NCHRP12-71_FR | Prestressed Concrete | Precast Concrete

NCHRP

Project Number
12-71
Design Specifications and Commentary
for Horizontally Curved
Concrete Box Girder Highway Bridges
Final Report
January 29, 2008
NRV
Nutt, Redfield & Valentine
A Professional Partnership
________________________________________________________________
in association with
David Evans & Associates, Inc. Zocon Consulting Engineers, Inc.
Page i
NCHRP Project 12-71
Design Specifications and Commentary for Horizontally Curved Concrete Box-Girder
Highway Bridges

Final Report
TABLE OF CONTENTS

1. INTRODUCTION 1
2. STATE-OF-PRACTICE REVIEW 3
a. Domestic Practice 3
b. Foreign Practice 6
c. Field Problems 8
3. PUBLISHED LITERATURE REVIEW 11
a. Codes and Design Standards 11
b. Design Philosophy 19
c. Response of Curved Concrete Box Girder Bridges 22
i. Global Analysis 22
ii. Laboratory Experiments 31
d. Design Issues 32
i. Bearings 32
ii. Diaphragms 33
iii. Flexure and Shear 33
iv. Torsion 37
v. Wheel Load Distribution 39
vi. Tendon Breakout and Deviation Saddles 39
vii. Time Dependency 45
viii. Vehicular Impact 45
ix. Seismic Response 46
x. Design Optimization 46
xi. Detailing 47
4. GLOBAL RESPONSE ANALYSIS STUDIES 49
a. Objective 49
b. Model Verification 49
c. Parameter Study 51
d. Special Studies 65
5. LOCAL RESPONSE ANALYSIS STUDIES 71
a. Local Analysis Validation 72
b. Local Analysis of Multi-cell Box Girders 88
c. Local Analysis of Single-cell Box Girders 111
d. Conclusions from Local Analysis 122
Page ii
6. CONCLUSIONS 129
7. REFERENCES 137

APPENDICES

A. PROPOSED LRFD DESIGN SPECIFICATIONS A-1
B. DESIGN EXAMPLES B-1
a. Example B-1 - Comprehensive Problem (Spine and Grillage) B-5
b. Example B-2 - Tendon Confinement B-121
c. Example B-3 - Tendon Confinement B-131
d. Example B-4 – Global plus Regional Combination (Menn) B-137
e. Example B-5 – Global plus Regional Combination (Poldony) B-147
f. Example B-6 - Deviation Saddle Design B-149
C. GLOBAL ANALYSIS GUIDELINES C-1
a. Three Dimensional Spine Beam Analysis C-1
b. Grillage Analogy Analysis C-4
D. STATE OF PRACTICE SUMMARY FOR UNITED STATES D-1
E. DETAILED GLOBAL ANALYSIS RESULTS E-1
a. Bridge Model Scattergrams – Spine vs. Grillage E-3
b. Bridge Model Ratio Plots – Spine vs. Grillage E27
c. Bridge Model Ratio Plots - Straight to Curve (spine and grillage) E39
d. Effect of Diaphragms – Grillage Analysis E51
e. Effect of Diaphragms – Ratio Plots E55
f. Effect of Hinges – Grillage Analysis E57
g. Effect of Hinges – Ratio Plots E61
h. Effect of Skews E63
F. DETAILED LOCAL RESPONSE ANALYSIS RESULTS F-1
a. Single Cell Box Girder F-1
b. Multi-Cell Box Girder F-63
Page 1
NCHRP 12-71
Design Specifications and Commentary for Horizontally Curved Concrete Box-Girder
Highway Bridges
Final Report

1. INTRODUCTION

At one time, bridges on curved alignments were rare. However, times have changed and modern
highway bridges and traffic separation structures are commonly built on a horizontal curve. This
has come about because of higher traffic volumes and speeds, geometric constraints of the urban
environment, and improved structural forms that lend themselves to curved construction.
The concrete box girder, particularly post-tensioned prestressed concrete that can span large
distances, is one such structural form. The cross section of these structures is inherently strong in
torsion. This is important, because curvature induces high torsion forces. Also, because concrete
can be easily molded into the required shape, it is ideal for curved construction. For these
reasons, prestressed concrete box girders have become the structure type of choice in many
jurisdictions. A common application of curved structures is in freeway interchanges where
connector ramps or “fly-overs” carry traffic from one freeway to another at relatively high speed.
Cross sections of curved box girders may consist of single cell, multi-cell or spread box beams as
shown in Figure 1-1. Because only a very few spread box beam bridges use curved beams, only
the first two types were considered in this study.

Single-cell Box Girder
Multi-cell Box Girder
Spread Box Beams

Figure 1-1
Types of Cross Sections
Page 2
It has become common practice to analyze and design these structures as if they were straight.
Live load distribution is often addressed using the whole-width design approach. Local
problems, such as the lateral forces induced by curved prestressing ducts, are often handled using
specific design rules and details that have been developed over the years. For the most part, this
design approach has been used successfully, but some recent cases of poor bridge performance
have made it clear that this approach has its limitations.
Because it is likely that the use of curved structures is going to increase, and that the geometries
of some of these structures will continue to push the envelope with respect to the degree of
curvature, span lengths and depths, the amount of required prestressing force, etc., it is evident
that better guidelines are required for their design. That is the purpose of this project.
There has been a significant body of research and development relative to the design and
analysis of curved prestressed concrete box girder bridges. Some of this information has found
its way into design specifications, while much of it has not. There is a need to collect and analyze
this information with the goal of evaluating its merit for nationwide design rules. While much of
this work has been conducted domestically, there is also a significant body of work conducted by
other countries that also needs to be considered.
Although most issues relative to design of curved concrete box girders have been studied to
some degree, there is a need to fill in the gaps in our understanding. With modern computer
programs and analytical models calibrated to existing physical and experimental results, it is
possible to do most of this through analytical studies. It is evident that additional physical testing
of existing structures or laboratory experimentation, while important in and of itself, are beyond
the scope of this project, and are not necessary to accomplish this project’s intended goals.
This report is intended to present the results of a review of the literature and the state-of-practice
with respect to curved concrete box girder bridges. In addition, detailed results from both global
and local response analysis studies are presented. Final recommendations are presented in the
final chapter and are implemented in the form of recommended changes to the AASHTO LRFD
Bridge Design Specifications and Commentary and in analysis guidelines for these types of
bridges. Example problems are also presented that illustrate the application of these
recommendations.
Page 3
2. STATE-OF-PRACTICE REVIEW

a. Domestic Practice
To obtain a better understanding of current US practice, telephone interviews were conducted
with representatives from key state Departments of Transportation (DOT’s). The states surveyed
included California, Colorado, Florida, Hawaii, Idaho, Oregon, Nevada, New York, Tennessee,
Texas, Washington and Wisconsin. Other states were contacted but elected not to participate.
Based on our interviews current US practice can be summarized as follows:
1. With the exception of California and Washington, most states interviewed have a fairly
small inventory of curved concrete box girder bridges (i.e. <1% each of reinforced and
prestressed) although many see the number increasing in the future.
2. Cast-in-place construction is most popular in the West. Many other states are tending
toward segmental construction (cantilever and span-by-span using both pre-cast and cast-
in-place concrete) in order to avoid conflict with traffic. Colorado has used pre-cast,
curved, spliced “U” girders with a cast-in-place deck. The use of pre-cast box-beams in
most other states is limited to straight girders. Curvature, when present, is provided by a
curved deck slab (i.e. variable overhangs). Northern and eastern states, where weather
conditions cause more rapid deck deterioration, tend to avoid prestressed boxes due to the
need to provide for future deck replacement.
3. Most curved box girders are relatively narrow continuous ramp structures. A few are
single span and a significant number of all structures (approx. 20% to 30%) have skewed
abutments. A small percentage of structures have skewed multi-column bents. Span
lengths are usually less than 160 feet but approximately 20% exceed this limit.
4. The trend for the future appears to be dictated by the requirements of a built-up urban
environment. More curved alignments, longer spans, more skewed supports, and more
segmental construction are expected. Curved precast girder systems may also increase,
particularly in Colorado where good success has been experienced with this type of
construction
5. Many states have experienced some problems with the performance of curved box girder
bridges, but not many as a percentage of the total. Cracking along the tendon and tendon
breakout problems are absent or minimal where sufficient space is provided between the
ducts. Torsion and flexural shear cracking seems to be rare and not necessarily limited to
curved bridges. A few bearing failures have occurred but have been avoided in states that
avoid bearings altogether or use conservative bearing designs. In some cases bearing
uplift at the abutments has been observed to occur over time and are thought to be due to
the time dependant behavior of concrete. Unexpected vertical or horizontal displacements
of the superstructure are rare, but California has had some problems on skewed multi-
column bents where movement about the c.g. of the column group has caused the
transverse shear keys to engage. Lateral displacement of columns has also been observed.
6. Some states have special design rules. A few of these are discussed below.
Many states either already use AASHTO LRFD (2004) or are in the process of adopting it.
Page 4
Most states have no special rules for when a 3 dimensional analysis such as a grillage or finite
element analysis should be performed and leave it to the discretion of the designer. Many states
use an 800 ft radius as the trigger where designers should consider 3-D analysis. Most states
have access to computer programs that are capable of such an analysis. Almost all states use
AASHTO wheel load distribution. California commonly uses the whole-width design approach.
No state had specific guidelines for varying the prestress force in the individual webs of curved
box girder bridges although at least two states recognize that stresses can vary transversely
across the section and encourage designers to take the initiative to specify varying prestress
force. It should be noted that the horizontal curvature of the tendons produce additional tension
on the girder toward the inside of the curve, thus mitigating the severity of stress distribution
across the section and the need to vary prestress force.
California recently experienced a tendon breakout failure on the 405/55 interchange (Seible, et.
al., 2003). Prior to that failure they had published guidelines for designers related to the design of
curved post tensioned bridges (Caltrans, 1996). These guidelines dealt with the need for special
detailing in curved webs including criteria for when these details are not needed. This memo
indicated that because of its relatively large radius, tendon ties were not required in the 405/55
structure. The problem resulted because of a separate Caltrans standard plan, not specifically
related to curved bridges, that allowed up to six tendons 4-1/2” or less in diameter to be stacked
on top of one another without any space in between. Because the 405/55 was a long span
structure there were several tendons that needed to be stacked resulting in radial forces being
applied over a relatively wide area of essentially unreinforced cover concrete. It is thought that
this was the primary cause of the failure. Caltrans indicates that they currently have no special
policy for prevention of tendon breakout failures except that designers are to provide tendon ties
under certain circumstances. Breakout failures have not occurred when these ties are present.
Some other states indicated that they used the Caltrans tendon tie details to prevent tendon
breakout failures.
Several states had requirements for providing space between tendons. Oregon requires that no
more than three 4” diameter or less ducts be stacked without a space of 1-1/2” between a
subsequent stack of ducts. The number of stacked ducts is reduced to two for duct diameters
exceeding 4”. It should be noted that the current AASHTO LRFD specifications have duct
spacing requirements that are similar to Oregon’s. Texas also indicated that they have similar
duct spacing requirements.
Colorado requires a duct spacing of 44% of the duct diameter or 1-1/2” minimum. This applies
to all ducts (i.e. no stacks). This is more conservative than AASHTO and most other states, but
Colorado reports no breakout failures due to web curvature.
It appears that duct ties and duct spacing requirements have been successful in preventing tendon
breakout failures. However, excessive duct spacing requirements can present problems at
midspan and over the bents in continuous concrete box girder superstructures due to the
reduction in prestressing eccentricity and the corresponding increase in prestress forces that
results. Because the action of the deck and soffit slabs tend to prevent breakout failures at points
of maximum tendon eccentricity in box girder structures, it is possible that spacing requirements
could be relaxed at these locations. Duct spacing requirements do not affect tendon eccentricities
where the ducts lie near the mid height of the webs. These are the location of most breakout
failures. These are also areas where actual duct curvatures may be amplified due to the horizontal
deviation of tendons to accommodate end anchorage systems. Therefore, it should be possible to
Page 5
refine guidelines regarding duct spacing to both facilitate prestressing economies and prevent
breakout failures.
Most states interviewed did not have specific guidelines for the design of bearings in curved box
girder bridges. Some states expressed a preference for certain types of bearings and others try to
avoid the use of bearings in curved box girder bridges.
Design for torsion in most states follows the AASHTO requirements. Colorado expressed a need
for better guidelines for combining shear and flexural stresses. Colorado also uses pre-cast “U”
girders, which are temporarily braced for torsion during the placement of the cast-in-place deck.
At least one state said they ignored torsion design, but this might be because they have only
designed large radius bridges.
One point of interest is the combination of global shear and regional transverse bending stresses
in the webs of curved box girder bridges. Caltrans, which uses mostly cast-in-place bridges
constructed on falsework, does not combine these stresses. The reasoning is that when the bridge
is stressed, and regional transverse bending stresses are first realized, the bridge is on falsework
and experiences no flexural or torsion shear stress. By the time falsework has been released, the
prestress force is reduced due to relaxation and is not as critical for regional transverse bending.
Other states have no specific guidelines and leave it to the designer to determine how these
stresses should be combined.
Several states have standard details for concrete box girder bridges. Most of these deal with
prestress duct patterns and web reinforcing. Some of these were discussed above.
The requirements for the number and spacing of interior diaphragms vary among the states. The
current AASHTO Standard Specifications for Highway Bridges (AASHTO, 1996) has specific
requirements for the number and spacing of interior diaphragms in concrete box girder bridges
and several states use these or similar requirements. Diaphragms are not required in curved
bridges with a radius of 800 feet or greater. For a radius between 400 and 800 feet the maximum
diaphragm spacing shall not exceed 80 feet, and when the radius falls below 400 feet the
maximum diaphragm spacing is 40 feet.
The AASHTO LRFD Bridge Design Specifications state that diaphragms are required in curved
concrete box girder bridges with a radius of 800 feet or less, but the code implies that their
number and spacing are to be determined by the designer and depends on the radius and the
dimensions of the cross section. Specific guidelines on how to determine the number and spacing
of interior diaphragms are not provided.
Colorado has standards for curved “U” girders with a cast in place deck that they have used with
good success. These bridges often consist of several pre-cast segments spliced together to form a
span. Colorado projects that they will use this form more often in the future. It has the advantage
of potentially eliminating falsework over traffic
A similar “U” section, although straight, was developed by NRV for use by a contractor on a
bridge originally designed as a cast-in-place multi-cell box. The original design required building
the superstructure on falsework set above final grade, lowering it to its final grade position and
then casting a monolithic bent cap to connect the bridge to the column. This approach overcame
falsework clearance problems but presented some challenges for the contractor. In the contractor
redesign, pre-cast prestressed “U” sections set at final grade spanned over the required falsework
opening. Cast-in-place bottom soffit and stem pours were made on either end on the “U” girders.
The cast-in-place pours were monolithic with the columns. A cast-in-place deck was then poured
Page 6
and continuity prestressing used to tie the entire structure together. This approach has some
advantages and, as in Colorado, could be used on curved structures.
The Oyster Point Off-ramp in California also had a span consisting of curved pre-cast girders
made continuous with a cast-in-place multi-cell box section. This span in question crossed over
railroad tracks where falsework was not allowed. As a cost savings measure, the contractor
elected to use curved pre-cast beams as opposed to straight beams with a curved cast-in-place
deck. Girder erection required pick points to be located so that the girders would not “roll” and a
temporary tie down system at the ends of the girders during the deck pour.
Despite these examples, curved box beams appear to be relatively rare.

b. Foreign Practice

Concrete box girder bridges are used around the world. Many of these are on horizontally curved
alignments. A partial survey of recently constructed curved concrete box girder bridges outside
the United States was conducted by reviewing material published in engineering magazines,
trade journals and from personal experience. Many of these bridges were built using segmental
construction techniques. While the survey was not exhaustive, its results were representative of
the bridges being built today.
Canada has been very active in developing design specifications that address the behavior of
concrete box girder bridges. The Ontario Highway Bridge Design Code (1983) was an early
attempt to codify design and analysis requirements for these types of bridges. Many of the
provisions developed in this early code have influenced the development of the current Canadian
Standards Association design specifications (2000) and the updated version of the Ontario Code
(1998). Although curved box girders are not specifically addressed in these codes, many of the
analysis techniques identified for box girders, such as the orthotropic plate analogy and the
grillage analogy methods may have some applicability to curved structures.
In Europe, most countries and most designers base their designs on the first principles of
structural analysis and design. Although design codes are used, they are generally brief and
bridge designers rely on traditional text books such as those by Menn (1990), Schlaich and
Scheef (1980) or Strasky (2001); specific course work published by their professors at their
university (Leonhardt, Menn, Strasky, etc.); or personal experience.
In Switzerland the structural code is split into separate booklets. One each for loadings, concrete
and prestressed concrete, steel, wood, etc. The two most applicable to the NCHRP 12-71 project
(Loadings, and Concrete and Prestressed Concrete) are relatively brief documents compared to
the current AASHTO LRFD. In general the Concrete and Prestressed Concrete booklet does not
or only briefly deals with special structural configurations such as horizontal curvature. Instead,
students at the two Universities, in Zurich and Lausanne, study structural design in a practical
manner, preparing them for the professional situation in their own country. The textbook
discussed above (Menn, 1990) is very similar to what students will encounter when at the
university. Menn, who was a Professor for many years, does provide some general guidance on
the design of horizontally curved beams and skewed bridges.
Page 7
Swiss bridges on a curved alignment with a large radius are often designed without considering
the curvature, except for the bearing design. Many design firms use methods they have
developed over the years involving graphs and influence surfaces.
U.S. engineers with experience designing bridges in France have been contacted. It is our current
understanding that the French favor pre-cast segmental construction. Typically these structures
utilize external tendons with deviator blocks. As in most European countries, their design
specifications are less prescriptive than the U.S. and designers rely on their own experience as
well as other published material to analyze and design these bridges.
Germany has been a leader in the design of concrete box girder bridges, and engineers like Fritz
Leonhardt have been considered pioneers for this type of construction. Germans tend to favor
cast-in-place construction. They have their own DIN code, but as is typical of most European
practice, they rely on the engineer to apply first principals in selecting analysis techniques and
design details.
Based on the published literature (Branco, et. al., 1984; Danesi, et. al., 1982 & 1983; Evans, et.
al., 1975; Goodall, 1971; Grant, 1993; Lim, et. al., 1971; Maisal, et. al., 1974; Perry, et. al.,
1985; Pickney, et. al., 1985; Rahai, 1996; Rasmussan, et. al., 1998; Trikha, et. al., 1972) it
appears the British have been quite active in researching the behavior of curved concrete box
girder bridges. The British use their own code (BS5400) for bridge design.
The new “Eurocode” is intended to supersede the codes of the major European Countries. It has
been developed over a number of years and is currently in use. However, this code has
appendices which direct the designer to special provisions by individual countries (e.g., the DIN
code for Germany, BS5400 for England, etc.) and for the most part practice still follows the
traditional codes of the countries involved.
In Asia the British BS5400 (India, Malasia and Hong Kong) and AASHTO (Thailand, Taiwan,
Korea and Philippines) are widely used. Japan, which has its own code, frequently builds curved
concrete box girder bridges.
The structural code in Brazil is quite brief and all encompassing. It is much more concise than
the current U.S. design codes. Curved beams are not covered directly although there is a section
on torsion, but only with general instructions found in most textbooks. Bridges that have
alignments with slight curvature are generally designed as straight bridges without consideration
for the curve, except that bridge bearings are designed for eccentric loads taking into account the
curve of the superstructures.
As of today, Brazil, Argentina, Chile, and Mexico as well as other Latin American Countries use
computers and similar programs as the U.S.
The AASHTO design specifications (not necessarily LRFD) are widely used by many other
countries around the world.
Page 8

c. Field Problems

Several failures of stem concrete due to the radial forces developed by curved prestress tendons
have occurred over the years. These include the Las Lomas Bridge in 1978, the Kapiolani
Interchange in 1981, and the 405/55 HOV Connector OC in 2002. Repair costs for some of these
structures were significant (the Kapiolani Interchange was $4,000,000).
Prestress breakout failures have been linked to a combination of the regional action of the web
acting as a beam between the top and bottom deck and the local slab action of the concrete cover
over the prestressing tendons. Global actions, although theoretically a factor, have been found
not to be important in these failures.
Many such failures have occurred even in straight bridges where local curvatures of prestressing
ducts occur near the prestress anchorages. These stresses can either add to or subtract from the
stresses developed from horizontal alignment of the bridge depending on the direction of the
tendon flare.
Several members of the project team were involved in the investigation of the 405/55 HOV
Connector OC failure. Although the curvature of this structure was less than many, there were
several other geometric characteristics of this structure that led to the failure. First, because
prestress radial forces per duct were under the limit beyond which Caltrans Memo 11-31
required tie reinforcement, none was placed. Also, because the structure had fairly long spans,
the structure depth was relatively large as were the prestressing forces in each web. The resulting
large number of ducts required (5 per stem) for the increased prestress force were placed one on
top of the other without any space in between. The combination of proportionately larger radial
prestress forces applied to a deeper web exacerbated regional concrete stresses. When these
stresses were combined with the local stresses generated in the concrete cover over the stacked
ducts, concrete cracking and spalling occurred. This particular design pushed the envelope for
Caltrans design requirements to prevent a breakout failure, and it is generally agreed that have
the Caltrans lateral tie detail been used, the failure could have been prevented.
Abutment bearing failure that is progressing with time was experienced on the I-5 NB to Hwy
217 NB ramp in Oregon. The single cell box girder is supported on two large bearings at the
south abutment. Over time the entire load at these bearings has shifted to only one of the
bearings while the other has experienced uplift. These problems are thought to be due to the
time-dependent behavior of concrete. This theory is corroborated somewhat by the results
observed in the time dependant analyses of similar structures, although it is thought that
currently commercially available software will tend to over predict the problem because torsion
creep is not considered.
Another recent bearing failure occurred on the bridge at Wildcat Road in Shasta County,
California. This is a single span curved prestressed concrete box girder bridge that was under
construction. When the falsework was being removed, the bearings at the abutments began to
fail. The outside elastomeric bearing was overloaded and was destroyed and the bearing at the
inside of the curve began to lift off. This problem was corrected at the abutments by retrofitting
the bridge with prestress bar tie-downs and eliminating the bearings. This essentially converted
the seat abutments to end diaphragm abutments. Fortunately the relatively short bridge length
and the fact that most of the prestress shortening had already taken place made this possible.
Page 9
Stirrups in the outside web were also inadequate to resist the combined effects of torsion and
flexural shear in this structure. The web was retrofitted with external prestressing tendons that
will correct the problem. This repair was deemed to be preferable to adding extra mild
reinforcement within a web overlay.
A recent problem with two box girder bridges in Coahuila, Mexico that is apparently due to the
curvature of the structures has developed. These bridges, which are relatively new, are cast-in-
place, post-tensioned, continuous concrete box girder bridges supported on single bearings at
each non-integral single column. These relatively narrow multi-span ramp structures are
experiencing on-going deflections and lateral movement. It is not clear what is causing this
behavior but it is fairly certain that curvature is a factor. The bearings have experienced uplift
due to rotation of the superstructure as shown in Figure 2-1. The movement was severe enough
to remove the superelevation that had been placed in the bridge at the time of construction.
Significant cracking of the superstructure was also observed. The bridge owner is attempting to
correct the problem by increasing the size of the piers in the transverse direction as shown in
Figure 2-2 and jacking and shimming the superstructure back to its original as-built position. The
wider piers will allow bearings to be placed eccentric to the centerline of the superstructure and
hopefully stabilize the situation. A lightweight overlay is also being considered to completely
restore the superelevation.


Figure 2-1
Uplift at Edge of Bearing

Page 10

Figure 2-2
Construction of Widened Column
Page 11
3. PUBLISHED LITERATURE REVIEW

There is a large body of published literature related to curved concrete box girder bridges. Some
of the documents most important to this project are discussed below.

a. Codes and Design Standards

Currently there is no national code within the U.S. specifically developed for the design of
curved concrete box-girder bridges. AASHTO LRFD Bridge Design Specifications provide
comprehensive specifications with commentary for the design of highway bridges. A summary
of our review of Codes and Design Standards are summarized below.

- American Association of State Highway and Transportation Officials (AASHTO), 2004,
AASHTO LRFD Bridge Design Specifications, 3
rd
Edition with Interims, AASHTO,
Washington, D.C.
A number of sections are applicable to the design issues associated with curved concrete box
girders. Selected specifications articles are summarized as follows:
Article 4.6.1.2.2 “Single-Girder Torsionally Stiff Superstructures” allows for the analysis of
horizontally curved, torsionally stiff single-girder superstructures for global force effects as a
curved spine beam.
Article 4.6.1.2.3 “Multicell Concrete Box Girders” allows for the design of horizontally curved
cast-in-place multicell box girders as single-spine beams with straight segments, for central
angles up to 34
o
within one span, unless concerns about force effects dictate otherwise.
Article 4.6.3.4 “Cellular and Box Bridges” allows for the refined analysis of cellular bridges by
any of the methods specified in Article 4.4 “Acceptable Methods of Structural Analysis,” except
the yield line method, which accounts for the two dimensions seen in plan view and for the
modeling of boundary conditions. Models intending to quantify torsional warping and/or
transverse frame action should be fully three-dimensional.
Article 4.7.4.3 “Multispan Bridges” specifies the minimum requirements for the seismic analysis
of multispan bridges. Analysis requirements are based on the classification of a bridge as
“regular” or “irregular.” The classification of a curved bridge include the maximum subtended
angle and whether the spans are continuous or are multiple simple-spans.
Article 5.4.6 “Ducts,” specifies the requirements for duct material and curvature.
Article 5.8 “Shear and Torsion” specifies comprehensive design procedures for flexural shear
and torsion. The modified compression field theory is specified for flexural regions. Strut-and-tie
models are specified for regions near discontinuities. Alternative design procedures are permitted
for segmental bridges.
Article 5.9.1.6 “Tendons with Angle Points or Curves” cross references Articles 5.4.6 and 5.10.4
for duct curvature and stress concentration considerations, respectively.
Page 12
Article 5.9.5.2.2.2 “Friction” specifies the friction loss due to curvature and includes specific
requirements for determination of the total three-dimensional angle change as typically found in
curved girders with draped tendons.
Article 5.10.3.3.3 “Curved Post-Tensioning Ducts” specifies the clear distance between curved
ducts as required for tendon confinement as specified in Article 5.10.4.3 but not less than that
required for straight ducts.
Article 5.10.4.3 “Effects of Curved Tendons” specifies that where tendons are placed in curved
webs, additional cover and/or confinement reinforcement shall be provided.
Article 5.10.4.3.1 “In-Plane Force Effects” defines the in-plane deviation force effects due to the
change in direction of the tendon as F
u-in
= P
u
/R where P
u
is the factored tendon force and R is
the radius of curvature of the tendon. Specific requirements for local lateral shear on the
unreinforced concrete cover are given and neglect any increase in lateral shear capacity for
widely spaced tendons. Where the factored in-plane deviation force exceeds the lateral shear
resistance of the concrete cover, tieback reinforcement is required. Where stacked ducts are used
in curved girders, the moment resistance of the concrete cover, acting in flexure, shall be
investigated but no specific methodology or stress requirements are provided. For curved girders,
the global flexural effects of out-of-plane forces shall be investigated.
Article 5.13.2.2 “Diaphragms” requires the use of diaphragms at abutments, piers, and hinge
joints. Intermediate diaphragms may be used in beams in curved systems or where necessary to
provide torsional resistance. Intermediate diaphragms shall be used in curved box girders with an
inside radius of less than 800-feet. Diaphragms may be omitted where tests or structural analysis
show them to be unnecessary.
Article 14.4.1 “General” specifies the movement requirements for joints and bearings. It includes
the requirement to consider the effects of curvature, skew, rotations, and support restraint. The
commentary includes additional discussion pertinent to curved bridges.
With respect to torsion design, a detailed review of the specifications was performed. The
following briefly describes the design methods, outlines the basic steps of designing a box
section for the combined actions of flexural shear, torsion, and moment, and includes a
discussion (in italics) where further guidance is required in interpreting or applying the LRFD
specification.

DESIGN METHODS
Two basic design methods are specified in Articles 5.8.3 and 5.8.6 that are dependent on
construction method and structure type. A sectional model using the modified compression field
theory with a variable angle truss model is the basis of Article 5.8.3 and is applicable in most
cases. Article 5.8.6 contains the flexural shear and torsion provisions specific to segmental post-
tensioned concrete box girder bridges. A conservative expression for the concrete contribution
along with a 45
O
degree truss model is assumed.
General Comment – There is a minor conflict between Articles 5.8.3 and 5.8.6. Article 5.8.3
states that Article 5.8.6 may be used for segmental post-tensioned concrete box girder bridges
while Article 5.8.6 states that Article 5.8.6 shall be used for segmental post-tensioned concrete
box girder bridges in lieu of Article 5.8.3. It needs to be clarified whether Article 5.8.6 is a
permissible or mandatory procedure for segmental bridges. Additionally, Commentary Article
Page 13
C5.8.6.1 states, “For types of construction other than segmental box girders, the provisions of
Article 5.8.3 may be applied in lieu of the provisions of Article 5.8.6.” It appears that the word
may should be replaced by shall unless the intent is to permit Article 5.8.6 as an alternative
design method for bridge types other than segmental.

