 days) and during subsequent axial tension tests. Davis et al. instrumented the bars with
a polyimide-cladded and a nylon-cladded DOFS bonded to the bar with a cyanoacrylate
along the longitudinal ridge. It was observed that the shrinkage at both specimens’ free
edges was greater than at its center with only small strain variations (± 20 µε) between
these two. Furthermore, with the passing of the days, the DOFS bonded to the rebar
and embedded in the concrete displayed progressively increasing compressive strains
(approximately 300 µε at 28 days).
A proper shrinkage strain measurement is imperative for a proper assessment of the
tension stiffening as it may lead to an apparent underestimation [82]. To remove such
effect, the authors used some guidelines provided in Fields and Bischoff [83] and found
a good agreement with the DOFS experimental results for the rebar 15 M specimens and
not so good agreement with the rebar 10 M specimens. In conclusion, two additional
observations are worthy of emphasis. First, the authors’ suggestion of bonding DOFS on
both sides of the reinforcement bars in order to mitigate the localized bending effects due
to non-straight bars and non-symmetric crack formations. Second, the observation of a
certain amount of slip between nylon claddings and the sensing silica cores could result in
incorrect measurements.

4.3. Reinforced Concrete Slabs

The present sub-section starts acknowledging 2013’s pioneering work by Villalba and
Casas [84] which represented one of the first structural applications of DOFS monitoring. In
particular, the publication was based on the OFDR DOFS instrumenting of an RC slab with
successful detection of both crack locations and widths as per a subsequent publication by
Rodríguez et al. [85].
More recently, in Nurmi et al. [86], a proof of concept study was undertaken that
involved the measurement through OFDR DOFS distributed strain profiles in two-way RC
slabs with varying reinforcement ratios and levels of axial restraint tested under a central
point load. The goal, shedding light on the impact of partial restraint and reinforcement
ratio on two-way slab behavior and comparing the load capacity and strain assumptions
to a yield-line analysis. Two slabs were subjected to axial restraint using a welded HSS
restraint frame whilst two others were not restrained. The slabs with less reinforcement
demonstrated a ductile flexural failure while the more heavily reinforced slabs failed due
to punching shear (see Figure 25).
The DOFS sampled strain data (Figure 26) enabled the successful detection of perfor-
mance differences for the two support conditions (vertical and axial) and the capturing of
the onset of reinforcement yielding. The authors additionally observed an enhanced slab
capacity and stiffness in the presence of axial restraint.
Sensors 2021, 21, x FOR PEER REVIEW 32 of 87

Sensors 2021, 21, 1818 31 of 83

Sensors 2021, 21, x FOR PEER REVIEW 32 of 87

Figure 25. Specimen failure due to punching shear and its crack pattern [86].

The DOFS sampled strain data (Figure 26) enabled the successful detection of perfor-
mance differences for the two support conditions (vertical and axial) and the capturing of
the onset of reinforcement yielding. The authors additionally observed an enhanced slab
capacity
Figure and
Figure25. stiffnessfailure
25.Specimen
Specimen in thedue
failure presence of axial
duetotopunching
punching restraint.
shear
shear anditsitscrack
and crackpattern
pattern[86].
[86].

The DOFS sampled strain data (Figure 26) enabled the successful detection of perfor-
mance differences for the two support conditions (vertical and axial) and the capturing of
the onset of reinforcement yielding. The authors additionally observed an enhanced slab
capacity and stiffness in the presence of axial restraint.

Figure
Figure26.
26.Slab
Slab3-C-45
3-C-45DOFS
DOFSstrains:
strains:(a)
(a)5050kN,
kN,(b)
(b)100
100kN,
kN,(c)
(c)150
150kN
kN(adapted
(adaptedfrom
from[86]).
[86]).

Sienko et al. [87] glued OFDR DOFS cables on the composite rods of a truck scale
platform slab in order to analyze the internal concrete strains and local nonlinearities
(cracks). The monitoring was ongoing throughout the concrete hydration process (thermal-
shrinkage strains), the prestressing tendons activation (strains regarding the transfer of
Figure 26. Slab 3-C-45 DOFS strains: (a) 50 kN, (b) 100 kN, (c) 150 kN (adapted from [86]).
Sensors 2021, 21, x FOR PEER REVIEW 33 of 87

Sienko et al. [87] glued OFDR DOFS cables on the composite rods of a truck scale
Sensors 2021, 21, 1818 32 of 83
platform slab in order to analyze the internal concrete strains and local nonlinearities
(cracks). The monitoring was ongoing throughout the concrete hydration process (ther-
mal-shrinkage strains), the prestressing tendons activation (strains regarding the transfer
of compression
compression forces
forces fromfrom
the the tendons
tendons to concrete)
to the the concrete)
and the andlaboratory
the laboratory
tests, tests,
whenwhenslabs
slabs were mechanically loaded until failure. The reported early-age
were mechanically loaded until failure. The reported early-age hydration-induced concrete hydration-induced
concrete strainsthe
strains reached reached
valuethe value
of ± of ±220
220 µε. με.
Later, Later,the
during during the first
first test test
steps, thesteps, the platform
platform worked
worked linearly
linearly and and elastically
elastically (deflection
(deflection of 10 mm) of whilst
10 mm) whilst
later later on
on DOFS DOFS the
reported reported
presencethe
presence of discontinuities
of discontinuities related torelated to the
the crack crack occurrence
occurrence which continued
which continued developingdeveloping
during
during subsequent
subsequent load steps.
load steps.
Liao et al.
al. [88]
[88] instrumented
instrumentedan anRCRCslab
slabwith
withOFDR
OFDR DOFS
DOFS in in
anan attempt
attempt to monitor
to monitor its
its thermal
thermal curling
curling process
process whenwhen subjected
subjected to external
to external temperature
temperature fluctuations.
fluctuations. Explic-
Explicitly,
a 50 a×5050× ×
itly, 505× cm
5 cm mortar
mortar slabwas
slab wasinstrumented
instrumentedwith withthe thecombined
combinedmonitoring
monitoring system
system
consisting of inclinometer, DOFS, and thermocouples. During the testing, the heat source
was a light heater with a 500 W halogen halogen bulb positioned above the slab except that the
heating was
wasnotnotdirect
directbut
butbuffered
buffered byby
double
doublepolymer
polymerinsulation (see Figure
insulation 27) positioned
(see Figure 27) posi-
in orderintoorder
tioned createtoacreate
relatively even temperature
a relatively field above
even temperature fieldthe slab.the slab.
above

Figure 27. Test setup: (a) instrumented test slab, (b) testing with surrounding
surrounding insulation
insulation [88].
[88].

According to
According to the
the authors,
authors, DOFS
DOFS demonstrated
demonstrated that that the
the top
top surface
surface ofof the
the slab
slab had
had aa
fast reaction rate to the heating/cooling process whilst the bottom one suffered
fast reaction rate to the heating/cooling process whilst the bottom one suffered a time-lag a time-lag
resulting in
resulting in aa top-vs-bottom
top-vs-bottom temperature
temperature difference.
difference. In conclusion, the
In conclusion, the authors
authors stated
stated that
that
DOFS can provide a comprehensive picture of the longitudinal temperature
DOFS can provide a comprehensive picture of the longitudinal temperature distributions, distributions,
the heat
the heat flux
flux transfer
transfer process
process and
and the
the consequent
consequent strain
strain distributions
distributions thus
thus returning
returning an an
immediate idea of the deformed shapes of the pavement
immediate idea of the deformed shapes of the pavement slab. slab.
Teguedy et
Teguedy etal.
al.[89]
[89]sustained
sustainedthat
thatthe
thestudy
studyofof
thethe
early-age behavior
early-age behaviorof Timber–Concrete
of Timber–Con-
Composite (TCC) structures is of great interest as it provides valuable information
crete Composite (TCC) structures is of great interest as it provides valuable information for manu-
for manufacturing specification development, quality control, and optimization design.
facturing specification development, quality control, and optimization of the formwork of the
As such, the authors decided to deploy BOTDA DOFS inside two 900 × 75 × 8500 mm TCC
formwork design. As such, the authors decided to deploy BOTDA DOFS inside two 900 ×
slabs in order to continuously monitor (during 30 days) their short-term behavior, i.e., the
75 × 8500 mm TCC slabs in order to continuously monitor (during 30 days) their short-
early-age temperature/strain variation in the fresh concrete and in the CLT slab. Additionally,
term behavior, i.e., the early-age temperature/strain variation in the fresh concrete and in
the multiplexing capabilities of the DOFS system allowed for the strain measurement at several
the CLT slab. Additionally, the multiplexing capabilities of the DOFS system allowed for
heights of the timber section with a single sensor (see Figure 28).
the strain measurement at several heights of the timber section with a single sensor (see
The above illustrated distributed sensing system allowed for the detection of the
Figure 28).
changes of the composite action between timber and concrete (using indicators such as
the neutral axis position and the curvature evolution within the CLT slab) in addition to
highlighting the considerable structural changes induced by mechanisms such as concrete
creep and environmental thermos-hygrometric variations.
Sensors 2021, 21, 1818 33 of 83
Sensors 2021, 21, x FOR PEER REVIEW 34 of 87

Figure 28. Fiber arrangement for the Timber–Concrete Composite (TCC) slab monitoring [89].

The above illustrated distributed sensing system allowed for the detection of the
changes of the composite action between timber and concrete (using indicators such as
the neutral axis position and the curvature evolution within the CLT slab) in addition to
highlighting the considerable structural changes induced by mechanisms such as concrete
creep and environmental thermos-hygrometric variations.

4.4. Corrosion
Figure
Figure 28. Tests
28. Fiber
Fiber of Reinforced
arrangement
arrangement for Concrete
forthe
the StructuresComposite
Timber–Concrete
Timber–Concrete Composite(TCC)
(TCC)slab
slabmonitoring
monitoring[89].
[89].
The corrosion of steel reinforcement bars decreases the load-carrying capacity of RC
4.4. Corrosion Tests of Reinforced Concrete Structures
The above
structures illustrated
by reducing distributed sensing
the reinforcement area andsystem allowed
decreasing thefor the detection
latter’s bond with of the
the
The corrosion
changes
surrounding of the
concrete.of steel
compositeWithreinforcement
action
regardsbetween
to thisbars decreases
timber the load-carrying
and concrete
deterioration capacity
(using indicators
phenomenon, DOFS of po-
such
has the RC as
structures
the neutral
tential by reducing
axis
of becoming position the
a crucial reinforcement
and the curvature
SHM area
tool thanks and
evolution
to decreasing
their within the
immunity the latter’s
CLT
to slab)bond
corrosion inandwith
addition the
the po- to
surrounding
highlighting concrete.
the With
considerable regards to
structural this
changesdeterioration
induced by phenomenon,
mechanisms DOFS
such
tential of providing an insight into the internal corrosion levels, areas of pitting corrosion, as has the
concrete
potential
andofenvironmental
creeplocations
crack becoming
and bond a crucial SHM tool
thermos-hygrometric
deterioration [90].thanksvariations.
to their immunity to corrosion and the
potential of providing an insight into the internal
Davis et al. [90], building upon their previous work [91]corrosion levels,
on areas of pitting
corroded corrosion,
RC ties, inves-
crack
tigatedlocations
4.4. Corrosion and
the useTests bond deterioration
of Reinforced
of OFDR DOFS for Concrete[90].
Structures
the detection of corrosion in RC beams. In particular,
six RCDavis
beams et al.
were[90], building upon
instrumented withtheir
nylonprevious
and work [91]coated
polyimide on corroded RC ties, on
fibers installed investi-
both
The corrosion of steel reinforcement bars decreases the load-carrying capacity of RC
gated
the the use
bottom by 20 Mof OFDR
bars andDOFS for the detection
onreinforcement
the top 10 M bars of corrosion in RC beams. In particular, six
structures reducing the area(see
andFigure 29) with
decreasing the alatter’s
cyanoacrylate
bond with adhe-
the
RC beams were instrumented with nylon and polyimide coated fibers installed on both the
sive.
surrounding concrete. With regards to this deterioration phenomenon, DOFS has the po-
bottom 20 M bars and on the top 10 M bars (see Figure 29) with a cyanoacrylate adhesive.
tential of becoming a crucial SHM tool thanks to their immunity to corrosion and the po-
tential of providing an insight into the internal corrosion levels, areas of pitting corrosion,
crack locations and bond deterioration [90].
Davis et al. [90], building upon their previous work [91] on corroded RC ties, inves-
tigated the use of OFDR DOFS for the detection of corrosion in RC beams. In particular,
six RC beams were instrumented with nylon and polyimide coated fibers installed on both
the bottom 20 M bars and on the top 10 M bars (see Figure 29) with a cyanoacrylate adhe-
sive.

(a) (b)
Figure 29.
Figure 29. (a) Specimen’s cross-sectional
(a) Specimen’s cross-sectionallayout
layoutand
and(b)
(b)comparison
comparisonofofthe
thenylon
nylonbottom
bottomDOFS
DOFSmeasured
measuredstrains
strainsbetween
betweena
a non-corrodedcontrol
non-corroded controlbeam
beam(BS-C)
(BS-C)versus
versusa acorroded
corrodedone
one(BS-01)
(BS-01)for
forthree
threeapplied
appliedload
loadsteps,
steps, 20
20 kN,
kN, 60
60 kN
kN and
and 100
100 kN
kN
(adapted from [90]).
(adapted from [90]).

The beams were subjected to varying levels of corrosion by placing them in a bath of
water with 3.5% of the weight NaCl and imposing a current to accelerate the oxidation
process of the reinforcement. The beams were finally tested under three-point bending.
(a)The authors’ conclusions were preceded by a warning on (b) the possible exacerbation of the
Figure 29. (a) Specimen’s cross-sectional layout and (b) comparison of the nylon bottom DOFSto
corrosion effects induced by the test setup when compared in-situ corrosion.
measured Overall,
strains between
a non-corroded control beam the(BS-C)
FOS monitoring system
versus a corroded onewas reportedly
(BS-01) for three able to detect
applied the strain
load steps, 20 kN, distribution
60 kN and 100changes
kN
(adapted from [90]).
 The beams were subjected to varying levels of corrosion by placing them in a bath of
water with 3.5% of the weight NaCl and imposing a current to accelerate the oxidation
process of the reinforcement. The beams were finally tested under three-point bending.
Sensors 2021, 21, 1818 The authors’ conclusions were preceded by a warning on the possible exacerbation 34 ofofthe
83
corrosion effects induced by the test setup when compared to in-situ corrosion. Overall,
the FOS monitoring system was reportedly able to detect the strain distribution changes
along the longitudinal reinforcement due to the corrosion-induced loss of bond between
along the longitudinal reinforcement due to the corrosion-induced loss of bond between
the concrete and reinforcement. Furthermore, it was found that 1.1% and 4.0% corrosion
the concrete and reinforcement. Furthermore, it was found that 1.1% and 4.0% corrosion
levels of the bottom longitudinal reinforcement with short watertight protective sleeves
levels of the bottom longitudinal reinforcement with short watertight protective sleeves
at its ends could cause a decrease in ultimate load of 13% and 39%, respectively. For the
at its ends could cause a decrease in ultimate load of 13% and 39%, respectively. For the
beams with long protective sleeves, instead, 1.4% and 3.8% corrosion lead to a decrease in
beams with long protective sleeves, instead, 1.4% and 3.8% corrosion lead to a decrease in
ultimate capacity by 38% and 37%, respectively.
ultimate capacity by 38% and 37%, respectively.
Lv et al. [92] attempted to evaluate the corrosion-induced damage degree on RC
Lv et al. [92] attempted to evaluate the corrosion-induced damage degree on RC
structures by using BOTDA DOFS to monitor the expansion strains of the former. The
structures by using BOTDA DOFS to monitor the expansion strains of the former. The
authors deployed the DOFS on the rebar in a particular way, baptized in its complexity as
authors deployed the DOFS on the rebar in a particular way, baptized in its complexity as
“corrosion sensor”.
“corrosion sensor”. ItsIts production
production steps steps werewere as
as follows.
follows. First of all,
First of all, aa 55 mm
mm thick
thick mortar
mortar
layer
layer was poured directly on the rebar and left curing for one day before demolding it.
was poured directly on the rebar and left curing for one day before demolding it.
Three days later
Three days later aa33mmlong
longDOFSDOFSwas wastightly
tightlywinded
windedfor forseveral
several turns
turns against
against thethe surface
surface of
of
thethe mortar
mortar with
with both
both its its ends
ends fixed
fixed withwith acrylate
acrylate adhesive.
adhesive. Finally,
Finally, afterafter having
having bilater-
bilaterally
ally led out the fiber, an additional layer of mortar was cast on top of
led out the fiber, an additional layer of mortar was cast on top of the latter, embedding it. the latter, embedding
it.
Two TwoRCRC140140 × 140
× 140 × 180× 180mm mm specimens
specimens (one(one
withwith stirrups
stirrups and and one without)
one without) werewere sub-
subjected
jected to accelerated corrosion and a subsequent finite element
to accelerated corrosion and a subsequent finite element analysis was performed. The analysis was performed.
The latter’s
latter’s resultsresults
werewere consistent
consistent withwith the fiber
the fiber monitored
monitored strainsstrains
provingproving the “rational-
the “rationality” of
ity” of the analysis
the analysis and confirming
and confirming that thethat the so-called
so-called “DOFS “DOFScorrosion corrosion
sensor” sensor”
could becould
used beto
used
monitorto monitor the expansion
the expansion strain ofstrain of the
the steel steel corrosion
corrosion in real-time.
in real-time.
Fan
Fan etet al.
al. [93]
[93] investigated
investigated PPP-BOTDA
PPP-BOTDA DOFS’ DOFS’ ability
ability to
to monitor
monitor the the evolution
evolution of of the
the
corrosion of steel bars embedded inside concrete with the sensors
corrosion of steel bars embedded inside concrete with the sensors deployed both along the deployed both along
the
rebarrebar
andand winding
winding around
around it (with
it (with spacings:
spacings: 0 mm,
0 mm, 2 mm,2 mm,
5 mm 5 mmandand 10 10
mm)mm) as as visible
visible in
in Figure
Figure 30.30.

Figure 30. Photograph of the DOFS deployment on the steel bars: (a) Reference bar (b) L4 longitudinal deployment, (c) S0
spacing 0 mm, (d) S2 spacing 2 mm, (e) S5 spacing 55 mm
mm and
and (f)
(f) S10
S10 spacing
spacing 10
10 mm
mm [93].
[93].

In the first place, the authors studied the potential influence of the sensing cable on the
steel–concrete bond strength through several pull-out tests. It was noticed that whilst no
bond alterations were observed between a fiber-less reference specimen and longitudinally
bonded fiber specimen, the spirally bonded one led to a certain bond reduction (for example,
as the spacing is reduced from 10 mm to 5 mm, the bond strength was reduced by 18%).
Following such, electrochemical testing was carried out to assess the influence of the fiber
 In the first place, the authors studied the potential influence of the sensing cable on
the steel–concrete bond strength through several pull-out tests. It was noticed that whilst
no bond alterations were observed between a fiber-less reference specimen and longitu-
Sensors 2021, 21, 1818 dinally bonded fiber specimen, the spirally bonded one led to a certain bond reduction 35 of 83
(for example, as the spacing is reduced from 10 mm to 5 mm, the bond strength was re-
duced by 18%). Following such, electrochemical testing was carried out to assess the in-
fluence of the fiber installation method on the corrosion resistance of the RC specimen. In
installation method on the corrosion resistance of the RC specimen. In particular, the
particular, the measured
measured strain strainwere
distributions distributions were used
used to quantify to quantify
the corrosion thethickness
layer corrosionand layerto
thickness and to estimate
estimate the corroded volume. the corroded volume.
The
The same
same teamteam published
published another
another research
research article
article [94]
[94] on
on the
the PPP-BOTDA
PPP-BOTDA DOFS DOFS
monitoring
monitoring of the corrosion evolution of a rebar embedded inside concrete with
of the corrosion evolution of a rebar embedded inside concrete with different
different
percentages
percentages of of steel
steelfiber
fiberreinforcement
reinforcement andand different
different depths
depths of a superficially
of a superficially incised incised
notch.
notch.
The rebars were instrumented as specimen S10 in the previously illustrated campaign.cam-
The rebars were instrumented as specimen S10 in the previously illustrated The
paign. The
resulting resulting
DOFS DOFS measurements
measurements allowed
allowed the authors to the authors
conclude thattounder
conclude that surface
identical under
identical
conditions surface
(notch conditions
depth) the (notch
steel depth) the steelthe
fibers reduced fibers reduced
corrosion ofthe
thecorrosion
steel bar of the steel
embedded
bar embedded in the concrete and delayed any corrosion-induced
in the concrete and delayed any corrosion-induced concrete cracking (provoked by concrete cracking (pro-
the
voked
pressurebyof thethepressure
corrosion of products
the corrosion
againstproducts against
the concrete the concrete
matrix). Indeed, matrix).
when the Indeed,
steel
when the steel
fiber content fiber content
increased from increased
0% to 1%, thefrom 0% to
strain 1%, therate
increase strain increase
before ratecracking
concrete before con-was
crete cracking was reduced by 49%, 50%, and 44% for the specimens
reduced by 49%, 50%, and 44% for the specimens with notch depths of 20 mm, 10 mm and with notch depths of
20 mm, 10 mm
0 mm, respectively. and 0 mm, respectively.
Additional
Additionalsimilarsimilarwork
workfromfromFanFanet et al.
al. [23]
[23] was
was aimed
aimed at at evaluating
evaluating through
throughPPP- PPP-
BOTDA
BOTDA DOFS DOFS the the corrosion
corrosion products
products volume
volume and and mass
mass loss
loss of
of steel
steel bars
bars inin addition
addition to to
predicting cracking of
predicting the cracking ofthe
theconcrete
concretecover
coverfor forRCRC specimens.
specimens. TheThe latter
latter areare similar
similar to
to the
the
onesones in the
in the previous
previous articles
articles (see(see Figure
Figure 31)31)withwith
thethe exception
exception of being
of being characterized
characterized by
by
thethe different
different covercover thickness
thickness (28(28
mm,mm, 35 mm,
35 mm, andand 43 mm)
43 mm) andand water-cement
water-cement ratio
ratio (0.4,(0.4,
0.5,
0.5,
andand
0.6).0.6).

Figure 31.
Figure (a) Photography
31. (a) Photography of
of the
the tested
tested specimen
specimen and
and (b)
(b) cut
cut section
section along
along the
the steel
steel corroded
corroded bar
bar [23].
[23].

Accordingtotothe
According theauthors,
authors,the
the DOFS
DOFS results
results showed
showed thatthat the steel
the steel bar’sbar’s corrosion
corrosion pro-
process occurred in three stages at each of which the strains increased approximately
cess occurred in three stages at each of which the strains increased approximately linearly
linearly but at increasing rates. The thickness of the concrete cover reportedly influenced
but at increasing rates. The thickness of the concrete cover reportedly influenced the du-
the duration of each of the three stages. Meanwhile, the increase of the water-to-cement
ration of each of the three stages. Meanwhile, the increase of the water-to-cement ratio led
ratio led to a shortening of the first stage and an increased volume of rust filling the pores
to a shortening of the first stage and an increased volume of rust filling the pores near the
near the interface between steel and the concrete.
interface between steel and the concrete.
4.5. Miscellaneous
4.5. Miscellaneous
The present subsection is intended to collect experimental campaigns that, despite
The present
carrying significantsubsection is each
weight in intended to collect
relative experimental
field, are campaigns
characterized that, despite
by more diverse traits
carrying significant weight in each relative field, are characterized by
when compared to the above reported DOFS-powered monitoring campaigns. Therefore,more diverse traits
when compared
the following to the above
assortment reported
presents DOFS-powered
a high monitoring campaigns. Therefore,
level of heterogeneity.
the following assortment presents a high level of heterogeneity.
Bao et al. [95] discussed the importance of research on the integrity and load-carrying
capacity degradation when a structure is exposed to fire. In particular, as summarized in
a comprehensive review article on the use of fiber optic sensors for Structural Fire Engi-
neering [96], the team published in 2017 two experimental works featuring PPP-BOTDA
monitoring campaigns on the residual performances of both steel and RC structures during
a fire. In the first [97], DOFS were installed along the top and bottom flanges of a hot-rolled
S-shaped A36 mild steel beam which was later exposed to fire. In particular, the fibers were
 Bao et al. [95] discussed the importance of research on the integrity and load-carrying
capacity degradation when a structure is exposed to fire. In particular, as summarized in
a comprehensive review article on the use of fiber optic sensors for Structural Fire Engi-
neering [96], the team published in 2017 two experimental works featuring PPP-BOTDA
Sensors 2021, 21, 1818 36 of 83
monitoring campaigns on the residual performances of both steel and RC structures dur-
ing a fire. In the first [97], DOFS were installed along the top and bottom flanges of a hot-
rolled S-shaped A36 mild steel beam which was later exposed to fire. In particular, the
fibers were
loosely loosely
passed passed
through through
steel steel tubes
tubes (which (which
in turn were in turn were
attached attached
to the to theforsteel
steel beam) the
beam) for the measurement
measurement of thedistributions.
of the temperature temperatureAdistributions. A yoke
yoke was attached atwas attached at
the mid-span of the
the
mid-span
beam andof theloaded
was beam and was loaded
through a pulleythrough
system a(see
pulley system
Figure 32a).(see Figure 32a).

Figure 32.
Figure Applications of
32. Applications of DOFS
DOFS in
in structural
structural fire
fire testing
testing of:
of: (a)
(a) Steel
Steel beam
beam and
and (b)
(b) RC
RC beams
beams [96].
[96].

The resulting DOFS measurements, combined with the measured temperatures and
The resulting DOFS measurements, combined with the measured temperatures and
the building code recommended material parameters, yielded an enhanced thermomechan-
the building code recommended material parameters, yielded an enhanced thermome-
ical analysis of simply supported steel beams subjected to the combined action of thermal
chanical analysis of simply supported steel beams subjected to the combined action of
and mechanical loading. The simulated strains and deflections were then validated using
thermal and mechanical loading. The simulated strains and deflections were then vali-
measurements from a second distributed DOFS strain sensor and two linear potentiometers
dated using measurements from a second distributed DOFS strain sensor and two linear
resulting in accurate predictions of the temperature-dependent material properties. Of the
potentiometers resulting in accurate predictions of the temperature-dependent material
four investigated strain predicting building codes, the European building code provided
properties. Of the four investigated strain predicting building codes, the European build-
the best predictions.
ing code provided the best predictions.
In Bao et al. [95], instead, PPP-BOTDA DOFS were embedded inside RC beams (see
In Bao et al. [95], instead, PPP-BOTDA DOFS were embedded inside RC beams (see
Figure 32b) for condition assessment during the exposition to fire. Here, the distributed
Figure 32b) for condition assessment during the exposition to fire. Here, the distributed
sensor measured temperature distributions with a 2 cm spatial resolution in the beams
sensor measured
thus detecting temperature
non-uniform distributions
temperature with a 2 cmconcrete
distributions, spatial cracks
resolution in the beams
and spalling until
thus detecting non-uniform temperature distributions,
complete failure occurring after excessive concrete spalling. concrete cracks and spalling until
complete
Saidifailure occurring
and Gabor after excessive
[98] presented concrete spalling.
an experimental study on the mechanical behavior of
Textile Reinforced Cementitious Matrix Composites subjected to direct tensile loading. The
specimens were instrumented with OFDR DOFS positioned in their core (see in Figure 33)
and for the study of the local and global behavior of the matrix, of the textile and of the
textile/matrix interface.
 Saidi and Gabor [98] presented an experimental study on the mechanical behavior of
Textile Reinforced Cementitious Matrix Composites subjected to direct tensile loading.
The specimens were instrumented with OFDR DOFS positioned in their core (see in Fig-
Sensors 2021, 21, 1818 37 of 83
ure 33) and for the study of the local and global behavior of the matrix, of the textile and
of the textile/matrix interface.

Figure33.
Figure Positioningof
33.Positioning ofthe
theoptical
opticalfiber
fiberon
onthe
thetextile
textilereinforcement
reinforcement [98].
[98].

