Citation
Analysis and testing of the historic Blue River Bridge subjected to wind

Material Information

Title:
Analysis and testing of the historic Blue River Bridge subjected to wind
Creator:
Hamedian, Shohreh
Publication Date:
Language:
English
Physical Description:
xiii, 121 leaves : illustrations ; 28 cm

Subjects

Subjects / Keywords:
Wind-pressure ( lcsh )
Bridges -- Aerodynamics -- Colorado ( lcsh )
Bridges -- Aerodynamics ( fast )
Wind-pressure ( fast )
Blue River Bridge (Colo.) ( lcsh )
Colorado ( fast )
Colorado -- Blue River Bridge ( fast )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Bibliography:
Includes bibliographical references (leaves 120-121).
General Note:
Department of Civil Engineering
Statement of Responsibility:
by Shohreh Hamedian.

Record Information

Source Institution:
|University of Colorado Denver
Holding Location:
Auraria Library
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
71632690 ( OCLC )
ocm71632690
Classification:
LD1193.E53 2006m H45 ( lcc )

Full Text
ANALYSIS AND TESTING OF THE HISTORIC BLUE RIVER BRIDGE
SUBJECTED TO WIND
by
Shohreh Hamedian
B.S., Shahrood University, 1999
A Thesis submitted to the
University of Colorado at Denver
in partial fulfillment
of the requirements for the degree of
Master of Science
Civil Engineering
2006


This thesis for the Master of Science
degree by
Shohreh Hamedian
has been approved
by
Stephan Durham
April 9S~,aoo£
Date


Hamedian, Shohreh (M.S., Civil Engineering)
Analysis and Testing of the Historic Bine River Bridge Subjected to Wind
Thesis directed by Associate Professor Kevin L. Rens, PhD, PE
ABSTRACT
The overall purpose of this historical bridge research at the University of
Colorado Denver, is to test and analyze the Blue River Bridge near
Silverthome / Dillon, Colorado. The bridges response to wind load was
investigated in order to determine the feasibility of converting the bridge from
vehicular use to pedestrian usage. The short spanned bridge, located in the
Rocky Mountains at elevation 9,000 feet (2,743 meters), has a timber deck with
relatively high stiffness in the lateral direction. The traditional method of
structural engineering truss analysis is based on a skeleton frame. Alternative
load paths are normally neglected. This thesis explores the stiffening effect of
the often omitted deck system, which reduces the forces in critical structural
members of the bridge when compared with the traditional skeleton model.
Analytical modeling was completed using RISA-3D software. The
American Association of State Highway and Transportation Officials
(AASHTO) standard wind load of 75 psf (3.59 kpa) was applied to the models
to demonstrate the stiffening effect of the deck. The deck analytical model was
verified by a field test under real wind conditions. In summary, the Blue River


Bridge was analyzed under AASHTO wind load for two different systems,
skeleton frame and skeleton with stringers and deck. It was again analyzed
under wind pressure measured experimentally on-site for the skeleton with
stringers and deck. The results are compared for critical structural members.
Despite existing distress in the truss and abutments, it was found that the lateral
stiffness of the deck was near its theoretical maximum. Therefore, this stiff
deck can reduce the forces in critical members under wind load. On the other
hand, in case after case observations reveal no physical evidence to suggest that
wind has caused damage or distress, even after a century of exposure. At this
age bridge have indeed weathered many severe windstorms. Therefore,
rehabilitation for pedestrian use is a practical way to preserve these historic
structures.
This abstract accurately represents the content of the candidates thesis. I
recommend its publication.
Signed
Kevin L. Rens


DEDICATION
I dedicate this thesis to my parents, Hooshmand Hamedian and Gity Parsa, and
to my husband, Tom Khosravi.
i


ACKNOWLEDGEMENT
A great many people offered me guidance and support during the pursuit
of my Masters Degree. I wish to acknowledge the efforts of my advisor, Kevin
L. Rens, who supported this research over the past year. Also, my great thanks
go to Frederick R. Rutz, overall principal investigator of the University of
Colorado Denver historic bridge research, who helped me to complete my
thesis. In fact, this research began with his doctoral dissertation bridge research
completed in 2004. Next, my thanks go to my graduate committee: Chengyu Li
and Stephan Durham. The instrumentation used for the field-testing was funded
in part by the National Center for Preservation Technology and Training and the
State Historical Fund of the Colorado Historical society. Further, the
cooperation of the bridge owner, Summit County, Colorado, is acknowledged.
Finally, I wish to acknowledge the following University of Colorado at Denver
students and others, who assisted in laboratory and field investigations
throughout the project: Mohammad Abu-Hassan, Paul Bountry, Sam Brown,
Jennifer Davis, Kazwan Elias, Aaron Erfman, Veronica Jacobson, Chris Kline,
Peter Marxhausen, and Helen Frey.


CONTENTS
Figures.............................................................x
Tables.............................................................. xiii
Chapter
1. Overview........................................................1
1.1 Introduction................................................... 1
1.2 Goals...........................................................2
1.3 Historical and Modem Analysis.................................. 8
1.4 Loads...........................................................9
2. Blue River Bridge..............................................10
2.1 Bridge History.................................................10
2.2 Blue River As-Built Dimensions.................................12
2.3 Bridge Geometry................................................15
2.4 Bridge Members............................................... 16
2.4.1 Top Chords.....................................................16
2.4.2 Portals........................................................17
2.4.3 Bottom Chords..................................................17
2.4.4 Vertical Posts.................................................18
2.4.5 Bracings.......................................................19
vn


2.4.6 Deck.............................................................20
2.4.7 Floor Beams and Stringers....................................... 21
2.4.8 Railing......................................................... 22
3. Field Experimentation............................................24
3.1 Test Set Up......................................................24
3.2 Instrumentation System...........................................25
3.3 Instrumentation Components.......................................26
3.3.1 Strain Transducers...............................................26
3.3.2 Wheatstone Bridge...............................................28
3.3.3 Anemometers......................................................29
3.3.4 Wind Direction Sensor............................................30
3.3.5 Interval Timer...................................................31
3.3.6 Data Logger......................................................32
3.3.7 Software.........................................................34
3.3.8 Cables...........................................................34
3.3.9 Temperature Measurement..........................................35
3.4 Wind Speed Data..................................................35
3.5 Strain Data......................................................36
3.5.1 Cable Resistance.................................................36
3.5.2 Strain Computation...............................................37
3.5.3 Rolling Average..................................................38
vm


3.6 Determination of Actual Wind Pressure...........................40
4. Modeling and Analysis...........................................46
4.1 Introduction....................................................46
4.2 Modeling Consideration..........................................47
4.3 Analysis and Results............................................48
4.4 Eyebar Condition................................................58
4.5 Conclusions.....................................................60
5. Verification Method.............................................61
5.1 Introduction....................................................61
5.2 Verification Results............................................61
5.3 Conclusions.....................................................71
6. Conclusions and Recommendations for Future Research.............73
6.1 Introduction....................................................73
6.2 Conclusions.....................................................74
6.3 Recommendations for Future Research.............................74
Appendix
A. Blue River Bridge Photographs...................................76
B. Program for Data Logger.........................................90
C. Gravity Load Study of Blue River Bridge........................100
D. Basics of the spreadsheet......................................114
References
120


FIGURES
Figure
1.1 Fruita Bridge over the Colorado River.......................... 3
1.2 Prowers Bridge over the Arkansas River......................... 4
1.3 Blue River Bridge over the Blue River.......................... 5
1.4 San Miguel Bridge over San Miguel River........................ 6
1.5 Rifle Bridge over the Colorado River........................... 7
2.1 Colorado Map................................................... 11
2.2 Location of Blue River Bridge.................................. 11
2.3 Blue River Bridge As-Built Dimensions. Plan and Elevation....... 13
2.4 Blue River Bridge As-Built Dimensions. Member Details....... 14
2.5 Blue River Bridge over Blue River.............................. 15
2.6 Top Chords in the Blue River Bridge............................ 16
2.7 Portals in the Blue River Bridge............................... 17
2.8 Bottom chord eyebars in the Blue River Bridge.................. 18
2.9 Vertical Posts in the Blue River Bridge........................ 19
2.10 Bracings in different locations of the Blue River Bridge....... 20
2.11 Timber Deck on the Blue River Bridge........................... 21
2.12 Floor Beams and Stringers in the Blue River Bridge............. 22
2.13 Railing in the Blue River Bridge................................23
3.1 Blue River Test Set-up..........................................25
x