DESIGN STEPS (GENERAL SECTIONAL MODEL)
The following outlines the basic steps of designing the exterior web of a box section for the
combined actions of flexural shear, torsion, and moment. It is based on the provisions of Article
5.8.3 and, therefore, does not cover the steps for a segmental post-tensioned concrete box girder
bridge.
Step 1 – Determine the Controlling Load Cases
Determine the controlling load cases for the applicable strength limit states. Consider the
concurrent actions on the section. As a minimum, consider the following two cases:
1. Maximum flexural shear and concurrent actions
2. Maximum torsion and concurrent actions
Perform steps 3 through 7 separately for each the above cases and any additional cases that may
potentially govern the design.
Step 2 – Determine the Cross Section Parameters:
A
cp
– total area enclosed by outside perimeter of concrete cross section (Article 5.8.2.1)
p
c
– the length of the outside perimeter of concrete cross section (Article 5.8.2.1)
A
cp
2
/ p
c
2 A
o
b
v
(Equation 5.8.2.1-5)
(Comment – LRFD Article 5.8.2.1 does not address the case when the thickness of the flange of a
non-segmental box section is less than the effective web width. This is addressed in Article
12.2.10 of the Segmental Guide Specification (Reference 2) and LRFD Article 5.8.6.3 for
segmental bridges.)
A
o
– area enclosed by the shear flow path (in.
2
) (Article 5.8.2.1)
d
s
– the length of the torsional shear flow path on the exterior web (in.) (Commentary 5.8.2.1)
p
h
– perimeter of the centerline of the closed transverse torsion reinforcement (in.) (Article
5.8.2.1)
b
v
– effective web width (in.) (Article 5.8.2.9)
d
v
– effective shear depth (in.) (Article 5.8.2.9)
Step 3 – Check the Web Width:
Verify the effective web width is adequate to prevent web crushing:
V
u
oV
n

V
n
0.25f’
c
b
v
d
v
+ V
p
(Equation 5.8.3.3-2)
(Comment – It may be prudent to reduce V
n
by the torsional shear when web crushing governs
the capacity. Note that provisions for considering the combined flexural shear and torsional
Page 14
shear are required by Article 12.3.1 of the Segmental Guide Specification and LRFD Equation
5.8.6.5-3 for segmental bridges.)
Increase web width if Equation 5.8.3.3-2 is not satisfied.
Step 4 – Calculate Shear Stress:
Check if torsion must be considered:
T
u
> 0.25oT
cr
(Equation 5.8.2.1-3)
'
2
'
125 . 0
1 125 . 0
c
pc
c
cp
c cr
f
f
p
A
f T + = (Equation 5.8.2.1-4)
For T
u
> 0.25oT
cr
, calculate the equivalent factored shear force, V
u
, acting on the web where the
flexural shear and the torsional shear are additive as follows:
V
u
= V
u (flexure)
+ T
u
d
s
/ (2A
o
) (Equation 5.8.2.1-7)
(Comment – Note that V
u
is determined for a single girder and T
u
is acting on the total cross-
section.)
General Comment – The fifth paragraph of Commentary C5.8.2.1 regarding the equivalent
factored shear force would benefit from additional clarification. There is a mention of a stress
limit for the principal tension at the neutral axis of the section but the specific code section was
not referenced. Article 5.8.5 provides limits on the principal tensile stress in the webs of
segmental concrete bridges at the Service III limit state and during construction. As the principal
stresses are checked using service load, V
u
is not applicable. It appears that the intent of the
equivalent factored shear force is to consider the increased shear force and resulting shear
stress due to torsion in calculating
x
, and V
c
. It appears that equivalent factored shear
force should not be considered in determining the required shear capacity, V
n
, or the required
tensile capacity specified in Article 5.8.3.5.
Using the equivalent factored shear force, V
u
, acting on the exterior web where the flexural shear
and the torsional shear are additive to calculate the shear stress as follows:
v
u
= (V
u
– oV
p
) / (ob
v
d
v
) (Equation 5.8.2.9-1)
(Comment – Note that v
u
includes the effects of flexural shear and torsional shear.)
Step 5 – Calculate v
u
/ f’
c
and 
x
and Find  and :
Calculate v
u
/ f’
c
using v
u
calculated in Step 4. This accounts for the increased shear stress due to
torsion and will be used to determine u and | from Table 5.8.3.4.2-1.
Calculate
( )
ps p s s
po ps p u u
v
u
x
A E A E
f A V V N
d
M
+ ·
|
|
.
|

\
|
÷ ÷ + +
=
2
cot 5 . 0 5 . 0
(Equation 5.8.3.4.2-1) using the
equivalent factored shear force, V
u
, determined in Step 4.
(Comment – This simply adds the tensions due to flexural shear and due to torsion for the
exterior web where the flexural shear and torsional shear are additive and appears to be
conservative. Collins and Mitchell, Reference 3, pages 399-400, use an equivalent longitudinal
Page 15
tension for combined flexural shear and torsion equal to the square root of the sum of the
squares of the individually calculated tensions for flexural shear and for torsion.)
If whole-width design is used, forces acting on the total section are applied. In this case, the
shear force is conservatively taken as the equivalent factored shear force, V
u
, determined in Step
4 multiplied by the total number of webs.
(Comment – Whole-width design is not specifically addressed and would benefit from
clarification in this area.)
This accounts for the increased longitudinal tensile force due to torsion and will be used to
determine u and | from Table 5.8.3.4.2-1.
Using the calculated values of v
u
/ f’
c
and c
x
, find u and | from Table 5.8.3.4.2-1.
Step 6 – Determine Required Spacing of Stirrups:
The amount of transverse reinforcement required for shear is found from:
V
u
o V
n

(Comment – Clarify that the factored flexural shear, not the equivalent factored shear force, is
used for V
u
, as the transverse reinforcement for torsion is determined separately.)
V
n
= V
c
+ V
s
+ V
p
(Equation 5.8.3.3-1)
Where:
V
c
= 0.0316|
'
c
f b
v
d
v
(Equation 5.8.3.3-3)
V
s
= (A
v
f
y
d
v
cotu) / s (Equation C5.8.3.3-1)
A
v
/ s = V
s
/ (f
y
d
v
cotu)
The amount of transverse reinforcement required for torsion is found from:
T
u
oT
n

Where:
T
n
= (2A
o
A
t
f
y
cotu) / s (Equation C5.8.3.6-2)
A
t
/ s = T
n
/ (2A
o
f
y
cotu)
For the exterior web of a box section, the combined area of both stirrup legs in the web, A
stirrups
,
contribute to A
v
and A
t
, therefore the maximum spacing of the stirrups, S
max
, is given by:
S
max
= A
stirrups
/ [(A
v
/ s)
flexural shear
+ (A
v
/ s)
torsion
]
Step 7 – Check the Longitudinal Reinforcement:
The required tensile capacity of the longitudinal reinforcement on the flexural tension side of the
member is found from Equation 5.8.3.5-1.
(Comment – Clarify that the flexural shear, not the equivalent factored shear force, is used for
V
u
, as the additional longitudinal reinforcement for torsion is determined separately.)
The longitudinal reinforcement required for torsion, in addition to that required for flexure, is
found from:
Page 16
A
l
= T
n
p
h
/ (2 A
o
f
y
) (Equation 5.8.3.6.3-2)
Comments:
Article 5.8.3.6.3 would benefit from a clarification that A
l
is also in addition to the required
tensile capacity from Equation 5.8.3.5-1 when Equation 5.8.3.5-1 exceeds the longitudinal
reinforcement required for flexure.
The distribution of A
l
within the cross section needs to be clarified. Article 5.3, “Notation,” and
Article 5.8.6.4, “Torsional Reinforcement,” define it as the area of longitudinal torsion
reinforcement in the exterior web of the box girder, which appears to be incorrect. LRFD
Equations 5.8.3.6.3-2 and 5.8.6.4-3 are essentially identical to the equation in Article 12.3.8 of
the Segmental Guide Specification, which specifies that A
l
shall be distributed around the
perimeter of the closed stirrups.
Prestressing steel should also be permitted to satisfy Equation 5.8.3.6.3-2 similar to Article
12.3.8 of the Segmental Guide Specification and LRFD Commentary Article C5.8.6.4 for
segmental bridges. The area of longitudinal torsion reinforcement in the flexural compression
zone should be permitted to be reduced similar to Segmental Guide Specification Article 12.3.9
and LRFD Equation 5.8.6.4-4 for segmental bridges.

DESIGN STEPS (SEGMENTAL BOX GIRDER)
The following outlines the basic steps of designing the exterior web of a box section for the
combined actions of flexural shear, torsion, and moment. It is based on the provisions of Article
5.8.6 and, therefore, applicable to a segmental post-tensioned concrete box girder bridge.
Step 1 – Determine the Controlling Load Cases
The design for flexural shear and torsion in segmental bridges shall be performed at the strength
limit state per Article 5.8.6.2. The shear component of the primary effective longitudinal
prestress force, V
p
, shall be added as a load effect with a load factor of 1.0. The component of
inclined flexural compression or tension, in the direction of the applied shear, in variable depth
members shall be considered when determining the design factored shear force.
In accordance with Article 5.8.5, principal stresses at the neutral axis of segmental bridges shall
not exceed the tensile stress limits of Table 5.9.4.2.2-1 at the Service III limit state and Table
5.14.2.3.3-1 during construction
Determine the controlling load cases for each of the three applicable limit states separately
(Strength, Service III, and during construction). Consider the concurrent actions on the section.
As a minimum, consider the following two cases for each of the limit states:
1. Maximum flexural shear and concurrent actions
2. Maximum torsion and concurrent actions
Perform steps 3 through 6 separately for each of the above Strength cases and any additional
Strength cases that may govern the design.
In step 4, check the principal stresses separately for each of the above Service III and
construction cases and any additional Service III and construction cases that may govern the
design.
Step 2 – Determine the Cross Section Parameters:
Page 17
A
o
– area enclosed by the shear flow path (in.
2
) (Article 5.8.6.3)
b
e
– effective width of shear flow path, but not exceeding the minimum thickness of the webs or
flanges comprising the closed box section.(in.) b
e
shall be adjusted to account for the presence of
ducts as specified in Article 5.8.6.1. (Article 5.8.6.3)
b
e
may be taken as A
cp
/ p
c
(Article 5.8.6.3)
A
cp
– area enclosed by outside perimeter of concrete cross section (in.
2
) (Article 5.8.6.3)
p
c
– the outside perimeter of concrete cross section (in.) (Article 5.8.6.3)
p
h
– perimeter of the centerline of the closed transverse torsion reinforcement (in.) (Article
5.8.6.4)
b
v
– effective web width (in.) (Article 5.8.6.5)
d
e
– effective depth from extreme compression fiber to the centroid of the tensile force in the
tensile reinforcement (in.) (Article 5.8.6.4)
d
v
– effective shear depth (in.) (Article 5.8.6.5)
Step 3 – Check if Torsion Must be Considered:
Check if torsion must be considered:
T
u
> 1 / 3oT
cr
(Equation 5.8.6.3-1)
T
cr
= 0.0632K
'
c
f 2A
o
b
e
(Equation 5.8.6.3-2)
K =
'
0632 . 0
1
c
pc
f
f
+ 2.0 (Equation 5.8.6.3-3)
f
pc
= Unfactored compressive stress in concrete after prestress losses have occurred either at the
centroid of the cross-section resisting transient loads or at the junction of the web and flange
when the centroid lies in the flange. (Article 5.8.6.3)
Step 4 – Check the Web Width:
Verify the effective web width is adequate to prevent web crushing:
V
u
oV
n

V
n
0.379f’
c
0.5
b
v
d
v
(Equation 5.8.6.5-2)
V
u
/ (b
v
d
v
) + T
u
/ (2A
o
b
e
) 0.474f’
c
0.5
(Equation 5.8.6.5-3)
(Comment – It appears that V
u
and T
u
should be replaced by V
n
and T
n
, respectively, in this
equation to be consistent with Article 12.3.1b of the Segmental Guide Specification.)
Increase web width if either Equation 5.8.6.5-2 or Equation 5.8.6.5-3 is not satisfied.
Check the allowable principal tensile stress for Service Limit State III and during construction in
accordance with Article 5.8.5. Consider the compressive stress due to vertical tendons in the
webs. Increase the web width or the vertical prestressing force in the web if the allowable
principal stresses are exceeded.
Page 18
Step 5 – Determine Required Spacing of Stirrups:
The amount of transverse reinforcement required for shear is found from:
V
u
oV
n

V
n
= V
c
+ V
s
(Equation 5.8.6.5-1)
Where:
V
c
= 0.0632K
'
c
f b
v
d
v
(Equation 5.8.6.5-4)
V
s
= A
v
f
y
d
v
/ s (Equation 5.8.6.5-5)
A
v
/ s = V
s
/ (f
y
d
v
)
The amount of transverse reinforcement required for torsion is found from:
T
u
oT
n

Where:
T
n
= 2A
o
A
v
f
y
/ s (Equation C5.8.6.4-2)
A
v
/ s = T
n
/ (2A
o
f
y
)
For the exterior web of a box section, the combined area of both stirrup legs in the web, A
stirrups
,
contribute to the transverse hoop reinforcement for flexural shear and torsion, therefore the
maximum spacing of the stirrups, S
max
, is given by:
S
max
= A
stirrups
/ [(A
v
/ s)
flexural shear
+ (A
v
/ s)
torsion
]
When vertical tendons are provided in the web, the design yield strength for flexural shear and
torsion design shall be taken in accordance with Article 5.8.2.8.
Step 6 – Check the Longitudinal Reinforcement:
The minimum additional longitudinal reinforcement required for torsion (in addition to that
required for other concurrent actions), shall satisfy:
A
l
= T
u
p
h
/ (2oA
o
f
y
) (Equation 5.8.6.4-3)
Comment: The distribution of A
l
within the cross section needs to be clarified. Article 5.3,
“Notation,” and Article 5.8.6.4, “Torsional Reinforcement,” define it as the area of longitudinal
torsion reinforcement in the exterior web of the box girder, which appears to be incorrect.
Article 5.8.6.4 contains a conflicting statement that A
l
shall be distributed around the perimeter
of the closed stirrups in accordance with Article 5.8.6.6 which appears to be correct. LRFD
Equation 5.8.6.4-3 is essentially identical to the equation in Article 12.3.8 of the Segmental
Guide Specification which specifies that A
l
shall be distributed around the perimeter of the
closed stirrups.
In determining the required amount of longitudinal reinforcement, the beneficial effect of
longitudinal prestressing is taken into account by considering it equivalent to an area of
reinforcing steel with a yield force equal to the prestressing force. (Commentary Article
C5.8.6.4)
Page 19
Subject to the minimum reinforcement requirements of Article 5.8.6.6, the area of additional
torsion reinforcement in the flexural compression zone may be reduced by an amount equal to:
M
u
/ (0.9d
e
f
y
) (Equation 5.8.6.4-4)

- AASHTO, 2003, Guide Specifications for Horizontally Curved Steel Girder Highway
Bridges, American Association of State Highway and Transportation Officials, Washington,
D.C.
This is a recently published AASHTO Guide Specification for curved steel bridges including box
girders. It has been suggested that the design specification contained therein be used as a model
for the NCHRP 12-71 project. This specification discusses design philosophy and limit states and
includes provisions for loads; structural analysis; design of flanges, webs, shear connectors,
bearings, splices and connections; deflections; and constructability. It also includes a
construction specification and design examples for both an “I” girder and box girder bridge.

- AASHTO, 1999, Guide Specifications for Design and Construction of Segmental Concrete
Bridges, 2
nd
Edition with Interims, American Association of State Highway and
Transportation Officials, Washington, D.C.
This second edition of the guide specifications for segmental concrete bridges was prepared for
use in conjunction with the Standard Specifications for Highway Bridges (non-LRFD), and
subsequent interim revisions to these specifications. This publication, which was developed by a
broad-based committee organized by the American Segmental Bridge Institute, embodies several
concepts, which are significant departures from previous design and construction provisions. It is
based on recent research in the United States and abroad. The committee included
representatives of state departments of transportation, the Federal Highway Administration,
academicians, consulting engineers, contractors, and suppliers. Some of the details of this
specification were discussed above.

b. Design Philosophy

A number of books and papers have been written about the design of concrete box girder
bridges. Many of these discuss the effect of horizontal curvature on the behavior of these
bridges. A few of these publications are discussed below.

- Menn, C., 1990, Prestressed Concrete Bridges, ISBN 3-7643-2414-7, Birkhauser Verlag,
Basel
This book provides engineers with a comprehensive overview of the fundamental principles
governing the design and construction of concrete bridges. The content is based on the author’s
direct experience gained from the design and construction of bridges in Switzerland. While much
of the content is based on Swiss standards and practices, the material is presented in a manner
that stresses fundamental principles. This book covers straight, skewed, and curved bridges
Page 20
consisting of both open sections and closed box sections. The book addresses a number of issues
that are applicable to horizontally curved box girder bridges.
Article 4.6.4, “Detailing,” discusses the deviation forces generated by curved post-tensioning
tendons. The deviation force per unit length is determined as the tendon force divided by the
radius of curvature of the tendon. An example of the regional transverse bending moment in the
web of a horizontally curved member is presented.
Article 5.1.4 “Structural Models for Bridge Superstructures,” provides guidance on developing
analytical models that can be applied to curved box girder bridges. For example, techniques for
developing grillage models of multiple cell box girders are presented.
Article 5.3.2(b) “Web Design for Shear and Transverse Bending,” presents a rational method for
the design of webs subject to combined shear and regional transverse bending, a condition that
occurs in horizontally curved post-tensioned box girder bridges. The method is based on Swiss
practice and neglects the concrete contribution to the shear capacity. Shear-regional transverse
bending interaction diagrams are presented.
Article 5.3.4 “Diaphragms,” discusses the function, necessity, and design of internal diaphragms.
Diaphragms are recommended at abutments and piers. The use of intermediate diaphragms is
usually not necessary in straight and lightly curved box girder bridges.
Article 6.1.3(b) “Influence of Girder Curvature,” discusses the qualitative difference in
superstructure displacements due to temperature and shrinkage vs. longitudinal prestressing in
horizontally curved bridges.
Article 7.6, “Curved Girder Bridges,” is devoted entirely to horizontally curved bridges. The
article includes subsections on Conceptual Design, Analysis, Transformation of Torque into
Torsional Sectional Forces, Prestressing, and Prestressing Concept and Tendon Layout.
The discussion on the conceptual design of curved box girders points out the role of torsion in
design and how at ultimate loads torsion and bending moment can be redistributed. The effect of
torsion on bearing forces may require the bearings to be placed away from the webs. Expansion
bearings must be able to accommodate both temperature and prestress shortening displacements
which will be in different directions.
The book presents a simplified method for analyzing curved bridges iteratively. The method does
not satisfy compatibility equations exactly, but greatly simplifies the computational effort. An
example is given. Simple vector diagrams are presented to illustrate how torsional section forces
are developed by a variation in the direction of longitudinal bending moments due to the
curvature of the superstructure and resisted by shear flow around the perimeter of the box
section.
Prestressing can produce longitudinal and transverse bending and shear forces as well as torque
in curved box girder bridges. Torsion, which can increase flexural stresses, must be considered in
determining the required prestressing force. Prestressing can also be used to enhance torsional
and transverse bending resistance although this is often avoided for economic reasons.

Page 21
- Sennah, K. M., and Kennedy, J. B., 2001, State-of-the-Art in Curved Box-Girder Bridges,
Journal of Bridge Engineering, ASCE, 6(3), 159-167
The objective of this paper is to provide highlights of the most important references related to the
development of current guide specifications for the design of straight and curved box-girder
bridges. As such it provided an excellent bibliography from which to identify other papers that
were reviewed in detail. Subjects discussed in this review include (1) different box-girder bridge
configurations; (2) construction issues; (3) deck design; (4) load distribution; (5) deflection and
camber; (6) cross-bracing requirements; (7) end diaphragms; (8) thermal effects; (9) vibration
characteristics; (10) impact factors; (11) seismic response; (12) ultimate load carrying capacity;
(13) buckling of individual components forming the box cross section; (14) fatigue; and (15)
curvature limitations provided by the codes for treating a curved bridge as a straight one. The
literature survey presented herein encompasses (1) the construction phase; (2) load distribution;
(3) dynamic response; and (4) ultimate load response of box-girder bridges.

- ASCE Committee on Construction Equipment and Techniques, 1989, Concrete Bridge
Design and Construction in the United Kingdom, Journal of Construction Engineering and
Management, Vol. 115, No. 4, pp. 618-635
The design and construction of concrete bridges in the United Kingdom has changed rapidly
during recent decades. Better analytical methods, and increased mechanization and better
planning in the construction of these bridges have brought this about. However these steps have
also brought new problems in their wake for the engineer, contractor, and supervisor. This paper
shows the different approaches on a number of factors. The paper is divided into three parts as
follows:
1. The design of bridges in classes for span and type with reference to the pertinent factor
for that design.
2. The contractor’s approach to construction that illustrates the need for flexibility in the
construction method in order to meet contract deadlines.
3. The views of the supervising engineer, and his means of achieving a balance between the
designer’s intentions and the contractor’s proposals.

- Schlaich, J., and Scheef, H., 1982, Concrete Box Girder Bridges, ISBN 3 85748 031 9,
International Association for Bridge and Structural Engineering, Zurich, Switzerland
This publication is the outcome of a comprehensive survey of concrete box girder bridges. The
publication is divided into three main parts, “Design”, “Structural Analysis”, and “Dimensioning
and Detailing.” A comprehensive reference list is included. The publication addresses straight,
skew, and curved bridges.
The “Design” section covers several aspects of curved bridges. Recommendations are given for
when the longitudinal bending moments can be determined as for a straight bridge and then
combined with the torsional effects without considering the coupling of the two effects upon
each other. Alternative substructure configurations are discussed for curved bridges.
The “Structural Analysis” section discusses the mutual influence of the longitudinal bending
moments and torsional moments for horizontally curved bridges. A simple table of equations
Page 22
based on classical curved beam theory is presented for several different loading conditions of a
curved single span bridge with fixed supports.
The “Dimensioning and Detailing” section covers several aspects of curved bridges.
Dimensioning and reinforcement of the web for flexural shear, torsion, and regional transverse
bending is addressed including a rational method of designing the web reinforcement for
combined shear and regional transverse bending. The influence of horizontal curvature on the
movements at bearings is also discussed.

c. Response of Curved Concrete Box Girder Bridges

i. Global Analysis

Most published research seems to be directed toward the global response of box girder bridges.
A number of analytical techniques have been studied. Many of these are relatively complex, but
many others are suitable for production design practice. Our current feeling is that a properly
applied grillage analogy method provides good results and may be most suitable for analyzing
bridges with significant curvature. The following paragraphs discuss some of the papers and
reports that were reviewed.

- Al-Rifaie, W. N., and Evans, H. R., 1979, An Approximate Method for the Analysis of Box
Girder Bridges that are Curved in Plan, Proc., Int. Association of Bridges and Structural
Engineering, Int. Association for Bridge and Structural Engineering (IABSE), pp 1–21.
An approximate method for analyzing curved box girder bridges using the nodal section method
is described. This method, which was originally developed for straight box girders, is adapted for
curved box girders and is useful for preliminary design of these structures. The method
developed is applicable to simple span, single cell box girders.
In this procedure, the transverse nodal section is idealized as a plane frame. Nodes on the frame
are assumed to be fixed against translation but free to rotate. Each frame is analyzed and the
reactions at the nodes determined. The reactions are then applied to longitudinal plates that
represent the components of the box girder. The plates are only capable of resisting in-plane
forces. The final step is to apply a “sway correction” procedure that will make the displacements
of the nodes in each of the transverse nodal sections compatible with the deflections of the nodes
along the edges of the longitudinal plates. This approach results in substantial computational
savings over the finite element method and is well suited to preliminary design studies.
The method was checked against finite element results for model bridges representing both
concrete and steel box girders. Several different span-to-radius ratios and loading conditions
were considered. In general, good correlation was found for critical stresses although some
discrepancy was found for non-critical stresses. The method as developed does not accurately
account for shear lag in the deck.
Studies have extended this method to multi-cell box girders and have developed methods to
account for shear lag in straight box girders. These refinements are also being considered for
curved box-girders.
Page 23

- Bazant, Z., and Nimeiri, M. E., 1974, Stiffness Method for Curved Box Girders at Initial
Stress, Journal of the Structural Division, Vol. 100, No. 10, pp. 2071-2090
A sophisticated numerical method of analysis of the global behavior of long curved or straight
single-cell girders with or without initial stress is presented. It is based on thin-wall beam
elements that include the modes of longitudinal warping and of transverse distortion of the cross
section. Deformations due to shear forces and transverse bi-moment are included, and it is found
that the well-known spurious shear stiffness in very slender beams is eliminated by virtue of the
fact that the interpolation polynomials for transverse displacements and for longitudinal
displacements (due to rotations and warping) are linear and quadratic, respectively, and an
interior mode is used. The element is treated as a mapped image of one parent unit element and
the stiffness matrix is integrated in three dimensions, which is numerical in general, but could be
carried out explicitly in special cases. Numerical examples of deformation of horizontally curved
bridge girders, and of lateral buckling of box arches, as well as straight girders, validate the
formulation and indicate good agreement with solutions by other methods. This method is most
applicable to steel box girders and is of little use to our project.

- Buragohain D. N., and Agrawal, B. L., 1973, Analysis of Curved Box Girder Bridges, Journal
of the Structural Division, Vol. 99, No. 5, pp. 799-819.
A discrete strip energy method is presented for the analysis of curved box girder bridges of
arbitrary cross section and various forms of curved folded plate structures simply supported at
the two ends and composed of elements that may, in general, be segments of conical frustra. The
method described is applicable to orthotropic material properties, arbitrary cross sections,
constant curvature and pinned supports at both ends. The method is based on harmonic analysis
in the circumferential direction. The total potential energy of the structure is discretized into
energy due to extension and bending and energy due to shear and twisting. The two types of
circumferential strip elements are obtained by using a modified finite difference discretization in
the transverse direction. The use of minimum energy principles yields two types of element
matrices that are assembled to form the overall stiffness matrix of the structure following
stiffness matrix procedures. Results of two examples obtained by the method are compared with
available solutions. The applicability of this paper to NCHRP 12-71 is limited because it is only
applicable to simply supported bridges, and the tool (software) to implement this method is not
readily available.

- Choudhury, D., and Scordelis, A. C., 1988, Structural Analysis and Response of Curved
Prestressed Concrete Box Girder Bridges, Transportation Research Record 1180, Transportation
Research Board, Washington, D. C., 72-86.
A numerical finite element analysis method for linear-elastic analysis and nonlinear material
analysis of curved prestressed concrete box girder bridges is demonstrated through two
examples. A curved nonprismatic thin-walled box beam element is used to model the bridges.
The cross section of the element is a rectangular single-cell box with side cantilevers. Eight
displacement degrees of freedom, including transverse distortion and longitudinal warping of the
cross section, are considered at each of the three element nodes. Prestressing, consisting of
posttensioned bonded tendons in the longitudinal direction, is considered. For nonlinear material
Page 24
analysis, the uniaxial stress-strain curves of concrete, reinforcing steel, and prestressing steel are
modeled. The shear and the transverse flexural responses of the box beam cross section are
modeled using trilinear constitutive relationships based on cracking, yielding, and ultimate
stages. The first example demonstrates the versatility of the numerical method in determining the
linear-elastic distribution of forces in a three-span prestressed box girder bridge of curved plan
geometry and variable cross section. Dead load, live load, and prestressing load cases are
analyzed. In the second example, overload behavior and ultimate strength of a three-span curved
prestressed concrete box girder bridge under increasing vehicular load are investigated. The
different response characteristics of the bridge induced by different transverse locations of the
overload vehicle are presented.
Although the finite element formulation might be detailed and comprehensive and conducive to
studying the ultimate behavior of concrete box girder bridges, its applicability to the planned
global elastic analysis studies is limited due to its complex nature.

- Chu, K. H. and Pinjarkar, S. G., 1971, Analysis of Horizontally Curved Box Girder Bridges,
Journal of the Structural Division, Vol. 97, No. 10, pp. 2481-2501
A finite element method for the analysis of simply supported curved girder bridges with
horizontal sector plates and vertical cylindrical shell elements is outlined. Stiffness coefficients
of sector plates are presented herein whereas stiffness coefficients of shell elements are based on
Hoff’s solution of Donnell’s equations. The authors claim that this analysis is much more
accurate than other methods of analysis.
Results of a sample bridge analysis are shown with stresses and deflections reported for a simply
supported multi-cell concrete bridge. Some interesting results, particularly those with respect to
the effect of radius of curvature, were obtained. Although a comparison is made between the
results of a curved twin box girder bridge obtained by the proposed method and another
approximate analytical method (Tung, 1967), no comparisons with other (simpler) analysis
methods are given. The FEM analysis tool itself is not readily available and thus is of limited
use; but the results of the analysis can serve as a comparison case for measuring the accuracy of
other methods, if more detail on the presented example can be obtained.

- Bridge Design System (BDS), 1986, A Computer Program for Analysis and Design of Multi-
Cell Box Girder Bridges, ECC, 1986
The described software program is the most commonly used software for design of multi-cell
box girder bridges. Bridges are modeled as a plane frames ignoring all horizontal curve effects.
This modeling technique is significant for the NCHRP 12-71 project as it is commonly used in
practice and its limits of applicability need to be investigated.

- Computers and Structures, Inc., 1998, “SAP2000 – Integrated Finite Element Analysis and
Design of Structures,” CSI, Berkeley, California, October 1998.
This reference constitutes the concrete structure portion of the SAP-2000 Manual, with emphasis
on design code check analysis. SAP2000 features integrated modules for design of both steel and
reinforced concrete structures. The program provides the user with options to create, modify,
analyze and design structural models. The program is structured to support a variety of design
Page 25
codes for the automated design and check of concrete frame members. The program currently
supports a number of foreign and domestic design codes. Given that the design code check
features of the program focus on frame analysis, these design code checks are of limited
usefulness for the specialized needs of curved concrete box girder bridge “local” or sectional
(“regional”) analysis. But the program is, of course, very useful for global analysis.
Chapter II outlines various aspects of the concrete design procedures of the SAP2000 program.
This chapter describes the common terminology of concrete design as implemented in SAP2000.
Each of six subsequent chapters gives a detailed description of a specific code of practice as
interpreted by and implemented in SAP2000. Each chapter describes the design loading
combination, column and beam design procedures, and other special consideration required by
the code.
Aside from the obvious usage as a SAP user reference, this document is useful as a summary
(and side-by-side comparison) of various design codes for concrete columns and beams. Other
than this, however, it is of limited direct utility to the NCHRP 12-71 project. It should also be
noted that there is no coverage of design or analysis of prestressing in this document.

- Fu, C. C., and Tang, Y., 2001, Torsional Analysis for Prestressed Concrete Multiple Cell
Box, ASCE Journal of Engineering Mechanics, 127(1), pp 45-51.
Using the Softened Truss Model, the authors present the formulation for calculation of torsional
effects in a multi-cell reinforced and prestressed concrete box girder bridge. This paper asserts
that because concrete box girder sections are not made of thin webs and flanges, the stress
distribution in these components are not constant, and vary through the thickness, causing the
effective stiffness of the member to be less than that observed at low values of load (torque). The
formulation is coded in a computer program and the results from an example problem are
presented. This research may be of some value to NCHRP 12-71 if the methodology can be
simplified and used as the basis of simplified methods for calculating torsional effects. However,
the presented paper, in its current form, is too complex to be used in practical design situations or
for parametric studies.

- Lopez, A., and Aparico, A.C., 1989, Nonlinear Behavior of Curved Prestressed Box-Girder
Bridges, IABSE Periodica, Zurich, 132(1), 13-28.
This paper describes an analytical study of the ultimate strength of horizontally curved
reinforced and prestressed concrete box girder bridges. The analysis was performed using
materially nonlinear plane stress finite elements (i.e. panels) that exhibit membrane action. The
material is assumed to have a variable modulus of elasticity that is strain dependent. Panel
behavior is based on the evolutive truss analogy with peak stress reduction (Vecchio and Collins,
1986). Reinforcing steel and prestress strand are stressed uniaxially according to an assumed
multi-linear stress strain relationship. Section warping is not considered. Classical matrix
analysis techniques are used to perform the analysis.
A 5-span bridge was selected to study the difference between linear and nonlinear response. Live
loads were located at various transverse positions and the behavior observed as the intensity of
these loads was increased. Based on these studies the following conclusions were drawn.
Page 26

1. The structural response is highly nonlinear at ultimate loads.
2. The form and degree of internal force redistribution at ultimate loads depends on the
loading case considered.
3. Internal forces are redistributed due to progressive cracking and structural coupling
between bending and torsion.
4. The type of failure depends on the loading case considered.
5. Ultimate internal force response can be evaluated accurately using plastic sectional
analysis.
6. Transverse prestressing significantly affects post-cracking response.
The following criteria are proposed for design of curved prestressed box girder bridges.
1. The response of the bridge under service loads can be accurately predicted using elastic
models.
2. Elastic models cannot accurately (and will often non-conservatively) predict the ultimate
limit state.
3. When using linear analysis to determine the factor of safety against failure, cracked
flexural and torsional section properties should be used to determine demands and plastic
sectional analysis should be used to determine capacities.