Jaafarietetal.
Jaafari al.[99]
[99]stated
statedthat that early-age
early-age shrinkage,
shrinkage, creep
creep andand thermal
thermal conditions
conditions or
or in-
initial cracking could have a significant impact on the dynamic
itial cracking could have a significant impact on the dynamic behavior of RC structures. behavior of RC structures.
Inorder
In ordertotoquantify
quantifythe theimpact
impact ofof early-age
early-age drying
drying shrinkage
shrinkage on onthe the latter’s
latter’s dynamic
dynamic be-
behavior, the authors conducted an experimental campaign
havior, the authors conducted an experimental campaign on two types of RC portal on two types of RC portal
framessubjected
frames subjectedtotopseudo-dynamic
pseudo-dynamic testing.
testing. TheThefirstfirst frame
frame was waskeptkept in endogenous
in endogenous con-
conditions (it was covered with plastic sheets at its early age to prevent
ditions (it was covered with plastic sheets at its early age to prevent any water exchange any water exchange
with the surrounding environment) thus limiting the drying effects. The second one,
with the surrounding environment) thus limiting the drying effects. The second one, in-
instead, was kept in non-endogenous conditions (thus allowing all water exchanges)
stead, was kept in non-endogenous conditions (thus allowing all water exchanges) simi-
similarly to any construction site conditions. The structures were both instrumented
larly to any construction site conditions. The structures were both instrumented with
with Mono-mode optical OFDR DOFS characterized by a silica glass core covered by a
Mono-mode optical OFDR DOFS characterized by a silica glass core covered by a Polyu-
Polyurethane cladding. The distributed sensors reported that, whilst the maximum strains
rethane cladding. The distributed sensors reported that, whilst the maximum strains were
were similar in both structures, the gradient of deformation was more important in the
similar in both structures, the gradient of deformation was more important in the endog-
endogenous portal frame than in the non-endogenous portal frame. As a matter of fact, in
enous portal frame than in the non-endogenous portal frame. As a matter of fact, in the
the former one evident drying cracks appeared. After their early age period, both RC portal
former one evident drying cracks appeared. After their early age period, both RC portal
frames were subjected to an equivalent seismic loading and certain behavioral differences
frames were subjected to an equivalent seismic loading and certain behavioral differences
between the two were observable when subjected to a moderate intensity earthquake.
between the two were observable when subjected to a moderate intensity earthquake.
Liu et al. [100] performed an OFDR DOFS monitoring of an RC wall-beam-strut
jointLiu et al. [100]
subjected to aperformed
monotonicanlateral
OFDRload.DOFSThe monitoring of ansensor
optical fiber RC wall-beam-strut
used in this study joint
subjected to a monotonic lateral load. The optical fiber sensor
is a polyurethane elastomer-coated fiber with a diameter of 0.9 mm without any special used in this study is a pol-
yurethane
protective elastomer-coated
coating in the sensing fiber with
cable. a diameter
The sensors of 0.9 mmglued
were without to aany special
groove protec-
incised on
tive
the steel bar and were positioned on different vertical and horizontal layers of thesteel
coating in the sensing cable. The sensors were glued to a groove incised on the RC
bar and wereDuring
specimen. positioned on different
the test, vertical
the authors and horizontal
reported an excellent layers of the RC
agreement specimen.
between the
During the test, the authors reported an excellent agreement
visually observable cracks against the measured tensile strain profiles. According between the visually observ-
to the
able cracks
authors, theagainst the measured
initiation of crackingtensile
and itsstrain profiles.
location can beAccording
identifiedtoby thelocal
authors,
spikes the
inini-
the
tiation
tensile of
straincracking
profileand(seeits location
Figure 34). can
Thisbe identified
process, by local
though, spikes in the
was mentioned to be tensile
muchstrain
easier
profile
at lower(seeload Figure
levels34).
thanThis process,
at higher onesthough,
due to was mentioned to
the overlapping be much easierspikes
of crack-induced at lowerand
load levels than at higher ones due to the overlapping of crack-induced
the rise of SRAs occurring at higher loads. The extraction of data on crack orientation spikes and and
the
rise of SRAs occurring at higher loads. The extraction of data
depth was possible thanks to the multiplicity of layers on which the DOFS was bonded. on crack orientation and
depth was possible
The authors concluded thanks to the multiplicity
evidencing the abilityof oflayers on which
the optical fibersthe DOFS
inside RCwas bonded.
members to
The authors
capture concluded
its cracking evidencing the ability of the optical fibers inside RC members to
pattern.
capture its cracking
Woods pattern.
et al. [101] tested an RC shear wall with externally-bonded OFDR DOFS-
instrumented Fiber Reinforced Polymer (FRP) sheets placed to improve its performance.
In order to simulate the lateral drift demand along with the kind of damage suffered
by an RC shear wall during a large earthquake event, the specimen was tested under a
predetermined reverse cyclic lateral load sequence. This was achieved by fixing the tested
wall to the laboratory strong floor and having a hydraulic actuator (fixed on a reaction
frame as visible in Figure 35) apply lateral loads to its top.
Sensors 2021, 21, 1818 38 of 83
Sensors 2021, 21, x FOR PEER REVIEW 39 of 87

(e)

Figure 34. (a) Damage status at 0.15% drift, (b–d) are the strain profiles measured in the top fibers A, B and C at drift levels below
0.15% and (e) picture of the cracked specimen (adapted from [100]).

Woods et al. [101] tested an RC shear wall with externally-bonded OFDR DOFS-in-
strumented Fiber Reinforced Polymer (FRP) sheets placed to improve its performance. In
order to simulate the lateral drift demand along with the kind of damage suffered by an
RC shear wall during a large earthquake event, the specimen was tested (e) under a prede-
termined reverse cyclic lateral load sequence. This was achieved by fixing the tested wall
Figure
Figure 34. (a)34. (a) Damage
Damage statusstatus at 0.15% drift,
at 0.15% (b–d)(b–d) arestrain
the strain profiles measured in top
the top fibers
A, BA, B andatCdrift
at drift levels
to thedrift,
laboratory arestrong
the profiles
floor measured
and having in the
a hydraulic fibers
actuator and
(fixedC on levels
a reaction below
frame
below 0.15% and (e) picture
0.15% and (e) picture of the cracked of the cracked
specimen specimen
(adapted (adapted
from [100]).from [100]).
as visible in Figure 35) apply lateral loads to its top.

Woods et al. [101] tested an RC shear wall with externally-bonded OFDR DOFS-in-
strumented Fiber Reinforced Polymer (FRP) sheets placed to improve its performance. In
order to simulate the lateral drift demand along with the kind of damage suffered by an
RC shear wall during a large earthquake event, the specimen was tested under a prede-
termined reverse cyclic lateral load sequence. This was achieved by fixing the tested wall
to the laboratory strong floor and having a hydraulic actuator (fixed on a reaction frame
as visible in Figure 35) apply lateral loads to its top.

Figure 35. Experimental test setup (adapted from [101]).

Figure 35. Experimental test setup (adapted from [101]).
The fibers were bonded horizontally (see Figure 35) to the surface of the outer-most
CFRPThe fibers
layer were
(using anbonded horizontally
epoxy resin) (see Figure
and vertically on the35) to of
back thethe
surface of the
wall. This outer-most
setup permit-
CFRP
ted an layer (usingand
interesting an well-illustrated
epoxy resin) and vertically on the
two-dimensional back strain
spatial of themeasurement
wall. This setup per-
analysis
over the entire area of the shear wall (see Figure 36), the study of the FRP contribution
mitted an interesting and well-illustrated two-dimensional spatial strain measurement to
the resistance of the shear wall and the capturing of the failure mechanisms. The authors
concluded by stating that DOFS was durable, reliable and consistently providing strain
measurements under repeated large strain reversals.
Figure 35. Experimental test setup (adapted from [101]).

The fibers were bonded horizontally (see Figure 35) to the surface of the outer-most
CFRP layer (using an epoxy resin) and vertically on the back of the wall. This setup per-
mitted an interesting and well-illustrated two-dimensional spatial strain measurement
Sensors 2021, 21, x FOR PEER REVIEW 40 of 87

Sensors 2021, 21, 1818
analysis over the entire area of the shear wall (see Figure 36), the study of the FRP contri-
bution to the resistance of the shear wall and the capturing of the failure mechanisms.39 The
of 83
authors concluded by stating that DOFS was durable, reliable and consistently providing
strain measurements under repeated large strain reversals.

Figure
Figure 36. 2Dspatial
36. 2D spatialstrain
straindistribution
distribution in
in the:
the: (a)
(a)vertical
verticaland
and(b)
(b)horizontal
horizontal directions
directions [101].
[101].

4.6. Sub-Section Conclusions

The present section was focused on the DOFS instrumentation of various structural
elements subjected to laboratory experimentation. The reported efforts were categorized
as RC beams, RC ties, RC slabs, corrosion testing and miscellaneous applications. Many
Sensors 2021, 21, 1818 40 of 83

of these, integrated DOFS monitoring with a novel technology named Digital Image
Correlation (DIC) for the observation of superficial strains and for the cracking patterns of
RC structures.
In the former category, the main observations were:
• The authors of most articles were successful in the identification of the flexural strains
inside RC beams and a Cemented Pavement Material (CPM) beam, in the detection of
the cracking patterns and in the monitoring of both the widening and the propagation
of cracks;
• Both thin coating-less DOFS and high-resistance coated DOFS cables were successfully
used to achieve the previous point;
• Polyimide-cladded fibers allowed for the early and accurate detection and localization
of micro-cracks;
• For the bonding of thin coating-less DOFS on the surface of steel bars, the fibers
were commonly positioned at the concavity created by the longitudinal ridge (after
removing the mill-scale and degreasing the area with acetone) before gluing them with
cyanoacrylate adhesive with the addition of a protective layer of a one-component
water-proof oxygen-free silicone rubber;
• The strain of steel rebars embedded inside RC beams measured by means of DOFS
and SG were found to be in good agreement both at crack locations and between
the latter;
• In Berrocal et al. [59], the DOFS extracted pattern was compared against the one
externally sampled by DIC with an agreement of ±3 cm between the two; a range
which is understandable considering the different depth at which the transversal
cracks were measured and their probable lack of perpendicularity compared to the
beams’ axis. A good agreement (±20 µm) was generally observed between DOFS and
DIC crack width;
• For the monitoring of shear failure in RC beams different but wholly valid solutions
were proposed. The first saw the bonding of DOFS not only on the longitudinal
rebars but also on 16 stirrups. The second saw the external deployment of an OFDR
DOFS in a 2D grid on the shear span of a beam in order to obtain its shear cracking
pattern. The third, deployed multiple fibers both internally and externally to the beam
including an upward guided one running perpendicularly to the oblique shear cracks.
All three deployment kinds successfully reported the steel strains, crack patterns and
crack widths;
• OFDR DOFS bonded to longitudinal RC beams’ rebars were able to monitor their
strains whilst withstanding cyclic loadings all the way to failure;
• OFDR DOFS were successfully employed for the study of the steel/concrete bond
degradation due to reinforcement yielding inside RC beams subjected to lateral load-
ing, thus demonstrating their monitoring potential even for high magnitude strains.
The DOFS monitoring of RC ties was mainly used to illustrate the cracking, defor-
mation and bond behavior of RC structures thanks to the latter simplicity and reasonably
good representation of the distribution of internal forces and strains in the tensile zone of
RC structures. Indeed, most authors reported being able to extract plausible concrete–steel
bond-stress values by means of RC ties double pull-out testing. Some key observations
follow here:
• Zhang et al. [79] reported that the DOFS measurement sensitivity parameter decreased
with an increase in the cable coating thickness due to both elastic and inelastic coating
deformations. The authors further reported a certain influence on the DOFS-concrete
bond exercised by the coating stiffness and the cables’ surface properties;
• Sienko et al. [80] raised the issue of the possible existence of an additional source of
inaccuracies in thick-coated fibers. Here, a lower crack-width assessment accuracy
was attributed to the slip between the DOFS core and its external jacket;
Sensors 2021, 21, 1818 41 of 83

DOFS integration to RC and Timber–Concrete Composite (TCC) slabs was commended

by several authors who successfully detected the internal strains up until the onset of reinforce-
ment yielding in addition to obtaining a comprehensive picture of longitudinal temperature
distributions, the heat flux transfer process and the consequent strain distributions.
On DOFS-integrated corrosion testing:
• DOFS has the potential of becoming a crucial SHM tool thanks to their immunity
to corrosion;
• DOFS monitoring was reportedly able to detect the strain distribution changes along
the longitudinal reinforcement due to the corrosion-induced loss of bond between the
concrete and reinforcement;
• Lv et al. [92] introduced the idea of a so-called “corrosion sensor”. It consisted first of
all, of a 5 mm thick mortar layer poured directly on the rebar around which a DOFS is
tightly winded for several turns with a final mortar layer cast on top;
• The influence of a winded DOFS on the steel–concrete bond strength was studied by
means of several pull-out tests. It was noticed that whilst no bond alterations were
observed between a fiber-less reference specimen and longitudinally bonded fiber
specimen, the spirally bonded one led to a certain bond reduction;
• The spiraling DOFS around a steel rebar deployment technique was used for the as-
sessment of the influence of various parameters (concrete cover, amount of embedded
reinforcement fibers) on the amount of corrosion suffered by the rebar.
Finally, particularly noteworthy miscellaneous experimental DOFS applications in-
cluded steel and RC structures fire testing, the testing of Textile Reinforced Cementitious
Matrix Composites and the shear testing of a Reinforced Polymer (FRP) RC wall.

5. Monitoring of the Built Environment: Buildings, Bridges and Roads

The present section focuses on the built environment branch of Civil Engineering of
which its most conspicuous examples are buildings, bridges and roads but which also
includes power plants, dams, railroads and airports amongst others. These structures
cover key roles in the proper functioning, security and comfort of any society, thus their
time-induced deterioration and serviceability-jeopardizing issues occurring throughout
their service lifetime (e.g., corrosion, fatigue, creep and shrinkage-induced shortening
and cracking) should be treated with the equivalent criticality. Revealing data, averagely
representative of the situation of most countries worldwide, is provided by the American
Society of Civil Engineers (ASCE)’s 2017 Infrastructure Report Card [102]. According to
the latter, the U.S. has 614,387 bridges, four in 10 of which are 50 years old or more. Of
the nation’s bridges, 9.1% were structurally deficient in 2016 and, on average, there were
188 million trips across a structurally deficient bridge each day. Furthermore, one out of
every five miles of highway pavement is in poor condition. The most recent estimate puts
the nation’s backlog of bridge rehabilitation needs at USD 123 billion and of highways
rehabilitation needs at USD 420 billion. As observed by Regier and Hoult [103], it is not
feasible to replace all the structures that have surpassed their intended service lives because
of the budget, logistical and environmental concerns that such widespread demolition
and reconstruction process would bring along. The only other possible approach consists
in keeping the assets that are still fit for purpose in service. Hence the importance of
determining their deterioration level, serviceability performance and public safety. Though
visual inspection is the most common assessing methodology [104], it inherently includes
a certain degree of subjectivity together with evident limitations when it comes to damages
occurring on the inside of a structure or located in unreachable locations. As such, the
SHM of Civil Engineering infrastructure proposes the definition of damage identification
strategies performed through accurate quantitative data sampling tools and followed up
by their continuous and real-time monitoring.
The applicability and performance of DOFS to the built environment SHM has been
tested and proved possible in pioneering publications such as Regier and Hoult [103]
Glišić et al. [105], Matta et al. [106] and more. Interestingly, Glišić et al. [105], reported
 occurring
occurring on onthethe inside
inside of of a structure
a structure or or located
located in in unreachable
unreachable locations.
locations. AsAs such,
such, thethe
SHM SHM of Civil
of Civil Engineering
Engineering infrastructure
infrastructure proposes
proposes thethe definition
definition of damage
of damage identification
identification
strategies
strategies performed
performed through
through accurate
accurate quantitative
quantitative data data sampling
sampling tools
tools and and followed
followed upup
Sensors 2021, 21, 1818 by by their
their continuous
continuous andand real-time
real-time monitoring.
monitoring. 42 of 83
TheThe applicability
applicability and and performance
performance of DOFS
of DOFS to the
to the built
built environment
environment SHM SHM hashas
beenbeen
tested
tested andand proved
proved possible
possible in in pioneering
pioneering publications
publications such suchas as Regier
Regier andand Hoult
Hoult [103]
[103]
Glišić
Glišić et al.
et al. [105],
[105], MattaMatta et al.
et al. [106]
[106] and and
more.more. Interestingly,
Interestingly, Glišić
Glišić et al.
et al. [105],
[105], reported
reported thethe
the Brillouin
Brillouin scattering-based
scattering-based DOFS DOFS monitoring
monitoring of a of
1000 a
Brillouin scattering-based DOFS monitoring of a 1000 m long bridge (composed by a con-1000
m m
long long
bridge bridge (composed
(composed by a by
con-
a
crete concrete
crete
slabslab slab
poured
poured poured
on on nine on
nine
steel nine
steel steel
girders
girders girders
thatthat
areare that
thenthen are then
supported
supported supported
by bymoremore by
thanthanmore
50 50 than
columns)
columns) 50
columns)
which
which is stillwhich
is still
currentlyis
currentlystill currently
ongoing,
ongoing, ongoing,
effectively
effectively effectively
making
making making
it, the
it, to to the it, to
authors’
authors’ the authors’
knowledge,
knowledge, knowledge,
thethe long-
long-
estthe
est longest
DOFS
DOFS DOFS monitoring
monitoring
monitoring implementation
implementation
implementation on on a real
a real onstructure.
a real structure.
structure.
Webb
WebbWebb etetal.
etal.
al. [107]
[107][107] picked
picked
picked upuponon
up on
thethe
the previous
previous
previous concepts
concepts
concepts tototo perform
perform
perform a aBOTDR
aBOTDR
BOTDR DOFS
DOFSDOFS
monitoring of the Nine Wells Bridge in Cambridge (UK) carrying a new road over main
monitoring
monitoring ofofthe
the Nine
Nine Wells
WellsBridge
Bridgein Cambridge
in Cambridge (UK) carrying
(UK) a
carrying new a road
new over
road the
overthethe
rail
mainmainline connecting
railrail
lineline Cambridge
connecting
connecting to
Cambridge
Cambridge London (represented
to London
to London in
(represented
(represented Figure 37).
in Figure
in Figure 37).37).

Figure
Figure
Figure 37.
37.37.
NineNine Wells
Wells
Nine Bridge
Bridge
Wells [107].
[107].
Bridge [107].

TheThe
The bridge,
bridge,
bridge, aathree-span,
three-span,
a three-span, pre-tensioned,
pre-tensioned,
pre-tensioned, prestressed
prestressed
prestressed concrete
concrete
concrete beam-and-slab
beam-and-slab
beam-and-slab bridge
bridge
bridge
waswas monitored
monitored
was monitored with
with the
the the
with goal
goalgoal of observing
of observing the long-term
the long-term
of observing behavior
behavior
the long-term of the structure
of the structure
behavior for dam-
for dam-
of the structure for
ageage detection
detection
damage purposes
purposes
detection andand
purposes in
in order
andorder to compare
intoorder
compare thethe
to compare in-situ
in-situ
the measurements
measurements
in-situ to the
to the
measurements various
various
to the
empirical
empirical
various creep
creep and
empirical and shrinkage
shrinkage
creep models
models
and shrinkage usedused inused
in the
models the design
design phase.
phase.
in the design Six beams
Sixphase.
beams were
were
Six instru-
instru-
beams were
mented
mented with
with
instrumented fiber-optic
fiber-optic
with cables
cables
fiber-optic (as(as
cables in
in (as Figure
Figure 38)38)
in Figure after
after
38) thethe
after prestressing
prestressing
the strands
strands
prestressing hadhad
been
had been
been
pre-tensioned
pre-tensioned
pre-tensioned butbut
but before
before that
before that
that steel
steel reinforcing
reinforcing
steel reinforcing stirrups
stirrups
stirrupshadhad
had been
been
been tied
tied inplace
inin
tied place
place inthe
inin
thethe mold.
mold.
mold.

(a)(a) (b)(b)
Figure
Figure
Figure 38.
38.38.
BeamBeam
Beam cross
cross section
section
cross showing
showing
section thethe
showing thefiber-optic
fiber-optic cable
cable
fiber-optic locations:
locations:
cable (a) (a) illustration
illustration
locations: (a) andand (b)
(b)and
illustration photograph
photograph (adapted
(adapted
(b) photograph from
from
(adapted
[107]).
[107]).
from [107]).

In particular, two types of fiber-optic cables were installed in the six beams; one for
the sensing of mechanical strains (defined as total strain fibers) and one for the temperature
strains (defined as temperature fibers). Interestingly, the latter cable was filled with the
gel inside which the sensing fibers were suspended (this way, the shear transfer between
the concrete and the sensing core was prevented). With such dual fiber deployment, the
temperature effects on the Brillouin frequency shift could be removed, such that the true
 In particular, two types of fiber-optic cables were installed in the six beams; one for
the sensing of mechanical strains (defined as total strain fibers) and one for the tempera-
ture strains (defined as temperature fibers). Interestingly, the latter cable was filled with
Sensors 2021, 21, 1818 the gel inside which the sensing fibers were suspended (this way, the shear transfer 43 of 83
be-
tween the concrete and the sensing core was prevented). With such dual fiber deployment,
the temperature effects on the Brillouin frequency shift could be removed, such that the
true strain in the fiber (due to applied loading as well as time-dependent effects and the
strain in the fiber (due to applied loading as well as time-dependent effects and the thermal
thermal response of the structure) could be calculated. The authors concluded by remark-
response of the structure) could be calculated. The authors concluded by remarking that
ingthe
thatmeasured
the measured
creepcreep and shrinkage-induced
and shrinkage-induced strains
strains werewere found
found to betoin
bereasonably
in reasonablygood
good agreement with both Eurocode 2 [108] and Collins and Mitchell
agreement with both Eurocode 2 [108] and Collins and Mitchell [109]’s models. [109]’s models.
Similar investigation
Similar investigation procedures
procedures were undertaken
were undertaken by by
Cong
Conget al. [110]
et al. whilst
[110] at- at-
whilst
tempting
temptingto evaluate
to evaluatethethe
prestress
prestress lossloss
of four 11.9
of four mm
11.9 prestressed
prestressedconcrete
concretebeams
beams of of
a a
newly-constructed railway bridge in Staffordshire (UK). Both immediate
newly-constructed railway bridge in Staffordshire (UK). Both immediate prestress losses prestress losses
(caused by by
(caused thethe
elastic shortening
elastic shorteningof concrete)
of concrete) andand
thethe
time-dependent
time-dependent prestress
prestresslosses (due
losses (due
to steel relaxation, concrete shrinkage and creep) were captured by means
to steel relaxation, concrete shrinkage and creep) were captured by means of distributed of distributed
(BOTDR
(BOTDR DOFS)
DOFS)andand
discrete (FBG)
discrete (FBG)fiber optic
fiber sensors
optic installed
sensors during
installed duringthethe
manufacturing
manufacturing
process (as(as
process visible in Figure
visible 39).39).
in Figure TheThelatter sensors
latter measured
sensors measured both strains
both andand
strains temperature
temperature
(for(for
compensation
compensation purposes).
purposes).

Figure 39. 39.

Figure (a) DOFS deployment
(a) DOFS layout
deployment and (b)
layout andprestressed concrete
(b) prestressed beam beam
concrete cross-sections—edge
cross-sections—edge
TYE7 beams (left) and internal TY7 beams (right) [110].
TYE7 beams (left) and internal TY7 beams (right) [110].

TheThe
measured
measured prestress losses
prestress werewere
losses compared with with
compared the predicted ones calculated
the predicted ac-
ones calculated
cording to both
according the European
to both the Europeanand and
American
Americanstandards.
standards.TheThe
outcome
outcome waswas
an an
apparent
apparent
underestimation
underestimation from
from thethe code’s
code’s part(possibly
part (possiblydue
duetotothe
theinaccurate
inaccurateestimations
estimations of vari-
various
ousinput
inputparameters)
parameters)which
whichwas wasmore
morepronounced
pronouncedininthe theearly-age
early-ageversus
versuscompared
comparedtotothe
thetwo
twototothree
threeyears’
years’mark.
mark.
An additional
An additional bridge bridge
OFDR OFDR
DOFSDOFS monitoring
monitoring is reported
is reported in et
in Barrias Barrias et al.
al. [111]. [111].
Here,
Here, Sarajevo Bridge (Barcelona, Spain) was monitored during
Sarajevo Bridge (Barcelona, Spain) was monitored during some deck-enlarging construc- some deck-enlarging
tionconstruction works
works in order in orderato
to perform performof
follow-up a the
follow-up
stressesofinduced
the stresses
in its induced in its three
three prestressed
prestressed concrete box girder beams. Thus, the shorter of the two bridge spans saw the
longitudinal gluing of a single 50 m DOFS inside one of its beams (as visible in Figure 40).
Sensors 2021, 21, x FOR PEER REVIEW 45 of 87

Sensors 2021, 21, 1818 44 of 83

concrete box girder beams. Thus, the shorter of the two bridge spans saw the longitudinal
gluing of a single 50 m DOFS inside one of its beams (as visible in Figure 40).

(a) (b)
Figure
Figure40.
40.(a)
(a)General
Generalscheme
schemeof
ofDOFS
DOFSmonitoring
monitoringand
and (b)
(b)its
its photograph
photograph inside
inside the
the box
box girder
girder (adapted
(adapted from
from [111]).
[111]).

The authors
The authors reported
reportedthe thestrain
strainevolution
evolutionduring
during thethe
construction
construction phase
phase andand
calculated
calcu-
that, despite the strain variation (biggest one being − 304 µε), excessive
lated that, despite the strain variation (biggest one being −304 με), excessive stresses were stresses were not
transmitted
not transmittedto the concrete
to the during
concrete duringthe the
construction
constructionworks (maximum
works (maximum 11.42 MPa).
11.42 MPa).
AAcritical
criticalscenario
scenarioduring
duringCivil CivilEngineering
Engineering works
works might
might ensue
ensue after
after thethe pouring
pouring of
of mass
mass concrete
concrete structures
structures (such (such as dams,
as dams, nuclear nuclear
plants,plants,
massivemassive foundations,
foundations, bridgebridge
piers,
piers,slabs
thick thickand
slabs and skyscraper
skyscraper columns).columns).
Indeed, Indeed, afterconcrete
after the the concrete is poured,
is poured, the cement
the cement hy-
hydration reaction produces a large internal temperature upsurge.
dration reaction produces a large internal temperature upsurge. Due to the poor thermal Due to the poor thermal
conductivityofofconcrete,
conductivity concrete,a huge
a huge gradient
gradient formsforms
soonsoon
afterafter between
between the extremely
the extremely hot corehot
core of the pouring and its quickly cooling outer surface. Considering
of the pouring and its quickly cooling outer surface. Considering the poor tensile strength the poor tensile
strength
of of early
early age age concrete,
concrete, the risingthe risingtensile
thermal thermal tensileatstresses
stresses at themight
the surface surface might
cause cause
thermal
thermal cracking.
cracking. This couldThis could potentially
potentially lead to aprematurely
lead to a structure structure prematurely
failing to reach failing to reach
its service-
its serviceability requirements or even suffering a structural failure.
ability requirements or even suffering a structural failure. Some solutions for the preven- Some solutions for
the prevention of this phenomenon are aggregates precooling,
tion of this phenomenon are aggregates precooling, pipe cooling, superficial thermal in-pipe cooling, superficial
thermal and
sulation insulation and alternative
alternative bay construction
bay construction [112]. Nevertheless,
[112]. Nevertheless, the importance the importance
of monitor- of
monitoring the concrete’s temperature during the construction
ing the concrete’s temperature during the construction period (pertaining to the SHM period (pertaining to the
SHM cannot
field) field) cannot be understated.
be understated.
On such a topic,
On such a topic, Li Li et
et al.
al. [20]
[20] reported
reported aa BOTDABOTDA DOFS DOFS temperature
temperature monitoring
monitoring of of aa
bearing platform
bearing platform of of aamain
mainpierpierpilepilecap
cappertaining
pertaining to to
thethe
1073 m long
1073 m longHezhang
Hezhang super-long-
super-
span bridge. The 58 m long optical fiber cable was deployed in a planar cross-section (as
long-span bridge. The 58 m long optical fiber cable was deployed in a planar cross-section
sketched in Figure 41) in areas resistant to damage and tied to the rebar using cable ties.
(as sketched in Figure 41) in areas resistant to damage and tied to the rebar using cable
The concrete temperature variations were measured for four consecutive days after pouring
ties. The concrete temperature variations were measured for four consecutive days after
and three extra days in the long-term by means of both DOFS, point temperature sensors
pouring and three extra days in the long-term by means of both DOFS, point temperature
and precision thermometers for comparison purposes. The DOFS readings (characterized
sensors and precision thermometers for comparison purposes. The DOFS readings (char-
by accuracy of 0.3 ◦ C) visible in Figure 42 were reported to be in satisfactory agreement
acterized by accuracy of 0.3 °C) visible in Figure 42 were reported to be in satisfactory
with those extracted with conventional temperature sensors (the maximum difference
agreement ◦with those extracted with conventional temperature sensors (the maximum
being 0.23 C).
difference being 0.23 °C).
Ouyang et al. [112] deployed a Raman-based DOFS to measure and monitor the
concrete temperature during the sequential construction of an intake tower of the Qianping
reservoir project, on the Huai River (China). In particular, the authors introduced a
framework for the evaluation and control of the concrete cracking risk by integrating the
DOFS measurements with an inverse analysis method based on temperature simulation and
consequently thermal stress simulation. The DOFS was deployed on a horizontal double S
layout (bonded on a steel frame) with the peculiarity that the single cable is passed through
the designed path twice. As a result, the temperature distribution along the cable always
ended up showing a mirrored profile which facilitated the mapping of the measuring
points on the cable. As previously mentioned, in order to expand the significances of
 Sensors 2021, 21, 1818 45 of 83

the local temperature measurements to the whole structure, numerical simulations of

temperature and thermal stresses were conducted by FEM analysis. The effectiveness of
the above methodology to the construction process, was verified when, upon the drilling
Sensors
Sensors 2021,
2021, 21,
21, xx FOR
FOR PEER
PEER REVIEW
REVIEW 46 of
of 87
out of several concrete cylinders (hydration around 90 days), their examination46 87
showed
hardly any concrete cracking.