3.2 Instrumentation system.............................................26
3.3 Strain Transducer..................................................27
3.4 Wheatstone Bridge..................................................28
3.5 Anemometer installed at Blue River Bridge..........................29
3.6 Wind Direction Sensor installed at Blue River Bridge...............31
3.7 Interval Timer.....................................................32
3.8 Campbell Scientific CR5000 Data Logger.............................33
3.9 Blue River Bridge strain measurements..............................39
3.10 Diagram of Blue River Bridge.......................................42
3.11 Blue River Bridge quadrants........................................43
3.12 Wind pressure applied to the four quadrants for analysis...........44
4.1 Illustration of the traditional skeleton...........................48
4.2 Representation of superimposed gravity loads.......................49
4.3 Relative axial forces in the bottom chord eyebars..................50
4.4 Representation of wind pressure on the bridge......................51
4.5 Graphical representation of axial forces...........................52
4.6 Shear and moment diagrams for propped cantilever condition.......52
4.7 Skeleton model with stringers and deck elements....................53
4.8 Rendering of steel stringers on steel floor beam...................54
4.9 Offset members and release locations................................55
4.10 Rendering of the timber deck on steel stringers on steel floor beam.55


4.11 Buckled eyebars in the Blue River Bridge......................59
4.12 Abutment failure in the Blue River Bridge.....................59
5.1 Diagram of Blue River Bridge................................. 62
5.2 Locations of anemometers and wind direction sensore...........63
5.3 Blue River Bridge quadrants................................. 66
5.4 Wind pressure applied to the four quadrants for analysis..... 67
5.5 Wind Speed as measured by the five anemometers............... 68
5.6 Wind direction as measured during the test................... 69
5.7 Strain measurements for bottom chords.........................69
5.8 Enlargement of the trace for windward and leeward bottom chord.70
5.9 Measured strains at the south portal........................ 70
xii


TABLES
Table
4.1 Maximum Axial Forces in Bottom Chord Eyebars....................57
5.1 Wind Velocities, Quadrant Average Velocities and Pressures.....64
5.2 Blue River Bridge Verification Summary.........................71
xiii


1.
Overview
1.1. Introduction
Historic bridges are defined as bridges older than 100 years. During the
late 19 and early 20 centuries, there were numerous companies that
constructed iron and steel truss bridges with spans over 150 feet (45.72-meters).
Some bridges were moved from their original locations when loading or
dimensional restrictions made them obsolete. Many have been removed from
service due to structural or functional deficiencies-i.e., to carry todays heavier
and wider trucking loads.
Engineers who study historic truss bridges find todays vertical design
live loads to be on the same order of magnitude as those used by the original
designers a century ago. The American Association of State Highway and
Transportation Officials (AASHTO 1997) Guide Specifications prescribe the
live load value. Live loads vary from 65 to 85 psf (3.11 to 4.07 kPa), depending
on the area of the walkway to which the load is applied (Rutz, 2004). However,
horizontal or lateral design wind loads are now significantly higher. Early
design wind load values varied, typically from 30 to 50 psf (1.44 to 2.39 kPa).
The AASHTO Guide Specifications mandates 75 psf (3.59 kPa) be applied to
1


the same area today (Rutz, 2004). Struetural engineers discovered that many
historic bridge structures do not have the lateral capacity to resist todays wind
loads specified in the AASHTO code. Therefore, conversion from vehicular to
pedestrian use, as opposed to demolition, is a practical way to preserve these
historic structures.
1.2 Goals
The focus of the historical bridge research at the University of Colorado
at Denver (UCD) has been on the identification of alternate load paths, with
particular attention to the stiffening effects of the deck and stringer system. The
overall study focused on six real not text book structures located
throughout Colorado. The study on two bridges was first completed by Rutz
(2004). The remaining four were funded by the National Center for
Preservation Technology and Training (NCPTT). All of the bridges are
through- trusses, all are former highway bridges, all are pin-connected with
moment-resisting portal frames at the ends, all are metal-either wrought iron or
steel, and all have horizontal trusses consisting of rod X- bracing intended to
resist lateral loads. Figures 1.1 through 1.5 show the illustrations of the four
bridges used in the recent study The captions for each figure contain a brief
summary of information relative to the bridge.
2


Figure 1.1. Fruita Bridge over the Colorado River, near Fruita Colorado. This
three-span steel Parker truss was built in 1907. Each 47-meter (155-feet) span
has eight bays. It has steel floor beams and timber stringers covered by a timber
deck. Steel eyebars serve as bottom chords and principal diagonals and steel
rods provide counterbracing. It served until a replacement bridge was built
about one half mile downstream in 1970, and has been abandoned since then.
The former wood railing has fallen away. The City of Fruita would like to
reopen the bridge for pedestrian and bicycle use as part of a connecting bikeway
leading to nearby tourist attractions, but has been stymied by the expense of
rehabilitation. The north span, which was instrumented, is at the far right.
3


Figure 1.2. Prowers Bridge over the Arkansas River, near Lamar,
Colorado. This bridge consists of one 5 panel Pratt pony truss built in
1921, three 9 panel Camelback Pratt through trusses built in 1909, two 6
panel Pratt through truss built in 1902 and 1906. The 49-meter (160-
feet) Camelback Pratt through truss span, seen in this photo, was
instrumented because it received the greatest wind exposure. It has steel
floor beams with steel stringers, covered by a corrugated metal deck with
asphalt pavement. It has steel eyebar bottom chords and diagonals with
steel rod counterbracing. The railing is a steel lattice with single angle
top and bottom rails. Virtually all paint has weathered away. It survived
a major flood of the Arkansas River in 1921 and served as a highway
bridge until its abandonment in 1994 when a nearby replacement bridge
was constructed.
4


Figure 1.3. Blue River Bridge over the Blue River near
Silverthome/Dillon, Colorado. This 24.38-meter (80 feet)-span steel
Pratt truss has five bays, with a timber deck on steel stringers. It has
steel eyebar bottom chords and diagonals and steel rod X-bracing at the
center bay. The railing is a steel lattice with double angle top and
bottom rails. It is believed to have been built approximately 1895 as the
Two-Mile Bridge near Breckenridge, Colorado and moved to this site at
a later, but unknown, date. It is known to have been in its present
location when the Dillon Dam was built immediately upstream in 1960.
Closed to vehicular use, it is still used as a pedestrian crossing of the
Blue River.
5


Figure 1.4. San Miguel Bridge over San Miguel River near Uravan,
Colorado. This 43-meter (142-feet) wrought iron Pratt truss was built in
1886 as part of a five-span Fifth Street Bridge over the then Grand (now
Colorado) River at Grand Junction, Colorado. One span was relocated to the
San Miguel river location in the 1930s. It has a roadway of gravel on an
unusual system of semi-circular lengths of corrugated metal pipe set between
steel stringers. It has wrought iron eyebar bottom chords and diagonals and
wrought iron rod counterbracing. A steel vehicular rail has replaced the
original railing. It served the mining industry in western Colorado until the
1980s. Abandoned since 1990, it remains the oldest bridge originally built
in Colorado.
6


