MoDOT

Research, Development and Technology

University of Missouri-Rolla RDT 05-003

In-situ Load Testing

of Bridge A6358 (Part 1)
and of Bridge A6101
and Bridge A6102 (Part 2)

RI 04-006

February, 2005
 TECHNICAL REPORT DOCUMENTATION PAGE

1. Report No. 2. Government 3. Recipient’s Catalog No.

RDT05-003 Accession No.

4. Title and Subtitle 5. Report Date

In-situ load testing of Bridge A6358 Osage Beach, MO February 2005
6. Performing Organization Code
UMR
7. Author/s 8. Performing Organization Report
F. Matta, F. Bastianini, N. Galati, P. Casadei, and A. Nanni No. RI04-006
9. Performing Organization Name and Address 10. Work Unit No. (TRAIS)
Center for Infrastructure Engineering Studies
University of Missouri - Rolla
223 ERL 11. Contract or Grant No.
Rolla, MO 65409
12. Sponsoring Organization Name and Address 13. Type of report and period covered
MODOT Research Development and Technology
1617 Missouri Blvd., Jefferson City, MO 65109
University Transportation Center 14. Sponsoring Agency Code
223 Engineering Research Lab., Rolla, MO 65409 MoDOT
15. Supplementary Notes
This investigation was conducted in cooperation with the U.S. Department of Transportation.
16. Abstract
This report presents the assessment of a High Performance Steel (HPS) bridge located at the Lake of the Ozarks
in Miller County, MO. The bridge number is A6358 and it is located on US Rt. 54/Osage River. Two different
modern techniques for Structural Health Monitoring (SHM) were applied for this purpose, in order to advance
their field validation process.
Displacements of the girders during the pouring of the concrete deck and a diagnostic load test, conducted after
the bridge completion, were evaluated using an optical non-contact measurement technique based on an
automated total station system. The setup included a total of 22 reflecting prisms mounted onto the bottom flange
of the girders, plus four reference targets placed outside the bridge superstructure.
Strain and temperature distributions along some of the girders were measured during the diagnostic load test using
a fiber optic distributed Brillouin sensing technique. The method is based on the “Brillouin” phenomenon, which
defines a linear relation between the axial strain in an optical fiber, and the corresponding shift in the frequency
distribution of the scattered Brillouin light generated when an optical impulse runs through a fiber. A single
optical circuit, comprised of bare fibers and a custom-made novel glass FRP tape with embedded sensing fibers,
was installed on the web of the girders at different depths, along two continuous spans.
Experimental deflection and strain data were verified to be consistent with each other and allowed to evaluate the
actual girder distribution factors. The comparison between experimental and theoretical results based on the
AASHTO LRFD Bridge Design Specifications (AASHTO, 1998) allowed establishing the safety of the structure,
although pointing out a significant drop in the expected behavior of one of the external girders, which may call for
further assessment.
17. Key Words 18. Distribution Statement
Assessment, bridge monitoring, concrete No restrictions. This document is available to the public through
deck, fiber optics, high performance steel, in- NTIC, Springfield, VA 22161
situ load test, , structural evaluation, thermal
effect, total station., structural health
monitoring.
19. Security Classification (of this report) 20. Security Classification (of this page) 21. No. of Pages
Unclassified Unclassified
Form DOT F 1700.7 (8-72) Reproduction of form and completed page is authorized
 DISCLAIMER
The contents of this report reflect the views of the author(s), who are responsible

CIESfor the facts and the accuracy of information presented herein. This document is
disseminated under the sponsorship of the Center for Infrastructure Engineering
05-57Studies
C (CIES), University of Missouri-Rolla, in the interest of information
exchange. CIES assumes no liability for the contents or use thereof.

ii
 The mission of CIES is to provide leadership in research and
education for solving society's problems affecting the nation's
infrastructure systems. CIES is the primary conduit for communication
among those on the UMR campus interested in infrastructure studies and
provides coordination for collaborative efforts. CIES activities include
interdisciplinary research and development with projects tailored to
address needs of federal agencies, state agencies, and private industry as
well as technology transfer and continuing/distance education to the
engineering community and industry.

Center for Infrastructure Engineering Studies (CIES)

University of Missouri-Rolla
223 Engineering Research Laboratory
1870 Miner Circle
Rolla, MO 65409-0710
Tel: (573) 341-4497; fax -6215
E-mail: [email protected]
http://www.cies.umr.edu/
 RESEARCH INVESTIGATION RI04-004

IN-SITU LOAD TESTING OF BRIDGE A6358

OSAGE BEACH, MO

PREPARED FOR THE

MISSOURI DEPARTMENT OF TRANSPORTATION

IN COOPERATION WITH THE

UNIVERSITY TRANSPORTATION CENTER

Written by:

Fabio Matta, Graduate Research Assistant at UMR

Filippo Bastianini, Adjunct Assistant Professor at UMR
Nestore Galati, Research Engineer at UMR
Paolo Casadei, Research Engineer at SGM S.r.l., San Mariano (PG) - Italy
Antonio Nanni, V. & M. Jones Professor of Civil Engineering

CENTER FOR INFRASTRUCTURE ENGINEERING STUDIES

UNIVERSITY OF MISSOURI – ROLLA

Submitted
January 2005

The opinions, findings and conclusions expressed in this report are those of the principal investigators. They are not
necessarily those of the Missouri Department of Transportation, U.S. Department of Transportation, Federal
Highway Administration. This report does not constitute a standard, specification or regulation.

II
 IN-SITU LOAD TESTING OF BRIDGE A6358
OSAGE BEACH, MO

EXECUTIVE SUMMARY

This report presents the assessment of a High Performance Steel (HPS) bridge located at the
Lake of the Ozarks in Miller County, MO. The bridge number is A6358 and it is located on US
Rt. 54/Osage River. Two different modern techniques for Structural Health Monitoring (SHM)
were applied for this purpose, in order to advance their field validation process.
Displacements of the girders during the pouring of the concrete deck and a diagnostic load test,
conducted after the bridge completion, were evaluated using an optical non-contact measurement
technique based on an automated total station system. The setup included a total of 22 reflecting
prisms mounted onto the bottom flange of the girders, plus four reference targets placed outside
the bridge superstructure.
Strain and temperature distributions along some of the girders were measured during the
diagnostic load test using a fiber optic distributed Brillouin sensing technique. The method is
based on the “Brillouin” phenomenon, which defines a linear relation between the axial strain in
an optical fiber, and the corresponding shift in the frequency distribution of the scattered
Brillouin light generated when an optical impulse runs through a fiber. A single optical circuit,
comprised of bare fibers and a custom-made novel glass FRP tape with embedded sensing fibers,
was installed on the web of the girders at different depths, along two continuous spans.
Experimental deflection and strain data were verified to be consistent with each other and
allowed to evaluate the actual girder distribution factors. The comparison between experimental
and theoretical results based on the AASHTO LRFD Bridge Design Specifications (AASHTO,
1998) allowed establishing the safety of the structure, although pointing out a significant drop in
the expected behavior of one of the external girders, which may call for further assessment.

III
 ACKNOWLEDGMENTS

The project was made possible with the financial support received from the Missouri
Department of Transportation (MoDOT), and the University Transportation Center (UTC) on
Advanced Materials. Several people and parties actively participated:
• Mr. John Wenzlick and Mr. Rick Simmons (MoDOT), in organizing and performing the
load tests;
• Leica Geosystems, in particular in the persons of Mr. Gerard Manley and Mr. Jim
Naismith, in the setup of the automated total station surveying system and providing
assistance during the load test;
• Pirelli Cavi e Sistemi S.p.A., in particular in the person of Dr. Enrico Consonni, by
donating the optical fibers;
• Rocafix/Isaberg Rapid Group, in particular in the person of Mr. Fabio Ruggeri, by
donating the low temperature glue;
• Yokogawa Corporation of America, Test & Measurement Division/ANDO Product
Group, in particular in the person of Dr. Nobuyuki Morita, in the cooperation and efforts
to make the strain analyzer unit available;
• SEAL S.p.A. (SAATI Group), and in particular Mr. Paolo Grati, in manufacturing the
custom-made smart GFRP tape;
• Rolla Technical Institute (RTI), in particular in the persons of Mr. Max Vath, the
personnel and students of the Machine Technology and Precision Manufacturing, in the
design and construction of the modular cart;
• Strongwell, in particular in the person of Mr. Richard Altice, by donating the GFRP
pultruded structural panels used in the modular cart.
All of these precious contributions are gratefully acknowledged.

IV
 TABLE OF CONTENTS

EXECUTIVE SUMMARY .......................................................................................................... III

ACKNOWLEDGMENTS ............................................................................................................ IV
TABLE OF CONTENTS................................................................................................................V
LIST OF FIGURES ..................................................................................................................... VII
LIST OF TABLES..........................................................................................................................X
NOTATIONS................................................................................................................................ XI
1. INTRODUCTION .....................................................................................................................1
1.1 Background ......................................................................................................................1
1.1.1 Historical background on high performance steel bridges ..........................................1
1.1.2 Live-load deflections ...................................................................................................2
1.1.3 Diagnostic load testing.................................................................................................2
1.2 Bridge description............................................................................................................3
1.3 Objectives ........................................................................................................................3
1.4 Methodology ....................................................................................................................5
1.5 Description of measurement technologies .......................................................................6
1.5.1 Automated Total Station system for deflection measurement.....................................6
1.5.2 Brillouin Optical Time-Domain Reflectometry for strain distribution
measurement ............................................................................................................................7
1.5.2.1 Theoretical foundations ...............................................................................................7
1.5.2.2 Practical applications..................................................................................................9
1.5.2.3 Sensing fibers and smart FRP tape............................................................................11
1.5.3 Data Acquisition System............................................................................................11
2. ATS BASED DISPLACEMENT MEASUREMENT: FIELD EVALUATION.....................13
2.1 Bridge instrumentation...................................................................................................13
2.2 Deck installation ............................................................................................................14
2.2.1 Test procedure............................................................................................................14
2.2.2 Test results .................................................................................................................14
2.2.3 Discussion of results ..................................................................................................16
2.3 Load test.........................................................................................................................21
2.3.1 Test procedure............................................................................................................21
2.3.2 Test results .................................................................................................................23
2.3.3 Discussion of results ..................................................................................................26
3. BOTDR BASED STRAIN MEASUREMENT: FIELD EVALUATION...............................32
3.1 Bridge instrumentation...................................................................................................32
3.1.1 Installation of optical fibers .......................................................................................32

V
 3.1.2 Modular cart for sensor installation ...........................................................................35
3.1.3 Connection and preliminary testing of the sensing elements.....................................36
3.1.4 Sensing system calibration.........................................................................................38
3.1.5 Installation of resistive strain gauges.........................................................................39
3.2 Test procedure................................................................................................................40
3.3 Test results .....................................................................................................................41
3.3.1 BOTDR strain measurement......................................................................................41
3.3.2 Resistive strain gauge measurements.........................................................................44
3.4 Discussion of results ......................................................................................................44
4. CONCLUSIONS......................................................................................................................51
5. REFERENCES ........................................................................................................................52
APPENDIX I – LOAD TEST STOPS...........................................................................................55
APPENDIX II – RAW DATA.......................................................................................................57
II.a Deflection data (ATS)....................................................................................................57
II.b Strain data (BOTDR) .....................................................................................................59

VI
 LIST OF FIGURES

Figure 1 – Bridge A6358, U.S. Rt. 54/Osage River, Lake of the Ozarks, Miller County, MO...... 3
Figure 2 – Bridge details (not to scale)........................................................................................... 4
Figure 3 – Framing plan (not to scale, all dimensions in inches) ................................................... 5
Figure 4 – ATR TS system equipment for non-contact deflection measurement........................... 6
Figure 5 – Representation of quantum energy level of Brillouin scattering process showing both
Stokes (left) and anti-Stokes (right) collisions........................................................................ 8
Figure 6 – simplified schematic diagram of self-heterodyne BOTDR. .......................................... 9
Figure 7 – Schematic of correlation between elapsed time (T) and scanned length (l). ................. 9
Figure 8 – Brillouin OTDR............................................................................................................. 9
Figure 9 – Brillouin technology for structural monitoring ........................................................... 10
Figure 10 – Smart GFRP tape....................................................................................................... 11
Figure 11 – Portable data acquisition system (“Orange Box”)..................................................... 12
Figure 12 – Target positions: plan view (not to scale, all dimensions in ft)................................. 13
Figure 13 – Reflecting prisms mounted on steel I-girders............................................................ 13
Figure 14 – Pouring of concrete deck and deflection measurement via ATS .............................. 14
Figure 15 – Progression of deck pouring at deflection measurement steps.................................. 14
Figure 16 – Measured longitudinal deflections (deck pouring, Girder 1) .................................... 15
Figure 17 – Measured longitudinal deflections (deck pouring, Girder 3) .................................... 15
Figure 18 – Measured longitudinal deflections (deck pouring, Girder 5) .................................... 16
Figure 19 – Measured transversal deflections (deck pouring)...................................................... 16
Figure 20 – Comparison between theoretical and experimental longitudinal deflections (deck
pouring, Girder 1) ................................................................................................................. 18
Figure 21 – Comparison between theoretical and experimental longitudinal deflections (deck
pouring, Girder 3) ................................................................................................................. 19
Figure 22 – Comparison between theoretical and experimental longitudinal deflections (deck
pouring, Girder 5) ................................................................................................................. 20
Figure 23 – Dump trucks used for load test.................................................................................. 21
Figure 24 – Transverse position of trucks during load test (not to scale)..................................... 22
Figure 25 – Total station in load test configuration...................................................................... 23
Figure 26 – Experimental longitudinal deflection (load test, Girder 1)........................................ 23
Figure 27 – Experimental longitudinal deflections (load test, Girder 3) ...................................... 24