Figure
Figure
Figure 41. (a)(a)
41.
41. (a) AAA3D3D
3D sketch
sketch
sketch and
and
and (b)
(b)
(b) across-section
cross-section
aa cross-section ofof
of the
the
the DOFS
DOFS
DOFS deployment
deployment
deployment layout
layout
layout for
for
for temperature
temperature
temperature measurements
measurements
measurements inin
in thethe
the
study case
study pile
case cap
pile [20].
cap [20].
study case pile cap [20].

Figure
Figure
Figure 42.42.
42. Temperature
Temperature
Temperature variation
variation
variation asas
as afunction
function
aa function ofof
of the
the
the DOFS
DOFS
DOFS coordinates
coordinates
coordinates specified
specified
specified inin
in Figure41
Figure
Figure 4141[20].
[20].
[20].

Moving on
Ouyang from deployed
the bridge’s SHM, interesting results are reported by De Battista
Ouyang et et al.
al. [112]
[112] deployed aa Raman-based
Raman-based DOFS DOFS to to measure
measure and and monitor
monitor the the con-
con-
et al. [113,114] regarding
crete the long-term BOTDA DOFS monitoring of two columns and two
crete temperature
temperature during during the the sequential
sequential construction
construction of of an
an intake
intake tower
tower ofof the
the Qianping
Qianping
core walls (as in Figure 43a) of the 163 m, 50 storeys tall Principal Tower (London, UK).
reservoir
reservoir project,
project, onon the
the Huai
Huai River
River (China).
(China). In In particular,
particular, the
the authors
authors introduced
introduced aa frame-
frame-
The goal, assessing the combined structural shortening caused by the superimposed
work
work for for the
the evaluation
evaluation and and control
control ofof the
the concrete
concrete cracking
cracking riskrisk by
by integrating
integrating the the DOFS
DOFS
load, creep and shrinkage during the construction phase.
measurements
measurements with an inverse analysis method based on temperature simulation and
with an inverse analysis method based on temperature simulation and
Indeed, the authors highlighted the importance of establishing the floor-to-floor axial
consequently
consequently thermal
thermal stressstress simulation.
simulation. The The DOFS
DOFS was was deployed
deployed on on aa horizontal
horizontal double
double
shortening of vertical load-bearing elements and their differential shortening (whenever
SS layout
layout (bonded
(bonded on on aa steel
steel frame)
frame) with
with thethe peculiarity
peculiarity thatthat the
the single
single cable
cable isis passed
passed
they are characterized by different stiffness) for the installation of finishes and partitions
through
through the
the designed
designed path
path twice.
twice. As
As aa result,
result, the
the temperature
temperature distribution
distribution along
along the
the cable
cable
on the lower floors. Thus the importance of integrating the designers’ empirical shortening
always
always ended
ended up
up showing
showing aa mirrored
mirrored profile
profile which
which facilitated
facilitated the
the mapping
mapping of
of the
the meas-
meas-
calculations (adjusting them if necessary) with in-situ continuous distributed measure-
uring
uring points
points on the
onwasthe cable.
cable.theAs previously
Asideal
previously mentioned,
mentioned, in
in order
order to
to expand
expand the significances
ments. DOFS then candidate for such goal, allowing for thethe significances
calculation of the
of the
the local
ofaxial local temperature measurements to the whole structure, numerical simulations of
deformation at any time during the construction and along the whole height of of
temperature measurements to the whole structure, numerical simulations the
temperature
temperature and
and thermal
thermal stresses
stresses were
were conducted
conducted by
by FEM
FEM analysis.
analysis. The
The effectiveness
effectiveness of
of
completed elements. Consequently, each monitored element was instrumented with two
the
the
DOFSabove
above methodology
methodology
(measuring to
to the
strains the
andconstruction
construction
temperature) process,
fixed was
process, was
to verified
theverified when,
when, upon
reinforcement upon
(as inthe
the drilling
drilling
Figure 43b)
out
out of
of several
several concrete
concrete cylinders
cylinders (hydration
(hydration around
around 90
90 days),
days),
before embedding them inside concrete. The DOFS monitoring started on 3rd September their
their examination
examination showed
showed
hardly
2016 any
hardly any concrete
and concrete
since then cracking.
cracking.
measurements (with a spatial resolution of 1 m) were performed
Moving
Moving on
on from
from
at least twice every hour. the
the bridge’s
bridge’s SHM,
SHM, interesting
Expectedly, authorsresults
interesting
the results
detectedare reported
arethat
reported by
by De
the columns De Battista
Battista et
shortenedet
al. [113,114]
[113,114] regarding
al.significantly regarding the
the long-term
long-term BOTDA
BOTDA DOFS
DOFS monitoring
monitoring of
more than the walls throughout the construction causing a non-negligibleof two
two columns
columns and
and two
two
core
core walls
walls (as
(as in
in Figure
Figure 43a)43a) of of the
the 163
163 m,m, 50
50 storeys
storeys tall
tall Principal
Principal Tower
Tower (London,
(London, UK). UK).
Sensors 2021, 21, 1818 46 of 83

differential shortening between the two. For example, as visible in Figure 44 whilst column
Sensors 2021, 21, x FOR PEER REVIEWC8 suffered a shortening of 60.3 mm at level 38, simultaneously wall W2 shortened47 byofonly
87

10.2 mm.

(a) (b)
Figure 43. (a) Location of the instrumented columns and walls and (b) DOFS tied to the reinforce-
ment prior to concreting [114].The goal, assessing the combined structural shortening
caused by the superimposed load, creep and shrinkage during the construction phase.
Indeed, the authors highlighted the importance of establishing the floor-to-floor axial
shortening of vertical load-bearing elements and their differential shortening (whenever
they are characterized by different stiffness) for the installation of finishes and partitions
on the lower floors. Thus the importance of integrating the designers’ empirical shorten-
ing calculations (adjusting them if necessary) with in-situ continuous distributed meas-
urements. DOFS was then the ideal candidate for such goal, allowing for the calculation
of the axial deformation at any time during the construction and along the whole height
of the completed elements. Consequently, each monitored element was instrumented
with two DOFS (measuring strains and temperature) fixed to the reinforcement (as in Fig-
ure 43b) before embedding them inside concrete. The DOFS monitoring started on 3rd
September 2016 and since then measurements (with a spatial resolution of 1 m) were per-
formed at least twice every hour. Expectedly, the authors detected that the columns short-
ened significantly more (a) than the walls throughout the construction (b)causing a non-negli-
gible differential shortening between the two. For example, as visible in Figure 44 whilst
Figure
Figure43.
column (a)
43.
C8 Location
(a) Locationaof
suffered the
the instrumented
ofshortening
instrumented columns
columns
of 60.3 and
mm atand walls
walls
level and
38,and (b)DOFS
(b) DOFStied
tiedtoto
simultaneously the
the
wall reinforce-
reinforcement
W2 short-
ment
priorprior
to to concreting
concreting
ened by only 10.2 mm. [114]. [114]. The goal, assessing the combined structural shortening
caused by the superimposed load, creep and shrinkage during the construction phase.
Indeed, the authors highlighted the importance of establishing the floor-to-floor axial
shortening of vertical load-bearing elements and their differential shortening (whenever
they are characterized by different stiffness) for the installation of finishes and partitions
on the lower floors. Thus the importance of integrating the designers’ empirical shorten-
ing calculations (adjusting them if necessary) with in-situ continuous distributed meas-
urements. DOFS was then the ideal candidate for such goal, allowing for the calculation
of the axial deformation at any time during the construction and along the whole height
of the completed elements. Consequently, each monitored element was instrumented
with two DOFS (measuring strains and temperature) fixed to the reinforcement (as48
Sensors 2021, 21, x FOR PEER REVIEW in of
Fig-
87
ure 43b) before embedding them inside concrete. The DOFS monitoring started on 3rd
September 2016 and since then measurements
(a) Column C8 (with a spatial resolution of 1 m) were per-
formed at least twice every hour. Expectedly, the authors detected that the columns short-
ened significantly more than the walls throughout the construction causing a non-negli-
gible differential shortening between the two. For example, as visible in Figure 44 whilst
column C8 suffered a shortening of 60.3 mm at level 38, simultaneously wall W2 short-
ened by only 10.2 mm.

(b) Core wall W1

Figure
Figure44.
44.The
Thetotal
totalcumulative
cumulativeaxial
axialdisplacement
displacement(negative
(negative==shortening)
shortening)ofofthe
theinstrumented
instrumentedcolumns
columnsC8 C8and
andwall
wallW1
W1
measured at the mid-height of every level during the first 12 months of construction (adapted from [114]).
measured at the mid-height of every level during the first 12 months of construction (adapted from [114]).

This
Thisalso
alsooccurred,(a)with
occurred, Column
with C8 magnitude,
aasmaller
smaller magnitude,between
betweentwo
twocolumns
columnswith
withdifferent
different
cross-sections
cross-sectionsfor
forexample
exampleatatthe
theend
endof
ofJuly
July2017,
2017,the
themaximum
maximumshortening
shorteningrecorded
recordedatat
level 31 on column C8 was 36.6 mm whilst that on column C9 was 32.6 mm, approximately
11% less.
Qi et al. [115] also instrumented in a similar manner a structural wall pertaining to a
deep foundation pit supporting system for the excavation of a subway station in Suzhou
Sensors 2021, 21, 1818 47 of 83

level 31 on column C8 was 36.6 mm whilst that on column C9 was 32.6 mm, approximately
11% less.
Qi et al. [115] also instrumented in a similar manner a structural wall pertaining to a
deep foundation pit supporting system for the excavation of a subway station in Suzhou
(China). Here, a PPP-BOTDA DOFS integrated more conventional monitoring methods,
namely steel-bar meters, inclinometer, and water-level observation hole. According to the
authors, at each excavation stage, the strains induced on the wall by the active and passive
earth pressure were easily obtained by means of DOFS, thus leading to the assertion that
the fiber could easily substitute the other classic tools.
Broth and Hoult [116] assessed the dynamic sensing capabilities of an OFDR DOFS
system by performing in-situ monitoring of a 9.24 m cast-in-place RC T-beam located below
a multi-purpose room located in the Queen’s University campus in Kingston, Ontario,
Canada. The application of the dynamic loads was achieved by having 38 students align
themselves across the full span of the classroom (distributed test) and tightly gather over
the midspan of the beam (concentrated test) before repeatedly jumping in unison over
the instrumented specimen. The DOFS output led the authors to confirm that, through
the use of the OFDR DOFS system, both static and dynamic concrete strains could be
successfully measured. Furthermore, the observed trends in the data from both the linear
potentiometers and distributed sensors were in good agreement, suggesting the latter’s
ability to accurately capture a beam’s true dynamic behavior. Finally, DOFS was reported
to successfully provide the dynamic response factor of the instrumented structure, the
distributed crack widths and the deflected shapes.
Brault and Hoult [117] investigated the use of OFDR DOFS to monitor three beams
of a newly constructed RC building subjected to load testing. The investigated structures
were three cast-in-place RC elements, namely a beam, a drop panel and a larger beam. The
elements were first sanded along the fiber pathway, the latter was then cleaned with water
and 99% isopropyl alcohol and, on it, the fiber was bonded with a two-part adhesive. The
load test was performed as follows. For the beam and the drop, the measurements were
performed prior to loading, during an initial application of the load (using two scissor
lifts), during the positioning of six scissor lifts at once and finally once all scissor lifts had
been removed. According to the authors, DOFS efficaciously captured inflection points,
moment transfer at the supports, crack locations and crack openings. Finally, the authors
reported the average crack spacings and the maximum measured crack openings.
Proceeding to the third and last topic of the present section, Wang et al. [118] used five
BOTDA DOFS for the detection of transversal and longitudinal strains inside a multilayered
asphalt pavement under heavy traffic and temperature loads. They pointed out that the
deflection-induced tensile strains in the bottom pavement layers could lead to potentially
permanent deformations under heavy traffic flow and at high temperatures. The DOFS
extracted results revealed that the paving of the second asphalt concrete course could bring
about large transversal and longitudinal strains on the previously built bottom layer due to
the compressive rolling action. Later, when the road became operational, different tensile
strain evolutions were observed in the transversal and longitudinal sections. For example,
the asphalt pavement adjacent to the intersection exhibited high strain levels induced by
vehicle turning and braking effects. It was therefore deemed prone to tensile strain-induced
cracking. Large tensile strains were also detected along the vehicle-track direction, even
though at a much lesser degree than the former. The daily temperature variations altered
both transversal and longitudinal strains increasing (or decreasing) the measured strains of
around 100 µε.
Rabaiotti et al. [119] acknowledged the presence of similar risks for airfield pavements,
especially when taking into consideration their considerably larger loading. The authors
further elucidated that, for such a particular application, the main challenge lied in avoiding
damage to DOFS during the asphalt compaction process which usually involves high
temperatures (exceeding 140 ◦ C). To tackle this issue, ad hoc strain (V3 and V9) and
 track direction, even though at a much lesser degree than the former. The daily tempera-
ture variations altered both transversal and longitudinal strains increasing (or decreasing)
the measured strains of around 100 με.
Rabaiotti et al. [119] acknowledged the presence of similar risks for airfield pave-
ments, especially when taking into consideration their considerably larger loading. The
Sensors 2021, 21, 1818 48 of 83
authors further elucidated that, for such a particular application, the main challenge lied
in avoiding damage to DOFS during the asphalt compaction process which usually in-
volves high temperatures (exceeding 140 °C). To tackle this issue, ad hoc strain (V3 and
V9) and temperature
temperature (T) cables
(T) sensing sensing cables
were were developed
developed (illustrated
(illustrated in Figurein Figure
45a,b) 45a,b)
using using
armored
armored DOFS
DOFS cables, cables,
metal metal
tubes andtubes and PA
PA outer outer sheaths.
sheaths.

(a) (b)

(c) (d)
Figure 45. (a)
(a) Photograph
Photograph ofof the
the DOFS
DOFS cables,
cables, (b)
(b) an
an illustration
illustration of
of their
their cross-sections
cross-sections and (c) of
of their
their designed
designed installation
installation
position and finally
position and finally (d)
(d) photography
photography of
of the
the deployment
deployment process
process (adapted
(adapted from
from [119]).
[119]).

In
In the present project,
project, BOTDR
BOTDRand andRayleigh-based
Rayleigh-basedSwept SweptWavelength
WavelengthInterferometry
Interferome-
try
DOFSDOFSwerewere embedded
embedded in in
two two pavement
pavement layersinin150
layers 150mmlong
longloops.
loops. The
The deployment
steps were
steps were performed
performed as as follows.
follows. Initially,
Initially, the
the fibers
fibers were
were laid
laid out
out on
on the
the substrate
substrate layer,
layer,
Sensors 2021, 21, x FOR PEER REVIEW
followed by
followed by their
their pre-straining
pre-straining and
and fixing in place (Figure 45c,d). After After such, for50better
of 87
protection against the rolling compactor machine, the sensors were covered
against the rolling compactor machine, the sensors were covered manually manually with
finer finer
with asphalt, before
asphalt, the overlying
before layerlayer
the overlying was poured.
was poured.The test
The started whenwhen
test started a plane was
a plane
pushed
was backward
pushed in a perpendicular
backward direction
in a perpendicular to the to
direction fiber
theloop
fiberresulting in sampled
loop resulting strains
in sampled
illustrated
strains in Figure
illustrated 46.
in Figure 46.

Figure
Figure 46.
46. Strains
Strains measured
measured along
along the
the sensors
sensors when
when the
the study
study case airplane rear landing gear was rolled on them [119].

The authors
The authors observed
observed that
that the
the stiffer
stiffer V3 cable seemed less capable of following the
pavement’s high strain gradients versus
pavement’s high strain gradients versus itsits
V9 V9
counterpart. Additionally,
counterpart. the strains
Additionally, mea-
the strains
measured in the lowest layer (13.5 cm depth) were expectedly smaller. In conclusion, this
DOFS-based testing methodology was commended for the detection of problematic zones
and quality issues in the realized pavement in addition to validating the design assump-
tions.
Sensors 2021, 21, 1818 49 of 83

sured in the lowest layer (13.5 cm depth) were expectedly smaller. In conclusion, this
DOFS-based testing methodology was commended for the detection of problematic zones
and quality issues in the realized pavement in addition to validating the design assump-
tions.
Finally, an additional transportation-focused SHM application is also reported in
Wheeler et al. [120] where OFDR DOFS (OBR Luna ODiSI-B model) were employed for the
assessment of distributed dynamic strains of a steel rail during the passage of rail traffic. A
preliminary laboratory test was aimed at studying the difference in monitoring performance
whenever the fiber was bonded to the rail with minimal and optimal (grinding and cleaning)
surface preparation. After having observed that both yielded similar results, the authors
commended the former as the best option for rapid field installations. Additionally, the
authors defended the option of bonding the fibers to the rail foot and the railhead which,
despite being prone to be affected by localized strains under heavy train wheels, allowed
for more accurate rail curvature calculations thanks to the greater distance between the
Sensors 2021, 21, x FOR PEER REVIEW two DOFS. The following reported step was a field test performed near Kingston (Canada)51 of 87
where the DOFS was bonded according to the guidelines established earlier as51visible
Sensors 2021, 21, x FOR PEER REVIEW of 87 in
Figure 47.

Figure
Figure47.
47.Installation
Installation of fiber on
on rail[120].
[120].
Figure 47. Installation offiber
fiber onrail
rail [120].

Vehicular
Vehicular loading
Vehicular loading under
loading under low
underlow vibration
lowvibration conditions
vibrationconditions
conditions (hi-rail
(hi-rail vehicle)
vehicle)
(hi-rail and
vehicle) and high
high
and vibration
vibration
high vibration
conditions
conditions (in-service
(in-service passenger
passenger train)
train) was
was applied
applied on
on the rail.
the Figure
rail. Figure48
conditions (in-service passenger train) was applied on the rail. Figure 48 represents represents
48 the the
represents the
strains
strains measured
strains measured by
measured by thethe upper
the upper and
upper and lower
and lower fibers
lower fibers for
fibers for the
for the hi-rail
the hi-rail vehicle
hi-rail vehicleat three
vehicle at different
at three
three different
different
positionsalong
positions
positions along the
the instrumented
instrumented section
sectionofofthe
thetrack.
track.

Figure 48. DOFS-sampled strains for three different hi-rail vehicle positions along the track (adapted from [120]).
Figure 48. DOFS-sampled strains for three different hi-rail vehicle positions along the track (adapted from [120]).
Figure 48. DOFS-sampled strains for three different hi-rail vehicle positions along the track (adapted from [120]).
The authors reported that it was feasible to study the variability in track support
The
along authors
a section reported
of rail on the that it was
grounds feasible
of DOFS railto study the
curvature variability in
measurements. track
Yet, theysupport
also
acknowledged,
along a section ofthat
rail despite dynamicofdistributed
on the grounds rail strainsmeasurements.
DOFS rail curvature producing lowYet, vibration
they also
could be measured,
acknowledged, that the same dynamic
despite thing could not be saidrail
distributed in the presence
strains of high low
producing vibrations.
vibration
Sensors 2021, 21, 1818 50 of 83

The authors reported that it was feasible to study the variability in track support
along a section of rail on the grounds of DOFS rail curvature measurements. Yet, they also
acknowledged, that despite dynamic distributed rail strains producing low vibration could
be measured, the same thing could not be said in the presence of high vibrations. Indeed,
the quality factor of the strains resulting from the transit of the passenger train was below
the DOFS manufacturer’s minimum recommended threshold.
Conclusively, the present sub-section started describing various DOFS applications
for the SHM assessment of bridges. The following key observations were reported:
• DOFS can be embedded inside prestressed concrete beams, preferably combining a
strain sensing fiber with a temperature sensing one in order to remove any temperature
effect on the Brillouin frequency shift, such that the true strain in the DOFS can be
calculated. Alternatively, for box girder beams, DOFS can be simply glued inside their
hollow cross-section.
• A reported application described the temperature sensing cable to be filled with gel
inside which the DOFS was suspended in order to prevent the shear transfer between
the concrete and the sensing core;
• DOFS was successfully implemented inside bridge RC beams for the purpose of
monitoring damages and both immediate and time-dependent effects (prestress losses
for instance).
Furthermore, the temperature monitoring during the pouring of mass concrete struc-
tures (dams, nuclear plants, massive foundations, etc.) was successfully performed by
means of both BOTDA DOFS and Raman backscattering-based DOFS with an accuracy of
0.3 ◦ C.
On the monitoring of buildings by means of DOFS:
• The shrinkage and creep-induced shortenings (and differential shortenings) of the
columns and structural walls of high-rise buildings can be monitored with a single
BOTDA DOFS stretching across all the building’s floors. The sampled differential
shortenings in such kind of structures can be very relevant such as was the case in the
reported assessment of the shortenings of the Principal Tower C8 column (60.3 mm)
and W2 wall (10.2 mm);
• Following both static and dynamic testing of multiple buildings’ RC beams, the
authors of the relative publications state that, through the use of the OFDR DOFS
system, both static and dynamic concrete strains could be successfully measured;
• The above mentioned DOFS-powered tests further reported an effective capturing of
structural inflection points, moment transfer at the supports, crack locations and crack
openings;
• In all cases, DOFS were bonded to the surface of RC elements after sanding the latter
along the fiber pathway, cleaning it with water and 99% isopropyl alcohol.
Finally, on the monitoring of roads by means of DOFS:
• The strains transferred to a multilayered asphalt pavement can be monitored by means
of embedded DOFS. This fiber deployment’s main criticality is during the asphalt
compaction process which further involves high temperatures. To counteract any such
adversity specific high-resistance DOFS cables were employed in addition to manually
covering the DOFS with finer asphalt, before pouring the overlying layer;
• OFDR DOFS (OBR Luna ODiSI-B model) was employed to successfully assess the
distributed dynamic strains of a steel rail during the passage of low-speed and low-
vibration rail traffic but were unsuccessful for high-speed high-vibration ones;
• OFDR DOFS reported very similar reading qualities between a DOFS cable bonded to
a steel rail with minimal and optimal surface preparation (grinding and cleaning);
• A suggested DOFS deployment methodology was bonding DOFS to the rail foot and
to the railhead which, despite being prone to be affected by localized strains under
heavy train wheels, allows for more accurate rail curvature calculations thanks to the
greater distance between the two DOFS.
Sensors 2021, 21, 1818 51 of 83

For the readers interested in further reading on DOFS-based SHM of rails, the authors
refer to two recent comprehensive literature review articles, namely Du et al. [32] and Sasi
et al. [121]. Instead, for further reading on DOFS for Civil and Infrastructure Engineering
sensing the authors recommend Soga and Luo [15].

6. Geotechnical Applications
Geotechnical Engineering covers a wide range of applications amongst which pile
foundations, soil movements (landslides and soil subsidence), stabilization anchors, mining
and more. Each of these has the potential for a DOFS-powered SHM integration and their
relative literature review will receive independent scrutiny in the present section.
Sensors 2021, 21, x FOR PEER REVIEW 53 of 87
6.1. Pile Foundations
Starting from pile foundations, these are commonly used whenever vertical loads are
too high to be carried by shallow foundations (like in the case of high-rise buildings) and/or
and/or
when the when thebearing
soil’s soil’s bearing
capacitycapacity is not suitable
is not suitable to carrytothe carry the designed
designed loads. Now, loads.consid-
Now,
considering the rapid urbanization process occurring nowadays
ering the rapid urbanization process occurring nowadays all over the world, skyscrapers all over the world, sky-
scrapers and mega infrastructure are becoming commonplace
and mega infrastructure are becoming commonplace solutions to the novel requirements solutions to the novel re-
quirements of our modern-day
of our modern-day society.
society. As such, pileAs such, pile foundations
foundations have seen
have seen increased useincreased
in the latestuse
in the latest years, bringing along a stronger focus on the methodologies
years, bringing along a stronger focus on the methodologies for testing and monitoring for testing and
monitoring their performances.
their performances. In particular,Inlarge
particular,
researchlarge research
efforts efforts
have been have been
dedicated dedicated
to surpassing
to surpassing the limitations intrinsic to the traditional monitoring
the limitations intrinsic to the traditional monitoring tools (such as SG) commonly tools (such as SG) com-em-
monly employed for performing the required tests. Indeed, traditional
ployed for performing the required tests. Indeed, traditional sensors are easily destroyed sensors are easily
destroyed
in complexingeotechnical
complex geotechnical
environments environments
and are prone andto are prone
rust damageto rust
and damage and elec-
electromagnetic
tromagnetic interference
interference [122]. [122].though,
Predictably Predictably
their though, their largest
largest limitation limitation
is the punctualisnature
the punctual
of their
nature of their measurements.
measurements. This is especially This is especially
off-putting when off-putting
compared when compared
against the much against
broaderthe
much
globalbroader
structuralglobal structural
behavior whose behavior
knowledge whose is knowledge
required byisanrequired
engineer bytoan engineer
draw conclu- to
draw conclusions on the carrying capacity of the structure. Instead,
sions on the carrying capacity of the structure. Instead, as foreshadowed in Section 2, OFS as foreshadowed in
Section 2, OFS are characterized
are characterized by corrosion
by corrosion immunity, immunity,
high durability,highresistance
durability,toresistance to elec-
electromagnetic
tromagnetic
interference,interference, small size, and,
small size, lightweight lightweight and, most
most crucially, theycrucially, they easily
easily surpass surpass
the punctual
the punctual measurement
measurement limitation. Inlimitation.
more detail, In as
more detail,
clearly as clearly
illustrated inillustrated
Sienko et al. in [123]
Sienko foretthe
al.
[123]
examplefor the example
of pile of pile traditional
monitoring, monitoring, traditionaltools
monitoring monitoring
can carry tools
out can carry out meas-
measurements only
on a punctual
urements only scale (as visiblescale
on a punctual in Figure 49a).inWhilst
(as visible Figurethe situation
49a). Whilst canthe be improved
situation can bybe
installing several sensors along the monitored axis (quasi-distributed
improved by installing several sensors along the monitored axis (quasi-distributed sens- sensing illustrated
in Figure
ing 49b),in
illustrated the resulting
Figure 49b),profile still lacks
the resulting sufficient
profile accuracy
still lacks to properly
sufficient accuracy represent
to properly the
actual pilethe
represent strain profile.
actual pileDistributed sensing,
strain profile. instead, sensing,
Distributed providesinstead,
completeprovides
access tocomplete
this data
(Figureto49c).
access this data (Figure 49c).

Figure 49.
Figure Measurement
49.Measurement schemes
schemes for concrete
for concrete piles:piles: (a)(b)
(a) spot, spot, (b) quasi-continuous,
quasi-continuous, (c) dis-
(c) distributed
[123].
tributed [123].

The two kinds

kinds of
of pile
pilewhose
whosetesting
testingmethodologies
methodologieshavehavereceived
receiveda significant perfor-
a significant per-
mance boost
formance from
boost the integration
from of DOFS
the integration are the
of DOFS areContinuous FlightFlight
the Continuous AugerAuger
(CFA) (CFA)
or Auger
or
Cast Piling
Auger Cast and theand
Piling Rotary
the Bored
RotaryPiling.
Bored Research articles on
Piling. Research their on
articles testing
theirprocedures are
testing proce-
hereinare
dures reported.
herein reported.
Starting from the CFA, when constructing piles under groundwater level or soft
ground conditions, in order to prevent the collapse of the soil after drilling the borehole,
continuous down–up concreting is performed as the auger drill is extracted (the concrete
is pumped through a hollow shaft in the auger). The concrete hinders the penetration of
Sensors 2021, 21, 1818 52 of 83

Starting from the CFA, when constructing piles under groundwater level or soft
ground conditions, in order to prevent the collapse of the soil after drilling the borehole,
continuous down–up concreting is performed as the auger drill is extracted (the concrete
is pumped through a hollow shaft in the auger). The concrete hinders the penetration of
groundwater and soil in the borehole. A reinforcement steel cage is then inserted in the
filled borehole as soon as the concrete injection is terminated. Yet, as stated by Seo [7], the
use of the CFA is not entirely failure-proof. Indeed, the force distribution along a column
and its base’s load-bearing capacity might be affected by issues such as cross-sections
reductions or bulge formations (soft soils), soil mining problems (loose gravel54and
Sensors 2021, 21, x FOR PEER REVIEW sand
of 87
can be mixed with concrete materials at the bottom of the pile), hole stability and auger
rates control problems (ground with voids or water pockets) [123]. The assessment of the
column force distribution and base load-bearing capacity is, therefore, a key aspect of SHM
Seo [7] instrumented the steel cages of a 25 m deep CFA pile with both strain and
and was, therefore, the topic of numerous experimental campaigns.
temperature BOTDR DOFS (as visible in Figure 50) in addition to standard SG and exten-
Seo [7] instrumented the steel cages of a 25 m deep CFA pile with both strain and
someters. Interestingly, a certain prestress was applied to the strain DOFS in order to fa-
temperature BOTDR DOFS (as visible in Figure 50) in addition to standard SG and ex-
cilitate the recognition of its frequency data location during the data analysis (its central
tensometers. Interestingly, a certain prestress was applied to the strain DOFS in order to
frequency was, therefore, higher than the temperature’s cable). The compression load was
facilitate the recognition of its frequency data location during the data analysis (its central
initially increased to 8 MN, it was then unloaded and finally loaded again up to 20 MN.
frequency
Seo assumed was,that
therefore, higher
if the soil than the
had mixed temperature’s
with the concretecable). The compression
at the bottom of the CFAload
pile,was
initially
large strains would be reported in that location. However, the maximum compressiveSeo
increased to 8 MN, it was then unloaded and finally loaded again up to 20 MN.
assumed
strain was that if the
only 200 soil
με, had
thus mixed with the
confirming the proper
concrete at the bottom
functioning ofpile.
of the the CFA pile,Seo
Finally, large
strains
studiedwould be reported
the pile’s in that
skin friction at location.
its bottomHowever,
reportingthe maximumincrease
a significant compressive strain
with the ap-was
only
plied200 µε, thus confirming the proper functioning of the pile. Finally, Seo studied the
load.
pile’s skin friction at its bottom reporting a significant increase with the applied load.

