Citation
Wind load analysis of a truss bridge at Rifle, Colorado

Material Information

Title:
Wind load analysis of a truss bridge at Rifle, Colorado
Creator:
Swigert, William Barry
Place of Publication:
Denver, Colo.
Publisher:
University of Colorado Denver
Publication Date:
Language:
English
Physical Description:
ix, 60 leaves : illustrations ; 28 cm

Thesis/Dissertation Information

Degree:
Master's ( Master of Science)
Degree Grantor:
University of Colorado Denver
Degree Divisions:
Department of Engineering, CU Denver
Degree Disciplines:
Civil Enginering
Committee Chair:
Rens, Kevin L.
Committee Members:
Durham, Stephen
Janson, Bruce

Subjects

Subjects / Keywords:
Truss bridges -- Aerodynamics -- Colorado -- Rifle ( lcsh )
Wind-pressure ( lcsh )
Rifle Bridge (Rifle, Colo.) ( lcsh )
Genre:
bibliography ( marcgt )
theses ( marcgt )
non-fiction ( marcgt )

Notes

Bibliography:
Includes bibliographical references (leaves 59-60).
General Note:
Department of Civil Engineering
Statement of Responsibility:
by William Barry Swigert.

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:
227989494 ( OCLC )
ocn227989494
Classification:
LD1193.E53 2007m S84 ( lcc )

Full Text
WIND LOAD ANALYSIS OF A TRUSS BRIDGE
AT RIFLE COLORADO
by
William Barry Swigert
BSCET Metropolitan State College, 1976
A thesis submitted to the
University of Colorado at Denver and Health Sciences Center
In partial fulfillment
Of the requirements for the degree of
Master of Science
Civil Engineering
2007


This thesis for the Master of Science
degree by
William Barry Swigert
has been approved
Date


i
Swigert, William B., MS, Department of Civil Engineering
Wind Load Analysis of a Truss Bridge at Rifle Colorado
Thesis directed by Dr. Kevin L. Rens, PhD.
ABSTRACT
Current AASHTO requirements for pedestrian bridges may indicate some
historic bridges to be under-strength when applying wind loads to simple skeleton
models. The Rifle Bridge over the Colorado River at Rifle, Colorado, is a historic
steel truss structure that was analyzed with actual wind loads on the existing structure.
The structure is a 240 ft span Pennsylvania truss, built in 1909 and in service until the
late 1960s. This paper discusses the effects of including stiffening elements in 3D
models by comparing actual and calculated wind loads. During the six week wind
study period, a total of 207 events were recorded, of which the three most responsive
were used. Analytical models included the metal deck combined with asphalt
pavement as a stiffening element, which is treated as plate elements with a modulus
representative of the composite section. An analysis of main members of this bridge
using AASHTO loads is included.
I
]
!
This abstract accurately represents the content of the candidates thesis. I recommend
its publication.
Signed
in


ACKNOWLEDGEMENTS
The University of Colorado at Denver has completed this work funded in part
by Grant #MT-2210-04-NC-12 from the National Center for Preservation Technology
Transfer and Grant #2004-Ml-019 from the State Historical Fund of the Colorado
Historical Society. Further, the cooperation of the bridge owner Garfield County,
CO, through Garfield County Commissioner John Martin is gratefully acknowledged.
Most of all, I would like to thank Dr. Kevin Rens for his direction on this
research, and to Dr. Fred Rutz for piquing my interest in the study of historic truss
bridges.
Lastly, to my wife Dieanna and family Will, Bob, Ruth, Heather, Makayla,
and young William for allowing me the time to complete my studies.
I
iv


CONTENTS
Figures................................................................ vii
Tables................................................................. ix
Chapter
1. Introduction.......................................................1
1.1 Purpose..................................................... 1
1.2 Goals........................................................... 1
1.3 Previous Research............................................... 2
2. Rifle Bridge.................................................... 3
2.1 Location........................................................3
2.2 History.........................................................4
2.3 Field Documentation of the North Span............................ 10
3. Modeling......................................................... 16
3.1 Introduction..................................................... 16
3.2 Skeleton Model................................................... 16
3.3 Deck Model....................................................... 19
3.4 Diaphragm Model.................................................. 24
3.5 Verification Model................................................25
4. Wind Load Analysis................................................27
4.1 Historic Wind Loads.............................................. 27
4.2 AASHTO Wind & Live Loads......................................... 27
v


4.3 Verification Loads.............................................. 27
4.3.1 Dead Load....................................................... 28
4.3.2 Field Measured Wind Load........................................ 29
5. Results..........................................................43
5.1 Verification Load Results....................................... 43
5.2 Deck Model Stiffness Comparison..................................45
5.3 AASHTO Code Conformance..........................................49
6. Conclusions & Recommendations................................... 51
6.1 Conclusions......................................................51
6.2 Recommendations for Future Research............................. 52
Appendices
A. Field Measurements North Truss...................................54
B. Historic Bridge Inventory....................................... 57
C. Data Collection................................................. 58
References.............................................................60
vi


LIST OF FIGURES
Fig. Description
2.1 Location of Rifle Bridge.............................................. 3
2.2 Rifle Bridge over the Colorado River at Rifle, Colorado................4
2.3 Satelite photo of the Rifle Bridge.....................................6
2.4 Historic Photo sunken pier, abutment...................................8
2.5 Plaque.................................................................9
2.6 Photo looking downstream..............................................10
2.7 Typical bridge framing members....................................... 12
2.8 Member sizes used in the FEA models.................................. 13
2.9 Rifle Bridge deck.................................................... 14
2.10 Rifle Bridge bearing at the northeast support.........................15
3.1 3D skeleton model of the Rifle Bridge................................ 18
3.2 Shear and moment diagrams for propped cantilever................... 18
3.3 Transverse wind loads applied to the skeleton model.................. 19
3.4 3D model..............................................................19
3.5 Skeleton model with stringers and deck................................21
3.6 Rendering of stringers on a steel floor beam..........................21
3.7 Rendering of the asphalt paving over steel deck planks................22
3.8 Representation of the diaphragm feature...............................25
4.1 Representation of superimposed gravity loads......................... 28
4.2 Forces in the bottom chord eyebars due to gravity loads.............. 29
4.3 Diagram illustrating the locations of anemometers.................... 30


4.4 Installation of the anemometers.....................................30
4.5 Photo of installed anemometers......................................31
4.6 Diagram illustrating the locations of the strain transducers........31
4.7 Strain Transducer...................................................32
4.8 Strain Transducer mounted on the top chord..........................32
4.9 Strain Transducers mounted on the bottom chord eyebars..............33
4.10 Record of wind events...............................................35
4.11 Graph of Event 1....................................................35
4.12 Graph of Event 2....................................................36
4.13 Graph of Event 3....................................................36
4.14 Wind distribution on bridge.........................................38
4.15 Wind pressure applied to the four quadrants.........................42
4.16 Axial forces in the bottom chord eyebars due to wind................42
viii


LIST OF TABLES
Table
4.1 Wind Velocities, Quadrant Velocities, and Pressures for Event 1.....39
4.2 Wind Velocities, Quadrant Velocities, and Pressures for Event 2.....40
4.3 Wind Velocities, Quadrant Velocities, and Pressures for Event 3.....41
5.1 Comparison of calculated and measured values for Event 1............43
5.2 Comparison of calculated and measured values for Event 2............44
5.3 Comparison of calculated and measured values for Event 3............44
5.4 Comparison of the analysis results from the three models studied....47
5.5 Stress ratio in percent of material yield stress....................49
i
!
IX