VII
Figure 28 – Experimental longitudinal deflections (load test, Girder 5) ...................................... 24
Figure 29 – Experimental transversal deflections at midspan of Span 1 (load test)..................... 25
Figure 30 – Experimental deflections for Girder 1, Stop 8 and Girder 5, Stop 6 (load test)........ 25
Figure 31 – Girder 1 fallen due to wind during erection of bridge superstructure ....................... 26
Figure 32 – Comparison between AASHTO provisions and experimental longitudinal deflections
(load test, Girder 1) ............................................................................................................... 28
Figure 33 – Comparison between AASHTO provisions and experimental longitudinal deflections
(load test, Girder 5) ............................................................................................................... 29
Figure 34 – Comparison between AASHTO and experimental girder load distribution factor at
midspan between Abutment 1 and Pier 2 in case of one lane loaded (Stop 3, 4, 5, 6 and 8) 30
Figure 35 – Comparison between AASHTO and experimental girder load distribution factor at
midspan between Abutment 1 and Pier 2 in case of two lanes loaded (Stop 1, 2 and 7) ..... 30
Figure 36 – Schematic of fiber circuit placement......................................................................... 32
Figure 37 – Detailed location of fiber optics on girder web (not to scale, dimensions in in)....... 33
Figure 38 – Smart FRP installed on Girder 1 and folded at Pier 2 location to extend circuit ...... 34
Figure 39 – Sensing fibers installed onto web plate of Girder 2 .................................................. 35
Figure 40 – Sensing fibers installed onto web plate of Girder 5 .................................................. 35
Figure 41 – Modular frame cart for sensor installation ................................................................ 36
Figure 42 – Optical fiber core-alignment fusion splicing field operation .................................... 37
Figure 43 – Optical loss measured along sensing circuit.............................................................. 37
Figure 44 – Strain calibration fixture............................................................................................ 38
Figure 45 – Linearity of measured and imposed strain for different fibers tested........................ 39
Figure 46 – Strain-temperature correlation................................................................................... 39
Figure 47 – Close-up of strain gauges installed on top flange of Girder 1................................... 40
Figure 48 – Field equipment for distributed strain measurement ................................................. 41
Figure 49 – Proprietary software for strain data processing ......................................................... 41
Figure 50 – Experimental strain distribution (Stop 1) .................................................................. 42
Figure 51 – Experimental strain distribution (Stop 2) .................................................................. 42
Figure 52 – Experimental strain distribution onn Girder 5 (Stop 4)............................................. 42
Figure 53 – Experimental strain distribution (Stop 7) .................................................................. 43
Figure 54 – Experimental strain distribution on Girder 1 and Girder 2 (Stop 8).......................... 43
Figure 55 – Time history of strain gauge readings ....................................................................... 44
Figure 56 – Strain gauge data computed for Stop 1-8 accounting for thermal compensation...... 44
Figure 57 – Comparison between experimental and predicted strain profiles (Stop 1)................ 45

VIII
Figure 58 – Comparison between experimental and predicted strain profiles (Stop 2)................ 46
Figure 59 – Comparison between experimental and predicted strain profiles (Stop 4)................ 47
Figure 60 – Comparison between experimental and predicted strain profiles (Stop 7)................ 48
Figure 61 – Comparison between experimental and predicted strain profiles (Stop 8)................ 49

IX
 LIST OF TABLES

Table 2.1 – Properties of steel girder sections (as in Figure 2b) for deflection computation during
deck pouring.......................................................................................................................... 17
Table 2.2 – Dump trucks used for load test: weight distribution assumed for analysis ............... 21
Table 2.3 – GDFs as per AASHTO LRFD Bridge Design Specifications (1998) ....................... 26
Table 2.4 – Input data for calculation of GDFs (AASHTO, 1998) .............................................. 27
Table 3.1 – Properties of Wabo MBrace saturant for fiber optics installation (Wabo, 2004) ...... 32
Table 3.2 – Type, identification and location of sensing elements installed ................................ 33
Table 3.3 – Type, identification and location of strain gauges installed ...................................... 40
 NOTATIONS

As : Area of steel girder section

Ab : Cross sectional area of concrete barrier
Ec : Modulus of elasticity of concrete
Es : Modulus of elasticity of steel
Is : Moment of inertia of steel girder section
Kg : Longitudinal stiffness parameter
L : Span length
Mp : Plastic moment capacity
My : Yield moment capacity
P1 : Dump truck front axle load
P2 : Dump truck rear axle load
S : Girder spacing
eg : Distance between center of gravity of girder section and concrete deck
fy : Steel yield strength
hweb : Clear distance between flanges of I-girder
hνa : Energy transfer to/from acoustic phonon
m : Multiple presence factor
tc : Thickness of concrete deck
woverhang : Overhang width from exterior girder center line
δi : Deflection of ith girder
ν0 : Frequency of monocromatic light source
νa : Frequency shift of scattered phonons

XI
 1. INTRODUCTION

1.1 Background

1.1.1 Historical background on high performance steel bridges

Collaboration between government and industry has led to the development of high performance
steel (HPS) for bridge applications. HPS offers increased yield strength, enhanced weldability,
and improved toughness, and it may lead to lighter and more economical structures. The
introduction of High Performance Steel (HPS) with minimum yield strength of 70 ksi (HPS
70W) in 1994 and its utilization in bridge design and construction coincided with various design
limitations imposed by the AASHTO Standard Specifications (AASHTO, 1996). At the time,
these limitations reflected the lack of test data to fully comprehend the behavior of bridges
constructed using HPS. The code limitations were addressed through multiple research projects
funded by the HPS steering committee and the Federal Highway Administration (FHWA). As a
result, a coordinated national effort was initiated by the steel industry to address these
limitations. The effort was coordinated by the American Iron and Steel Institute’s (AISI) steel
bridge task force and HPS Design Advisory Group, in close cooperation with the AASHTO T-14
committee (Steel Bridges).
The 1998 version of the AASHTO LRFD Bridge Design Specifications marked the first time
where design limitations were imposed (AASHTO, 1998). The following is a summary of the
various design issues that either had HPS related limitations in the 1998 version of the AASHTO
code or were a concern for designers (Azizinamini et al., 2004).
• For continuous plate girders with compact negative sections, the maximum moment
capacity was limited to yield moment capacity of the section, My, rather than the plastic
moment capacity, Mp, as used for lower grade steels.
• Inelastic methods of analysis and design were not permitted for HPS plate girders with
yield strength equal to or exceeding 70 ksi. It should be noted that this also disallows the
use of the 10% moment redistribution for girders comprised of compact sections.
• Information to check the ductility of composite plate girders in the positive section, per
requirements stated in section 6.10.4.2.2 of the AASHTO LRFD Bridge Design
Specifications, was incomplete. As a result, one could conclude that HPS could not be
used in the positive sections since ductility check provisions were not available.
• In the initial stages of introducing HPS, there was a concern that HPS plates may not be
able to sufficiently develop large tensile strains without fracture. Some of the examples
included use of HPS as tension flanges of plate girders in positive sections or use of HPS
as tension members in trusses. These concerns were mainly a result of work by
McDermott (1969) who conducted tests on A514 100 ksi steel girders. When these
girders were loaded in three-point bending, the high strength steel tension flange
fractured very shortly after yielding.
• For years, AASHTO codes had limited the shear capacity of the hybrid plate girders to
the elastic buckling capacity of the panel. Hybrid girders consist of using higher grades of
steel for flanges and lower grades for the web. For instance, using 50 ksi steel for flanges

1
 and 36 ksi steel for webs, or using 70 ksi steel for flanges and 50 ksi steels for webs. This
limitation was not confined to HPS. It also applied to hybrid girders fabricated with 50
and 36 ksi steels. What made this limitation important was the fact that it has been shown
that the best use of HPS in plate girders is in the hybrid form where flanges are
constructed using 70 ksi steels and webs are constructed using 50 ksi steels (Horton et al.,
2000). Design studies have indicated that for many typical situations, hybrid girders will
produce the most economical designs (Barker and Schrage, 2000; Horton et al., 2000;
Clingenpeel and Barth, 2003).
• Current specifications place limits on the maximum allowable live load deflections.
These requirements have not typically controlled the geometry of sections designed with
steel having fy < 50 ksi. However, due to the reduced section geometries required when
HPS 70W steel is incorporated, these limits may be the controlling limit state for some
design situations. Therefore, research has been initiated to investigate the rationale
behind the current limits and to assess their influence on the serviceability and economy
of HPS bridges.
• In addition to incorporating HPS 70W in traditional I girder configurations, it was felt
that innovative concepts capitalizing on both the increased strength and improved
toughness of the steel may also produce cost effective structures. Some of the design
innovations that have been developed may prove to be beneficial for various steel grades
(e.g., Grade 50, HPS 70W and HPS 100W, with minimum yield strength of 100 ksi).

1.1.2 Live-load deflections

The AASHTO Standard Specifications (1996) place a limit on the maximum allowable live-load
deflection of L / 800 for most bridges and L / 1000 for bridges subject to pedestrian use. Similar
specifications are given in the AASHTO LRFD Specifications (1998), which are specifically
written as optional criteria; however, many state transportation departments will continue to view
them as mandatory requirements (Azizinamini et al., 2004). For traditional steel bridges
comprised of steels with fy < 50 ksi, these limits have rarely been found to control girder
geometries. However, recent studies have shown that for some design situations with HPS 70W,
particularly in cases with high span-to-depth ratios, these limits may have a significant influence
on the final section requirements (Barth et al., 2001; Roeder et al., 2001; Roeder et al., 2004).
A recent research study focusing on examining the influence of these limits on steel I girder
bridge design showed that to date there is not a relationship between either reported bridge
damage or objectionable vibration characteristics with a direct check of live load deflections
(Barth et al., 2001; Roeder et al., 2001; Roeder et al., 2004).

1.1.3 Diagnostic load testing

Field testing is an increasingly important topic in the effort to deal with new infrastructures using
new technologies as well as deteriorating infrastructure, in particular bridges and pavements.
There is a need for accurate and inexpensive methods for diagnostics, verification of load
distribution and determination of the actual load carrying capacity.
Recent studies indicate that over one quarter of the national bridges are either structurally
deficient or functionally obsolete, with about half of these involving steel superstructure
(FHWA, 2004). The major factors that have contributed to the present situation are ageing,

2
substantial increase in traffic on existing life-lines, lack of routine inspection and inadequate
maintenance in addition to the effect of the harsh and changing environment. The deficient
bridges are posted, repaired or replaced, with the disposition of bridges involving clear
economical and safety implications. To avoid high costs of replacement or repair, the evaluation
must accurately reveal the present load carrying capacity of the structure and predict loads and
any further changes in the capacity (deterioration) in the applicable time span (Deza, 2004).
Frequently, diagnostic load tests reveal strength and serviceability characteristics that exceed the
predicted codified parameters. Usually, codified parameters are very conservative in predicting
lateral load distribution characteristics and the influence of other structural attributes. As a result,
the predicted rating factors are typically conservative (Chajes et al., 1997).

1.2 Bridge description

The bridge under investigation (see Figure 1) was recently open to traffic. The bridge is built
with five continuous symmetrical spans: the two external are 147 ft and 185 ft long, respectively,
while the central one has a length of 200 ft, resulting in a total bridge length of 864 ft. Each
internal support consists of reinforced concrete (RC) bents supported by two RC circular piers
having a 6 ft diameter. The cross section comprises five composite, equally spaced, HPS I-
girders acting compositely with a 8.5 in thick RC deck, with an out-to-out deck and clear
roadway width of 40.7 ft and 38 ft, respectively. Cross section and elevation of the girders are
provided in Figure 2, while the framing is illustrated in Figure 3.

a) Side view b) Bridge superstructure under construction

Figure 1 – Bridge A6358, U.S. Rt. 54/Osage River, Lake of the Ozarks, Miller County, MO

1.3 Objectives
The scope of this project is the assessment of a High Performance Steel (HPS) bridge located at
the Lake of the Ozarks in Miller County. The bridge number is A6358 and it is sited on the U.S.
Rt. 54/Osage River.
In the project two different innovative techniques for Structural Health Monitoring (SHM) have
been applied:

3
1. An optical non-contact measurement technique, based on an automated target recognition
Total Station (TS) system, used to measure the vertical deflections of the girders during
the pouring of the concrete deck and the diagnostic load test, conducted after the
completion of the bridge;
2. A distributed Brillouin sensing technique, utilizing a “smart” glass FRP (GFRP) tape
with embedded sensing fibers and bare optical fibers, used to determine the strain
distribution at different depths of the web for some of the girders. Strains were measured
during the diagnostic load test only, since at the time when the deck was poured the
optical sensors had not been installed yet.

16" 38' 16"

8 1/2"

3' 8' 8" 8' 8" 8' 8" 8' 8" 3'

a) Cross section
PL. 18x1 1/8 PL. PL. PL. PL. 16x1 1/8 PL. PL.
HPS 485 W 18x1 1/8 14x3/4 16x1 1/8 HPS 485 W 16x1 1/8 16x3/4

PL. 5/8 PL. 9/16 PL. 5/8 PL. 5/8 PL. 9/16 84

PL. 17x1 1/2 PL. 16x3/4

18 28 98 21 20 20 19 108 0.92

185 147
Span 2 Span 1
C
L C
L C
L
Pier 3 Pier 2 Abutment 1

PL. 18x1 1/8 PL. PL. PL. PL. 18x1 1/8

HPS 485 W 18x1 1/8 18x3/4 18x1 1/8 HPS 485 W

PL. 5/8 PL. 5/8 PL. 5/8

Lengths in ft,
plate dimensions in in
PL. 22x2

20 19 122 19 20

200
Span 3
C
L C
L
Pier 4 Pier 3

b) Girder elevation (structural steel ASTM A709 Grade HPS 345 W unless specified)

Figure 2 – Bridge details (not to scale)

4
 Span 3 Span 2 Span 1

Girder 1
Girder 2
Girder 3 4 spa.
@ 8' 8"
Girder 4
Girder 5
2 spa. 2 spa.
4 spa. @ 22' 24' 4 spa. @ 22' @ 20' 9" 4 spa. @ 24' 6" @ 22' 9" 22' 5 spa. @ 25'

CL CL CL CL
Pier 4 Pier 3 Pier 2 Abutment 1

a) Layout

3.5 x 3.5 x 15/16

3.5 x 3.5 x 15/16

3/8 PL. 3/8 PL. 7'

3.5 x 3.5 x 15/16

8' 8"

b) Typical cross frame (structural steel ASTM A709 Grade HPS 70 W unless specified)

Figure 3 – Framing plan (not to scale, all dimensions in inches)

1.4 Methodology
To evaluate the response of the bridge, a nondestructive field test was conducted. Experimental
load testing on bridges can be either categorized as diagnostic or proof test. In a diagnostic test, a
predetermined load, typically near the rated capacity of the structure, is placed at several
different locations along the bridge. The corresponding response is measured, and then used to
develop a numerical model of the bridge to allow estimating of the maximum allowable load. In
a proof test, incremental loads are applied to the bridge until either a target load is reached or a
predetermined limit state is exceeded. Using the maximum load achieved, the capacity of the
bridge can be determined. Although diagnostic tests provide only an estimate of the load
carrying capacity of a bridge, they have several practical advantages including lower cost,
shorter testing time, and less disruption to traffic. Because of these advantages, diagnostic testing
was used in this case.
The bridge was tested under different static loads using six fully loaded dump trucks, and the
comparison between theoretical displacement and strain profiles based on the AASHTO LRFD
Bridge Design Specifications (1998) and experimental results allowed determining the safety of
the structure.
A major difficulty in the testing and evaluation of bridges in the field is the measurement of
vertical deflections. The use of instruments such as mechanical dial gauges, linear
potentiometers, linear variable differential transducers (LVDTs), and other similar types of
deflection transducers is not always feasible, because a fixed base is needed from which relative

5
displacements are measured. This often requires access under the bridge to erect a temporary
scaffolding to mount the instrument or for running a wire from the instrument to the ground. In
the present case, these difficulties were overcome by means of a non-contact deflection
measurement technique using an automated target recognition TS system. The solution allows to
measure the spatial coordinates of discrete points on a bridge in three dimensions without need
of touching the structure. The reliability of this method has been previously demonstrated in a
companion report (Galati et al., 2004), where the percentage variation of the total station with
respect to LVDTs displacement readings was estimated in the range between 0.1% and 9.7%.