Figure
Figure 50.50. Installation
Installation of the
of the DOFS
DOFS andand the pile
the pile [7]. [7].

A similar testing methodology was presented in Kania et al. [124] who monitored, by
means of OFDR DOFS, the strains and temperature inside different kinds of piles sub-
jected to static load testing. The authors provided a comprehensive look at the proper way
 Sensors 2021, 21, 1818 53 of 83

A similar testing methodology was presented in Kania et al. [124] who monitored, by
means of OFDR DOFS, the strains and temperature inside different kinds of piles subjected
to static load testing. The authors provided a comprehensive look at the proper way to
perform a DOFS structural instrumentation. Such work is commendable as it can provide
Sensors 2021, 21, x FOR PEER REVIEW 55 of 87
an initial quality boost to similar future tests and, as such, some key aspects are reported
in the following. First of all, if the pile may potentially be subjected to bending, Kania
et al. suggested the mounting of a pair of strain sensors in opposite locations in order
order to measure
to measure both itsboth its compressive
compressive and strains.
and tensile tensile Furthermore,
strains. Furthermore, they recom-to
they recommended
mended
always to always
install twoinstall
fiberstwo fibers
at each at eachfor
location location for both redundancy
both redundancy purposes purposes
and, in case and,of
indamage
case of or damage or unexpected
unexpected results, forresults,
havingfor having
access access to independent
to independent verificationverification
data. Then,
similarly
data. Then,tosimilarly
what was done was
to what in [7], theyinintroduced
done the idea ofthe
[7], they introduced a pre-tension of the DOFS
idea of a pre-tension of
cables to an amount larger than the expected compressive
the DOFS cables to an amount larger than the expected compressive strain in order to strain in order to align its
length
align its scale.
lengthFinally, once theonce
scale. Finally, sensing cables are
the sensing fixedare
cables to fixed
a structure, they proposed
to a structure, they pro- the
marking
posed the of the points
marking of interest
of the points of (e.g., pile head,
interest specific
(e.g., pile head,depths, and
specific pile toe)
depths, andbypile
pressing
toe)
byorpressing
heating/cooling the cable inthe
or heating/cooling thatcable
specific point/s
in that andpoint/s
specific recordingandthe locationthe
recording of the signal.
location
ofInthe
[124], the In
signal. piles in question
[124], the pileswere a steel pile,
in question werea aCFAsteelpile and
pile, a precast
a CFA pile which
pile and a precastwerepileall
instrumented with different fiber deployment methodologies.
which were all instrumented with different fiber deployment methodologies. For the For the bonding to the steel
pile, additional
bonding steel
to the steel rebars
pile, were welded
additional to itswere
steel rebars surface as a guide
welded and protection
to its surface as a guide forand
the
DOFS (see Figure 51a). The fibers were bonded to the CFA pile’s
protection for the DOFS (see Figure 51a). The fibers were bonded to the CFA pile’s rein- reinforcement cage with
epoxy glue
forcement (see
cage Figure
with epoxy51b,c)
glueand,
(seefor the precast
Figure 51b,c) and,concrete
for the pile, they concrete
precast were installed into
pile, they
previously
were installed incised grooves and
into previously then fixed
incised with
grooves andepoxy.
then fixed with epoxy.

Figure 51. Typical cross-section layout of DOFS cables attached to (a) a steel pile and to (b) a steel cage; photography of
Figure 51. Typical cross-section layout of DOFS cables attached to (a) a steel pile and to (b) a steel cage; photography of the
the latter is represented in (c) (adapted from [124]).
latter is represented in (c) (adapted from [124]).

The
Theauthors
authorsstudied
studiedandandreported
reportedthe
thesteel
steelpile’s
pile’sstrains
strainsinduced
inducedbybyits
itsdriving
drivingininthe
the
ground (Figure 52a), the CFA pile’s strain distribution during a test
ground (Figure 52a), the CFA pile’s strain distribution during a test which includedwhich included
twelve
twelve incremental
incremental load (Figure
load steps steps (Figure 52b),
52b), the the precast
precast concrete
concrete piles’ operational
piles’ operational strainsstrains
(which
(which were influenced
were influenced by the evolution
by the evolution of cracks
of several severalascracks asinvisible
visible Figurein52c),
Figure 52c), the
the short and
short and long term influence of temperature on the sampled strains and finally
long term influence of temperature on the sampled strains and finally the influence of the influ-
ence of lateral
lateral forces.forces.
Sienko et al. [123] also performed a 12 m long strain and temperature OFDR DOFS
monitoring on a CFA pile. The article reported the pile strains sampled during a multi-load
steps compression test in addition to its accumulated shortening. The strain profiles clearly
showed some local extremes which were attributed to a reduction of the column stiffness
(reportedly due to the reduction of the pile’s cross-sectional area or the concrete’s modulus
of elasticity).
Moving on to Rotary Bored Piling, differently than CFA, after drilling these require
additional ground support such as casing or drilling fluids but can reach much deeper
depths ~60 m (versus its CFA counterpart ~25 m) thus being able to accommodate higher
loads, bypass underground impediments and infiltrate the ground which could otherwise
be too hard for CFA drilling.

(a) (b) (c)

 The authors studied and reported the steel pile’s strains induced by its driving in the
ground (Figure 52a), the CFA pile’s strain distribution during a test which included
twelve incremental load steps (Figure 52b), the precast concrete piles’ operational strains
Sensors 2021, 21, 1818
(which were influenced by the evolution of several cracks as visible in Figure 52c), the
54 of 83
short and long term influence of temperature on the sampled strains and finally the influ-
ence of lateral forces.

(a) (b) (c)

Figure 52.
Figure 52. Distribution
Distribution of
of strains
strainsalong
alongindividual
individualDOFS
DOFS(a)
(a)after
afterthe
thedriving
drivingofofa asteel
steel pile
pile (b)(b) during
during a static
a static load
load testing
testing of
of the CFA pile and (c) the crack evolution inside a precast concrete pile (adapted from
the CFA pile and (c) the crack evolution inside a precast concrete pile (adapted from [124]). [124]).

Cheng et al. [125] reported the outcome of a Bi-Directional Static Load Test (BDSLT) on
a BOTDA DOFS instrumented bored pile with the aim of verifying the geotechnical design
parameters for the highly fractured limestone formation on which the pile’s residential
building project was located. BDSLT represents a deviation from the classic but reportedly
inferior (time, cost and risk wise) Maintained Load pile Tests (MLT) as it avoids the
latter’s massive reaction systems (i.e., dummy piles, platforms, reaction beams or concrete
blocks) in favor of jacks incorporated inside the pile shaft itself. Whilst the conventional
MLT systems apply the testing load from the top, the BDSLT’s bidirectional jacks expand
separating the pile’s body into the upper pile section and lower pile section (shafts) pushing
the former upwards (mobilizing the shaft friction) and the latter downwards (mobilizing
the shaft resistance and end bearing). The study case pile was 57.3 m long with a 1.35 m
diameter and the bidirectional jack embedded at 5.8 m from the pile toe. Its axial strains
were sampled with four single-core DOFS, reinforced by several strands of steel wires and
a polyethylene jacket, bonded to four sides of the reinforcement cage at 90-degrees from
each other. Figure 53a displays the continuous axial strain profile at several loading stages
of the working load (WL).
The authors reported witnessing a consistent DOFS strain increment as the load in-
creased along with unexpected spikes (between 16 m to 43 m depth) on all four independent
fibers likely due to the pile’s stiffness variation (similarly to [123]). Despite acknowledging
the strong possibility of the bored pile’s shape non-uniformity and concrete stiffness in-
consistency, due to the inaccessibility of the stiffness data, the authors assumed it constant
for the calculation of its axial forces (Figure 53b). The sampled DOFS strains successfully
diagnosed an unexpected and potentially disruptive low shaft friction, hypothized to be
caused by unaccounted features of the karstic limestone and of the inclined bedrock.
Xu et al. [126] and Pei et al. [21] presented a new bored piles BOTDA DOFS installation
methodology and reported the outcome of its application to three bored piles (20.9 m
embedded depth and 3.0 m diameter) pertaining to a retaining wall reinforcing a soil slope
subjected to multi-stage excavations. Similar to [124], in order to obtain both the axial
and bending strains along an instrumented pile, two sets of DOFS sensors were installed
symmetrically with respect to their axis (as in Figure 54).

Sensors 2021, 21, 1818
expand separating the pile’s body into the upper pile section and lower pile section
(shafts) pushing the former upwards (mobilizing the shaft friction) and the latter down-
wards (mobilizing the shaft resistance and end bearing). The study case pile was 57.3 m
long with a 1.35 m diameter and the bidirectional jack embedded at 5.8 m from the pile
toe. Its axial strains were sampled with four single-core DOFS, reinforced by several
strands of steel wires and a polyethylene jacket, bonded to four sides of the reinforcement 55 of 83
cage at 90-degrees from each other. Figure 53a displays the continuous axial strain profile
at several loading stages of the working load (WL).

Sensors 2021, 21, x FOR PEER REVIEW 57 of 87

Figure 53. (a) Changes in strains along the pile during Bi-Directional Static Load Test (BDSLT) and (b) the calculated load
distribution along it (adapted from [125]).

The authors reported witnessing a consistent DOFS strain increment as the load in-
creased along with unexpected spikes (between 16 m to 43 m depth) on all four independ-
ent fibers likely due to the pile’s stiffness variation (similarly to [123]). Despite acknowl-
edging the strong possibility of the bored pile’s shape non-uniformity and concrete stiff-
ness inconsistency, due to the inaccessibility of the stiffness data, the authors assumed it
constant for the calculation of its axial forces (Figure 53b). The sampled DOFS strains suc-
cessfully diagnosed an unexpected and potentially disruptive low shaft friction, hy-
pothized to be caused by unaccounted features of the karstic limestone and of the inclined
bedrock.
Xu et al. [126] and Pei et al. [21] presented a new bored piles BOTDA DOFS installa-
tion methodology and reported the outcome of its application to three bored piles (20.9 m
embedded depth and 3.0 m diameter) pertaining to a retaining wall reinforcing a soil slope
(a)
subjected to multi-stage excavations. Similar to [124], in (b)order to obtain both the axial and
bending strains along an instrumented pile, two sets of DOFS sensors were installed sym-
Figure 53. (a) Changes in strains along the pile during Bi-Directional Static Load Test (BDSLT) and (b) the calculated load
metrically with respect to their axis (as in Figure 54).
distribution along it (adapted from [125]).

Figure Fiber
54. 54.
Figure optic
Fiber opticdeployment layout
deployment layout in the
in the study
study case pile
case bored bored pile
[126]. (a)[126]. (a) Total
Total view;(b) Cross section Ⅰ-Ⅰ;
view;(b) Cross section I-I; (c)
(c) Enlarged
view.view.
Enlarged

The maximum lateral wall deflection was assessed to be 0.05–0.1% and the measure-
ments obtained from the BOTDA DOFS were found to be in good agreement with the one
coming from parallelly installed inclinometers.
Sensors 2021, 21, 1818 56 of 83

The maximum lateral wall deflection was assessed to be 0.05–0.1% and the measure-
ments obtained from the BOTDA DOFS were found to be in good agreement with the one
coming from parallelly installed inclinometers.
Guo and Zhao [127] performed a BOFDA DOFS monitoring of two 39 m long and
1.5 m diameter bored piles buried in a sloping terrain thus also subjected to both vertical
and horizontal loading. The DOFS was bonded to the pile’s steel cage to form a U-loop
monitoring its two sides and its bottom. Once again, strain profile peaks were present and
were attributed to inhomogeneities of the pile’s material. The pile’s axial force calculated
per DOFS was compared with the ones extracted by an embedded steel stress meter with
good agreement between them.
Finally, Gao et al. [18] proceeded in a very similar way instrumenting a bored pile with
U-looping single-mode GFRP-covered BOTDR DOFS fixed to its steel cage. Additionally,
the results were filtered using three different noise removal methods.

6.2. Soil and Rock Deformations

Ground deformations are key study points for Geotechnical Engineering (land subsi-
dence, ground fissure and surface collapse amongst others) as serious geological disasters
may occur on their back. Their monitoring and the development of related early warn-
ing systems are crucial for the prevention and control of such disasters. Conventional
technologies such as tiltmeters, inclinometers and multipoint extensometers have been
adopted for this purpose for decades but the high costs and logistical issues associated
with the creation of automated systems make them impractical for long term and massive
deployments [128]. In response, the industry is slowly shifting to alternative and more
performant ground deformation monitoring technologies such as DOFS.
The first scrutinized geotechnical issue will be land subsidence, explicitly the gentle
down warping and/or sudden sinking of discrete segments of the ground surface [129].
One of its main causes is the intense extraction and exploitation of groundwater (but also
of crude oil and natural gas) which leads to the compaction and collapse of underground
aquifer systems. Soil subsidence has been the topic of an intense DOFS-powered investiga-
tion effort spanning multiple articles concisely presented in the following paragraphs.
Liu et al. [128] performed a long-term BOTDA DOFS monitoring of subsidence-
induced ground fissure deformations of a field in Wuxi City, China. A special high-strength
optical fiber was used with a measurable strain limit of 15,000 µε. The DOFS was placed in
a 32 m long, 75 cm deep and 45 cm wide trench excavated at a 45 degree angle compared
to the fissure direction. In order to compensate for the effect of seasonal temperature
variation, a temperature-sensing optical fiber was also inserted in the DOFS-containing
tube (placed for protection purposes). In the span of almost 5 years, nine measurements
were carried out indicating a ground subsidence-induced total cable stretch of 11.3 mm.
The latter was mainly concentrated in two areas (Figure 55a,b) where the largest strain
evolutions were reported.
Wu et al. [130] applied DOFS technology to a laboratory experimental study on land
subsidence caused by seasonal water level fluctuations and dewatering. In particular, the
specimen was a small-scale sand–clay interbedded model box built for the purpose of
carrying out consolidation and rebound tests during two drainage–recharge cycles. In
order to analyze the deformative response of each of the specimen’s layers to the water
level variations, the authors placed inside it two different sensing cables (a special tight-
buffered PPP-BOTDA single-mode DOFS and a carbon fiber tube with FBGs) with small
Plexiglas disks fixed perpendicularly to the cables in order to increase the soil/fiber bond.
The PPP-BOTDA DOFS-extracted results indicated that the layers were compressed during
drainage and rebounded during recharge, with larger deformations in the clay layer than in
the sand. Furthermore, the deformation of the sand layer was synchronous with the water
content variations, whilst the clay’s lagged behind due to its lower permeability coefficient.
Sensors 2021, 21, 1818 57 of 83
Sensors 2021, 21, x FOR PEER REVIEW 59 of 87

(a) (b)
Figure55.
Figure 55.(a)
(a)Monitoring
Monitoring area plan
plan and
and(b)
(b)resulting
resultingstrains
strains(adapted from
(adapted [128]).
from [128]).

Wu same
The et al. [130] applied
authors DOFSin
published technology
Zhang et toal.a[131]
laboratory experimental
the results study
of an in-situ on land
experimental
Sensors 2021, 21, x FOR PEER REVIEW 60 of 87
subsidence
campaign caused by
(Suzhou, seasonal
China) water
aimed level fluctuations
at acquiring data on theandsubsidence
dewatering.andIn particular, the
strata subsurface
specimen
vertical was a small-scale
deformations sand–clay
by means interbedded
of a BOTDR DOFS model box built
embedded for the purpose
in a borehole of car-56).
(as in Figure
rying out consolidation and rebound tests during two drainage–recharge cycles. In order
to analyze the deformative response of each of the specimen’s layers to the water level
variations, the authors placed inside it two different sensing cables (a special tight-buff-
ered PPP-BOTDA single-mode DOFS and a carbon fiber tube with FBGs) with small Plex-
iglas disks fixed perpendicularly to the cables in order to increase the soil/fiber bond. The
PPP-BOTDA DOFS-extracted results indicated that the layers were compressed during
drainage and rebounded during recharge, with larger deformations in the clay layer than
in the sand. Furthermore, the deformation of the sand layer was synchronous with the
water content variations, whilst the clay’s lagged behind due to its lower permeability
coefficient.
The same authors published in Zhang et al. [131] the results of an in-situ experimental
campaign (Suzhou, China) aimed at acquiring data on the subsidence and strata subsur-
face vertical deformations by means of a BOTDR DOFS embedded in a borehole (as in
Figure 56).

Figure
Figure 56.
56. Illustration
Illustration of
of soil
soil subsidence
subsidence and DOFS vertical
and DOFS vertical deployment
deployment in
in aa borehole
borehole as
as performed
per-
formed
in [131].in [131].

After the 200 m deep borehole was excavated, the DOFS cable was attached to a cone
guide and lowered into the former. The hole was then backfilled
backfilled with
with a mixture of sand,
clay and gravel. Then,
Then, by
by means
means of direct integration, the authors converted the DOFS
strains sampled between December 2012 and November 2014 into a numerical estimation
of the compaction or rebound that occurred in that time. The results highlighted the rise
of positive strains between 60.3 m and 67.9 m depth indicative of an overall rebound of
the pumped aquifer.
Liu et al. [22] performed a very similar monitoring campaign in a 130 m deep bore-
Sensors 2021, 21, 1818 58 of 83

of the compaction or rebound that occurred in that time. The results highlighted the rise of
positive strains between 60.3 m and 67.9 m depth indicative of an overall rebound of the
pumped aquifer.
Liu et al. [22] performed a very similar monitoring campaign in a 130 m deep borehole
in the Yangtze River Delta region (China) using BOFDA DOFS technology. The authors
specify having backfilled the drilled borehole in accordance with the surrounding stratum
type for consistency purposes. The monitoring, which lasted 2 years, successfully detected
the occurrence of compression-rebound cycles simultaneous to the draining–recharging
one. Additionally, being the soil rebound smaller than that of compression it was assessed
that irreversible deformations had occurred. Similar to all the previous articles, the authors
concluded suggesting that DOFS provides clear advantages for the analysis and early
warning of land subsidence mechanisms.
According to Cheng et al. [17], an additional purpose that DOFS could cover is the
monitoring of drying-induced shrinkage and desiccation cracking in clayey soils. In order
to investigate such possibility, the authors parallelly placed three differently coated optical
fibers (thermoplastic polyester elastomer jacket, nylon jacket and acrylate coating) in a mold
before pouring the soil slurry. The authors reported BOTDR DOFS-sampled strains in the
order of several micro strains when the soil was over-saturated, their gradual increase with
the drop of water content (and the increase in the soil/fiber interfacial shear stresses) and
finally their drop after the development of desiccation cracks (reportedly due to fiber/soil
decoupling). The observations on the performance of each fiber and their strong impact
Sensors 2021, 21, x FOR PEER REVIEWon strain measurement are herein reported. The optical fiber with acrylate coating was
61 of 87
deemed not suitable for the test due to its fragility and the poor interfacial coupling with
the soil. Oppositely, the optical fibers covered with TPEE jackets and nylon were deemed
suitable with the former being more sensitive to water content variations.
Soil shear deformations are a supplementary geotechnical parameter to closely mon-
Soil shear deformations are a supplementary geotechnical parameter to closely mon-
itor for stability assessments and for early warnings of potential geo-hazards [132]. The
itor for stability assessments and for early warnings of potential geo-hazards [132]. The
following articles present DOFS-powered SHM campaigns developed in this framework.
following articles present DOFS-powered SHM campaigns developed in this framework.
Wang et al. [133] reported the results of a BOTDA DOFS monitoring of the 408 km
Wang et al. [133] reported the results of a BOTDA DOFS monitoring of the 408 km long
long Trans-Andean Pipeline (Peru). Given that soil shear movements could lead to move-
Trans-Andean Pipeline (Peru). Given that soil shear movements could lead to movements,
ments, deformations and even fracture of the pipeline, an IEC 794-1 standard DOFS cable
deformations and even fracture of the pipeline, an IEC 794-1 standard DOFS cable (capable
(capable of monitoring movements as small as 5 cm) was positioned in the former’s trench
of monitoring movements as small as 5 cm) was positioned in the former’s trench in order
in order to detect the appearance of any such phenomena. Early warnings would trigger
to detect the appearance of any such phenomena. Early warnings would trigger a patrol’s
adamage
patrol’sassessment
damage assessment and thestrengthening
and the prompt prompt strengthening of thewhether
of the pipeline pipelinenecessary.
whether nec-
The
essary. The system reportedly warned of the occurrence of rockfall and a
system reportedly warned of the occurrence of rockfall and a water seepage-induced water seepage-
induced
collapse collapse of the pipeline
of the pipeline trench
trench (as (asinvisible
visible Figurein57).
Figure 57).

Figure
Figure57.
57.Monitoring
Monitoringchart
chartand
andsite
sitemap
mapof
ofpipeline
pipelinetrench
trench collapse
collapse accident
accident [133].
[133].

Zheng
Zhenget etal.
al.[134]
[134]attempted
attemptedto todetermine
determinethe
theperformance
performanceof ofaaBOTDR
BOTDRDOFS DOFSmoni-
moni-
toring
toringsystem
systemforforthe
thedetection
detectionof ofdeformations
deformationsinduced
inducedby byshear
shearsliding
slidingononrock
rockslopes.
slopes.
An
Aninitial
initialdirect-shear
direct-shearlaboratory
laboratoryexperiment
experimentwaswasperformed
performedin inorder
ordertotoinvestigate
investigateDOFS’
DOFS’
monitoring effectiveness in
monitoring effectiveness inthis
thisenvironment,
environment, finding
finding a high
a high correlation
correlation between
between pre-
predicted
dicted and tested
and tested resultsresults (average
(average ratio between
ratio between 0.974 0.974 to 1.081).
to 1.081). Following
Following such,such, the setup
the setup was
was applied in an in-situ shallow artificial slope monitoring test whose results were, once
again, reportedly similar to the predicted ones. According to the authors, these results
suggested OFSs applicability and accuracy for slope deformation.
On the same topic, Wu et al. [132] raised the question of whether the slip between
Sensors 2021, 21, 1818 59 of 83

applied in an in-situ shallow artificial slope monitoring test whose results were, once again,
reportedly similar to the predicted ones. According to the authors, these results suggested
OFSs applicability and accuracy for slope deformation.
On the same topic, Wu et al. [132] raised the question of whether the slip between
OFDR DOFS cables and soil, that could potentially occur during soil shear deformation
applications, would yield strain measurement errors. To answer the matter, the authors
performed a laboratory direct shear test with different DOFS/soil anchorage method-
ologies. These included surface-mounted heat-shrinking tubes (V1 in Figure 58) whose
post-shrinking local corrugations supposedly enhanced the cable-soil interfacial bonding,
aluminum block anchors (V2) and no anchoring (V3) for reference purposes.
The test results showed that for small soil shear displacement (9–15 mm), good
soil displacement estimations could be extracted even with no additional anchorage (V3).
Beyond such displacements, though, the non-anchored cables slipped from the surrounding
soil (leading to smaller detected strains) whilst the block and tube anchored ones still
maintained a well-coupled state. The authors concluded by suggesting the use of block
Sensors 2021, 21, x FOR PEER REVIEW 62 of 87
anchors for shallowly-embedded field applications and tube anchors for high overburden
pressure scenarios.

Figure 58. Different kinds of anchors used in the test from top to bottom: V1 tube-anchored cable,
Figure 58. Different kinds of anchors used in the test from top to bottom: V1 tube-anchored cable,
V2 aluminum
V2 aluminum block
block anchored
anchored cable, V3smooth
cable,V3 smoothcable
cable[132].
[132].

Shear deformations were also studied in Schenato et al. [135] in the framework of a
The test results showed that for small soil shear displacement (9–15 mm), good soil
laboratory large-scale physical landslide model monitored with an OFDR DOFS deployed
displacement estimations could be extracted even with no additional anchorage (V3). Be-
in the latter’s body. The shallow landslide was triggered by artificial rainfall which was
yond such displacements, though, the non-anchored cables slipped from the surrounding
sustained until complete slope failure. The authors’ most important takeaway was the
soil (leading to smaller detected strains) whilst the block and tube anchored ones still
possibility of observing through DOFS precursory failure signs well before the slope
maintained a well-coupled state. The authors concluded by suggesting the use of block
collapse thus being able to gain insights into its failure dynamic.
anchors for shallowly-embedded field applications and tube anchors for high overburden
Zhang et al. [136], Chen et al. [137] and Xu et al. [138] remarked that adverse phenom-
pressure scenarios.
ena decreasing the carrying capacity of a railway subgrade soil (soil cave or karst sinkhole
Shearcould
collapse) deformations were also
severely impact the studied in Schenato
latter’s viability andettrain
al. [135] in the framework
derailment risk. Thus, of
thea
laboratory large-scale physical landslide model monitored with an OFDR
importance of a train track subgrade SHM. The articles report a common BOTDR DOFS DOFS deployed
in the latter’s
subgrade body. The
settlement shallowtest
monitoring landslide
aimed atwas triggered by
determining artificial
their rainfall
viability which
for such kindwas
of
sustained until complete slope failure. The authors’ most important takeaway
monitoring. In this test, a 4 m long DOFS was embedded into a soil subgrade crossing the was the
possibility
roof of a 100of cm
observing through DOFS
wide pre-made precursory
soil cave. The fourfailure
stagessigns well of
collapse before the slope
the latter col-
led to a
lapse thus being able to gain insights into its failure dynamic.
147 cm wide superficial collapse measured by means of both DOFS and micrometer. Their
Zhang etmaximum
comparative al. [136], relative
Chen eterror
al. [137]
was and XuThe
5.28%. et al. [138] remarked
common conclusion that adverse
was phe-
that DOFS
nomena decreasing the carrying capacity of a railway subgrade
can effectively monitor a sinkhole collapse with relatively high precision. soil (soil cave or karst
sinkhole collapse) could severely impact the latter’s viability and train derailment risk.
Thus, the importance of a train track subgrade SHM. The articles report a common
BOTDR DOFS subgrade settlement monitoring test aimed at determining their viability
for such kind of monitoring. In this test, a 4 m long DOFS was embedded into a soil sub-
grade crossing the roof of a 100 cm wide pre-made soil cave. The four stages collapse of
the latter led to a 147 cm wide superficial collapse measured by means of both DOFS and
micrometer. Their comparative maximum relative error was 5.28%. The common conclu-
Sensors 2021, 21, 1818 60 of 83

6.3. Soil Stabilization Anchors

After discussing soil deformations and landslide DOFS monitoring, the logical next
step is the monitoring of stabilization anchors. Starting from Cola et al. [139], the authors
set themselves to monitor the strains of an OFDR DOFS embedded with the tendons inside
a composite anchor. These are special hollow steel bars coupled with tendons cemented
in the inner hole to increase the global anchor tensile strength for the remediation of
unstable slopes. The installation was reportedly performed in two steps. The first saw
the installation of the composite anchors using a self-drilling technique whilst the second
saw the insertion of one or more harmonic steel tendons instrumented with OFDR DOFS
positioned in the bar’s inner cavity as in Figure 59.
Once the instrumented anchors were installed the system permitted a millimetric
Sensors 2021, 21, x FOR PEER REVIEW 63 of 87 actions
evaluation of the axial force distribution in the anchor, of the soil–anchor interface
and of the stabilizing capability associated with the specific hydrogeological conditions
of the site. According to the authors, the extracted strain profiles had similar profiles. All
had their respective maximum strains sampled just behind the external plate and slowly
decreasing strain values when moving inside the ground. The strains decreased to almost
Sensors 2021, 21, x FOR PEER REVIEW 63 of 87
zero at about 16 to 17 m from the slope due to the activation of distributed shear stress
along the soil–bar interface.

Figure 59. DOFS installation by means of plastic hooks for coupling the DOFS in an anchor’s (a) extremity bar and (b)
internal bar [139].