- Meyer, C., 1970, Analysis and Design of Curved Box Girder Bridges, Structural Engineering
and Structural Mechanics Report No. UC SESM 70-22, University of California, Berkeley,
December 1970
The history of curved bridges and the highway geometric requirements of these structures are
discussed. The report outlines the methods developed over the years for analyzing curved
bridges. These include straight beam approximation, curved beam theory, refined curved beam
theories, plate and grillage analysis methods, finite element analysis, and the finite strip method
analysis of curved folded plates. Refined curved beam theories are required to analyze thin-
walled box sections that can experience warping of the cross-section in the transverse direction.
Because concrete box sections have relatively thick walls, warping is generally small and
ordinary curved beam theory can be used successfully.
Two methods of analysis are developed in the form of computer programs. The first program,
FINPLA2, uses the finite element method. The second program, CURSTR, uses the finite strip
method of curved folded plates. The solution methodology requires that loadings be applied in
the form of Fourier series. Both programs yield essentially the same results.
The CURSTR program was used to study wheel load distribution in 1, 2, 3 and 4 cell concrete
box-girder bridges. Several parameter studies were conducted with different curvatures, span
lengths, deck widths, depth to span ratios and loading configurations.
With respect to single cell boxes, the following observations were made:
1. The girder on the inside of the curve is stiffer than the girder on the outside of the curve
and will attract more load.
Page 27
2. Load distribution improves with an increase in curvature. This behavior is independent of
span, cell width, and depth to span ratio.
3. The girder on the outside of the curve has a larger statical moment because of its longer
span.
4. The combination of items 1 and 3 result in nearly equal moments in the two girders.
5. The influence of span length on load distribution is similar to straight girders.
6. The influence of depth to span ratio is also similar.
For two-cell boxes:
1. The moments in the middle girder and the girder on the inside of the curve increase with
curvature.
2. The moment in the girder on the outside of the curve decreases with curvature up to a
certain level and then starts to decrease.
3. The influence of span on load distribution is small.
4. Cell width accelerates the curvature effects.
The response of three and four cell box-girders exhibit similar characteristics to one and two cell
boxes.
With respect to negative moments over a “fixed” support:
1. The girders may be assigned moments proportional to their moments of inertia.
2. Load distribution is generally worse in continuous bridges.
For design, approximate methods are justified and even preferred in most cases. A girder
moment distribution factor is developed:

'
÷ + = o
600
L
1 . 2
R
B
1
Bridges with curvatures radii large than 1000 ft. may be considered straight for analysis
purposes.

- Nakai, H., and Heins, C. P., 1977, Analysis Criteria for Curved Bridges, Journal of the
Structural Division, ASCE, Vol. 103, No. 7, pp 1419-1427.
The paper reports on a series stiffness equations and limiting angle equations developed for
determination of need for analysis of a bridge as a curved structure. The equations consider the
type of supporting element, (i.e. open girder, spread box, or single cell box), bending and
torsional stiffness and central angles, and the induced stresses and deformation. It appears that
these equations are specific to steel girders and warping torsion is a part of this methodology.
However, the overall approach may be utilized or modified to be applicable to concrete bridges
and NCHRP 12-71, especially for flowcharting the decision path for analysis.
The paper provides equations for moment, stress, and deflection of curved and straight bridges.
Design criteria for curved bridges have been formulated using these equations along with
parametric studies. The range for the parameter , which “relates the cross-sectional geometry
Page 28
and the spacing between the outside girders, is determined for multiple I, twin box, and monobox
girders. Data for multicell girders are not available from this paper. The bounds for the torsional
stiffness parameter are derived and are dependent on “the central angle,” which is the total
horizontal angle the girder passes through between supports, the torsional rigidity of the cross-
section, and EI. The deflection ratio is primarily dependent on , which “reflects the bending and
torsional stiffness of the girders.” The relationship between and the central angle were also
found for the three bridge types studied.
Conclusions made in the paper are as follows:
A series of stiffness parameter equations and limiting angle equations have been presented,
which provide information to the designer in determining the need for a curved girder analysis.
The expressions are functions of the girder types, bending and torsional stiffness, and central
angles.
The evaluation of gives the following criteria: “when is less than 0.4 evaluation of stresses
due to pure torsion may be omitted. When is greater or equal to 10 evaluation of stresses due
to warping may be omitted.”

- Sennah, K. M., and Kennedy, J. B., 2002, Literature Review in Analysis of Curved Box-
Girder Bridges, Journal of Bridge Engineering, ASCE, 7(2), 134-143.
The curvilinear nature of box girder bridges along with their complex deformation patterns and
stress fields have led designers to adopt approximate and conservative methods for their analyses
and design. Recent literature on straight and curved box girder bridges has dealt with analytical
formulations to better understand the behavior of these complex structural systems. Few authors
have undertaken experimental studies to investigate the accuracy of existing methods. This paper
presents highlights of references pertaining to straight and curved box girder bridges in the form
of single-cell, multiple-spine, and multicell cross sections. The literature survey presented herein
deals with: (1) elastic analysis, and (2) experimental studies on the elastic response of box girder
bridges.
The elastic analysis techniques that are discussed include:
1. Orthotropic Plate Theory Method
2. Grillage-Analogy Method
3. Folded Plate Method
4. Finite-Strip Method
5. Finite-Element Method
The orthotropic plate method lumps the stiffness of the deck, webs, soffit and diaphragms into an
equivalent orthotropic plate. The Canadian Highway Bridge Design Code (CHBDC, 2000)
recommends limiting this method to straight bridges with multi-spine cross-sections. Parameter
studies indicated that acceptable results are given for bridge with three or more spines.
In the grillage analogy method, the multi-cellular structure is idealized as a grillage of beams.
The CHBDC does not recommend that this method be used for sections with less than three cells
or box beams. This method requires special attention to the modeling of shear lag and the
Page 29
torsional stiffness of closed cells. When modeling is properly done, this method yields results
that compare well with finite element techniques.
The folded plate method uses plates to represent the deck, webs and soffit of box girders.
Diaphragms are not modeled. The plates are connected along their longitudinal edges and loads
are applied as harmonic load functions. The method is time consuming and only applicable to
restrictive support conditions.
The finite-strip method has been widely researched. It is essentially a special case of the finite-
element method but requires considerably less computational effort because a limited number of
finite strips connected along their length are used. Its drawback is that it is limited to simply
supported bridges with line supports and thus not applicable as a general use analysis tool for
production design.
With the advent of powerful personal computers and computer programs, the finite element
method has become the method of choice for complex structural problems. Many researchers
have applied this technique to the analysis of curved box girder bridges. A problem that occurs is
that a large number of flat plate elements are required to properly model the curved elements of a
curved bridge. Several researchers have attempted to overcome this difficulty by developing
special elements or using special substructuring techniques. The versatility of this method has
allowed researchers to investigate several aspects of bridge behavior, including dynamics, creep,
shrinkage, temperature, etc.
Curved box girder structures cannot be accurately analyzed using the classical curved beam
theory developed by Saint-Venant because it does not account for warping, distortion and
bending deformations of the individual wall elements of the box. Vlasov first developed an
adaptation of Saint Venant theory to thin-walled sections. Even this adaptation does not account
for all warping and bending stresses. Considerable research effort has been expended over the
years to develop computational techniques to over come shortcomings in the present theory.
Several laboratory experiments involving model box girder bridges have been conducted over
the years. In general, these experiments have shown good agreement with analytical results;
particularly those obtained using the finite element method of analysis.
In conclusion, the finite-element method, though more difficult to apply, accounts for all relevant
behavior in curved box girder bridges and yields the most reliable analysis results. Many
computer programs have been developed specifically for box girder bridges, but most of these
are not commercially available.

- Turkstra, C. J., and Fam, A. R. M., 1978, Behavior Study of Curved Box Bridges, Journal of
the Structural Division, Vol. 104, No. 3, pp. 453-462
A numerical analysis of a number of single-cell curved box-girder sections with variable
curvature, length, web spacing, number of diaphragms, and loading was performed. The effects
of these parameters on longitudinal stresses are considered, based on selected numerical results.
Implications for preliminary design are presented for both concrete and composite concrete/steel
sections

Page 30
- Reilly, R. J., 1972, Stiffness Analysis of Grids Including Warping, Journal of the Structural
Division, Vol. 98, No. 7, pp. 1511-1523
Two methods of including warping effects in the stiffness method of analysis are presented.
Method B seems to be superior to Method A for cases where the warping constant is not large. In
the limiting case where I
w
= 0, the warping effects disappear and leave only the familiar GI
x
/L.
When I
w
is small relative to I
x
(approximately pL > 5 for each element) computational errors
grow, because the stiffness matrix tends to become singular as the elements on the main diagonal
associated with warping approach zero. This is not a serious practical problem, as good solutions
can be obtained using an ordinary grillage analysis neglecting warping for structures where
warping stiffness is small and p is large. Composite bridges seem to fall near the borderline
where warping can be neglected. The bridge used in the example above was noncomposite so
that warping would be significant. Bimoment and warping torsion are obtained for grillage
structures. Results of computer programs based upon these methods are shown to agree closely
with published solutions for straight beams, a curved beam and a curved highway bridge.

- Meyer, C., and Scordelis, A. C., 1971, Analysis of Curved Folded Plate Structures, Journal
of the Structural Division, Vol. 97, No. 10, pp. 2459-2480
A finite strip method of analysis is presented which can be used to analyze curved folded plate
structures simply supported at the two ends and composed of elements that may in general be
segments of conical frustra. The method is based on a harmonic analysis in the circumferential
direction, with the loadings expressed by Fourier series, and on a finite element stiffness analysis
in the transverse direction. The direct stiffness method is used to assemble the structure stiffness
matrix and to determine displacements and element stresses. A description of a general computer
program developed for the analysis and the results of several examples are also given.

- Okeil, Ayman M., and El-Tawil, Sherif, 2004, Warping Stresses in Curved Box Girder
Bridges: Case Studies, Journal of Bridge Engineering, Vol. 9, No. 5, ASCE, September 1, 2004.
This paper discusses a number of case studies that were performed on 18 actual composite steel-
concrete box girder bridges. These analytical studies were conducted using the computer
program ABAQUS and a special 7-degree of freedom beam-column element that can account for
warping. These studies, which were designed to investigate warping related stresses in these
bridges, found that in all cases the effect of warping stress was insignificant. The 1997 AASHTO
curved girder specifications places limitations on the span to radius ratio for designing the
bridges as straight. These ratios were found to be conservative by a factor of 2 or more when it
comes to the need to consider warping. (Since concrete box girders will have thicker webs and
soffits, they are even less vulnerable to warping, and it is likely that the effects of warping can be
ignored in almost all of these bridges.)

Page 31
ii. Laboratory Experiments

Most, although not all, laboratory experiments related to curved concrete box girder bridges have
been conducted on small scale plexiglass or metal models of these bridges. A large-scale test of a
concrete structure was performed at the University of California at Berkeley during the 1970’s.
In general, these tests have shown that refined analytical techniques predict structural behavior
quite well. The following paragraphs discuss published papers and reports on these tests in
greater detail.

- Aneja, I. K., and Roll, F., 1971, Model Analysis of Curved Box-Beam Highway Bridge,
Journal of the Structural Division, Vol. 97, No. 12, pp. 2861-2878
Fabrication, preparation and instrumentation of a plexiglass model of a horizontally curved box-
beam highway bridge are described. The model was extensively instrumented with rosette strain
gages at three cross sections. Experimental data for three lane-loading conditions was obtained.
An approximate theoretical analysis of the model was obtained by using finite element method.
It showed that finite element models with curved shell elements provide better predictions than
those with straight plate elements. A typical comparison between the experimental and
theoretical stress distribution across the mid-span gage section for one of the loading conditions
is shown graphically. The comparison shows a good agreement between the shapes but not the
magnitudes of the stress plots obtained experimentally and theoretically. Experimental data at the
three gage sections for each load condition is also given.

- Aslam, M., and Godden, W. G., 1975, Model Studies of Multicell Curved Box-Girder
Bridges, Journal of the Engineering Mechanics Division, Vol. 101, No. 3, pp. 207-222
A model study on the static response of curved box-girder bridges is presented, and a close
agreement is found between the test and analytical results. The prototype bridge was a four-cell
reinforced concrete design, 33 ft 10 in. (10.31 m) wide, 4 ft 10 in. (1.47 m) deep, and had a
radius of curvature of 282 ft (86 m). A 1/29 scale aluminum model was studied for spans of 60
in. (1,500 mm), 45 in. (1,140 mm), and 30 in. (760 mm), with or without a midspan radial
diaphragm. The quantities measured were: (1) Boundary reactions; (2) strains at a radial section
close to midspan; and (3) deflections at selected points. The data were reduced by computer, and
distribution graphs of tangential plate forces, radial-bending moments, and deflections were
plotted by Calcomp plotter. Based on the model data, some general observations are made
regarding the behavior of curved box-girder bridges.

- Fam, A. R.M. and Turkstra C. J., 1976, Model Study of Horizontally Curved Box Girder,
Journal of the Structural Division, Vol. 102, No. 5, pp. 1097-1108
This paper describes an experimental study of a single span horizontally curved plexiglass box-
girder beam with diaphragms and flange overhangs. Static loads were applied at midspan to
cause a complex pattern of membrane and bending stresses with the effects of diaphragms
clearly evident. Experimental results in typical cases are shown graphically and compared to the
results of a special purpose finite element program developed especially for curved box analysis.
Page 32
This program used the softened truss model theory applied to a prestressed concrete multiple cell
box. In this theory, the concrete torsional problem is solved by combining equilibrium and
compatibility conditions and constitutive laws of materials. Until now the theory has been
applied only to the case of pure torsion with a single cell section. An algorithm is presented to
deal with the torsional problem for reinforced concrete and prestressed concrete box girder
bridge superstructures with multiple cell sections. Results are compared with previous theoretical
and experimental work for single cell cases. Good agreement was obtained between
experimental and analytical results.

- Heins, C. P., Bonakdarpour, B. P., and Bell, L. C., 1972, Multicell Curved Girder Model
Studies, Journal of the Structural Division, Vol. 98, No. 4, pp. 831-843
The behavior of a single two-span, three-cell plexiglass model is predicted by the Slope
Deflection Fourier Series Technique. This analytical technique had previously been applied to
only open cross-sectional, I-type, bridge systems. The model was tested under various static
concentrated loads. The resulting experimental deflection, rotation, and strain data for some
loadings are reported. Effects of single and multicell torsional properties are examined. Results
indicate that single cell properties can be applied in the analysis, and warping effects may be
neglected.

- Scordelis, A. C., Elfgren, L. G., and Larsen, P. K., 1977, Ultimate Strength of Curved RC
Box Girder Bridge, Journal of the Structural Division, Vol. 103, No. 8, pp. 1525-1542
Results obtained in a study of a large-scale curved two-span four-cell reinforced concrete box
girder bridge model are presented. The model, which was a 1:2.82 scale replica of a prototype,
had overall plan dimensions of 72 ft (21 m) long by 12 ft (3.7 m) wide. The radius of curvature
was 100 ft (30.5 m). This represents the sharpest curvature normally used for bridges in the
California highway system. Experimental and theoretical results are considered for reactions,
steel and concrete strains, deflections, and moments due to conditioning overloads producing
stress values as high as 2.5 times the nominal design stress. The loading to failure and the
ultimate strength behavior is examined. The excellent live-load overload capacity of the bridge is
evaluated and comparisons are made with the similar behavior of an earlier tested straight bridge
model. Conclusions appropriate for the design of this type of bridge are given.

d. Design Issues

i. Bearings

Although several bearing failures consisting of uplift, overload or binding have been experienced
in curved box girder bridges, no published research exclusively addressing this issue were found.
However, because an accurate 3-D analysis will account for differences in bearing forces and
displacements, several references that deal with global analysis and laboratory experimentation
deal with this issue. (Aslam, et. al., 1975, Scordelis, et. al., 1977, Choudhury, et. al., 1988,
Sennah, et. al., 2002). This issue is also discussed in some textbooks (Menn, 1990).
Page 33

ii. Diaphragms

Diaphragms help prevent excessive distortions of the cross section, facilitate wheel load
distribution, and distribute transverse load. The following two papers discuss research directed
toward determining the number and spacing of interior diaphragms in box girder bridges.

- Oleinik, J. C. and Heins, C. P., 1975, Diaphragms for Curved Box Beam Bridges, Journal of
the Structural Division, Vol. 101, No. 10, pp. 2161-2178
A finite difference procedure is used to determine the response of a single span curved single box
beam bridge with any number of interior diaphragms. The bending and torsional distortions as
well as cross-sectional distortions can then be determined throughout the curved box girder. The
forces that are determined include bending moment and flexural shear, pure torsion, warping
torsion, and bi-moment. These forces, in addition to distortional functions, yield resulting normal
bending, normal warping, and normal distortional stresses. The technique is then used to
determine the dead load and live load response of a series of typical curved box beams. A study
of the data has resulted in a series of empirical design equations.
- Abendroth, R. E., Klaiber, F. W., and Shafer, M. W., 1995, Diaphragm Effectiveness in
Prestressed-Concrete Girder Bridges, Journal of Structural Engineering, Vol. 121, No. 9, pp.
1362-1369
Each year many prestressed-concrete (P/C) girder bridges are damaged by overheight vehicles or
vehicles transporting overheight loads. The effects of this type of loading on P/C bridge behavior
were investigated for various types and locations of intermediate diaphragms. The research
included a comprehensive literature review; a survey of design agencies; the testing of a full-
scale, simple-span, P/C girder-bridge model with eight intermediate diaphragm configurations,
as well as a model without diaphragms; and the finite-element analyses of the bridge model
assuming both pinned- and fixed-end conditions. The vertical load distribution was determined
to be essentially independent of the type and location of the intermediate diaphragms, while the
horizontal load distribution was a function of the intermediate diaphragm type and location.
Construction details at the girder supports produced significant rotational-end restraint for both
vertical and horizontal loading. Both the vertical and horizontal load distributions were found to
be affected by the girder-end restraint. A fabricated intermediate structural steel diaphragm was
determined to provide essentially the same type of response to lateral and vertical loads that was
provided by the reinforced-concrete intermediate diaphragms presently used by the Iowa
Department of Transportation.

iii. Flexure and Flexural shear

Beyond the issue of global analysis, the mechanism for resisting flexural and shear stresses in
box girders is important. The mechanisms of shear resistance and its interaction with flexural
stresses in reinforced and prestressed concrete have been well researched (Marti, 1999 and
Vecchio, et. al., 1986). Also, the effectiveness of the deck and soffit slabs in resisting flexural
Page 34
compressive forces has been studied. This includes the phenomenon commonly known as shear
lag. Several published papers and reports have dealt with these issues. Some of these are
discussed below.

- Chang, S. T., and Zheng, F. Z., 1987, “Negative Shear Lag in Cantilever Box Girder with
Constant Depth,” J. Struct. Eng., 113(1), 20–35.
This paper addresses the classical phenomenon of shear lag in box girders, and draws attention to
distinguishing between positive and negative shear lag. Shear Lag and negative shear lag effect
in cantilever box girders are analyzed through a variation approach and finite element
techniques. Expressions are derived to determine the region of negative shear lag effect with the
interrelation of span/width parameters involved. The theoretical results obtained are compared
with a plexiglass model test. Finally, conclusions are drawn with regard to further study and
research.
Positive shear lag is the phenomenon in which, near the support of a cantilever, flange
longitudinal stresses near the web are larger than away from the web. But for a cantilever box
girder with constant depth under a uniform load, away from the support, the bending stress in the
deck near the webs is smaller than the stress away from the webs. This is a result of negative
shear lag. Using the principle of minimum potential energy, following Reissner’s procedure
with slight modifications, shows that the additional moment created by flange shear deformation
plays an important role in both positive and negative shear lag. For a single point load at the free
end of the cantilever, only positive shear lag is created. When there is a uniformly distributed
load along the full span of the cantilever box girder however, negative shear lag occurs. The
region of the cantilever affected by negative shear lag is from the free end to more than ¾ L from
free end. Negative shear lag affects a larger region than positive shear lag.
With a finite element model analysis, three load cases were considered; a distributed load, a
point load, and a combination of a downward point load and an upward distributed force. This
analysis showed that negative shear lag occurred only with the first load case of a distributed
load. This model was consistent with the results from the minimum potential energy method.
Negative shear lag is dependent not only on the load case but also the boundary conditions. The
ratio of the length of the cantilever to the width of the box girder affects the amount of moment
caused by shear. As the ratio increases, both positive and negative shear lag decreases.
Actual testing using a plexiglass model confirmed the theoretical results. When a uniform load is
applied, not only is positive shear lag more severe compared to a point load, but negative shear
lag is also present. A cross sectional analysis of shear stress in the flange is taken at several
locations. Near the fixed end where shear lag is greatest, the bending stress near the web is much
larger than the stress away from the webs. At a cross section where negative shear lag is
significant, the bending stress away from the webs is greater than the stress near the webs.
This paper is only indirectly applicable to this project because the paper does not deal
specifically with curved girders. However, since shear lag effects are an important consideration
in developing analysis and design strategies, the conclusions in this paper, and the theoretical
solutions are noteworthy. In short, the relevant conclusions are:
Page 35
1. Positive shear lag may occur under both point and uniform load, but negative shear lag
occurs only under uniform load.
2. Negative shear lag is also dependent upon the ratio of L/b, where b is the net width of the
box section. The smaller the ratio, the more severe are the effects of positive and
negative shear lag.
3. Negative shear lag depends upon the boundary condition of displacement as well as on
the external force applied to the girder.
4. In cantilever box girders, although the negative shear lag yield in the region of the
bending stress is small, the relative additional stress induced by this effect is often
considerably greater. It cannot be neglected. It should never be believed that in all cases
only positive shear lag is produced.

- Chang, S. T., and Gang, J. Z. ,1990, “Analysis of Cantilever Decks of Thin-Walled Box
Girder Bridges,” J. Struct. Eng., 116(9), 2410 –2418.
This paper addresses the cantilever decks (“wings”) of single-cell box girder bridges, and does
not make any distinction about the effects of horizontal curves. But it does present some useful
qualitative information about cantilever deck evaluation, in general.
The paper reports on a spline finite strip approach used to analyze the cantilever decks. Effects of
distortion of thin-walled box sections are taken into account by treating the cantilever deck as a
slab with horizontally distributed spring supports along the cantilever root. Perspex model tests
were conducted in the model structural laboratory at Tong Ji University. The results based on the
spline finite strip method are compared with those of the model test. Simplified solutions are also
given for the distribution of transverse moment along the cantilever root.
A plexiglass model of a single-cell box beam was evaluated. As a point load moved transversely
across the box girder, the bending stress and membrane stress at the root of the overhang of the
deck were obtained. From this analysis, it was observed that it is reasonable to treat the
cantilever decks as cantilever slabs with horizontally distributed spring support along the
cantilever root with the spring constant K depending on dimensions and material properties of
the deck.
The spline finite strip analysis was shown to be in close agreement with actual model test results.
Tests showed that the moment along the cantilever root approaches zero as the longitudinal
distance from the point load increases. The maximum moment at the root is when the point load
is at the end of the cantilever and approximately zero when the load is not on the cantilever.
There is also sagging in the cantilever around the point load. Although sagging moment is only
local, many point loads at the same longitudinal location can cause significant sagging moment,
so it should not be ignored.
Conclusions made in this paper are as follows:
1. Cantilever decks of thin-walled box girder bridges can be treated as cantilever slabs with
horizontally distributed spring support along the cantilever root, taking into account the
influence of local distortion of the box section.
2. Spline finite strip results, based on the simplified idealization of the cantilever decks of
thin-walled box girder bridges, are in close agreement with test results. The sagging
Page 36
moment at the cantilever root can be obtained in conjunction with information tabulated
in the article.
3. In cantilever slabs of infinite length and large cantilever length, sagging moment cannot
be ignored. The sagging moment may be taken from information also provided in the
article.
Other than the general information provided for evaluation of cantilever decks of straight box-
girder bridges, there is no specific information pertaining to curved box-girders.

- Hasebe, K., Usuki, S., and Horie, Y., 1985, “Shear Lag Analysis and Effective Width of
Curved Girder Bridges,” J. Eng. Mech., 111(1), 87–92.
This paper develops and guidelines and graphs for estimating the effective width of curved girder
bridges. The methodology is formulated by substituting the flange stress derived from present
theory into the equation of effective width definition for the curved girders. The required
information in formulating the effective width rule for design of curved girder bridges is
provided. The actual longitudinal stress distributions for the curved girders are evolved from the
present theory for shear lag in order to determine the effective width. The thin-walled curved
girders used in this investigation are based on box and channel cross sections, and are analyzed
for a uniform lateral load and for a concentrated load.
Numerical examples are shown for several problems to investigate the effect on effective width
of curved girder bridges. The values of the effective width obtained by the present theory are
compared with those of the straight girder bridges. According to the results, it is OK to say the
values of effective width of curved girder bridges are the same (approximately) as the values of
the straight girder bridges. This is a very important conclusion for the NCHRP 12-71 project.
The inner and outer effective widths are denoted as
1
and
2
, respectively, for a curved girder.
0
is half of the effective width for a straight beam and 2b = width of the flange. The term
effective width ratio is defined as the ratio of an effective width to the actual breadth of the
flange. The parametric study involves calculating the effective width ratio for simply supported
girders and comparing results of; a) point load at mid span vs. uniformly distributed load, b)
present theory vs. folded plate theory, c) inner and outer effective width ratio for curved girders
vs. effective width ratio for straight girders, and d) box cross section vs. channel cross section.
In the test, the curved beam has a radius of curvature R = 4L and b/L = 0.1. In the case of a
concentrated loading, the values of the effective width ratio are minimum at the center of the
span length and increase rapidly towards the supports. However, in the case of a uniformly
distributed loading, the values are maximum at the mid span and decrease towards the supports.
The values of effective width ratio of the present theory are lower than the folded plate theory
curves, and the inside and outside effective width ratio agree with the values of the straight
beams, irrespective of cross-sectional shapes or types of load distributions.
The authors analyze the inner and outer effective widths at mid span relative to the curvature
R/L. For a box girder under a distributed load, the effective widths remain fairly constant for b/L
= 0.1 when R/L > 2 and only decreases slightly for b/L=0.2. With a point load and b/L = 0.1, the
effective width also remains fairly constant for R/L > 4. When b/L = 0.2 with a point load
however, the effective widths deviate significantly for small R/L.
Page 37
For a curved girder, as b/L increases, the difference between the inner and outer effective widths
becomes significant, especially for box girders. For small values of b/L however, those two
values are almost identical.
One needs to be careful when R/L is small, b/h is small, or b/L is large, especially under
concentrated loads. Otherwise, it is reasonable to assume the effective width of a curved girder
is equal to that of a straight girder for practical applications.

iv. Torsion

Torsion design is currently addressed by the AASHTO LRFD code and in the case of box girders
is integrated with shear design. Recent studies (Rahal, et. al., 2003) indicate that this approach
yields good results when the correct value of u is used. A number of other papers on the subject
of torsional resistance have been reviewed and it is concluded that the current design methods
are acceptable. Only minor clarifications of the current specifications and guidelines for applying
them to box girders are required. Following is a summary of the papers that were reviewed on
torsion design.

- Collins, M., and Mitchell, D., 1980, Shear and Torsion Design of Prestressed and Non-
Prestressed Concrete Beams, Journal of the Prestressed Concrete Institute, V. 25, No. 5
This document presents design proposals for flexural shear and torsion of prestressed and non-
prestressed concrete beams based on the compression field theory. A rational method is proposed
which addresses members subject to flexural shear, torsion, combined flexural shear and flexure,
and combined torsion, flexural shear and flexure. Early design procedures using truss models are
also presented. The compression field theory, which is a development of the traditional truss
model for flexural shear and torsion, considers in addition to the truss equilibrium conditions,
geometric compatibility conditions and material stress-strain relationships. The compression
field theory is capable of predicting the failure load as well as the complete load-deformation
response. Measured and predicted response of numerous members is presented. The proposed
design recommendations are provided in code format. Comparisons to the provisions of the ACI
318-77 and CEB codes are provided. Numerous worked examples are provided that demonstrate
the proposed method.

- Rahal, K. N., and Collins, M. P., 1996, Simple Model for Predicting Torsional Strength of
Reinforced and Prestressed Concrete Sections, ACI Structural Journal, Vol. 93, No. 6, pp 658-
666.
A noniterative method for calculating the ultimate torsional strength and the corresponding
deformations of reinforced and concentrically prestressed concrete sections is presented. This
method is based on the truss model. It avoids the need for iterations by making simplifying
assumptions about the thickness of the concrete diagonal, the softening of the concrete due to
diagonal cracking, and the principal compressive strains at ultimate conditions. A simple check
on the spalling of the concrete cover is implemented. The calculated torsional capacities of 86
beams are compared with the experimental results and very good agreement is obtained.
Page 38

- Rahal, K. N., and Collins, M. P., 2003, Experimental Evaluation of ACI and AASHTO-LRFD
Design Provisions for Combined Shear and Torsion, ACI Structural Journal, Vol. 100, No. 3, pp
277-282.
The experimental results from four large nonprestressed specimens loaded in combined flexural
shear and torsion are used to evaluate the torsion design procedures of ACI 318-02 and
AASHTO-LRFD. Both sets of procedures calculate the required amounts of hoop reinforcement
for torsion based on a space truss model with compression diagonals inclined at an angle of u to
the longitudinal axis. It is shown that the ACI provisions give very conservative results if the
recommended value of 45 degrees is used for u. If the lower limit of 30 degrees is used,
however, some unconservative results are possible. The AASHTO-LRFD provisions predicted
values of u of approximately 36 degrees for these specimens and gave accurate estimates of the
strengths.

- Fu, Chung C, and Yang, Dailli, 1996, Designs of Concrete Bridges with Multiple Box Cells
Due to Torsion Using Softened Truss Model, ACI Structural Journal, Vol. 93, No. 6, pp 696-702.
Using a softened truss model, this paper presents a method for torsional design of multicell
concrete bridges. Earlier researchers have successfully shown the development for solving single
cell torsion by combining the equalibrium, compatibility, and the softened constitutive laws of
concrete. By solving the simultaneous equations based on the membrane analogy,
multicontinuous or separate cells can be solved.