1.5 Description of measurement technologies

1.5.1 Automated Total Station system for deflection measurement

The Automated Total Station (ATS) system used in this project is based on the Leica TCA2003
TS, shown in Figure 4a. The instrument sends a laser ray to reflecting prisms mounted on the
structure (Figure 4b) to be monitored. By triangulation with fixed reference points placed outside
the structure, the built-in computer can determine the movements of each prism with an accuracy
of 0.5" on angular measurements and 0.04 in + 1 ppm on distance measurements, in average
weather conditions.

a) Total station b) Reflecting prism mounted on girder

Figure 4 – ATR TS system equipment for non-contact deflection measurement

TSs have been used to measure the movement of structures and natural processes with good
results (Hill and Sippel, 2002; Kuhlmann and Gläser, 2002). Leica Geosystems quotes accuracies
of better than 1 mm for their bridge and tunnel surveys. They use a remote system that logs
measurements 6 times daily via a modem, with measurements still possible at peak times.
Kuhlmann and Gläser (2002) used a reflectorless TS to monitor the long term deformation of
bridges. Measurements on a whole bridge were taken every six years and statistical tests were
used to confirm whether the points had moved over time. Hill and Sippel (2002) used a TS as
part of a multiple-sensor system for landslide monitoring. Merkle (2004) used the TS as part of a
5-year monitoring program for the in-situ load testing prior to and after the strengthening of five
existing concrete bridges, geographically spread over three Missouri Department of

6
Transportation (MoDOT) districts. The five bridges were strengthened using five different Fiber
Reinforced Polymer (FRP) technologies as part of a joint MoDOT – University of Missouri-
Rolla (UMR) initiative (Lynch, 2004).
There are advantages and disadvantages of using a TS for dynamic deformation monitoring. The
advantages include the high accuracy as quoted above, the automatic target recognition which
provides precise target pointing (Hill and Sippel, 2002) and the possibility of measuring indoors
and in urban canyons in kinematic operating modes (Radovanovic and Teskey, 2001). The
disadvantages include the low sampling rate (Radovanovic and Teskey, 2001; Meng, 2002),
problems with measurement in adverse weather conditions (Hill and Sippel, 2002) and the need
for a clear line of sight between the TS and the prism.
Radovanovic and Teskey (2001) conducted experiments to compare the performance of a robotic
TS with GPS. This experiment was conducted because GPS is not an option in many application
areas such as indoors. TSs are now capable of automatic target recognition and they can track a
prism taking automatic measurements of angles and distances once lock has been established
manually. It was found that the TS performed better than GPS in a stop and go situation, where
measurements were taken of a moving object only when it was stationary. In a completely
kinematic situation, GPS performed the best. It was found that there were two main problems
with the TS in kinematic mode; these were a low electronic distance measurement (EDM)
accuracy due to a ranging error that was linearly dependent upon the line of sight velocity, and
an uneven sampling rate over time worsened by no time tagging.

1.5.2 Brillouin Optical Time-Domain Reflectometry for strain distribution measurement

1.5.2.1 Theoretical foundations

The Brillouin effect is an anelastic scattering process that arises from the interaction between
optical and acoustic waves propagating in the same physical medium. The process, sketched in
Figure 5, is characterized by a partial energy transfer between the colliding electromagnetic
waves (photons) and mechanical wave (phonons). When the medium is illuminated with a
monochromatic light source of frequency ν0, there is a spontaneous production of scattered
photons characterized by a frequency shift νa with respect to the incident ones. The shift can be
either negative, when the energy hνa is transferred to the created acoustic phonon, also known as
Stokes process, or positive, when it is absorbed from an annihilated acoustic phonon, also known
as anti-Stokes process (Brillouin, 1922). Being the energy of the acoustic phonon related to the
sound speed in the medium in which the process takes place, the Brillouin wavelength shift is
related to the density of the medium, and therefore both with its temperature and mechanical
strain. The Stokes process is the most probable due to entropy considerations (Hill et al., 1976;
Yariv, 1989); it can be observed in silica single-mode optical fibers (Pelous and Vacher, 1975;
Thomas et al., 1979).
With the development of novel techniques capable of scanning the Brillouin scattered light
spectrum with better resolution, a correlation allowing the evaluation of the strain level from the
Brillouin frequency shift was experimentally found (Horiguchi et al., 1989).
Brillouin distributed strain sensing is characterized by a certain number of advantages that make
it a very attractive technique for SHM of civil infrastructures.
The main advantage is in the availability of a fully-distributed sensing technique using standard

7
telecom fiber optics as fully-intrinsic sensors. This results in the following practical
considerations:
• The sensor is a standard, inexpensive telecom-grade optical fiber, with easy and
immediate off-the-shelf availability for any required length;
• At each point, the fiber acts as a strain sensor, allowing detection and evaluation of
phenomena that are hard to locate in advance, such as cracks, debonding, fatigue and
defect-triggered faults, along with local discontinuities that cannot be detected on real
structures by means of point sensors;
• Low cost, in addition to distributed sensitivity, makes possible long-haul sensors to be
installed extensively over the length, area or volume of large structures, thus allowing
better understanding of the structural behavior and producing a global overview of the
damage distribution, which cannot be retrieved even with long measure-basis sensors;
• Low cost and distributed sensitivity make the system suitable for disposable sensors. For
instance, in case of monitoring of tendons during pre-stressing operations, or for
emergency sensors that can be left “sleeping” in the structure without heavy additional
costs in all cases when it is not possible to know in advance whether the sensor will be
used during the life-cycle of the structure or not;
• The whole fiber length can be scanned from the same end point (up to several tenths of
km), a characteristic that dramatically simplifies installation and management of cabling;
• Using an Optical Time Domain Reflectometer (OTDR) based on spontaneous Brillouin
scattering, no end-to-end full cable continuity is required, which allows operations even
with damaged sensing fibers, performing separate scans for the two survival half-lengths.

Figure 5 – Representation of quantum energy level of Brillouin scattering process showing

both Stokes (left) and anti-Stokes (right) collisions.

In addition to these peculiarities, the Brillouin reflectometry maintains all the advantages that are
typical of fiber optic sensing technologies (Casas and Cruz, 2003), i.e.
• Absolute insensitivity to Electro-Magnetic Interferences (EMI);
• Long-term durability even in chemically aggressive or harsh environments;
• Current technology top miniaturization grade;
• Complete compatibility with FRP composite materials.

8
1.5.2.2 Practical applications
Even though primarily developed for telecom cable testing, Brillouin Optical Time-Domain
Reflectometry (BOTDR) can be effectively applied for strain distribution monitoring in large-
scale civil structures, with some limitations (Komatsu et al., 2002). BOTDR combines the self-
heterodyne analysis of the spontaneous Brillouin backscattered light spectrum (Horiguchi et al.,
1994), illustrated in Figure 6, with OTDR (Barnoski and Jensen, 1976). The first allows the
evaluation of the strain level of a certain length of fiber, while OTDR enables determining the
distance along the fiber where the strain information is retrieved by exploiting the relation
between the speed of light (c), the sum of the distances traveled by the injected and backscattered
light pulses, and the delay time at which the backscattered pulse is detectable (Figure 7).
For the purpose of this test program, the evaluation of the strain distribution has been performed
using a Brillouin OTDR AQ8603 optical fiber strain analyzer manufactured by Yokogawa-Ando
Corporation of Kawasaki, Japan (Figure 8). The unit allows measurements of strain distribution
along a single-mode optical fiber from one end of the optical circuit.
In addition, recent experimental tests have evidenced that performances can overcome the
declared limits when the strain distribution under evaluation has no steep discontinuities and has
positive peak strain values of at least 100 µε. Under these conditions the relative error versus
conventional resistive strain gauges has been experimentally shown to be lower than 5%
(Bastianini et al., 2003).

Figure 6 – simplified schematic diagram of Figure 7 – Schematic of correlation between

self-heterodyne BOTDR. elapsed time (T) and scanned length (l).

Figure 8 – Brillouin OTDR.

9
Figure 9 shows the flow chart for structural testing and monitoring using BOTDR technique. It
requires the following steps:

Figure 9 – Brillouin technology for structural monitoring

1. Preliminary structural analysis to identify the areas that have strategic relevance for the
structural behavior and are subject to strain value and distributions that are within the
detectable strain range of the BOTDR equipment;
2. Design of the optical fiber circuit, including strain sensitive hauls, thermal compensation
hauls and access points for future bypass patches in accordance with the areas located
during step 1;
3. Installation of the optical fiber circuit providing suitable strain transfer and fiber
protection, typically using fiber-equipped “smart” composite materials and other products
developed for the specific purpose, and then connecting the various circuit branches
through fusion splices and other suitable joints;
4. Optical loss testing in order to ensure that the total optical loss is within the maximum
acceptable losses that allow BOTDR operation within the desired accuracy limits;
5. Identification of the 3-D structure points / 1-D sensing fiber correspondence, needed to
translate the strain distribution measured along the optical fiber to the strains in the
specific areas of the structure equipped with the same fiber. Brillouin co-sensitivity to
temperature may be conveniently used in this phase to locate specific temporarily heated
points;
6. Load testing of the structure with BOTDR, retrieving the strain distribution along the
fiber circuit in different condition, i.e. without any load applied and during the various
steps of static load tests;

10
 7. Data interpretation, based on the 3-D/1-D correspondence retrieved during step 5;
8. Health status monitoring, performing Brillouin analysis on a periodical basis, or, if
required, continuously.

1.5.2.3 Sensing fibers and smart FRP tape

The application of Brillouin fiber optic sensors to a bridge structure introduces a number of
problems related to the small order of magnitude of the strains typically achieved during a
diagnostic load test, and to the large temperature variations due to prolonged exposure to
sunlight. Therefore, it was chosen to use different sensing fibers for strain detection and thermal
compensation:
a) 9/125 µm silica single mode fiber with ∅ 900 µm tight PA buffer coating, manufactured
by Pirelli Cavi e Sistemi S.p.A., F.ID. 100005827586/FOS_Bast;
b) 9/125 µm silica single mode fiber with ∅ 900 µm loose dry coupled PVC buffer coating,
manufactured by Pirelli Cavi e Sistemi S.p.A., F.ID. 100009866510/PIR_Bast;
c) 9/125 µm silica single mode fiber with ∅ 900 µm silicone oil filled PE buffer coating,
manufactured by H&S, 01-E9/CW-EW 22600050-110562 F.ID. 1641/2003.
In order to facilitate the installation and reduce the risk of damaging the fibers, an FRP tape with
embedded sensing fibers was designed and manufactured. The tape, which is woven from E-
glass fiber strands with a 435 g/m2 warp and a 68 TEX weft, has a width of 100 mm and carries
two strain-sensing and two temperature-sensing fibers (Figure 10).

a) Schematic with location of embedded sensing fibers b) Manufacturing

(dimensions in mm)
Figure 10 – Smart GFRP tape

1.5.3 Data Acquisition System

A portable data acquisition unit, custom-manufactured at UMR and suitable for use in field
testing of structures (see Figure 11), was used for signal conditioning of the strain gauges
installed on the girders to provide a reference data for comparison with the results from BOTDR.
The “Orange Box” is capable of recording 32 high-level channels of data, 16 strain channels, and

11
32 thermocouple channels, as well as interfacing with TS surveying. The high-level channels
may receive DC LVDT's, string transducers, linear potentiometers, or any other +/− 10 Volt DC
signal. The strain channels can be used to monitor and record strain gauge signals, load cells,
strain-based displacement transducers, or any strain based signal. The 32 thermocouple channels
are configured for type T thermocouples.

Figure 11 – Portable data acquisition system (“Orange Box”)

The unit consists of a shock-mounted transport box, with removable front and rear covers.
Removal of the front cover exposes the computer keyboard and LED display, as well as the front
panel of the data acquisition equipment. Removal of the back panel exposes the connector bay,
where cables from all the transducers terminate.
The data acquisition system is comprised of National Instruments equipment, listed below:
a) A PXI-1010 SCXI combination unit, which houses the industrial-grade 2.2 GHz Pentium
4 computer, floppy drive, and CDR/W module;
b) A PXI-6030E Analog to Digital converter module for doing the A/D conversion in the
system;
c) A pair of SCXI-1520 modules to interface strain based sensors;
d) A SCXI-1102B module for multiplexing high-level sensors;
e) A SCXI-1102B module for multiplexing thermocouple sensors;
f) I/O devices in order to connect additional peripherals and other data acquisition systems
such as a Leica TS surveying instrument.
The data acquisition system is controlled by a custom made LabVIEW program installed on a
built-in computer, which allows control of data rate, sensor selection and calibration, and display
of the data.

12
 2. ATS BASED DISPLACEMENT MEASUREMENT: FIELD EVALUATION

2.1 Bridge instrumentation

A total of 22 prisms were adhesively bonded under the bottom flange along the girders using a
cherry picker. Figure 12 depicts the position of the prisms. Targets T1 to T5, T6 to T12 and T13
to T20 along Girder 1, 3 and 5 provided the data to evaluate the longitudinal deformed shape,
while the remaining targets were distributed on Girders 2 and 4 to determine the transversal load
distribution. Figure 13 shows the reflecting prisms mounted along the first span, between
Abutment 1 and Pier 2.

