You are on page 1of 5

See

discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/252874906

Fiber Bragg grating system for continuous large-


scale monitoring of convergence in Rossio
Tunnel

Article in Proceedings of SPIE - The International Society for Optical Engineering October 2009
DOI: 10.1117/12.835420

CITATIONS READS

3 116

8 authors, including:

Luis Ferreira Francisco Arajo


HBM FiberSensing, S.A. Institute for Systems and Computer Engineerin
155 PUBLICATIONS 2,039 CITATIONS 150 PUBLICATIONS 1,812 CITATIONS

SEE PROFILE SEE PROFILE

Lus P. C. Costa Gonalves


EPOS, SA
7 PUBLICATIONS 3 CITATIONS

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

PROTEU View project

Prospeco mineral no Polgono Albernoa (Faixa Piritosa Ibrica, Portugal) View project

All content following this page was uploaded by Luis Ferreira on 24 August 2014.

The user has requested enhancement of the downloaded file. All in-text references underlined in blue are added to the original document
and are linked to publications on ResearchGate, letting you access and read them immediately.
Fiber Bragg grating system for continuous large-scale monitoring of
convergence in Rossio Tunnel
Cristina B. Barbosaa, Lus A. Ferreiraa,b, Francisco M. Arajoa,b, Lus Gonalvesc, Carlos D. Gamad,
Rui Malvae, ngela Silvae, Vitor Freitase

a
FiberSensing, S.A, Rua Vasconcelos Costa, 277, 4470-640 Maia, Portugal
b
INESC Porto, Rua Dr. Roberto Frias, 378, 4200 - 465 Porto, Portugal
c
EPOS Empresa Portuguesa de Obras Subterrneas, Lda, Lagoas Park, Ed1, 2740-264 Porto Salvo,
Portugal
d
Geotechnical Center, Instituto Superior Tcnico, Av. Rovisco Pais, 1049-001 Lisboa, Portugal
e
REFER EPE Departamento de Tneis, Av. D. Joo II, lote 1.18.0, bloco A, 1049-001 Lisboa,
Portugal

ABSTRACT

On this paper, the implemented solution for continuous monitoring of Rossio Train Tunnel in Lisbon (Portugal) is
described. The monitoring system is a complete solution that comprises measurements of strain and temperature with
more than 850 fiber Bragg grating sensors, data acquisition, processing, storage and easy access through a web platform.
The applied sensing technology has several advantages such as fiber Bragg sensors being immune to electrical
interferences and suited to harsh environments. The used method for convergence monitoring (MEMCOT) makes it
possible to determine tunnel convergences based on strain measurements around the tunnel contour. An optoelectronic
measurement unit and optical switch are deployed at the entrance of the tunnel and remotely connected to a server that
saves and displays information to authorized users in web interface.

Keywords: Fiber Bragg Sensors; Convergence; Tunnel; Continuous Monitoring; Monitoring System.

1 INTRODUCTION

Rossio Tunnel is a centenary train tunnel between Rossio and Campolide Train Stations in Lisbon that has recently been
subjected to a major rehabilitation. Since this is a critical infrastructure, an automated sensing system for continuous
monitoring of the tunnel structural behavior during its service life was installed. The system used an innovative method
for determining the tunnel convergence based on measurements of strain along the section contour. The Extensometric
Method for Monitoring Convergences in Tunnels1 (MEMCOT) is suited for continuous monitoring of the structural
stability during construction as well as during regular service of common tunnels. The 2.6 km tunnel is instrumented on
109 sections, reaching a total of 872 fiber Bragg grating based sensors interrogated every day. This sensing technology
was preferred to the conventional strain gauges for several reasons, going from total cost to technical characteristics
(EMI/RFI immunity, signal integrity even on long transmission distances, and suitability to harsh environments1).
Moreover, this system reduces only a few centimeters to the tunnel clearance gauge.

2 MONITORING SYSTEM

2.1 The method

Traditionally, assessment of tunnels and underground structures safety consists only in measuring displacements of the
contour that may indicate the collapse. This is normally preformed using mechanical or optical methods such as

20th International Conference on Optical Fibre Sensors, edited by Julian Jones, Brian Culshaw,
Wolfgang Ecke, Jos Miguel Lpez-Higuera, Reinhardt Willsch, Proc. of SPIE Vol. 7503, 75035S
2009 SPIE CCC code: 0277-786X/09/$18 doi: 10.1117/12.835420

Proc. of SPIE Vol. 7503 75035S-1


extensometric wires or total station surveys. Yet, these operations are seldom and require the interruption of the normal
operation of the tunnel or of the construction works.