- Hsu, T.T.C., 1997, ACI Shear and Torsion Provisions for Prestressed Hollow Girders, ACI
Structural Journal, Vol 94, No. 6, pp 787-799.
New torsion design provisions have been proposed for the 1995 ACI Building Code. As
compared to the 1989 provisions, these generalized 1995 provisions have three advantages: First,
they are applicable to closed cross sections of arbitrary shapes. Second, they are applicable to
prestressed concrete. Third, they are considerably simplified by deleting the “torsional concrete
contribution” and its interaction with flexural shear. These new provisions are suitable for
application to concrete guideways and bridges, because these large structures are always
prestressed and are often chosen to have hollow box sections of various shapes. This paper
discussed the background of the new code provisions, suggests modifications to code formulas,
and illustrates the application of the code provisions to prestressed hollow girders by way of a
guideway example.

Page 39
v. Wheel Load Distribution

Wheel load distribution has been the subject of many of the analytical studies cited earlier. Other
studies are discussed in the following paragraphs.

- Song, S.T., Y.H. Chai, and S.E. Hida, 2001, Live Load Distribution in Multi-Cell Box-Girder
Bridges and its Comparison with the AASHTO LRFD Bridge Design Specifications, Final
Report to Caltrans for Contract Number 59A0148, July 2001
This report presents the results of a series of analyses done on box girder bridges with normal
and skewed supports and straight and curved geometry. The analysis models included 3-D finite
element as well as grillage, and single line of elements for superstructure. The goal of the project
was to evaluate the LRFD live load distribution factors for box girder bridges. However, the
modeling techniques used and verified for this project can be used as guidelines for NCHRP 12-
71.

- Zokaie, T., K.D. Mish, and R.A. Imbsen, 1993, Distribution of Wheel Loads on Highway
Bridges, Phase 3, Final Report to NCHRP 12-26 (2), December 1993
This report presents a computer program (LDFac) that was developed for modeling bridge
superstructure with straight or skewed supports and obtaining live load distribution factors.
Although this computer program did not consider curved geometry specifically, the modeling
process and load placement guidelines may be used for analysis of curved bridges as well. One
of the key issues discussed in this report is the modeling of distortion of box girders in a grillage
analysis via an equivalent shear deformation parameter.

- Zokaie, T., T.A. Osterkamp, and R.A. Imbsen, 1991, Distribution of Wheel Loads on
Highway Bridges, Final Report to NCHRP 12-26 (1), 1991
This report presents a series guidelines for analysis of various bridge types. The guidelines
include calculation of equivalent section property parameters to be used in plate and grillage
analyses, as well as guidelines for setting the boundary conditions. Although the research did not
specifically consider curved bridge geometry, many of the guidelines for modeling and analysis
using common analysis tools is applicable to modeling that will be needed in NCHRP 12-71
global analysis studies.

vi. Tendon Breakout and Deviation Saddles

Prestress tendon breakout in curved bridges has occurred on a number of bridges over the years.
It is evident from observing the reasons for these failures, that they can be prevented through
close attention to details such as tendon spacing and tie back reinforcement. The following
paragraphs summarize the references reviewed on the subject.

Page 40
- Beaupre, R.J., Powell, L.C., Breen, J.E., and Kreger, M.E., 1988, Deviation Saddle Behavior
and Design for Externally Post-Tensioned Bridges, Research Report 365-2. Center for
Transportation Research, University of Texas at Austin, Austin, Texas.
This report is the second in a series outlining a major study of the behavior of post-tensioned
concrete box girder bridges with post-tensioning tendons external to the concrete section. It
presents the results of an experimental program in which ten very accurately sealed reinforced
concrete models of typical tendon deviators were tested. Detailed instrumentation led to a very
good understanding of the behavior of the various patterns of reinforcement in the deviators. The
models included two very different patterns of detailing, several arrangements of tendon
inclinations, and both normal and epoxy coated reinforcement.
The report evaluates the results with respect to both simplified conventional analysis methods
and strut-and-tie models. The results provide the basis for deviator design recommendations and
several examples are presented to illustrate the practical use of these recommendations.
- Caltrans, 1996, Bridge Memo to Designers Manual, Memo 11-31 Curved Post-Tensioned
Bridges, California Department of Transportation, Sacramento, California
Memo 11-31 addresses the design of curved post-tensioned concrete box girders for lateral
prestress forces. The force effects considered are tendon confinement and web regional
transverse bending. The lateral prestress force, F, is determined by dividing the jacking force (P
j
)
per girder by the horizontal radius (R) of the web. A standard detail for tendon confinement (see
Figure 3-1) is required for all webs with a P
j
/ R > 100 kN per m or a horizontal radius (R) of 250
m or less. The regional transverse bending moment in the web is taken as M
u
= 0.20 F h
c
where
h
c
is the clear distance between the top and bottom slabs. This assumes the web to act as a simple
beam spanning between the top and bottom slabs with a concentrated load, F, acting at mid-
height of the web. The resulting simple beam moment is reduced 20% for continuity. The load
factor is taken as 1.0. The design of stirrup reinforcement does not combine regional transverse
bending and shear requirements. Graphs are provided to check webs for containment of tendons
and adequate stirrup reinforcement to resist regional transverse bending.

Figure 3-1 - Caltrans Detail A

A review of the 405/55 failure has led to the identification of several issues related to the
Caltrans Memo to Designers 11-31. These are discussed below.
Page 41



Page 42

- Cordtz, K., 2004, Design of Curved Post-Tensioned Bridges for Lateral Prestress Forces,
David Evans and Associates, Inc., Roseville, California
This document presents the internal guidelines of David Evans and Associates, Inc. for the
design of horizontally curved post-tensioned concrete box girders for lateral prestress forces. The
primary focus is on the regional beam action of the webs and local slab action of the cover
concrete over the tendons. The document provides a discussion of the actions on curved post-
tensioned girders, identifies those actions not completely addressed in current design codes and
guidelines, and recommends design procedures that reflect current best practice.
Local slab action of the concrete cover over the tendons has been identified as the major cause of
failure in several curved post-tensioned bridges that did not have duct or web ties. For a web
without duct or web ties, the cover concrete is the only element restraining the lateral prestress
force. The cover concrete acts as a plain concrete beam to restrain the lateral prestress force. The
local slab is subject to lateral shear and bending from the lateral prestress force. Specific
requirements for local lateral shear are given in AASHTO LRFD 5.10.4.3.1 “In-Plane Force
Effects”. No specific design methodology for local flexure is given by AASHTO. The document
provides interim recommendations for the tensile stresses in the cover concrete. Where duct ties
are required, the document recommends the use of a rational method for design such as a strut
and tie model.
The vertical reinforcement in the web is subjected to combined global shear and regional
transverse bending due to regional beam action. No specific design methodology for these
combined actions is given by AASHTO. The document provides interim recommendations for
the design of the stirrups.
A flowchart is presented that outlines the recommended procedures. Numerous worked examples
are also provided.

- Podolny, W., Jr., 1985, The Cause of Cracking in Post-Tensioned Concrete Box Girder
Bridges and Retrofit Procedures, Journal of the Prestressed Concrete Institute, March-April 1985
This article discusses the types of problems that lead to cracking in post-tensioned concrete box
girder bridges and have been encountered in both Europe and the United States. Cracking is
attributed in a broad sense to the following factors: inadequate flexural and shear capacity, non-
consideration of thermal stresses, insufficient attention to stresses developed by curvature of
tendons, improper or inappropriate construction techniques, lack of quality workmanship to meet
the tolerances necessary for problem free structures, and understrength materials. It is noted that
in general the cause of cracking can be attributed to the superposition of stresses of multiple
effects.
The article discusses the pullout of horizontal curved tendons that occurred on several cast-in-
place post-tensioned concrete box girder bridges. In two of these structures, there was a
combination of relatively sharp horizontal curvature; thin concrete cover over the tendons, and
the bundling of a number of large sized tendons close together. Podolny divides the analysis of
the failures of these bridges in three separate actions:
Page 43

1. The global or overall girder action of the bridge together with it’s supporting piers and
abutments.
2. Regional beam action of each web supported at the top and bottom flanges as a beam.
3. Local slab action of the concrete cover over the tendons.
It appears that for both of these bridges local slab action of the concrete cover over the tendons
was the primary cause of the failure, but the regional beam action could have been a contributory
cause, and could by itself have overstressed some of the stirrups. The global action had a very
small effect on these failures.

- Seible, F., Dameron, R., and Hansen, B., 2003, Structural Evaluation of the 405-55 HOV
Connector and the Curved Girder Cracking/Spalling Problems. StD&A, San Diego, California.
This document is a detailed (70 pages, including illustrations) project report on results of
structural evaluation of the 405-55 HOV Connector’s curved girder cracking/spalling problems.
The report provides background on the observed cracking and spalling (caused by horizontal
breakout of web tendons in the multi-cell boxes), summary of occurrences of similar problems at
Las Lomas (California) and Kapiolani (Hawaii), and then documents a step-by-step analysis of
the 405-55 bridge. The steps followed include global, regional, and local analysis, as suggested
by Podolny. The regional and local analysis utilize detailed finite element modeling with
cracking concrete constitutive models, but various hand calculation methods for evaluating
regional and local effects are also described and demonstrated as a check on the detailed
analysis. The report also provides in depth summary of Caltrans’s Memo to Designers (MTD)
11-31, and points to some potential shortcomings (under the right set of unusual circumstances)
of the Memo. The report then lists eight contributory causes for the damage that occurred, but
concludes that the one cause which, if corrected would have prevented the damage, was the
omission of duct ties from the design.
While this report is very structure-specific and contains detailed project information that should
not be directly referenced in a set of design Specifications, it does provide a useful reference for
summarizing the various analyses (both hand calculation and finite element) appropriate to this
class of problems, especially at the “regional” and “local” level for girder cross-section and web
evaluation.

- Strasky, J., 2001, Influence of Prestressing in Curved Members, Betonve Mosty, Report
TK21, Prauge, Czech Republic.
The influence of prestressing in curved members is discussed. To mitigate the effect of radial
prestressing forces in the webs of box girders, it is recommended that prestress tendons be
separated vertically where they pass near the middle of the web. This will spread out the radial
forces and result in lower stresses tending to rip the strand out of the side of the web. Care is also
required for tendons in the soffit of segmentally constructed bridges (straight or curved) where
tendons are anchored in blisters along the length of the bridge and deviated across the width of
the soffit resulting in transverse tensile stresses that can fail the soffit. Such a failure occurred in
Page 44
a bridge constructed with a gantry in Austria. The vertical curvature of these tendons in a
haunched bridge can also present a problem.

Van Landuyt, D., and Breen, J., 1997, Tendon Breakout Failures in Bridges, Concrete
International, American Concrete Institute, Farmington Hills, MI, November 1997
The article discusses the pullout failure of horizontal curved tendons that occurred on several
cast-in-place post-tensioned concrete box girder bridges. In both of these structures, there was a
combination of relatively sharp horizontal curvature, thin concrete cover over the tendons, and
the bundling of a number of large sized tendons close together. It appears that for both of these
bridges local slab action of the concrete cover over the tendons was the primary cause of the
failure.
The article discusses the theory of transverse stresses in curved box girder cross sections due to
post-tensioning including “distributed radial force arch action.” Horizontal curved post-tensioned
bridges are subjected to the following three separate actions:
1. The global or overall girder action of the bridge together with it’s supporting piers and
abutments.
2. Regional beam action of each web supported at the top and bottom flanges as a beam.
3. Local slab action of the concrete cover over the tendons.
The article discusses the current design philosophy of the California Department of
Transportation, the Texas Department of Transportation, the AASHTO Guide Specifications for
the Design and Construction of Segmental Concrete Bridges, and the AASHTO Load and
Resistance Factor Design Bridge Design Specifications for tendon confinement.
A test program was carried out by the authors on webs without tendon confinement
reinforcement. Both closely and widely spaced ducts (duct spacing less than or greater than one
duct diameter respectively) were tested and different lateral shear failure modes were observed.
Recommendations for lateral shear capacity are proposed that are more conservative than the
AASHTO LRFD Specifications and yielded a consistently narrow range of factors of safety for
the four test specimens (1.99 to 2.34). A recommendation for a design methodology for the
flexure of the concrete cover is not proposed due to the lack of understanding of its behavior.
The test specimens did not fail in regional transverse bending of the web due to the formation a
second load path after formation of flexural cracks in the web. The load path is envisioned
primarily as a vertical one with the lateral prestress load carried through flexural bending of the
web until cracking. Once cracking occurred, the stiffness was reduced and the load was carried
primarily through longitudinal arching until a local lateral shearing failure occurred. A
University of Texas theses (Van Landuyt, 1991) that discusses this research in greater detail was
also reviewed.

Page 45
vii. Time Dependency

The redistribution of stresses due to creep and shrinkage may be important in curved concrete
bridges. This issue is discussed in at least one of the references previously described (Menn,
1990). A paper further exploring this issue is described below.

- Zhang, L., Liu, M. and Huang, L., 1993, Time-Dependent Analysis of Nonprismatic Curved
PC Box Girder Bridges, Conference Proceeding Paper, Computing in Civil and Building
Engineering, pp. 1703-1710
This proceeding consists of papers presented at the Fifth International Conference on Computing
in Civil and Building Engineering held in Anaheim, California, June 7-9, 1993. The proceedings
cover five major areas of concern: 1) Computing in construction, 2) geographic information
systems, 3) expert systems and artificial intelligence, 4) computing in structures, and 5)
computing in transportation. Within these broad topics, subjects such as: 1) Computer analysis of
cable-stayed bridges; 2) artificial intelligence in highway CAD systems; 4) automated systems
for construction bidding; 5) effect of automation on construction changes; 6) optimal seismic
design of structures; and 7) CAD instruction for civil engineering students. The book also
presents several papers discussing different aspects of multimedia information systems and
geographic information systems.

viii. Vehicular Impact

Vehicular impact in curved bridges is different than in straight bridges. A paper addressing this
subject is discussed below.

- Rabizadeh, R. O., and Shore, S., 1975, Dynamic Analysis of Curved Box-Girder Bridges,
Journal of the Structural Division, Vol. 101, No. 9, pp. 1899-1912
The finite element technique is used for the forced vibration analysis of horizontally curved box
girder bridges. Annular plates and cylindrical shell elements are used to discretize the slab,
bottom flanges and webs. Rectangular plate elements and pin-jointed bar elements are used for
diaphragms discretization. A vehicle, as the applied time varying forcing function, is simulated
by two sets of concentrated forces, having component in the radial and transverse directions and
moving with constant angular velocities on circumferential paths of the bridge. The effect of
centrifugal forces is considered and the effect of damping of the bridge is neglected in the
analysis. The mass condensation technique is used to reduce the number of coupled differential
equations obtained from the finite element method. The resulting differential equations are
solved by the linear acceleration method. A number of bridges with practical geometries are
analyzed and impact factors are calculated.

Page 46
ix. Seismic Response

Several papers and reports have addressed the seismic response of curved box girder bridges.
Most of these deal with substructure response and are beyond the scope of this study. However,
at least one presentation (no paper available) has implications for superstructure design. This
study is described below.

- Ibrahim, A. M. M., Soin, P., Masroor, T., and Quiogue, J., 2005, Torsional Analysis and
Design of Curved Bridges with Single Columns - LFD vs. LRFD Approach, Paper presented at
the Western Bridge Engineers Conference, Portland, OR.
This unpublished study compared torsional superstructure design of a curved concrete box girder
bridge subjected to seismic loading using the Load Factor Design approach currently used by
Caltrans and the AASHTO LRFD method. Torsion is induced in the superstructure not only by
curvature, but also by column plastic hinging of the single column bent during an earthquake.
Caltrans practice is to design the superstructure in bridges with monolithic columns to remain
elastic. In this case the limitation on seismically induced torsional forces resulting from column
yielding causes a redistribution of superstructure forces in the box girder.

x. Design Optimization

Several combinations of slab and web width can be selected to resist the applied loads. Although
design optimization is generally not the subject of design specifications, at least one paper
reviewed addressed this issue.

- Ozakea, M., and Tavsi, N., 2003, Analysis and Shape Optimization of Variable Thickness
Box Girder Bridges in Curved Platform, Electronic Journal of Structural Engineering
International, 3(2003), Queensland, Australia.
This paper deals with the development of reliable and efficient computational tools to analyze
and find optimum shapes of box girder bridges in curved planform in which the strain energy or
the weight of the structure is minimized subject to certain constrains. The finite strip method is
used to determine the stresses and displacements based on Mindlin-Reissner shell theory. An
automated analysis and optimization procedure is adopted which integrates finite strip analysis,
parametric cubic spline geometry definition, automatic mesh generation, sensitivity analysis and
mathematical programming methods. It is concluded that the finite strip method offers an
accurate and inexpensive tool for the optimization of box girder bridges having regular
prismatic-type geometry with diaphragm ends and in curved planform.

Page 47
xi. Detailing

The detailing of prestressed concrete in bridges is addressed by several agencies that commonly
use this structural form. A general reference published by VSL, a large prestresser with
international experience, is discussed below.

- Rogowsky, D.M. and Marti, P., 1991, Detailing for Post-Tensioning, VSL International Ltd.,
Bern, Switzerland
Detailing for Post-Tensioning includes discussions and examples demonstrating the forces that
are produced by post-tensioning, in particular, those in anchorage zones and regions of tendon
curvature. Emphasis is placed on the use of strut-and-tie models to determine the tensile
reinforcement requirements. Article 4.4 “Tendon Curvature Effects” deals with special issues
associated with curved tendons including: in-plane deviation forces, out-of-plane bundle
flattening forces, minimum radius requirements, and minimum tangent length requirements. The
radial force generated by a curved tendon is given as P/R where P is the tendon force and R is the
radius of curvature of the tendon. Methods for preventing tendon breakout in thin curved webs
include adequate lateral shear capacity of the concrete cover (adequate cover) or providing
tieback reinforcement.

SUMMARY:

A considerable body of research has been conducted on box girder bridges. Much of this is
useful to this project.
With respect to global response analysis, research can be broadly divided between steel and
concrete bridge. Concrete bridges have been found to be stiff enough so that torsion warping can
generally be ignored. Sophisticated elastic analysis techniques such as finite element methods
have been shown to produce excellent results that compare well with physical testing. It is
therefore not necessary to do any more sophisticated research on this subject. It is necessary for
our project to explore the accuracy of less sophisticated methods such as grillage analysis. If
grillage analysis methods can be shown to produce reliable results, than they can be used both in
design, and as a verification tool for even less sophisticated analysis methods. The goal is to
identify the simplest methods that can be safely used.
It also seems that there are several potential configurations of curved box girder bridges that
need further study from the designer’s point of view. Although some research work has been
performed on skewed bridges, bearings and interior diaphragms, most of it has not found its way
into design specifications. Part of our goal is to develop design procedures to handle these issues.
Conventional reinforced and prestress concrete design methods can be used for curved concrete
box girder design provided accurate global demands can be established. Considerable work has
been performed over the years in these areas. Torsion design, particularly as it applies to box
girders is well established and further refinement of these methods is beyond the scope of this
project.
Page 48
The local behavior of prestess tendons in curved concrete box girder bridges is an issue to be
addressed by this project. Although excellent research has been conducted at the University of
Texas )Van Landuyt, 1991) this needs to be studied further using available analytical techniques.
Page 49

4. GLOBAL RESPONSE ANALYSIS STUDIES

This chapter is a summary of the work performed for global response analysis and the results and
conclusions obtained from this study. Additional detailed results are presented in Appendix E.

a Objective:

The global behavior of a curved bridge can be distinctly different from a straight bridge. The
curvature results in off-center placement of loads and subsequently such loads induce torsion
into the superstructure. The torsion in turn causes the shear stresses on the outside of the curve to
increase. Also, the curved geometry of the bridge will result in development of transverse
moments, which can cause an increase the normal stresses on the outside edges of the bridge, and
can result in higher tension and/or compression stresses. Post-tensioned bridges also have an
additional equivalent transverse load, which can result in significant tension on the inside of the
curve and compression on the outside edge. The magnitudes of such effects depend on radius of
curvature, span configuration, cross sectional geometry and load patterns among other
parameters. The structural analysis required to capture such effects, in most cases, is beyond the
scope of day-to-day normal bridge design activities.

The objective of the global analysis study in this research is to quantify the effect of increased
shear force and normal stresses in the cross section, identify common trends, and find
approximate modeling methods to obtain accurate results for design purposes. In particular, shear
forces and normal stresses due to dead loads, live loads, and post-tensioning are studied to obtain
analysis modeling limitations and develop empirical adjustment factors for simplified analysis. A
set of special studies is also performed to review the effect of diaphragms, pier connection
(bearing vs. monolithic), and skewed abutments on curved bridges.

b Model Verification:

Three-dimensional finite element analysis using plate and shell elements is accepted as the most
accurate level of analysis available for box girder bridges. However, the magnitude of analysis
cases desirable for parametric studies in this project is such that a more simplified model is
desirable. Given that the parametric studies are based on bridges with radial supports, available
guidelines for grillage analysis were used. In order to make sure that the grillage and finite
element models produced similar results, a detailed model of a three span bridge on a tight curve
(400 ft radius) was created and results were compared for various load effects. The models are
shown in Figures 4-1 and 4-2. The results for superstructure dead load and a concentrated
midspan load obtained from the two models (grillage and finite element) for a two-cell box are
compared in Table 4-1. These results indicate a very close comparison. As a result, grillage
models were used throughout the rest of the study as the basis of comparisons and such results
are deemed to be accurate for all practical purposes. Guidelines for performing an analysis with a
grillage analogy are included in Appendix C. These guidelines may be used for design purposes
when the bridge configuration requires it.

Page 50


Figure 4-1 – Finite Element Model of the Bridge used for Model Verification


Figure 4-2 – Grillage Model of the Bridge used for Model Verification




Page 51
Table 4-1 – Comparison of Results, Grillage vs. FEM – Two Cell Box

Action Location
Grillage FEM Grillage/FEM Grillage FEM Grillage/FEM
Left -4.40 -4.31 1.02 -0.23 -0.26 0.91
Right -3.87 -3.80 1.02 -0.25 -0.23 1.10
Gross 52103 53523 0.97 3791 3829 0.99
Left Girder 14016 16290 0.86 1197 1190 1.01
Center Girder 21523 21735 0.99 1566 1504 1.04
Right Girder 16565 15499 1.07 1028 1135 0.91
Gross -91261 -94623 0.96 -3095 -3235 0.96
Left Girder -27469 -27309 1.01 -1007 -845 1.19
Center Girder -38265 -39494 0.97 -1342 -1526 0.88
Right Girder -25528 -27820 0.92 -746 -864 0.86
Midspan Span 2 Deflections
(inches)
Midspan Span 2 Moments
(ft. kips)
Negative Bent 3 Moments
(ft. kips)
DL LL2C


c Parameter Studies

Analysis Cases – In order to study the effect of various bridge parameters on the response of
curved bridges, a parametric study was performed which was focused on the variation of span
configuration and length, bridge cross-section geometry, and loading. The study included the
following cases:

Four bridge cross sections shapes were considered:

- Single-cell box based on a typical cast-in-place cross section
- Single-cell box based on a typical pre-cast cross section
- Two-cell box based on a typical cast-in-place cross section
- Five-cell box based on a typical cast-in-place cross section

The typical cross sections are shown in figures 4-3, 4-5, 4-7 and 4-9. Note that the section
geometry remains the same for various bridge span lengths with the exception of the overall
section depth. Also, in order to simplify the analysis cases and automate the model generation,
without jeopardizing the global behavior of the bridge, the cross section was idealized as shown
in Figures 4-4, 4-6, 4-8 and 4-10. The cross section properties used for the analytical models are
shown in Tables 4-2 through 4-5



Page 52


Figure 4-3 – Typical Single-Cell Cast-in-Place Cross Section




Figure 4-4 – Idealized Single-Cell Cast-in-Place Cross Section
Page 53


Figure 4-5 – Typical Single-Cell Pre-cast Cross Section




Figure 4-6 – Idealized Single-Cell Pre-cast Cross Section







Page 54


Figure 4-7 – Typical Two-Cell Cast-In-Place Cross Section




Figure 4-8 – Idealized Two-Cell Cast-In-Place Cross Section








Page 55



Figure 4-9 – Typical Five-Cell Cast-In-Place Cross Section




Figure 4-10 – Idealized Five-Cell Cast-In-Place Cross Section


For each cross-section type, five typical bridge models were considered:

- Single Span – 200 ft long
- Single Span – 100 ft long
- Three Span – 200’-300’-200’
- Three Span – 150’-200’-150’
- Three Span – 75’-100’-75’

Each of the three span bridges was analyzed with two types of piers to identify the effect of
softer vs. stiffer transverse pier stiffness. The stiffer columns are 6’x18’ which can also be
considered similar in behavior to a multi-column pier. The softer columns are 6’x6’, 7’x7’ and
8’x8’ for the short, medium and long bridges respectively. All column heights were assumed to
be 50ft to the point of fixity at the base. It was assumed that the pier and abutment diaphragms
are relatively stiff, i.e., they have a moment of inertia of 5000 ft
4
.

Each of the above 32 bridges was configured as a straight bridge and as curved bridges with radii
of 200’, 400’, 600’, 800’, and 1000’, resulting in 192 bridge configurations.

Structural Analysis – Each bridge configuration was modeled as a spine model (in which one
line of elements was used for superstructure, located along the centerline of the bridge), and also
Page 56
as a grillage analogy (grid) model (where each web line was modeled as a line of elements along
its centerline and transverse elements along radial lines were used to connect them in the
transverse direction). Typical spine and grillage modeling techniques are shown in figures 4-11
and 4-12. Note that each model uses several elements per span in the longitudinal direction of the
bridge. Each span element is limited to a central angle of 3.5 degrees as recommended in
Appendix C.


Figure 4-11 – Typical Curved Line (Spine Beam) Bridge Model (showing 3-span unit)



Figure 4-12 – Typical Curved Grillage Bridge Model (showing 3-span unit with 3 webs)



Page 57
The section properties for the spine model were based on the entire section. An example of these
for the two-cell cross section is shown in Table 4-2.

Table 4-2 – Section Properties for Two-cell Curved Line (Spine) Bridge Model


200 ft. Single
Span
100 ft. Single
Span
200/300/200
ft. Multi-Span
150/200/150
ft. Multi-Span
75/100/75 ft.
Multi-Span
CG (ft) 4.53 2.24 5.46 4.06 2.01
Area (ft2) 81.73 66.73 87.73 78.73 65.23
Avy (ft2) 56.60 56.60 56.60 56.60 56.60
Avz (ft2) 30.00 15.00 36.00 27.00 13.50
Iyy (ft4) 1,326.25 256.23 2,028.29 1,037.00 197.80
Izz (ft4) 9,544.20 7,488.61 10,366.44 9,133.08 7283.05
Jxx (ft4) 1,559.82 365.08 2,213.30 1,267.24 286.74
Depth (ft) 10.0 5.0 12.0 9.0 4.5


The section properties for the grillage model included a set of properties for the exterior girder, a
set of properties for the interior girder, and a set of properties for the transverse element.
Furthermore, the transverse element shear area is calculated via a special formula to account for
the warping on the cells. An example of these properties for the two-cell section is shown in
Table 4-3. Note that the transverse element properties may be different for each element based
on its width and therefore is shown here for a unit width. Note that the members on the outside of
the curve are longer than the ones on the inside. Likewise, the widths of transverse members
along the outside of the curve are larger than those of the members along the inside. The actual
member property used in analysis was calculated by multiplying the values shown here by the
average width of each element, except that “Izz” is multiplied by L
3
.
Page 58


Table 4-3 – Section Properties for Two-cell Grillage Bridge Model
EXTERIOR
GIRDER
200 ft.
Single
Span
100 ft.
Single
Span
200/300/200 ft.
Multi-Span
150/200/150 ft.
Multi-Span
75/100/75 ft.
Multi-Span
CG (ft) 4.23 2.06 5.13 3.78 1.85
Area (ft2) 25.03 20.03 27.03 24.03 19.53
Avy (ft2) 18.87 18.87 18.87 18.87 18.87
Avz (ft2) 10.00 5.00 12.00 9.00 4.50
Iyy (ft4) 386.50 73.28 594.75 301.19 56.45
Izz (ft4) 216.76 208.19 219.34 215.31 207.11
Jxx (ft4) 519.94 121.69 737.77 422.41 95.58
INTERIOR
GIRDER
200 ft.
Single
Span
100 ft.
Single
Span
200/300/200 ft.
Multi-Span
150/200/150 ft.
Multi-Span
75/100/75 ft.
Multi-Span
CG (ft) 5 2.50 6.00 4.50 2.25
Area (ft2) 31.67 26.67 33.67 30.67 26.17
Avy (ft2) 18.87 18.87 18.87 18.87 18.87
Avz (ft2) 10.00 5.00 12.00 9.00 4.50
Iyy (ft4) 541.74 106.59 823.13 425.04 82.44
Izz (ft4) 399.43 399.02 399.60 399.35 398.97
Jxx (ft4) 519.94 121.69 737.77 422.41 95.58
Transverse
Element
200 ft.
Single
Span
100 ft.
Single
Span
200/300/200 ft.
Multi-Span
150/200/150 ft.
Multi-Span
75/100/75 ft.
Multi-Span
CG (ft) 5 2.50 6.00 4.50 2.25
Area (ft2) 1.63 1.63 1.63 1.63 1.63
Avy (ft2) 0.01 0.01 0.01 0.01 0.01
Avz (ft2) 1.63 1.63 1.63 1.63 1.63
Iyy (ft4) 34.38 7.21 50.94 27.32 5.61
Izz (ft4) 0.14 0.14 0.14 0.14 0.14
Jxx (ft4) 68.58 14.25 101.69 54.47 11.05
Page 59

The bent cap and abutment diaphragms were assumed to be relatively rigid. Cap and column
element properties are shown in Table 4-4.

Table 4-4 – Section Properties for cap and column elements


Cap
Section
6' x 18'
Column
8' x 8'
Column
7' x 7'
Column
6' x 6'
Column
Area
(ft
2
) 108 108 64 49 36
A
vy
(f
t
2
) 108 108 64 49 36
A
vz (ft
2
)
108 108 64 49 36
I
yy
(ft
4
)
5000 324 341.33 200.08 108
Izz

(ft
4
)
5000 2916 341.33 200.08 108
Jxx

(ft
4
) 5000 1296 1365.33 800.33 432


Loads – Each bridge configuration was subjected to dead load (self weight), live load and Post-
tensioning loads. A concentrated load of 100 kips was used to simulate the live load. This
loading captures the effect of concentrated axle loads and may magnify its effect on curved
bridges to some extent; therefore it is justified as a simplification for a conservative upper bound.
This load was applied at the middle of the bridge and transversely was located at various
positions: i) on outside web, ii) on inside web, iii) on center of bridge, and iv) on all webs, i.e.,
equally distributed over the bridge width. Maximum stresses occur when the entire bridge width
is loaded, therefore, the results from this case were studied in more detail when developing
guidelines for design. Post-tensioning was also applied along all webs of the section. The
prestress effects are modeled as equivalent loads at nodes, i.e., the axial forces along the
prestress path are applied at each end of each element which in effect is the same as modeling the
prestress tendon as a series of straight lines. In case of grillage models, additional load cases
were studied by loading only the inside and only the outside webs.