Figure 1.5. Rifle Bridge over the Colorado River at Rifle, Colorado. This 73
meter (240-foot) span Pennsylvania truss comprises the longer of two different
spans at that location. It has steel floor beams with steel stringers, covered by
a corrugated metal deck with asphalt pavement. It has steel eyebar bottom
chords and diagonals and steel rod counterbracing. The railing is a steel lattice
with double angle top and bottom rails. It has been abandoned since the late
1960s, when a replacement bridge was constructed.
As mentioned previously, this research began around the year 2000 with
the doctoral dissertation research completed by Rutz (2004). The goal was to
investigate the feasibility of providing the option to convert or preserve historic
truss bridges into pedestrian bridges using modem building codes and analysis
methods. At a minimum, it is hoped that the UCD research will prove to bridge
designers and historical preservationalists that options other than demolition
exist. This thesis is based on the subject bridge called Blue River near
7


Silverthome/Dillon, Colorado shown in Figure 1.3. The history, geometry, and
structural members of this bridge are described in Chapter 2. A photo gallery of
the bridge members and field investigation is also included in Appendix A. The
research includes laboratory testing, in-situ field-testing, and traditional and
non-traditional structural analysis. Several measurement instruments have been
utilized in the field experimentation, as described in Chapter 3. The scope of
this research is to measure forces in different bridge members under actual wind
loading, followed by a comparison of the measured forces with forces calculated
using AASHTO wind load criteria (75 psf/3.59 kPa). The scope also includes
the investigation of the stiffening effect of the deck and stringer system in
comparison with the skeleton model. The original skeleton analysis of the Blue
River Bridge was completed earlier by Carrol (2003), who used RISA 3D
software to analyze the bridge truss members under AASHTO wind load.
Carrolls model has been utilized in this thesis. All data from this RISA 3D
analysis, as well as the field measurements are included in Chapters 4 and 5.
1.3 Historic and Modern Analysis
Until the middle 19lh century, there were no accurate or accepted
methods to calculate the force for each member in a truss bridge. The classical
method was based on a skeleton frame analysis where alternative load paths
were neglected. Modem engineers use computer software to analyze and
8


calculate forces in each member of a truss bridge. Computer analysis is also
based on a skeleton frame in which alternative load paths are considered in the
design. Therefore, tradition and modem methods follow the same fundamental
basis with different techniques (Rutz, 2004).
1.4 Loads
The actual loading criteria used to design truss bridges in the late 19th
century and early 20th century varied according to the bridges proposed use.
Superimposed dead load, superimposed live load, and self-weight were
computed manually. Design wind loads varied from 1.44 Pa to 2.39 Pa (30 psf
to 50 psf) and were applied to the projected area of the exposed components
(Rutz, 2004).
Todays bridge engineers use the AASHTO Guide Specification for the
Design of Pedestrian Bridges (AASHTO, 1997) to determine the live load and
wind load values. According to AASHTO, live load may vary between 3.11 kPa
and 4.07 kPa (65 psf and 85 psf), depending on the area of the walkway. Design
wind load should be taken as 3.59 kPa (75 psf) applied to the projected area of
the components.
9


2.
Blue River Bridge
2.1 Bridge History
In 1895, Colorado House Bill No, 37 provided funds for the construction
of a vehicular bridge over the Blue River, north of the town of Breckenridge,
Colorado. The contract to construct the bridge was awarded to the Kansas City
Bridge Company of Kansas City, Missouri (Biennial Reports, 1916). The bridge
design consisted of a Pratt through truss configuration with five approximately
equal-length panels. The bridge spanned 80 feet (24.38 meters) with a nominal
clear road width of 14 feet (4.27 meters). The bridge was constructed using
steel and consisted entirely of pinned connections. All four of the bearings were
connected to the abutments. At a later date, this bridge was relocated until it
was moved to its current location near the tailrace of the Dillon Dam. This
structure no longer carries vehicular traffic, but carries regular pedestrian traffic.
Figures 2.1 and 2.2 show the location of the Blue River Bridge.
10



v**._
w. Minted *
**SpN'1 * "' B*dCM? *.
,jiJ*Mt / K*^
san* jSrv?wm>*$ / ; .
iwmaM 5'^;
v1*?,*. - >
Vhits Rfirer National Fores! Twin La* "
JrWV'l <*'*
ttod 9 ' \
uK* /Mount Crested {Mi*t
&Utt# flany
*, ^-nr>Jsor Isabel
- V;- ' /Nanonol Natioi
W; .^*: Fores! - .Forest
auS tjunmson ^ Wtei
Figure 2.1. Colorado Map where the black star shows the current location of
the Blue River Bridge.
Figure 2.2. Location of Blue River Bridge. It is located over the Blue River
near Silverthome/Dillon, Colorado (Delorme 1997b). The bridge was built
approximately 1895 as the Two-Mile bridge near Breckenridge, Colorado and
moved to this site at a later date.
11


2.2 Blue River As-Built Dimensions
The dimensional data and member sizes for this structure were obtained
through field measurements made by Carroll (2003). Figures 2.3 through 2.4
show dimensional information and member sizes.
12


80'-0" (24.38 m)
PLAN
symmetrical
U1 U2 j U3 U4
15'0
(4.57 m)
140"
(4.27 m)'
Figure 2.3. Blue River Bridge As-Built Dimensions. Plan, Elevation and
Section.
13


14


2.3 Bridge Geometry
The Blue River Bridge is a steel Pratt truss bridge that has an 80 foot
(24.38 meters) span with five 16 feet (4.88 meters) bays. The bridge road width
is 16 feet (4.88 meters) and the truss height is also 16 feet (4.88 meters). The
bridge truss consists of double channel lattice posts, top chords, and portals.
The bottom chord eyebars are double plate and the diagonal members are either
rods or double plates. The top and bottom chord x-bracings are also rods. The
bridge deck consists of longitudinal and transverse timbers on steel stringers that
bear on the floor beams. Figure 2.5 shows the bridge geometry.
15


2.4 Bridge Members
2.4.1 Top Chords
Top chords consist of double lattice channel Cl x 2 inches (17.78 x 5.08
centimeters). The channels are spaced 12 inches (30.48 centimeters) apart and
connected together by VSWVa (0.64 x 4.45 centimeters) lacing bars. Figure 2.6
shows the top chord for the Blue River Bridge.
Figure 2.6. Top Chords in the Blue River Bridge.
16


2.4.2 Portals
Portals are diagonal members located at the ends of the truss frame. The
portal geometry and cross-section was the same as the top chord described in
Section 2.4.1. Figure 2.7 depicts a portal from the Blue River Bridge.
Figure 2.7. Portals in the Blue River Bridge.
2.4.3 Bottom Chords
Bottom chord eyebars are located at the bottom of the truss. These
members consist of a double plate lx 2 (2.54 x 5.08 centimeters). For the
17


Blue River Bridge, the bottom chords on both sides had buckled significantly.
Figure 2.8, shows the bottom chord eyebars that are buckled in the bridge truss.
Figure 2.8. Bottom chord eyebars in the Blue River Bridge.
2.4.4 Vertical Posts
Posts are oriented vertically and located on top of floor beams at the end
of each bay. Posts consist of double lattice channels C5 x 1 % (12.7 x 4.45
centimeters). The channels are 8 (20.32 centimeters) spaced apart and
18


connected together by %xl % (0.64 x 4.45 centimeters) lacing bars. Figure 2.9
shows the vertical posts in the Blue River Bridge.
Figure 2.9. Vertical Posts in the Blue River Bridge.
2.4.5 Bracings
The bridge truss includes several braces at different locations. The
middle bays have a double 7/8 (2.22 centimeters) diameter steel rod x-bracing.
11" 9"
All other bays have double x 1 (1.75 x 3.97 centimeters) steel plates
19


between the top and bottom chords. The top chord bracings consist of a %
(1.91 centimeters) diameter steel rod x-bracing and the bottom chord bracing
consists of 1 (2.54 centimeters) diameter steel rod x-bracing. Figure 2.10
shows all types of bracing at the various locations.
Figure 2.10. Bracings in different locations of the Blue River Bridge.
2.4.6 Deck
The Blue River bridge deck consists of longitudinal running boards on
transverse timbers over steel stringers. The orthogonal crisscrossing of the
running boards and the deck timbers creates a relatively continuous stiff deck.
Both the longitudinal and transverse boards have a 3xl2 (7.62 x 30.48
centimeters) cross-section. Figure 2.11 shows the deck geometry as viewed
20