1. Introduction
1.1 Purpose
This study is intended to aid in the preservation of historic truss bridges. As
the steel truss bridges in this country, built in the late 1800s and early 1900s,
become unfit for modem vehicular traffic, there is typically only one other function
they can serve pedestrian bridges. Modem codes for this usage may calculate the
bridges to be under strength. The field work and analyses of this study primarily
focuses on lateral (wind) loads applied to the truss bridge at Rifle. It is intended to
provide not only specific information for the re-use of this bridge, but also others
similar to it throughout the country.
1.2 Goals
The goal of this study is to provide designers with additional methods to
analyze wind loads on truss bridges. This is achieved by the following:
1. Evaluate several types of deck models. Previous research has shown that
the deck does add stiffness, and thereby assists in resisting loads.
Different models were used to evaluate how much stiffness can and should
be assigned to this model.
2. Evaluate three different wind events. Since field recorded wind load data
can vary from analytical models, several wind events were studied to help
verify analysis results, and eliminate anomalies.
3. Develop a model that correlates with the field data. Comparison between
analytical models and field recorded data should result in an accurate
finite analysis element model.
4. Provide a general evaluation of the bridge with regard to current
pedestrian bridge code requirements. The Rifle Bridge has been closed to
all uses for many years. This study may assist in its re-opening.
1


1.3 Previous Research
Studies that were the basis for this work include Lateral Load Paths in
Historic Truss Bridges, a PhD thesis completed by Dr. Fred Rutz (Rutz 2004). In
it, he studied the Bridge at Fruita, Colorado, which is similar to the Rifle Bridge.
The general deck analysis methodology and wind load application was the basis
for several studies, including this one.
The second study is Load Paths in Historic Truss Bridges (Rutz, Rens,
Jacobson, Hamedian, Elias, Swigert, 2005), which was a report prepared for and
funded by the National Center for Preservation Technology Transfer. This
document evaluated five bridges, which include the Rifle Bridge and the Fruita
Bridge, and compared results, which assisted in standardizing the results for
bridges with different deck types, spans, and dead loads.
The last is a study completed by T.V. Herrero entitled Wind Pressure and
Strain Measurements on Bridges Part II: Strain Transducer Development
(Herrero 2003). This study developed the strain transducer by analysis and field
testing, and allowed subsequent studies to easily install and measure strains in
bridge members.
2


Part 2 Rifle Bridge
2.1 Location
The bridge is located between the City of Rifle, CO, on the north, and
Interstate 70 on the south (USGS 1982). It is oriented 10 to the west from north,
so local north was assumed to be global north. The winds are predominately from
the west, and the Colorado River Valley in this area is very open and flat to the east
and west. The bridge has been barricaded to prevent access by vehicles and
pedestrians for the last 20 years.
3


Figure 2.2. 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.
2.2 History
The City of Rifle was established in 1882 along the Denver & Rio Grande
Railroad, just north of the Grand (now Colorado) River. During this decade, Garfield
County erected a timber wagon bridge across the river. Since the bridge was located
at a tight bend in the river, the flow continually undermined the bridges substructure.
In 1901, the north abutment was replaced with concrete, which would probably
indicate the original foundations were masonry or timber. Continual scour occurred,
as repairs were reported almost every year from 1903 thru 1908 (Colorado Historical
Society 2000).
4


In November of 1908, the County Board of Commissioners voted to replace
the bridge, awarding a contract to Denver-based C.G. Sheely for $26,872. The
superstructure was fabricated by Minneapolis-Moline Power Implement Company,
and was completed in the autumn of 1909 (Colorado Historical Society 2000). There
is no record of whether the portions of the previous foundations were used, or
completely replaced.
In 1922, a concrete jetty or training wall was constructed, extending from
the south abutment into the main channel, upstream of the bridge. This wall was
intended to divert the flow north towards the bend in the channel in an effort to
protect the south abutment.
5


Figure 2.3. Satelite photo of the Rifle Bridge (Google 2007). The Colorado
River flows from right to left (east to west). The arrow in the center of the
photo points to the concrete jetty.
The Glenwood Post reported in the May 23, 1929 article that the county
bridge across the Colorado river was in danger of collapse, due to the high water
which had underlashed the cement pier in the center of the structure. (Glenwood
Post 1929). The article continues to report that the newly installed concrete wall had
6


diverted the current so that the main force of the same was directed against the pier,
which had settled more than two feet, resulting in a severe strain on the truss and
seriously threatening entire structure. In addition to it providing a river crossing, the
bridge also supported a pipe which provided the main water supply for the town of
Rifle. Loss of this water line provided the greatest apprehension to the town.
It is not known what types of repairs were instituted, but today the center pier
is approximately plumb. In the description for the photo below, there is also
discussion of deterioration of the northeast shoe at the abutment. It is also
questionable what repair was performed at this, or any of the other supports, or what
stresses in the bridge superstructure the pier settlement created, and to what extent
they were relieved as a result of the repair process.
7


Figure 2.4. In 1929, twenty years after it was built, bridge inspectors at the
Rifle Bridge in Garfield County documented a sunken pier, top, and
deterioration and corrosion on the northeast shoe of the abutment, bottom.
(CDOT 2004). Note the angled water mark on the pier, which would indicate a
settlement of more than 1 ft.
In 1958, a new prestressed concrete girder bridge was constructed lA mile
upstream to replace this bridge. For the next 25 years, the truss bridge was closed to
vehicular traffic, and used for pedestrian purposes only. In the early 1980s Garfield
County barricaded both ends of the bridge to discontinue pedestrian use. The deck
had deteriorated to a point that the County felt it unsafe for walking traffic.
8


Figure 2.5. Plaque at the south end of the bridge commemorating a fatality at
the bridge.
Although owned by Garfield County, the City of Rifle considered purchasing
the bridge, repairing, and re-opening it for pedestrian and possibly commercial use.
As of this writing, the bridge remains closed.
The Rifle Bridge has a Parker thru-truss at the south span, and a Pennsylvania
thru-truss at the north span. Both are today distinguished as the longest pin-
connected wagon trusses of their type found in Colorado, and each is one of only two
such trusses identified by the Colorado Historical Society bridge inventory (2000).
9


Figure 2.6. Photo looking downstream, depicting both north and south spans. The
jetty can be seen at the left.
2.3 Field Documentation of the North Span
Since the bridge owner, Garfield County, does not have records of the original
design or drawings of the superstructure, this researcher performed field
measurements of all the members on the north span, and documented them in
Appendix A. The two main vertical trusses supporting the edges of the bridge span
from center pier to abutment, and are typical of those fabricated in this era, in that
dimensions and member sizes are modular and repeated. There are thirteen panel
points measuring 20 ft on center, for an overall length of 240 ft. The maximum
10


height of the trusses is 34 ft, the deck is 18 ft wide, and the center to center spacing
between trusses is 19.17 ft.
Tension members (bottom chords and x-bracing) are rectangular bars.
Compression members (top chords and vertical web members) are built-up sections
of channels, plates, and/or lacing riveted together. Floor beams and purlins are I
beams, which are currently designated as S shapes by the American Institute of Steel
Construction (2005). X-bracing is located in the vertical plane of the main trusses,
and in the horizontal plane of both the floor and top chord panels. The horizontal x-
bracing is discontinued between the top chords at the ends of the bridge to allow
wagon or vehicle access. At these locations, there is a horizontal truss, configured as
a lattice, which in combination with the top chords, acts as a rigid frame, and is
normally referred to as a portal frame. Horizontal loads from wind or seismic
events collected by the x-bracing and the portal frame are transferred to the supports
in this manner.
11