Results Review – The following results were obtained for each load case and compared from
spine and grillage models.

- Midspan Deflection at middle of center span
- Midspan Rotation at middle of center span
- Midspan Longitudinal Bending Moment at middle of center span
- Midspan Transverse Bending Moment at middle of center span
- Midspan End Shear at first abutment in single span and at start of middle span in 3-span
case.
- Midspan Normal Stress at bottom outside corner of cross section, at middle of center
span
- Midspan Normal Stress at bottom inside corner of cross section, at middle of center span

The graphical review of results included scatter-grams of each response quantity from spine and
grillage models. An example of such graphs is shown in Figure 4-13.

Page 60
Two-Cell CIP Dead Load: Midspan Moment
-70000
-60000
-50000
-40000
-30000
-20000
-10000
0
-70000 -60000 -50000 -40000 -30000 -20000 -10000 0
Grid Model


Figure 4-13 – Scatter-gram Comparison of Results from Spine and Grillage Models

The ratio of stresses and shear forces from the spine model to those of the grillage model were
also obtained to review the effect of radius of curvature and modeling technique. An example of
such graphs is shown in Figure 4-14a and Figure 4-14b.



0.9
0.905
0.91
0.915
0.92
0.925
0.93
0.935
0.94
0.945
0 1 2 3 4 5 6 7
Radius (ft)
Midspan Outside Corner Stress (ksi)

a) Midspan Outside Girder Shear Ratio b) Longitudinal Stress Ratio

Figure 4-14 – Line Graphs of Shear and Stress Ratios of Results from Spine and Grillage Models

Summary of Results – Numerous scatter-grams were plotted for the various items described on
page 58 and bridge types shown in Figures 4-4 thru 4-10. Figure 4-14 shows the plots with the
most divergent results (i.e. with values furthest away from 1.00) between the spine and grillage
models. As a result, the outside web shear force and longitudinal stress are the main effects that
need attention in design. To better quantify these results and the effects of curvature on various
Page 61
bridge types, ratios of results for various bridge models are combined and plotted as shown in
Figures 4-15 through 4-18. Average ratios and standard deviations are also calculated for each
group of bridges with same cross-section type and loading. These results are the primary source
of the recommendations for design at the end of this study. Figures 4-15 and 4-17 show the ratios
of Curve to Straight Bridge spine models. This figure reveals the limit of radius of curvature
beyond which the curvature effects may be ignored altogether and the bridge may be analyzed as
if it was straight. The ratios of “Spine to Grillage Model” shown in Figures 4-16 and 4-18 reveal
if a curve spine model can be used in cases of tighter curvatures and the limit of this type of
model to obtain accurate results for design purposes. Note that in the case of the spine models,
shear forces are obtained by considering the shear flow from torsion in addition to the vertical
shear. The stresses are also obtained from the combined effect of axial force, and longitudinal
and transverse moments.


All 3-Span V
r
/V
s
Dead Load
1
1.02
1.04
1.06
1.08
1.1
1.12
1.14
1.16
1.18
0 0.5 1 1.5 2
Length/Radius (L/R)
CIP1_6x18_sp3l_dl
CIP1_6x18_sp3m_dl
CIP1_6x18_sp3s_dl
CIP1_6x6_sp3l_dl
CIP1_6x6_sp3m_dl
CIP1_6x6_sp3s_dl
PC1_6x18_sp3l_dl
PC1_6x18_sp3m_dl
PC1_6x18_sp3s_dl
PC1_6x6_sp3l_dl
PC1_6x6_sp3m_dl
PC1_6x6_sp3s_dl
CIP2_6x18_sp3l_dl
CIP2_6x18_sp3m_dl
CIP2_6x18_sp3s_dl
CIP2_7x7_sp3l_dl
CIP2_7x7_sp3m_dl
CIP2_7x7_sp3s_dl
CIP5_6x18_sp3l_dl
CIP5_6x18_sp3m_dl
CIP5_6x18_sp3s_dl
CIP5_8x8_sp3l_dl
CIP5_8x8_sp3m_dl
CIP5_8x8_sp3s_dl


Figure 4-15 – Ratio of Outside Web Dead Load Shear Forces
in Curved (V
r
) to Straight (V
s
) Bridges where Length Equals Middle Span Length


Figures 4-15 thru 4-19 are for the four bridge cross-sections (cast in place single cell (CIP1), pre-
cast single cell (PC1), cast in place two cell (CIP2), and cast in place five cell (CIP5) with
different pier configurations and span lengths. Therefore, points designated as
PC1_6x6_sp3m_dl are for a single cell pre-cast bridge (PC1) with a 6 x 6 pier (6x6), 3 span
configuration of medium length spans (sp3m) and dead load response (dl).
Page 62
All 3-Span V
line
/V
grid
Dead Load Total
0.9
0.92
0.94
0.96
0.98
1
1.02
1.04
1.06
1.08
0 0.5 1 1.5 2
Length/Radius (L/R)
CIP1_6x18_sp3l_dl
CIP1_6x18_sp3m_dl
CIP1_6x18_sp3s_dl
CIP1_6x6_sp3l_dl
CIP1_6x6_sp3m_dl
CIP1_6x6_sp3s_dl
PC1_6x18_sp3l_dl
PC1_6x18_sp3m_dl
PC1_6x18_sp3s_dl
PC1_6x6_sp3l_dl
PC1_6x6_sp3m_dl
PC1_6x6_sp3s_dl
CIP2_6x18_sp3l_dl
CIP2_6x18_sp3m_dl
CIP2_6x18_sp3s_dl
CIP2_7x7_sp3l_dl
CIP2_7x7_sp3m_dl
CIP2_7x7_sp3s_dl
CIP5_6x18_sp3l_dl
CIP5_6x18_sp3m_dl
CIP5_6x18_sp3s_dl
CIP5_8x8_sp3l_dl
CIP5_8x8_sp3m_dl
CIP5_8x8_sp3s_dl


Figure 4-16 – Ratio of Outside Web Dead Load Shear Forces from Spine to Grillage Model
Page 63

All 3-Span f
r
/f
s
Dead Load
0.96
0.98
1
1.02
1.04
1.06
1.08
1.1
0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6
Length/Radius (L/R)
CIP1_6x18_sp3l_dl
CIP1_6x18_sp3m_dl
CIP1_6x18_sp3s_dl
CIP1_6x6_sp3l_dl
CIP1_6x6_sp3m_dl
CIP1_6x6_sp3s_dl
PC1_6x18_sp3l_dl
PC1_6x18_sp3m_dl
PC1_6x18_sp3s_dl
PC1_6x6_sp3l_dl
PC1_6x6_sp3m_dl
PC1_6x6_sp3s_dl
CIP2_6x18_sp3l_dl
CIP2_6x18_sp3m_dl
CIP2_6x18_sp3s_dl
CIP2_7x7_sp3l_dl
CIP2_7x7_sp3m_dl
CIP2_7x7_sp3s_dl
CIP5_6x18_sp3l_dl
CIP5_6x18_sp3m_dl
CIP5_6x18_sp3s_dl
CIP5_8x8_sp3l_dl
CIP5_8x8_sp3m_dl
CIP5_8x8_sp3s_dl

Figure 4-17 – Ratio of Outside Corner Dead Load
Longitudinal Stress in Curved to Straight Bridge
Page 64

All 3-Span f
line
/f
grid
Dead Load
0.94
0.95
0.96
0.97
0.98
0.99
1
0.00 0.20 0.40 0.60 0.80 1.00 1.20 1.40 1.60
Length/Radius (L/R)
CIP1_6x18_sp3l_dl
CIP1_6x18_sp3m_dl
CIP1_6x18_sp3s_dl
CIP1_6x6_sp3l_dl
CIP1_6x6_sp3m_dl
CIP1_6x6_sp3s_dl
PC1_6x18_sp3l_dl
PC1_6x18_sp3m_dl
PC1_6x18_sp3s_dl
PC1_6x6_sp3l_dl
PC1_6x6_sp3m_dl
PC1_6x6_sp3s_dl
CIP2_6x18_sp3l_dl
CIP2_6x18_sp3m_dl
CIP2_6x18_sp3s_dl
CIP2_7x7_sp3l_dl
CIP2_7x7_sp3m_dl
CIP2_7x7_sp3s_dl
CIP5_6x18_sp3l_dl
CIP5_6x18_sp3m_dl
CIP5_6x18_sp3s_dl
CIP5_8x8_sp3l_dl
CIP5_8x8_sp3m_dl
CIP5_8x8_sp3s_dl


Figure 4-18 – Ratio of Outside Corner Dead Load
Longitudinal Stress from Spine to Grillage Model

Detailed results from the parameter studies are presented in Appendix E.

Conclusions of the Parametric Study – The above results have been studied and the following
general conclusions are drawn:

- Bridges with L/R less than 0.2 may be designed as if they were straight. Figures 4-15 and
4-17 reveal results are within 4% of a curved spine model when this is done.
- Bridges with L/R less than 0.8 may be modeled with single girder spine model using a
curved (spine) model and lateral effects shall be included in the analysis. Figure 4-18
shows that longitudinal stresses will be unconservative by less than 4% up to this limit
unless the bridge has a low span length to width ratio (i.e. short span 5-cell sections)
- Bridges with L/R larger than 0.8 shall be analyzed with more detailed analysis models
such as grillage or finite element model. Figures 4-16 and 4-18 reveal spine beam
analysis will generally become increasingly unconservative beyond the 0.8 L/R limit.
- Curved bridges with length/width ratios of less than 0.2 and an L/R larger than 0.2 also
require detailed analysis as revealed by the unconservative results for tightly curved short
span 5-cell bridges in Figures 4-16 and 4-18.
- Bearing forces (i.e. Shear at the abutments) obtained from a spine model must consider
the effect of torsion. Bearings must be designed considering the curvature effect.
Page 65
d Special Studies

In addition to the above parametric study, a set of special studies were performed to get a better
understanding of the effect of diaphragms, bearing conditions, skewed abutments and long term
creep when combined with curved geometry.

Diaphragm – All five-cell grillage bridge models that were used in parametric study were also
modified to have a stiff diaphragm in the center of each span. The results from each model with
and without diaphragms were compared for dead load, live load and post-tensioning. These
results were compared using scatter-grams and ratios (line diagrams) similar to the results shown
in Figures 4-19 and 4-20. The overall conclusion from these results is that interior diaphragms
have minimal effect on the shear and stress responses, and therefore may be eliminated
altogether.



Figure 4-19 – Scatter-gram Comparison of Live Load Results from Grillage Models With and
Without Interior Diaphragm

Page 66


Figure 4-20 – Line Graph of ratio of Results from Grillage Models With and Without Interior
Diaphragm

To verify the conclusions relative to interior diaphragms, the two-cell finite element model used
in the model verification studies was modified to include interior diaphragms. These diaphragms
were placed at the center of one of the end spans and at one of the third points in the center span.
These diaphragms were located on one side of the midpoint of the bridge, which was otherwise
symmetrical. Results were compared on both sides of the bridge and found to be nearly
symmetrical. Therefore, the diaphragms were shown to have very little effect on the global
response of the bridge.

Bearings at the Bents – The purpose of this study is to determine if bridges with integral and
non-integral bents respond differently to curve effects. All three-span five-cell spine and grillage
bridge models that were used in the parametric study were also modified to have a point bearing
support at the piers; i.e., free to move in transverse or longitudinal directions. The results were
studied using scatter-grams and ratio (line) diagrams comparing the results from spine to grillage
models. It was found that, in general, magnitudes of curve to straight results are in the same
order as the integral bridges. Therefore, the final conclusion is that as long as the support
conditions are modeled correctly, the same guidelines for modeling limitations are equally
applicable to integral and non-integral conditions.
Page 67
Skewed Abutments – A two-cell single span (200 ft long) and a two-cell three span bridge
(200ft-300ft-200ft) were modified to have 30 degree skew support at the left support and another
case with 30 degree skew support at both ends, see Figure 4-21. The abutments are supported on
rollers. The obtuse and acute abutment shear values were compared in straight and skewed
bridges with and without curved alignment. The simple span bridge results were compared for
dead load and live load and the three span results were compared for dead load responses. The
results are shown in Tables 4-5 through 4-7. The final outcome of these results is that the curved
alignment does not aggravate the effect of skewed abutments and therefore, any consideration
taken for straight bridges can be equally valid for curved bridges. These corrections will be
necessary when the bridge is designed using a spine beam analysis. Skew can be automatically
accounted for in a grillage analysis approach as shown in Figures 4-21.


Skew at One Abutment Only


Skew at Both Abutments

Figure 4-21 – Skew Configurations Studied

Table 4-5 - Dead Load Shear Results for 200 ft/Single Span Skew

Obtuse (Element 1) Acute (Element 21)
Straight Value Ratio Value Ratio
Radial -389.35 1 -387.84 1
Skew-Left -605 1.55387184 -190.65 0.49156869
Skew-Both -392.94 1.0092205 -342.23 0.88239996
Obtuse (Element 1) Acute
400’ Radius Value Ratio Value Ratio
Radial -577.89 1 -189.78 1
Skew-Left -782.91 1.3547734 -58.279 0.30708715
Skew-Both -614.14 1.0627282 -183.66 0.96775213

Page 68
Table 4-6 – Live Load Shear Results for 200 ft/Single Span Skew

Obtuse (Element 1) Acute
Straight Value Ratio Value Ratio
Radial -577.89 1 -189.78 1
Skew-Left -782.91 1.3547734 -58.279 0.30708715
Skew-Both -614.14 1.0627282 -183.66 0.96775213
Obtuse (Element 1) Acute
400’ Radius Value Ratio Value Ratio
Radial -81.618 1 -12.007 1
Skew-Left -116.15 1.42309295 12.526 -1.0432248
Skew-Both -85.079 1.04240486 -9.1316 0.76052303


Table 4-7 – Dead Load Shear Results for 400 ft Radius, 200/300/200 ft Multi-Span Skew

Obtuse (Element 1) Acute
Straight Value Ratio Value Ratio
Radial -297.35 1 -296.59 1
Skew-Left -378.79 1.27388599 -197.35 0.66539668
Skew-Both -378.02 1.27129645 -197.66 0.66644189
Obtuse (Element 1) Acute
400’ Radius Value Ratio Value Ratio
Radial -370 1 -202.2 1
Skew-Left -452.84 1.22389189 -135.81 0.67166172
Skew-Both -453.89 1.22672973 -135.81 0.67166172


Long Term Creep – The effect of the time dependant properties of concrete (principally creep)
on the response of curved bridges was investigated using the LARSA 4D computer program,
which can consider both the three-dimensional geometry of the bridge plus the time dependant
behavior of concrete. The same structure used in the comprehensive example problem was
analyzed over a period of 10 years. The bridge was modeled as a 3-D spine beam. In addition,
the model was modified by changing the curve radius in two cases and changing the length of the
end span in another case. Therefore, four models were evaluated. In all cases the abutments were
fixed against torsion. The end span and radii of these bridges are shown in Table 4-8

Table 4-8 – Models of Two-cell Bridge Used to Study the Effect of Creep

Model Number Radii (ft) End Span Length (ft)
1 400 200
2 800 200
3 1600 200
4 400 140
Page 69

The long-term deflection of models 1 through 3 did not appear to be affected by the radius of the
bridge. Therefore, methods used for adjusting cambers for straight bridges would appear to be
applicable to curved bridges analyzed as three-dimensional spine beams.
The major concern, however, was abutment bearing reactions. These will change over time. This
is manifested by the change in the torsion reaction at the abutment, which in turn will affect the
bearing reactions. Table 4-9 summarizes the dead load and prestress results from the four models
investigated.

Table 4-9 – Time Dependence of Abutment Torsion Moment

Model Number Torsion Moment
7 Days (ft-kips)
Torsion Moment
3600 Days (ft-kips)
% Difference
1 -8826 -10034 13.7
2 -7357 -9262 25.9
3 -6395 -8500 32.9
4 -4012 -5842 45.6

The following conclusions can be drawn from this limited study:
- The torsion reaction and thus the relative bearing forces in continuous curved concrete
box girder bridges vary over time and are dependent on both the radius of the curve and
the relative length of the end span with respect to the center span. The forces in outside
bearings will increase and inside bearings will decrease. Given the many possible bridge-
framing configurations, it is difficult to make an accurate assessment of time dependent
bearing forces.
- The above study assumed bridges constructed on falsework. Segmentally constructed
bridges are expected to behave differently.
- The LARSA 4D program does not consider torsion creep, as is the case with most other
commercially available software. Torsion creep is expected to mitigate long-term
changes in bearing forces, as would modeling the flexibility of bearing systems subjected
to torsion loads from the superstructure.
Given these conclusions, it would appear to be safe to analyze curved concrete box girder bridges
using commercially available software. This should definitely be done for segmentally
constructed bridges. It is recommended that the vertical flexibility of bearings be considered in
these analyses.
In the absence of such an analysis, it is recommended that dead load torsion reactions at the
abutment from a three-dimensional spine beam analysis be increased by approximately 20% for
the final condition and bearings or bearing systems be designed to envelope both the initial and
final conditions. This is a crude recommendation, but given the mitigating factors not considered
Page 70
in this limited study, it should provide a reasonable hedge against bearing failure. In the case of a
grillage analysis, the same adjustment can be made by resolving bearing reactions into a
torsional moment, increasing that moment by 20% and recalculating the new bearing forces.
Page 71
5. LOCAL RESPONSE ANALYSIS STUDIES

There are three types of actions, as shown in Figure 5-1, that have been considered in the
analysis work for this project.



Figure 5-1 Types of Actions Considered

1. Global or overall girder action of the bridge together with its supporting piers and
abutments.
2. Regional beam action of each web supported at the top and bottom flanges as a beam.
3. Local action of the concrete cover over the tendons, and/or local lateral shear/breakout
failure adjacent to the ducts. This is sometimes referred to as Lateral Tendon Breakout
(LTB)

Global or overall girder action of the bridge together with its supporting piers and abutments is
covered in Chapter 4. This chapter focuses on “regional” and “local” action.

Regional beam action considers each web as a beam supported at the top and bottom flanges.
The regional moments can be determined from a 2-D frame analysis of the cross section. The
prestress lateral force is determined individually for each web with due consideration for the
allowable variation in prestress force between webs. The compressive reactive forces on the
concrete are applied as distributed loads on the webs and as concentrated loads at the centerlines
of the slabs. Note that the system is in static equilibrium and the support reactions will be zero.

Local slab action of the concrete cover over the tendons has been identified as a major cause of
failure in several curved post-tensioned bridges that did not have duct or web ties. For a web
without duct or web ties, the cover concrete is the only element restraining the lateral prestress
Page 72
force. The cover concrete acts as a plain concrete beam to restrain the lateral prestress force, F,
as shown in Figure 5-2. The “local slab” is subject to lateral shear and bending from the lateral
prestress force. The web is subject to regional transverse bending which results in tensile stresses
on the local slab.

d
M
d
min
min
T C
C T
LOCAL
SLAB
W = F/L
L
LOCAL
MOMENT
L F
V
V
M
M
n
n
V ~ F/2 T ~ M /(jd) M ~ FL/12
jd
n L


Figure 5-2 Regional and Local Actions on a Web

Detailed local analysis models were used to evaluate the local stresses resulting from
longitudinal tendons generating transverse forces on curved webs. Finite element (FE) models
were developed to perform parameter studies, investigate capacities and damage modes, and
prescribe a methodology to prevent such damage. Detailed 3-D, nonlinear analyses were
performed using ABAQUS Version 6.5 ‘damaged plasticity’ concrete cracking model. Such
models provided tendon horizontal force plots versus deformation, and these in combination with
strain contour and crack pattern plots show the evolution of damage with increasing force.
Comparison of such plots amongst different geometries and reinforcing schemes provide
quantitative and qualitative parameter sensitivity evaluation. It should be emphasized that the
focus of the FE models was on the comparison of parameter sensitivities rather than on the
prediction of “post failure” behavior.

a Local Analysis Validation / Demonstration Case (UT Test Case)

The local analyses were begun with a validation case simulating Specimen “BC” from prior
research conducted at the University of Texas (Van Landuyt, 1991). The case studied is a 1/3
scale representation of the configuration of Las Lomas, which is a well-known bridge that failed
in lateral tendon breakout.
Page 73

Test Model and Test Conduct

The following test model and test conduct description comes from the test report and thesis:
“The Effect of Duct Arrangement on Breakout of Internal Post-tensioning Tendons in
Horizontally Curved Concrete Box Girder Webs,” by D.W. Van Landuyt, 1991.

The box cross-section was a scaled version of Las Lomas with changes made for simplifying
construction (Figure 5-3). To avoid having to build cantilever forms, the girder was built and
tested in an inverted position. This did not significantly affect results (it was assumed by the
researchers that gravity loads are unimportant to breakout). The model top slab therefore
represents the bottom slab at Las Lomas, etc. The top slab thickness at Las Lomas varied
transversely and the bottom varied longitudinally. Average thicknesses were calculated, scaled
and rounded off to an integral number of inches. At Las Lomas, the centerline distance between
interior and exterior webs was 11’. This scaled to 3’-8”. The model was constructed with a
centerline web-to-web distance of 4’ so that the radius of each web was a whole number. To save
on materials and labor, the cantilevers were shortened slightly from a scaled value of 1’-10” to
1’-4”. (The actual cantilever length has little effect on web behavior.) The dimensions of the
web were considered important. The exact scaled values of 4” for the thickness and 3’ for the
overall height were maintained (Figure 5-4).

The web radii were chosen to be small enough so that the tendon breakout in the web would
occur before failure of any other part of the girder or testing apparatus. If the curve was not sharp
enough, anchorage zone failure may have resulted or the strands would have been loaded to an
unsafe level. Duct arrangement controlled the design of the curve radius. The capacity to resist
lateral shear failure was calculated for each tendon assuming two failure planes would form and
that the maximum concrete strength would be 5000 psi.

( ) ( ) ft k F
r
/ 8 . 3 5 12 . 1 2 1 2 5000 2 = ' ' ' ' =

where F
r
is the lateral (radially oriented relative to the curve) prestress force.

Therefore a total F
r
of 15.2 k/ft was required for all four tendons. A jacking force of 372 kips
could be delivered from the loading apparatus. An 18’ radius for the inside of web was selected,
as it would permit a total F
r
of 20.6 k/ft. This was more than a third larger than the anticipated
failure load. The previously determined cell width mandated that the companion web radius be
22’. The curve length of 5’ was chosen for the 18’ radius. The curved region was the most
difficult part of the model to construct and was kept as short as possible. A 5’ curve was more
than twice the clear height of the web and was thought to be sufficient to allow regional
transverse bending. A 5’ straight transition zone on each end insulated the curve from the
complex stresses at anchorages.

Las Lomas was reinforced with GR40 #5 stirrups spaced at 15”. No standard bars match this on a
1/3 scale. The closest match was 6mm, 75 ksi bars from Sweden already available in the lab.
This is nearly equivalent to a #2 bar. The stirrup spacing needed to be adjusted to reflect the
imprecise scaling of stirrup sizes. Equivalent spacing of #6’s at 21.3 inches was scaled to 6mm
Page 74
bars at 7”. The spacing was not increased to account for the greater yield strength of the Swedish
bar. Stirrup spacing was reduced in the anchorage zones to 5½”.

A four-tendon bundle and three other promising arrangements were tested. Only duct positioning
varied from web to web; all other details remained the same. Duct size at Las Lomas was not
given although based on the maximum number of strands (28), ducts of approximately 4½” O.D.
should have been used. Scaling required 1½” ducts for the model, but the major manufacturers of
post-tensioning duct do apparently not make this size. The nearest available duct size (1.75”) was
used.

Specimen BC is the duct arrangement similar to the one at Las Lomas. A slight modification was
made for the model (Figure 5-4). A straight vertical formation was used in lieu of the zigzag
because it was considered more universal. The relative horizontal offsets between ducts in a
zigzag pattern can change from bridge to bridge depending on the clear distance between the
stirrup legs and the diameter of the ducts. No large difference in behavior between the two
arrangements was anticipated. The vertical bundle height at Las Lomas was approximately 16”; a
vertical stacked bundle would have been 17½”. All ducts are centered on the web vertical axis.

Specimen 1.0DC follows the Texas State Department of Highway and Public Transportation
design of the San Antonio “Y” project with an arrangement that maintained a clear spacing
between ducts equal to the diameter of the duct. It is believed that this allows for better
consolidation and, more importantly, eliminates the single large discontinuity found at Las
Lomas. This arrangement is conservative and would be considered an upper limit beyond which
further spacing of ducts would provide no benefit. The scaled vertical spacing was 1.75”. All
ducts were centered on the web vertical axis.

All test concrete strengths were much greater than the 28-day design strength of 3500 psi. The
slab and web concrete had higher overall strengths and faster strength gains as is typical of
concrete containing super-plasticizers. The web concrete strength was 5300 psi.

6mm diameter hot rolled bars were used for all reinforcement. Tensile tests on bars conducted at
the lab showed average yield strength of 75 ksi.

Galvanized, corrugated, folded metal ducts were used in all instances. The outside ridge-to-
outside ridge dimension was 1.75”. The inside diameter was 1.60” and the gauge was 0.035”.

Post-tensioning was applied to the specimens with 7-wire, ½”Ø, 270 ksi, low relaxation strands.
Test data provided with the strand showed an actual yield of 276 ksi and ultimate of 289.5 ksi.

A loading system was developed to apply gradually increasing load simultaneously to all tendons
with an equal force in each tendon. It was necessary to consider how the strand or strands that
would constitute a tendon would bear on a duct. The ducts at Las Lomas were nearly filled to
capacity with strands. This meant that almost the entire 180 degrees of the duct on the inside of
the curve was in contact with the strands. That same type of load distribution could be
approximated with a minimum of three ½” diameter strands per duct (Figure 5-5). Since there
were four ducts, a total of twelve strands could be safely stressed to 0.75f
pu
to develop a
maximum force of 372 kips.

Page 75
Regional beam behavior was monitored by deflection of the web relative to the top and bottom
slabs. U-shaped frames were mounted to the web face on the outside of the curve (Figure 5-6).
The actual attachment points were about 2” below the top slab and 2” above the bottom slab.
This permitted the construction of only one type of frame, which could be mounted on either
web. Ideally, the frame should have been mounted on the slabs. However the deflection
anticipated in the first 2” is negligible. A single potentiometer was mounted at the mid-height of
the frame (which is also the c.g. of the tendon group). A small mirror glued to the specimen
provided a smooth surface on which the potentiometer stem could bear.

Mounting the potentiometers on the outside face of the curve meant that deflections should not
be influenced by local beam action; regional beam behavior should be solely responsible for
measured deflections. Also potentiometers attached to the back face were protected from
exploding concrete. The web-slab potentiometer nomenclature is given in Figure 5-6. The
description begins with the letters WS to signify that deflections of the web relative to the slab
are being measured. The number of the stirrup nearest the potentiometer follows these letters.
Web delaminations were measured by wires/potentiometers placed in tubes cast through the
webs above and below the duct group. Sudden movements in these measurements were good
indicators of imminent failure.

Figure 5-7 shows a photo of the test specimen and Figure 5-8 shows a sketch of how the concrete
cracked and failed during the test. Figure 5-9 shows a comparison of tendon horizontal force
versus deflections for the four different webs tested.

Finite Element (FE) Model and Analysis

A FE model of Girder BC and 1.0DC was developed for a typical slice of the test model in the
curved region, as was shown in Figure 5-3. The cross-section and duct geometry (as was shown
in Figure 5-4, 5-5) were modeled per the test configuration and dimensions. The FE model and
boundary conditions are illustrated in Figure 5-10. This shows how a 3-dimensional slice was
modeled with horizontal tendon loads applied directly to the inner surfaces of the ducts. The
wedge slice model configuration and “symmetry” boundary conditions allowed horizontal
(radial) displacement to occur naturally in the FE model, and this in turn, causes “longitudinal”
compression prestress in the concrete in a manner similar to the actual test specimen. What is
precisely mathematically represented with such boundary conditions is a wedge-slice of a
complete circle, but the boundary conditions are also reasonably accurate for a section of a
curve.

The tendon forces were applied in incremental fashion with enough equilibrium iterations and
small load increments to achieve solution convergence at every increment. This solution
procedure is often referred to as Full-Newton, where upon the next load increment; the structure
stiffness is updated, based on the material damage, which has occurred (concrete
cracking/crushing and steel yielding). The analysis was run out to large web displacements and
significant damage (failure), i.e., well beyond the displacements plotted in Figure 5-9.

A typical deformed shape from the analysis (at the 1.0 DC web displacement of 0.07 inches) is
shown in Figure 5-11. It can be noted that the amount of deflection and the sharpness of the
flexural curvature is significantly more severe in Girder-BC than in Girder-1.0DC. This agrees
with test observations. In the illustration, displacements are magnified by 50.
Page 76

Figure 5-12 shows the same deformed shape, but with contours of maximum principal strain. In
reinforced concrete analysis, these are one of the most effective ways to show damage and
deformation distributions in the concrete and rebar. Concrete stress contours are generally not
helpful, because after concrete cracks, the stress reduces to nearly zero, so a zero or small tensile
stress may be displayed in a zone that is already highly damaged. But maximum principal
strains (generally, the maximum tension found in any orientation for a given point in the
continuum) can indicate concrete cracking, which occurs at strain of approximately 1.50E-4 to
2.00E-4. Figure 5-12 shows widespread cracking in the vicinity of the ducts, and much more
severe cracking at the BC configuration than at the 1.0DC.

Figure 5-13 shows similar strain contours, but in an end view at a lower displacement. Here the
crack zones are also fully formed, but are easier to see in cross-section. Figure 5-14 shows the
same deformed shape as Figure 5-13, but viewed in conjunction with principal strains, is used to
estimate the extent of cracks. Using this information, and judgment from experience with many
similar analyses, actual cracks implied by the analysis are drawn onto the figure. Comparing to
Figure 5-8 shows similar trends in the location of the cracking as compared to the test. In
making this conclusion, it is assumed that the test also showed other small cracks in the vicinity
of the ducts, but the cracks shown in Figure 5-8 were the primary failure planes.

Another significant analysis versus test comparison is shown in Figure 5-15: the lateral force
versus deflection at the mid-height of the duct bank relative to the slab. The comparison shows
that for both the BC and 1.0DC girder webs, the analysis is simulating the response (initial
stiffness and stiffness degradation with accumulating damage) and the lateral force capacity
reasonably well.

It is concluded from these comparisons, and from experience with many similar reinforced
concrete analyses, that the FE representation and modeling strategy is appropriate for moving on
to the prototype models.