185 147

97.5 46.5 37 32 37.5 23.5 47 3

Target 22 Target 21 Target 20 Target 19 Target 18

Girder 1
Target 17
Girder 2
Target 16 Target 15 Target 14 Target 13 Target 12 Target 11 Target 10
Girder 3
Target 9
Girder 4
Target 8 Target 7 Target 6 Target 5 Target 4 Target 3 Target 2 Target 1
Girder 5
Pier 2 Pier 1 Abutment 1

Figure 12 – Target positions: plan view (not to scale, all dimensions in ft)

a) Installation operation b) Targets in place

Figure 13 – Reflecting prisms mounted on steel I-girders

An additional prism was mounted on Pier 2 in order to evaluate possible settlements. Finally,
two additional prisms were mounted on two piers of the existing bridge running parallel to the
one under investigation (Figure 1), and two more prisms were mounted on tripods fixed to the
ground. These four additional targets were used as reference points for triangulation.

13
2.2 Deck installation

2.2.1 Test procedure

Prior to the final load test, the ATS system was utilized to measure the deflections induced by the
pouring of the concrete deck (Figure 14). Data were recorded after pouring of progressive deck
portions, starting with the first span on the opposite side of Span 1 and 2, where the targets were
mounted. The steps are sketched in Figure 15.

a) Casting of concrete b) Total station running

Figure 14 – Pouring of concrete deck and deflection measurement via ATS

Day 1 (Apr. 13, 2004) Day 2 (Apr. 14, 2004)

1 1/4 spans

2 spans

2 3/4 spans

4 spans

5 spans

Abutment 6 Pier 5 Pier 4 Pier 3 Pier 2 Abutment 1

Span 5 Span 4 Span 3 Span 2 Span 1

Figure 15 – Progression of deck pouring at deflection measurement steps

Data were retrieved without any interruption of the pouring or interference with the personnel on
the bridge. Therefore, the accuracy of the measurements was expected to be influenced also by
potential settlements and loading due to the crane and equipment moving along the bridge. In
addition, it is envisaged that the deflections during the second day were slightly affected by a
certain amount of deck/girder composite action in Span 3, 4 and 5.

2.2.2 Test results

The longitudinal deflections measured during pouring of the concrete deck are summarized in

14
Figure 16, Figure 17 and Figure 18 for Girder 1, Girder 3 and Girder 5, respectively. Figure 19
shows the transversal midspan deflections in Span 1 measured on Girder 1 to 5 through Target
20, 17, 12, 9 and 3, respectively, wherein data from the same reading have been piecewise
connected with straight lines for clarity. Raw data are provided in Appendix II.a.

2.0
1.5 1 1/4 spans - Exp.
1.0 2 spans - Exp.
Deflection [in]

0.5 2 3/4 spans - Exp.

0.0 4 spans - Exp.

-0.5 5 spans - Exp.

-1.0
Bolted joint

Bolted joint
Abutment 1

Bolted joint
-1.5 Pier 1

Pier 2
-2.0
-2.5
0 30 60 90 120 150 180 210 240 270 300 330

x [ft]

Figure 16 – Measured longitudinal deflections (deck pouring, Girder 1)

2.0
1.5 1 1/4 spans - Exp.

1.0 2 spans - Exp.

Deflection [in]

0.5 2 3/4 spans - Exp.

0.0 4 spans - Exp.

-0.5 5 spans - Exp.

-1.0
Bolted joint
Bolted joint
Abutment 1

-1.5
Bolted
Pier 1

Pier 2

-2.0
joint

-2.5
0 30 60 90 120 150 180 210 240 270 300 330

x [ft]

Figure 17 – Measured longitudinal deflections (deck pouring, Girder 3)

15
 2.0
1.5 1 1/4 spans - Exp.

1.0 2 spans - Exp.

0.5
Deflection [in]
2 3/4 spans - Exp.

0.0 4 spans - Exp.

-0.5 5 spans - Exp.

-1.0

Bolted joint
Bolted joint
Abutment 1

Bolted joint
-1.5

Pier 1

Pier 2
-2.0
-2.5
0 30 60 90 120 150 180 210 240 270 300 330

x [ft]

Figure 18 – Measured longitudinal deflections (deck pouring, Girder 5)

2.0

1.5

1.0

0.5

0.0

-0.5

-1.0

-1.5

-2.0 1 1/4 spans 2 spans 2 3/4 spans 4 spans 5 spans

-2.5

1 2 3 4 5
Girder No.

Figure 19 – Measured transversal deflections (deck pouring)

2.2.3 Discussion of results

The experimental deflections were compared with predictions based on 1-D beam analysis. The
vertical load per girder was estimated using a tributary area approach and assuming an average
thickness of 5.5 in for the concrete layer poured, increased by 3 in along the 3 ft wide overhangs.
The actual geometry of each steel girder section was accounted for, as referred to Figure 2b and
Table 2.1.

16
Table 2.1 – Properties of steel girder sections (as in Figure 2b) for deflection computation during
deck pouring
Flexural Stiffness,
Length Interval Moment of Inertia, Is
EsIs (Es = 29 msi)
(ft) (Span) (in4) (kip·in2)
0-108 (1) 70,879 2.06E+09
108-188 (1,2) 96,090 2.79E+09
188-286 (2) 88,245 2.56E+09
286-371 (2,3) 104,243 3.02E+09
371-432 (C.L., 3) 108,323 3.14E+09

Figure 20, Figure 21 and Figure 22 show the comparison between analytical and representative
experimental results for Girder 1, 3 and 5. It can be observed that the predictions fairly describe
the measured deflections. Some apparent inconsistencies due to relatively high deflections for
Girders 1, 3 and 5, as the concrete was poured on Span 4 in particular, and Span 5, may be
attributed to the following variables:
• Additional loading due to the steel crane, loading the two external girders directly, and
the workers and equipment moving along the bridge during the pouring operations (see
Figure 14);
• Deflections induced by thermal loads;
• Settlements, due to the additional uniform dead load of the concrete poured;
• Deck/girder composite behavior in Span 3, 4 and 5, as far as the measurement of the
second day are concerned, as the concrete poured during the first day was already
considerably hardened;
• Approximation of the loading steps considered in the calculations (see Figure 15) versus
the actual load conditions, as the pouring operations were never interrupted for deflection
measurement purposes, in order not to interfere with the personnel on the bridge.
However, it is noted that a comparison between Figure 20d and Figure 22d points out a rather
significant midspan deflection of Girder 1 (1.193 in), which is 32% and 14% greater than the
correspondent in Girder 5 (0.898 in) and the theoretical value (1.025 in), respectively. Due to the
variety of factors that may affect measurements, as specified above, more thoughtful conclusions
cannot be drawn herein.

17
 2.5 2.5
2.0 Analytical 2.0 Analytical
1.5 Experimental 1.5 Experimental
1.0 1.0
Deflection [in]

Deflection [in]
0.5 0.5
0.0 0.0
-0.5 -0.5
-1.0 -1.0
Bolted joint

Bolted joint
Abutment 1

Abutment 1
Bolted joint

Bolted joint
Bolted joint

Bolted joint
-1.5 -1.5
Pier 2

Pier 2
Pier 3

Pier 3
-2.0 -2.0
-2.5 -2.5
0 30 60 90 120 150 180 210 240 270 300 330 0 30 60 90 120 150 180 210 240 270 300 330
x [ft] x [ft]

a) 2 spans b) 2 3/4 spans

2.5 2.5
Bolted joint

2.0 2.0
1.5 1.5 Analytical
Analytical
1.0 1.0 Experimental
Experimental
Deflection [in]

Deflection [in]
0.5 0.5
0.0 0.0
-0.5 -0.5
-1.0 -1.0
Bolted joint

Bolted joint
Abutment 1

Abutment 1
Bolted joint

Bolted joint

Bolted joint
-1.5 -1.5
Pier 2

Pier 2
Pier 3

Pier 3
-2.0 -2.0
-2.5 -2.5
0 30 60 90 120 150 180 210 240 270 300 330 0 30 60 90 120 150 180 210 240 270 300 330
x [ft] x [ft]

c) 4 spans d) 5 spans

Figure 20 – Comparison between theoretical and experimental longitudinal deflections (deck

pouring, Girder 1)

18
 2.5 2.5
2.0 2.0 Analytical
1.5 Analytical 1.5 Experimental
1.0 Experimental 1.0
Deflection [in]

Deflection [in]
0.5 0.5
0.0 0.0
-0.5 -0.5
-1.0 -1.0
Bolted joint

Bolted joint
Abutment 1

Abutment 1
Bolted joint

Bolted joint

Bolted joint

Bolted joint
-1.5 -1.5
Pier 2

Pier 2
Pier 3

Pier 3
-2.0 -2.0
-2.5 -2.5
0 30 60 90 120 150 180 210 240 270 300 330 0 30 60 90 120 150 180 210 240 270 300 330
x [ft] x [ft]

a) 2 spans b) 2 3/4 spans

2.5 2.5
Bolted

2.0 2.0
joint

1.5 1.5 Analytical

1.0 Analytical 1.0 Experimental
Deflection [in]

Deflection [in]
0.5 Experimental 0.5
0.0 0.0
-0.5 -0.5
-1.0 -1.0
Bolted joint

Bolted joint
Abutment 1

Abutment 1

Bolted joint

Bolted joint
-1.5 -1.5
Bolted
Pier 2

Pier 2
Pier 3

Pier 3
-2.0 -2.0
joint

-2.5 -2.5
0 30 60 90 120 150 180 210 240 270 300 330 0 30 60 90 120 150 180 210 240 270 300 330
x [ft] x [ft]

c) 4 spans d) 5 spans

Figure 21 – Comparison between theoretical and experimental longitudinal deflections (deck

pouring, Girder 3)

19
 2.5 2.5
2.0 2.0
1.5 Analytical 1.5 Analytical

1.0 Experimental 1.0 Experimental

Deflection [in]

Deflection [in]
0.5 0.5
0.0 0.0
-0.5 -0.5
-1.0 -1.0
Bolted joint

Bolted joint
Abutment 1

Abutment 1
Bolted joint

Bolted joint

Bolted joint

Bolted joint
-1.5 -1.5
Pier 2

Pier 2
Pier 3

Pier 3
-2.0 -2.0
-2.5 -2.5
0 30 60 90 120 150 180 210 240 270 300 330 0 30 60 90 120 150 180 210 240 270 300 330
x [ft] x [ft]

a) 2 spans b) 2 3/4 spans

2.5 2.5
Bolted joint

Bolted joint

2.0 2.0
1.5 1.5 Analytical
1.0 1.0 Experimental
Deflection [in]

Deflection [in]
0.5 Analytical 0.5
0.0 Experimental 0.0
-0.5 -0.5
-1.0 -1.0
Bolted joint

Bolted joint
Abutment 1

Abutment 1

Bolted joint
Bolted joint
-1.5 -1.5
Pier 2

Pier 2
Pier 3

Pier 3
-2.0 -2.0
-2.5 -2.5
0 30 60 90 120 150 180 210 240 270 300 330 0 30 60 90 120 150 180 210 240 270 300 330
x [ft] x [ft]

c) 4 spans d) 5 spans

Figure 22 – Comparison between theoretical and experimental longitudinal deflections (deck

pouring, Girder 5)

20
2.3 Load test

2.3.1 Test procedure

The load test was carried out using six fully loaded 10-wheel three-axle dump trucks. Axles
spacing and width are shown in Figure 23.

5' 15'

5' 11"

6' 11"
P2 P2 P1

Rear axles Front axle

(a) Trucks on site (b) Spacing and width of axles

Figure 23 – Dump trucks used for load test

The trucks were weighed before testing and coded with a number from 1 to 6. Table 2.2
summarizes weight and assumed load distribution between front and rear axles.

Table 2.2 – Dump trucks used for load test: weight distribution assumed for analysis

Truck Total Weight Front Axle ( P1 ) Each Rear Axle ( P2 )

No. (kips) (kips) (kips)
1 45.4 17.0 14.2
2 43.3 16.2 13.5
3 53.1 19.9 16.6
4 54.8 20.6 17.1
5 60.2 22.6 18.8
6 56.4 21.2 17.6

A total of eight stops were planned, and details are reported in Appendix I. Marks were made on
the concrete deck to indicate the trucks stops, and the distance between the front axle of each
truck and the rear axle of the truck in front in the line load was 9 ft.
Stop 1, 2 and 7 were intended to produce the maximum deflections on Girder 1 and 5 using both
directions of traffic with two symmetrical lanes of two (Stop 1) and three (Stop 2 and 7) trucks
each. Stop 3, 4 and 6 were intended to produce the maximum deflections on Girder 5 in Span 3,

21
2 and 1, respectively, using a train of six trucks. Stop 5 was planned to induce the maximum
negative moment on Girder 5, at Pier 2. Finally, Stop 8 was intended to produce the maximum
deflections on Girder 1 in Span 1 using a single lane of six trucks. Figure 24 shows the
transverse position of the trucks for the different stops.

3' 5' 11"

Girder 1 Girder 2 Girder 3 Girder 4 Girder 5

(a) Stop 8
3' 5' 11" 5' 11" 3'

Girder 1 Girder 2 Girder 3 Girder 4 Girder 5

(b) Stop 1, 2 and 7
5' 11" 3'

Girder 1 Girder 2 Girder 3 Girder 4 Girder 5

(c) Stop 3, 4, 5 and 6
Figure 24 – Transverse position of trucks during load test (not to scale)

In order to minimize the influence of systematic errors in the displacement measurements, and
during the load test only, the TS was mounted on a leveling steel plate fixed on a concrete pile
(Figure 25).
Once the TS was leveled and acclimatized, initial readings were taken for each prism. Then, the
trucks drove to the first stop. At each stop, 15 minutes were allowed for potential settlements,
before acquiring the data. To ensure stable measurements, four readings were taken for each
target, two direct and two inverse, in order to average out possible errors. Once the reading was
terminated, the trucks moved to the next stop and the same procedure was repeated.

22
 Leveling steel plate

Concrete pile

Figure 25 – Total station in load test configuration

2.3.2 Test results

Figure 26, Figure 27 and Figure 28 show the longitudinal deflections of Girder 1, Girder 3 and
Girder 5, respectively, corresponding to all the eight stops during the load test, while Figure 29
summarizes the transversal distribution of deflections at the mid-span between Abutment 1 and
Pier 2. Data from the same stop have been piecewise connected with straight lines for clarity.
Raw data are provided in Appendix II.a.

1.0
Bolted joint

Bolted joint

Bolted joint
Abutment 1

0.5

0.0 Stop 1 - Exp.

Deflection [in]

Stop 2 - Exp.
-0.5 Stop 3 - Exp.
Stop 4 - Exp.
-1.0 Stop 5 - Exp.
Stop 6 - Exp.
-1.5 Stop 7 - Exp.
Pier 2

Pier 3

Stop 8 - Exp.
-2.0
0 30 60 90 120 150 180 210 240 270 300 330
x [ft]

Figure 26 – Experimental longitudinal deflection (load test, Girder 1)

23
 1.0

Stop 1 - Exp.
0.5 Stop 2 - Exp.
Stop 3 - Exp.
0.0 Stop 4 - Exp.
Deflection [in]

Stop 5 - Exp.
-0.5 Stop 6 - Exp.
Stop 7 - Exp.
-1.0 Stop 8 - Exp.