Once the instrumented anchors were installed the system permitted a millimetric
evaluation of the axial force distribution in the anchor, of the soil–anchor interface actions
and of the stabilizing capability associated with the specific hydrogeological conditions of
the site. According to the authors, the extracted strain profiles had similar profiles. All had
Figure theirmeans
respectiveplastic
maximum strains sampled just behindanthe external(a)
plate and slowly de- (b)
Figure 59. DOFS installationby
59. DOFS installation by meansof of plastichooks
hooksfor
forcoupling
couplingthetheDOFS
DOFSinin an anchor’s
anchor’s (a)extremity
extremitybarbarand
and (b)
internal creasing strain values when moving inside the ground. The strains decreased to almost
internal bar
bar [139].
[139]. zero at about 16 to 17 m from the slope due to the activation of distributed shear stress
along the soil–bar interface.
Once the et
Lienhart instrumented anchorsinwere
al. [140] proceeded installed
a similar way tothemonitor
system multiple
permitted a millimetric
Single Borehole
Lienhart et al. [140] proceeded in a similar way to monitor multiple Single Borehole
evaluation
Multiple of the axial force distribution in the anchor, of the soil–anchor interface actions
Multiple strand Anchors instrumented with two OFDR DOFS and subjected to a pull-out pull-out
strand Anchors instrumented with two OFDR DOFS and subjected to a
and
test.of
test. thefibers
The
The stabilizing
fibers were
were capability
positioned
positioned associated
along
along thewith the
tendons
the tendons of specific
each hydrogeological
of individual
each individual alongconditions
anchor anchor of
along with
with
the
twosite. According
fiber loops in to
the the authors,
grout the
material extracted
(see strain
Figure
two fiber loops in the grout material (see Figure 60). 60). profiles had similar profiles. All had
their respective maximum strains sampled just behind the external plate and slowly de-
creasing strain values when moving inside the ground. The strains decreased to almost
zero at about 16 to 17 m from the slope due to the activation of distributed shear stress
along the soil–bar interface.
Lienhart et al. [140] proceeded in a similar way to monitor multiple Single Borehole
Multiple strand Anchors instrumented with two OFDR DOFS and subjected to a pull-out
test. The fibers were positioned along the tendons of each individual anchor along with
two fiber loops in the grout material (see Figure 60).

Figure 60.
Figure 60.Schematic
Schematic representation of single
representation borehole
of single with multiple
borehole anchors anchors
with multiple and with and
two DOFS
with two DOFS
deployed along
deployed alongthe tendons
the of each
tendons individual
of each anchor.
individual Two fiber
anchor. Two loops were
fiber alsowere
loops deployed
also in the
deployed in the
grout material [140].
grout material [140].
The load tests included a stepwise load increase. Initially, the authors reported al-
most uniform strains along the free length of the anchor and decreasing strains in the
tendons with increasing depth (due to the load transfer from the anchor to the soil). As
the load increased, the load transfer area widened suggesting a progressive failure of the
Sensors 2021, 21, 1818 61 of 83

The load tests included a stepwise load increase. Initially, the authors reported
almost uniform strains along the free length of the anchor and decreasing strains in the
tendons with increasing depth (due to the load transfer from the anchor to the soil). As
the load increased, the load transfer area widened suggesting a progressive failure of the
tendon/grout interface. The same article also reported a DOFS monitoring of the strain
variation along the top and bottom of soil nails placed into a slope in a road construction
site (as in Figure 61).
On the grounds of the distance between the top and bottom fibers, curvature changes
Sensors 2021, 21, x FOR PEER REVIEW 64 of 87
were derived from their strain differences and a 7 mm vertical displacement at the front
end of the nail was observed after roughly one month.

(a) (b)
Figure
Figure61.
61.(a)
(a)Three
Threefiber
fiberoptic-instrumented
optic-instrumented soil
soil nails
nails at
at aa construction
construction site
site and
and (b)
(b) their
their cable
cable layout
layout [140].
[140].

6.4. Mining
On the grounds of the distance between the top and bottom fibers, curvature changes
were The last geotechnical
derived sub-topic
from their strain under scrutiny
differences and a 7 ismm DOFS monitoring
vertical for mining
displacement at theapplica-
front
tions.
end of In
theZhang et al.
nail was [24] a similarity
observed modelone
after roughly wasmonth.
designed to simulate the rock overburden
movement during the mining of a coal seam in Huainan, China. The original 800 m min-
ingMining
6.4. depth and its 2.5 m thick panel were simulated by means of a similarity model test
(geometrically, kinematically
The last geotechnical and dynamically
sub-topic under scrutinysimilar to themonitoring
is DOFS real-life case)
for carried
mining out on
appli-
a 4000 × 400 × 1607 mm plane stress model shelf instrumented with
cations. In Zhang et al. [24] a similarity model was designed to simulate the rock overbur-four DOFS cables
connected
den movement to a PPP-BOTDA
during the mininganalyzer.
of a DOFS successfully
coal seam in Huainan,detected
China.peak
Thetensile strains
original 800 min
the soft strata and the interface of soft and hard strata.
mining depth and its 2.5 m thick panel were simulated by means of a similarity model test
A similar test
(geometrically, setup wasand
kinematically useddynamically
in Chai et al. [25] intoorder
similar to establish
the real-life case)ground
carried surface
out on
movement law for deep and extra-thick coal seam mining
a 4000 × 400 × 1607 mm plane stress model shelf instrumented with four DOFS under the ultra-thick conglom-
cables
erate. Similar to the previous article, the authors reported that the PPP-BOTDA DOFS
connected to a PPP-BOTDA analyzer. DOFS successfully detected peak tensile strains in
technology could successfully and effectively capture the surface movements and deforma-
the soft strata and the interface of soft and hard strata.
tions caused by mining.
A similar test setup was used in Chai et al. [25] in order to establish ground surface
Zhang and Sun [141] suggested that coal mining caused the movement of the overlying
movement law for deep and extra-thick coal seam mining under the ultra-thick conglom-
strata, thus leading to advance abutment pressure and consequently to the deformation
erate. Similar to the previous article, the authors reported that the PPP-BOTDA DOFS
and damage of the floor rock strata. As such, the authors developed a two-dimensional
technology could successfully and effectively capture the surface movements and defor-
geological model of the advanced abutment pressure on the grounds of BOTDR DOFS
mations caused by mining.
strain readings performed in a borehole situated in Inner Mongolia. Such a model was
Zhang and Sun [141] suggested that coal mining caused the movement of the over-
then successfully validated by means of another borehole monitoring performed in the
lying strata, thus leading to advance abutment pressure and consequently to the defor-
same mining area.
mation and damage of the floor rock strata. As such, the authors developed a two-dimen-
Finally, Lanciano and Salvini [142] presented an innovative combination of DOFS (as
sional
visiblegeological
in Figure model of the
62), digital advanced abutment
photogrammetry throughpressure on the Aerial
Unmanned grounds of BOTDR
Vehicle, topo-
DOFS
graphic and geotechnical monitoring systems for the monitoring of the rock massastability
strain readings performed in a borehole situated in Inner Mongolia. Such model
was then successfully
conditions in marble validated
quarries inbyorder
means to of anotherthe
eliminate borehole monitoring
risks of rock falls. performed
By means in of
the same mining area.
a BOTDA DOFS monitoring system, some displacements were successfully detected de-
spiteFinally, Lanciano
their values fell and Salvini
within [142] presented
the tolerance an innovative
threshold range. Nocombination
other critical of situations
DOFS (as
visible in Figure
were detected. 62), digital photogrammetry through Unmanned Aerial Vehicle, topo-
graphic and geotechnical monitoring systems for the monitoring of the rock mass stability
conditions in marble quarries in order to eliminate the risks of rock falls. By means of a
BOTDA DOFS monitoring system, some displacements were successfully detected de-
spite their values fell within the tolerance threshold range. No other critical situations
were detected.
Sensors 2021,
Sensors 21, x1818
2021, 21, FOR PEER REVIEW 6562ofof87
83

Figure
Figure62.
62.Geotechnical
Geotechnical sensors (in(in
sensors blue) andand
blue) fiber optics
fiber cable
optics (in red)
cable on the
(in red) onwestern (a) and
the western (a) and
southern-eastern (b) sides of the buttress [142].
southern-eastern (b) sides of the buttress [142].

6.5.Sub-Section
6.5. Sub-SectionConclusions
Conclusions
The present
The present sub-section
sub-sectionpresented
presentedvarious
variousDOFS
DOFS SHMSHM applications
applicationsfor for
the the
assessment
assess-
of theofperformance
ment the performanceof foundation
of foundationpiles, piles,
for theformonitoring of soilofmovements
the monitoring soil movements(i.e., subsi-
(i.e.,
dence, drying-induced shrinkage, soil shear deformation, landslides),
subsidence, drying-induced shrinkage, soil shear deformation, landslides), for the moni- for the monitoring
of soilof
toring anchors and mining-induced
soil anchors and mining-induced deformations. On the
deformations. On assessment of theof
the assessment performance
the perfor-
of foundation piles:
mance of foundation piles:
•• The Themost
mostcommonly
commonlyemployed
employed deployment
deployment technique
technique waswas bonding
bonding DOFS
DOFS to to aa pile’s
pile’s
steel cage before embedding the latter inside concrete. DOFS deployment
steel cage before embedding the latter inside concrete. DOFS deployment techniques techniques
were also
were also suggested
suggested for
for steel
steel pipes
pipes (bonding
(bonding thethe fiber
fiber on
on its
itssurface
surfacewith
withthe
theaddition
addition
of steel rebars welded to its surface as a guide and protection for
of steel rebars welded to its surface as a guide and protection for the DOFS) and the DOFS) and on on
pre-cast piles (DOFS were installed into previously incised grooves
pre-cast piles (DOFS were installed into previously incised grooves and then fixed and then fixed
with epoxy);
with epoxy);
•• Oppositely mounted strain
Oppositely mounted strain sensors
sensors (or
(or aasingle
singleU-looping
U-loopingone) one)were
wereused
usedto tomeasure
measure
both compressive and tensile strains in those situations where
both compressive and tensile strains in those situations where flexural strains areflexural strains are
expected (piles pertaining to retaining walls for
expected (piles pertaining to retaining walls for example); example);
•• The Theinstallation
installationofoftwo
twofibers
fibersat ateach
eachlocation
locationwaswasrecommended
recommendedfor forboth
bothredundancy
redundancy
purposes and, in case of damage or unexpected results, in order to have access to
purposes and, in case of damage or unexpected results, in order to have access to
independent verification data;
independent verification data;
• In more than one article, DOFS reportedly provided crucial data on the pile stiffness
• In more than one article, DOFS reportedly provided crucial data on the pile stiffness
reductions possibly due to the decrease of its cross-sectional area and/or loose gravel
reductions possibly due to the decrease of its cross-sectional area and/or loose gravel
and sand mixing with the concrete.
and sand mixing with the concrete.
On the monitoring of soil movements:
On the monitoring of soil movements:
• For the monitoring of soil subsidence, multiple publications positioned DOFS inside
• For the monitoring of soil subsidence, multiple publications positioned DOFS inside
previously excavated boreholes before backfilling them (preferably in accordance
previously excavated boreholes before backfilling them (preferably in accordance
with the surrounding stratum type) and assessing through the extracted profiles the
with the surrounding stratum type) and assessing through the extracted profiles the
occurrence of soil rebound and compaction;
occurrence of soil rebound and compaction;
• Superficial soil deformations, induced for example by soil subsidence, by a landslide
or by a rockfall, were successfully monitored positioning a high-strength DOFS inside
a superficially excavated trench.
Sensors 2021, 21, 1818 63 of 83

• On the DOFS soil-embedment, low performance was detected for acrylate coated
DOFS due to their fragility and the poor interfacial coupling with the soil.
• On the use of DOFS for the monitoring of large shear displacements, the suggested
procedure was the bonding of DOFS to the soil/rock with heat-shrinking tubes or
aluminum block anchors.
On the monitoring of soil stabilization anchors and nails, the most commonly used
method was the coupling of one or more DOFS to their tendons. On one occasion, two
fiber loops were additionally deployed in the grout material surrounding the anchors.
Overall the reported articles testified to a good DOFS strain monitoring performance and
the extraction of crucial data on the anchor load-carrying capacity.
Finally, PPP-BOTDA and BOTDR DOFS were successful in the monitoring of move-
ments in the overlying strata and of the advanced abutment pressure during coal seam
mining. DOFS was also successfully employed for the monitoring of the rock mass stability
conditions in a marble quarry.
For the readers interested in further reading, the authors refer them to the following
comprehensive literature review articles on geotechnical [143], geophysical [13] and geo-
hydrological [14] DOFS-based SHM.

7. Tunnels
There is no doubt that Geotechnical Engineering and Tunnel Engineering go hand
in hand. This statement can be even stretched as far as asserting that they are one and
the same especially considering that proper geotechnical monitoring systems are crucial
for safe underground construction and operations. Nevertheless, keeping in mind the
difference in boundary conditions and vast application diversity, for simplicity’s sake the
present article distinguishes their DOFS monitoring in two different sections, the previous
and the present one.
The ever-growing urbanization throughout the world has been the catalyst for conges-
tion, air pollution and surface scarceness issues. The accompanying awareness that this
situation could be alleviated by means of an underground placement of some key facilities
(rapid transit, parking, utilities, storage, etc.) has led to an increased tunneling activity.
The amount of tunnels is not the only growth factor, though. Their design complexity
is rising too, encompassing longer and wider tunnels, excavations through increasingly
difficult ground conditions, under densely populated zones (thus requiring very high
safety standards) and in highly seismic areas [144].
The need for a performant tunnel SHM is particularly self-explanatory, for example,
in the aftermath of a seismic event. especially when attempting to assess its residual
capacity to survive consecutive seismic aftershocks. This withstanding, SHM is equally
crucial during a tunnel’s construction and service-life. As a matter of fact, both new and
existing tunnels suffer an inherent structural risk connected to the surrounding geotechnical
and hydrological conditions (new ground level constructions and excessive rain periods
amongst others) as well as to unknowns related to design assumptions and construction
materials [145]. These risks may translate into the appearance of abnormal horizontal
and vertical tunnel deformations, convergence deformations (relative displacement of two
diametrically-opposed points), cracks, joint movements, water leaks and, in the worse
cases, collapse. The SHM of tunnels and underground structures is not only a key aspect
for their correct functioning and safety of use but also for the security and integrity of
its surroundings. Indeed, the latter can potentially experience geotechnical condition
alterations due to tunneling (especially in the case of shallow tunneling) such as settlement
troughs and holes at ground level, groundwater level disruption, urbanizations’ drainage
system jeopardizing and more [146].
Whilst the common practice for the detection of the above-mentioned tunnel structural
issues (or of their precursor signs) are regular visual inspections, these have very logical
performance limitations and often require to stop the tunnel traffic, thus strongly limiting
their frequency. Alternatively, geodetic target points can be mounted to a tunnel linings’
Sensors 2021, 21, x FOR PEER REVIEW 67 of 87

Whilst the common practice for the detection of the above-mentioned tunnel struc-
Sensors 2021, 21, 1818 tural issues (or of their precursor signs) are regular visual inspections, these have64very of 83
logical performance limitations and often require to stop the tunnel traffic, thus strongly
limiting their frequency. Alternatively, geodetic target points can be mounted to a tunnel
linings’ surface in order to measure their displacements by means of total stations. How-
surface in order to measure their displacements by means of total stations. However, these
ever, these measurements also miss out on the internal strain distribution data in addition
measurements also miss out on the internal strain distribution data in addition to being
to being taken averagely only once per day [147]. Finally, discrete and punctual sensors
taken averagely only once per day [147]. Finally, discrete and punctual sensors also pack
also
a verypack a very
limited limited performance
performance being where
being the location the location where the
the structural structural
failures might failures
appear
might appear unknown a priori thus impossible to
unknown a priori thus impossible to accurately instrument. Distributed sensing,accurately instrument. Distributed
instead,
sensing,
surpassesinstead, surpasses
all the above all the above
limitations limitations
and positions and
itself as positions
a permanent itselfandas autonomous
a permanent
and autonomous
monitoring solution monitoring
capable of solution capable
measuring of measuring
physical parametersphysical
alongparameters
the whole alonglengththe of
whole length of the tunnel
the tunnel with a single transducer. with a single transducer.
DOFS’ potential application
DOFS’ potential application to tunneling was
to tunneling proved feasible
was proved feasible andand satisfactory
satisfactory in in aa
few pioneering publications such as 2010’s Shi et al. [148] (who
few pioneering publications such as 2010’s Shi et al. [148] (who monitored with a BOTDR monitored with a BOTDR
DOFS
DOFS the the deformations
deformations of of the
the extra-long
extra-long Taiwan
Taiwan StraitStrait Tunnel running over
Tunnel running over 150
150 km km ofof
seafloor geologic body)
seafloor geologic body) andand 2015’s
2015’s Gue
Gue et et al.
al. [16]
[16] (who
(who monitored through BOTDR
monitored through BOTDR DOFS DOFS
London’s Royal Mail tunnel during the construction of the large
London’s Royal Mail tunnel during the construction of the large cross-rail platform tunnel cross-rail platform tunnel
located
located below it).
below it).
Starting
Starting the present
the presentdiscussion
discussionfrom fromthe thetunnel
tunnel monitoring
monitoring during
during their construction
their construc-
phase,
tion phase, a common building methodology is the New Austrian Tunneling (NATM)
a common building methodology is the New Austrian Tunneling Method Method
which iswhich
(NATM) executed through
is executed excavation
through excavation sequences
sequences thatthat
allow
allowfor forthethe surrounding
ground/rock
ground/rockto tosupport
supportitselfitself[147].
[147].The
Themethod’s
method’s first phase
first phaseconsists
consists of of
thetheexcavation
excavation of
the top-heading
of the top-heading areaareaof of
thethetunnel,
tunnel,later
latersupported
supportedby bybolts,
bolts,lattice
latticegirders
girdersand and aa rein-
forced shotcrete
shotcrete (concrete
(concretespraying)
spraying)lining.
lining.The Thesecond
secondphase phaseconsists
consists ofofthetheexcavation
excavation of
the bench/invert
of the bench/invert areaarea
whichwhichis later supported
is later supported by more reinforced
by more shotcrete
reinforced to form
shotcrete a load-
to form a
bearing ring system.
load-bearing ring system.
Monsberg
Monsberg et et al.
al. [147]
[147]reported
reportedaacomprehensive
comprehensivework workonon thethe design
design and and realization
realization of
an an
of OFDR
OFDR DOFSDOFS monitoring
monitoring system where
system the fiber
where was installed
the fiber in theinabove-mentioned
was installed the above-men-
shotcrete
tioned linings situated
shotcrete in a faultinzone
linings situated a faultof the
zoneSemmering Base railway
of the Semmering BaseTunnel
railway(Austria),
Tunnel
at the time under construction following the NATM. The
(Austria), at the time under construction following the NATM. The described campaign described campaign saw the
distributed monitoring of strains and temperature by means
saw the distributed monitoring of strains and temperature by means of a 230 m long DOFSof a 230 m long DOFS located
on two layers
located on twooflayers
a study of case tunnel
a study casecross-section (rock-side(rock-side
tunnel cross-section and cavity-side as visible in
and cavity-side as
Figure 63).
visible in Figure 63).

Figure
Figure 63.
63. DOFS
DOFS installation
installation system
systeminside
insidethe
thetunnel
tunnellining:
lining:(a)
(a)Schematic
Schematicrepresentation
representationand
and(b)
(b)
practical realization [147].
practical realization [147].

The several weeks-long measurements began from the moment the shotcrete started
curing and continued over the advancement of the tunnel drive construction. In order to
prove the suitability of the DOFS-based monitoring system, the results were compared
with the ones coming from Vibrating Wire Sensors (VWS in Figure 63) and geodetic
measurements. Approximately 12 h after the construction of the top-heading, the authors
Sensors 2021, 21, x FOR PEER REVIEW 68 of 87

Sensors 2021, 21, 1818 65 of 83

prove the suitability of the DOFS-based monitoring system, the results were compared
with the ones coming from Vibrating Wire Sensors (VWS in Figure 63) and geodetic meas-
urements. Approximately 12 h after the construction of the top-heading, the authors re-
ported observing
reported compression
observing zones
compression at the
zones atcrown and the
the crown andshoulders (as reported
the shoulders in Figure
(as reported in
64).
Figure 64).

Figure
Figure 64.
64. Measured
Measured strain
strain distribution
distributionat atthe
therock-side
rock-sidelayer
layerin
inperipheral
peripheraldirection:
direction:(a)
(a)12
12hours
h after
after installation
installation at top-heading,
at top-heading, (b) 48 (b) 48 hours
h after after installation
installation at top-heading,
at top-heading, (c) 12 h after(c) 12 hours after
installation at invert
installation
and (d) 48 hatafter
invert and (d) 48
installation athours
invertafter
[147].installation at invert [147].

Then,
Then, 48
48 hhours
after after the construction
the construction of theofbench/invert,
the bench/invert, the deformations
the deformations were re-
were reported
ported to be almost
to be almost constant constant along
along the the entire
entire supporting
supporting ring suggesting
ring suggesting that itthat it formed
formed a load-a
load-bearing
bearing system system capable
capable of coping
of coping withwith the remaining
the remaining loads.
loads. TheThe authors
authors concluded
concluded by
by remarking
remarking thatthat
DOFS DOFS allowed
allowed to precisely
to precisely localize
localize the maximum
the maximum strainstrain
withinwithin the
the cross-
cross-section instantly
section instantly (differently
(differently than than conventional
conventional geodetic
geodetic measurements)
measurements) along
along with with
its
its deformation
deformation history.
history. Yet,Yet,
theythey
also also
notednoted thatposition
that the the position
of theofsensor
the sensor
versusversus the di-
the direction
rection of the drive
of the tunnel tunnelsignificantly
drive significantly
influencedinfluenced the results.
the results.
Whilst
Whilst the
thepresence
presenceofoftemporary
temporary shorings
shorings ensure
ensurethethe
stabilization of large
stabilization cross-sec-
of large cross-
section
tion tunnels
tunnels excavated
excavated according
according to the
to the NATM,
NATM, their
their consequent
consequent dismantling
dismantling during
during the
the secondary lining’s construction process is amongst the most dangerous
secondary lining’s construction process is amongst the most dangerous steps of the whole steps of the
whole construction
construction development
development due todue to alteration
alteration of the of the tunnel
tunnel lining’slining’s original
original structural
structural equi-
equilibrium.
librium. As such,
As such, Li [149]
Li et al. et al. performed
[149] performeda BOTDRa BOTDR
DOFSDOFS monitoring
monitoring in order in to
order to
deter-
determine
mine the Beijing
the Beijing Rail Transit
Rail Transit airport airport line tunnel’s
line tunnel’s deformation
deformation behaviorbehavior
during during such
such shor-
shoring
ing dismantling
dismantling stage.stage.
A DOFS A DOFS
was usedwasin used in combination
combination with a with a metal-reinforced
metal-reinforced strain-
sensing optical cable (purposefully designed for harsh environments). Three DOFSDOFS
strain-sensing optical cable (purposefully designed for harsh environments). Three were
were positioned
positioned in 1three
in three 1 cm longitudinal
cm deep deep longitudinalgrooves grooves
incisedincised
in the in the tunnel
tunnel primary primary
lining
lining which were later sealed with quick setting high-strength mortar. The sampled
measurements showed large settlement variations during all the dismantling steps, the
largest of which was 5.7 mm corresponding to a tensile strain variation of 50 µε in the
crown. Furthermore, the authors were faced with the limitation of being unable to measure
 Sensors 2021, 21, x FOR PEER REVIEW 69 of 87

Sensors 2021, 21, 1818

which were later sealed with quick setting high-strength mortar. The sampled 66 measure-
of 83
ments showed large settlement variations during all the dismantling steps, the largest of
which was 5.7 mm corresponding to a tensile strain variation of 50 με in the crown. Fur-
thermore, the authors were faced with the limitation of being unable to measure directly
directly the tunnel’s
the tunnel’s horizontal
horizontal deformation
deformation due to due to theDOFS
the axial axialdeployment
DOFS deploymentalong the along
tunnel.
theThe
tunnel. The limitation
limitation was surpassed
was surpassed by extracting
by extracting an inversion
an inversion model ofmodel of the
the vault vault
settlement
settlement curveon
curve based based on a numerical
a numerical simulation
simulation and the and the strains
strains obtained
obtained usingusing
DOFS. DOFS. In
In conclu-
conclusion, the authors estimated a 0.9 mm maximum deviation of the
sion, the authors estimated a 0.9 mm maximum deviation of the tunnel’s settlement. tunnel’s settlement.
BOTDR
BOTDR waswas
alsoalso
employed
employed by Fajkus
by Fajkuset al.
et[146] in order
al. [146] to perform
in order a security
to perform moni-
a security mon-
toring of the
itoring of structural loads
the structural carried
loads by an
carried by867 m long
an 867 m longhighway
highwaytunnel under
tunnel construction
under construction
in Žilina (Slovakia)
in Žilina also
(Slovakia) following
also following thethe
NATM.
NATM. In In
this case,
this too,
case, thethe
too, DOFS
DOFS was
was placed
placedonon a
a tunnel’s braced girder which would later be incorporated in the first
tunnel’s braced girder which would later be incorporated in the first of two linings of two linings by by
means of concrete spraying (see Figure
means of concrete spraying (see Figure 65). 65).

(a) (b)
Figure
Figure 65. Tunnel
65. (a) (a) Tunnel
and and distributed
distributed temperature
temperature andand strain
strain system
system layout
layout andand (b) optical
(b) optical fiberfiber implementation
implementation on iron
on iron
barsbars of braced
of the the braced girder
girder of primary
of the the primary lining
lining [146].
[146].

TheThe authors
authors reported
reported thethe
loadload changes
changes (referred
(referred to the
to in in the article
article as changes
as changes of the
of the
Brillouin
Brillouin frequency)
frequency) occurring
occurring over
over a period
a period of time
of time of 5ofmonths
5 months andand hypothesize
hypothesize their
their
origin
origin to betoabe a combination
combination of theoftemperature
the temperature variation
variation (ofconcrete
(of the the concrete and
and the the surround-
surrounding
ing and
rocks) rocks) and concrete
concrete shrinkage.shrinkage.
As As mentioned
mentioned beforehand,
beforehand, thethe monitoring
monitoring of tunnels
of tunnels under
under service
service is asisimportant
as important
as the oneone
as the during their
during construction
their construction phase. A recent
phase. A recent example
example of the former
of the is reported
former in in
is reported
Gómez
Gómez et al.et[150]. This This
al. [150]. article reported
article the implementation
reported the implementation of an OFDR
of an DOFS
OFDRonDOFS the TMBon the
L-9TMB
metroL-9 tunnel
metro in tunnel
Barcelona in order toinassess
in Barcelona orderwhether
to assessit whether
was affected byaffected
it was the construction
by the con-
of astruction
residential of building 14 mbuilding
a residential above it 14 (of m
which
above roughly 37 m2 roughly
it (of which could be37 found right above
m2 could be found
the right
tunnel). In particular, a 50 m long fiber was glued to the tunnel’s
above the tunnel). In particular, a 50 m long fiber was glued to the tunnel’s cross-cross-section expected
to suffer
section theexpected
largest deformations
to suffer the (thus
largestdefined as critical(thus
deformations section), to a segment
defined as criticalbeneath theto a
section),
tracks and tobeneath
segment a segment the of the linen
tracks and to parallel to theoftunnel
a segment (as represented
the linen parallel to thein Figure
tunnel 66). In
(as repre-
thissented
case, the asymmetrical
in Figure construction-induced
66). In this case, the asymmetrical loading and unloading effects
construction-induced wouldand
loading notun-
be directly
loading translated
effects would in a not
uniform pressure
be directly around in
translated thea ring
uniformdue pressure
to the additional
around the presence
ring due
of atobending effect. This was observed during the in-situ monitoring
the additional presence of a bending effect. This was observed during the in-situ mon- which reportedly
detected
itoringcompression
which reportedlyin the linen’s
detectedintrados (where the
compression DOFS
in the was actually
linen’s intradosbonded)
(where during
the DOFS
thewas
excavation
actually bonded) during the excavation phase and tension during the the
phase and tension during the building’s construction (when externalcon-
building’s
load was negative).
struction (when the external load was negative).
For proper analysis and visualization of the data, the authors mentioned having used
complex data treatment algorithms [151] to get rid of a significant number of anomalous
readings in the raw strain data. The authors concluded by highlighting the need for general,
 Sensors 2021, 21, 1818 67 of 83

Sensors 2021, 21, x FOR PEER REVIEWstandardized 70 of 87

guidelines which also account for the structure’s nature, for the temperature
effects and for the adopted post-processing techniques.

Figure66.
Figure Monitoredtunnel
66.Monitored tunnelsection
section[150].
[150].