from the top.
w" -
^:.i-;^^r T.-i&r * >-T',ffv -a-'- -r ' . -.-:-i>;- -
rjp^_!
Figure 2.11. Timber Deck on the Blue River Bridge.
2.4.7 Floor Beams and Stringers
Two types of stringers support the deck boards in the Blue River Bridge.
Channel stringers (C7x 2) (17.78 x 5.08 centimeters) are located at both ends
r
and (S6x 4 ) (15.24 x 10.48 centimeters) stringers are located between the
8
ends. The stringers span 6.5 feet (1.98 meters) and bear on the floor beams.
There are a total of four floor beams (S15 x 5 Vz') (38.1 x 13.97 centimeters) at
21


the end of each bay. Figure 2.12 shows the stringers and floor beams in the
bridge. Also shown in figure 2.12 is the bottom layer of the deck running
orthogonal to the upper deck members shown in figure 2.11.
Figure 2.12. Floor Beams and Stringers in the Blue River Bridge.
2.4.8 Railing
Railing consists of steel lattice diagonal
centimeters) bars with double angle 1 x 1
4 4
1" 1"
-x 1- (0.32x3.18
8 4
x (4.45 x4.45x0.32
8
22


centimeters) top and bottom rails. Figure 2.13 shows the railing in the Blue
River Bridge.
23


3.
Field Experimentation
3.1 Test Set Up
The university research team utilized several instruments to obtain
measurements of wind speed, wind direction, and strains from selected members
during windy conditions.
A total of five anemometers were used to obtain wind speed data. Wind
direction with respect to the bridge orientation was determined by a wind
direction sensor. Wind direction was assumed to be consistent over the bridge
span. A total of sixteen strain transducers were installed on different members
to obtain the strain data. These members were selected because of relatively
high axial forces or moments that were expected due to lateral load. Strain rings
were used because of the easy installation in comparison with mounting an
actual gage on the bridge member. However, strain measurement obtained from
the actual gage applied to a structural member is more accurate. On the Blue
River Bridge, eight strain transducers were installed back-to-back on each
eyebar at mid-span. Four transducers were installed on end stringers at mid-
span, and four on top and bottom of the south portals. Figures 3.1 shows the test
24


set-up on the Blue River Bridge. Appendix A shows various photographs of
different stages of the test set up.
Figure 3.1. Blue River Test Set-up.
3.2 Instrumentation System
The instrumentation consists of sixteen strain transducers, five
anemometers, a data logger, a generator, modem, antenna and laptop computer.
Figure 3.2 shows a diagram of the instrumentation system.
25


STRAUS TRANSDUCERS
WIND OTECIKTU
ANEMOMETERS
LEEWARD
EEESAR
WINDWARD
ETEBAR
SOUTH
PORTAL
STRINGER
VVD WS1 WS2 WS3 WS4 WS5
SCHEMATIC PIACRAU
TIM TERMINAL INPUT MODULE (WHEATSTONE BRIDGE)
UPS UNINTERUPTABIE POWER SUPPLE
Figure 3.2. Instrumentation system
3.3 Instrumentation Components
All individual instruments that were used in the field work are described
below.
3.3.1 Strain Transducers
A transducer is a device that transforms one type of energy into another.
A strain transducer is an instrument that can be used to determine the force in
structural members by measuring deformation. For this research, each strain
transducer consisted of a 3 (7.62 cm) steel ring with a strain gage adhered to
26


the inside face as shown in Figure 3.3. Model CEA-06-250VW-12c strain
gages, manufactured by Vishay Micro Measurements Group, were used. The
gage factor for strain gages was either 2.065 or 2.095. The ring was attached to
two steel angles, which were used to clamp the transducer to the bridge member.
Axial strain of the members, measured by the transducer was obtained by
flexural deformation of the ring. The true strain is obtained by multiplying the
transducer strain by a factor determined by Herrero (Herrero, 2003). The strain
gage is considered a quarter bridge strain gage because it is one of four resistors
in the Wheatstone bridge circuit, which is described below.
Figure 3.3. Strain Transducer
27


3.3.2 Wheatstone Bridge
Each of the sixteen channels for the strain signals had an in-line
Wheatstone bridge. The circuitry for the bridges was provided by Campbell
Scientific model 4WFB120 Terminal Input Modules (TIMs), shown in Figure
3.4. The bridge consists of four wires with small modules that connect to the
channel terminals on the data logger. The resistor in the bridge circuit had a
resistance of 120 ohms plus or minus a tolerance of 0.01% (Campbell Scientific,
1996b). A quarter bridge strain gage is named because the strain gage is one
of four resistors that make up a lull bridge. The other three resistors are
provided by the TIM module.
Figure 3.4. Wheatstone Bridge. The variable resistor (Vx) represents the strain
gage and the other three resistors (H, L, and AG) are contained within the TIM
that connects to the data logger. Data logger provides excitation voltage via Vx
lead wire (Rutz, 2004).
28


3.3.3 Anemometers
An anemometer is an instrument that measures wind speed. The R.M.
Young model 0310-5 anemometers, shown in Figure 3.5, were used to obtain
wind data in this experiment. They have three cups connected to a wheel on a
vertical shaft. The shaft drives an AC generator, which produces a sine wave
output voltage signal directly proportional to the wind speed. Each cup wheel
revolution produces one complete sine wave cycle (Campbell Scientific, 1996a).
Figure 3.5. Anemometer installed at Blue River Bridge.
29


3.3.4 Wind Direction Sensor
The wind direction sensor used in the field was a R.M. Young Model
03301-5 wind Sentry Vane. The vane rotates with the wind and positions itself
parallel to the wind direction. This vane is connected to a shaft that turns a
potentiometer. The potentiometer changes electrical resistance when the vane
rotates.
An excitation voltage is applied, and from measurement of the voltage
drop the wind direction can be determined (Campbell Scientific, 1996a). The
wind direction sensor was located on the bridge such that the local bridge north
(0 degrees and 360 degrees) was aligned with the longitudinal direction of the
bridge. Thus, a wind from the local west direction (270 degrees) would be
transverse to the bridge. The terms north and west are the local directions.
Figure 3.6 shows the wind direction sensor.
30


Figure 3.6. Anemometer and Wind Direction Sensor installed at Blue River
Bridge.
3.3.5 Interval Tinier
The AC sine wave voltage signals from the anemometers were received
and processed by an interval timer. A Campbell Scientific model SDM-INT8
interval timer, shown in Figure 3.7 was used to download the processed data to
the data logger. The procedure used to process data is described below:
The rising edges of the sine wave voltage signals were timed.
The time intervals between rising sine waves were then downloaded to
the data logger.
31


The data logger processed this information into wind velocity in units of
either meters per second or miles per hour.
Figure 3.7. Interval Timer
3.3.6 Data Logger
A data logger was used to log the data from the strain transducers and
anemometers. A Campbell Scientific model CR5000 data logger was used for
this research. In the logger, 17 channels were used: 16 channels for strain
measurements from the transducers and one channel for wind direction
measurements. The five anemometers were connected to one of two pulse
32


counters on the data logger via the interval timer. The logger provided
excitation voltage for all of these devices and measured the voltage drop. The
collected data was stored in a separate table and then downloaded to a laptop
computer. Figure 3.8 shows the Campbell Scientific data logger.
Figure 3.8. Campbell Scientific CR5000 Data Logger
33