Bolted joint
Bolted joint

Bolted joint
Abutment 1

-1.5
Pier 2

Pier 3
-2.0
0 30 60 90 120 150 180 210 240 270 300 330
x [ft]

Figure 27 – Experimental longitudinal deflections (load test, Girder 3)

1.0

Stop 1 - Exp.
0.5 Stop 2 - Exp.
Stop 3 - Exp.
0.0 Stop 4 - Exp.
Deflection [in]

Stop 5 - Exp.
-0.5 Stop 6 - Exp.
Stop 7 - Exp.
-1.0 Stop 8 - Exp.
Bolted joint
Bolted joint

Bolted joint
Abutment 1

-1.5
Pier 2

Pier 3

-2.0
0 30 60 90 120 150 180 210 240 270 300 330
x [ft]

Figure 28 – Experimental longitudinal deflections (load test, Girder 5)

24
 1.5
Stop 1 Stop 2 Stop 3 Stop 4
1.0
Stop 5 Stop 6 Stop 7 Stop 8
0.5
Deflection [in]

0.0

-0.5

-1.0

-1.5

-2.0
1 2 3 4 5
Girder No.

Figure 29 – Experimental transversal deflections at midspan of Span 1 (load test)

It can be observed that the readings are consistent from stop to stop, while the curves generally
exhibit a smooth transition from point to point. However, a comparison between the deflections
of Girder 1 and 5 during Stop 8 and 6, points out that the first reached values 28% higher when
subjected to absolutely similar load conditions, as shown in Figure 30. Such difference cannot be
explained by accounting for possible concurrent factors that played a role in determining the
measurement results. A preliminary, yet reasonable explanation may lie in the fact that Girder 1
fell during construction (Figure 31), due to the action of wind, and likely experienced some
damage either during the falling or re-positioning. This calls for further verification by means of
the results from BOTDR based strain measurement, along with a comparison with analytical and
numerical predictions.

1.0
Abutment 1

Bolted joint

Bolted joint
joint
Bolted
Pier 2

Pier 3

0.5

0.0
Deflection [in]

-0.5 Girder 1, Stop 8 - Exp.

Girder 5, Stop 6 - Exp.

-1.0

28%
-1.5

-2.0
0 30 60 90 120 150 180 210 240 270 300 330
x [ft]

Figure 30 – Experimental deflections for Girder 1, Stop 8 and Girder 5, Stop 6 (load test)

25
 Figure 31 – Girder 1 fallen due to wind during erection of bridge superstructure

2.3.3 Discussion of results

The experimental deflections of Girder 1, 2, 4 and 5 were compared with results based on 1-D
beam analysis. The girder distribution factors, GDF (defined as the ratio of the maximum
moment on a girder at a given live load condition, and the maximum theoretical moment
determined by applying the entire load truck to a single composite girder), were determined
according to the AASHTO Bridge Design Specifications (1998), [Table 4.6.2.2.2d-1]. A multiple
presence factor m = 1 was assumed when using the lever rule. Adjustments in the negative
moment region near the interior support (Pier 2) due to continuity effects have been neglected.
Details of calculations are summarized in Table 2.3 and Table 2.4.

Table 2.3 – GDFs as per AASHTO LRFD Bridge Design Specifications (1998)
Girder No. of Lanes AASHTO GDF GDF
No. Loaded Formula Span 1 Span 2
1, 5 1 Lever rule 0.658 0.658
0.4 0.3 0.1
⎛ S ⎞ ⎛ S ⎞ ⎛ Kg ⎞
2, 4 1 GDF = 0.06 + ⎜ ⎟ ⎜ ⎟ ⎜ 3⎟ 0.432a 0.406a
⎝ 168 ⎠ ⎝ L ⎠ ⎝ Ltc ⎠
de
1, 5 2c GDF = e ⋅ GDF2,4 , e = 0.77 + ≥1 0.641a,b 0.613a,b
109.2
0.6 0.2 0.1
c ⎛ S ⎞ ⎛S⎞ ⎛ Kg ⎞
2, 4 2 GDF = 0.075 + ⎜ ⎟ ⎜ ⎟ ⎜ 3⎟ 0.641a 0.613a
⎝ 114 ⎠ ⎝L⎠ ⎝ Ltc ⎠
a
See Table 2.4 for parameters and units
b
de = distance from center of exterior beam and inside edge of barrier = 36 – 16 = 20 in, therefore e = 1
c
In the case of two load lanes (Figure 24b), the closer to the girder considered was assumed for computation

26
 Table 2.4 – Input data for calculation of GDFs (AASHTO, 1998)

Value Value
Parameter (unit) Definition
(Span 1) (Span 2)
S (in) Girder spacing 104 104
L (in) Span length 1,764 2,220
Kg (in4) Longitudinal stiffness parameter 1.836·106 2.189·106

=
Es
Ec
( I s + As eg2 )

Es (ksi) Modulus of elasticity of steel (girder) 29,000 29,000

Ec (ksi) Modulus of elasticity of concrete (deck) 3,605 3,605
Is (in4) Moment of inertia of steel girder section 70,879 88,245
As (in2) Area of steel girder section 71.25 83.25
eg (in) Distance between center of gravity of 47.0 47.0
girder section and concrete deck
tc (in) Thickness of concrete deck 8.5 8.5

The width of the concrete slab section was selected as the average spacing of adjacent beams
[4.6.2.6.1], i.e. S = 104 in, for both interior and exterior beams. It is noted that the contribution of
the structurally continuous concrete barrier has been accounted for [C4.6.2.6.1-1], thus
determining a theoretical flange width for the external girder as

S A 104 328
+ woverhang + b + 36 + 107.3 in,
2 2tc 2 2 ⋅ 8.5

wherein woverhang is the width of the overhang (in), Ab the cross sectional area of the barrier (in2),
and tc the deck thickness (in). Since the difference between the flexural stiffness of the
equivalent interior and exterior girders is of the order of 0.5%, the 104 in flange width was used
for all the beams.
Figure 32 and Figure 33 show the comparison between the deflections measured on Girder 1 and
5, respectively, and the analytical deformed shapes. The deflections measured through the ATS
system are generally smaller than the theoretical values, and clearly well below the L / 800 limit
[2.5.2.6.2] of 2.2 in and 2.8 in on Span 1 and Span 2, respectively, as shown in Figure 32d and
Figure 33d and f, thereby demonstrating the safety of the structure. Nevertheless, comparison of
Figure 32d and Figure 33d and f points out the tendency of Girder 1 to approach the theoretical
limits, differently from Girder 5, when directly subjected to essentially the same load conditions.
As previously mentioned, the falling of the girder occurred during construction may have
determined some alteration of the structural response, which may call for further assessment.

27
Conversely, the response of Girder 5 confirms the conservativeness of the design provisions.

2.0 2.0
1.5 AASHTO 1.5 AASHTO
1.0 Experimental 1.0 Experimental
0.5 0.5
Deflection [in]

Deflection [in]
0.0 0.0
-0.5 -0.5
-1.0 -1.0
-1.5 -1.5
Bolted joint

Bolted joint

Bolted joint

Bolted joint
Abutment 1

Bolted joint

Abutment 1

Bolted joint
-2.0 -2.0
Pier 2

Pier 3

Pier 2

Pier 3
-2.5 -2.5
-3.0 -3.0
0 30 60 90 120 150 180 210 240 270 300 330 0 30 60 90 120 150 180 210 240 270 300 330
x [ft] x [ft]

a) Stop 1 b) Stop 2
2.0 2.0

Abutment 1

Bolted joint

Pier 2
1.5 AASHTO 1.5
1.0 Experimental 1.0
0.5 0.5
Deflection [in]

Deflection [in]

0.0 0.0 AASHTO

-0.5 -0.5 Experimental
-1.0 -1.0 L/800 Span 1
-1.5 -1.5

Bolted joint
Bolted joint
Bolted joint

Bolted joint
Abutment 1

Bolted joint

-2.0 -2.0

Pier 3
Pier 2

Pier 3

-2.5 -2.5
-3.0 -3.0
0 30 60 90 120 150 180 210 240 270 300 330 0 30 60 90 120 150 180 210 240 270 300 330
x [ft] x [ft]

c) Stop 7 d) Stop 8
Figure 32 – Comparison between AASHTO provisions and experimental longitudinal deflections
(load test, Girder 1)

28
 2.0 2.0
1.5 AASHTO 1.5 AASHTO
1.0 Experimental 1.0 Experimental
0.5 0.5
Deflection [in]

Deflection [in]
0.0 0.0
-0.5 -0.5
-1.0 -1.0
-1.5 -1.5
Bolted joint

Bolted joint

Bolted joint

Bolted joint
Abutment 1

Abutment 1
Bolted joint

Bolted joint
-2.0 -2.0

Pier 2

Pier 3

Pier 2

Pier 3
-2.5 -2.5
-3.0 -3.0
0 30 60 90 120 150 180 210 240 270 300 330 0 30 60 90 120 150 180 210 240 270 300 330
x [ft] x [ft]

a) Stop 1 b) Stop 2
2.0 2.0

Bolted

Bolted

Pier 3
joint

joint
1.5 AASHTO 1.5
1.0 Experimental 1.0
0.5 0.5 AASHTO
Deflection [in]

Deflection [in]
0.0 0.0 Experimental
-0.5 -0.5 L/800 Span 2

-1.0 -1.0
-1.5 -1.5
Bolted joint

Bolted joint

Bolted joint
Abutment 1

Abutment 1
Bolted joint

-2.0 -2.0
Pier 2

Pier 3

Pier 2
-2.5 -2.5
-3.0 -3.0
0 30 60 90 120 150 180 210 240 270 300 330 0 30 60 90 120 150 180 210 240 270 300 330
x [ft] x [ft]

c) Stop 3 d) Stop 4
2.0 2.0
Abutment 1

joint

joint

joint
Pier 2
Bolted

Bolted

Bolted
1.5 AASHTO 1.5
1.0 Experimental 1.0
0.5 0.5
Deflection [in]

Deflection [in]

0.0 0.0
-0.5 -0.5
-1.0 -1.0 AASHTO
-1.5 -1.5 Experimental
Bolted joint
Bolted joint

Bolted joint
Abutment 1

-2.0 -2.0 L/800 Span 1

Pier 3
Pier 2

Pier 3

-2.5 -2.5
-3.0 -3.0
0 30 60 90 120 150 180 210 240 270 300 330 0 30 60 90 120 150 180 210 240 270 300 330
x [ft] x [ft]

e) Stop 5 f) Stop 6
2.0
1.5 AASHTO
1.0 Experimental
0.5
Deflection [in]

0.0
-0.5
-1.0
-1.5
Bolted joint

Bolted joint
Bolted joint
Abutment 1

-2.0
Pier 2

Pier 3

-2.5
-3.0
0 30 60 90 120 150 180 210 240 270 300 330
x [ft]

g) Stop 7
Figure 33 – Comparison between AASHTO provisions and experimental longitudinal deflections
(load test, Girder 5)

29
Since all the beams considered for the 1-D analysis were assumed to have essentially the same
geometrical and material characteristics, i.e. an equivalent flexural stiffness, the GDF for the ith
girder at the midspan in Span 1 was estimated from the deflection data retrieved as

5
GDFi = δ i / ∑δ
j =1
j ,

wherein δi = deflection of the ith girder. Figure 34 and Figure 35 shows the distribution factors
calculated at each stop with one and two lanes loaded, respectively.

0.8
AASHTO LRFD Limit for Girder 1 and 5 = 0.658
Load Distribution Factor [Load Lanes]

0.6
AASHTO LRFD Limit for Girder 2, 3 and 4 = 0.432
0.4

0.2

0.0

Stop 3 Stop 4 Stop 5

-0.2
Stop 6 Stop 8

-0.4
1 2 3 4 5
Girder No.

Figure 34 – Comparison between AASHTO and experimental girder load distribution factor at
midspan between Abutment 1 and Pier 2 in case of one lane loaded (Stop 3, 4, 5, 6 and 8)

0.8
Load Distribution Factor [Load Lanes]

AASHTO LRFD Limit for All Girders = 0.641

0.6

0.4

Stop 1
0.2 Stop 2
Stop 7

0.0
1 2 3 4 5
Girder No.

Figure 35 – Comparison between AASHTO and experimental girder load distribution factor at
midspan between Abutment 1 and Pier 2 in case of two lanes loaded (Stop 1, 2 and 7)

30
The load distribution factors determined experimentally at the midspan section between
Abutment 1 and Pier 2 (Figure 12) are always smaller than those based on the AASHTO
provisions, which are proved to be conservative for the case under consideration. The main
reason typically lies in the inherent conservativeness of neglecting the transverse stiffening
action of the diaphragms when using either the design formulas, wherein the thickness of the
deck is the only parameter that explicitly accounts for a degree of transverse stiffness in the
structural system, or the lever rule.

31
 3. BOTDR BASED STRAIN MEASUREMENT: FIELD EVALUATION

3.1 Bridge instrumentation

3.1.1 Installation of optical fibers

The sensing elements were installed on Girder 1, 2, 4 and 5. Both the bare optical fibers and the
FRP tape with embedded sensing fibers were adhesively bonded onto the girder webs using a
two-part epoxy encapsulation resin, after manual surface roughening with steel brushes and
degreasing with lacquer thinner. Table 3.1 summarizes the properties of the resin, as provided by
the manufacturer (Wabo, 2004). Detailed location and characteristics of the fibers are
summarized in Figure 36, Figure 37 and Table 3.2.