Page 77


Figure 5-3
Plan and End View of U.T. Girder Test Specimen (from Van Landuyt, 1991)

Page 78
Figure 5-4
Girder #1 Cross Section in Curved Region (from Van Landuyt, 1991)
Page 79





Figure 5-5
Strand Positions in Curve (from Van Landuyt, 1991)




Figure 5-6
Web Potentiometer Configuration (from Van Landuyt, 1991)

Page 80


Figure 5-7
Photo of 1.0 DC and BC (from Van Landuyt, 1991)

Page 81
Figure 5-8
Specimen Failure Plans (form Van Landuyt, 1991)
Page 82
Figure 5-9
Comparison of Web Deflections (from Van Landuyt, 1991)
Page 83

Figure 5-10
Finite Element Model and Boundary Conditions
Page 84


Figure 5-11
Deformed Shape (Displacements x 50) at 1.0 DC Deflection of 0.07 inches.
Page 85


Page 86




















Figure 5-13
Contours for Max Principle Strains at BC Deflection of 0.03 inches (displ x 25)


























Figure 5-14
Estimated Cracking in the Webs Based on Strains (Displ. X 25)

Page 87


Figure 5-15
Force vs. Deflection (FE model compared with U.T. Test)















Page 88
b Local Analysis of Multi-cell Box Girders

For the NCHRP 12-71 project study, there are two basic local model types:

o Multi-cell box (three cell box with super-elevation / vertical interior webs,
and inclined exterior webs; also ran with vertical exterior webs)
o Single-cell box (prototypical pre-cast box with super-elevation and inclined webs)

The prototype geometry for the multi-cell box girder was super-elevated since this is how curved
girders often occur in practice. In the multi-cell studies conducted, the only effect of super-
elevation was a small difference in the end conditions of the regional moment calculation in the
sloping exterior webs. Though not studied in depth, there was no apparent effect on the behavior
of vertical webs caused by super-elevation so it was not included in the single cell section that
was studied.

The variables that may significantly influence local behavior are listed below.

o Web depth
o Web thickness
o Web slope
o Cover thickness
o Number and configuration of tendon ducts
o Number and configuration of duct ties
o Stirrups
o Concrete material properties, especially assumed tensile strength

A last variable, P
t
/R, is evaluated by the analyses producing Pt/R versus deformation curves. So
each analysis includes the full range of P
t
/R up to failure. The analysis matrix to study the
parameters for the multi-cell series is shown in Table 5-1.

Model Prototype: Three-Cell CIP Box Girder

A multi-cell, cast-in-place box configuration was used as shown in Figure 5-16. The basic model
(shown in Figure 5-17) uses 3d elements and a slice, with out-of-page thickness equal to one
stirrup spacing. The stirrups (and other rebar) are modeled explicitly, (unlike in two dimensions,
where the stirrup is ‘smeared’). This allows introduction of the out-of-page compression due to
prestress. Using this model framework, geometry variabilities were introduced directly into the
models – i.e., one web can have one thickness, and another have a different thickness, etc. Also,
by using this model prototype, the effects of web slope are included and can be compared. Webs
A and D have different slopes and can be compared to B and C, which are vertical. And Webs A
and D demonstrate the differences related to web sloping away from the radius of curvature
versus sloping toward the radius of curvature. (Two additional cases were later added with
vertical-web exterior webs to provide additional comparisons.)

The tendon duct arrangements and local reinforcements are shown in Figure 5-18. In
configuration “Type 4”, analyses were run with (“4b”) and without the center web ties (“4a”).
This refers to the two rebar ties in the middle of the group of ducts. In configuration Type 3,
separation of the ducts by 1-1/2” was found to provide increased resistance to lateral tendon
Page 89
breakout. Further separation may provide even higher resistance, but it was the opinion of the
project team that enforcing even larger separations between the ducts begins to pose a very
significant limitation on the effectiveness of the prestressing. Designers do need to be allowed
some flexibility in duct placement in order to achieve a range of vertical positions of prestressing
within the webs, for typical designs.

The boundary conditions for the models were the same as for the UT Test simulation, which
produced reasonable correlation between analysis and test. The model is a sector slice taken
from a curve. The dimension longitudinally varies between inside and outside edge, but on
average is equal to 1.5 ft. This is also the stirrup spacing for the baseline model. This bridge
example is assumed to be on an 800-foot curve radius, so the sector width varies slightly from
the inside of the curve (Web D) to the outside (Web A).

The concrete properties were f
c
’ = 5,000 psi, and Young’s Modulus = 4,030.5 ksi. The rebar was
Grade 60. Plots of material stress-strain curves for the concrete and steel are shown in Figure 5-
19. Tensile strengths (f
t
) for concrete when tested in direct uniaxial tension can show large
variations, but most results fall within the range 4 ' fc to 6 ' fc ; 5 ' fc is considered a
reasonable average.
275.75” 275.75”
70.86” 47.25” 102” 72.83” 72.83” 102” 47.25” 70.86”
9”
8.5”
& Var
12”
& Var
13”
& Var
10% & Var
I-405 - 55 HOV Connector O/C
General Cross-Section (inches)

Figure 5-16. Multicell Girder Cross-section Geometry

Page 90


Figure 5-17. Example Finite Element Mesh for Multi-Cell Local Analysis Prototype. All Duct
Diameters are 4-1/2”

Type 1 Type 2 Type 3 Type 4
12” Typ
@ all sections
Inside of Curve,
Typ @all
sections
3” Clr
To Duc t
Typ
Duct Typ
Stirrups Typ
3” Clr
To Duc t
Typ
3” Clr
To Du c t
3” Clr
To Du c t
2” c lr
2” c lr
Case B
Only
# 4
12”
# 4 @12”, Alternative
Sides for 135 hooks
o



Figure 5-18. Tendon Duct and Local Reinforcement for Multi-Cell Box Local Analysis
Prototypes (Note that this figure is an idealization of bar placements that
are implemented in the FE analysis)


Page 91



Figure 5-19. Stress vs Strain Curves Concrete and Steel in Local FE Analysis

Page 92
Table 5-1 – Parameters for Multi-cell Girder Local Response Analysis

Analysis # Model Type Web #
Web
Thickness
Duct/Tie
Config.
Bundle
Vert.
Pos.
Stirrup
Spacing (in.)
Cover
Thickness
Concr.
Tens.
Str.
(x fc`)
1M
"baseline"
Multi-cell A, D
B
C
12
12
12
1
1
1
midheight
midheight
midheight
18
18
18
3
3
3
4
4
4
2M Multi-cell A, D
B
C
12
12
12
2a
2a
2a
midheight
midheight
midheight
18
18
18
3
3
3
4
4
4
3M Multi-cell A, D
B
C
A-10, D-12
10
14
2b
2a
2a
midheight
midheight
midheight
18
18
18
3
3
3
4
4
4
4M Multi-cell A, D
B
C
12
12
12
2a
2a
2a
1/4 height
1/4 height
bottom
18
18
18
3
3
3
4
4
4
5M Multi-cell A, D
B
C
12
12
12
2b
2b
2a
midheight
midheight
midheight
18
18
18
3
3
3
4
4
6
6M Multi-cell A, D
B
C
12
12
12
2a
2a
2a
midheight
midheight
midheight
12
12
18
3
3
3
4
4
2
7M Multi-cell A, D
B
C
12
12
12
2a
2a
2a
midheight
midheight
midheight
27
27
18
3
3
2
4
4
4
8M Multi-cell A, D
B
C
12
12
12
2a
2b
2a
midheight
midheight
midheight
18
18
18
2
2
4
4
4
4
9M Multi-cell A, D
B
C
12
12
12
3a
2a
2a
1/4 height
1/4 height
bottom
18
18
18
2
2
2
4
4
4
10M Multi-cell A, D
B
C
12
12
12
3a
3a
3a
midheight
midheight
midheight
27
27
18
3
3
2
4
4
4
11M Multi-cell A, D
B
C
A-10, D-12
10
14
4a
4a
4b
midheight
midheight
midheight
18
18
18
3
3
3
4
4
4
12M Multi-cell A, D
B
C
12
12
12
4a
4a
4b
Midheight
midheight
midheight
18
18
18
3
3
3
2
4
6
13M Multi-cell A, D
B
C
12
12
12
4a
4a
4a
midheight
midheight
midheight
27
27
18
3
3
2
4
4
4
14M Multi-cell A, D
B
C
12
12
12
4b
4b
4b
1/4 height
1/4 height
bottom
18
18
18
3
3
3
4
4
4
2m-Vert Multi-cell A, D
B
C
12
12
12
2a
2a
2a
midheight
midheight
midheight
18
18
18
3
3
3
4
4
4
11M-Vert Multi-cell
A, D
B
C
A-10, D-12
10
14
4a
4a
4b
midheight
midheight
midheight
18
18
18
3
3
3
4
4
4

Page 93

Multi-Cell Models – Analysis Results

The 16 different multi-cell girder local models were developed, and analyses were completed.
The results are shown in this chapter through the following plots and tables. These allow for
qualitative and quantitative assessment and comparison of the cases analyzed.

- Lateral Force vs. Deflection of Web Mid-height
- Lateral Force vs. Deflection of Web Quarter-height
- Maximum Principal Strain Contours in Concrete at 75%, 100%, 125%, 150% Pc
- Strains in stirrup rebar at 3 locations along duct bank at 75%, 100%, 125%, 150% Pc
- Distortions (change in web width) 3 locations along duct bank 75%, 100%, 125%, 150%
Pc

P
c
refers to a lateral force applied to the web that will cause theoretical web failure calculated
using conventional means, and removing various safety factors. This creates a framework for
comparing the results of the detailed FE analysis to a baseline capacity. The displacements were
measured at the “outside curve” edge of the webs.

Although the four individual webs in each multi-cell model tend to act independently, the local
analyses required a decision as to loading of the individual webs. In planning this loading it was
found that the interaction of the webs with their end conditions (i.e., the stiffness characteristics
of the top and bottom slabs) was important to how the webs behaved. An initial study using the
baseline geometry (Model 1M) and all-elastic material properties showed that web mid-height
deflections varied as shown when equal loads (1,000 lbs per web) were applied to the webs:


Figure 5-20. Deformed shape of Typical Cross-section with the
Same Force Applied to Each Web

Page 94


Figure 5-21. Deformed Shape/Strain Contour When the Same Force is Applied to Each Web

These figures show how the exterior web ends are freer to rotate than the interior webs. For the
two extremes of fixed-fixed versus pinned-pinned, the ratios of mid-height moment to applied
tendon force (P) would be h/8 versus h/4 or a ratio of one-half. But the web end conditions are
neither fixed-fixed nor pinned-pinned. One way to quantify these differences associated with
end effects was to apply the tendon forces to a beam model, as shown below.



Figure 5-22. Tendon Forces Applied to a Beam Model of Typical Cross-Section
Page 95

This exercise produced the following ratios of web mid-height moments to applied force.

Web

A B C D
0.186h 0.145h 0.147h 0.171h

h = 92.75 in. (7.73 ft.)

Normalizing to the pinned-pinned condition (M=P x h/4) gives ratios of:

0.744 0.580 0.588 0.684

Considered as coefficients, these can be compared to the Caltrans Memo-to-Designers Formula:

M
u
= 0.8 (1/4) (P
j
/R) h
c


Where P
j
is the tendon force (j is for “jacking force”)

Thus, for this case, Caltrans uses a continuity factor of 0.8 for design. Based on the work
performed for this project, factors of 0.6 for interior webs and 0.7 for exterior webs are proposed.

The overall distribution of these moments to webs agrees well with damage trends observed in
the FE analysis, so one finding from the Local Analysis study may be that designers should
account for this effect in more detail than to just apply a single formula to calculate mid-height
moment demand.

Further study of this using fully nonlinear properties showed that once concrete cracking begins
to occur, the differences between webs become even larger. So it was eventually decided to
choose a baseline prestress force divided by the four webs, then increase this force for the
interior webs, and reduce this force for the exterior webs. Using this as a basis, and choosing a
prestress force large enough to cause significant damage in all of the parametric models, led to
the following total applied forces in kips/ft. This is analogous to P
j
/R. Of course in some cases,
the webs failed prior to reaching this total load.

Web
A B C D
13 k/ft 17 k/ft 17 k/ft 15 k/ft

In order to establish a baseline for comparison with design calculations, as previously mentioned,
P
c
is defined as a “Capacity” calculated using conventional means, but removing safety factors,
so to make direct comparison to finite element analyses. For the interior (B or C webs) of the
multi-cell geometry prototype, P
c
was calculated as follows.

o M
n
= 8.7 k-ft/ft

Page 96
Removing the resistance factor o=0.9.

M
n
= 9.7 k-ft/ft

Applying an over-strength factor for rebar strain hardening (which is included in the FE
analysis),

M
o
= M
n
x 1.125 = 10.9 k-ft/ft

The moment-fixity effect is rounded off at 0.6. It is also recognized that the duct forces are not
applied at one point in mid-height, but are instead, applied at five points distributed along 18” of
the height. This decreases the moment to 88.4% of that caused by a single point load at mid-
height. So the baseline P
c
becomes

P
c
= [10.9 / (h/4)] / 0.6 / 0.884 = 10.6 k/ft


This is used as “100%” in the results tables and plots. But note that for the exterior webs, the
“100%” force is less, because the load is applied in the proportions previously listed. The result
of all this is to proportion loads so that webs will be “failing” more-or-less simultaneously,
therefore maintaining proper flexural stiffnesses in webs and flanges at all loading stages.

The results of all the multi-cell analyses are shown in the following tables and figures. The
figures are shown in Appendix E-b and are numbered as (Fig. E-b-1, E-b-2, etc.). The maximum
principal strain contours illustrate the general level of damage to the concrete surrounding the
tendon ducts (maximum tensile strain regardless of orientation).

The following strain thresholds are important quantities to compare to when viewing these
results.

c
11
= 1.6 x 10
-4
first visible cracking (micro-cracking occurs at about half of this strain)
2.0 x 10
-3
first rebar yield
1.0 x 10
-2
1% strain in rebar; typically wide open cracks / sometimes spalling concrete

In Table 5-4, delaminations (“distortion”) information is provided at “Duct 1”, “Duct 3”, and
“Duct 5”. This refers to a measure across the bottom edge, top edge, and centerline of the duct
assembly.
Page 97
Table 5-2 - Deflections


Web A Web B Web C Web D
Percent Mid Quarter Mid Quarter Mid Quarter Mid Quarter
Model # Capacity

1m 75% 0.0290 0.0287 0.0360 0.0262 0.0271 0.0210 0.0199 0.0139
100% 0.0830 0.0686 0.0962 0.0555 0.0797 0.0420 0.0554 0.0225
125% 0.1851 0.1553 0.2249 0.1371 0.1981 0.1115 0.1390 0.0625
150% 0.4067 0.3567 0.5017 0.3382 0.4859 0.3083 0.3379 0.1867

2m 75% 0.0309 0.0340 0.0242 0.0262 0.0237 0.0259 0.0254 0.0216
100% 0.0872 0.0891 0.0702 0.0690 0.0673 0.0672 0.0673 0.0573
125% 0.1966 0.2026 0.1702 0.1629 0.1676 0.1623 0.1566 0.1366
150% 0.4455 0.4509 0.3821 0.3756 0.4080 0.3897 0.3477 0.3067

3m 75% 0.0554 0.0520 0.0411 0.0403 0.0225 0.0293 0.0309 0.0266
100% 0.1524 0.1395 0.1323 0.1191 0.0615 0.0792 0.0863 0.0751
125% 0.3711 0.3367 0.3314 0.2972 0.1571 0.1974 0.2103 0.1841
150% 2.2610 2.2670 1.6490 1.5490 0.9640 1.2280 0.9860 0.6550

4m 75% 0.2148 0.3214 0.2289 0.3171 0.1890 0.2885 0.2141 0.2803
100% 0.8136 1.2218 0.9112 1.2538 0.7589 1.1392 0.8323 1.0922
125% 2.3248 3.5099 2.7380 3.7348 2.2520 3.3725 2.4692 3.2460
150% 11.1900 16.3600 11.3300 15.5600 9.9500 15.1500 12.4100 14.2900

5m 75% 0.0251 0.0269 0.0190 0.0202 0.0172 0.0195 0.0179 0.0142
100% 0.0665 0.0687 0.0518 0.0539 0.0409 0.0481 0.0511 0.0415
125% 0.1385 0.1408 0.1103 0.1114 0.0938 0.0986 0.1082 0.0870
150% 0.2707 0.2731 0.2248 0.2240 0.1954 0.1983 0.2109 0.1726

6m 75% 0.0352 0.0386 0.0278 0.0304 0.0506 0.0385 0.0257 0.0231
100% 0.1240 0.1290 0.1051 0.1048 0.1610 0.1313 0.0887 0.0840
125% 0.3047 0.3182 0.2684 0.2603 0.3685 0.3123 0.2182 0.2071
150% 0.7176 0.7433 0.6550 0.6354 0.8475 0.7286 0.5231 0.4981

7m 75% 0.0279 0.0298 0.0220 0.0230 0.0207 0.0225 0.0206 0.0170
100% 0.0789 0.0799 0.0583 0.0606 0.0548 0.0587 0.0596 0.0486
125% 0.1674 0.1702 0.1295 0.1325 0.1283 0.1305 0.1332 0.1074
150% 0.2980 0.3033 0.2425 0.2393 0.2493 0.2427 0.2385 0.1935

8m 75% 0.0313 0.0337 0.0217 0.0253 0.0214 0.0253 0.0235 0.0202
100% 0.0922 0.0916 0.0545 0.0660 0.0621 0.0680 0.0671 0.0558
125% 0.2020 0.2030 0.1338 0.1560 0.1639 0.1630 0.1563 0.1285
150% 0.4180 0.4180 0.2815 0.3252 0.3478 0.3461 0.3340 0.2714

9m 75% 0.1442 0.1983 0.1270 0.1825 0.1067 0.1682 0.1481 0.1654
100% 0.5575 0.7819 0.5310 0.7502 0.4394 0.6840 0.5756 0.6430
125% 1.6680 2.3691 1.6878 2.3781 1.3931 2.1697 1.7823 1.9963
150% 9.7190 13.7340 8.8200 12.3520 7.6240 11.8020 11.0610 11.3050

10m 75% 0.0230 0.0254 0.0178 0.0197 0.0178 0.0195 0.0156 0.0122
100% 0.0591 0.0627 0.0397 0.0480 0.0390 0.0463 0.0427 0.0363
125% 0.1350 0.1428 0.0892 0.1096 0.0924 0.1098 0.0990 0.0847
150% 0.2458 0.2636 0.1778 0.2091 0.1755 0.2090 0.1939 0.1626

11m 75% 0.0388 0.0390 0.0253 0.0273 0.0183 0.0213 0.0195 0.0154
100% 0.1024 0.1019 0.0683 0.0765 0.0453 0.0566 0.0575 0.0491
125% 0.2601 0.2544 0.1764 0.1961 0.1160 0.1450 0.1534 0.1270
150% 1.1300 1.2390 0.7370 0.7760 0.6300 0.6890 0.5420 0.3130

Page 98
12m 75% 0.0212 0.0236 0.0168 0.0181 0.0161 0.0180 0.0153 0.0113
100% 0.0493 0.0537 0.0355 0.0414 0.0343 0.0412 0.0407 0.0328
125% 0.1051 0.1118 0.0761 0.0864 0.0705 0.0860 0.0991 0.0748
150% 0.1949 0.2089 0.1515 0.1688 0.1332 0.1642 0.1939 0.1454

13m 75% 0.0190 0.0202 0.0161 0.0169 0.0162 0.0175 0.0154 0.0117
100% 0.0372 0.0392 0.0285 0.0319 0.0288 0.0327 0.0324 0.0256
125% 0.0750 0.0780 0.0533 0.0610 0.0541 0.0627 0.0630 0.0516
150% 0.1201 0.1258 0.0986 0.1089 0.1059 0.1164 0.1122 0.0903

14m 75% 0.1016 0.1346 0.0841 0.1233 0.0842 0.1228 0.1207 0.1239
100% 0.3100 0.4193 0.2772 0.4126 0.2657 0.3889 0.3656 0.3779
125% 0.8000 1.0955 0.7350 1.0775 0.7158 1.0229 0.9539 0.9897
150% 2.1159 2.8840 1.8586 2.7535 1.9289 2.7251 2.5100 2.6583

2MVert 75% 0.0153 0.0160 0.0170 0.0172 0.0162 0.0171 0.0154 0.0162
100% 0.0338 0.0318 0.0400 0.0343 0.0398 0.0348 0.0319 0.0302
125% 0.0913 0.0748 0.0896 0.0731 0.1033 0.0796 0.0806 0.0665
150% 0.1824 0.1475 0.1917 0.1481 0.2198 0.1664 0.1831 0.1401

11MVert 75% 0.0314 0.0295 0.0245 0.0254 0.0146 0.0182 0.0191 0.0206
100% 0.0686 0.0594 0.0474 0.0479 0.0255 0.0320 0.0325 0.0355
125% 0.1449 0.1190 0.1026 0.0977 0.0489 0.0614 0.0677 0.0697
150% 0.2706 0.2221 0.2076 0.1937 0.0935 0.1189 0.1365 0.1337

Page 99
Table 5-3 – Stirrup Strain (%) Adjacent to Mid-height of Ducts
(Duct 1 – bottom, Duct 3 – middle, Duct 5 – top)

Web A Web B Web C Web D
Percent Duct 1 Duct 3 Duct 5 Duct 1 Duct 3 Duct 5 Duct 1 Duct 3 Duct 5 Duct 1 Duct 3 Duct 5
Model # Capacity
1m 75% 0.0002 0.0003 0.0002 0.0002 0.0005 0.0004 0.0002 0.0004 0.0002 0.0002 0.0003 0.0002
100% 0.0007 0.0011 0.0005 0.0006 0.0017 0.0010 0.0005 0.0015 0.0006 0.0006 0.0011 0.0005
125% 0.0015 0.0023 0.0013 0.0012 0.0045 0.0016 0.0012 0.0033 0.0013 0.0013 0.0026 0.0013
150% 0.0026 0.0057 0.0018 0.0017 0.0104 0.0026 0.0020 0.0085 0.0026 0.0024 0.0071 0.0019

2m 75% 0.0002 0.0002 0.0002 0.0002 0.0002 0.0002 0.0002 0.0002 0.0002 0.0002 0.0002 0.0001
100% 0.0007 0.0003 0.0003 0.0003 0.0005 0.0008 0.0003 0.0012 0.0003 0.0006 0.0004 0.0003
125% 0.0013 0.0008 0.0006 0.0007 0.0013 0.0018 0.0008 0.0018 0.0008 0.0013 0.0019 0.0007
150% 0.0026 0.0015 0.0012 0.0018 0.0021 0.0039 0.0019 0.0083 0.0014 0.0024 0.0019 0.0012

3m 75% 0.0003 0.0003 0.0002 0.0003 0.0002 0.0003 0.0002 0.0001 0.0001 0.0003 0.0005 0.0001
100% 0.0009 0.0008 0.0005 0.0012 0.0008 0.0008 0.0004 0.0004 0.0003 0.0010 0.0012 0.0003
125% 0.0023 0.0019 0.0013 0.0027 0.0019 0.0018 0.0010 0.0009 0.0006 0.0018 0.0032 0.0009
150% 0.0209 0.0186 0.0207 0.0225 0.0153 0.0218 0.0178 0.0170 0.0172 0.0242 0.0244 0.0164

4m 75% 0.0011 0.0019 0.0007 0.0008 0.0013 0.0014 0.0009 0.0015 0.0010 0.0016 0.0013 0.0003
100% 0.0025 0.0029 0.0025 0.0019 0.0045 0.0080 0.0027 0.0068 0.0046 0.0056 0.0021 0.0011
125% 0.0093 0.0235 0.0090 0.0065 0.0153 0.0246 0.0107 0.0171 0.0158 0.0197 0.0048 0.0021
150% 0.0789 0.0978 0.0787 0.0793 0.0756 0.0994 0.0870 0.0796 0.0876 0.1241 0.0923 0.0653

5m 75% 0.0002 0.0002 0.0001 0.0002 0.0001 0.0002 0.0001 0.0001 0.0001 0.0003 0.0003 0.0001
100% 0.0007 0.0003 0.0003 0.0007 0.0004 0.0003 0.0003 0.0003 0.0002 0.0007 0.0012 0.0002
125% 0.0015 0.0009 0.0006 0.0016 0.0009 0.0006 0.0009 0.0008 0.0003 0.0014 0.0020 0.0005
150% 0.0029 0.0015 0.0012 0.0033 0.0015 0.0012 0.0019 0.0017 0.0007 0.0025 0.0040 0.0010

6m 75% 0.0003 0.0002 0.0002 0.0002 0.0002 0.0002 0.0007 0.0006 0.0004 0.0004 0.0004 0.0001
100% 0.0016 0.0009 0.0004 0.0007 0.0010 0.0004 0.0018 0.0016 0.0011 0.0014 0.0004 0.0003
125% 0.0041 0.0022 0.0015 0.0014 0.0027 0.0012 0.0042 0.0036 0.0023 0.0028 0.0026 0.0008
150% 0.0099 0.0065 0.0037 0.0027 0.0026 0.0031 0.0110 0.0029 0.0054 0.0081 0.0020 0.0017

7m 75% 0.0002 0.0002 0.0001 0.0002 0.0002 0.0002 0.0002 0.0002 0.0002 0.0002 0.0002 0.0001
100% 0.0006 0.0006 0.0003 0.0005 0.0005 0.0003 0.0005 0.0005 0.0003 0.0006 0.0006 0.0003
125% 0.0013 0.0012 0.0005 0.0010 0.0010 0.0007 0.0010 0.0010 0.0007 0.0013 0.0013 0.0007
150% 0.0026 0.0017 0.0010 0.0017 0.0019 0.0013 0.0017 0.0019 0.0013 0.0023 0.0019 0.0014

8m 75% 0.0002 0.0002 0.0002 0.0002 0.0001 0.0002 0.0001 0.0001 0.0001 0.0003 0.0003 0.0001
100% 0.0009 0.0008 0.0003 0.0003 0.0004 0.0003 0.0004 0.0004 0.0004 0.0008 0.0009 0.0003
125% 0.0019 0.0016 0.0005 0.0008 0.0010 0.0007 0.0012 0.0010 0.0008 0.0015 0.0019 0.0008
150% 0.0040 0.0035 0.0012 0.0015 0.0022 0.0017 0.0024 0.0020 0.0017 0.0024 0.0046 0.0018

9m 75% 0.0010 0.0005 0.0002 0.0007 0.0009 0.0006 0.0008 0.0008 0.0004 0.0013 0.0006 0.0001
100% 0.0042 0.0022 0.0010 0.0029 0.0042 0.0025 0.0027 0.0036 0.0022 0.0062 0.0013 0.0003
125% 0.0147 0.0083 0.0019 0.0101 0.0133 0.0105 0.0100 0.0118 0.0070 0.0203 0.0038 0.0012
150% 0.0747 0.0664 0.0559 0.0708 0.0742 0.0753 0.0721 0.0713 0.0688 0.1066 0.0592 0.0514

10m 75% 0.0001 0.0001 0.0001 0.0001 0.0001 0.0001 0.0001 0.0001 0.0001 0.0002 0.0001 0.0001
100% 0.0004 0.0004 0.0003 0.0002 0.0002 0.0002 0.0002 0.0001 0.0002 0.0008 0.0003 0.0002
125% 0.0009 0.0009 0.0007 0.0003 0.0003 0.0005 0.0004 0.0003 0.0007 0.0016 0.0006 0.0004
150% 0.0016 0.0016 0.0011 0.0008 0.0007 0.0012 0.0008 0.0007 0.0013 0.0027 0.0014 0.0008

11m 75% 0.0002 0.0002 0.0002 0.0002 0.0001 0.0002 0.0001 0.0001 0.0001 0.0003 0.0001 0.0001
100% 0.0006 0.0005 0.0004 0.0003 0.0002 0.0005 0.0002 0.0001 0.0002 0.0008 0.0002 0.0002
125% 0.0016 0.0013 0.0011 0.0008 0.0005 0.0012 0.0004 0.0002 0.0005 0.0021 0.0006 0.0005
150% 0.0103 0.0091 0.0109 0.0108 0.0048 0.0117 0.0098 0.0042 0.0096 0.0154 0.0070 0.0086

Page 100
12m 75% 0.0001 0.0001 0.0001 0.0001 0.0001 0.0001 0.0001 0.0001 0.0001 0.0002 0.0001 0.0001
100% 0.0002 0.0003 0.0003 0.0002 0.0002 0.0002 0.0002 0.0001 0.0002 0.0008 0.0002 0.0002
125% 0.0005 0.0008 0.0007 0.0004 0.0004 0.0005 0.0004 0.0003 0.0005 0.0017 0.0007 0.0006
150% 0.0009 0.0013 0.0014 0.0009 0.0008 0.0010 0.0008 0.0005 0.0009 0.0034 0.0013 0.0011

13m 75% 0.0001 0.0001 0.0001 0.0001 0.0001 0.0001 0.0001 0.0001 0.0001 0.0002 0.0001 0.0001
100% 0.0002 0.0002 0.0003 0.0002 0.0001 0.0002 0.0001 0.0001 0.0002 0.0006 0.0002 0.0002
125% 0.0006 0.0005 0.0006 0.0003 0.0002 0.0004 0.0002 0.0002 0.0003 0.0011 0.0004 0.0003
150% 0.0009 0.0008 0.0009 0.0006 0.0005 0.0007 0.0005 0.0005 0.0007 0.0017 0.0007 0.0005

14m 75% 0.0005 0.0005 0.0002 0.0001 0.0001 0.0004 0.0002 0.0001 0.0003 0.0011 0.0002 0.0001
100% 0.0017 0.0015 0.0009 0.0004 0.0007 0.0021 0.0010 0.0004 0.0013 0.0034 0.0008 0.0002
125% 0.0045 0.0048 0.0019 0.0011 0.0013 0.0063 0.0032 0.0014 0.0032 0.0111 0.0011 0.0004
150% 0.0125 0.0143 0.0046 0.0017 0.0022 0.0182 0.0113 0.0036 0.0081 0.0273 0.0018 0.0012

2MVert 75% 0.0001 0.0001 0.0001 0.0002 0.0001 0.0001 0.0001 0.0001 0.0001 0.0001 0.0001 0.0001
100% 0.0003 0.0003 0.0002 0.0005 0.0004 0.0003 0.0004 0.0004 0.0003 0.0003 0.0003 0.0002
125% 0.0010 0.0009 0.0006 0.0009 0.0010 0.0008 0.0011 0.0010 0.0006 0.0008 0.0010 0.0007
150% 0.0019 0.0017 0.0011 0.0019 0.0021 0.0016 0.0023 0.0020 0.0015 0.0020 0.0021 0.0016

11MVert 75% 0.0002 0.0001 0.0002 0.0002 0.0001 0.0002 0.0001 0.0001 0.0001 0.0001 0.0001 0.0001
100% 0.0006 0.0002 0.0007 0.0002 0.0002 0.0006 0.0002 0.0002 0.0002 0.0002 0.0002 0.0003
125% 0.0013 0.0007 0.0014 0.0006 0.0005 0.0011 0.0003 0.0003 0.0004 0.0005 0.0005 0.0006
150% 0.0024 0.0015 0.0026 0.0013 0.0011 0.0019 0.0006 0.0006 0.0008 0.0011 0.0011 0.0012

Page 101
Table 5-4 – Distortion (Web Thickness Change) Across the Mid-height of Ducts
(Duct 1 – bottom, Duct 3 – middle, Duct 5 – top)