3.3.7 Software
PC 9000 software was used in conjunction with the data logger to create
the data tables (Campbell Scientific 2001). This software was accessed from
another program, VALNEW5, which was developed with the aid of Campbell
Scientific for this research. The collected tables contained the date and time of
logging, the five wind speeds, the wind direction, and the sixteen strain
transducer outputs. All of this data was logged at a sampling rate of 0.1
seconds. A program listing for VALNEW5 is included in Appendix B for the
Blue River Bridge.
3.3.8 Cables
Each strain gage had copper lead wires of 26 American Wire Gauge
(AWG) soldered to its terminals. The length of each wire varied between 5 ft
(1.52 m) and 15 ft (4.57 m). These lead wires were connected to the data logger
by 18 AWG copper wires with PVC insulation, all of which was wrapped inside
a foil shield and bundled into a PVC sheath. The lengths of these wires varied
between 78 ft (23.77 m) and 250 ft (76.2 m). The foil shield was grounded to
the data logger to provide shielding from spurious electrical signals. The
resistance of all wires was accounted for in the strain computations.
34


The cables for the anemometers and wind direction sensor were
Campbell Scientific model 9661 with 22 AWG twisted pair lead wires in
santoprene. All wires were 125 ft (38.1 m) long.
3.3.9 Temperature Measurement
In this test, temperature was measured by a Campbell Scientific
thermocouple and logged with all other data.
3.4 Wind Speed Data
In this research, the wind speed was logged at intervals of 0.1 second.
The data with the highest wind speed and the appropriate wind direction (east to
west or west to east with +30 changes) was chosen to use in the verification
model. Wind directions were logged by the wind direction sensor, which was
installed parallel to the bridge orientation on the local west side of the bridge.
Rather than using the maximum wind velocity, the difference between
high and low velocities was used in this study due to potential slack in the many
component member connections in the bridge. The minimum wind pressure can
displace the members and remove the slack from the connections. After the
slack is taken up, the strains would be proportional to wind velocity. Both low
and high wind velocities correspond to the average of 20 consecutive data
35


measurements over a two second interval. The two second average is used
because wind speeds fluctuate considerably over very short time intervals.
3.5 Strain Data
Data from the strain transducers required the following corrections:
Make a temperature correction for the gage factor.
Make a correction for the resistance of the lead wire that was
outside of the arms of the Wheatstone bridge.
Filter out signal noise by rolling averages.
Zero the data.
The following methods were used to make these corrections.
3.5.1 Cable Resistance
The lead cables that connect the strain gages to the data logger were
much longer than the strain gage wires. Thus, the resistances of the lead cables
were significant. In order to account for the resistance of these cables, the gage
factor is modified as follows:
where:
OP rip
revised ^ initial
R
(Rs+Rl)
(3.1)
GF'inmai = Gage factor of the stand alone strain gage
after the temperature correction
36


GFrevised = Adjusted gage factor that accounts for lead
wire resistance
R = Resistance of strain gage (120 ohms for this case)
Rl = Resistance of lead wire
The resistance of the 26 AWG copper strain gage lead is 0.041
ohn
meter
.) and the resistance of the 18 AWG copper cable
is 0.006385 o/lm'j400, (0.021 ohms/meter) (Rutz, 2004).
The lead wires were connected with one wire to one gage terminal and
two wires to the other terminal. By using the wiring method shown in Figure
3.4, the resistance of the two lead wires of the Wheatstone bridge and the
resistance changes due to temperature fluctuations were self-compensating
(Campbell Scientific, 1996b).
3.5.2 Strain Computation
The strain was computed as shown in equations 3.2 and 3.3:
GF{ \-2Vr)
(3.2)
with:
(3.3)
37


where:
c = strain
GF = Gage Factor, after temperature adjustment and
adjustment accounting for lead wire resistance
Vout = Measured bridge output voltage
Vex = Excitation voltage (Campbell Scientific, 1996b)
The ratio of output voltage to excitation voltage, Vr, was used because:
The data logger uses a technique that allows this ratio to be more
accurate than a simple measurement of output voltage.
This ratio can be conveniently used in the strain calculation.
The derivation of Equations 3.2 and 3.3 can be found in the Instruction Manual
for 4WFB 120 (Campbell Scientific, 1996c).
3.5.3 Rolling Average
Strain signals logged by the data logger varied significantly because of
electrical noise. Since the strain measurements were in the milli-volt and micro-
volt range, even a slight disturbance could cause a lot of noise in the strain
signals. In contrast, the anemometer measurements were in the 0.1 -0.3 volt
range. In comparison with the strain measurements, there were smaller changes
in the wind speed signals during the same period of time.
38


In order to reduce the effect of electrical noise in the strain signals, the
raw data was filtered by taking a rolling average of 20 consecutive data
measurements and advancing the average by the sampling interval of 0.1
seconds. A 2-second rolling average has been applied to the Blue River bridge
strain data for all strain gages, including the 20 consecutive data points at 0.1
second intervals adding up to 2 seconds. Figure 3.9 shows an example of a 2-
second rolling average of the Blue River bridge data.
60
BOTTOM CHORDS
Blue River Bridge 4/20/05
90
120
Time (seconds)
150
180
|WSavg Leeward Strain Windward StrainLeeward filteredWindward filtered j
Figure 3.9. Blue River Bridge. Strain measurements for the windward and
leeward bottom chord eyebars. The average wind speed, to an arbitrary scale, is
also shown as the bold line at the top. Both the raw data and a filtered line that
removes much of the signal noise by taking a 2- second rolling average are
shown.
39


3.6 Determination of Actual Wind Pressure
After choosing the data with highest wind pressure from all data logged
in a few different days, the following procedure was used to determine the wind
pressure for use in the verification model.
1. Determine the wind pressure from the wind speed data for each
anemometer using the following method.
P = 0.00256F2 (3.4)
where: P- Stagnation pressure (psf) at sea level
V = Velocity (mph)
Based on Equation 3.4, the pressure at the location of each
anemometer is determined from:
P = CC 0.00256F2 (3.5)
where: V = Wind velocity (mph) for each quadrant.
Cd= 2, which is the drag coefficient (also known as the
shape factor) for most bluff bodies such as
structural shapes (ASCE, 1961).
Ca = Altitude coefficient = ratio of the average ambient
air density for the elevation of the bridge to the average
ambient air density at sea level. Ca is determined for the
40


information in ASCE 7-02, Table C6-1 (ASCE, 2002).
Ca is simply a correction for actual air density.
Because gust conditions are inherent in field measurements of actual
wind velocity, the gust response factor was not applied to this equation.
The altitude of the Blue River Bridge was approximately 9000 feet (2743
meters). From Table C6-1 in ASCE 7-02, the ambient air density at this
elevation is 0.0584 ^J/^3 (0.9344 3) and the ambient air density for sea
level is 0.0765 ^J/^3 (1.2240 ^/3) Ca is calculated below:
c = = 0.7634
0.0765
For example, for quadrant 1, the wind pressure was calculated as
below:
P = CdCa 0.00256E2
P = (2)(0.7634)(0.00256)(39.84)2 = 6.20//=0.297kPa
2. Figure 3.10 shows a diagram was illustrating the locations of the
anemometers and the wind direction sensor.
41


WS5
WS2
WS4 *WS3
Figure 3.10. Diagram of Blue River Bridge, illustrating the locations of
anemometers (WS1-WS5) and wind direction sensor (WD).
3. At the Blue River Bridge, calculations of wind pressure were
completed at four quadrants. The wind pressure on quadrant 1 was
determined from a weighted average from the velocities measured at
fVSj and WS2. The wind pressures at the other quadrants were
similarly determined. Figure 3.11 shows the four quadrants of the
bridge.
42


Figure 3.11. Blue River Bridge. Quadrants subjected to different uniformly
distributed wind pressures. Wind pressure on quadrant 1 was determined from a
weighted average from the velocities measured at WS1 and WS2. Wind
pressure at the other quadrants were similarly detennined.
4. Apply the wind pressures to the 3D verification model. Figure 3.12
shows the wind pressures on all quadrants.
43