Table 3.1 – Properties of Wabo MBrace saturant for fiber optics installation (Wabo, 2004)

Property (Unit) Value

Tensile strength (psi) 8,000
Tensile strain (%) 3.5
Tensile modulus (ksi) 440
Poisson’s ratio 0.4
Compressive strength (psi) 12,500
Compressive strain (%) 5
Compressive modulus (ksi) 380
Coefficient of thermal expansion (ºF-1) 20·10-6
Glass transition temperature (ºF) 163

bare fibers
Smart GFRP
tape

Span 1 Span 2
Abutment 1 Pier 2 Pier 3
Figure 36 – Schematic of fiber circuit placement

32
 3 bare fibers
C on Girder 5
Strain sensing fiber

53 Thermal compensation fiber

h web =
84
bare fibers
B on Girder 2 4 3.1 "Smart" FRP tape
hweb /3 =
28 "smart" FRP tape
A on Girder 1, 2, 4 and 5

Figure 37 – Detailed location of fiber optics on girder web (not to scale, dimensions in in)

Table 3.2 – Type, identification and location of sensing elements installed

Number of Fibers × Circuit

Girder Span Location Fiber type
Active length (ft) ID
(Instrumented tape)
tight buffer fiber 2 × 148 Ia, Ib
1 1 A
loose (dry) buffer fiber 1 × 148 Ic
loose (wet) buffer fiber 1 × 148 Id
(Bare fibers)
B
2 1, 2* loose (dry) buffer fiber 1 × (148, 137) IIa
(hweb/3 = 28 in)
tight buffer fiber 1 × (148, 137) IIb
(Instrumented tape)
2 1, 2* A tight buffer fiber 1 × (148, 137) IIIa
loose (dry) buffer fiber 1 × (148, 137) IIIb
(Instrumented tape)
4 1, 2* A
tight buffer fiber 1 × (148, 137) IVa
(Instrumented tape)
C
5 1 loose (dry) buffer fiber 1 × 148 Va
(hweb – 3 = 81 in)
tight buffer fiber 1 × 148 Vb
(Instrumented tape)
5 1, 2* A tight buffer fiber 1 × (148, 137) VIa
loose (dry) buffer fiber 1 × (148, 137) VIb
* fibers are installed up to bolted joint prior to Pier 3

33
The smart GFRP tape was applied over the web plate of all the girders monitored. The tape was
drawn through the corner holes between the welded vertical stiffeners and the plate girder and
then installed about 0.4 in above the bottom flange, in order to allow detection of a strain level
reasonably close to the actual maximum attained.
Girder 1 was instrumented on Span 1 using smart tape. The tape circuit end was also lengthened
for 32.8 ft after Pier 2. The additional FO installed on the web, at a distance of 15 in from the
adjacent surface of the top flange (Figure 38), was used to gain additional data and therefore to
estimate the position of the neutral axis. Since it was not possible to keep the tape perfectly flat
in the vicinity of the cross frames, a certain disturbance of the data retrieved in those areas was
expected.
Girder 2 was instrumented on Span 1 and 2. The smart tape was installed at 0.4 in from the
bottom flange surface. In addition, two bare fibers, one for strain sensing and the other for
thermal compensation, were bonded at 28 in from the bottom flange (i.e. at one third of the
height of the plate web), (see Figure 39a). This was done in an attempt to evaluate the position of
the neutral axis; however, the need to draw the fibers through the holes between the welded
vertical stiffeners and the plate girder required several sharp bending areas of the fibers, as
shown in Figure 39b, which are believed to be the cause of an undesired increase of the optical
losses and significant perturbations of the strain distribution measured.
Girder 4 was instrumented on Span 1 and 2 with FRP tape only, bonded at the web base.
Girder 5 was instrumented on Span 1 and 2 with optical fiber-embedded composite tape at the
web base. In addition, two bare fibers, one for strain sensing and the other for thermal
compensation, were installed at 3 in from the upper flange surface (Figure 40), in order to
retrieve data to estimate the temperature gradient throughout the web height, and of the strain
profile, especially in the negative moment zone.

“Smart” GFRP tape

Rubber band

Figure 38 – Smart FRP installed on Girder 1 and folded at Pier 2 location to extend circuit

34
 a) Smart FRP (lower blue stripe) and bare fibers (upper b) Bending of fibers to pass
blue stripe) through transversal stiffeners

Figure 39 – Sensing fibers installed onto web plate of Girder 2

Figure 40 – Sensing fibers installed onto web plate of Girder 5

3.1.2 Modular cart for sensor installation

The fact that the 863 ft long bridge spans over the Osage River for about 571 ft, while the
concrete deck had not been cast yet, precluded the use of a cherry-picker or any such vehicle to
allow the personnel to access the desired locations along the steel girders and install the sensors.
This was accomplished by the aid of a modular frame cart, designed on-purpose at the University
of Missouri-Rolla, and built at the Rolla Technical Institute (Figure 41). The cart was conceived
to move along the whole bridge rolling over the bottom flanges of two adjacent girders, using a
set of built-in devices to by-pass the vertical stiffeners, cross frames and diaphragms at the pile
locations. The vehicle is composed of three 5 ft wide and 4 ft long demountable modules, for a
total working surface of about 60 ft2. The total weight of the cart was limited to 423 lbs by
extensive use of 6061-T6 aluminum profiles bolted together to form the frame, and structural
GFRP pultruded planks manufactured by Strongwell Corporation (2004) as the flooring system.
The transversal support members are made of AISI 1018 steel plates, which were cut to length
and welded, and contain the AISI 1045 steel axles, with diameter up to 1.5 in. 3.25 in by 3.25 in
polyurethane lift truck wheels with bearings were secured at one end of each axle by means of

35
safety pins, while a total of eight electric actuators allowed the axles to extend out of the
transverse member for 18.5 in on each side of the vehicle, as shown in Figure 41c.

a) Schematic rendering b) Cart frame as built

PTFE lift
truck wheel

Transverse stiffener
Electrically-actuated
steel axle

c) Electrically-actuated axles for obstacle by-passing

d) Along Girder 1/2, Span 2 e) Installation operations

Figure 41 – Modular frame cart for sensor installation

3.1.3 Connection and preliminary testing of the sensing elements

Upon completion of the installation procedure, the different fiber hauls were tested with the
AQ8603 strain analyzer in coherent OTDR loss measurement mode. The check-up evidenced

36
some unwanted sources of optical loss, likely due to fiber damages which occurred during
installation. The amount of optical losses in the sensing circuit is of crucial importance in
BOTDR systems, since it may be highly detrimental to the strain sensitivity and accuracy. The
hauls were series-connected in the field using a fiber optic fusion splicer (Figure 42), obtaining a
single circuit with a total length of 3,802 ft. The solution allowed containing the total optical loss
within a value of 6 dB in the first 3,366 ft from the optical test end (Figure 43), thus enabling
strain measurement with a minimum accuracy of 0.004% (±40 µε) on a length resolution of 6.5
ft, setting the AQ8603 analyzer for laser pulses with a length of 20 ns. Since a significant optical
loss is detected before the last portion of the circuit, i.e. Girder 4, Location A, the correspondent
measured strain profiles were expected to have a lower degree of accuracy.

Figure 42 – Optical fiber core-alignment fusion splicing field operation

46

44
Optical loss @ 1550 nm (dB)

6 dB @ 3,366 m
42

40

38

36

34
0 300 600 900 1,200 1,500 1,800 2,100 2,400 2,700 3,000 3,300 3,600 3,900

Distance (ft)

Figure 43 – Optical loss measured along sensing circuit

37
3.1.4 Sensing system calibration
The correlation between the strain and the Brillouin frequency shift (characteristic of the fiber
used), was determined using a strain calibration fixture that allowed to induce a predetermined
uniform strain in the fiber tested, over a reference length of 24.34 ft (Figure 44a).
Using the strain data collected with the calibration fixture (Figure 44b), a specific set of
correction coefficients was computed for each type of fiber used, in order to improve accuracy
and linearity with respect to the standard equipment settings. The best linearization was obtained
for the sensing fiber embedded into the custom designed and manufactured GFRP tape, as shown
in Figure 45.
In addition, since the AQ8603 unit does not allow direct temperature estimate, a suitable strain-
temperature correlation was established by testing an optical fiber spool placed in an
environmental chamber at different temperature levels. The results are summarized in Figure 46.

a) Schematic

b) Data recording using AQ8603 strain analyzer

Figure 44 – Strain calibration fixture

38
 1,300
Pirelli
Pirelli tight ∅ 900 bare
tight 900um µm B
1,100 Pirelli tight ∅ 900 µm T
Pirelli tight 900um tape
Pirelli
Pirelli tight ∅ 900
tight tape µm T (corr.)
corrected
Corning ∅ 900 µbare
Corning tight 900um mB
900
semiloose 900 µ
∅
Measured strain (µε)

H&S semiloose 900um tape

mT
Pirelli loose
Pirelli ∅ 900 bare
loose 900um µm B
700 Pirelli loose
Pirelli ∅ 900 tape
loose 900um µm T
imposed strain
Imposed strain
500

300

100

-100
0 200 400 600 800 1,000
Imposed strain (µε)

Figure 45 – Linearity of measured and imposed strain for different fibers tested

10,000

8,000
Experimental
Apparent strain (µε)

Least square interpolation

6,000

4,000

2,000

0
20 30 40 50 60 70 80 90 100 110 120 130
o
Fiber temperature ( F)

Figure 46 – Strain-temperature correlation

3.1.5 Installation of resistive strain gauges

A total of four resistive strain gauges were installed on Girder 1 and 2, in order to have a strain
detection reference independent from the BOTDR system, while three additional gauges were
used to evaluate the thermal drift compensation (Figure 47). Table 3.2 summarizes the type and

39
location of the strain gauges installed.

Table 3.3 – Type, identification and location of strain gauges installed

Gauge Distance from

Type* Girder Location
ID Abutment 1 (ft)

G1 TC 1 Top flange 65.5

G2 SM 1 Top flange 66.5
G3 SM 1 Bottom flange 65.5
G4 TC 1 Bottom flange 65.5
G5 SM 2 B (hweb / 3) 75.5
G6 SM 2 Bottom flange 75.0
G7 TC 2 Bottom flange 75.0
* TC = thermal compensation; SM = strain measurement

Figure 47 – Close-up of strain gauges installed on top flange of Girder 1

3.2 Test procedure

The load test procedure is described in 2.3.1. The equipment for strain measurement and
recording is shown in Figure 48. The raw data exported by the strain analyzer were elaborated
through proprietary software (Figure 49), in order to properly account for the distributed thermal
compensation, thereby retrieving the experimental load-induced strain profile along each
monitored girder. Strain data from the strain gauges were recorded continuously at a frequency
of 1 Hz, using the data acquisition system shown in Figure 11.

40
 Figure 48 – Field equipment for distributed strain measurement

Figure 49 – Proprietary software for strain data processing

3.3 Test results

3.3.1 BOTDR strain measurement

The experimental strain profiles along the monitored girders during the load test at Stops 1, 2, 4,
7 and 8 are shown in Figure 50 to Figure 54, along with the strain gauge readings detailed in
Section 3.3.2. The data from the remaining stops are currently not available due to a temporary
hardware problem. The strain profile history for each monitored girder throughout the load test is
provided in Appendix II.b.

41
 -300 -300

Bolted joint

Bolted joint
-200 -200

Pier 3
-100 -100

0 0
Strain [µε]

Strain [µε]
100 100 Girder 4 - bottom - Exp.
Girder 1 - bottom - Exp.
Girder 5 - bottom - Exp.
200 Girder 2 - bottom - Exp. 200
Girder 5 - top - Exp.
Girder 2 - h/3 - Exp.
300 G3 (Girder 1 - bottom) 300

Bolted joint
Abutment 1

Abutment 1
Bolted joint

Bolted joint

Bolted joint
G2 (Girder 1 - top)
400 400

Pier 3
Pier 2

Pier 2
G5 (Girder 2 - h/3)

500 500
0 30 60 90 120 150 180 210 240 270 300 330 0 30 60 90 120 150 180 210 240 270 300 330
x [ft] x [ft]

a) Girder 1 and 2 b) Girder 4 and 5

Figure 50 – Experimental strain distribution (Stop 1)

-300 -300
Bolted joint

Bolted joint

-200 -200
Pier 3

-100 -100

0 0
Strain [µε]

Strain [µε]

100 100 Girder 4 - bottom - Exp. (low accuracy)

Girder 1 - bottom - Exp. Girder 5 - bottom - Exp.
200 200
Girder 2 - bottom - Exp.
Girder 5 - top - Exp.
Girder 2 - h/3 - Exp.
300 300
G3 (Girder 1 - bottom)
Abutment 1

Bolted joint

Bolted joint

Bolted joint
Abutment 1

Bolted joint

400 G2 (Girder 1 - top) 400

Pier 2

Pier 3
Pier 2

G5 (Girder 2 - h/3)
500 500
0 30 60 90 120 150 180 210 240 270 300 330 0 30 60 90 120 150 180 210 240 270 300 330
x [ft] x [ft]

a) Girder 1 and 2 b) Girder 4 and 5

Figure 51 – Experimental strain distribution (Stop 2)

-300

-200 Girder 5 - bottom - Exp.

-100 Girder 5 - top - Exp.

0
Strain [µε]

100

200

300
Abutment 1

Bolted joint
Bolted joint

Bolted joint

400
Pier 3
Pier 2

500
0 30 60 90 120 150 180 210 240 270 300 330
x [ft]

Figure 52 – Experimental strain distribution on Girder 5 (Stop 4)

42
 -300

Bolted joint

Bolted joint
-200

Pier 3
-100

Strain [µε]
100 Girder 1 - bottom - Exp.
Girder 2 - bottom - Exp.
200
Girder 2 - h/3 - Exp.

300 G3 (Girder 1 - bottom)

Abutment 1

Bolted joint
G2 (Girder 1 - top)
400
G5 (Girder 2 - h/3)

Pier 2
500
0 30 60 90 120 150 180 210 240 270 300 330
x [ft]

a) Girder 1 and 2
-300

-200

-100

0
Strain [µε]

100 Girder 4 - bottom - Exp.

Girder 5 - bottom - Exp.

200
Girder 5 - top - Exp.
300
Bolted joint
Abutment 1

Bolted joint

Bolted joint
400

Pier 3
Pier 2

500
0 30 60 90 120 150 180 210 240 270 300 330
x [ft]

b) Girder 4 and 5

Figure 53 – Experimental strain distribution (Stop 7)

-300
Bolted joint

Bolted joint

-200
Pier 3

-100

0
Strain [µε]

100 Girder 1 - bottom - Exp.

Girder 2 - bottom - Exp.
200
Girder 2 - h/3 - Exp.

300 G3 (Girder 1 - bottom)

Abutment 1

Bolted joint

G2 (Girder 1 - top)
400
G5 (Girder 2 - h/3)
Pier 2

500
0 30 60 90 120 150 180 210 240 270 300 330
x [ft]

Figure 54 – Experimental strain distribution on Girder 1 and Girder 2 (Stop 8)

43
3.3.2 Resistive strain gauge measurements
Of the seven strain gauges installed, only G1, G2, G3, G4 and G5 worked properly during the
load test. The strain reading history throughout the test and the values at each stop, computed
accounting for thermal compensation, are provided in Figure 55 and Figure 56, respectively.