Web A Web B Web C Web D
Percent Duct1 Duct 3 Duct 5 Duct1 Duct 3 Duct 5 Duct1 Duct 3 Duct 5 Duct1 Duct 3 Duct 5
Model # Capacity

1m 75% 0.0020 0.0029 0.0016 0.0029 0.0088 0.0020 0.0020 0.0078 0.0010 0.0020 0.0116 0.0016
100% 0.0033 0.0092 0.0041 0.0368 0.0283 0.0039 0.0039 0.0303 0.0029 0.0041 0.0251 0.0035
125% 0.0075 0.0209 0.0078 0.0254 0.0693 0.0107 0.0078 0.0732 0.0078 0.0062 0.0251 0.0073
150% 0.0101 0.0409 0.0150 0.0508 0.1387 0.0234 0.0156 0.1621 0.0244 0.0089 0.0591 0.0133

2m 75% 0.0016 0.0012 0.0014 0.0010 0.0049 0.0020 0.0010 0.0039 0.0020 0.0034 0.0032 0.0014
100% 0.0026 0.0035 0.0018 0.0020 0.0146 0.0029 0.0020 0.0137 0.0039 0.0067 0.0098 0.0019
125% 0.0051 0.0943 0.0044 0.0010 0.0342 0.0059 0.0068 0.0332 0.0059 0.0158 0.0221 0.0053
150% 0.0113 0.1010 0.0087 0.0029 0.0742 0.0146 0.0293 0.0791 0.0117 0.0443 0.0488 0.0089

3m 75% 0.0019 0.0013 0.0010 0.0029 0.0029 0.0020 0.0020 0.0049 0.0020 0.0024 0.0033 0.0011
100% 0.0023 0.0031 0.0032 0.0098 0.0117 0.0020 0.0039 0.0107 0.0039 0.0064 0.0087 0.0017
125% 0.0052 0.0055 0.0087 0.0244 0.0342 0.0059 0.0078 0.0264 0.0107 0.0112 0.0242 0.0052
150% 0.1803 0.1789 0.1843 0.1748 0.3154 0.1875 0.2803 0.4268 0.2910 0.2535 0.3563 0.2105

4m 75% 0.0019 0.0170 0.0021 0.0078 0.0127 0.0049 0.0078 0.0137 0.0059 0.0104 0.0110 0.0022
100% 0.0048 0.0133 0.0042 0.0293 0.0596 0.0039 0.0264 0.0459 0.0068 0.0380 0.0437 0.0083
125% 0.0249 0.0081 0.0074 0.0947 0.2266 0.0010 0.0752 0.1211 0.0117 0.0918 0.1244 0.0292
150% 0.9973 0.8594 0.8428 1.1494 1.7051 0.9990 1.1553 1.5244 1.1211 1.0413 1.2327 0.9574

5m 75% 0.0014 0.0013 0.0021 0.0020 0.0029 0.0020 0.0010 0.0029 0.0020 0.0034 0.0024 0.0006
100% 0.0013 0.0025 0.0026 0.0068 0.0059 0.0029 0.0020 0.0059 0.0029 0.0065 0.0037 0.0016
125% 0.0042 0.0045 0.0056 0.0186 0.0146 0.0049 0.0049 0.0176 0.0049 0.0133 0.0059 0.0029
150% 0.0065 0.0088 0.0093 0.0361 0.0312 0.0107 0.0117 0.0371 0.0098 0.0253 0.0130 0.0046

6m 75% 0.0016 0.0019 0.0013 0.0029 0.0039 0.0020 0.0059 0.0117 0.0039 0.0026 0.0032 0.0014
100% 0.0044 0.0149 0.0020 0.0068 0.0186 0.0029 0.0225 0.0381 0.0107 0.0075 0.0117 0.0028
125% 0.0080 0.0174 0.0046 0.0146 0.0479 0.0098 0.0547 0.0801 0.0234 0.1099 0.0321 0.0105
150% 0.0193 0.0282 0.0110 0.0391 0.1055 0.0186 0.1113 0.1660 0.0537 0.0447 0.0764 0.0140

7m 75% 0.0021 0.0007 0.0014 0.0020 0.0039 0.0010 0.0029 0.0049 0.0020 0.0027 0.0040 0.0014
100% 0.0021 0.0036 0.0034 0.0059 0.0088 0.0029 0.0049 0.0098 0.0049 0.0082 0.0073 0.0025
125% 0.0035 0.0064 0.0046 0.0117 0.0195 0.0068 0.0156 0.0244 0.0088 0.0184 0.0169 0.0052
150% 0.0068 0.0130 0.0093 0.0283 0.0391 0.0146 0.0303 0.0498 0.0195 0.0358 0.0369 0.0109

8m 75% 0.0014 0.0014 0.0011 0.0020 0.0020 0.0020 0.0029 0.0010 0.0010 0.0040 0.0037 0.0012
100% 0.0010 0.0049 0.0029 0.0029 0.0049 0.0029 0.0029 0.0098 0.0029 0.0083 0.0057 0.0022
125% 0.0039 0.0077 0.0056 0.0078 0.0137 0.0049 0.0244 0.0303 0.0068 0.0197 0.0174 0.0047
150% 0.0075 0.0167 0.0103 0.0156 0.0321 0.0107 0.0527 0.0625 0.0176 0.0478 0.0438 0.0084

9m 75% -0.0001 0.0014 0.0013 0.0049 0.0117 0.0039 0.0029 0.0098 0.0029 0.0031 0.0032 0.0011
100% 0.0012 0.0015 0.0031 0.0215 0.0361 0.0078 0.0078 0.0352 0.0059 -0.0040 0.0089 0.0035
125% -0.0140 -0.0261 0.0055 0.0801 0.1162 0.0059 0.0352 0.1143 0.0146 -0.0002 0.0296 0.0111
150% 0.7518 0.6562 0.6789 0.9424 1.1504 0.7197 0.7256 1.1426 0.8164 0.7270 0.9038 1.8598
Page 102
10m 75% 0.0014 0.0019 0.0013 0.0010 0.0020 0.0020 0.0010 0.0029 0.0010 0.0021 0.0013 0.0016
100% 0.0026 0.0035 0.0013 0.0020 0.0039 0.0020 0.0010 0.0039 0.0029 0.0035 0.0025 0.0024
125% 0.0041 0.0046 8.0000 0.0049 0.0088 0.0049 0.0020 0.0117 0.0049 0.0061 0.0062 0.0031
150% 0.0060 0.0095 9.0000 0.0127 0.0176 0.0078 0.0059 0.0225 0.0088 0.0082 0.0146 0.0064

11m 75% 0.0015 0.0010 0.0015 0.0020 0.0020 0.0010 0.0010 0.0020 0.0010 0.0019 0.0016 0.0008
100% 0.0560 0.0010 0.0021 0.0029 0.0029 0.0010 0.0010 0.0039 0.0020 0.0033 0.0021 0.0024
125% 0.0036 0.0033 0.0032 0.0049 0.0078 0.0029 0.0029 0.0107 0.0049 -0.0090 0.0086 0.0049
150% 0.0809 0.0963 0.0978 0.0977 0.1387 0.0781 0.1211 0.1963 0.1328 0.0939 0.1518 0.1160

12m 75% 0.0014 0.0014 0.0014 0.0008 0.0008 0.0008 0.0020 0.0020 0.0020 0.0018 0.0018 0.0018
100% 0.0020 0.0020 0.0020 0.0016 0.0016 0.0016 0.0039 0.0039 0.0039 0.0030 0.0030 0.0030
125% 0.0041 0.0041 0.0041 0.0041 0.0041 0.0041 0.0078 0.0078 0.0078 0.0038 0.0038 0.0038
150% 0.0078 0.0078 0.0078 0.0093 0.0093 0.0093 0.0127 0.0127 0.0127 0.0071 0.0071 0.0071

13m 75% 0.0015 0.0017 0.0008 0.0010 0.0020 0.0020 0.0000 0.0020 0.0020 0.0014 0.0014 0.0017
100% 0.0023 0.0026 0.0021 0.0020 0.0029 0.0020 0.0000 0.0029 0.0020 0.0026 0.0027 0.0017
125% 0.0022 0.0027 0.0021 0.0039 0.0068 0.0029 0.0000 0.0068 0.0039 0.0041 0.0058 0.0028
150% 0.0047 0.0045 0.0021 0.0078 0.0137 0.0059 0.0000 0.0117 0.0059 0.0059 0.0117 0.0017

14m 75% 0.0018 0.0018 0.0012 0.0010 0.0022 0.0004 0.0010 0.0020 0.0020 0.0016 0.0022 0.0011
100% 0.0030 0.0029 0.0015 0.0042 0.0043 0.0065 0.0010 0.0049 0.0059 0.0013 0.0053 0.0019
125% 0.0072 -0.0008 0.0017 0.0097 0.0109 0.0086 -0.0010 0.0059 0.0117 0.0060 0.0104 0.0047
150% 0.0228 -0.0302 0.0064 0.0268 0.0279 -0.0193 -0.0195 0.0146 0.0098 -0.0168 0.0364 0.0140

2MVert 75% 0.0015 0.0028 0.0020 0.0020 0.0029 0.0010 0.0010 0.0029 0.0020 0.0021 0.0029 0.0020
100% 0.0021 0.0057 0.0020 0.0059 0.0088 0.0020 0.0010 0.0088 0.0029 -0.0017 0.0050 -0.0233
125% 0.0065 0.0204 0.0059 0.0000 0.0000 0.0039 0.0049 0.0264 0.0000 0.0046 0.0175 0.0074
150% 0.0132 0.0397 0.0109 0.0332 0.0488 0.0068 0.0117 0.0557 0.0117 0.0074 0.0371 0.0163

11MVert 75% 0.0010 0.0010 0.0010 0.0010 0.0020 0.0010 0.0020 0.0020 0.0010 0.0022 0.0010 0.0019
100% 0.0000 0.0020 0.0000 0.0020 0.0029 0.0010 0.0039 0.0039 0.0029 0.0024 0.0030 0.0019
125% 0.0000 0.0039 0.0010 0.0039 0.0049 -0.0013 0.0059 0.0068 0.0049 0.0050 0.0069 0.0038
150% 0.0000 0.0107 0.0020 0.0068 0.0146 0.0000 0.0127 0.0107 0.0088 0.0091 0.0148 0.0068

Page 103

Discussion of Results

The analysis results have been used to compare the web design parameters. The first general
observation is the comparison of the mid-height cases to the “1/4 height” and “bottom” cases. It
was generally observed that when the ducts occur near the bottom of the web (either “quarter-
height” or “bottom” as was tested in Configurations 4M and 14M) the force at “failure” is
substantially lower than when the ducts are placed at the mid-height, i.e., on average as much as
25% to 40% lower when comparing these cases to other similar cases. The reason for this is a
tendency toward lateral shear failure of the overall web. When the ducts are located at the mid-
height, the lateral shear is divided equally between the top and bottom of the web. But when the
ducts move down, the bottom of the web carries most of the lateral shear, and this is a different
mechanism than the failure modes observed for tendon ducts at mid-height. So the “quarter-
height” cases can be compared to each other (and the “bottom” cases), but should not be
compared directly to the “mid-height” cases. “P
c
” only applies to the “mid-height” cases.

For purposes of interpreting and comparing the results, the following damage criteria should be
considered:

- Stirrup rebar strain exceeding yield (i.e., 0.2% strain for Grade 60 steel); note that for
Load Factor Design, concrete reinforcement is designed to yield at the ultimate member
forces.

- Visible concrete cracking occurs at strains of approximately 0.016%, but this is not
necessarily web failure; concrete with maximum principal strains of 0.3% can be
considered to be heavily cracked. Concrete with strains in excess of 1.0% will generally
show wide-open cracks and potential spalling from the section.

- Significant distortion or delamination (change of width of the webs) would also represent
an upper bound on serviceable capacity for webs; the delamination is evidence of a local
splitting or lateral shear failure within the web. It was arbitrarily assumed that a crack
width of 1/16” is an indicator of such a failure. For 12” webs, this represents a distortion
of 0.06 inches, and a distortion ratio (average strain through the section) of 0.5%. For
sections with web ties, this means the web ties have yielded; for sections without web
ties, the section is at a web splitting or a web lateral shear failure condition.

Two of the criteria, Stirrup Yield and Web Delamination, have been summarized in Table 5-5
and 2-6. These are the total forces (sum of all tendon ducts in the web) applied when any part of
the stirrup reaches yield, and when the web distortion reaches 0.06 inch.
Page 104

Table 5-5 – Lateral Prestress Force at Stirrup Yield (0.2% Strain) for Stirrups on Inside of Curve

Total Force
(in % of Pc and in K/ft)
Model #
Web A
Web A
Web B
Web B
Web C
Web C
Web D
Web D
1M 119.55% 9.93 107.36% 11.38 111.04% 11.77 122.41% 11.18
2M 141.66% 11.77 128.85% 13.66 128.67% 13.64 129.38% 11.82
3M 121.11% 10.06 118.77% 12.59 144.72% 15.34 134.24% 12.26
4M - 6.24 - 8.59 - 8.57 - 7.42
5M 134.96% 11.21 133.11% 14.11 150.47% 15.95 126.25% 11.53
6M 105.54% 8.77 118.54% 12.57 105.75% 11.21 117.92% 10.77
7M 140.98% 11.71 154.06% 16.33 153.87% 16.31 147.93% 13.51
8M 124.72% 10.36 148.49% 15.74 145.38% 15.41 129.86% 11.86
9M - 7.28 - 9.59 - 9.58 - 7.75
10M 157.23% 13.06 162.45% 17.22 161.60% 17.13 142.01% 12.97
11M 131.45% 10.92 141.51% 15.00 148.15% 15.70 126.77% 11.58
12M 160.55% 13.34 165.81% 17.58 165.58% 17.55 135.53% 12.38
13M 180.74% 15.01 184.97% 19.61 184.00% 19.50 166.09% 15.17
14M - 8.65 - 10.47 - 11.81 - 8.48
2MVert 151.05% 12.55 148.79% 15.77 145.06% 15.38 151.69% 13.85
11MVert 140.59% 11.68 152.49% 16.16 165.41% 17.53 167.19% 15.27
- Percentages not shown for cases other than ducts placed at midheight


Table 5-6 – Lateral Prestress Force at Web Delamination of 0.5% (0.06” for 12” Web)
* Never reached delamination limit, so the delamination at 200% of P
c
is shown

Model # Web A Web B Web C Web D
% Pc
Total
Force Deflection % Pc
Total
Force Deflection % Pc
Total
Force Deflection % Pc
Total
Force Deflection
(K/ft) (in) (K/ft) (in) (K/ft) (in) (K/ft) (in)
1M 165.03% 13.708 3.440 121.05% 12.831 0.193 118.90% 12.604 0.156 155.08% 14.163 0.425
2M 130.38% 10.830 0.249 143.12% 15.170 0.301 142.06% 15.059 0.305 159.45% 14.563 0.446
3M 136.29% 11.321 0.568 120.84% 12.809 0.278 123.99% 13.143 0.148 124.87% 11.405 0.190
4M - 9.458 1.573 - 10.620 0.914 - 10.677 0.787 - 9.337 0.853
5M 200% * 16.613 0.830 184.92% 19.601 0.513 172.64% 18.300 0.384 200% * 18.266 0.688
6M 162.10% 13.464 3.103 131.33% 13.921 0.332 114.65% 12.153 0.257 142.79% 13.041 0.365
7M 200% * 16.613 0.753 173.24% 18.364 0.407 161.42% 17.111 0.346 179.26% 16.372 0.429
8M 200% * 16.613 14.363 178.17% 18.886 1.810 142.28% 15.082 0.274 167.97% 15.341 1.032
9M - 10.058 1.504 - 10.939 0.602 - 11.042 0.528 - 10.767 1.181
10M 200% * 16.613 10.122 200% * 21.200 7.139 200% * 21.200 7.533 200% * 18.266 3.868
11M 145.14% 12.056 0.519 137.96% 14.624 0.259 138.95% 14.729 0.183 140.55% 12.836 0.235
12M 200% * 16.613 7.883 200% * 21.200 7.160 200% * 21.200 5.334 200% * 18.266 3.725
13M 200% * 16.613 2.717 200% * 21.200 2.575 200% * 21.200 2.697 200% * 18.266 2.230
14M - 16.830 9.277 - 20.186 6.473 - 27.218 15.109 - 16.288 6.762
2MVert 166.40% 13.822 0.289 158.62% 16.814 0.250 153.77% 16.299 0.248 171.58% 15.670 0.315
11MVert 200% * 16.613 10.559 200% * 21.200 8.047 200% * 21.200 6.917 200% * 18.266 7.757
Page 105

Using these criteria and the results tables and plots has resulted in the following observations:

Web Depth

This parameter was varied indirectly by subjecting the webs to wide ranges of moments and
horizontal shears. Based on observations of the analysis results, web depth can be adequately
accounted for by considering and designing for web moments.

Web Thickness

Web thickness was varied in Model 3M (A – 10”, B – 10”, C – 14”), and similarly in Model
11M. Model 11M included web ties. The results compared to their respective baselines are
shown in Table 5-7:

Table 5-7 – Effect of Web Thickness – Thin Webs

Model-Web Force at stirrup yield (kips/ft) Difference
2M-A vs. 3M-A 11.77 vs. 10.06 -15% change between 12” vs. 10”
2M-B vs. 3M-B 13.66 vs. 12.59 -8% change between 12” vs. 10”
2M-C vs. 3M-C 13.64 vs. 15.34 12% change between 12” vs. 14”

12M-A vs. 11M-A 13.34 vs. 10.92 -18% change between 12” vs. 10”
12M-B vs. 11M-B 17.58 vs. 15.00 -15% change between 12” vs. 10”

Model-Web Force at web delam. (kips/ft) Difference
2M-B vs. 3M-B 15.17 vs. 12.81 -16% change between 12” vs. 10”

12M-A* vs. 11M-A 16.61 vs. 12.06 -27% change between 12” vs. 10”
12M-B* vs. 11M-B 21.20 vs. 14.62 -31% change between 12” vs. 10”
* Never reached delamination limit


These results demonstrate significant influence on resistance to lateral bending and tendon
pullout caused by web thickness. Stirrups yielded sooner, and concrete damage and web
delamination was more extensive with the thinner webs.

For stirrup yield, capacity formulae based on regional flexure considerations appear to be
appropriate for design. Differences in stirrup yield and especially web delamination were also
significantly influenced when the web ties were added because the web ties tended to eliminate
the web delamination failure mode. Moving the ducts toward the curve outside face within the
webs also contributes to resistance against delamination and local lateral shear damage.

Web Slope

As described earlier, the sloped webs in this analysis series were found to be significantly weaker
(30-40%) than the vertical webs, but part of this difference is caused by being exterior webs
rather than interior. Exterior webs have more flexible end conditions at their connection with the
top and bottom slab, and this produces larger mid-height moments.

Page 106
Comparison of Webs A to D for the inclined webs show that Web A is generally weaker than D
by about 10%, due to orientation of slope relative to the direction of the tendon force.

In order to examine the differences between sloped webs and vertical webs more directly, two
additional analyses were performed with exterior webs converted to vertical webs. The strain
contour and Force versus Deflection plots are included in the Appendix. Models 2M and 11M
were chosen for these comparisons because these have the baseline values for all properties, but
they investigate duct-tie configurations 2a and 4a/b. Direct comparisons of Force versus
Deflection are shown in Figures 5-23a and 5-23b below.

0
2
4
6
8
10
12
14
16
0.00E+00 5.00E-02 1.00E-01 1.50E-01 2.00E-01 2.50E-01 3.00E-01 3.50E-01 4.00E-01
Defl (in)
Model 2M Web A Model 2M Web D Model 2MVert Web A Model 2MVert Web D


Fig. 5-23a. Model 2M and 2MVert
Force vs. Deflection Comparison for Webs A and D (quarter height)
Page 107
0
2
4
6
8
10
12
14
16
18
20
0.00E+00 5.00E-02 1.00E-01 1.50E-01 2.00E-01 2.50E-01 3.00E-01 3.50E-01 4.00E-01
Defl (in)
Model 11M Web A Model 11M Web D Model 11MVert Web A Model 11MVert Web D


Fig. 5-23b. Model 11M and 11MVert
Force vs. Deflection Comparison for Webs A and D (mid height)

These figures show the vertical webs to be stiffer and stronger than the sloping webs, but show
that the differences in force capacity (strength) are negligibly small. A comparison of the
occurrence of rebar yield and web delamination is shown in Table 5-8:

Table 5-8 – Effect of Web Slope – Thin Webs

Model-Web Force at stirrup yield (kips/ft) Difference
2M-A vs. 2Mvert-A 11.77 vs. 12.55 7% change with vertical web
2M-D vs. 2Mvert-D 11.82 vs. 13.85 17% change with vertical web
11M-A vs. 11Mvert-A 10.92 vs. 11.68 7% change with vertical web
11M-D vs. 11Mvert-D 11.58 vs. 15.27 32% change with vertical web

Model-Web Force at web delam. (kips/ft) Difference
2M-A vs. 2Mvert-A 10.83 vs. 13.82 28% change with vertical web
2M-D vs. 2Mvert-D 14.56 vs. 15.67 8% change with vertical web
11M-A vs. 11Mvert-A* 12.06 vs. 16.61 38% change with vertical web
11M-D vs. 11Mvert-D* 12.84 vs. 18.27 42% change with vertical web
* Never reached delamination limit


Page 108
Cover Thickness

Cover thickness was varied in models 7M, 8M, and 13M. Table 5-9 summarizes sample strength
comparisons:

Table 5-9 – Effect of Cover Thickness – Thin Webs

Model-Web Force at stirrup yield (kips/ft) Difference
8M-A vs. 2M-A 10.36 vs. 11.77 14% change between 2" vs. 3"

8M-C vs. 2M-C 15.41 vs. 13.64 -11% change between 4" vs. 3"
8M-D vs. 2M-D 11.86 vs. 11.82 -0.3% change between 2" vs. 3"

Model-Web Force at web delam. (kips/ft) Difference
8M-C vs. 2M-C 15.08 vs. 15.06 -0.2% change between 4" vs. 3"
8M-D vs. 2M-D 15.34 vs. 14.56 -5% change between 2" vs. 3"
Qualitative evidence can also be obtained by examining the strain plots in the Appendices for the models where cover thickness
was varied.

The conclusion reached is that cover thickness influences lateral pullout resistance, but it is not
the only driver of pullout resistance. The results of the parameter study are influenced by the
fact that when the cover is reduced, for the same overall web thickness, the moment arm for the
stirrups is increased, and this is an off-setting influence on pullout resistance. As will be
discussed further in the conclusions, it appears appropriate to check cover concrete thickness for
resistance to initial cracking, but not to include cover concrete tensile strength in the calculation
of regional transverse bending strength.

Number and Configuration of Tendon Ducts

This was evaluated by comparing Configurations 1, 2, and 3 (from Figure 5-18), which involve
comparing Models 1M vs. 2M, 3M-D vs. 2M-D, 5M vs. 2M, and 10M-C vs. 7M-C. Results are
shown in Table 5-10.

Table 5-10 – Effect of Different Duct Configurations – Thin Webs

Model-Web Force at stirrup yield (kips/ft) Difference
1M-A vs. 2M-A 9.93 vs. 11.77 19% change - Config. 1 vs. 2A
1M-B vs. 2M-B 11.38 vs. 13.66 20% change - Config. 1 vs. 2A
1M-C vs. 2M-C 11.77 vs. 13.64 16% change - Config. 1 vs. 2A

3M-D vs. 2M-D 12.26 vs. 11.82 -4% change - Config. 2B vs. 2A

5M-B vs. 2M-B 14.11 vs. 13.66 -3% change - Config. 2B vs. 2A

10M-C vs. 7M-C 17.13 vs. 16.31 -5% change - Config. 3A vs. 2A

Model-Web Force at web delam. (kips/ft) Difference
1M-B vs. 2M-B 12.83 vs. 15.17 18% change - Config. 1 vs. 2A
1M-C vs. 2M-C 12.60 vs. 15.06 19% change - Config. 1 vs. 2A

5M-B vs. 2M-B 19.60 vs. 15.17 -23% change - Config. 2B vs. 2A

10M-C* vs. 7M-C 21.20 vs. 17.11 * change - Config. 3A vs. 2A
* Never reached delamination limit

Page 109

Clearly, when the ducts are spread apart, the performance significantly improves. Roughly 20%
resistance force improvement was demonstrated by separating the 5-duct bundle into two
bundles (Config. 2A versus Config. 1), and an additional 4% improvement was demonstrated by
spreading the bundles further apart (4.5” versus 1.5” separation). So in general, a prudent
recommendation is to require a maximum of 3 ducts per bundle. When the individual ducts were
separated (i.e., Config. 3A), and moved toward the curve’s outside edge of the web, performance
further improved. In fact, as measured by the delamination criteria, Configuration 3A exceeded
200% P
c
, so the improvement in delamination performance was very large. It should be noted,
however, that the influence on stirrup yield performance by spreading individual ducts apart was
only 5%, and it is often impractical for designers to spread individual ducts apart due to lack of
space in the web, and due to requirements on location of C.G. of the tendon group.

Number and Configuration of Duct Ties

This was evaluated by comparing Configurations 4A, 4B, to Configurations 1, 2, 3. This is
covered by comparing Model/Webs 11M-D to 10M-D, 12M-B to 10M-B, 13M-A-D to 10M-A-
D, 12M-C to 12M-B, and 14M-B, D to 9M-B. This comparison is shown in Table 5-11.

Table 5-11 – Effect of Duct Tie Arrangements – Thin Webs

Model-Web Force at stirrup yield (kips/ft) Difference
10M-B vs. 12M-B 17.22 vs. 17.58 2% change - Config. 3A vs. 4A
10M-C vs. 12M-C 17.13 vs. 17.55 2% change - Config. 3A vs. 4B
10M-A vs. 13M-A 13.06 vs. 15.01 15% change - Config. 3A vs. 4A
10M-B vs. 13M-B 17.22 vs. 19.61 14% change - Config. 3A vs. 4A
10M-C vs. 13M-C 17.13 vs. 19.50 14% change - Config. 3A vs. 4A
10M-D vs. 13M-D 12.97 vs. 15.17 17% change - Config. 3A vs. 4A

12M-B vs. 12M-C 17.58 vs. 17.55 -0.2% change - Config. 4A vs. 4B

9M-B vs. 14M-B 9.59 vs. 10.47 9% change - Config. 2A vs. 4B

Model-Web Force at web delam. (kips/ft) Difference
9M-B vs. 14M-B 10.94 vs. 20.19 85% change - Config. 2A vs. 4B

The conclusions from these comparisons are that web/duct ties make a significant contribution to
the resistance to lateral tendon breakout.
Stirrups

Stirrup spacing was evaluated by comparing Model-Webs 6M-A,B, D to 2M-A,B, D, comparing
7M-A,B, D to 2M-A,B, D, and comparing 13M-A,B, D to 12M-A,B, D. This comparison is
shown if Table 5-12.
Page 110

Table 5-12 – Effect of Stirrup Spacing – Thin Webs

Model-Web Force at stirrup yield (kips/ft) Difference
6M-A vs. 2M-A 8.77 vs. 11.77 34% change with 50% more stirrup steel
6M-B vs. 2M-B 12.57 vs. 13.66 9% change with 50% more stirrup steel
6M-D vs. 2M-D 10.77 vs. 11.82 10% change with 50% more stirrup steel

7M-B vs. 2M-B 16.33 vs. 13.66 -16% change with 33% less stirrup steel
7M-D vs. 2M-D 13.51 vs. 11.82 -13% change with 33% less stirrup steel

13M-A vs. 12M-A 15.01 vs. 13.34 -11% change with 33% less stirrup steel
13M-B vs. 12M-B 19.61 vs. 17.58 -10% change with 33% less stirrup steel
13M-D vs. 12M-D 15.17 vs. 12.38 -18% change with 33% less stirrup steel

Model-Web Force at web delam.(kips/ft) Difference
6M-B vs. 2M-B 13.92 vs. 15.17 9% change with 50% more stirrup steel
6M-D vs. 2M-D 13.04 vs. 14.56 12% change with 50% more stirrup steel

7M-B vs. 2M-B 18.36 vs. 15.17 -17% change with 33% less stirrup steel
7M-D vs. 2M-D 16.37 vs. 14.56 -11% change with 33% less stirrup steel

This indicates that web section strength is significantly influenced by the stirrup spacing only
when web/duct tie reinforcement is NOT used or when the web-splitting/lateral shear-failure
does not occur. In other words, if the failure mode is tending toward local duct pullout, stirrups
are not a very effective deterrent against this failure mode. But if the duct layout and duct ties
are properly detailed to eliminate the local pullout failure mode, the stirrup spacing does define
the web “regional” beam strength.

Concrete Material Properties, Especially Assumed Tensile Strength

This was evaluated by comparing Model-Webs 5M-C to 2M-C, 6M-C to 2M-C, 12M-D to 11M-
D, and 12M-C to 12M-B. This comparison is shown in Table 5-13.

Table 5-13 – Effect of Concrete Strengths – Thin Webs

Model-Web Force at stirrup yield (kips/ft) Difference
2M-C vs. 5M-C 13.64 vs. 15.95 17% change with 50% larger concr. strength

2M-C vs. 6M-C 13.64 vs. 11.21 -18% change with 50% smaller concr. strength

12M-B vs. 12M-C 17.58 vs. 17.55 -0.2% change with 50% larger concr. strength

Model-Web Force at web delam. (kips/ft) Difference
2M-C vs. 5M-C 15.06 vs. 18.30 22% change with 50% larger concr. strength

2M-C vs. 6M-C 15.06 vs. 12.15 -19% change with 50% smaller concr. strength



Page 111
So the web section strength tends to be significantly influenced by the concrete strength only
when the section is prone to web-splitting/local-lateral shear-failures, i.e., when vulnerable duct
placement is used or web/duct tie reinforcement is NOT used. When web/duct tie reinforcement
is used, concrete tensile strength has little effect on the section strength. This tends to further
underscore the importance of providing web/duct tie reinforcement, because of the various
parameters involved in reinforced concrete design, concrete tensile strength has wide variability,
and low reliability. Thus designers should be directed toward design rules that will ensure good
performance regardless of variabilities in concrete tensile strength.

c Local Analysis of Single-cell Box Girders

Model Prototype: Single-Cell CIP Box Girder

A single-cell box configuration was used as shown in Figure 5-24, and with the tendon duct
arrangements and local reinforcements that were shown in Figure 5-17 and further illustrated in
Figure 5-24. The duct and tie configurations are referred to as #6 and #6a, to differentiate from
the configurations of the multi-cell. The only difference between these is that #6 has no duct
ties, and #6a has the two rebar ties as shown. The basic model uses 3d elements and a slice, with
out-of-page thickness equal to one stirrup spacing. The stirrups (and other rebar) are modeled
explicitly, (unlike in two dimensions, where the stirrup is ‘smeared’). This allows introduction
of the out-of-page compression due to prestress. Using this model framework, geometry
variabilities were introduced directly into the models – i.e., one web can have one thickness, and
another have a different thickness, etc. Webs 1 and 2 demonstrate the differences related to web
sloping away from the radius of curvature versus sloping toward the radius of curvature.