Figure 2.7. Left photo depicts the typical built-up top chord and vertical web
members with channels and plates riveted together. The photo on the right depicts
the bottom chord, which are two bars pinned at the panel points. The main floor
beam is supported by and hangs below this connection.
12


Figure 2.8. Indicates the member sizes used in the FEA models. Shown here is the
Skeleton model; hence the deck is not present. Members not labeled are repetitive of
those labeled.
The built-up sections, even though labor intensive to build, did save weight
and were probably constructed as such due to the expense of materials and
availability of a relatively cheap labor force. Connections at panel points are made by
pins, which truly make them a pinned joint in a structurally analytical sense.
The steel typically used from 1905 thru 1930 had a yield strength of 30 ksi
(Rutz, F. R. 2004). All the members used in the Rifle bridge are assumed to be steel,
as was reported by the inventory document in Appendix B.
13


The bridge was last inventoried by Clayton B. Fraser on March 31, 2000
(Appendix B). In his report, he notes that the condition is fair. At the time of this
study, the general condition of bridge would appear to be unchanged. The bridge
overall is in need of painting, as corrosion is visually evident in several locations.
The metal deck has rusted out in several locations, so holes in the deck thru the
asphalt approximately 2 ft. square are evident. Bearings have corrosion, and may not
allow for movement as intended with the sliding and pinned connections. An under
deck rod x-brace is disconnected at about midspan, and hangs in the river.
Figure 2.9. Rifle Bridge deck, of which the asphalt surface is notably weathered.
Stringers in the right photo are steel, with wood and steel members at the edge of the
deck.
14


Figure 2.10. Rifle Bridge bearing at the northeast support. The concrete the shoe
rests on appears to be newer; however the front face has cracked allowing moisture to
enter.
!
15


3. Modeling
3.1 Introduction
This study constructed four finite element models: a skeleton model, a
deck model, a diaphragm model, and a verification model. Each is used for a
specific purpose. The skeleton, deck, and diaphragm models are analyzed with
current code loads (AASHTO 1997), and the verification model is analyzed with
actual field recorded wind loads.
The Rifle bridge was analyzed using RAM Advanse software (RAM
2005). RAM Advanse is similar to other 3D codes in that it is a tool that is readily
available to engineers.
3.2 Skeleton Model
The skeleton model is assembled based on a simple 2D framing
system (longitudinal trusses framing each side of the bridge), which would probably
be used by most engineers for a simplified or textbook analysis. It was constructed
using traditional truss elements composed of chords and web members longitudinally,
and floor beams together with x-bracing and lateral trusses in the transverse direction.
Members are pinned at their end connections where there is a true pin, and continuous
where plates or members are continuous thru the joint. For example, web (vertical)
members are continuous about a vertical axis and a horizontal axis longitudinal to the
bridge, and hinged about a horizontal axis transverse to the bridge. Boundary
conditions use a pinned support at one end (south), and roller support at the opposite
16


end (north). In a 3D analysis, using the x axis transverse to the bridge, the z axis
longitudinal to the bridge, and the y axis vertical, the resulting input would show
the following:
Pier (south) Support Abutment (north) Support
Axis Translation Rotation Translation Rotation
X fixed free fixed free
Y fixed free fixed free
Z fixed free free free
Figures 3.1, 3.3, and 3.4 portray the Skeleton Model with applied loads and
resulting stresses. Figure 3.1 reflects the 3D model. Figures 4.1 illustrates how
vertical (gravity loads) were applied to the main transverse beams, the ends of which
are supported by the vertical trusses at panel points, and the resulting stress in the
bottom chords. Figure 3.3 illustrates the application of wind load, which per code, is
applied to the vertical projection of the members. Note in Figure 3.4 the stress in the
bottom chords reverses in sign near the pinned supports. This effect is due to the
truss in the horizontal plane cantilevering from the supports fixed in translation as a
propped cantilever shown in Figure 3.2, and would not be readily apparent to
designers considering only a vertical truss in a 2D analysis.
17


Figure 3.1. 3D model of the Rifle Bridge, illustrating the traditional skeleton
model. The 3D structural engineering software RAM Advanse was used for
this bridge.
i
i
*v T 1 w l
|| || || iP
Th>^ m
Shear Ch l 4 Xr 7i
Unmonf
Mmax
_1
Figure 3.2. Shear and moment diagrams for propped cantilever
condition, from the AISC Manual of Steel Construction (AISC 2001).
The moment at the fixed end is analogous to a couple induced by lateral
loads at the pinned ends of the skeleton truss shown in Figure 3.1.
18


Figure 3.3. Transverse wind loads applied to the skeleton model. The
model used for analysis included longitudinal loads as well, not shown
for clarity.
i
I
Figure 3.4. 3D model indicating stresses due to wind load in the
bottom chords of the main trusses. Note the stress reversal near the
supports at the top of the figure, which are fixed for translation.
I 3.3 Deck Model
i
i
i
1 The deck model is derived from the skeleton model. Geometry,
member sizes, member to member connections, and boundary conditions are the
same. This model then adds the deck as a structural element. In this bridge, as with
many truss bridges, the deck is not located in the same horizontal plane as the bottom
19


chord panel points. The Rifle Bridge deck is framed with 18 I beams that are hung
by plates extending below the panel points. The 8 I beams at 2-3 on center frame
as stringers over the 18 beams, and are connected with (2) 5/8 bolts at the flange-to-
flange connection. The outer beams are bolted to 3 wood beams, which may
indicate the steel replaced original wood construction. A 2-1/4 metal deck was then
attached to the 8 beams. The wearing surface is a 2 layer of asphalt paving. Most
bridges in this period were originally framed with wood decking; it is not known
whether the current deck system replaced a wood deck, or if the asphalt/metal deck
was original.
The resulting deck system has its centroid located 5 above the bottom chord
panel points. The 3D model accounts for this eccentricity by placing rigid offset
elements at each beam/stringer intersection. The rigid offset elements are fixed at the
beam centerlines, and pinned at their intersection to model the flexibility of the
flange-to-flange bolted connection. The deck is then modeled with plate elements,
approximately square in dimension. The plate elements are also supported by rigid
offset elements, fixed at centerline of the stringers, and pinned at their intersection
with the deck. The pin was to model an assumed puddle weld. In a model with the
deck located at, and attached to the panel points, the resulting model would provide
more stiffness to resist lateral loads than would actually occur. This model, with its
rigid offset elements connecting the deck, more accurately allows for the loss in
stiffness due to the eccentricity.
20


Figure 3.5. Skeleton model with stringers and deck
Figure 3.6. Rendering of stringers on a steel floor beam. The rendering
was produced by RAM Advanse 3D. The dummy offset elements, used to
connect the floor beam and stringer, have a rotational release at the beam-
to-stringer interface.
21