Strain (µε)

Time (sec)

Figure 55 – Time history of strain gauge readings

350

300 Girder 1 - Bottom flange (G3)

Girder 1 - Top flange (G2)
250
Girder 2 - B (G5)
200
Strain (µε)

150

100

50

-50

-100
1 2 3 4 5 6 7 8
Stop #

Figure 56 – Strain gauge data computed for Stop 1-8 accounting for thermal compensation

3.4 Discussion of results

The experimental results were compared with analytical strain profiles based on the AASHTO
LRFD Bridge Design Specifications (AASHTO, 1998), according to the theoretical approach
previously detailed in 2.3.3, pp. 27-28. Figure 57, Figure 58, Figure 59, Figure 60 and Figure 61
show the comparison between the experimental and analytical strain profiles at Stop 1, Stop 2,
Stop 4, Stop 7 and Stop 8, respectively. Strain gauge readings are also included.

44
 -150

Bolted joint

Bolted joint
-100

Pier 3
-50

0
Strain [µε]

50 Girder 1 - bottom - Exp.

Girder 1 - bottom - AASHTO
100
Girder 2 - bottom - Exp.

150 Girder 2 - bottom - AASHTO

Girder 2 - h/3 - Exp.
200 Girder 2 - h/3 - AASHTO
Abutment 1

Bolted joint
G3 (Girder 1 - bottom)
250

Pier 2
G5 (Girder 2 - h/3)
300
0 30 60 90 120 150 180 210 240 270 300 330
x [ft]
a) Girder 1 and 2
-150
Bolted

Bolted

Pier 3
joint

joint
-100

-50

0
Strain [µε]

50 Girder 4 - bottom - Exp.

100 Girder 4 - bottom - AASHTO

Girder 5 - bottom - Exp.

150
Girder 5 - bottom - AASHTO
200 Girder 5 - top - Exp.
Abutment 1

Bolted joint

250 Girder 5 - top - AASHTO

Pier 2

300
0 30 60 90 120 150 180 210 240 270 300 330
x [ft]
b) Girder 4 and 5

Figure 57 – Comparison between experimental and predicted strain profiles (Stop 1)

45
 -150

Bolted

Bolted joint
joint
-100

Pier 3
-50

0
Strain [µε]

50
Girder 1 - bottom - Exp.
100 Girder 1 - bottom - AASHTO
Girder 2 - bottom - Exp.
150 Girder 2 - bottom - AASHTO
Girder 2 - h/3 - Exp.
200
Girder 2 - h/3 - AASHTO
Abutment 1

Bolted joint

250 G3 (Girder 1 - bottom)

Pier 2
G5 (Girder 2 - h/3)
300
0 30 60 90 120 150 180 210 240 270 300 330
x [ft]
a) Girder 1 and 2
-150
Abutment 1

-100

-50

0
Strain [µε]

50 Girder 4 - bottom - Exp.

Girder 4 - bottom - AASHTO
100
Girder 5 - bottom - Exp.
150 Girder 5 - bottom - AASHTO
Girder 5 - top - Exp.
200 Girder 5 - top - AASHTO
Bolted joint

250
Bolted

Bolted

Pier 3
Pier 2

joint

joint

300
0 30 60 90 120 150 180 210 240 270 300 330
x [ft]
b) Girder 4 and 5

Figure 58 – Comparison between experimental and predicted strain profiles (Stop 2)

46
 -300
Abutment 1

-200

-100

0
Strain [µε]

100
Girder 5 - bottom - Exp.
200 Girder 5 - bottom - AASHTO
Girder 5 - top - Exp.
300
Girder 5 - top - AASHTO
Bolted joint

Bolted joint
400
Bolted

Pier 3
Pier 2
joint

500
0 30 60 90 120 150 180 210 240 270 300 330
x [ft]

Figure 59 – Comparison between experimental and predicted strain profiles (Stop 4)

47
 -150

Bolted

Bolted joint
joint
-100

Pier 3
-50

0
Strain [µε]

50 Girder 1 - bottom - Exp.

Girder 1 - bottom - AASHTO
100
Girder 2 - bottom - Exp.
150 Girder 2 - bottom - AASHTO
Girder 2 - h/3 - Exp.
200 Girder 2 - h/3 - AASHTO
Abutment 1

Bolted joint
G3 (Girder 1 - bottom)
250

Pier 2
G5 (Girder 2 - h/3)
300
0 30 60 90 120 150 180 210 240 270 300 330
x [ft]
a) Girder 1 and 2
-150

Bolted joint
Bolted
joint

-100

Pier 3
-50

0
Strain [µε]

50
Girder 4 - bottom - Exp.
100
Girder 4 - bottom - AASHTO
150 Girder 5 - bottom - Exp.
Girder 5 - bottom - AASHTO
200
Abutment 1

Bolted joint

Girder 5 - top - Exp.

250
Pier 2

Girder 5 - top - AASHTO

300
0 30 60 90 120 150 180 210 240 270 300 330
x [ft]
b) Girder 4 and 5

Figure 60 – Comparison between experimental and predicted strain profiles (Stop 7)

48
 -300

Bolted joint
Bolted
joint
-200

Pier 3
-100

0
Strain [µε]

Girder 1 - bottom - Exp.

100
Girder 1 - bottom - AASHTO
Girder 2 - bottom - Exp.
200
Girder 2 - bottom - AASHTO

300 Girder 2 - h/3 - Exp.

Girder 2 - h/3 - AASHTO
Abutment 1

Bolted joint
400 G3 (Girder 1 - bottom)

Pier 2
G5 (Girder 2 - h/3)
500
0 30 60 90 120 150 180 210 240 270 300 330
x [ft]

Figure 61 – Comparison between experimental and predicted strain profiles (Stop 8)

The strain readings from the optical fibers embedded in the GFRP tape fairly follow the
analytical profiles in all the girders in case of one load lane (Stop 4 and 8), and in the exterior
members, Girder 1 and Girder 5, in case of two load lanes (Stop 1, 2 and 7), when the
distribution factor in the interior girders is not representative of the actual load test conditions.
The offset is generally consistent with the strain gauge readings and the deflection results from
ATS measurements, thus verifying the conservativeness of the AASHTO design provisions
(1998).
As expected, the strain measurement on Girder 4, Location A (adjacent to the bottom flange),
yields less accurate values, due to the significant optical loss before the interested portion of the
circuit. The experimental results also point out the relation between measurement accuracy and
strain level, with reliable outputs generally provided at strain levels beyond 100 µε.
The BOTDR experimental data confirm the concerns raised by the ATS measurements on Girder
1. In particular, the strain profile at Stop 8 (Figure 61a) shows the clear tendency to approach the
theoretical values, in accordance with the deflection results shown in Figure 32d This indicates
that the structural member deforms as under a vertical load significantly close to that associated
with the theoretical conservative distribution factor of 0.658 load lanes. Conversely, the strain in
Girder 2 at Stop 8 (Figure 61a) remains well below the analytical threshold.
Local effects, such as that at the bolted joints, may result in disturbances that affect the strain
readings in a relatively large adjacent portion of the circuit, although the profiles are qualitatively
correct as they often reflect the presence of local strain gradients. The issue is evident in Girder
5, Stop 1 (Figure 57b) and Stop 2 (Figure 58b) on Span 1, and Girder 5, Stop 4 (Figure 59) on

49
Span 1, whereas it seems not of concern in case of relatively high strain levels (beyond 100 µε)
for positive moment regions. This suggests that localized inconsistencies may be essentially due
to initial fiber misalignment and/or local disbond, with reduced effect as the fibers are
progressively stretched.
It is finally noticed that the data retrieved from the bare fibers did not have the same quality of
that from the GFRP tape, in particular in Girder 2, Location B (at hweb/3, Figure 37), where
practically zero strain was detected. This may be principally due to the presence of several
vertical circuit portions, sometimes partially unbonded, in the vicinity of the transverse
stiffeners, which likely caused undesired perturbations in the signals detected.

50
 4. CONCLUSIONS

Conclusions based on the deflection measurement via ATS during pouring of the deck slab and
diagnostic load test, strain measurement via BOTDR during the load test, and comparison with
analytical results based on the AASHTO LRFD Bridge Design Specifications (1998), can be
summarized as follows:
1. The structural assessment confirms the safety of the bridge, since the experimental
deflection and strain results were found to be smaller than the theoretical values
determined according to the design provisions (AASHTO, 1998);
2. A significant drop in the behavior of one of the exterior girders (Girder 1) was assessed,
with differences in the Span 1 midspan deflection between Girder 1 and Girder 5 of the
order of 30%, under similar load conditions. In the case of the load test, such difference
cannot be justified considering concurrent factors that typically affect the experimental
results. A preliminary explanation may account for a possible alteration of the structural
response as a consequence of the falling of the steel girder during erection of the
superstructure. The issue calls for further verification and monitoring during the life of
the bridge;
3. Experimental results from both techniques employed fairly describe the global response
of the bridge both in terms of deflections and strains, and were consistent with each other;
4. ATS confirms its cost-effectiveness for deflection measurement. The fact that the
technology allows for non-contact measurement significantly enhances its versatility;
5. BOTDR is a promising, novel method for global structural assessment, with the unique
ability to detect the distributed strain along a structural member. In particular, the
adhesively bonded custom-made GFRP tape with embedded optical fibers for strain
measurement provided a far better performance than bare fibers.

51
 5. REFERENCES

Bastianini, F., Cargnelutti, M., Di Tommaso, A. and Toffanin, M. (2003), “Distributed

Brillouin Fiber Optic Strain Monitoring Applications in Advanced Composite Materials”, Proc.
SPIE, 5057, pp. 478-485.
Brillouin, L. (1922), “Diffusion de la Lumière et des Rayonnes X par un Corps Transparent
Homogéne; Influence del’Agitation Thermique”, Annales des Physique, 17, pp. 88-122.;
Casas, J.R. and Cruz, P.J.S. (2003), “Fiber Optic Sensors for Bridge Monitoring”, ASCE
Journal of Bridge Engineering, 8[6], pp. 362-373.
American Association of State Highway and Transportation Officials (1996), “LFD Standard
Specifications”, 16th ed., AASHTO, Washington, D.C.
American Association of State Highway and Transportation Officials (1998), “LRFD Bridge
Design Specifications”, 2nd ed., AASHTO, Washington, D.C.
American Association of State Highway and Transportation Officials (2002), “Standard
Specifications for Highway Bridges”, 17th ed., AASHTO, Washington, D.C.
Azizinamini, A., Barth, K. E., Dexter, R. and Rubeiz, C. (2004), “High Performance Steel:
Research Front – Historical Account of Research Activities”, ASCE Journal of Bridge
Engineering, 9[3], pp. 212-217.
Barker, M.G. and Schrage, S.D. (2000), “High Performance Steel: Design and Cost
Comparisons”, Modern Steel Construction, American Institute of Steel Construction, 16, pp. 35-
41.
Barnoski, M.K. and Jensen, S.M. (1976), "Fiber Waveguides: A Novel Technique for
Investigating Attenuation Characteristics”, Applied Optics, 15[9], pp. 2112-2115.
Barth, K.E., Roeder, C.W., Christopher, R.A. and Wu, H. (2001), “Evaluation of Live Load
Deflection Criteria for I-shaped Steel Bridge Design Girders”, Proc. International Conference on
High Performance Materials in Bridges, Kona, HI, Jul. 29 – Aug. 3, 2001, A. Azizinamini, A.
Yakel and M. Abdelrahman, eds., American Society of Civil Engineers, pp. 193-208.
Casadei, P. and Nanni, A. (2003), “In-Situ Load Testing of Reinforced Concrete Structures:
Case Studies”, Proc. Concrete Technology Today, Singapore, Vol. 3, pp. 10-16.
Chajes, M.J., Mertz, D.R. and Commander, B. (1997), “Experimental Load Rating of a
Posted Bridge”, ASCE Journal of Bridge Engineering, 2[1], pp. 1-10.
Clingenpeel, B.F. and Barth, K.E. (2003), “Design Optimization Study of a Three-span
Continuous Bridge Using HPS70W”, AISC Engineering Journal, 39[3], pp. 121-126.
Deza, U. (2004) “Evaluation of the In-Service Performance of Bridge Decks Built with Fiber
Reinforced Polymer (FRP) Composite Systems”, Master Thesis, University of Missouri-Rolla,
Rolla, MO, U.S.A.
Federal Highway Administration Bridge Programs NBI Data (2004), “Material Type of
Structure by State”, http://www.fhwa.dot.gov/bridge/mat03.xls.
Galati, N., Casadei, P. and Nanni, A. (2004), “In-situ Load Testing of Bridge A6101

52
Lexington, MO”, CIES Report 04-002, University of Missouri-Rolla, Rolla, MO, U.S.A.
Hill, C.D. and Sippel, K.D. (2002), “Modern Deformation Monitoring: a Multi Sensor
Approach”, Proc. XXII FIG International Congress, International Federation of Surveyors, April
19-22, 2002, Washington, D.C., 12 pp., http://www.fig.net/pub/fig_2002/Ts6-3/TS6_3_hill
_sippel.pdf.
Hill, K.O., Johnson, D.C. and Kawasaki, B.S. (1976), “CW Generation of Multiple Stokes
and Anti-Stokes Brillouin-Shifted Frequencies”, Applied Physics Letters, 29, pp. 185-187.
Horiguchi, T., Kurashima, T. and Tateda, M. (1989), “Tensile Strain Dependence of
Brillouin Frequency Shift in Silica Optical Fibers”, IEEE Photonics Technology Letters, 1[5], pp.
107-108.
Horiguchi T., Kurashima, T. and Koyamada, Y. (1994), “1-m Spatial Resolution
Measurement of Distributed Brillouin Frequency Shift in Single-Mode Fibers”, Proc. NIST/IEEE
Symposium on Optical Fiber Measurements, Boulder, CO, Sep. 1994, pp. 73-76.
Horton, R., Power, E., Van Ooyen, K. and Azizinamini, A. (2000), “High Performance Steel
Cost Comparison Study.’’ Proc. Steel Bridge Design and Construction for the New Millennium
with Emphasis on High Performance Steel, Baltimore, MD, pp. 120-137.
Komatsu, K., Fujihashi, K. and Okutsu, M. (2002), “Application of the Optical Sensing
Technology to the Civil Engineering Field with Optical Fiber Strain Measurement Device (B-
OTDR)”, presented at the SPIE Smart Sensor Technology and Measurement Systems Conference
(SPIE 4694), San Diego, CA, Mar. 17-21, 2002.
Kuhlmann, H. and Gläser, A. (2002), “Investigation of New Measurement Techniques for
Bridge Monitoring”, Proc. 2nd Symposium on Geodesy for Geotechnical and Structural
Engineering, Berlin, Germany, pp. 123-132, http://www.uni-stuttgart.de/iagb/publikation/
newmeasure_bridge.pdf.
Leica Geosystems, Automated High Performance Total Station TCA2003, http://www.leica-
geosystems.com/surveying/product/totalstations/tc2003.htm.
Lynch, R.J. (2004), “Provisional Design Guide in AASHTO Language for FRP Bridge
Strengthening”, Master Thesis, University of Missouri-Rolla, Rolla, MO, U.S.A.
McDermott, J.F. (1969), “Local Plastic Buckling of A514 Steel Members”, ASCE Journal of
the Structural Division, 5[9], pp. 1837-1850.
Merkle, W.J. (2004), “Load Distribution and Response of Bridges Retrofitted with Various
FRP Systems", Master Thesis, University of Missouri-Rolla, Rolla, MO, U.S.A.
Meng, X. (2002), “Real-Time Bridge Deflection and Vibration Monitoring Using an
Integrated GPS/Accelerometer/Pseudolite System”, Proc. 11th FIG Symposium on Deformation
Measurements, Santorini, Greece, May 25-28, 2003, www.fig.net/commission6/santorini/I-
Monitoring%20Static%20and%20 Dyn/I5.pdf.
Pelous, J. and Vacher, R. (1975), “Thermal Brillouin Scattering Measurements of the
Attenuation of Longitudinal Hypersounds in Fused Quartz from 77 to 300 K”, Solid State
Communications, 16[3], p. 279.
Roeder, C.W., Barth, K.E. and Bergman, A. (2004), ‘‘Effect of Live-Load Deflections on