Minardo
For proper et al. [152]
analysis andemployed distributed
visualization sensing
of the data, for the long-term
the authors mentionedmonitoring
having usedof
a railway tunnel (originally built in 1880 and reconstructed in
complex data treatment algorithms [151] to get rid of a significant number of anomalous 1992) in the accumulation
zone of in
readings an the
activerawearthflow
strain data. in the
TheSouthern Italian Apennines.
authors concluded The 1 cm/year
by highlighting the needdisplacing
for gen-
eral, standardized guidelines which also account for the structure’s nature, thin
landslide had caused the opening of the joints and the appearance of some for thecracks
tem-in
the tunnel. The latter had been monitored in
perature effects and for the adopted post-processing techniques.recent years by means of a 200 m long BOTDA
DOFS hand-glued to its walls. The paper presented the results of such monitoring and in
Minardo et al. [152] employed distributed sensing for the long-term monitoring of a
particular the increase in joint opening/misalignment.
railway tunnel (originally built in 1880 and reconstructed in 1992) in the accumulation
Further examples are provided by Inaudi and Walder [145] who reported two dis-
zone of an active earthflow in the Southern Italian Apennines. The 1 cm/year displacing
tributed sensing-powered tunnel monitoring. The first was deployed in 1968’s San Sal-
landslide had caused the opening of the joints and the appearance of some thin cracks in
vatore Tunnel (Italy) where water accumulation behind a concrete lining had caused its
the tunnel. The latter had been monitored in recent years by means of a 200 m long
collapse. To prevent this from reoccurring, two Brillouin scattering-based DOFS were
BOTDA DOFS hand-glued to its walls. The paper presented the results of such monitoring
bonded to the tunnel linings allowing for the detection and localization (with a 1 m spatial
and in particular the increase in joint opening/misalignment.
resolution) of any hydrostatic pressure-induced deformation and cracking. The effects of
Further examples are provided by Inaudi and Walder [145] who reported two dis-
temperature variations (in the order of magnitude of 100–200 µε) were accounted for and
tributed sensing-powered tunnel monitoring. The first was deployed in 1968’s San Salva-
compensated by means of an extra temperature sensing cable. In the second study case,
tore Tunnel (Italy) where water accumulation behind a concrete lining had caused its col-
the average strain distributions along five longitudinal lines of a high-speed rail tunnel
lapse. To prevent
in Barcelona (Spain)thiswerefrom reoccurring,
monitored. Here, twotwoBrillouin
DOFS were scattering-based
deployed on the DOFS sidewere
walls
bonded
and one toontheitstunnel
floor, linings
between allowing
the twofor railthe detection
lines (summing andup localization (with a 1 mlength
to a total monitored spatialof
resolution)
12 km). of any hydrostatic pressure-induced deformation and cracking. The effects of
temperature
As seen variations
before in(in the order
[149,150] theof magnitude
tunnel of 100–200
monitoring με)accompanied
is often were accounted by afor and
certain
compensated by means of an extra temperature sensing cable.
amount of computational and algorithmic work whether this is for validation, inverse- In the second study case,
the average strain
calculation distributions
or forecasting along five longitudinal lines of a high-speed rail tunnel in
purposes.
Barcelona (Spain)
In Zhao were
et al. monitored.
[153] the authorsHere, two DOFS
performed a were
BOTDR deployed
DOFS on the side walls
monitoring systemandof
one on its
1984’s floor,Baiyun
Dalian between the two
Tunnel. rail nine
Here, lines6.6(summing up to a totalstretches
m long optical-fiber monitored length
were of 12to
bonded
km).
the tunnel vault with polyimide adhesive and covered with fiber cloth. The authors used
the As seen before instrains
DOFS-extracted [149,150] the tunnel
to study monitoring
the vault is oftenwhich,
displacement accompanied by a certain
in combination with
amount
the surrounding rock structure’s data, yielded a Differential Evolution Algorithminverse-
of computational and algorithmic work whether this is for validation, capable
calculation
of carryingorout forecasting purposes.
inverse analyses on the major mechanical impact parameters of a tunnel’s
In Zhao
lining et al. [153] the authors performed a BOTDR DOFS monitoring system of
structure.
1984’s Piccolo
Dalian Baiyun Tunnel.
et al. [154], Here,proposed
instead, nine 6.6 m long optical-fiber
a DOFS-based stretches were
methodology bonded to
for convergence
the tunnel vault with
measurements over polyimide adhesive
time of existing and covered
tunnels. with fiber
In particular, cloth. from
starting The authors used
experimental
the DOFS-extracted strains to study the vault displacement
orthoradial PPP-BOTDA and Tunable Wavelength Coherent Optical Time Domain Reflec- which, in combination with
the surrounding
tometry (TW-COTDR)rock structure’s
DOFS strain data,measurements,
yielded a Differential
an inverseEvolution
analysisAlgorithm
associated capable
with a
offinite
carrying out inverse
element model wouldanalyses on the
allow majorinsight
gaining mechanical
into aimpact
tunnel’s parameters
deformedof a tunnel’sIn
geometry.
lining
orderstructure.
to understand the practicability of this procedure, the authors performed a laboratory
Piccolo et al. [154], instead, proposed a DOFS-based methodology for convergence
measurements over time of existing tunnels. In particular, starting from experimental or-
Sensors 2021, 21, 1818 68 of 83

test on a 762 mm diameter steel ring, horizontally fixed at different points to an external
reaction frame. The orthoradial strain values were sampled by means of the two DOFS
along with displacement sensors. The authors concluded by stating that the DOFS seemed
to be sufficiently performant to obtain convergence measurement. They also noticed that, as
the testing force grew, DOFS soldered to the steel ring’s surface provided better agreement
with reference values compared to the simply glued ones.
The present sub-section saw the DOFS SHM monitoring of both new and older tunnels.
In the first scenario most applications were aimed at establishing the deformations of a
study case cross section by means of DOFS positioned inside the steel lattice girders during
the various stages of NATM construction. The monitoring system allowed to instantly and
precisely localize the maximum cross-sectional strains along with its deformation history
(over multiple months-long in one of the reported cases). The orientation of the sensors
compared to the direction of the tunnel drive was seen to have a large influence. For
instance, in one of the reported campaigns, due to the DOFS longitudinal deployment, the
use of an inversion model of the vault settlement curve based on a numerical simulation
was necessary to establish the tunnel’s horizontal displacements occurring during the
removal of temporary shoring. Furthermore, the use of expressly designed resistant DOFS
was necessary for a successful sensor embedment inside concrete. This precaution was
not necessary for externally bonded DOFS, as was the case for most monitoring of already
existent tunnels where the fibers were simply glued to the linens’ intrados. For the latter,
the main takeaway was the successful monitoring of minute but relevant deformations
induced, for example, by the construction above the tunnel of a residential building or by
water accumulating behind a tunnel’s concrete linings or even by the continuous effect of
an earthflow occurring above the tunnel.

8. Pipelines
Pipelines, whether underground, above ground or subsea, are one of the most impor-
tant ways to transport across extensive distances key resources such as water, oil, natural
gases and telecommunication cables. Furthermore, pipeline transportation is substantially
more efficient (financially, environmentally and labor-intensely wise) compared to other
ground transportation alternatives such as trucks and trains. Yet, the economic importance
of laying and maintaining pipelines is absolutely non-neglectable and should be pondered
seriously. Maintenance, in particular, is crucial in those common situations where the
pipelines are located in complex and hazardous environments (exposed, for instance, to
biological and/or chemical hazards, landslides, soil subsidence and third party interfer-
ence) which might lead to issues such as pitting and patch corrosion, cracking and to an
overall degradation of a pipeline’s structural integrity. The above could even represent a
great safety hazard and risk of environmental pollution in case of pipe leaks and bursts.
Thus, the constant monitoring of a pipeline’s structural health is crucial to guaranteeing
its safe functioning and to averting the substantial economic, environmental and social
consequences of its failure.
Generally, the SHM for pipelines demands the observation of their deformation,
curvature, eventual leaks and the movements of their surrounding ground. Yet, considering
their kilometric lengths and beforehand-unknown failure location, accurate structural
monitoring of their entire extension is challenging for discrete sensors. This is not the case
for DOFS which completely bypasses the need to install thousands of discrete sensors (with
their complex cabling) in order to cover all the potential failure locations. This by itself
makes DOFS the ideal pick for a pipeline’s SHM.
This was clearly established by pioneering publications such as Glisic and Yao [155]
(who developed a novel Brillouin scattering-based DOFS-based method for real-time,
automatic or on-demand assessment of the health condition of buried pipelines after an
earthquake) and Inaudi and Glisic [156] (who monitored a 35 years old gas pipeline near
Rimini, Italy through a Brillouin scattering-based DOFS).
 This was clearly established by pioneering publications such as Glisic and Yao [155]
(who developed a novel Brillouin scattering-based DOFS-based method for real-time, au-
Sensors 2021, 21, 1818 tomatic or on-demand assessment of the health condition of buried pipelines after 69 ofan
83
earthquake) and Inaudi and Glisic [156] (who monitored a 35 years old gas pipeline near
Rimini, Italy through a Brillouin scattering-based DOFS).
More recently, Wong et al. [157] attempted to experimentally demonstrate the ability
of an More
OFDRrecently,
DOFS to Wong et al.the
monitor [157] attempted
growth to experimentally
of fatigue cracks along demonstrate the ability
a cast-iron pipeline. A
of an OFDR DOFS to monitor the growth of fatigue cracks along a cast-iron
study case pipe was subjected to internal cyclic water pressure loadings after its thickness pipeline. A
study case pipe was subjected to internal cyclic water pressure loadings
was artificially reduced in a small elliptical patch (leaving a minimum final wall thickness after its thickness
was
of 3.5artificially reduced ainseverely
mm) to simulate a small elliptical
corrodedpatch area.(leaving
DOFS wasa minimum
deployedfinal
on wall thickness
the surface of
such a patch from the center of which a longitudinal crack appeared and started toofprop-
of 3.5 mm) to simulate a severely corroded area. DOFS was deployed on the surface such
a patchDOFS
agate. from effectively
the center of which athe
detected longitudinal
latter two crack appeared
phenomena butand
onlystarted
when to thepropagate.
distance
DOFS effectively detected the latter two phenomena but only
between the sensor and the damage was less than 40 mm. The authors concluded when the distance betweenby
the sensor and the damage was less than 40 mm. The authors concluded
stressing the importance of the orientation of DOFS and suggesting that it hinges on the by stressing the
importance
minimal of the
critical orientation
patch and crackof DOFS
lengthandthatsuggesting
is intendedthat
to it hinges on the minimal critical
detect.
patchRenandetcrack length that is intended to detect.
al. [158] also attempted to monitor both corrosion and leakage by means of
OFDR. RenIn et
theal.corrosion
[158] alsotest,
attempted to monitor
several optical fiberboth corrosion
sensors and leakage
were bonded by means
to a pipe’s surfaceof
OFDR. In the corrosion test, several optical fiber sensors were bonded
in order to form a sensor array capable of monitoring and reporting any corrosion-in- to a pipe’s surface in
order to form a sensor array capable of monitoring and reporting any corrosion-induced
duced variation in the hoop strains, explicitly strains along the circumferential direction.
variation in the hoop strains, explicitly strains along the circumferential direction. In the
In the leakage test, instead, the bonded DOFS were intended to detect any leakage-in-
leakage test, instead, the bonded DOFS were intended to detect any leakage-induced
duced internal pressure drop along the pipe. According to the authors, the successful de-
internal pressure drop along the pipe. According to the authors, the successful detection of
tection of both corrosion (see Figure 67) and leaks implies OFDR’s high efficiency and
both corrosion (see Figure 67) and leaks implies OFDR’s high efficiency and accuracy for
accuracy for pipeline SHM.
pipeline SHM.

(a) (b)

(c) (d)
Figure
Figure 67.
67. Hoop
Hoop strain
strain nephogram
nephogram when
when thethe pipe
pipe was
was corroded
corroded at
at 200
200 hh (a)
(a) 2D
2D and
and (b)
(b) 3D
3D and
and wall
wall thickness
thickness nephogram
nephogram
when the pipe was corroded at 200 h. (c) 2D and (d) 3D. [158].
when the pipe was corroded at 200 h. (c) 2D and (d) 3D. [158].

In recent years, pipeline leak detection has also been investigated by means of Dis-
tributed optical fiber base Acoustic Sensing (DAS). The latter, also referred to as Distributed
Vibration Sensing (DVS), belongs to the latest developments of DOFS technologies and are
part of the Rayleigh-based DOFS family. DAS is an ideal fit for pipeline SHM as their base
principle, heterodyne coherent optical time-domain reflectometry (Φ-OTDR), grants them
the ability to detect highly-dynamic vibration signals up to the kHz range. Here, the elastic
waves phase change produced by perturbing objects in the fiber’s proximity is detected by
 In recent years, pipeline leak detection has also been investigated by means of Dis-
tributed optical fiber base Acoustic Sensing (DAS). The latter, also referred to as Distrib-
uted Vibration Sensing (DVS), belongs to the latest developments of DOFS technologies
and are part of the Rayleigh-based DOFS family. DAS is an ideal fit for pipeline SHM as
Sensors 2021, 21, 1818 70 of 83
their base principle, heterodyne coherent optical time-domain reflectometry (Φ-OTDR),
grants them the ability to detect highly-dynamic vibration signals up to the kHz range.
Here, the elastic waves phase change produced by perturbing objects in the fiber’s prox-
imity is detected
an Acoustic OpticbyModulator.
an AcousticThus,
Opticthe
Modulator. Thus, the
phase-sensitive OTDRphase-sensitive
can be usedOTDR can be
for real-time
used for real-time
vibration vibration
monitoring monitoring
of extended of extended
objects, objects,
e.g., bridges, e.g., highways,
dams, bridges, dams, highways,
railway tracks,
railway
pipelines, tracks,
pavedpipelines, paved
paths, and many paths, and many others.
others.
In
In2018,
2018,Stajanca
Stajancaet
etal.
al.[159]
[159]performed
performedaahelical
helicalDAS
DASwrapping
wrappingaroundaroundfour
foursegments
segments
of
of aa 38m
38m long
long DN100
DN100 pipe
pipe and
and simulated
simulated leaks
leaks using
using holey
holey adapter
adapter caps
caps (with
(with circular
circular
holes of different diameters) inserted in the pipe’s side adapter
holes of different diameters) inserted in the pipe’s side adapter (as in (as in Figure 68).

Figure
Figure68.
68.(a)
(a)Schematic
Schematicillustration
illustrationof
ofthe
thefiber
fiberhelical
helicalwrapping
wrappingaround
around the
the pipe
pipe segments
segments and
and (b)
(b) photo of one of the pipe’s side adapters
photo of one of the pipe’s side adapters [159].[159].

The
The goal,
goal,studying
studyingDAS’ DAS’potential
potential toto
detect
detecta pipe’s
a pipe’s weakweakpinhole
pinhole leak-induced
leak-induced vi-
brations
vibrationsforfor
various
variouscombinations
combinations of leak sizes sizes
of leak and pipeline internal
and pipeline pressures.
internal The results
pressures. The
revealed that the
results revealed DAS
that deployment
the DAS deployment was was
capable
capableof detecting
of detecting thethepipeline’s
pipeline’svibration
vibration
modes
modesinduced
inducedby bythe
thebroadband
broadband leak-noise
leak-noise excitation.
excitation.In particular,
In particular,according
according to the
to au-
the
thors, leaking
authors, leakingthrough
through increasingly
increasingly larger
largerhole
holesizes
sizesled
ledtotoexcitation
excitationofofhigher
higherfrequencies
frequencies
vibrational
vibrational modes.
modes.
A different
A different DAS deployment
deployment methodology
methodologyisisproposed
proposedininZuo Zuoetetal.al.[29].
[29].InInparticular,
particu-
thethe
lar, authors
authors freely
freelysuspended
suspended thethe
fiber outside
fiber outsidea a12.9
12.9mmlonglongØ0.3
Ø0.3 m m gas
gas pipeline with with
the goal of studying the former’s leak detection performance. As
the goal of studying the former’s leak detection performance. As a consequence of the a consequence of the
deployment methodology, the experiment had to be carried out
deployment methodology, the experiment had to be carried out in a fully muffled cham-in a fully muffled cham-
ber with
ber with aa background
background noise noise of
of less
less than
than 4040 dB.
dB. The
The authors
authors reported
reported aa successful
successful leak
leak
detection and
detection and localization
localization by by means
means of of detailed
detailed algorithms
algorithms based based onon frequency
frequency domaindomain
cumulativeaveraging.
cumulative averaging.
As previously mentioned, along
As previously mentioned, along with
with aa pipes’
pipes’ leak
leak monitoring,
monitoring, its its deformation
deformation mon- moni-
toring is also of crucial importance. As such, Feng et al. [160] reported
itoring is also of crucial importance. As such, Feng et al. [160] reported a methodology for a methodology for
monitoring the upheaval buckling of subsea buried pipelines by
monitoring the upheaval buckling of subsea buried pipelines by means of BOTDA DOFS. means of BOTDA DOFS.
The authors were faced with the challenge of separately detecting a pipe’s buckling-related
bending-induced strains and axial force-induced strains instead of their compound strain.
The proposed solution was the longitudinal deployment of three DOFS sensors along a
pipe at a radial distance of 2π/3 from each other in addition to a proposed mathematical
process. An experimental test was then designed in order to check the performance of such
a monitoring scheme. A DOFS-instrumented 5.71 m polyvinyl chloride (PVC) pipe was
buried inside standard compacted sand with one of its extremities axially loaded by means
of a hydraulic jack. According to the authors, the preliminary results demonstrated the
Sensors 2021, 21, 1818 71 of 83

efficiency of the proposed monitoring method despite noting that further research should
be dedicated to its performance analysis when including the complex thermomechanical
circumstances of buried pipelines.
The deformation of pipelines was also studied by Zhang et al. [161] who instrumented
a 4 m long PVC pipeline with two opposite longitudinal runs of BOTDA DOFS and multiple
electrical resistance SG in order to verify the former’s monitoring performance superiority
for long-distance pipelines. The pipe was gradually subjected to five different loads in its
midspan and its displacements were calculated on the grounds of the DOFS-extracted and
SG-extracted strains by means of the Conjugate Beam Method. The results led the authors
to the conclusion that the measurement accuracy of a pipeline’s displacement is correlated
to the number of monitored points thus proving the superiority of DOFS over SG-based
for such applications.
Conclusively, multiple DOFS SHM studies were dedicated to the detection of pipeline
leaks. This was performed both by means of standard DOFS and DAS. When a standard
strain sensing DOFS was bonded locally, in the proximity of a corrosion-damaged patch,
the results showed the need of the fiber to be close no more than 40 mm from the damage
to actually be able to detect it. Alternatively, DOFS or DAS could be wrapped along the
entire extension of the pipe to detect any potential peaking of hoop strains and vibration,
respectively, indicative of corrosion and/or leaking. The authors reported successfully
detecting and localizing leaks by means of this technique. Alternatively, DAS could
be suspended outside of the pipe in order to detect any leak-induced vibrations (if the
requirement of low environmental noise can be met). Finally, in order to monitor the
occurrence of pipeline deformation, DOFS was deployed on the latter’s surface by means
of a double and opposite fiber run or by means of a triple fiber run at a radial distance
of 2π/3 from each other. Tests performed with these methodologies both yielded key
deformation data.
For further reading, the author suggests the following recently published review
article, Inaudi [162].

9. Wind Turbines
The popularity of wind turbines as an additional/alternative renewable source of
power is steadily increasing over the years, contributing to the reduction of fossil fuel
reliance. According to a report from the World Wind Energy Association [163] the wind
power capacity worldwide has reached a total energetic value of 597 GW in 2018 with
a growth of 9,1%, compared to the previous year and 10,8% versus the one before that.
Worth noting is the potential brought forwards by offshore wind power plants. These
have the advantages of high wind speed, low wind edge cutting, low turbulence, and high
output [164] despite requiring considerable engineering efforts for their design. In addition
to these, novel engineering challenges are arising in the past few years as a consequence
of the increased size characterizing the newly designed/produced wind turbines (for the
sake of increased productivity). Thus, in order to reduce the risk of failure induced by the
consequent greater forces acting on the wind turbines’ structures, a comprehensive SHM
has become of paramount importance.
Coscetta et al. [164], for example, reported the results of two BOTDA DOFS powered
tests aimed at demonstrating the possibility of performing a static and dynamic strain
and vibration monitoring of a 14 m long composite wind turbine blade. In the first test,
the blade was stressed as a simple cantilever beam (the load was applied in the flap-wise
direction) and it was monitored by a fiber attached along the middle of the blade’s upper
and lower surfaces. The resulting non-linear strain profiles were explained to be the fruit of
the combined effect of the cantilever configuration and the varying geometry of the blade
itself. This aside, the authors reported a high level of linearity between the applied load
and the measured strains. In the dynamic test, a 100 kg load was applied on the blade’s
free end and the latter was put into vibration by means of an electrical motor positioned on
 blade was stressed as a simple cantilever beam (the load was applied in the flap-wise di-
rection) and it was monitored by a fiber attached along the middle of the blade’s upper
and lower surfaces. The resulting non-linear strain profiles were explained to be the fruit
Sensors 2021, 21, 1818 of the combined effect of the cantilever configuration and the varying geometry 72 ofofthe
83
blade itself. This aside, the authors reported a high level of linearity between the applied
load and the measured strains. In the dynamic test, a 100 kg load was applied on the
blade’s free end and the latter was put into vibration by means of an electrical motor po-
it. Expectedly,
sitioned the amplitude
on it. Expectedly, theof the extracted
amplitude of thestrains wasstrains
extracted larger was
on the clamped
larger on theend of the
clamped
blade
end ofthan on itsthan
the blade free one.
on its free one.
In Raman
In Raman et et al.
al. [165],
[165], the
the authors
authors stated
stated the
the importance
importance of of performing
performing aa real-time
real-time
SHM of the adhesives present in a blade’s trailing edges (joining the
SHM of the adhesives present in a blade’s trailing edges (joining the upper
upper andand lower
lower
composite structures),
composite structures), aacommon
commonfailurefailurelocation.
location.TheThe article’s
article’s research
research efforts
efforts werewerefo-
focused
cused onon establishingthe
establishing theoptimal
optimalDOFSDOFSpositioning
positioningtechnique
technique for for multi-direction
multi-direction strain
strain
monitoring within
monitoring within thethe carbon
carbon composite
composite plies. In particular,
plies. In particular, thethe multi-linear
multi-linear placement
placement
was discouraged due to its limited strain sensing direction (only
was discouraged due to its limited strain sensing direction (only the lateral one) the lateral one) and, in its
and, in
stead, a dual sinusoidal placement was commended. Indeed, the
its stead, a dual sinusoidal placement was commended. Indeed, the latter is reportedlylatter is reportedly able
to detect
able strain
to detect signals
strain bothboth
signals in the
inlongitudinal and and
the longitudinal in the
in lateral directions
the lateral allowing
directions allowingthe
simultaneous characterization of a blade’s bending and torsional
the simultaneous characterization of a blade’s bending and torsional loads. loads.
The above-mentioned
The above-mentioned wind wind turbine
turbine blade’s
blade’s size
size surge
surge has
has the
the additional
additional side
side effect
effect
of increasing the overturning moment withstood by a pile’s foundation. Zhang et al. [26]
of increasing the overturning moment withstood by a pile’s foundation. Zhang et al. [26]
used aa Differential
used Differential Pulse
Pulse Pair
Pair Brillouin
Brillouin Optical
Optical Time-Domain
Time-Domain Analysis
Analysis (DPP-BOTDA)
(DPP-BOTDA) to to
monitor a 69 m tall offshore wind turbine pile (belonging to the Xinghua Bay Wind Farm
monitor a 69 m tall offshore wind turbine pile (belonging to the Xinghua Bay Wind Farm
in Fujian) in order to obtain the bearing capacity of its foundation (see Figure 69).
in Fujian) in order to obtain the bearing capacity of its foundation (see Figure 69).

(a) (b)
Figure
Figure 69.
69. (a)
(a) Study
Study case
case offshore
offshore turbine (b) DOFS
turbine (b) DOFS bonded
bonded with
with epoxy
epoxy to
to the
the sample
sample pile
pile (adapted
(adapted from
from [26]).
[26]).

The monitored concrete-filled/steel composite pile was a test sample mounted on the
same offshore platform as the actual study study case
case one.
one. The
The author’s
author’s aim waswas pin-pointing
pin-pointing
the pile
pilefoundation’s
foundation’smaximum
maximum stress location
stress under
location horizontal
under loading
horizontal (applied
loading by means
(applied by
of a horizontal
means reaction
of a horizontal frame).frame).
reaction The test,
Thesubdivided into 15into
test, subdivided consecutive load stages
15 consecutive with
load stages
load load
with increments of 60of
increments kN,
60 yielded the maximum
kN, yielded the maximum strain (1436
strain at the
µε)με)
(1436 lastlast
at the load stage
load at
stage
y coordinate − 19.63 m (as visible in Figure 70). Expectedly, the maximum
at y coordinate −19.63 m (as visible in Figure 70). Expectedly, the maximum displacement displacement
(0.79 m) occurred at the uppermost
uppermost endend of
of the
the fiber
fiber and
and the
the load/maximum
load/maximum strain and the
load/maximumdisplacement
load/maximum displacement profiles
profiles were both found to be linear in nature. The authors
concluded commending
concluded commending the the proposed
proposed DPP-BOTDA
DPP-BOTDA for for any offshore wind turbine pile
foundation SHM.
foundation SHM.
Conclusively, it can be stated that few DOFS wind turbines SHM were performed in
the latest years. The reported research efforts saw three different DOFS bonding conditions,
one on the surface of turbine blades, one inside them (through an innovative deployment
technique named dual sinusoidal placement) and one more on the pile. Throughout the
articles, both static and dynamic tests were performed, all successfully yielding the sought-
after results. As such, all the authors commended DOFS for the SHM of wind turbines.
Sensors 2021,21,
Sensors2021, 21,1818
x FOR PEER REVIEW 73 of
76 of 83
87

Figure 70.
Figure 70. Strain
Strain values
values per
per corresponding
corresponding elevation
elevation from
from00kN
kNto
to900
900kN
kN[26].
[26].

10. An Eye towardsitthe

Conclusively, canFuture
be stated that few DOFS wind turbines SHM were performed in
The present
the latest section
years. The intends
reported to concisely
research report
efforts saw novel
three DOFS DOFS
different applications/studies
bonding condi-
that may very well pioneer new ways of performing SHM. Whilst not
tions, one on the surface of turbine blades, one inside them (through an innovativeall the followingde-
applications can be tagged
ployment technique named as dual
strictly Civil Engineering-focused,
sinusoidal placement) and one the potential
more on thatthe they
pile.
bring forth could
Throughout definitely
the articles, benefit
both staticthe field
and in the near
dynamic testsfuture.
were performed, all successfully
For example,
yielding Fernandez-Ruiz
the sought-after results. Aset al. [28]
such, introduced
all the the idea of using
authors commended DOFS the already-
for the SHM
existing
of wind world-wide
Sensors 2021, 21, x FOR PEER REVIEW turbines. telecommunication fiber optic network (Figure 71) for continuous, 77 of 87
real-time monitoring of the global surface on the lookout for any seismic activity.
10. An Eye towards the Future
The present section intends to concisely report novel DOFS applications/studies that
may very well pioneer new ways of performing SHM. Whilst not all the following appli-
cations can be tagged as strictly Civil Engineering-focused, the potential that they bring
forth could definitely benefit the field in the near future.
For example, Fernandez-Ruiz et al. [28] introduced the idea of using the already-ex-
isting world-wide telecommunication fiber optic network (Figure 71) for continuous, real-
time monitoring of the global surface on the lookout for any seismic activity.

Figure 71.Submarine
Figure71. Submarinecable
cablemap
mapinin2015
2015[166].
[166].