Figure 3.7. Rendering of the 2 asphalt paving over 2.25 steel deck
planks on steel floor beam. For clarity, near stringers are intentionally
not shown. Offset elements, from the beam centerline, to the
beam/stringer intersection, to the stringer centerline, to the deck/stringer
interface were used. The rendering was produced by RAM Advanse.
The modeled representation of the deck shown in Figure 3.5 represents the
actual deck shown in Figure 1.13. The 2.25 x 0.125 corrugated metal bridge deck
(Vulcraft 1993) is topped with an average of 1 of asphalt pavement, all of which was
modeled as approximately square plate elements. This analysis included the asphalt
as a contributing factor to the stiffness. This was accomplished by calculating a
composite stiffness by using a transformed modulus composed of both steel deck and
asphalt.
This study analyzed several cases, of which the assumptions are believed to
bound the results presented in 5.2 Deck Model Stiffness Comparison, since the
stiffnesses used for the cases range from essentially no stiffness, to a case with the
22


maximum amount of stiffness a light gage steel deck could possess. The cases are as
follows:
o Case 1 Steel Plate Deck. This model uses the steel deck as a flat
steel plate with a thickness of 0.125, and E = 29,000x103 psi.
o Case 2A, 2B, 2C Steel Deck and Asphalt. These models use a
modulus of elasticity of the 2.25 x 0.125 corrugated steel deck,
combined with a 1 asphalt topping, with the moduli of the asphalt
transformed to that of the steel. The stiffness of the steel deck was
calculated by a finite element analysis, accounting for the effect of
the corrugations. The asphalt modulus was determined from
published sources (Park, S.W. and Lytton, R.L. 2004), and (Huang,
Y.H., 2004). A composite value of E = 6,231 xlO3 psi for the
asphalt is based on studies evaluating pavements subjected to a
principal stress; a value of E = 4,412 xlO3 psi and E = 1,212 xlO3
psi are recommended for typical pavements at 70F and 100F
respectively, assuming they are in service, but not subjected to
stress. The value that would be most appropriate for this structure
would require additional field studies. In lieu of this, the above
values bound the probable exact value.
o Case 3 SDI Steel Deck. This model used a modulus of E = 2,309
x 103 psi, which is calculated using Steel Deck Institute (Vulcraft
23


1993) for stiffness of corrugated metal deck. This stiffness takes
into account deflections due to shear, flexure, and warping of a
light-guage corrugated metal deck section. It should be noted that
these values are developed assuming low span to depth ratios, and
that shear deflection is the major contributor. There is no
contribution of stiffness assumed to be provided by the asphalt.
o Case 4 No Deck. This model used a modulus of E = 1 psi, ie. virtually
zero, to arrive at a lower bound, assuming the deck adds no stiffness to the
model. The calculated stiffness is slightly greater than that of the skeleton
model due to the presence of the 8 I beam stringers.
3.4 Diaphragm Model
The diaphragm model used a rigid plate member at the floor elevation. The
diaphragm model was used as an upper-bound for stiffness at the floor level. The
diaphragm feature is a standard element in most 3D programs, and attributes infinite
stiffness to a plate in an-plane direction, and represents the maximum possible
contribution to the deck as a stiffening element.
24


I
I
I
I
I
Figure 3.8. Representation of the diaphragm feature. The rendering was
produced by RAM Advanse.
3.5 Verification Model
The Verification Model incorporates the Deck Model (E = 2,309 x 103 psi)
with modifications to reflect actual conditions of the real bridge. This stiffness model
was chosen because 1) the stiffness is calculated using the Steel Deck Institute
methodology, which is a well-recognized authority on steel deck design, 2) results in
Section 5 indicate the asphalt could add from 20% to 37% addition stiffness; the
asphalt on the Rifle bridge is in poor shape, and may add minimal stiffness as a result.
The Verification Model is used in the field wind load studies to justify the
assumptions that additional elements exist to strengthen and stiffen the bridge beyond
those which the skeleton model assumes. Hence, it is one of the bases for this study.
These modifications are listed and discussed below.
I
25


o Member to member fixity. It was presented in previous studies (Rutz et.al
2005) that due to the magnitude of loading, ie: loads not sufficient to
overcome fixity due to friction, that all joints act as fixed. The model
assumes all connections as fixed, except for the ends of pinned members.
o Boundary Conditions. This study assumes bearings on one side of the
bridge (upwind) to be partially fixed with a spring constant of 150 k/in. in
a longitudinal translational direction, while the other side (downwind) to
be free. This assumption resulted from the review of analysis output, and
will be discussed in Section 5.1. Additionally, all bearings are assumed
fixed for rotation, due to the relatively low level of wind load compared to
the design wind load.
o Floor X-bracing. The Rifle Bridge was observed to have one x-brace
disconnected in the floor near a midspan bay. This brace was deleted in
the model. The brace would normally resist tension from an easterly
wind. The wind on this bridge during the study period blew
predominately from the west; therefore, the absence of the brace in the
model for this study has little or no effect on the results.
26


4. Wind Load Analysis
4.1 Historic Wind Loads
The late 19th century and early 20th century designer selected design wind
loads on a case-by-case basis, which varied typically from 1.44 kPa to 2.39 kPa (30
psf to 50 psf) applied to the projected area of the components (Smith 1881, Waddell
1898, and Cooper 1905).
4.2 AASHTO Wind & Live Loads
Todays typical analysis for truss bridges used for pedestrian crossings would
be based on The AASHTO Guide Specifications for the Design of Pedestrian Bridges
(AASHTO August 1997), which prescribes the wind load value, and equates to a
pressure of 3.59 kPa (75 psf) applied to the projected vertical area of all
superstructure elements, including exposed truss members on the leeward truss.
This value is almost twice of that probably used for the original design.
The prescribed live load according to the AASHTO guide is 65 psf for this
span length, which is used in a general analysis for this bridge to discuss future use as
a pedestrian crossing.
4.3 Verification Loads
The verification loads, or rather the procedure, are the basis of this study.
Much time and effort was spent to erect monitoring equipment, record wind events
twice a week for a 6 week period, then disassembling the equipment. Additionally,
each piece of equipment was either designed, or specifically chosen, to perform its
27


intended task. Indeed, this was the fifth bridge of an overall study (Rutz et al 2005),
of which this methodology was applied.
4.3.1 Dead Load
The dead load of the Rifle bridge is composed of the weight of the steel
skeleton, the metal deck and asphalt, and appurtenances, which include the railing,
walkway, and waterline. The total weight of the bridge is calculated to be 302.7
kips, of which the metal deck and asphalt contribute almost half, or 144 kips. Most
bridges of this era were constructed with a wood deck, which at a 4 thickness would
result in a dead load of 53 kips. This difference, approximately 91 kips, would result
in a net compression in the bottom chords at midspan, which would result in an
overstress condition.
Figure 4.1. Representation of superimposed gravity loads (DL + LL).
28


Figure 4.2. Relative axial forces in the bottom chord eyebars due to gravity
loads for the skeleton structure.
4.3.2 Field Measured Wind Load
The wind loads used in this study were derived from field measurements taken
over a six week period. The measurements were obtained in the following manner:
o seven anemometers that recorded wind speed were installed at locations that
would best represent each area of the bridge, Figure 4.3.
o strain transducers were installed on bottom chords at midspan, and at portal
members at each end of the bridge, which are locations of maximum stress,
o a Campbell Scientific model CR5000 data logger, mounted to the bridge in a
secure box, logged data as it was received from the anemometers and strain
gauges.
29


Figure 4.3. Diagram illustrating the locations of anemometers (WS1-WS7) and
wind direction sensor. North is to the left. WS1 was positioned directly
upwind of the centroid of the wind intercept area. WS2 and WS7 were located
approximately 3 meters above the top of the second diagonal frame from the
end, near the portals. WS3 and WS4 were positioned 2.5 meters below the
bridge deck, at an elevation mid-height between the bridge deck and the water
surface below.
Figure 4.4. Installation of the anemometers. They were mounted on the
predominately upwind side of the bridge, approximately 3 meters out from the
plane of the truss.
30