The Extensometric Method for Monitoring Convergences in Tunnels (MEMCOT) continuously assesses the
displacement of the tunnel support by measuring deformations along the contour. It allows the quantification of
convergences (relative displacements) of the support and its geometric evolution throughout time1. Since MEMCOT
measures absolute displacements of each point, geometrical changes can be detected, whereas with conventional
convergence measurements one have to assume that movements are symmetric. The mathematical model is based on the
Theory of Bending Initially Curved Bars, which allows the quantitative modeling of radial displacement using: the axial
strain that is being measured; the geometrical shape of the bar (its radius); and the distance from the point to the neutral
axis2. If a beam with an initial radius R1 is submitted to an external load or moment, it bends to create an axial strain x at
a section located at a distance y from the neutral axis. Consider the curved bar unloaded and subjected to bending in
Figure 1. The strain at a small segment CD, CD, at a distance y from the neutral axis is given by equation (1). Since there
is no change in length in an element on the neutral axis (2), one can find the radial displacement using equation (3).

l CD (R 2 + y ) (R 1 + y )
CD = = (1)
l CD (R1 + y )
R 2 = R 1 (2)
x R 1 (y + R 1 )
= R1 R 2 = (3)
x R1 + y(1 + x )

For determining convergence of several points along a tunnel contour, MEMCOT uses strain gauges deployed along the
perimeter of the tunnel section (Figure 2). Having each points radial displacement one can find several variables,
namely the new coordinates, the tunnel deformation and hence the convergence between two points.



R2

A B A B
y
y

C D C D
Figure 1 Bending of bars. Figure 2 Sensor position on section. Figure 3 Production.

2.2 Sensing system

On Rossio Tunnel fiber Bragg grating sensors were used as strain gauges. These are well-known microstructures of small
dimensions obtained photo-inducing periodic refraction-index changes in the core of an optical fiber. Such modulation
makes this microstructure a wavelength selective mirror reflecting a narrow band of an incoming broadband spectrum
that is centered at the Bragg wavelength. The grating wavelength variation is essentially dependant on temperature and
mechanical stress3. Fiber Bragg sensors have high sensitivity, reduced dimensions and small weight. On the Rossio
Tunnel this is an important feature for it makes it possible to monitor each section with marginal clearance gauge
reduction. Bragg sensors are also immune to EMI/RFI hence suitable for use near train catenaries, wheras electrical
sensors would suffer from interferences. The fiber Bragg grating multiplexing capability and self reference
characteristics allow the connection of many sensors in series on a single fiber and with a considerably smaller amount
of cables4,5. In Rossio Tunnel, each optical channel addresses eight sensors in series that correspond to one profile
seven strain sensors and one temperature sensor for thermal compensation.

In this project, weldable fiber Bragg strain and temperature sensors, developed by FiberSensing, have been used6. These
sensors have been installed in stainless steel bars with 40x4 mm section so that the tunnel contour could be defined as a

Proc. of SPIE Vol. 7503 75035S-2


continuous perimeter. Since the tunnel perimeter is around 19 m, there was the need to divide the complete profile in
shorter lengths for transport, storage, production and handling purposes, meaning that every bar is an array of sensors
which is prepared with a connector to be joined inside a protection box (Figure 3).

3 VALIDATION, CALIBRATION AND PREPARATION.

In order to calibrate the method and also to test the


Real Scale Model - Sensor S0
system, a real scale model has been implemented.
8000 The profiles correspond to the main existing cross-
Concrete: y = 0.9722x - 316.17; R2 = 0.9733
7000 sections in the Rossio Tunnel: the masonry section
Masonry: y = 0.9021x - 315.08; R2 = 0.9746
and the concrete section. Tests consisted on the
6000
application of concentrated load to each model.
Method (10-6 m)

5000 During each load-unloading cycle, strain was


4000 measured simultaneously with relative displacements
of the seven instrumented points. For signal
3000
acquisition, a portable BraggMETER was used. The
2000 real displacement was measured using a total station
Masonry Model
1000 Concrete Model
and reflecting targets fixed on the sensors location.
0
From test results, a straight relationship between the
0 1000 2000 3000 4000 5000 6000 7000 8000
measured relative displacement and the weldable
-6
Total Station (10 m)
sensor readings was observed. On Figure 4 the result
Figure 4 Test results. of vertical loading on the models is presented,
showing a linear correlation coefficient near one. For
these two tests, one can calculate the maximum difference between the expected value (displacement given by the total
station) and the value calculated with MEMCOT using sensor readings. For all seven sensors, this difference was always
smaller than 2.0 mm and the average value of the difference is below 1.0 mm (0.7 mm and 0.6 mm for Masonry and
Concrete models, respectively), which is a result in line with the accuracy of the total station itself (1 mm).