Figure 3.12. Wind pressure applied to the four quadrants for analysis. North is
to the right.
5. Modify the deck in RISA 3D as follows:
a. Change the four main bearings to fixed because the actual
bearings were assumed frozen in rust and dirt.
b. Leave the bearings at the stringers as rollers.
c. Change the internal end releases of all members except rods
to fixed because it can be assumed that there is no rotation
at the ends of riveted or bolted members under the relatively
low wind pressures measured.
6. Run the verification model in RISA 3D.
44


7. Generate a table of the forces measured in the bottom chords and
portals.
8. Compare the results from both the verification model and the
measured data.
45


4. Modeling and Analysis
4.1 Introduction
In this chapter, the Blue River Bridge is modeled and analyzed using a
commercial three-dimensional structural analysis program, RISA 3-D (Risa
Technologies, 2001). RISA 3-D is a finite element software program that
includes both frame elements and plate elements. When dealing with bridges of
all types in the United States, the AASHTO code is the primary design code
used. AASHTO provides design requirements for pedestrian bridges, as well as
vehicular bridges.
There are three main loads that a pedestrian bridge structure will be
exposed to. These loads are gravity, wind, and seismic loads. Gravity loads
consist of the self-weight of the structure (dead load), a pedestrian loading (live
load), and depending on its location, the weight of snow and ice. According to
the AASHTO code, the design live load values varies from 65 psf to 85 psf
(3.11 kPa to 4.07 kPa), depending on the area of the walkway to which the load
is applied. The design wind load is also 75 psf (3.59 kPa) (AASHTO, 1997).
46


In this research, the bridge was analyzed under dead load, wind load, and
wind plus dead load for two different model configurations: a skeleton model
and a deck model. The results were compared at the end of this chapter.
A gravity load study of the Blue River was also completed by Davis
(2005). Three analysis methods-classical method of joints, Risa 3D modeling
software and field experimentation- were used in order to gather concrete data to
verify that the bridge is still performing correctly when gravity loads apply. The
results of this research are included in Appendix C.
4.2 Modeling Consideration
Two different model configurations were analyzed: a skeleton model and
a deck model, under the same considerations.
The external boundary conditions for both configurations include pinned
joints at one end and roller joints at the other. The roller joints are restrained
from lateral translation but are free to translate in the longitudinal direction the
bridge. All internal member connections were assumed to be pinned and
released to rotate. These assumed internal boundary conditions are normal
assumptions typically made in the bridge design industry.
47


4.3 Analysis and Results
On a pin-connected truss bridge, there are certain members where a
failure, if one is to occur, is expected. One such location is the bottom chord
eyebars in the center panel of the truss. Failures are expected to occur in these
members first because under gravity loads, the highest tensile forces occur at
these locations and under wind loads these elements will be the first members to
go into compression on the windward side.
Figure 4.1 shows the skeleton frame of the Blue River Bridge modeled in
RISA-3D. The external bearing conditions, as previously described, include a
pin at one end and a roller at the other end.
Figure 4.1. Illustration of the traditional skeleton based on the primary
members only.
48


Figures 4.2 and 4.3 illustrate the superimposed vertical dead loads and
their axial load effects on the bottom chord eyebars for the skeleton structure.
Live load is not included. As shown in Figure 4.3, the maximum tensile force
under gravity loads occurs at mid span.
Figure 4.2. Representation of superimposed gravity loads.
49


Figure 4.3. Relative axial forces in the bottom chord eyebars due to
gravity loads for the skeleton structure.
Figures 4.4 and 4.5 illustrate lateral wind load and its effect on the
skeleton structure. Figure 4.5 shows the axial forces in the bottom chord
eyebars due to the wind load. As expected, the windward bottom chords go into
compression and the leeward bottom chords go into tension, with the exception
of those on the panel closest to the pinned end on the leeward side. On the
windward side, however, this member did not go into tension. The compression
force reduced significantly compared to the other panels. This reversal of the
sign in the member force on the leeward side is very similar to the basic
propped cantilever shown in Figure 4.6. However, a pure cantilever condition
does not exist because while the bottom chords are free to move longitudinally
along the axis of the bridge, they are laterally restrained. The bottom chord of
50


the truss acts like a beam rigidly connected at the pin-pin end and simply
supported on the roller-roller end. As a result, the maximum compressive force
on the windward side and the maximum tensile force on the leeward side shifted
toward the free end (instead of the center span), which is less stiff. Also, the
maximum compression force due to wind load was greater than the maximum
tension force due to the dead load only on the windward side that caused
significant buckling in the eyebars. A logical engineering solution to solve the
Kl
bucking problem would be to decrease the - (slenderness) of these members.
Figure 4.4. Representation of wind pressure on the bridge
51


Figure 4.5. Graphical representation of axial forces in the bottom chord
eyebars due to wind for the skeleton structure.
Mi
wl
Ml 1 J
w -s-X-sH
?-
Shear
UM
'1th,
l
4
Momenf
XT
N
A
max
JL
Figure 4.6. Shear and moment diagrams for propped cantilever
condition, from the AISC Manual of Steel Construction (AISC
2001)
52


The next step was to add plate elements to the skeleton model with
stringers. The timber deck, as described in Chapter 2, consists of longitudinal
running boards on transverse timbers on steel stringers. The transverse deck
boards and the longitudinal running boards oriented perpendicular to them were
treated as a solid single plate instead of as individual boards. This was because
the two mutually perpendicular layers, well spiked together, were believed to act
as a single element. The deck was modeled using 245 plate elements. In this
model, both timber deck and steel stringers were included as shown in Figure
4.7. The model used frame elements for all members except for the deck, which
is modeled with RISA plate/shell elements.
Figure 4.7. Skeleton model with stringers and deck elements
53


The steel stingers bear on the floor beams as shown in Figure 4.8. The
floor beam frame elements were offset by the offset elements as shown in Figure
4.9 to represent the stacking of actual stringers on the floor beam. The
connection between stringers and floor beams were pinned. The plate elements
were offset in a similar manner, again to represent the stacking of actual deck on
the stringers. The connections between the deck and stringers were also pinned.
Figure 4.10 illustrates the rendering of the timber deck on steel stringers on steel
floor beams.
Figure 4.8. Rendering of steel stringers on steel floor beam
54


LONGITUDINAL DECK TIMBERS
TRANSVERSE DECK TIMBERS
FIXED
OFFSET ELEMENT
RELEASE
OFFSET ELEMENT
FLOOR BEAM
STRINGER
FRAME ELEMENT
STRINGER
Figure 4.9. Offset members and release locations
Figure 4.10. Rendering of the timber deck on steel stringers on steel
floor beam. The mutually orthogonal deck timbers are treated as a single
monolithic solid.
55


The final case was to treat the deck as a diaphragm. The diaphragm
model is a very stiff model that is rigid in the plane of the deck. As a result, the
axial forces in the bottom chord eyebars under wind are lower, compared to the
skeleton model values, as shown in Table 4.1. Note that the negative sign
represents a compression force and the positive sign represents a tension force.
56


Table 4.1. Summary of Maximum Axial Forces in Bottom Chord
eyebars. Forces are for windward side and are expressed are in
kN (kips), followed by percent reduction in compression (or
increase in tension) compared to the traditional skeleton value.
(Positive = tension; negative = compression). Note the deck and
diaphragm values are virtually identical, suggesting the deck, as
modeled, is about as stiff as possible._______________________
Model Axial force due to dead load only Axial compression due to wind load only Net axial force due to wind plus dead load*
Case 1:
Skeleton 57 -60 -1.4
(Figure 26) (12.6) (-13.3) (-0.3)
Case 2:
Deck 56 -50 9.4
(Figure 31) (12.5) (-11.2) (2.1)
Case 3: 1% 16% 600%
Diaphragm 56 -46 13.2
(12.5) (-10.4) (3.0)
1% 22% 900%
The percent change from the skeleton case was determined for the deck
model from:
% change =100 x
Fskeleton Fdeck
Fskeleton
(2.1)
and for the diaphragm model from:
% change=100 x
Fskeleton Fdiaphragm
Fskeleton
(2.2)
where:
57