Figure 4.5. Photo of installed anemometers at the north end of the bridge.
One below the deck and one 2.5 meters above the deck can be seen, with the
one above the top chord just out of the picture.
Figure 4.6. Diagram illustrating the locations of the strain transducers. North
is to the left. The wind direction was from the west, orthogonal to the bridge.
Strain transducer numbers G1-G4 were clamped to the leeward bottom chord
eyebars. G5-G8 were clamped to the windward bottom chord eyebars. G9,
G20, G11, and G12 were clamped to the end diagonals at the south portal.
G13, G18, G15, and G16 were clamped to the end diagonals at the north portal
Although G18 replaced G14, and G20 replaced G10, the nomenclature of G10
and G14 were used in subsequent graphs for consistency.
31


I
Figure 4.7. Strain Transducer. It consists of a 3 diameter steel ring, bolted to steel
angles. The strain gage is on the inside ring surface, 90 degrees from the axis of the
bolts. This was developed by a graduate student at the University of Colorado
specifically for this multi-bridge study.
Figure 4.8. Strain Transducer mounted on the top (diagonal) chord
immediately below the portal truss.
32


Figure 4.9. Strain Transducers mounted on the bottom chord eyebars.
Note the reflective insulation wrap which protected the transducers
from direct sunlight, and ensured temperatures of the transducers
remained ambient.
The wind events recorded are summarized in Appendix C. The software,
developed by CU graduate student Veronica Jacobs (Jacobson 2006), sorted the
events by velocity in increments of 10 mph. When the velocity recorded by WS1
reached 20 mph, the logger began recording the event, with the event data including
the previous one minute and subsequent three minutes of wind velocity. If the
velocity increased to 30 mph, a new event was initiated and recorded, which may
include information recorded with the previous event. The software saved events to
the appropriate file named STRAIN 20, STRAIN 30, and so on to STRAIN 60. In
this manner, this researcher was able to view the events with the greatest velocities,
which tend to provide the maximum measured strains. Since there were over 100
events recorded over the study period, significant time was saved.
33


Wind events were reviewed, with the primary focus on a maximum velocity,
coupled with a significant change in strain. In a perfect scenario, the wind would be
exactly transverse to the bridge. Almost every measured wind event originated from
approximately 30 degrees to the northwest. No significant recorded events originated
from a direction 90 degrees to the bridge, which resulted in an analysis that included a
longitudinal component. This was accomplished by applying the wind load both
transverse and perpendicular to the bridge, multiplied by the cosine and sine
respectively of the average angle of attack of the event.
Three events were chosen for this study. Figure xx portrays the time period
that the most promising events were recorded, which occurred within a 17.5 minute
period on July 3, 2005, at 2 pm. Events 1 & 2 increased in velocity, and Event 3
decreased in velocity. Wind directions for the events were 298, 305, and 308, with
270 being transverse to the bridge.
34


WIND SPEED
Rifle Bridge 7/03/05
Figure 4.10. Portrays the 1047 second (17.5 minute) period producing the best
data. Events beginning at times 39.0 seconds (Event 1), 623.9 seconds (Event
2), and 717.3 seconds (Event 3) produced measureable strains in all members
instrumented, and were therefore used in this study. Events 1 & 2 increase in
velocity, while Event 3 decreases in velocity.
BOTTOM CHORDS: Change in strain
Rifle Bridge 7/03/05
(/)
WS avg Delta Leeward Strain filtered Delta Windward Strain- filtered
Figure 4.11. Graph of Event 1. Portrays Average Wind Speed to an arbitrary scale as
the top graph line, and Change in Strain (ue) in the bottom chords in the two bottom
graph lines. The event occurs between the two arrows as shown.
35


BOTTOM CHORDS: Change in strain
Rifle Bridge 7/03/05
Figure 4.12. Graph of Event 2. Portrays Average Wind Speed to an arbitrary scale as
the top graph line, and Change in Strain (ue) in the bottom chords in the two bottom
graph lines. The event occurs between the two arrows as shown.
50
40
r 30
_3
£ 20
ra
W 10
0
-10
120 130 140 150 160 170 180
Time (seconds)
WS avg Delta Leeward Strain filtered Delta Windward Strain-filtered
Figure 4.13. Graph of Event 3. Portrays Average Wind Speed to an arbitrary scale as
the top graph line, and Change in Strain (ue) in the bottom chords in the two bottom
graph lines. The event occurs between the two arrows as shown.
BOTTOM CHORDS: Change in strain
Rifle Bridge 7/03/05
Wind speeds, as recorded by the data logger, were transformed to pressures by
a formula similar to the standard formula currently used by ASCE 7 Natural
Phenomena Loadings.... This study used the formula p=Cd*Ca*0.00256 V2,
36


where:
p = stagnation pressure (psf)
V = velocity (mph)
Cd = 2, Drag Coefficient for structural shapes
Ca = 0.85 ambient air density correction for the site elevation of 5280 ft MSL
The pressures were then applied to the model as shown in Figure 4.14. Each wind
event exhibited different speeds from each of the anemometers. Since each reading is
averaged with those obtained from adjacent sensors, high and low readings are
balanced. This bridge, due to its long span, could arguably have used more sensors to
gain additional accuracy. However, in review of the events studied, each varied
where the maximum speeds were measured, ie: maximum speeds in one event were
measured at the lowest sensor, while another measured it at the highest point.
Because of this, this researcher believes the events chosen portray a reasonably
accurate model of the actual wind loads.
37


Figure 4.14. Quadrants subjected to different uniformly distributed wind pressures.
Wind pressure on Quadrant 1 was determined from a weighted average from the
velocities measured at WS7, WS5, and WS1. Wind pressure at the other quadrants
were similarly determined.
Pressures used in analysis were averaged from adjacent quadrants. Wind
velocities varied from 1.5 m/s to 14.9 m/s at the different anemometer locations;
although it is the difference in the maximum and minimum velocity at a location
for the period of time studied that was used to establish pressure on the model.
38


Table 4.1. Wind Velocities, Quadrant Average Velocities, and Quadrant Pressures
for Event 1.
Anemometer Location Velocity m/s (mph) Average velocity for quadrant m/s (mph) Average pressure for quadrant Pa (psf)
WS1 Central 9.0 (20.1) n/a n/a
WS2 South 5.2 5.1 123
upper (11.6) (11.3) (2.56)
WS3 South 1.7 3.9 84
lower (3.8) (8.7) (1.76)
WS4 North -0.8 3.8 94
lower (-1.7) (8.4) (1.97)
WS5 North 3.1 Not used in Not used in
central (6.9) any quadrant any quadrant
WS6 South 1.0 Not used in Not used in
central (2.2) any quadrant any quadrant
WS7 North 1.2 4.4 114
upper (2.7) (9.9) (2.38)
39


Table 4.2. Wind Velocities, Quadrant Average Velocities, and Quadrant Pressures
for Event 2.
Anemometer Location Velocity m/s (mph) Average velocity for quadrant m/s (mph) Average pressure for quadrant Pa (psf)
WS1 Central 7.9 (17.6) n/a n/a
WS2 South 6.7 6.4 253
upper (14.9) (14.2) (5.29)
WS3 South 8.3 6.9 240
lower (18.6) (15.4) (5.01)
WS4 North 0.5 3.4 124
lower (1.1) (7.7) (2.58)
WS5 North 1.9 Not used in Not used in
central (4.2) any quadrant any quadrant
WS6 South 4.5 Not used in Not used in
central (10.1) any quadrant any quadrant
WS7 North -0.6 3.1 112
upper (-1.4) (6.9) (2.34)
40