ByBy
means
meansofof
the optical
the opticalfibers’ ubiquity
fibers’ and
ubiquity andability
abilitytoto
perform
performdistributed
distributedand highly
and highly
sensitive DAS vibration monitoring, it would be possible to recognize patterns of potentially
sensitive DAS vibration monitoring, it would be possible to recognize patterns of poten-
hazardous seismicseismic
tially hazardous activity.activity.
In particular, the authors
In particular, tested the
the authors feasibility
tested of suchof
the feasibility a plan
such a
plan by monitoring a simulated seismic wave with both a seismometer and DAS sensing
fibers. The comparison of the extracted seismometer spectrogram and DAS’s output (com-
bined with a linear 2D bandpass filtering technique and a 1D adaptive LMS filter) showed
an encouraging time–frequency dynamics similarity (as visible in Figure 72).
 tially hazardous seismic activity. In particular, the authors tested the feasibility of such a
plan by monitoring a simulated seismic wave with both a seismometer and DAS sensing
fibers. The comparison of the extracted seismometer spectrogram and DAS’s output (com-
bined with a linear 2D bandpass filtering technique and a 1D adaptive LMS filter) showed
an encouraging time–frequency dynamics similarity (as visible in Figure 72).
Sensors 2021, 21, 1818 74 of 83
The potential of cabled seafloor observatory networks was also discussed in Hartog
et al. [167]. Here, the authors debated the potential role of DOFS technology in evolving
the traditional single-point isolated oceanographic measurements (performed on the sea-
floor
by or on the surface
monitoring with seismic
a simulated moorings, landers
wave with and
botharrays of surfaceand
a seismometer floats)
DAS bysensing
finally filling
fibers.
in the
The extensive spatial
comparison and information
of the extracted gaps existing
seismometer between
spectrogram andthe latter.
DAS’s This way,
output DOFS
(combined
with
couldahelp
linear 2D bandpass
improve filtering
the modern technique and
understanding a 1D
of the adaptive
world’s LMS
oceans filter)
and showed
of the an
role they
encouraging
play on issuestime–frequency
such as climatedynamics similarity
regulation, (as visible
global food supplyinand
Figure 72). production.
energy

(a) (b) (c)

Figure 72. (a)72.
Figure Spectrogram of theof
(a) Spectrogram seismic signalsignal
the seismic measured by a reference
measured seismograph
by a reference (b) Spectrogram
seismograph of the DAS
(b) Spectrogram of thedata
DAS denoised
data
with denoised
a 2D linearwith a 2D linear bandpass filter and a (c) Spectrogram of the DAS data denoised with a 1D adaptive LMS filter[28])
bandpass filter and a (c) Spectrogram of the DAS data denoised with a 1D adaptive LMS filter (adapted from
(adapted from [28]).
Additionally, DOFS could permit a vaster-deep seafloor monitoring for the detection
of hazards such as earthquakes,
The potential seafloor
of cabled seafloor remobilizations,
observatory networksslope
wasinstability that may
also discussed trigger
in Hartog
tsunamis,
et al. [167]. turbidity
Here, the currents and seafloor
authors debated expulsion
the potential roleof
of fluids
DOFS which all pose
technology a threatthe
in evolving to
traditional single-point
critical seafloor isolated oceanographic
infrastructure measurements (performed
including telecommunication networks, on the seafloor
offshore infra-
or on the surface
structure, oil andwith moorings,and
gas pipelines landers and arrays
umbilicals, windoffarm
surface floats) by
interarray finally
cables andfilling in
coastal
the extensive spatial and information gaps existing between the latter. This way, DOFS
communities.
could help improve the modern understanding of the world’s oceans and of the role they
play on issues such as climate regulation, global food supply and energy production.
Additionally, DOFS could permit a vaster-deep seafloor monitoring for the detection
of hazards such as earthquakes, seafloor remobilizations, slope instability that may trigger
tsunamis, turbidity currents and seafloor expulsion of fluids which all pose a threat to criti-
cal seafloor infrastructure including telecommunication networks, offshore infrastructure,
oil and gas pipelines and umbilicals, wind farm interarray cables and coastal communities.
DAS was also the focus of Huang et al. [168] who proposed a DOFS-based acoustic
environmental safety monitoring scheme for the security of a building’s exterior glass
walls/windows. In order to improve the accuracy of the disturbing event recognition, the
authors suggested the use of the Wigner bispectrum analysis combined with an extreme
gradient boosting tree algorithm. The recognition of eight different 25.3 ms long vibra-
tion events was then tested (no disturbance, wind blowing, knocking, window opening,
watering, beating, dog barking and aircraft sound) resulting in a recognition rate of 93.3%.
Amongst the many potential applications of DAS, Merlo et al. [30] proposed a unique
Φ-OTDR monitoring system buried under different tracks and runways of an airport for
the positioning and trajectory calculation of airplanes and other vehicular transportation.
This system was intended to integrate the current radar surveillance systems operating
in the so-called Precision Runway Monitoring. The gist was that the intensity and spatial
extension of an externally-induced fiber scattering perturbation phenomenon could allow
distinguishing the originating cause (plane, bus, car and human) and its trajectory. With
such purpose, the authors reported the design, development, and characterization of a
narrow-linewidth solid-state laser with extra-low-noise properties. The latter feature was
aimed at preventing any confusion from the processing algorithm’s part between the laser
phase noise and an event occurring along the fiber.
Zhang et al. [52] and Liu et al. [169]’s scope was very similar in nature to the above
article. This time, though, for vehicular traffic flow monitoring. In the former, multi-scale
 extension of an externally-induced fiber scattering perturbation phenomenon could allow
distinguishing the originating cause (plane, bus, car and human) and its trajectory. With
such purpose, the authors reported the design, development, and characterization of a
narrow-linewidth solid-state laser with extra-low-noise properties. The latter feature was
Sensors 2021, 21, 1818 aimed at preventing any confusion from the processing algorithm’s part between the laser 75 of 83
phase noise and an event occurring along the fiber.
Zhang et al. [52] and Liu et al. [169]’s scope was very similar in nature to the above
article. This time, though, for vehicular traffic flow monitoring. In the former, multi-scale
Distributed
Distributed OpticalFiber
Optical FiberVibration
VibrationSensors,
Sensors,essentially
essentiallyDAS,
DAS,and
andFBG
FBG were
were installed
installed on
on a
road pavement to form a load identification system (as in Figure
a road pavement to form a load identification system (as in Figure 73). 73).

(a) (b)
Figure 73.73.
Figure (a)(a)
Fiber Bragg
Fiber Grating
Bragg (FBG)-based
Grating acceleration
(FBG)-based accelerationsensor
sensorand
and(b)
(b)Distributed
DistributedOptical
Optical Fiber
Fiber Vibration Sensor
Vibration Sensor [52]. [52].

TheTheauthors
authorsreported
reportedthat
thatthe
thetest,
test, carried
carried out forfor four
fourvibration
vibrationloadsloads(rockfall
(rockfall im-
impact,
pact, motorcycle, automobile and concrete tanker), showed that
motorcycle, automobile and concrete tanker), showed that both the DAS and the FBG-based both the DAS and the
FBG-based
accelerationacceleration
sensors couldsensors could successfully
successfully identify the identify
different theloads.
different loads.
InIn
LiuLiuet et
al.al.
[169], instead,
[169], instead,thethe
authors
authorssuggested
suggested the the
useuse of DOFS
of DOFSfor the
for detection
the detectionof
traffic vibration
of traffic signals
vibration combined
signals with novel
combined wavelet
with novel threshold
wavelet and dual-threshold
threshold algo-
and dual-threshold
rithms. The efficiency
algorithms. of such
The efficiency a method
of such a methodwas was
laterlater
tested with
tested a DAS-powered
with a DAS-powered vehicle-
vehicle-
counting
counting andand vehicle-speed
vehicle-speed determination
determination test inin
test thetheNanshan
Nanshan Iron
IronMine
Mine (China)
(China) where
where
thethe optical
optical fiber
fiber was
was located
located onon the
the road
road shoulder
shoulder closest
closest totothethe traffic
traffic lane.
lane. The
The test
test results
results
were
were reportedly
reportedly encouragingwith
encouraging withananaccuracy
accuracyerror
errorofoflesslessthan
than6%.6%.
Moving
Moving ononfromfrom macroscale
macroscale DOFS DOFS applications,
applications, Li etLi al. et al. collected
[170] [170] collected
the basethecon-
base
concepts
cepts behind behind
severalseveral
novelnovel optical
optical fibers
fibers for for distributed
distributed sensing.
sensing. Among
Among these,
these, worth
worth
mentioning
mentioning is is the
the idea
idea (originally
(originally introduced
introduced byby BaoBao et et
al.al. [171])
[171]) ofofusing
usingsensing
sensingcables
cables
with two embedded fibers with different Brillouin frequency
with two embedded fibers with different Brillouin frequency shifts. This would permit shifts. This would permit
thethe separation
separation ofof
thethe temperature
temperature and
and strain
strain effects
effects onon thethe Brillouin
Brillouin frequency
frequency shiftthus
shift thus
preventing
preventing their
their stacking
stacking inin
thethe output
output analysis
analysis phase.
phase.
You et al. [172] also brought forth a very interesting proposal for a novel OFDR-DOF
sensing tape aimed at guaranteeing the survival of the DOFS in harsh construction and
service environments. In particular, the authors suggested a packaging technique that
traps DOFS between two fiberglass tapes glued with an adhesive. The strain transfer
mechanism was analyzed in detail followed by a calibration test and a real-life test that
saw its application to the surface of an RC beam subjected to three-point bending. The
extracted results reportedly showed the correct functioning of the sensing tape on concrete
and its ability to localize cracks.
When it comes to modern-day OBR interrogators and BOTDA, the use of post-
processing algorithms is an important step for the proper study of the extracted strain data.
Three examples of these are provided by Bado et al. [151], Heinze and Echtermeyer [27]
and Song et al. [173]. In the former, the authors put into practice two previously established
algorithms and a novel one for the elimination of SRAs risen in two different monitoring
campaigns. The first of these was a laboratory experimental campaign performed on two
RC ties whilst the second was in-situ monitoring of a tunnel deformation. The authors
reported the presence of large performance discrepancies between the three analyzed
algorithms. Indeed, the classic methods (Spectral Shift Quality and Geometrical Threshold
Method) provided results with questionable quality despite their quick and small computa-
tional taxing analysis. Instead, the novel but more computationally taxing one (Polynomial
 lished algorithms and a novel one for the elimination of SRAs risen in two different mon-
itoring campaigns. The first of these was a laboratory experimental campaign performed
on two RC ties whilst the second was in-situ monitoring of a tunnel deformation. The
authors reported the presence of large performance discrepancies between the three ana-
Sensors 2021, 21, 1818 76 of 83
lyzed algorithms. Indeed, the classic methods (Spectral Shift Quality and Geometrical
Threshold Method) provided results with questionable quality despite their quick and
small computational taxing analysis. Instead, the novel but more computationally taxing
one (Polynomial Interpolation Comparison Method) provided much smoother and relia-
Interpolation Comparison Method) provided much smoother and reliable output plots as
ble output
evident plots as
in Figure 74.evident in Figure 74.

(a) (b)
Figure 74. DOFS extracted strain evolution in time: (a) before post-processing and (b) after post-processing with PICM—
Figure 74. DOFS
Polynomial extracted
Interpolation strain evolution
Comparison in time: (a) before post-processing and (b) after post-processing with PICM—
Method.
Polynomial Interpolation Comparison Method.
In Heinze and Echtermeyer [27], a novel post-processing methodology was intro-
ducedInasHeinze andtoEchtermeyer
a solution the issue of [27], a novel post-processing
“meaningless results” (SRAs) methodology
being reportedwas intro-
by OBR
duced aswhenever
analyses a solutionOFDR
to the DOFS
issue of “meaningless
pinching results” (SRAs)
or micro-bending beingThe
occurred. reported
latter by OBR
report-
analyses
edly is thewhenever
typical caseOFDR DOFSembedded
for DOFS pinching or micro-bending
inside hardening and occurred. The materials.
shrinking latter report-
In
edly is thethe
particular, typical
newcase
datafor DOFSmethod
analysis embedded inside“running
baptized hardeningreference
and shrinking
analysismaterials.
method”, In
suggested
particular,that
theeach
new measurement
data analysis should
methodbebaptized
compared to their reference
“running previous one and that
analysis their
method”,
strain differences
suggested should
that each be addedshould
measurement up tobethe absoluteto
compared strain
their value (theone
previous original
and that0 load
their
reference measurement).
strain differences should be added up to the absolute strain value (the original 0 load ref-
Finally,
erence Song et al. [173] argued that whenever BOTDA DOFS are employed for the
measurement).
detection of large
Finally, Songcracks
et al. inside any study
[174] argued thatmaterial,
whenever these lead by
BOTDA definition
DOFS to large Crack
are employed for the
Opening
detectionDisplacements
of large crackswhich
insidecananybe easily
study detected
material, in the
these fiber’s
lead output astostrain
by definition largepeaks
Crack
or singularities. Yet, when it comes to microcracks, their formation does not
Opening Displacements which can be easily detected in the fiber’s output as strain peaks bring along
any pronounced or
or singularities. even
Yet, visible
when peak. to
it comes Onmicrocracks,
the contrary,their
a microcrack-induced
formation does not peak could
bring be
along
easily lost in the midst of measurement noise due to its low signal-to-noise ratio. In order
to prevent this from happening and in return improve the crack detection, the authors
proposed a novel deep learning method trained and validated by means of an experimental
campaign performed on a 15 m-long wide-flange steel beam with artificial defects. The
authors concluded by reporting that Crack Opening Displacements as small as 23 microns
were successfully detected with the proposed method.

11. Conclusions
The present work was intended to introduce, inform and advise the readers on various
Distributed Optical Fiber Sensors (DOFS) deployment methodologies for Structural Health
Monitoring (SHM) purposes, i.e., the assessment of the residual ability of a structure to
continue serving its intended purpose. By collecting in a single article the most recent
efforts, advancements and findings on such DOFS SHM integrations the authors intend to
contribute to the goal of collective growth towards an efficient SHM.
The current work analyzed numerous DOFS applications such as laboratory experi-
mentation, the built environment (bridges, buildings, roads, etc.), geotechnical construc-
tions, tunnels, pipelines and wind turbines. Relevant specific conclusions are reported at
the end of each application section. A common conclusion shared by the largest portion
of the reported articles is the efficiency and performance of DOFS for the monitoring of
the structures under both serviceability and ultimate conditions. To a certain extent, this
Sensors 2021, 21, 1818 77 of 83

shows the instrumental maturity already achieved in some Civil Engineering areas. For
instance, in the case of concrete structures, it was shown how the technique is fully opera-
tive in the field of crack monitoring both in the controlled environment of a laboratory and
in-situ. Crack detection, mapping and width quantification was achieved both in bending
and shear. The technique’s efficiency was also fully demonstrated for the calculation of
deflections by the integration of measured curvatures and for the distribution of tempera-
tures inside the concrete. Furthermore, DOFS deployments were also shown to provide
relevant results in the monitoring of corrosion in the reinforcing steel. Overall, DOFS were
commendable SHM tools for the monitoring of geotechnical and underground structures
too (foundation piles, soil anchors and tunnels amongst others).
Multiple deployments techniques were described for each kind of structure and for
the varied sought-after results. For example, different deployment techniques were sug-
gested for the monitoring of cast-in-situ RC piles versus prefabricated ones or for the
cross-sectional versus longitudinal monitoring of a tunnel linen’s intrados or even for
the detection of steel pipe leaks versus axial deformations. Furthermore, the increasingly
numerous types of commercially available DOFS cables allowed for the surpassing of
the problem of fiber fragility and the need of protecting it during the concrete pouring
operations as well as increasing their resistance to the high temperatures that may ap-
pear during the concrete hardening process or in the case of their embedding into hot
asphalt pavements.
The increasingly larger number of applications in each of the reported fields showed a
growing interest and recognition of DOFS’ potential, yet, a large number of such publica-
tions called for a deeper understanding of the mechanical performance (short-term and
long-term) of the multi-layered system with which the fibers are bonded to a structure.
Indeed, the strain transfer from a host structure to a DOFS core was seen to depend on the
mechanical and geometrical properties of the different intermediate layers, i.e., coating,
jacket, reinforcements, adhesives, etc. Therefore, whilst extensive recommendations are
now available on optimal bonding/protective materials and on embedding procedures,
further work is required for the interpretation of the extracted strain profiles. Additionally,
more research is still needed on the DOFS long-term applicability. Indeed, few studies have
addressed their performance to fatigue and under cyclic loading, a condition that may
affect the behavior of the bonding agents and of the fiber itself. Durability aspects are also
under current investigation branching in two main directions: (1) the ability to monitor
material degradation (2) the degradation of the fiber and bonding agents themselves when
subject to harsh and aggressive environments (chloride and alkaline contamination, etc.).
Overall, though, with the already existent potential of DOFS, their number of routine
industrial applications for SHM is certain to increase.
Author Contributions: Conceptualization, M.F.B. and J.R.C.; Methodology, M.F.B. and J.R.C.; Writing-
Original Draft Preparation, M.F.B.; Writing—Review and Editing, M.F.B. and J.R.C. All authors have
read and agreed to the published version of the manuscript.
Funding: This research received no external funding.
Institutional Review Board Statement: Not applicable.
Informed Consent Statement: Not applicable.
Data Availability Statement: Not applicable.
Conflicts of Interest: The authors declare no conflict of interest.

References
1. Barrias, A.; Casas, J.R.; Villalba, S. A Review of Distributed Optical Fiber Sensors for Civil Engineering Applications. Sensors 2016,
16, 748. [CrossRef] [PubMed]
2. Brownjohn, J.M.W. Structural health monitoring of civil infrastructure. J. Philosofical Trans. A Math. Phys. Eng. Sci. 2006, 365,
589–622. [CrossRef] [PubMed]
3. Housner, G.W.; Bergman, L.A.; Caughey, T.K.; Chassiakos, A.G.; Claus, R.O.; Masri, S.F.; Skelton, R.E.; Soong, T.T.; Spencer, B.F.;
Yao, J.T.P. Structural control: Past, present and future. J. Eng. Mech. 1997, 123, 897–971. [CrossRef]
Sensors 2021, 21, 1818 78 of 83

4. Cawley, P. Structural health monitoring: Closing the gap between research and industrial deployment. Struct. Health Monit. 2018,
17, 1225–1244. [CrossRef]
5. Baker, M. Sensors Power Next-Generation SHM. Available online: https://www.sensorland.com/HowPage131.html (accessed
on 20 October 2020).
6. Glišić, B.; Hubbell, D.; Sigurdardottir, D.H.; Yao, Y. Damage detection and characterization using long-gauge and distributed
fiber optic sensors. Opt. Eng. 2013, 52, 087101. [CrossRef]
7. Seo, H. Monitoring of CFA pile test using three dimensional laser scanning and distributed fiber optic sensors. Opt. Lasers Eng.
2020, 130, 106089. [CrossRef]
8. Kurzweil, R. The Singularity Is Near; Palgrave Macmillan: London, UK, 2014.
9. Li, H.; Li, D.; Song, G. Recent applications of fiber optic sensors to health monitoring in civil engineering. Eng. Struct. 2004, 26,
1647–1657. [CrossRef]
10. Cantarero, A. Raman Scattering Applied to Materials Science. Procedia Mater. Sci. 2015, 9, 113–122. [CrossRef]
11. Gillibert, R.; Huang, J.Q.; Zhang, Y.; Fu, W.L.; de la Chapelle, M.L. Food quality control by Surface Enhanced Raman Scattering.
TrAC Trends Anal. Chem. 2018, 105, 185–190. [CrossRef]
12. Gillibert, R.; Huang, J.Q.; Zhang, Y.; Fu, W.L.; Lamy de la Chapelle, M. Explosive detection by Surface Enhanced Raman Scattering.
TrAC Trends Anal. Chem. 2018, 105, 166–172. [CrossRef]
13. Miah, K.; Potter, D.K. A review of hybrid fiber-optic distributed simultaneous vibration and temperature sensing technology and
its geophysical applications. Sensors 2017, 17, 2511. [CrossRef]
14. Schenato, L. A Review of Distributed Fibre Optic Sensors for Geo-Hydrological Applications. Appl. Sci. 2017, 7, 896. [CrossRef]
15. Soga, K.; Luo, L. Distributed fiber optics sensors for civil engineering infrastructure sensing. J. Struct. Integr. Maint. 2018, 3, 1–21.
[CrossRef]
16. Gue, C.Y.; Wilcock, M.; Alhaddad, M.M.; Elshafie, M.Z.E.B.; Soga, K.; Mair, R.J. The monitoring of an existing cast iron tunnel
with distributed fibre optic sensing (DFOS). J. Civ. Struct. Health Monit. 2015, 5, 573–586. [CrossRef]
17. Cheng, Q.; Tang, C.S.; Zhu, C.; Li, K.; Shi, B. Drying-induced soil shrinkage and desiccation cracking monitoring with distributed
optical fiber sensing technique. Bull. Eng. Geol. Environ. 2020, 79, 3959–3970. [CrossRef]
18. Gao, L.; Han, C.; Xu, Z.; Jin, Y.; Yan, J. Experimental Study on Deformation Monitoring of Bored Pile Based on BOTDR. Appl. Sci.
2019, 9, 2435. [CrossRef]
19. Wang, H.; Xiang, P.; Jiang, L. Strain transfer theory of industrialized optical fiber-based sensors in civil engineering: A review on
measurement accuracy, design and calibration. Sens. Actuators A Phys. 2019, 285, 414–426. [CrossRef]
20. Li, C.; He, W.; Luo, Y.; Zou, Y. Temperature Control Measurement of Bridge Foundation Concrete based on the Optical Fiber
Sensing Technology. J. Phys. Conf. Ser. 2019, 1288, 012082. [CrossRef]
21. Pei, H.; Yin, J.; Wang, Z. Monitoring and analysis of cast-in-place concrete bored piles adjacent to deep excavation by using
BOTDA sensing technology. J. Mod. Opt. 2019, 66, 703–709. [CrossRef]
22. Liu, J.; Song, Z.; Lu, Y.; Bai, Y.; Qian, W.; Kanungo, D.P.; Chen, Z.; Wang, Y. Monitoring of vertical deformation response to water
draining–recharging conditions using BOFDA-based distributed optical fiber sensors. Environ. Earth Sci. 2019, 78. [CrossRef]
23. Fan, L.; Tan, X.; Zhang, Q.; Meng, W.; Chen, G.; Bao, Y. Monitoring corrosion of steel bars in reinforced concrete based on helix
strains measured from a distributed fiber optic sensor. Eng. Struct. 2020, 204, 110039. [CrossRef]
24. Zhang, D.; Wang, J.; Zhang, P.; Shi, B. Internal strain monitoring for coal mining similarity model based on distributed fiber
optical sensing. Measurement 2017, 97, 234–241. [CrossRef]
25. Chai, J.; Lei, W.; Du, W.; Yuan, Q.; Zhu, L.; Zhang, D.; Li, H. Experimental study on distributed optical fiber sensing monitoring
for ground surface deformation in extra-thick coal seam mining under ultra-thick conglomerate. Opt. Fiber Technol. 2019, 53,
102006. [CrossRef]
26. Zhang, Z.; Guan, P.; Xu, J.; Wang, B.; Li, H. Horizontal Loading Performance of Offshore Wind Turbine Pile Foundation Based on
DPP-BOTDA. Appl. Sci. 2020, 10, 492. [CrossRef]
27. Heinze, S.; Echtermeyer, A.T. A Running Reference Analysis Method to Greatly Improve Optical Backscatter Reflectometry Strain
Data from the Inside of Hardening and Shrinking Materials. Appl. Sci. 2018, 8, 1137. [CrossRef]
28. Fernández-Ruiz, M.R.; Soto, M.A.; Williams, E.F.; Martin-Lopez, S.; Zhan, Z.; Gonzalez-Herraez, M.; Martins, H.F. Distributed
acoustic sensing for seismic activity monitoring. APL Photonics 2020, 5, 030901. [CrossRef]
29. Zuo, J.; Zhang, Y.; Zhao, Z.; Wei, X.; Wang, X.U. Pipeline Leak Detection Technology Based on Distributed Optical Fiber Acoustic
Sensing System. IEEE Access 2020, 8, 30789–30796. [CrossRef]
30. Merlo, S.; Malcovati, P.; Pavia, U.; Ferrata, V.; Norgia, M.; Pesatori, A.; Svelto, C.; Pniov, A.; Zhirnov, A.; Nesterov, E. Runways
Ground Monitoring System by Phase-Sensitive Optical-Fiber OTDR. In Proceedings of the 2017 IEEE International Workshop on
Metrology for AeroSpace (MetroAeroSpace), Padua, Italy, 21–23 June 2017; IEEE: Padova, Italy, 2017; pp. 523–529.
31. Ferdinand, P. The Evolution of Optical Fiber Sensors Technologies During the 35 Last Years and Their Applications in Structure
Health Monitoring. In Proceedings of the EWSHM-7th European Workshop on Structural Health Monitoring, Nantes, Frances,
8–11 July 2014.
32. Du, C.; Dutta, S.; Kurup, P.; Yu, T.; Wang, X. A review of railway infrastructure monitoring using fiber optic sensors. Sens.
Actuators A. Phys. 2020, 303, 111728. [CrossRef]
33. Udd, E. An overview of fiber-optic sensors. Rev. Sci. Instrum. 1995, 66, 4015–4030. [CrossRef]
Sensors 2021, 21, 1818 79 of 83

34. Kersey, A.D. A review of recent developments in fiber optic sensor technology. Opt. Fiber Technol. 1996, 2, 291–317. [CrossRef]
35. Grattan, K.T.V.; Sun, T. Fiber optic sensor technology: An overview. Sens. Actuators A Phys. 2000, 82, 40–61. [CrossRef]
36. InformationGatekeepers 2019 Distributed and Single Point Fiber Optic Sensing Systems Forecast. Available online: http:
//www.igigroup.com/st/pages/photonic_sensor_report.html (accessed on 20 October 2020).
37. Bao, X.; Chen, L. Recent Progress in Distributed Fiber Optic Sensors. Sensors 2012, 12, 8601–8639. [CrossRef] [PubMed]
38. Tennyson, R.C.; Mufti, A.A.; Rizkalla, S.; Tadros, G.; Benmokrane, B. Structural health monitoring of innovative bridges in Canada
with fiber optic sensors. Smart Mater. Struct. 2001, 10, 560. [CrossRef]
39. Whelan, M.P.; Albrecht, D.; Capsoni, A. Remote structural monitoring of the Cathedral of Como using an optical fiber Bragg
sensor system. In Proceedings of the Smart Structures and Materials and Nondestructive Evaluation for Health Monitoring and
Diagnostics, San Diego, CA, USA, 18–21 March 2002; Volume 4694, pp. 242–252.
40. Her, S.C.; Huang, C.Y. Effect of coating on the strain transfer of optical fiber sensors. Sensors 2011, 11, 6926–6941. [CrossRef]
[PubMed]
41. Feng, X.; Zhou, J.; Sun, C.; Zhang, X.; Ansari, F. Theoretical and Experimental Investigations into Crack Detection with BOTDR-
Distributed Fiber Optic Sensors. J. Eng. Mech. 2013, 139, 1797–1807. [CrossRef]
42. Alj, I.; Quiertant, M.; Khadour, A.; Grando, Q.; Terrade, B.; Renaud, J.C.; Benzarti, K. Experimental and numerical investigation on
the strain response of distributed optical fiber sensors bonded to concrete: Influence of the adhesive stiffness on crack monitoring
performance. Sensors 2020, 20, 5144. [CrossRef]
43. Henault, J.M.; Salin, J.; Moreau, G.; Quiertant, M.; Taillade, F.; Benzarti, K.; Delepine-Lesoille, S. Analysis of the strain transfer
mechanism between a truly distributed optical fiber sensor and the surrounding medium. In Proceedings of the 3rd International
Conference on Concrete Repair, Rehabilitation and Retrofitting, ICCRRR 2012, Cape Town, South Africa, 3–5 September 2012; pp.
733–739.
44. Bassil, A.; Chapeleau, X.; Leduc, D.; Abraham, O. Concrete Crack Monitoring Using a Novel Strain Transfer Model for Distributed
Fiber Optics Sensors. Sensors 2020, 20, 2220. [CrossRef]
45. Zhang, S.; Liu, H.; Cheng, J.; DeJong, M.J. A mechanical model to interpret distributed fiber optic strain measurement at
displacement discontinuities. Struct. Health Monit. 2020. [CrossRef]
46. Alj, I. Durability of Truly Distributed Optical Fiber Sensor in Cementitious Environment; Université Gustave Eiffel: Marne de la Vallée,
France, 2020.
47. Alj, I.; Quiertant, M.; Khadour, A.; Grando, Q.; Benzarti, K. Durability of distributed optical fiber sensors used for SHM of
reinforced concrete structures. In Proceedings of the 12th International Workshop on Structural Health Monitoring 2019: Enabling
Intelligent Life-Cycle Health Management for Industry Internet of Things (IIOT), Stanford, CA, USA, 10–12 September 2019; pp.
1696–1703.
48. Bado, M.F.; Casas, J.R.; Barrias, A. Performance of Rayleigh-Based Distributed Optical Fiber Sensors Bonded to Reinforcing Bars
in Bending. Sensors 2018, 20, 3125. [CrossRef]
49. Bado, M.F.; Casas, J.R.; Dey, A.; Berrocal, C.G. Distributed optical fiber sensing bonding techniques performance for embedment
inside reinforced concrete structures. Sensors 2020, 20, 5788. [CrossRef]
50. Barrias, A.; Casas, J.R.; Villalba, S. Distributed optical fibre sensors in concrete structures: Performance of bonding adhesives and
influence of spatial resolution. Struct. Control Health Monit. 2019, 26, 1–16. [CrossRef]
51. Barrias, A.; Casas, J.R.; Villalba, S. Fatigue performance of distributed optical fiber sensors in reinforced concrete elements. Constr.
Build. Mater. 2019, 218, 214–223. [CrossRef]
52. Zhang, S.; He, J.; Yu, Q.; Wu, X. Multi-scale load identi fi cation system based on distributed optical fi ber and local FBG-based
vibration sensors. Optik 2020, 219, 165159. [CrossRef]
53. McCormick, N.; Lord, J. Digital image correlation. Mater. Today 2010, 13, 52–54. [CrossRef]
54. Mata-Falcón, J.; Haefliger, S.; Lee, M.; Galkovski, T.; Gehri, N. Combined application of distributed fibre optical and digital image
correlation measurements to structural concrete experiments. Eng. Struct. 2020, 225. [CrossRef]
55. Bado, M.F.; Kaklauskas, G.; Casas, J.R. Performance of Distributed Optical Fiber Sensors (DOFS) and Digital Image Correlation
(DIC) in the monitoring of RC structures. IOP Conf. Ser. Mater. Sci. Eng. 2019, 615. [CrossRef]
56. Huo, L.; Cheng, H.; Kong, Q.; Chen, X. Bond-slip monitoring of concrete structures using smart sensors—A review. Sensors 2019,
19, 1231. [CrossRef] [PubMed]
57. Bassil, A. Distributed Fiber Optics Sensing for Crack Monitoring of Concrete Structures; Université de Nantes: Nantes, France, 2020.
58. Zhang, J. Distributed Sensing to Assess the Shear Behaviour of Reinforced Concrete Beams; Queen’s University: Kingston, ON, Canada,
2017.
59. Berrocal, C.G.; Fernandez, I.; Rempling, R. Crack monitoring in reinforced concrete beams by distributed optical fiber sensors.
Struct. Infrastruct. Eng. 2020. [CrossRef]
60. Berrocal, C.G.; Fernandez, I.; Bado, M.F.; Casas, J.R.; Rempling, R. Assessment and visualization of performance indicators of
reinforced concrete beams by distributed optical fiber sensing. Struct. Health Monit. 2021. [CrossRef]
61. Brault, A.; Hoult, N.A. Monitoring reinforced concrete serviceability performance using fiber-optic sensors. ACI Struct. J. 2019,
116, 57–70. [CrossRef]
62. Brault, A.; Hoult, N.A. Distributed Reinforcement Strains: Measurement and Application. ACI Struct. J. 2019, 116. [CrossRef]
Sensors 2021, 21, 1818 80 of 83