Fskeleton = calculated force in windward bottom chord from the skeleton model
Fdeck = calculated force in windward bottom chord from the deck model
Fdiaphragm = calculated force in windward bottom chord from the diaphragm
model
4.4 Eyebar Condition
As can be observed in Figure 4.11, the eyebars were buckled
significantly under the applied loads. This might be because of the abutment
failure shown in Figure 4.12. The bridge abutment was buried in the soil and
retained by wood boards and steel plates. Since the abutment wall failed, soil
pressure was applied to the bridge laterally, placing the bridge into axial
compression. This axial force adds to the compression due to wind loads. The
combination of these compressive forces caused buckling in the bottom chords.
Despite the eyebar buckling, the bridge has not failed yet. This might be
because of the stiff deck which could carry tensile forces. Alternatively, the
bridge may be exhibiting arch behavior, causing compression forces to transfer
to the portals.
58


Figure 4.11. Buckled eyebars in the Blue River Bridge.
Figure 4.12. Abutment failure in the Blue River Bridge.
59


4.5 Conclusions
According to Table 4.1, adding the deck to the skeleton model does not
change the tensile force in the bottom chord eyebars significantly for the dead
load case (1.75 percent reduction). However, for the wind load case, the decks
presence reduces the compression force by 16.67 percent because of the bridges
increased lateral stiffness. However, the compression force was not completely
eliminated.
For the wind plus dead load case, the bottom chord eyebars were still in
compression in the skeleton model. However, they went completely into tension
in the deck model, which means the deck was stiff enough to eliminate the
compression force.
Table 4.1 also illustrates the differences between the deck and diaphragm
models. In the dead load only models, the tension forces in the bottom chord
eyebars are identical and for the wind load models, the forces are within 8
percent. Since a diaphragm model is a rigid model with a very large stiffness,
the deck is about as laterally stiff as theoretically possible.
60


5.
Verification Method
5.1 Introduction
In this chapter, strain data collected during the field testing was
compared to the analytical strain data under the actual wind pressure to verify
the analysis which was completed in Chapter 4.
In order to collect the data in the field, the bridge was instrumented as
described in Chapter 3, such that the strain data from selected members could be
collected simultaneously with the wind speed and direction data. The forces
from the collected data were compared to the forces calculated using the RISA-
3D model under the actual wind pressure. If the calculated and measured data in
this chapter are identical or close, one can make conclusions about the lateral
stiffness of the deck.
5.2 Verification Results
The instruments were set up on the Blue River Bridge as described in
Chapter 3, on April 16, 2005. Data used in the verification model was collected
during the afternoon of April 20, 2005. Data was collected continuously for 28
days. The collected data contained strain in milli-volt per volts, wind speed in
61


mph, wind direction, date and time. Total of 2400 wind events was used in the
verification method. Wind speeds greater than 20 mph and wind direction from
east to west (90+30) or west to east (270+30) were the parameters that
triggered the data logger to begin collecting. The data was then transferred form
the logger to the laptop computer using the cellular modem and antenna.
A total of sixteen strain transducers were installed on the eyebars where
the maximum axial forces occur and also on the south portals where the
maximum moments occur in order to measure the strain in these members under
the wind loads. Figure 5.1 shows the location of the strain transducers.
Figure 5.1. Diagram of Blue River Bridge, illustrating the locations of
the strain transducers. North is to the left. The wind direction was from
east to west, orthogonal to the bridge. Strain transducers Gl, G2, G3 and
G5 were clamped to the windward bottom chord eyebars. G4, G6, G7
and G8 were clamped to the leeward bottom chord eyebars. G9 and Gl 1
were clamped to the top of the south portals. G12 and Gl 8 were
clamped to the bottom of the south portals. G13 and G14 were clamped
to the east stringers in the middle bay and G15 and G17 were clamped to
the west stringers.
62


A total of five anemometers were installed to obtain the wind speed data.
Figure 5.2 shows the locations of the anemometers and the wind direction
sensor. Rather than use the absolute values of the wind velocity, the average of
the high and low velocities were used in this study. This was done to reduce the
potential for slack as much as possible (refer to Chapter 3, Section 3.4). Both
the low and high wind velocities correspond to the average of 20 consecutive
data measurements over a two-second interval. The averages were utilized
because wind speeds fluctuate considerably over very short time intervals (0.1
seconds). Table 5.1 shows the wind velocities for each anemometer installed on
the Blue River Bridge for April 20, 2005 wind event.
WS5
*WS2
Figure 5.2. Diagram of Blue River Bridge, illustrating the locations of
anemometers (WS1-WS5) and wind direction sensor (WD). North is to
the left. WS1 was positioned at the approximate center of the wind
intercept area. WS2 and WS5 were located above the end diagonal in
the portals. WS3 and WS5 were located below the bridge an the
elevation mid-height between the bridge deck and the water surface
below.
63


Table 5.1. Wind Velocities, Quadrant Average Velocities, and Quadrant
Pressures for April 20, 2005 wind event
Anemometer Location Velocity m/s (mph) Average velocity for quadrant m/s (mph) Average pressure for quadrant Pa (psf)
WS1 Central 14.06 (31.46)
WS2 South upper 15.31 (34.24) 14.69 (32.85) 230.30 (4.81)
WS3 South lower 11.16 (24.96) 12.61 (28.21) 110.12 (2.30)
WS4 North lower 10.64 (23.81) 12.35 (27.64) 121.62 (2.54)
WS5 North upper 16.69 (37.33) 15.38 (34.40) 205.89 (4.30)
Rather than apply a uniformly distributed wind pressure on the bridge
members, an attempt was made to account for different pressures at different
parts of the bridge. Since the data from the five anemometers were not identical,
the bridge truss was considered as four different quadrants as shown in Figure
5.3. The wind pressure on each quadrant is determined by using Equation 3.5,
described in Chapter 3, Section 3.6. For example, for quadrant 1 which is
located on the upper south part of the bridge, the wind pressure was calculated
as below:
Ap = CdCa0.00256(Vhigh2-Vj)
64


where: Ap = The difference between high wind pressure and low
wind pressure
Vhigh = 2-second average of high wind velocity measure by
the anemometer
Cd- 2 (refer to Chapter 3, Section 3.6)
Ca = 0.7634 (refer to Chapter 3, Section 3.6)
For WS{: Ap = 2x0.7634*0.00256(39.842 23.0892) = 4.12psf
For WS2:Ap = 2x0.7634jc0.00256(44.492 -23.992) = 5.49psf
The wind pressure on quadrant 1 was determined from a weighted
average of calculated wind pressures at WS: and WS2.
' Quad. 1
4.12 + 5.49
2
4-81 psf
The wind pressures at the other quadrants were similarly determined.
The results are shown in Table 5.1.
65


Figure 5.3. Blue River Bridge. Quadrants subjected to different
uniformly distributed wind pressures. Wind pressures on quadrants are
determined by weighted averages.
Finally, the lateral forces applied to the bridge members on each
quadrant were determined by multiplying the wind pressure by the depth of each
member. Figure 5.4 shows the actual wind loads applied to the bridge members
on all quadrants. These wind loads were then incorporated into the RISA-3D
model, described in Chapter 4, by making a new load case (WQuad 1+ +QuadA).
66


Figure 5.4. Wind pressure applied to the four quadrants for analysis.
North is to the right.
Modifications to the RISA-3 D model were made to more accurately
model the actual conditions (see Chapter 3, Section 3.6). Such modification is
to change both end and internal boundary conditions. The bridge was then
analyzed under the actual wind loads. The results from the RISA-3D model for
the bottom chord eyebars and the south portals are summarized in Table 5.2.
This table also includes the measured forces at the same members. These
measured data, collected during the field test, were downloaded to a laptop
computer and converted into an analysis spreadsheet (Rutz, 2005). The basic of
this spreadsheet is included in Appendix D. The measured wind speed, wind
direction, strains at the bottom chord eyebars, and strains at the south portals
67