Table 4.3. Wind Velocities, Quadrant Average Velocities, and Quadrant Pressures
for Event 3.
Anemometer Location Velocity m/s (mph) Average velocity for quadrant m/s (mph) Average pressure for quadrant Pa (psf)
WS1 Central 5.9 (13.1) n/a n/a
WS2 South 2.4 3.5 129
upper (5.4) (7.9) (2.70)
WS3 South 14.9 7.7 249
lower (33.3) (17.2) (5.19)
WS4 North 1.5 4.4 144
lower (3.3) (9.8) (3.01)
WS5 North 5.8 Not used in Not used in
central (13) any quadrant any quadrant
WS6 South 2.3 Not used in Not used in
central (5.2) any quadrant any quadrant
WS7 North 5.6 5.8 212
upper (12.5) (12.9) (4.44)
41


Figure 4.15. Wind pressure applied to the four quadrants for
verification analysis. North is to the left. Another component of wind
pressure from the north (not shown here) was also applied to simulate
the measured wind direction from the northwest.
Figure 4.16. Graphical representation of axial forces in the bottom
chord eyebars due to wind for the skeleton structure. Note the reversal
in sign for the bottom chords near the pinned ends.
42


5. Results
5.1 Verification Load Results
The verification loads produced by Wind Events 1,2, and 3, were applied to
the verification model. Results are presented below in a tabular format, Tables 5.1
thru 5.3, summarizing calculated force or moment for each of the member types
(bottom chords, portal frames), compared with measured forces/moments, and
percent difference.
Table 5.1. EVENT 1. Compares the calculated forces or moments to those measured
in the field.
Calculated Force/Moment Measured Force/Moment Correlation
Member kN (kips)/m-kN (foot kips) kN (kips)/m-kN (foot kips) % difference

Windward bottom chord -5.96 kN (-1.34 k) -6.36 kN (-1.43 k) 7%
Leeward bottom chord 11.12 kN (2.50 k) 14.41 kN (3.24 k) 30%
North portal upper 3.20 m-kN (2.36 ft k) 1.15 m-kN (0.85 ft k) 64%
North portal lower 6.59 m-kN (4.86 ft k) 1.94 m-kN (1.43 ft k) 71%
South portal upper 3.47m-kN (2.56 ft k) 1.72 m-kN (1.27 ft k) 50%
South portal lower 7.04 m-kN (5.19 ft k) 1.65 m-kN (1.22 ft k) 76%
43


Table 5.2. EVENT 2. Compares the calculated forces or moments to those measured
in the field.
Calculated Force/Moment Measured Force/Moment Correlation
Member kN (kips)/m-kN (foot kips) kN (kips)/m-kN (foot kips) % difference

Windward bottom chord -11.48 kN (-2.58 k) -11.74 kN (-2.64 k) 2%
Leeward bottom chord 19.3 kN (4.34 k) 14.54 kN (3.27 k) 25%
North portal upper 4.42 m-kN (3.26 ft k) 1.41 m-kN (1.04 ft k) 68%
North portal lower 8.96 m-kN (6.61 ft k) 2.74 m-kN (2.02 ft k) 69%
South portal upper 6.32 m-kN (4.66 ft k) 3.43 m-kN (2.53 ft k) 46%
South portal lower 13.13 m-kN (9.68 ft k) 3.36 m-kN (2.48 ft k) 74%
Table 5.3. EVENT 3. Compares the calculated forces or moments to those measured
in the field.
Calculated Force/Moment Measured Force/Moment Correlation
Member kN (kips)/m-kN (foot kips) kN (kips)/m-kN (foot kips) % difference

Windward bottom chord -10.01 kN (-2.25 k) -8.41 kN (-1.89 k) 16%
Leeward bottom chord 16.5 kN (3.71 k) 16.81 kN (3.78 k) 2%
North portal upper 4.15 m-kN (3.06 ft k) 2.24 m-kN (1.65 ft k) 46%
North portal lower 8.46 m-kN (6.24 ft k) 3.97 m-kN (2.93 ft k) 53%
South portal upper 4.18 m-kN (3.08 ft k) 2.13 m-kN (1.57 ft k) 49%
South portal lower 8.92 m-kN (6.58 ft k) 5.15 m-kN (3.80 ft k) 42%
Note that percent correlation for the bottom chords of 7% and 30% for Event
1, 2% and 25% for Event 2, and 16% and 2% for Event 3 would indicate close
correlation. Overall, the correlation for Event 3 is the best, and coincidentally this
event had the largest magnitude of velocity, and the highest applied wind load force.
44


Events 1 & 2 were increasing in velocity, and Event 3 was decreasing in velocity;
correlations indicate this had negligible effect.
Also, the leeward bottom chord was measured to have almost twice the force
of the windward bottom chord. As discussed in Section 3.5, the boundary conditions
of the verification model were input to provide partial fixity to the windward support
for translation, while free at the leeward support. This correlates well with the
measured data, and the fact that the bearings could have been damaged as a result of
settlement as discussed in Section 2.2.
Correlations for the portals were not as close, with percentages ranging from
46% to 76%. Wind loads were applied to the top chord, vertical and diagonal web
members assuming they are solid, since these members are connected with plate
lacing. Additionally, the Drag Coefficient for structural shapes Cd = 2 was used. The
fact that the calculated moments are approximately twice the measured results may
indicate an analytically conservative approach, as there may actually be less loaded
area due to the open construction of the built up members.
5.2 Deck Model Stiffness Comparison
As presented in Section 4.2, the 1997 AASHTO Guide Specifications for the
Design of Pedestrian Bridges would dictate the design live load to be 65 psf,
reduced from 85 psf due to span length, and the design wind load 75 psf. This guide
also specifies the design load combinations to be used shall be in conformance with
the AASHTO Standard Specifications for Highway Bridges, August 1997, Table
45


3.22.1 A. This Table specifies the following combinations be used for dead load
(DL), live load (LL), and wind load (WL) in the combinations listed as follows:
o DL + LL
o 0.8DL + 0.8WL
Note that DL, LL and WL are not combined, as this section of the code specifies
Wind on Live Load equals zero.
This study uses the dead load plus wind load combinations to compare the
deck models developed. Only one member is used: the bottom chord eyebar on the
windward side. The reasoning is that this member is the simplest to study, since it is
typically stressed in compression (for wind loading), is analyzed at midspan (which
is the location of maximum stress), is composed of a simple rectangular section (not
built up of a combination of members), and can become a critical member if
loadings cause stress to reverse from tension to compression. Shown below is Table
5.4 with the calculated stresses for each of the models, and percent change in force
for each of the basic elements from the basic skeleton model. The percent change
reflects how much additional lateral load is resisted by the deck in each of the cases.
As presented above, the stiffness increases for each of the deck types studied, and is
upper-bounded by the Diaphragm Model.
46