63. Poldon, J.J.; Hoult, N.A.; Bentz, E.C. Distributed Sensing in Large Reinforced Concrete Shear Test. ACI Struct. J. 2019, 116, 235–245.
[CrossRef]
64. Rodriguez, G.; Casas, J.R.; Villalba, S. Shear crack pattern identification in concrete elements via distributed optical fibre grid.
Struct. Infrastruct. Eng. 2019, 15, 1630–1648. [CrossRef]
65. Rodriguez, G.; Casas, J.R.; Villalba, S. Shear crack width assessment in concrete structures by 2D distributed optical fiber. Eng.
Struct. 2019, 195, 508–523. [CrossRef]
66. Fischer, O.; Thoma, S.; Crepaz, S. Distributed fiber optic sensing for crack detection in concrete structures. Civ. Eng. Des. 2019, 1,
97–105. [CrossRef]
67. Mehdi Mirzazadeh, M.; Green, M.F. Fiber Optic Sensors and Digital Image Correlation for Measuring Deformations in Reinforced
Concrete Beams. J. Bridg. Eng. 2018, 23, 04017144. [CrossRef]
68. Sawicki, B.; Bassil, A.; Brühwiler, E.; Chapeleau, X.; Leduc, D. Detection and measurement of matrix discontinuities in uhpfrc by
means of distributed fiber optics sensing. Sensors 2020, 20, 3883. [CrossRef] [PubMed]
69. Broth, Z.; Hoult, N.A. Dynamic distributed strain sensing to assess reinforced concrete behaviour. Eng. Struct. 2020, 204, 110036.
[CrossRef]
70. Barrias, A.; Casas, J.; Villalba, S. Embedded Distributed Optical Fiber Sensors in Reinforced Concrete Structures—A Case Study.
Sensors 2018, 18, 980. [CrossRef]
71. Bassil, A.; Wang, X.; Chapeleau, X.; Niederleithinger, E.; Abraham, O.; Leduc, D. Distributed fiber optics sensing and coda wave
interferometry techniques for damage monitoring in concrete structures. Sensors 2019, 19, 356. [CrossRef]
72. Wu, J.; Liu, H.; Yang, P.; Tang, B.; Wei, G. Quantitative strain measurement and crack opening estimate in concrete structures
based on OFDR technology. Opt. Fiber Technol. 2020, 60, 102354. [CrossRef]
73. Chen, Z.; Zheng, D.; Shen, J.; Qiu, J.; Liu, Y. Research on distributed optical-fiber monitoring of biaxial-bending structural
deformations. Measurement 2019, 140, 462–471. [CrossRef]
74. Malek, A.; Scott, A.; Pampanin, S.; Hoult, N.A. Postyield Bond Deterioration and Damage Assessment of RC Beams Using
Distributed Fiber-Optic Strain Sensing System. J. Struct. Eng. 2019, 145, 1–17. [CrossRef]
75. Sounthararajah, A.; Wong, L.; Nguyen, N.; Bui, H.H.; Kodikara, J. Evaluation of flexural behaviour of cemented pavement
material beams using distributed fibre optic sensors. Constr. Build. Mater. 2017, 156, 965–975. [CrossRef]
76. Ruiz, M.F.; Muttoni, A.; Gambarova, P.G. Analytical modeling of the pre- and postyield behavior of bond in reinforced concrete. J.
Struct. Eng. 2007, 133, 1364–1372. [CrossRef]
77. Bado, M.F.; Casas, J.R.; Kaklauskas, G. Distributed Sensing ( DOFS ) in Reinforced Concrete members for reinforcement strain
monitoring, crack detection and bond-slip calculation. Eng. Struct. 2021, 226, 111385. [CrossRef]
78. Cantone, R.; Fernández Ruiz, M.; Muttoni, A. A detailed view on the rebar–to–concrete interaction based on refined measurement
techniques. Eng. Struct. 2021, 226, 111332. [CrossRef]
79. Zhang, S.; Liu, H.; Coulibaly, A.A.S.; DeJong, M. Fiber optic sensing of concrete cracking and rebar deformation using several
types of cable. Struct. Control Health Monit. 2020, 1–23. [CrossRef]
80. Sieńko, R.; Zych, M.; Bednarski, Ł.; Howiacki, T. Strain and crack analysis within concrete members using distributed fibre optic
sensors. Struct. Health Monit. 2018, 18, 1510–1526. [CrossRef]
81. Davis, M.B.; Hoult, N.A.; Bajaj, S.; Bentz, E.C. Distributed Sensing for Shrinkage and Tension Stiffening Measurement. ACI Struct.
J. 2017, 114. [CrossRef]
82. Dey, A.; Vastrad, A.V.; Bado, M.F.; Sokolov, A.; Kaklauskas, G. Long-Term Concrete Shrinkage Influence on the Performance of
Reinforced Concrete Structures. Materials 2021, 14, 254. [CrossRef]
83. Fields, K.; Bischoff, P.H. Tension Stiffening and Cracking of High-Strength Reinforced Concrete Tension Members. ACI Mater. J.
2004, 101, 447–456.
84. Villalba, S.; Casas, J.R. Application of optical fiber distributed sensing to health monitoring of concrete structures. Mech. Syst.
Signal Process. 2013, 39, 441–451. [CrossRef]
85. Rodríguez, G.; Casas, J.R.; Villaba, S. Cracking assessment in concrete structures by distributed optical fiber. Smart Mater. Struct.
2015, 24. [CrossRef]
86. Nurmi, S.; Hoult, N.A.; Howell, S.D. Distributed strain monitoring of two-way slabs. Eng. Struct. 2019, 189, 580–588. [CrossRef]
87. Sienko, R.; Bednarski, L.; Howiacki, T. About Distributed Internal and Surface Strain Measurements within Prestressed Concrete
Truck Scale Platforms. IOP Conf. Ser. Mater. Sci. Eng. 2019, 471. [CrossRef]
88. Liao, W.; Zhuang, Y.; Zeng, C.; Deng, W.; Huang, J.; Ma, H. Fiber optic sensors enabled monitoring of thermal curling of concrete
pavement slab: Temperature, strain and inclination. Meas. J. Int. Meas. Confed. 2020, 165, 108203. [CrossRef]
89. Teguedy, M.C.; Joly-Lapalice, C.; Sorelli, L.; Conciatori, D. Optical fiber sensors implementation for monitoring the early-age
behavior of full-scale Timber-Concrete Composite slabs. Constr. Build. Mater. 2019, 226, 564–578. [CrossRef]
90. Davis, M.; Hoult, N.A.; Scott, A. Distributed strain sensing to assess corroded RC beams. Eng. Struct. 2017, 140, 473–482.
[CrossRef]
91. Davis, M.; Hoult, N.A.; Scott, A. Distributed strain sensing to determine the impact of corrosion on bond performance in
reinforced concrete. Constr. Build. Mater. 2016, 114, 481–491. [CrossRef]
92. 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]
Sensors 2021, 21, 1818 81 of 83

93. Fan, L.; Bao, Y.; Chen, G. Feasibility of distributed fiber optic sensor for corrosion monitoring of steel bars in reinforced concrete.
Sensors 2018, 18, 722. [CrossRef] [PubMed]
94. Fan, L.; Bao, Y.; Meng, W.; Chen, G. In-situ monitoring of corrosion-induced expansion and mass loss of steel bar in steel fiber
reinforced concrete using a distributed fiber optic sensor. Compos. Part B Eng. 2019, 165, 679–689. [CrossRef]
95. Bao, Y.; Hoehler, M.S.; Smith, C.M.; Bundy, M.; Chen, G. Temperature measurement and damage detection in concrete beams
exposed to fire using PPP-BOTDA based fiber optic sensors. Smart Mater. Struct. 2017, 26. [CrossRef] [PubMed]
96. Bao, Y.; Huang, Y.; Hoehler, M.S.; Chen, G. Review of fiber optic sensors for structural fire engineering. Sensors 2019, 19, 877.
[CrossRef] [PubMed]
97. Bao, Y.; Chen, Y.; Hoehler, M.S.; Smith, C.M.; Bundy, M.; Chen, G. Experimental Analysis of Steel Beams Subjected to Fire
Enhanced by Brillouin Scattering-Based Fiber Optic Sensor Data. J. Struct. Eng. 2017, 143, 04016143. [CrossRef]
98. Saidi, M.; Gabor, A. Experimental analysis of the tensile behaviour of textile reinforced cementitious matrix composites using
distributed fibre optic sensing (DFOS) technology. Constr. Build. Mater. 2020, 230, 117027. [CrossRef]
99. Jaafari, C.; Bertrand, D.; Guillot, T.; Prudhomme, E.; Tardif, N.; Georgin, J.F.; Delhomme, F.; Trunfio, R.; Chateur, N.; Bruyere,
E.; et al. Effect of early age drying shrinkage on the seismic response of RC structures. Mater. Struct. 2020, 53. [CrossRef]
100. Liu, T.; Huang, H.; Yang, Y. Crack Detection of Reinforced Concrete Member Using Rayleigh-Based Distributed Optic Fiber Strain
Sensing System. Adv. Civ. Eng. 2020, 2020. [CrossRef]
101. Woods, J.E.; Lau, D.T.; Bao, X.; Li, W. Measuring strain fields in FRP strengthened RC shear walls using a distributed fiber optic
sensor. Eng. Struct. 2017, 152, 359–369. [CrossRef]
102. (ASCE), A.S. of C.E. Infrastructure Report Card; The American Society of Civil Engineers: Reston, VA, USA, 2017.
103. Regier, R.; Hoult, N.A. Distributed Strain Behavior of a Reinforced Concrete Bridge: Case Study. J. Bridg. Eng. 2014, 19, 1–9.
[CrossRef]
104. Moore, M.; Phares, B.M.; Graybeal, B.; Rolander, D.; Washer, G.; Wiss, J. Reliability of Visual Inspection for Highway Bridges; TFHRC:
McLean, VA, USA, 2001; Volume I.
105. Gliši, B.; Posenato, D.; Inaudi, D. Integrity monitoring of old steel bridge using fiber optic distributed sensors based on Brillouin
scattering. In Proceedings of the Nondestructive Characterization for Composite Materials, Aerospace Engineering, Civil
Infrastructure, and Homeland Security 2007, San Diego, CA, USA, 20–22 March 2007.
106. Matta, F.; Bastianini, F.; Galati, N.; Casadei, P.; Nanni, A.; Asce, F. Distributed Strain Measurement in Steel Bridge with Fiber
Optic Sensors: Validation through Diagnostic Load Test. J. Perform. Constr. Facil. 2008, 22, 264–273. [CrossRef]
107. Webb, G.T.; Vardanega, P.J.; Hoult, N.A.; Fidler, P.R.A.; Bennett, P.J.; Middleton, C.R. Analysis of Fiber-Optic Strain-Monitoring
Data from a Prestressed Concrete Bridge. J. Bridg. Eng. 2017, 22, 1–14. [CrossRef]
108. CEN. (European Committee for Standardization), Eurocode 2: Design of Concrete Structures; CEN: Brussels, Belgium, 2004.
109. Collins, M.P.; Mitchell, D. Prestressed Concrete Structures; Response Publications, 1997.
110. Ye, C.; Butler, L.J.; Elshafie, M.Z.E.B.; Middleton, C.R. Evaluating prestress losses in a prestressed concrete girder railway bridge
using distributed and discrete fibre optic sensors. Constr. Build. Mater. 2020, 247, 118518. [CrossRef]
111. Barrias, A.; Rodriguez, G.; Casas, J.R.; Villalba, S. Application of distributed optical fiber sensors for the health monitoring of two
real structures in Barcelona. Struct. Infrastruct. Eng. 2018, 14, 967–985. [CrossRef]
112. Ouyang, J.; Chen, X.; Huangfu, Z.; Lu, C.; Huang, D.; Li, Y. Application of distributed temperature sensing for cracking control of
mass concrete. Constr. Build. Mater. 2019, 197, 778–791. [CrossRef]
113. de Battista, N.; Harvey, R.; Cheal, N. Distributed fibre optic sensor system to measure the progressive axial shortening of a
high-rise building during construction. In Proceedings of the 39th IABSE Symposium-Engineering the Future, Vancouver, BC,
Canada, 19–23 September 2017.
114. de Battista, N.; Cheal, N.; Harvey, R.; Kechavarzi, C. Monitoring the axial displacement of a high-rise building under construction
using embedded distributed fibre optic sensors. In Proceedings of the 8th International Conference on Structural Health
Monitoring of Intelligent Infrastructure, Brisbane, Australia, 5–8 December 2017.
115. Qi, J.; Wang, B.; Wang, X.; Song, Z.; Cui, Z. Application of optical-fiber sensing to concrete support and continuous wall strain
monitoring. In Proceedings of the The Second International Workshop on Environment and Geoscience, Hangzhou, China, 17–19
July 2021.
116. Broth, Z.E.; Hoult, N.A. Field Monitoring of RC-Structures under Dynamic Loading Using Distributed Fiber-Optic Sensors. J.
Perform. Constr. Facil. 2020, 34, 04020070. [CrossRef]
117. Brault, A.; Hoult, N.A.; Greenough, T.; Trudeau, I. Monitoring of Beams in an RC Building during a Load Test Using Distributed
Sensors. J. Perform. Constr. Facil. 2019, 33, 04018096. [CrossRef]
118. Wang, H.; Xiang, P.; Jiang, L. Optical Fiber Sensor Based In-Field Structural Performance Monitoring of Multilayered Asphalt
Pavement. J. Light. Technol. 2018, 36, 3624–3632. [CrossRef]
119. Rabaiotti, C.; Hauswirth, D.; Fischli, F.; Facchini, M. Structural health monitoring of airfield pavements using distributed
fiber-optics sensing. In Proceedings of the 4th Conf. on Smart Monitoring, Assessment and Rehabilitation of Civil Structures
(SMAR 2017), Zürich, Switzerland, 13–15 September 2017.
120. Wheeler, L.N.; Pannese, E.; Hoult, N.A.; Take, W.A.; Le, H. Transportation Geotechnics Measurement of distributed dy-namic rail
strains using a Rayleigh backscatter based fiber optic sensor: Lab and field evaluation. Transp. Geotech. 2018, 14, 70–80. [CrossRef]
Sensors 2021, 21, 1818 82 of 83

121. Sasi, D.; Philip, S.; David, R.; Swathi, J. A review on structural health monitoring of railroad track structures using fiber optic
sensors. Mater. Today Proc. 2020. [CrossRef]
122. Zhu, H.; Shi, B.; Zhang, C. FBG-Based Monitoring of Geohazards: Current. Sensors 2017, 17, 452. [CrossRef] [PubMed]
123. Sienko, R.; Bednarski, L.; Kanty, P.; Howiacki, T. Application of Distributed Optical Fibre Sensor for Strain and Temperature
Monitoring within Continuous Flight Auger Columns. IOP Conf. Ser. Earth Environ. Sci. 2019, 221, 012006. [CrossRef]
124. Kania, J.G.; Sorensen, K.K.; Fellenius, B.H. Application of Distributed Fibre Optic Cables in Piles Application of Distributed Fibre
Optic Cables in Piles. Geotech. Eng. J. Seags Agssea 2020, 51, 3.
125. Cheng, L.S.; Pin, T.B.; Fai, C.M.; Mohamad, H.; An, A.K.; Rahardjo, P.P. Application of Distributed Fibre Optic Sensor (DFOS) in
Bi-directional Static Pile Load Tests. Geotech. Eng. J. Seagc 2018. Available online: https://smartsensing.com.my/wp-content/
uploads/2018/12/Application-of-Distributed-Fibre-Optic-Sensor-DFOS-in-Bi-directional-Static-Pile-Load-Tests.pdf (accessed
on 5 March 2021).
126. Xu, D.; Yin, J.; Liu, H. A new measurement approach for deflection monitoring of large-scale bored piles using distributed fiber
sensing technology. Measurement 2018, 117, 444–454. [CrossRef]
127. Guo, Z.; Zhao, Z. Application of distributed optical fiber sensing technique in pile foundation monitoring Application of
distributed optical fiber sensing technique in pile foundation monitoring. IOP Conf. Ser. Earth Environ. Sci. 2018, 189, 052074.
[CrossRef]
128. Liu, J.; Wang, Y.; Lu, Y.; Wei, J.; Kanungo, D.P. Application of distributed optical fiber sensing technique in monitoring the ground
deformation. J. Sensors 2017, 2017. [CrossRef]
129. Galloway, D.L.; Burbey, T.J. Review: Regional land subsidence accompanying groundwater extraction. Hydrogeol. J. 2011, 19,
1459–1486. [CrossRef]
130. Wu, J.; Shi, B.; Cao, D.; Jiang, H.; Wang, X.; Gu, K. Model test of soil deformation response to draining-recharging conditions
based on DFOS. Eng. Geol. 2017, 226, 107–121. [CrossRef]
131. Zhang, C.C.; Shi, B.; Gu, K.; Liu, S.P.; Wu, J.H.; Zhang, S.; Zhang, L.; Jiang, H.T.; Wei, G.Q. Vertically Distributed Sensing of
Deformation Using Fiber Optic Sensing. Geophys. Res. Lett. 2018, 45, 11732–11741. [CrossRef]
132. Wu, H.; Zhu, H.H.; Zhang, C.C.; Zhou, G.Y.; Zhu, B.; Zhang, W.; Azarafza, M. Strain integration-based soil shear displacement
measurement using high-resolution strain sensing technology. Measurement 2020, 166, 108210. [CrossRef]
133. Wang, T.; Xu, B.; He, B.; Du, M. Research on Application of Distributed Optical Fiber Sensing Technology in the Safety Monitoring
of Pipeline Transportation. In Proceedings of the 7th International Conference on Energy, Environment and Sustainable
Development (ICEESD 2018), Shenzhen, China, 30–31 March 2018.
134. Zheng, Y.; Zhu, Z.W.; Li, W.J.; Gu, D.M.; Xiao, W. Experimental research on a novel optic fiber sensor based on OTDR for landslide
monitoring. Meas. J. Int. Meas. Confed. 2019, 148, 106926. [CrossRef]
135. Schenato, L.; Camporese, M.; Bersan, S.; Cola, S.; Galtarossa, A.; Pasuto, A.; Simonini, P.; Salandin, P.; Palmieri, L. High density
distributed strain sensing of landslide in large scale physical model. In Proceedings of the 25th International Conference on
Optical Fiber Sensors, Jeju, Korea, 24–28 April 2017.
136. Zhang, D.; He, J.; Xue, Y.; Xu, J.; Xu, X. Investigation of settlement monitoring method based on distributed Brillouin fiber optical
sensor. Measurement 2019, 134, 118–122. [CrossRef]
137. Chen, H.; He, J.; Xue, Y.; Zhang, S. Experimental study on sinkhole collapse monitoring based on distributed Brillouin optical
fiber sensor. Optik 2020, 216, 164825. [CrossRef]
138. Xu, J.; He, J.; Zhang, L. Collapse prediction of karst sinkhole via distributed Brillouin optical fiber sensor. Measurement 2017, 100,
68–71. [CrossRef]
139. Cola, S.; Schenato, L.; Brezzi, L.; Pangop, F.C.T.; Palmieri, L.; Bisson, A. Composite anchors for slope stabilisation: Monitoring of
their in-situ behaviour with optical fibre. Geosciences 2019, 9, 240. [CrossRef]
140. Lienhart, W.; Monsberger, C.M.; Kalenjuk, S.; Woschitz, H. High Resolution Monitoring of Retaining Walls with Distributed Fibre
Optic Sensors and Mobile Mapping Systems 2. New Monitoring Approaches. In Proceedings of the Asian Pacific Workshop
Structural health Monitoring APWSHM 2018, Hong Kong, China, 12–15 November 2018.
141. Zhang, P.; Sun, B. Distribution characteristics of the advance abutment pressure in a deep stope. J. Geophys. Eng. 2020, 686–699.
[CrossRef]
142. Lanciano, C.; Salvini, R. Monitoring of strain and temperature in an open pit using brillouin distributed optical fiber sensors.
Sensors 2020, 20, 1924. [CrossRef] [PubMed]
143. Hong, C.-Y.; Zhang, Y.-F.; Li, G.-W.; Zhang, M.-X.; Liu, Z.-X. Physical Recent progress of using Brillouin distributed fiber optic
sensors for geotechnical health monitoring. Sens. Actuators A Phys. 2017, 258, 131–145. [CrossRef]
144. Loupos, K.; Kanellos, G.; Bimpas, M.; Amditis, A.; Bursi, O.; Orfanoudakis, A. Fiber Sensors Based System for Tunnel Linings ’
Structural Health Monitoring. In Proceedings of the 2nd Conference On Smart Monitoring, Assessment And Rehabilitation of
Civil Structures, Istanbul, Turkey, 9–11 September 2013; pp. 9–11.
145. Inaudi, D.; Walder, R. Full-Length Tunnel Structural Monitoring. In Proceedings of the Structural Health Monitoring, Stanford,
CA, USA, 10–12 September 2019.
146. Fajkus, M.; Nedoma, J.; Mec, P.; Hrubesova, E.; Martinek, R.; Vasinek, V. Analysis of the highway tunnels monitoring using an
optical fiber implemented into primary lining. J. Electr. Eng. 2017, 68, 364–370. [CrossRef]
Sensors 2021, 21, 1818 83 of 83

147. Monsberger, C.M.; Lienhart, W.; Kluckner, A. Continuous strain measurements in a shotcrete tunnel lining using distributed fibre
optic sensing. In Proceedings of the 9th European Workshop on Structural Health Monitoring, Manchester, UK, 10–13 July 2018;
pp. 1–13.
148. Shi, B.I.N.; Xu, H.; Chen, B.I.N.; Zhang, D.A.N.; Ding, Y.; Gao, J.; Shi, B.I.N. A Feasibility Study on the Application of Fiber- Optic
Distributed Sensors for Strain Measurement in the Taiwan Strait Tunnel Project. Mar. Georesources Geotechnol. 2010, 25, 575–586.
[CrossRef]
149. Li, Z.; Hou, G.; Hu, T.; Zhou, T.; Xiao, H. A study on the application of the distributed optical fiber sensing monitoring technology
in the process of dismantling temporary tunnel shoring. Arab. J. Geosci. 2020, 13, 1–11. [CrossRef]
150. Gómez, J.; Casas, J.R.; Villalba, S. Structural Health Monitoring with Distributed Optical Fiber Sensors of tunnel lining affected by
nearby construction activity. Autom. Constr. 2020, 117, 103261. [CrossRef]
151. Bado, M.F.; Casas, J.R.; Gómez, J. Post-processing algorithms for distributed optical fiber sensing in structural health monitoring
applications. Struct. Health Monit. 2020. [CrossRef]
152. Minardo, A.; Catalano, E.; Coscetta, A.; Zeni, G.; Zhang, L.; Di Maio, C.; Vassallo, R.; Coviello, R.; Macchia, G.; Picarelli, L.; et al.
Distributed Fiber Optic Sensors for the Monitoring of a Tunnel Crossing a Landslide. Remote Sens. 2018, 10, 1291. [CrossRef]
153. Zhao, J.; Feng, J.; Ren, L. Inversion Calculation Analysis of Operational Tunnel Structure Based on the Distributed Optical-Fiber
Sensing System. Adv. Civ. Eng. 2017, 2017. [CrossRef]
154. Piccolo, A.; Lecieux, Y.; Delepine-Lesoille, S.; Leduc, D.; Bumbieler, F.; Teixeira, P.; Zghondi, J. Tunnel convergence analysis
by distributed optical fiber strain sensing with means of finite element—Inverse analysis method. In Proceedings of the
EWSHM-European Workshop on Structural Health Monitoring, Manchester, UK, 10–13 July 2018; pp. 1–12.
155. Glisic, B.; Yao, Y. Fiber optic method for health assessment of pipelines subjected to earthquake-induced ground movement.
Struct. Health Monit. 2012, 11, 696–711. [CrossRef]
156. Inaudi, D.; Glisic, B. Distributed Fiber optic Strain and Temperature Sensing for Structural Health Monitoring. In Proceedings
of the Proceedings of the 3rd International Conference on Bridge Maintenance, Safety and Management—Bridge Maintenance,
Safety, Management, Life-Cycle Performance and Cost, Porto, Portugal, 16–19 July 2006; p. 964.
157. Wong, L.; Rathnayaka, S.; Chiu, W.K.; Kodikara, J. Fatigue Damage Monitoring of a Cast Iron Pipeline Using Distributed Optical
Fibre Sensors. Procedia Eng. 2017, 188, 293–300. [CrossRef]
158. Ren, L.; Jiang, T.; Jia, Z.; Li, D.; Yuan, C.; Li, H. Pipeline corrosion and leakage monitoring based on the distributed optical fi ber
sensing technology. Measurement 2018, 122, 57–65. [CrossRef]
159. Stajanca, P.; Chruscicki, S.; Homann, T.; Seifert, S.; Schmidt, D. Detection of Leak-Induced Pipeline Vibrations Using Fiber—Optic
Distributed Acoustic Sensing. Sensors 2018, 18, 2841. [CrossRef]
160. Feng, X.; Wu, W.; Meng, D.; Ansari, F.; Zhou, J. Distributed monitoring method for upheaval buckling in subsea pipelines with
Brillouin optical time-domain analysis sensors. Adv. Struct. Eng. 2017, 20, 180–190. [CrossRef]
161. Zhang, S.; Liu, B.; He, J. Pipeline deformation monitoring using distributed fiber optical sensor. Measurement 2019, 133, 208–213.
[CrossRef]
162. Inaudi, D. Distributed optical fiber sensors for strain and deformation monitoring of pipelines and penstocks. Geotech. Eng. 2019,
50, 64–70.
163. WWEA. Wind Power Capacity Worldwide Reaches 597 GW, 50,1 GW Added in 2018. Available online: https://wwindea.org/
blog/2019/02/25/wind-power-capacity-worldwide-reaches-600-gw-539-gw-added-in-2018/ (accessed on 18 November 2020).
164. Coscetta, A.; Minardo, A.; Olivares, L.; Mirabile, M.; Longo, M.; Damiano, M.; Zeni, L. Wind Turbine Blade Monitoring with
Brillouin-Based Fiber-Optic Sensors. J. Sens. 2017, 2017. [CrossRef]
165. Raman, V.; Drissi-habti, M.; Limje, P.; Khadour, A. Finer SHM-Coverage of Inter-Plies and Bondings in Smart Composite by Dual
Sinusoidal Placed Distributed Optical Fiber Sensors. Sensors 2019, 19, 742. [CrossRef]
166. Chesnoy, J. Undersea Fiber Communication Systems; Academic Press: Cambridge, MA, USA, 2015.
167. Hartog, A.H.; Belal, M.; Clare, M.A. Advances in Distributed Fiber-Optic Sensing for Monitoring Marine Infrastructure, Measuring
the Deep Ocean, and Quantifying the Risks Posed by Sea fl oor Hazards. Mar. Technol. Soc. J. 2018, 52, 58–73. [CrossRef]
168. Huang, L.; Li, Y.; Chen, S.; Zhang, Q.; Song, Y.; Zhang, J.; Wang, M. Optical Fiber Technology Building safety monitoring based
on extreme gradient boosting in distributed optical fi ber sensing. Opt. Fiber Technol. 2020, 55, 102149. [CrossRef]
169. Liu, H.; Ma, J.; Yan, W.; Liu, W. Traffic Flow Detection Using Distributed Fiber Optic Acoustic Sensing. IEEE Access 2018, 6,
68968–68980. [CrossRef]
170. Li, M.; Li, S.; Stone, J.S. Novel Optical Fibers for Distributed Sensor Applications. In Proceedings of the 25th Optical Fiber Sensors
Conference (OFS), Jeju, Korea, 24–28 April 2017; pp. 1–4.
171. Bao, X.; Webb, D.J.; Jackson, D.A. Combined distributed temperature and strain sensor based on Brillouin loss in an optical fiber.
Opt. Lett. 1994, 19, 141–143. [CrossRef] [PubMed]
172. You, R.; Ren, L.; Song, G. A novel OFDR-based distributed optical fiber sensing tape: Design, optimization, calibration and
application. Smart Mater. Struct. 2020, 29. [CrossRef]
173. Song, Q.; Zhang, C.; Tang, G.; Ansari, F. Deep learning method for detection of structural microcracks by brillouin scattering
based distributed optical fiber sensors. Smart Mater. Struct. 2020, 29. [CrossRef]