4 INSTALLATION

The system installation was performed during the final period of the
tunnel rehabilitation. After marking the verticality of the defined
sections, by means of radial lasers, the three stainless steel instrumented
bars were raised, hanged and screw-connected one to each other. The
non-permanent optical connection between the three instrumented bars
was systematically controlled. Whenever significant signal losses were
detected, the connection would be replaced by a fusion splice. This
operation was performed in approximately 8 % of these connections.
After optical connection and sensor verification, the instrumented profile
was fixed onto the tunnel surface by means of an appropriated epoxy
resin, chosen for having suitable properties in terms of tensile strength,
pot-life and curing time. The resin was applied on the metallic profile,
which was then pressed against the tunnel surface by jacks located on its
base. The profile was afterwards anchored on the base. Some pictures of
the installation are shown in Figure 5. Since MEMCOT needs an initial
calibration to determine the coordinates of the sensors, a topographic Figure 5 Installation
survey was performed after installation. The sensors connection to the
measurement unit was made through 8 optical cables with sixteen fibers
each. The cables were deployed along the technical gutter while one fiber was pulled out and connected to the
corresponding sensor branch near each instrumented section. At the Campolide end of the tunnel the technical cabinet
with the measurement unit and the multiplexer was installed. The multiplexer is an optical switch that makes it possible

Proc. of SPIE Vol. 7503 75035S-3


to address each optical channel at a time. A complete turn making one acquisition in each channel takes approximately 8
minutes (averages included).

At the end of the installation, some anomalies were detected in the readings of 13 sections, which were repaired in place.
The majority of these situations (77%) corresponded to malfunction caused by the construction works that were yet
taking place. The other 23% were common problems generated during installation.

5 MONITORING RESULTS

The information gathered by the measurement unit on all sensors is transmitted throughout the internet to a dedicated
server installed on REFER control room. The Data Server application was developed in order to receive, process and
save the acquired data. The mathematical algorithm (MEMCOT) is subroutined in LabVIEW so that every wavelength
shift is transformed into convergence and other relevant variables. The developed application is also capable of detecting
and treating alarms according to previous user definitions.

All the information is presented in web format to authorized users. Online, the user can search for results in various
combinations: search from the tunnel plant, from the section or sensor name, correlate sensors from different sections,
search for aligned sensors on a certain extent of the tunnel, and define My Search for faster searches. The results are
then showed as a graph and/or as a table that can be saved as a .txt file format. Since every variable is a function of time,
it is also possible to have information and graphical representation of its velocity and acceleration. When searching on
sections, data can be shown as the deformation of the tunnel contour.

6 CONCLUSION

The successful conception, implementation and installation of a large scale fiber optic convergence monitoring system
along the 2.6 km of the Rossio Tunnel were demonstrated. The system, which has no active elements inside the tunnel,
requires only a few centimeters of the tunnel clearance gauge and provides, in a continuous and automated way, remote
information about the tunnel structural evolution. This information is fundamental for assessing the health of the
structure and can be used for better programming maintenance actions. The system also generates alarms every time a
singular event occurs allowing its localization, characterization and surveillance. It has been proved that the combination
of the convergence method (MEMCOT) with the fiber Bragg sensing technology is a valid and competitive solution for
long term monitoring of tunnels.

7 REFERENCES

1. Gama, C. Dinis, 2004, Mtodo Extensomtrico para Medir Deformaes e Convergncias em Tneis, nas Fases de
Construo e de Explorao, Patente n 103.058, Instituto Nacional da Propriedade Industrial, Lisboa - Portugal
2. Gama, C. Dinis, 2006, A Continuous tunnel monitoring system for both construction and service stages, USRMS,
Golden Rocks 2006, Golden Colorado - USA
3. Othonos, A, Kalli, K, 1999, Fiber Bragg gratings: fundamentals and applications in telecommunications and sensing,
Artech House Inc, Norwood.
4. Ansari, F, 2005, Sensing issues in civil structural health monitoring, Springer, Dordrecht.
5. Lpez-Higuera, J M, 2002, "Handbook of optical fibre sensing technology, John Wiley & Sons, Chichester.
6. C. Barbosa, N. Costa, L. A. Ferreira, F. M. Arajo, H. Varum, A. Costa, Monitoring the new circular pedestrian
steel bridge over the So Roque and Botires channels with weldable fibre-Bragg grating sensors, 18th Int.
Conference on Optical Fibre Sensors, Cancun, 23-27 October 2006.

Proc. of SPIE Vol. 7503 75035S-4

View publication stats

You might also like