Table 5.4. Comparison of the analysis results from the three models studied:
Skeleton, Deck, and Diaphragm. Several types of deck, with their associated
stiffnesses, are shown for the Deck model. Forces are for the windward bottom
chord eyebar, followed by percent reduction in compression (or increase in tension)
compared to the traditional skeleton value. (Positive = tension; negative =
compression). Results are listed in increasing percent of change or stiffness
compared to the skeleton model.
Model Axial Load, kN (kips) Load / Load Combination
DL WL 0.8DL + 0.8WL % change
Skeleton Model 649 kN (146 k) -447 kN (-101 k) 190 kN (43 k) N/A
Deck Models Case 4 No Deck E=1 psi 649 kN (146 k) -438 kN (-99 k) 168 kN (38 k) 4
Case 2C Steel Deck+ Asphalt E=1,212 x 103 psi 648 kN (146 k) -406 kN (-91 k) 194 kN (44 k) 20
Case 3 SDI Steel Deck E=2,309 x 103 psi 649 kN (146 k) -398 kN (-89 k) 201 kN (45 k) 24
Case 2B Steel Deck + Asphalt E=4,412 x 103 psi 650 kN (146 k) -384 kN (-86 k) 213 kN (48 k) 32
Case 2A Steel Deck + Asphalt E=6,231 x 103 psi 650 kN (146 k) -374 kN (-84 k) 221 kN (50 k) 37
Case 1 Steel Plate E=29,000 x 103 psi 655 kN (147 k) -320 kN (-72 k) 268 kN (60 k) 66
Diaphragm Model 649 kN (146 k) -292 kN (-66 k) 305 kN (69 k) 77
The percent change from the skeleton case was determined for the model
from:
% change =100 x
Fskeleton Fdeck
Fskeleton
(2.1)
47


and for the diaphragm model from:
% change = 100 x
Fskeleton ~ Fdiaphragm
Fskeleton
(2.2)
where:
Fskeleton = calculated force in windward bottom chord from the skeleton model
Fdeck= calculated force in windward bottom chord from the deck model
F'diaphragm = calculated force in windward bottom chord from the diaphragm model
The results indicate modifications in the stiffness of the deck models can
range from 4% to 66%. Typical metal decks are fabricated in an accordion fashion,
allowing the greatest stiffness to weight ratio, and would not be as stiff as the steel
plate model. Additionally, the asphalt has been included in three of the cases, and
does calculate to add stiffness; however as the deck becomes worn, as at the Rifle
bridge, may not be depended upon for long term contributions. This study uses Case
3 SDI Steel Deck as a reasonable approach to deck stiffness. The modulus is based
on the Steel Deck Institutes recommendation for properties. It adds 24% to the
overall stiffness of the system, thereby reducing loads to the bottom chords. This
inclusion does require proper connection to the supporting elements, which should be
verified during the analysis. The Rifle bridge deck was observed to be welded to
supporting beams.
48


5.3 AASHTO Code Conformance
As discussed in Section 5.2, the Deck Model using the stiffness of a
corrugated metal deck (Case 3) provides reasonable results in comparison with the
field wind load study, and is therefore used to analyze the bridge. The code required
loads are therefore applied to this model, to check compliance with the current code.
In applying the loads, stresses are calculated for the members under study, and are
shown below along with the percent of yield stress, assuming Fy = 30 ksi., as noted
in Section 2.3. As current design code dictates, allowable stresses would be less
than yield, as allowance for type of stress, and local/global stability must be
evaluated for each member.
Table 5.5. Indicates the stress ratio in percent of material yield stress, of certain
members of the Rifle bridge using load combinations of the 1997 AASHTO code for
pedestrian bridges, applied to the Case 3 Deck Model. Wind load is applied
transverse to the bridge.
DL + L 0.8 DL + 0.8 WL
Member Stress Stress Ratio Stress Stress Ratio
MPa (ksi) % MPa (ksi) %

Windward bottom chord 198 MPa (28.7 ksi) 96 34 MPa (4.9 ksi) 16
Leeward bottom chord 153 MPa (22.2 ksi) 74 118 MPa (17.1 ksi) 57
North portal 447 MPa (64.8 ksi) 216 249 MPa (36.1 ksi) 120
South portal 354 MPa (51.3 ksi) 171 570 MPa (82.7 ksi) 276
Additionally, the code requires deflection due to the design live load and wind
load, applied separately, not exceed the ratio of the span divided by 500 (L/500).
49


Analyses calculate these values to be 8.49 inches vertical deflection, or L/339 for
live load, and 24.3 inches lateral drift, or L/l 19 for wind load, which do not meet
code required limitations for serviceability.
In summary, the members included in this study, which would generally
exhibit the maximum stresses, would not meet current code requirements.
Therefore, other means would need to be instituted to use of this bridge under the
current AASHTO code. Methods could include reduction of dead load, reduction in
the accessible (and therefore loaded) area, and reinforcement of members at critical
locations.
50


6. Conclusions & Recommendations
6.1 Conclusions
The Verification Model provides reasonable correlation with field data,
particularly the bottom chords. Inclusion of boundary conditions with partial fixity
accurately reflects measured results. Wind loads used for the analysis varied in
magnitude, direction, and application (increasing or decreasing in velocity), which
eliminated anomalies in wind events chosen. The deck does calculate to assist in
resisting lateral loads, and its inclusion in the overall analysis is justified.
Calculated stresses at portals did not correlate well with those measured.
Anemometers were distributed across the bridge at three locations horizontally,
which resulted in a spacing of 40 ft to 80 ft on center. Since the Rifle bridge is 240
ft long, additional measuring devices may have provided better information, and
therefore better correlation, for the members located at the upper reaches of the
bridge.
Reasonable assumptions of deck stiffness, within a range of magnitude of E =
6,231 xlO3 psi and E = 1,212 xlO3 psi, indicate the difference in calculated bottom
chord force is 32 kN (7 kips), a difference of 8% of the total force. This would
suggest that any reasonable assumption of deck stiffness, whether it is based on the
steel deck alone, or acting compositely with an asphalt topping, will not have major
effects on the final force. This analysis indicates the forces in the bottom chords are
51


not sensitive to changes in deck stiffness within the ranges discussed here, but are
sensitive to deck vs. no deck (Case 4).
Results support the inclusion of the bridge deck in modeling. This analysis
would indicate the deck contributes between 25% to 40% to the overall stiffness of
the system, when comparing a reasonable deck model to that of no deck at all.
Inclusion of the asphalt topping, while arguably helps, is not an important
contributor to stiffness. Skeleton models alone do not account for the stiffnesses
measured in actual models. The type of deck assumed, connections to supporting
members, and connections between deck members all play a part in the stiffness of
the model to transfer forces to supports, and are left to the designer to estimate;
however provided the deck is included, the model is not overly sensitive to the
range of deck stiffness values that are typically assumed.
The Rifle Bridge does not calculate to meet current AASHTO code
requirements for pedestrian bridges. Efforts, in addition to the maintenance and
repair currently needed, would be required to allow for pedestrian use.
6.2 Recommendations for Future Research
Analysis of Boundary Conditions most designers will assume bearings to be
pinned or free. Fixity does develop, particularly in older bearings.
Investigations as to the range of loads and stresses required to overcome fixity
would be beneficial.
Investigation of Member Area for Wind Loads the historic truss bridges are
typically comprised of built-up members that have open area. Testing to
52


establish actual loads on these members may prove to allow reduced loads in
wind calculations.
This work stems from earlier work completed by Carrol (2003), Herrero
(2003) Rutz (2004), and Rutz and Rens (2004). For additional information about the
Rifle Bridge, refer to Rutz et al (2005).
53