Wind Speed (mph)
were plotted. Figures 5.5 through 5.9 are the graphical representatives of the
collected data from the Blue River Bridge.
WIND SPEED
Blue River Bridge 4/20/05
Figure 5.5. Wind Speed as measured by the five anemometers. The
average of all five anemometers is shown in bold line.
68
Wind Speed (m/s)


WIND DIRECTION
Blue River Bridge 4/20/05
240
180
o

5 120
o
c
t
60
0
0 60 120 180 240
Time (seconds)
f- Wind Direction |
Figure 5.6. Wind direction as measured during the test.
60
BOTTOM CHORDS
Blue River Bridge 4/20/05
90
120
Time (seconds)
150
180
- WS avg Leeward Strain Windward Strain Leeward filtered Windward filtered
Figure 5.7. Strain measurements for the windward and leeward bottom
chord eyebars. The average wind speed, to an arbitrary scale, is also
shown as the bold line at the top. Strain in the leeward eyebar is shown
above strain in the windward eyebar. Both the raw data and a filtered
line that removes much of the signal noise are shown.
69


BOTTOM CHORDS: Change in strain
Blue River Bridge 4/20/05
Figure 5.8. Enlargement of the trace for windward and leeward bottom
chord eyebar measured strains. Both are baseline traces of the filter data.
Thus, they represent the change in measured strain starting from the
same point in time as the corresponding change in wind velocity. The
wind velocity is shown for reference at the top of the graph to an
arbitrary scale.
SOUTH PORTAL
Blue River Bridge 4/20/05
WSavg
_.JUJ__
Gy G1U
..-..£12...................... G9 2-S^c Ayg
Figure 5.9. Measured strains at the south portal are shown.
70


Table 5.2. Blue River Bridge Verification Summary. Comparison of
calculated forces to measured forces expressed in kN (kips) and kN-m
(ft-kips). The north portal was not instrumented.
Member Calculated Force Measured Force Correlation: % difference
Windward bottom chord -0.44 kN (-0.1 kips) -0.54 kN (-0.122 kips) 23%
Leeward bottom chord 1.44 kN (0.32 kips) 0.78 kN (0.18 kips) 45%
South portal upper 0.77 kN-m (0.56 ft-k) 0.70 kN-m (0.51 ft-k) 15%
South portal lower 1.25 kN-m (0.91 ft-k) 0.55 kN-m (0.40 ft-k) 56%
* E = 10.34 x 106 KPa (1.5 x 103ksi)andE = 13.79 x 106 KPa (2.0 x
103 ksi) were both used in RISA-3D modeling and both gave the same
results, therefore the deck stiffness had little or no effect on member
forces. E = 13.79 x 106 KPa (2.0 x 103 ksi) is considered an upper bound
on the stiffness of the wood in the deck, so higher stiffnesses were not
considered.
5.3 Conclusions
Good correlations were achieved for the windward bottom chord and the
south upper portal, suggesting that the deck model is a reasonable
approximation. The weaker correlation for the leeward bottom chord may have
been influenced by the buckled condition of the existing eyebars which may
have changed the measured force. Additionally, the wind pressure was not
directly perpendicular to the bridge during the entire event (between 78 and
71


97), which could reduce the measured force in the leeward bottom chord. The
weaker correlation for the south lower portal may be because the calculated
value is strongly influenced by the portal bases, boundary conditions, which
may not be entirely accurate. The actual bearings could not be observed because
they were buried in soil. The bearings were presumed to be rusted to a frozen
condition and were treated as fixed in the calculations. However, they may not
be truly fixed, and may have permitted some small rotation, that is, they may
have exhibited some small degree of partial fixity, which would have altered the
calculated results. Lower portions of the south portal members were also
shielded from wind due to an existing fence on the east side of the bridge.
Therefore, the wind pressure in the RISA-3D model may be higher than the
bridge actually experienced in the field due to the shielding effect of the fence.
72


6.
Conclusions and Recommendations for Future Research
6.1 Introduction
The overall purpose of this research was to study the effect of the lateral
load paths in the historic Blue River Bridge. While the AASHTO specified
wind pressure of 75 psf (3.59 kPa), which corresponds to approximately 121
mph (194 km/hr), did not occur, the actual wind velocity was measured in the
field and the wind pressure calculated. Strains at the south portals and bottom
chord eyebars were also measured during the test. The bridge was then modeled
using RISA-3D finite element software under the actual wind load measured
during the test. The deck model was modified to reflect the actual
configurations. Such modifications were to change the boundary conditions to
better reflect the actual conditions at the supports. Removing the internal
member-to-member release to better reflect the actual end rotations of internal
members was also utilized.
The results from the deck model, calculated strains under actual wind
pressure, were then compared to the measured strains to verify the analysis of
the bridge. The results from the RISA-3D analysis and field test correlate well,
so the deck model was a reasonable approximation of actual conditions.
73


6.2 Conclusions
The conclusions of this research are summarized below:
The presence of a deck stiffens a truss bridge laterally.
The deck model reduces calculated forces in the bottom chords
compared to the skeleton model. Therefore, deck provides more lateral
stiffness for the bridge truss.
Calculated forces correlate well with measured forces. Therefore, the
deck model is a reasonable approximation of actual conditions.
The actual wind pressure, measured at five different locations, was not
constant. Therefore, zonal wind analysis is appropriate.
6.3 Recommendations for Future Research
Account for the stiffening effect of the deck in historic bridge
preservation projects. This will reduce the axial compression forces in
the bottom chords. Note that there is no construction cost.
Add dead load to the deck. This will decrease the compression and
increase the tension in the bottom chords. Note that there is an upper
bound to the amount of additional dead load (member stresses are
limited by the allowable stress). There will be construction cost.
Analyze the bridge for different boundary conditions. The bearings may
not be truly fixed, and may have permitted some small rotation, that is,
74


they may have exhibited some small degree of partial fixity. This will
affect the calculated moments in the portals and axial forces in the
bottom chords.
Investigate bridges with different stringers and deck arrangements. This
includes the steel stringers welded (fixed) to floor beams, concrete decks
on steel stringers, gravel road base on corrugated metal decks on steel
stringer, etc.
Apply uplift force to the bridge deck. The difference between the wind
speed on top and bottom of the bridge deck can cause an uplift force.
According to American Association of State Highway and
Transportation Officials (AASHTO 1996) an upward force shall be
applied at the windward quarter point of the transverse superstructure
width. This force shall be either 20 or 6 (depending on the group
combination) pounds per square foot (0.958 or 0.287 kPa) of deck and
sidewalk plan area. At the Blue River Bridge, wind velocity was
reduced on top of the deck because of the existing fence and railing.
Therefore, uplift force was not considered in the analysis of the bridge.
75


APPENDIX A
BLUE RIVER BRIDGE PHOTOGRAPHS
A.l Introduction
The following is a photo gallery of the Blue River Bridge
members and field investigation taken by the author. The captions
contain brief information about photographs.
76


Dillon dam, built in 1960, at the local east side of the bridge.
Blue River Bridge, front view.
i
77


Blue River Bridge, side view.
Blue River Bridge, floor beams and stringers.
78


Blue River Bridge, curb corrosion .
Blue River Bridge, deck-curb connections.
79


Vertical post corrosion at mid span.
Buckling at the west bottom chord eyebars.
80


Abutment failure, north side.
81


82


Blue River test set-up in progress.
Strain transducer installation on stringers.
83


Strain transducer installation on eyebars.
Strain transducer, installed on upper south portal.
84


Anemometer mounting.
85


i
1
Anemometers, installed at west mid span (under the bridge).
Anemometer, installed at south-east end span (above the bridge).
86


CR5000 Data logger with all cables connected to channels.
Laptop computer, used to upload the program to the data logger and
download the data tables from the data logger.
87