53
Steel Bridge Performance”, ASCE Journal of Bridge Engineering, 9[3], pp. 259-267.
Roeder, C.W., Barth, K.E., Bergman, A. and Christopher, R.A. (2001), “Improved Live-Load
Deflection Criteria for Steel Bridges”, NCHRP Interim Report No. 20-07/133, National
Cooperative Highway Research Program, Washington, D.C.
Radovanovic, R.S. and Teskey, W.F. (2001), ‘‘Dynamic Monitoring of Deforming
Structures: GPS versus Robotic Tacheometry Systems”, Proc. 10th FIG Symposium on
Deformation Measurements, Orange, CA, May 19-22, 2001, http://rincon.gps.caltech.edu/
FIG10sym/pdf/Session%20II_Paper%204.pdf.
Strongwell Corporation (2004), Fiberglass Flooring and Decking System brochure,
http://www.strongwell.com/PDFfiles/SafdeckSafplank/FlooringandDecking0304.pdf.
Thomas, P.J., Rowell, N.L., van Driel, H.M. and Stegeman, G.I. (1979), “Normal Acoustic
Modes and Brillouin Scattering in Single-Mode Optical Fibers”, Physical Review B, 19, pp.
4986-4998.
Watson Bowman Acme Corporation (2004), Wabo® MBrace Saturant Data Sheet,
http://www.wbacorp.com/downloads/DataSheets/Composites/MBraceSaturant.pdf.
Yariv, A. (1989), “Quantum Electronics”, 3rd ed., John Wiley & Sons.

54
 APPENDIX I – LOAD TEST STOPS

L3 = 200 ft L2 = 185 ft L1 = 147 ft

L3 / 2 L3 / 2 L2 / 2 L2 / 2 L1 / 2 L1 / 2

Girder 1
2 1
Girder 2
Girder 3
Girder 4
4 3 Girder 5
C
L CL C
L CL
Pier 4 Pier 3 Pier 2 Abutment 1

Stop 1

L3 = 200 ft L2 = 185 ft L1 = 147 ft

L3 / 2 L3 / 2 L2 / 2 L2 / 2 L1 / 2 L1 / 2

Girder 1
5 2 1
Girder 2
Girder 3
Girder 4
6 4 3 Girder 5
C
L C
L C
L C
L
Pier 4 Pier 3 Pier 2 Abutment 1

Stop 2

L3 = 200 ft L2 = 185 ft L1 = 147 ft

L3 / 2 L3 / 2 L2 / 2 L2 / 2 L1 / 2 L1 / 2

Girder 1
Girder 2
Girder 3
Girder 4
3 4 1 2 6 5 Girder 5
C
L C
L C
L C
L
Pier 4 Pier 3 Pier 2 Abutment 1

Stop 3

L3 = 200 ft L2 = 185 ft L1 = 147 ft

L3 / 2 L3 / 2 L2 / 2 L2 / 2 L1 / 2 L1 / 2

Girder 1
Girder 2
Girder 3
Girder 4
3 4 1 2 6 5 Girder 5
C
L C
L CL C
L
Pier 4 Pier 3 Pier 2 Abutment 1

Stop 4

55
 L3 = 200 ft L2 = 185 ft L1 = 147 ft

L3 / 2 L3 / 2 L2 / 2 L2 / 2 L1 / 2 L1 / 2

Girder 1
Girder 2
Girder 3
Girder 4
3 4 1 2 6 5 Girder 5
CL C
L C
L C
L
Pier 4 Pier 3 Pier 2 Abutment 1

Stop 5

L3 = 200 ft L2 = 185 ft L1 = 147 ft

L3 / 2 L3 / 2 L2 / 2 L2 / 2 L1 / 2 L1 / 2

Girder 1
Girder 2
Girder 3
Girder 4
3 4 1 2 6 5 Girder 5
CL C
L CL CL
Pier 4 Pier 3 Pier 2 Abutment 1

Stop 6

L3 = 200 ft L2 = 185 ft L1 = 147 ft

L3 / 2 L3 / 2 L2 / 2 L2 / 2 L1 / 2 L1 / 2

Girder 1
5 2 4
Girder 2
Girder 3
Girder 4
6 1 3 Girder 5
C
L C
L CL C
L
Pier 4 Pier 3 Pier 2 Abutment 1

Stop 7

L3 = 200 ft L2 = 185 ft L1 = 147 ft

L3 / 2 L3 / 2 L2 / 2 L2 / 2 L1 / 2 L1 / 2

Girder 1
6 1 3 5 2 4
Girder 2
Girder 3
Girder 4
Girder 5
C
L C
L C
L C
L
Pier 4 Pier 3 Pier 2 Abutment 1

Stop 8

56
 APPENDIX II – RAW DATA

II.a Deflection data (ATS)

Concrete-deck pouring
x 1 1/4 spans 2 spans 2 3/4 spans 4 spans 5 spans
[ft] [in] [in] [in] [in] [in]
Girder 1
3.0 1.74E-02 2.53E-02 8.27E-03 6.57E-02 -9.76E-02
50.0 1.02E-01 1.26E-01 -9.96E-03 5.16E-01 -1.13E+00
73.5 1.16E-01 1.32E-01 -1.02E-01 4.98E-01 -1.19E+00
111.0 5.50E-02 6.75E-02 -1.46E-01 3.23E-01 -5.93E-01
143.0 1.97E-04 -2.09E-02 -3.48E-02 6.76E-02 -3.87E-02
Girder 2
73.5 7.14E-02 1.39E-01 -1.04E-01 7.11E-01 -1.05E+00
Girder 3
3.0 1.22E-02 2.75E-02 4.17E-02 1.30E-01 -2.22E-02
50.0 2.22E-02 1.42E-01 -2.43E-02 9.12E-01 -9.54E-01
73.5 2.22E-02 1.35E-01 -1.51E-01 9.00E-01 -1.03E+00
111.0 9.80E-03 6.55E-02 -1.72E-01 6.15E-01 -4.95E-01
143.0 7.40E-03 1.51E-02 -2.76E-02 8.24E-02 -2.30E-02
188.0 -2.50E-02 -6.09E-02 3.75E-01 -1.21E+00 -6.01E-01
234.5 -4.62E-02 -1.34E-01 9.67E-01 -1.77E+00 -1.14E+00
Girder 4
73.5 -3.90E-02 1.21E-01 -7.80E-02 1.12E+00 -8.82E-01
Girder 5
3.0 1.26E-02 1.59E-02 N.A. N.A. N.A.
50.0 -1.19E-01 8.03E-02 -1.26E-01 1.11E+00 -8.62E-01
73.5 -9.86E-02 8.47E-02 -1.70E-01 1.19E+00 -8.98E-01
111.0 -4.94E-02 5.15E-02 -1.93E-01 8.04E-01 -4.27E-01
143.0 1.97E-04 2.11E-02 -5.35E-03 1.00E-01 9.13E-03
151.0 1.62E-02 5.51E-03 4.89E-02 -1.21E-01 -1.15E-02
188.0 5.86E-02 -6.25E-02 4.01E-01 -1.34E+00 -6.66E-01
234.5 5.79E-03 -1.67E-01 9.08E-01 -2.03E+00 -1.35E+00

57
 Diagnostic load test
x Stop 1 Stop 2 Stop 3 Stop 4 Stop 5 Stop 6 Stop 7 Stop 8
[ft] [in] [in] [in] [in] [in] [in] [in] [in]
Girder 1
3.0 -5.56E-02 -9.90E-02 -1.78E-02 -1.26E-02 7.64E-03 7.95E-03 -6.41E-02 -1.55E-01
50.0 -4.99E-01 -8.19E-01 -1.50E-01 -2.11E-01 -1.24E-01 4.67E-02 -8.29E-01 -1.37E+00
73.5 -4.79E-01 -8.25E-01 -1.37E-01 -1.95E-01 -6.80E-02 7.81E-02 -8.48E-01 -1.50E+00
111.0 -2.20E-01 -4.75E-01 -7.78E-02 -1.23E-01 -6.54E-02 5.33E-02 -4.57E-01 -9.85E-01
143.0 1.34E-03 -2.92E-02 2.30E-02 1.08E-02 3.80E-02 2.83E-02 -1.57E-02 -1.26E-01
Girder 2
73.5 -4.19E-01 -7.17E-01 -1.00E-01 5.76E-02 -4.24E-02 -1.76E-01 -7.47E-01 -9.57E-01
Girder 3
3.0 -1.64E-02 -5.19E-02 9.72E-03 4.98E-02 3.40E-02 -1.46E-02 -4.34E-02 -3.03E-02
50.0 -3.67E-01 -6.07E-01 -5.47E-02 2.20E-01 -3.07E-02 -3.88E-01 -6.03E-01 -4.57E-01
73.5 -3.84E-01 -6.41E-01 -6.20E-02 2.79E-01 -4.46E-02 -4.35E-01 -6.55E-01 -5.18E-01
111.0 -2.18E-01 -3.76E-01 -4.75E-02 2.34E-01 7.09E-03 -2.78E-01 -3.80E-01 -3.27E-01
143.0 3.94E-05 -3.43E-02 6.30E-04 6.14E-02 2.76E-02 -3.00E-02 -2.16E-02 -2.62E-02
188.0 1.59E-01 2.27E-01 1.38E-01 -3.82E-01 -1.22E-01 1.61E-01 2.23E-01 1.82E-01
234.5 1.75E-01 2.50E-01 2.81E-01 -6.56E-01 -2.21E-01 1.76E-01 2.44E-01 1.86E-01
Girder 4
73.5 -5.26E-01 -7.62E-01 -1.65E-01 3.38E-01 -2.26E-01 -8.98E-01 -7.62E-01 -2.36E-01
Girder 5
3.0 -4.68E-02 -7.66E-02 -8.66E-03 7.24E-02 -1.46E-03 -9.99E-02 -4.90E-02 2.30E-02
50.0 -5.97E-01 -8.10E-01 -1.48E-01 5.19E-01 -2.53E-01 -1.10E+00 -7.35E-01 1.26E-01
73.5 -5.23E-01 -7.72E-01 -9.66E-02 7.16E-01 -2.23E-01 -1.17E+00 -7.05E-01 1.88E-01
111.0 -3.06E-01 -4.65E-01 -1.23E-01 5.90E-01 -1.34E-01 -8.24E-01 -4.20E-01 1.21E-01
143.0 2.99E-03 -2.75E-02 -6.14E-03 9.46E-02 5.87E-03 -7.16E-02 -2.47E-02 2.07E-02
151.0 4.40E-02 3.18E-02 3.53E-02 -7.43E-02 -3.19E-02 4.79E-02 4.31E-02 6.61E-03
188.0 1.38E-01 1.90E-01 2.27E-01 -1.11E+00 -5.19E-01 2.85E-01 1.67E-01 -6.39E-02
234.5 1.37E-01 1.90E-01 4.91E-01 -1.85E+00 -7.61E-01 2.61E-01 1.38E-01 -1.08E-01

58
II.b Strain data (BOTDR)

Girder 1 (Location A, Table 3.2)

-300

Bolted joint

Bolted joint
-200

Pier 3
-100

0
Strain [µε]

Stop 1

100 Stop 2

200 Stop 7

300 Stop 8
Abutment 1

Bolted joint

400
Pier 2

500
0 30 60 90 120 150 180 210 240 270 300 330
x [ft]

Girder 2 (Location A, Table 3.2)

-300
Bolted joint

Bolted joint

-200
Pier 3

-100

0
Strain [µε]

100
Stop 1

200 Stop 2

300 Stop 7
Abutment 1

Bolted joint

Stop 8
400
Pier 2

500
0 30 60 90 120 150 180 210 240 270 300 330
x [ft]

59
 Girder 2 (Location B, Table 3.2)
-300

Bolted joint

Bolted joint
-200

Pier 3
-100

0
Strain [µε]

100
Stop 1

200 Stop 2

300 Stop 7
Abutment 1

Bolted joint

Stop 8
400 Pier 2

500
0 30 60 90 120 150 180 210 240 270 300 330
x [ft]

Girder 4 (Location A, Table 3.2)

-300
Bolted joint

Bolted joint

-200
Pier 3

-100

0
Strain [µε]

100 Stop 1

200 Stop 2

300 Stop 7
Abutment 1

Bolted joint

Stop 8
400
Pier 2

500
0 30 60 90 120 150 180 210 240 270 300 330
x [ft]

60
 Girder 5 (Location A, Table 3.2)
-300

Stop 1 Stop 2 Stop 4

-200
Stop 7 Stop 8
-100

0
Strain [µε]

100

200

300

Bolted joint
Abutment 1

Bolted joint
Bolted joint

400

Pier 3
Pier 2

500
0 30 60 90 120 150 180 210 240 270 300 330
x [ft]

Girder 5 (Location C, Table 3.2)

-300
Bolted joint

Bolted joint

-200
Pier 3

-100

0
Strain [µε]

100 Stop 1

Stop 2
200
Stop 4
300
Stop 7
Abutment 1

Bolted joint

400 Stop 8
Pier 2

500
0 30 60 90 120 150 180 210 240 270 300 330
x [ft]

61