APPENDIX A
Field Measurements
I
Historic Truss Bridge Research
Bridge Across the Colorado River at Rifle, CO
6/5/05, 6/8/05
Longitudinal (taken at west truss)
dim (x.xx ft) member descr. lacing
panel pt to panel pt
(bottom chord) 1 to 3 20.00 N/A
3 to 5 20.02 N/A
5 to 7 20.00 N/A
7 to 9 20.00 N/A
9 to 11 20.04 N/A
11 to 13 20.00 N/A
13 to 15 20.00 (2)4x1.1875 N/A
15 to 17 20.00 (2)4x1.1875 N/A
17 to 19 20.04 (2) 4x1.00 N/A
19 to 21 20.00 (2) 4x1.00 N/A
21 to 23 20.00 (2)3x1.125 N/A
23 to 24 20.02 (2)3x1.125 N/A
(top chord) 1 to 2 #4 at bottom
2 to 4 #4 at bottom
4 to 6 #4 at bottom
6 to 8 #4 at bottom
18 to 20 #4 at bottom
20 to 22 #4 at bottom
22 to 24 #4 at bottom
(web members vert) 2 to 3 #3 inside & outside face
4 to 5 #2 inside & outside face
6 to 7 #3 inside & outside face
8 to 9 #2 inside & outside face
10 to 11 #3 inside & outside face
12 to 13 #2 inside & outside face
14 to 15 34.29 #3 inside & outside face
16 to 17 34.17 #2 inside & outside face
18 to 19 30.92 #3 inside & outside face
20 to 21 27.81 #2 inside & outside face
22 to 23 20.04 #3 inside & outside face
54


Members dimension (in.)
number/type #2 b 10.00
d 6.063
bf 1.938
tf
tw
quantity 2
lacing 1.5x0.1875
rivet spac 10" vert
8.25" horiz
#3 b 9.625
d 5.00
bf 1.8125
tf
tw 0.219
quantity 2
lacing 1.25x0.219
rivet spac 10" vert
7.25" horiz
# 4/top chord
PL top C sides 14x0.27
b 13.375
d 10.875
bf 2.875
tf 0.27
tw
quantity 2
lacing 1.5x0.219 at bot
rivet spac 12" at bot
6" at top
floor beam (S shape) d 18.00 A (calc) = 17.13
bf 6.00 VWft(calc) = 58.3
tf 0.547
tw 0.625
stringers (S shape) d 8.125 A (calc) = 3.98
bf 3.9275 Wt/ft(calc) = 13.5
tf 0.328
tw 0.188
55


APPENDIX B
COLORADO HISTORICAL SOCETY
Offer of Archaeology and Hmorie Preservation
1300 Broadway Denver. Colorado 80203
a t
Q E G m ¥
T
V
Site No. SGFWSt
467
county

Office ui only
Garfield
ttructire name
Hi fie
COOT No
GAR019
Rifle Bridge
highway location
mfepoat
highway carried ABD. COUNTY HOAD
feature htercected COLORADO RIVER
dractions IN RIFLE
eigUe for National Register ______________yes _______no
date______________________ initials________________________
Criteria _A __B _C _D
contrfcutes to a potential Notional Rerpster district
___yos __no dstrictname
efcgble for State Regster _____yes _____no
date___________________nitiais_______________________
Criteria _____a ____b c d . e
areas of siytifcance
period of significance .
needs data______ dale.
consultant's evaluation
City of Rifle
dmensorw
National Register etgblrty listed
Criterion A ] Criterion C S
Matters gterc ter Criterion A rater Id hotoric centrxb y**n te ite
report L*tmt lot both Cnter* rater to *S* tigafaant and Tf norvagSBon
structure location
townshp 6S range 93W section NE1/4 S16
1JTM tone 13 easing 2609? northng 4378880
USGS quad Rifle
hrstorcal nfor motion
meet span number 2
appr span number 0
nan span length 240.00
structural reformation
stiuctiae tirngth 434.00
structiae width 19.00
roadway vndth 17.30
supersTuclire
substructure
ftoor/deckng
ether feetses
bcation map
steel, pin-connectod Pennsylvania and Parker through trusses
concrete abutments, wingwalls and pier
corrugated steel deck with asphalt overlay
upper chord: 2 channels with cover plate and ladng; lower
chord: 2 rectangular eyebars; vertical: 2 channels with lacing:
diagonal: 2 rectangular eyebars; counter: 1-2 square eyebars
with turnbackles; floor beam / stringer: I-beam; lattice guards
erected 1909
designer Garfield County Surveyor
fabreator Minneapolis-Moline Power
Implement Company
contiactor Charles G. Sheely Construction Co.
hfo scarce Garfield County records
testorc/presant use
histone u roadway bridge
present use pedestrian bndge
xedent good j far i x r
location
origin! j x moved j
altar aborts
type bridge closed 1o vehicular traffic
datefs)
PRASERDESIO
56


APPENDIX C
Historic Truss Bridge Research Bridge Across the Colorado River at Rifle, CO Data Collection
File Name Date Collected 20 mph 30 mph 40 mph 50 mph 60 mph Remarks
STRAIN 01 6/20/2005 test
STRAIN 02 6/20/2005 test
STRAIN 03 6/20/2005 test
STRAIN 04 6/29/2005 25 events 6/19 event 1 6/20 event 2 thru 5 6/21 event 6 thru 9 6/22 event 10 thru 19 6/23 event 20 thru 23 6/25 event 24 thru 25
STRAIN 05 6/29/2005 24 events
STRAIN 06 6/29/2005 7 events
STRAIN 07 7/9/2005 25 events event on July 3 premier event
STRAIN 08 7/9/2005 25 events
STRAIN 09 7/9/2005 13 events
STRAIN 10 7/9/2005 6 event s 6/20 9:53 no results 6/28 16:59 ok 7/3 14:08 ok 7/3 14:14 ok
57


STRAIN
11
7/9/2005
4
event
s
strain
12
strain
13
strain
14
7/17/2005
7'17/2005
7'17/2005
25
events
14
events
strain
15
strain
16
7/24/2005
7/24/2005
25
events
11
events
2
events
6/20 9.49
6/20 9.53
7/3 14-J9
7/3 14 23
no results
no results
ok
ok
strain
17
7/24/2005
1


REFERENCES
AASHTO (1997). Guide Specifications for Design of Pedestrian Bridges, American
Association of State Highway and Transportation Officials, Washington, D.C.
American Institute of Steel Construction (2005) Steel Design Manual
Carroll, D. (2003). Analysis of historic pin connected through truss bridges for
conversion into pedestrian use, Dept, of Civil Engineering, Univ. of Colorado at
Denver, Denver, CO.
r rraser uesign,
Inventoried by
What Wind Pressure Should be Assumed in the Design of Long
mg News, 5 Jan., 15-16.
Cooper, T. (1905). What Wind Pressi
Bridge Spans?, Engineering News,
Glenwood Post (1929). County Bridge at Rifle in Danger, Glenwood Post, May 23,
1929
Google Earth (2007). Satellite Photo, Digital Globe @ 2007 Europa Technologies
Herrero, T. (2003). Development of strain transducer prototype for use in field
determination of bridge truss member forces, Dept, of Civil Engineering, Univ. o
Colorado at Denver, Denver.
, vh 17004) Pavement Analysis and Design, Second Edition, Pearson
- November 0981), Verco Steel Decks, PO Box 14667,
Huang,
Education, Inc.
1CB0 Report No. 2078, November (1981), Verco,
Phoenix, Arizona, 85603
ideal Techniques and Field Verification Method foi
r ad of Prowers Bridge, July 25,2006
, Denve
IX, /tr *''-
,onn6\ Analytical Techniques and tieia r c,.
Jacobson, V. ^ j^|ySjs of the Histr Prowers Bridge, July 25,2006
fd ^ sign Lecture and Worbhop Notes,
n nendent Modulus at
V\ Effect of Mess fPe nts, Jo
ry
*