21’- 6” 21’- 6”
10’- 0” 11’- 6”
5’-0”
20”
9’- 0” 9’- 0”
3’-6”
~ 3.5”
2” Cl 1.5” Cl
Tendons assumed
6-31 0.6” 1 web
Duc ts 5”00
F
20”
# 4 at 12” Typ
# 7 at 12” Typ
# 6 at 12”

Figure 5-24. Tendon Duct and Local Reinforcement for
the Local Analysis Prototype for a Single-Cell Box

Page 112
Web 1 - left, Web 2 - right. (Duct / Tie configuration “6”. “6a” would include horizontal
crossties.)

Similar to the multi-cell analysis series, in order to establish a baseline for comparison with
design calculations, P
c
was defined as a “Capacity” calculated using conventional means, but
removing resistance factors. This provides the best direct comparison to finite element analyses.
For the webs of the single-cell geometry prototype, P
c
was calculated as follows.

h = 9.67 feet

oM
n
= 42.1 k-ft/ft

Removing the safety factor o = 0.9.

M
n
= 46.8 k-ft/ft

Applying an over-strength factor for rebar strain hardening (which is included in the FE
analysis),

M
o
= M
n
x 1.125 = 52.6 k-ft/ft

As described earlier, moment-fixity effects are approximately 0.6 for interior webs and 0.7 for
exterior webs. Since there are only three ducts distributed vertically, the increase to capacity
caused by load distribution is small, so no capacity increase is applied for this. So the baseline P
c

becomes

P
c
= [52.6 / (h/4)] / 0.7 = 31.1 k/ft


This is used as “100%” in the results tables and plots. The results of the single-cell analyses are
shown in the following tables and figures. The figures are included in Appendix Fa and are
numbered (Fig. B-1, B-2, etc.). The maximum principal strain contours illustrate the general
level of damage to the concrete surrounding the tendon ducts (maximum tensile strain regardless
of orientation).
Page 113

Table 5-14 – Single-cell Box Variations / Parameter Study

Analysis # Model Type
Web
#
Web-
ties
Web
Thickness
Duct/Tie
Config.*
Bundle
Vert. Pos.
Stirrup
Spacing(in.)
Cover
Thickness
Concr.
Tens.
Str.
(x fc`)
1S
"baseline"
Single-cell 1
2
N
N
20
20
6
6
midheight
midheight
12
12
1.5"/2"
1.5"/2"
4
4
2S Single-cell 1
2
N
N
20
20
6
6
1/4 height
bottom
12
12
1.5"/2"
1.5"/2"
4
4
3S Single-cell 1
2
N
N
20
20
6
6
midheight
midheight
12
12
1.5"/2"
1.5"/2"
4
6
4S Single-cell 1
2
N
N
20
20
6
6
midheight
midheight
8
12
1.5"/2"
1.5"/2"
4
2
5S Single-cell 1
2
N
N
20
20
6
6
midheight
midheight
18
12
1.5"/2"
1.5"/2"
4
4
6S Single-cell 1
2
N
N
20
20
6
6
midheight
midheight
12
12
2.5"/3"
3:/3.5"
4
4
7S Single-cell 1
2
Y
Y
20
20
6a
6a
midheight
midheight
12
12
1.5"/2"
1.5"/2"
4
4
8S Single-cell 1
2
Y
Y
20
20
6a
6a
midheight
midheight
12
12
1.5"/2"
1.5"/2"
2
6
9S Single-cell 1
2
Y
Y
20
20
6a
6a
midheight
midheight
18
8
1.5"/2"
1.5"/2"
4
4
10S Single-cell 1
2
Y
N
20
20
6a
6a
1/4 height
bottom
12
12
1.5"/2"
1.5"/2"
4
4


Single-Cell Models – Analysis Results

The results of the 10 different single-cell girder local models allow for qualitative and
quantitative assessment and comparison of the cases analyzed.

- Lateral Force vs. Deflection of Web Mid-height
- Lateral Force vs. Deflection of Web Quarter-height
- Maximum Principal Strain Contours in Concrete at 75%, 100%, 125%, 150% Pc
- Strains in stirrup rebar at 3 locations along duct bank at 75%, 100%, 125%, 150% Pc
- Distortions (change in web width) 3 locations along duct bank 75%, 100%, 125%, 150%
Pc

Detailed results are included in Appendix F-a.

Table 5-15 – Deflections (inches) Measured at Mid-height of Webs on Backside

Web 1 Web 2
Percent Mid Quarter Mid Quarter
Model # Capacity

1S 75% 0.176 0.174 0.179 0.158
100% 0.404 0.402 0.418 0.360
125% 1.076 1.090 1.208 0.978
150% 2.496 2.499 2.903 2.224
Page 114
2S 75% 0.888 1.251 0.724 1.068
100% 2.340 3.365 1.924 2.852
125% 4.521 6.484 3.647 5.410
150% 26.702 36.174 22.345 31.707

3S 75% 0.156 0.152 0.135 0.131
100% 0.366 0.357 0.338 0.307
125% 0.918 0.906 0.871 0.778
150% 2.239 2.161 2.246 1.868

4S 75% 0.272 0.274 0.314 0.256
100% 0.678 0.681 0.764 0.609
125% 1.805 1.775 2.050 1.570
150% 4.030 3.904 4.542 3.445

5S 75% 0.140 0.140 0.159 0.137
100% 0.295 0.298 0.351 0.293
125% 0.685 0.701 0.961 0.722
150% 1.728 1.815 2.526 1.835

6S 75% 0.195 0.190 0.192 0.170
100% 0.462 0.457 0.463 0.405
125% 1.212 1.204 1.288 1.066
150% 2.805 2.691 2.974 2.362

7S 75% 0.168 0.168 0.174 0.154
100% 0.372 0.377 0.400 0.355
125% 0.953 0.987 1.056 0.923
150% 2.286 2.335 2.600 2.169

8S 75% 0.205 0.192 0.155 0.154
100% 0.435 0.420 0.357 0.341
125% 1.095 1.046 0.866 0.846
150% 2.588 2.447 2.127 1.996

9S 75% 0.146 0.148 0.182 0.151
100% 0.320 0.335 0.472 0.365
125% 0.849 0.915 1.409 1.011
150% 2.102 2.292 3.413 2.463

10S 75% 0.808 1.131 0.659 0.972
100% 2.301 3.242 1.877 2.785
125% 4.462 6.274 3.493 5.187
150% 23.358 31.548 19.232 27.371

Page 115

Table 5-16 – Stirrup Strain (%) on Curve Inside - Face

Web 1 Web 2
Percent Duct 2 Duct 4 Duct 6 Duct 2 Duct 4 Duct 6
Model # Capacity
1S 75% 0.00166 0.00137 0.00097 0.00150 0.00150 0.00150
100% 0.00421 0.00222 0.00173 0.00310 0.00310 0.00310
125% 0.01851 0.01023 0.00275 0.01506 0.01506 0.01506
150% 0.04107 0.03204 0.01021 0.04032 0.04032 0.04032

2S 75% 0.01429 0.00345 0.00187 0.00680 0.00311 0.00188
100% 0.03765 0.01701 0.00450 0.02543 0.01638 0.00724
125% 0.07417 0.04288 0.01750 0.04969 0.03558 0.02096
150% 0.24986 0.15104 0.06873 0.21465 0.10079 0.05699

3S 75% 0.00152 0.00127 0.00090 0.00158 0.00097 0.00049
100% 0.00376 0.00218 0.00176 0.00326 0.00236 0.00164
125% 0.01529 0.00902 0.00279 0.01341 0.00712 0.00421
150% 0.04083 0.03034 0.00937 0.03334 0.02640 0.01838

4S 75% 0.00332 0.00190 0.00134 0.00165 0.00172 0.00175
100% 0.01269 0.00563 0.00244 0.00361 0.00580 0.00520
125% 0.03219 0.02447 0.00713 0.01271 0.02075 0.02036
150% 0.06683 0.05776 0.01953 0.03541 0.04853 0.04659

5S 75% 0.00126 0.00112 0.00076 0.00140 0.00141 0.00110
100% 0.00189 0.00177 0.00137 0.00244 0.00332 0.00187
125% 0.00942 0.00452 0.00200 0.01017 0.01499 0.00759
150% 0.02689 0.01755 0.00329 0.03406 0.03425 0.03028

6S 75% 0.00178 0.00137 0.00109 0.00150 0.00132 0.00113
100% 0.00642 0.00204 0.00190 0.00415 0.00284 0.00215
125% 0.02163 0.00915 0.00323 0.01628 0.01319 0.00910
150% 0.04799 0.02955 0.00911 0.03942 0.03223 0.02922

7S 75% 0.00159 0.00135 0.00092 0.00160 0.00140 0.00124
100% 0.00329 0.00215 0.00160 0.00525 0.00335 0.00259
125% 0.01451 0.01098 0.00257 0.01872 0.01541 0.00807
150% 0.03987 0.03238 0.00851 0.04636 0.04096 0.02690

8S 75% 0.00162 0.00149 0.00134 0.00154 0.00106 0.00059
100% 0.00387 0.00339 0.00277 0.00407 0.00243 0.00158
125% 0.01690 0.01528 0.01066 0.01612 0.00684 0.00332
150% 0.04019 0.03653 0.02766 0.04088 0.02560 0.01352

Page 116
9S 75% 0.00129 0.00110 0.00072 0.00207 0.00174 0.00151
100% 0.00207 0.00178 0.00129 0.00841 0.00630 0.00387
125% 0.01140 0.00698 0.00196 0.02662 0.02184 0.01958
150% 0.03106 0.02321 0.00373 0.06473 0.05138 0.04844

10S 75% 0.01253 0.00783 0.00210 0.00703 0.00380 0.00184
100% 0.03730 0.02768 0.00920 0.02789 0.01708 0.00750
125% 0.07189 0.05973 0.02640 0.05278 0.03636 0.01851
150% 0.23078 0.19175 0.09182 0.19842 0.09676 0.04984


Page 117

Table 5-17 – Distortion (Web Thickness Changes – inches) at Mid-height of Ducts

Web A Web B
Percent Duct 2 Duct 4 Duct 6 Duct 2 Duct 4 Duct 6
Model # Capacity

1S 75% 0.0126 0.0104 0.0026 0.0128 0.0247 0.0095
100% 0.0351 0.0297 0.0053 0.0442 0.0630 0.0316
125% 0.0732 0.0666 0.0179 0.0987 0.1464 0.0876
150% 0.1398 0.1446 0.0673 0.1702 0.3469 0.2093

2S 75% 0.0169 0.0278 0.0064 0.0472 0.0849 0.0226
100% 0.0635 0.0779 0.0303 0.1207 0.2037 0.0671
125% 0.1959 0.2289 0.1114 0.2546 0.4035 0.1122
150% 0.4625 0.4472 0.2166 0.4946 0.7850 0.2093

3S 75% 0.0165 0.0084 0.0031 0.0051 0.0108 0.0034
100% 0.0348 0.0277 0.0092 0.0176 0.0455 0.0137
125% 0.0658 0.0512 0.0196 0.0374 0.0875 0.0412
150% 0.1347 0.1356 0.0813 0.1102 0.2152 0.1084

4S 75% 0.0245 0.0141 0.0039 0.0339 0.0496 0.0332
100% 0.0569 0.0392 0.0132 0.0928 0.1219 0.0903
125% 0.1079 0.0971 0.0463 0.1995 0.2876 0.2064
150% 0.2377 0.2341 0.1441 0.4056 0.5930 0.4169

5S 75% 0.0077 0.0094 0.0031 0.0112 0.0223 0.0087
100% 0.0247 0.0212 0.0058 0.0356 0.0521 0.0250
125% 0.0459 0.0462 0.0133 0.0862 0.1145 0.0600
150% 0.0734 0.0898 0.0346 0.1594 0.3204 0.1824

6S 75% 0.0158 0.0129 0.0021 0.0106 0.0235 0.0048
100% 0.0366 0.0328 0.0077 0.0406 0.0632 0.0175
125% 0.0935 0.0748 0.0302 0.0965 0.1555 0.0424
150% 0.1985 0.2021 0.1147 0.2057 0.3781 0.1055

7S 75% 0.0106 0.0093 0.0026 0.0104 0.0217 0.0075
100% 0.0213 0.0237 0.0053 0.0274 0.0479 0.0153
125% 0.0425 0.0405 0.0085 0.0573 0.0939 0.0261
150% 0.0567 0.0824 0.0253 0.1254 0.2166 0.0563

8S 75% 0.0936 0.0157 0.0069 0.0106 0.0173 0.0052
100% 0.0207 0.0273 0.0164 0.0246 0.0503 0.0127
125% 0.0413 0.0574 0.0461 0.0442 0.0872 0.0250
150% 0.0735 0.1541 0.0933 0.0922 0.1443 0.2584

9S 75% 0.0057 0.0095 0.0032 0.0390 0.0227 -0.0218
100% 0.0189 0.0190 0.0062 0.0298 0.0480 0.0126
125% 0.0345 0.0361 0.0139 0.0630 0.1241 0.0217
150% 0.0386 0.0645 0.0274 0.1379 0.3115 0.0543
Page 118
10S 75% 0.0132 0.0250 0.0064 0.0399 0.0661 0.0072
100% 0.0124 0.0600 0.0154 0.0800 0.1449 0.0163
125% 0.0751 0.1321 0.0279 0.1604 0.2918 0.0202
150% 0.1939 0.2563 0.0306 0.2885 0.5176 0.0376



Discussion of Results

Analyses of these models showed similar trends as the multi-cell model analyses, and there were
no surprises as to the performance of the sections. As expected, the sections with duct ties
performed better than without. Having the double row of tendons was found to concentrate the
local damage area within the web, but having the 20-inch web thickness with local reinforcement
is quite adequate to accommodate this.

Similar to the multi-cell studies, the analysis results can be used to compare the web design
parameters. For purposes of interpreting the 3d finite element analysis results, the following
damage limit criteria is suggested:

- Stirrup rebar strain exceeding yield (i.e., 0.2% strain for Grade 60 steel); note that for
Load Factor Design, concrete reinforcement is designed to yield, yield should be
considered an upper bound criteria for unfactored loads.

- Visible concrete cracking occurs at strains of approximately 0.016%, but this is not
necessarily web failure; concrete with maximum principal strains of 0.3% can be
considered to be heavily cracked. Concrete with strains in excess of 1.0% will generally
show wide-open cracks and potential spalling from the section.

- Significant distortion or delamination (change of width of the webs) also represents an
upper limit on capacity for webs; the delamination is evidence of a local splitting or
lateral shear failure within the web; it was again assumed that an upper bound on crack
width of 1/16” as an indicator of such a failure. For 20” webs, this represents a distortion
ratio (average strain through the section) of 0.3%. For sections with web ties, this means
the web ties have yielded; for sections without web ties, the section is at a web splitting
or a cover concrete spalling condition.

One of the criteria, Stirrup Yield, has been summarized in Table 5-18. These are the total forces
(sum of all tendon ducts in the web) applied when any part of the stirrup reaches yield.
Page 119

Table 5-18 – Force at Stirrup Yield (0.2% Strain)


Total Force (K/ft)
Model # Web 1 Web 2
1S 25.83 28.18
2S 16.05 17.88
3S 26.21 28.6
4S 20.1 26.17
5S 31.18 28.77
6S 24.16 28.04
7S 26.21 26.64
8S 26.12 27.95
9S 29.87 23.78
10S 16.09 18.38

Using these criteria, and examining the results tables and plots has resulted in the following
observations.

Web Slope

Similar to the multi-cell series, the inside radius web, sloping toward the center of the curve was
found to be roughly 10% stronger than the outside radius web, sloping away from the center of
the curve. After further study, a possible reason for this is that loading the inside radius web
(Web 2) creates positive transverse moment in the top slab adjacent to the web (tension on the
bottom of the slab), whereas loading Web 1 creates negative moment. The slab resistance to
these moments is stronger (about 2 times stronger, based on typical deck reinforcing) in positive
moment than in negative moment, and this translates to more strength in the associated web.

Cover Thickness

Cover thickness was varied in model 6S, where increases for Web 1, 2 were by 50% and 75%,
respectively. The following summarizes the relevant strength comparisons:

Table 5-19 – Effect of. Cover Thickness – Thick Webs

Model-Web Force at Stirrup Yield (kips) Difference
6S-1 vs. 1S-1 24.16 vs. 25.83 7% increase with 3” vs. 2”
6S-2 vs. 1S-2 28.04 vs. 28.18 0% increase with 3.5” vs. 2”


The local concrete damage was less severe with thicker cover, but the comparisons for stirrup
yield are inconclusive. This is because a) for 2” cover and above, cover failure does not control
the failure mode, and b) when the cover is less, the “moment arm” between the stirrups is more,
and this increases capacity rather than decreasing it.
Page 120

Number and Configuration of Tendon Ducts

The only variations studied in the single cell case was the positioning of the duct group, studied
in Configurations 2S and 10S. The number and relative position of the ducts to each other was
held constant. But it was again observed that when the ducts occurred near the bottom of the
web (either “quarter-height” or “bottom” as was tested in Configurations 4M and 14M) the force
at “failure” is substantially lower than when the ducts are placed at the mid-height, i.e., on
average as much as 25% to 40% lower when comparing these cases to other similar cases. The
reason for this is a tendency toward lateral shear failure of the overall web, and a tendency
toward flexural damage in the top slab, thus weakening the whole system. When the ducts are
located at the mid-height, the lateral shear is divided equally between the top and bottom of the
web. But when the ducts move down, the bottom of the web carries most of the lateral shear. So
this is a different mechanism than the failure modes observed for tendon ducts at mid-height, but
one that still warrants consideration in design.

Number and Configuration of Duct Ties

This was evaluated by comparing duct tie Configurations 6a, which had duct ties, to 6, which
had no duct ties. Table 5-20 compares Model/Webs 7S to 1S, 8S to 3S, and 9S-1 to 5S-1.

Table 5-20 – Effect of Duct Ties – Thick Webs

Model-Web
Force at Stirrup Yield (kips) (and
delamination at 100%P
c
)
Difference
26.21 vs. 25.83 2% incr. with Duct ties
7S-1 vs. 1S-1
(0.024” vs. 0.035”) (31% less delamination)
26.64 vs. 28.18 -5% change with Duct ties
7S-2 vs. 1S-2
(0.048” vs. 0.063”) (24% less delamination)
27.95 vs. 28.60 -2% change with Duct ties
8S-2 vs. 3S-2
(0.050” vs. 0.046”) (little change in delamination)
29.87 vs. 31.18 -4% change with Duct ties
9S-1 vs. 5S-1
(0.019” vs. 0.025”) (24% less delamination)

These comparisons show that for the wider web (20”) and double row of tendon ducts, the ties do
not make a significant difference in the force to cause stirrup yield, but they do make a large
difference in the delamination-damage that can occur within the web. Delaminations (width
changes in the web) are reduced by 24-31% with duct ties compared to without duct ties.

Stirrups

Stirrup spacing was evaluated by comparing Model-Webs 4S-1 to 1S-1, comparing 5S-1 to 1S-1,
and comparing 9S-1, 2 to 7S-1, 2. The results are shown in Table 5-21
Page 121

Table 5-21 – Effect of Stirrup Spacing – Thick Webs

Model-Web
Force at stirrup
yield (kips)
Difference
4S-1 vs. 1S-1 20.1 vs. 25.83 29% increase with 50% more stirrup
steel
5S-1 vs. 1S-1 31.18 vs. 25.83 21% decrease with 33% less stirrup
steel
9S-1 vs. 7S-1 29.87 vs. 26.21 14% decrease with 50% less stirrup
steel
9S-2 vs. 7S-2 23.78 vs. 26.64 21% increase with 50% more stirrup
steel

So the web section strength tends to be significantly influenced by the stirrup spacing for this
geometry also, perhaps even more so than for the multi-cell geometry. Here again, stirrup
spacing is a driver of web “regional” beam strength.

Concrete Material Properties, Especially Assumed Tensile Strength

The effect of concrete strength was evaluated by comparing Model-Webs 3S-2 to 1S-2, 4S-2 to
1S-2, and 8S-1, 2 to 7S-1, 2. The results are shown in Table 5-22

Table 5-22 – Effect of Material Strength – Thick Webs

Model-Web
Force at stirrup
yield (kips)
Difference
3S-2 vs. 1S-2 28.60 vs. 28.18 2% increase with 50% larger
concrete tensile strength
4S-2 vs. 1S-2 26.17 vs. 28.18 7% decrease with 50% smaller
concrete. strength
8S-1 vs. 7S-1 26.12 vs. 26.21 0% change with 50% smaller
concrete. strength
8S-2 vs. 7S-2 27.95 vs. 26.64 5% increase with 50% larger
concrete. strength

So, repeating the trend observed in the multi-cell analysis, the web section strength tends to be
only marginally influenced by the concrete strength, and mostly this influence occurs when
web/duct tie reinforcement is NOT used. When web/duct tie reinforcement is used, concrete
tensile strength has less effect on the section strength. Of the various parameters involved in
reinforced concrete design, concrete tensile strength has wide variability, and low reliability. So
designers should use design rules that will ensure good performance regardless of variabilities in
concrete tensile strength.
Page 122

d Conclusions from Local Analyses

General observations on “Capacity”

Using the “capacity” definitions described in this section (P
c
, developed based on regional
transverse bending considerations), it was found that all of the multi-cell box girders achieved
this target capacity. The baseline (Model 1M) interior webs achieved it marginally (i.e., stirrup
yield was reached at 107% of P
c
), while stronger details which utilize spreading apart the ducts,
adding duct ties, or moving the ducts toward the curve-outside-face of the web reached as high
as 185% of P
c
. The variations in force to cause local duct bank breakout (either local shearing of
web delamination) were even larger, depending on the detailing used, so it is clear that the
detailing has very significant influence on resistance to lateral pullout. For the single-cell
example, with the 20-inch webs and double row of ducts, the finite element analysis showed
capacities which were mostly lower than the hand-calculated regional transverse bending
capacity (i.e., stirrup yield was reached at a range from 52% P
c
up to 100% P
c
), but this is
explained by the fact that for the thicker web, failures were dominated by local lateral shearing.

Summary of Influences from Detailing Parameters

Web Depth - can be adequately accounted for by considering and designing for web moments.

Web Thickness – has significant influence on resistance to regional transverse bending and
tendon pullout caused by web thickness. For stirrup yield, capacity formulae based on regional
flexure considerations appear to be appropriate for design.

Web Slope - the sloped webs were found to be significantly weaker (roughly 30%) than the
vertical webs, but much of this difference is caused because these are exterior webs rather than
interior ones. Exterior webs have more flexible end conditions at their connection with the top
and bottom slab, and this produces larger mid-height moments. Comparison of Webs A to D for
the inclined webs show that Web A is generally weaker than D by about 10%. It is believed this
is due to the difference in positive bending versus negative bending strength of the top slab.
Lateral force for Web D applies positive moment to the top slab, and the positive moment
reinforcement is approximately 2 times that of the negative moment reinforcement.

Cover Thickness – Inside face duct cover influences lateral pullout resistance, but it is not the
only driver of pullout resistance. The results of the parameter studies are influenced by the fact
that when the cover is reduced, for the same overall web thickness, the moment arm for the
stirrups is increased, which is an off-setting influence on pullout resistance. It appears
appropriate to check cover concrete thickness for resistance to initial cracking, but not to include
cover concrete tensile strength in the calculation of regional transverse bending strength.

Number and Configuration of Tendon Ducts - when ducts are spread apart, the performance
significantly improves. Roughly 20% resistance force improvement was demonstrated by
separating the 5-duct bundle into two bundles, and an additional 4% improvement was
demonstrated by spreading the bundles further apart (4.5” versus 1.5” separation). It is believed
that a prudent recommendation is to require a maximum of 3 ducts per bundle. When individual
ducts were separated and moved toward the curve’s outside face of the web, performance further
Page 123
improves. When measured by the delamination/local-lateral shear criteria, Duct Configuration
3A exceeded 200% P
c
, so the improvement in delamination performance was very large. It
should be noted, however, that it is often impractical for designers to spread individual ducts
apart due to lack of space in the web, and due to requirements on location of the C.G. of the
tendon group.

Number and Configuration of Duct Ties - make a significant contribution to the resistance to
lateral tendon breakout.

Stirrups –regional transverse bending strength is directly tied to stirrup area, but it controls the
design only when web/duct tie reinforcement is NOT used or when the web-splitting/lateral
shear-failure does not occur. In other words, if the failure mode is tending toward local duct
breakout, stirrups are not a very effective deterrent against this failure mode. But if the duct
layout and duct ties are properly detailed to eliminate the local pullout failure mode, the stirrup
spacing does define the web “regional” beam strength.

Concrete Material Properties, Especially Assumed Tensile Strength - web section strength
can be significantly influenced by concrete tensile strength only when the section is prone to
web-splitting/local-lateral shear-failures, i.e., when vulnerable duct placement is used or
web/duct tie reinforcement is NOT used. When web/duct tie reinforcement is used, concrete
tensile strength has little effect on the section strength. Thus designers should be directed toward
design rules that will ensure good performance regardless of variabilities in concrete tensile
strength.

Recommendations for Web Capacity Design

Web capacity design for lateral tendon force resistance should be a three-step calculation:
Regional flexure check, local-lateral shear/breakout check, and cover concrete cracking check

Regional transverse bending

The regional mechanism is the web acting as a vertical beam loaded laterally near its center.
Fundamentally, the calculation follows the equation:

M
u
= (Load Factor) (Moment Fixity Factor) (1/4) (Pj/R) h
c


This equation (a modified version of the Caltrans Equation) and the corresponding stirrup
spacing should be evaluated for each web of a box girder separately – not for the total box
divided by the number of webs. The radius is different for each web, and it was found that the
Moment Fixity Factor is also different. AASHTO LRFD currently applies a load factor of 1.2 to
the Pjack tendon force, which is judged to be reasonable. Appropriate moment fixity factors are
0.6 for interior webs and 0.7 for exterior webs.

The stirrup sizing and spacing should then be calculated using Ultimate Strength design such
that:


u n
M M >
Page 124

However, the V
s
stress in the stirrups due to vertical shear in the web should be added to the
stress due to flexure in the sizing and spacing of the stirrups. At the midheight of the web, on the
inside-curve side of the web, these stresses are directly additive.

Local Lateral shear Check

The local lateral shear mechanism involves the complex behavior that develops in the concrete
and stirrup region immediately adjacent to the duct bank. This may be checked by the following
equations developed by the University of Texas (Van Landuyt, 1991):

For a strip of web one foot long, the applied lateral shear demand along a plane d
eff
long is:

V
d
= P
j
/R ÷ 2

V
c
capacity of the cover-beam along this plane may be taken as:

V
c
= 24d
eff
c
f '

Where 75 . 0 = (reduced due to uncertainties in concrete quality within the cover-beam)

When the spacing between ducts is greater than or equal to the duct diameter:

Duct d d
c eff
( + = 4 / .) Diam + ¯s/2

or

d
eff
= t
w
– (Duct Diam)/2

whichever is least.

where:

s = space between ducts (assume 0 if s< 1.5” or for single ducts)
t
w
= thickness of web

When the spacing between ducts is less than the duct diameter or for single ducts:

d
eff
= d
c
+ (Duct Diam)/4

where d
c
= cover over the ducts
Figure 5-25 shows what is intended by the above equations for d
eff
.

Page 125


Figure 5-25 Definition of d
eff
(after Van Landuyt, 1991)

There has been discussion within the industry as to selecting d
eff
(some refer to this as the
“lateral shearing plane depth). Some say this should be no greater than d
c
(the cover concrete
depth) due to uncertainties in the concrete interaction with the ducts, but the local analyses
conducted here allow for the extra width of ¼ of a duct diameter.

If this lateral shear is exceeded, the most effective design remedy is the addition of duct-tie
reinforcement.

Cover Concrete Cracking Check

Evaluating the cracking of the cover concrete is a check that is made to insure serviceability
since it is recommend that the lateral tendon forces be completely carried by the strength
elements of the above two checks. But this serviceability check remains critical to achieving a
good design, because significant cover cracks running along the tendons should be avoided for
long term structure durability.

The flexure on the cover beam involves a complex mechanism because it is uncertain what the
level of adhesion is of the cover concrete to the duct bank and to the concrete surrounding the
duct bank. Assuming there is no adhesion between the metal duct and the web concrete in the
radial direction of the duct, the flexure calculation proceeds as follows. The cover-beam acts as
a vertical beam “built-in” or fixed at top and bottom. Thus the following moments are produced:
Page 126

12 / ) / Pj/R (
12
2
2
L L
wL
M
ends
= =
24
2
wL
M
center
=

L is the height of the duct bank

12
3
c
bd
= I

and
u n
M M >

Where M
n
is defined by an allowable tensile stress for concrete of
c
f ' 5 , and o = 0.55. The
allowable tensile stress should also be reduced by the tensile stress in the concrete at the critical
point due to regional transverse bending. While this may appear to be quite conservative in terms
of choice of tensile strength and choice of o, it should be noted that once cracking begins within
the interior of the cover concrete near the top and bottom of the duct bank, the moment at the
center of the duct banks quickly becomes

8
2
wl
M
center
=

So these factors and conservative tensile strengths are judged to be appropriate to prevent this
progressive cracking mechanism from occurring.

Other Local Detailing Guidelines

A further guideline, which has come out of the local analysis work, and from examination of
some local breakout failures in various bridges and test structures, is to limit the number of ducts
of a “sub-bundle” to no more than three. Sub-bundles should then be separated by either a duct-
tie rebar, or by a minimum of 1/3 of one duct diameter (for example, 1-1/2” for the analyses
performed here).

Duct ties should be well anchored with hooks around stirrup reinforcement. A generic duct tie
detail is shown in Figure 5-26


Page 127


Figure 5-26 – Generic Web and Duct Tie Detail

Construction Tolerances

Designers should consider the practical aspects of construction tolerances when checking and
implementing their designs. Construction tolerances should be held to industry standards, and it
is not the point of this design recommendation to modify these, but designers may wish to
consider conservatively allowing for field variations in web width and in rebar placement of up
to ± 0.5” when evaluating issues of web regional transverse bending strength, local breakout
resistance, and particularly cover-beam strength. Dimensional changes of 0.5” can make
considerable difference in the stresses in the web concrete and reinforcing steel.
The following is an example of how design and construction issues can affect conditions for
lateral tendon breakout. As a box girder gets deeper, the stirrup cage gets deeper. As the stirrup
cage gets deeper, it becomes more flexible laterally, especially is areas of low lateral shear
demand where designers often specify stirrup spacing as large as 24 inches. During the web and
soffit pour, the stirrup cage has been shown to deflect laterally within the web form due to
unbalanced concrete placement, vibration process, and duct float. Duct float, in combination
with sloped exterior webs, can often lead to a substantial reduction in concrete cover between the
stirrup cage and the interior face of web. This may be mitigated somewhat by rebar spacer
requirements at mid-height of webs to help control stirrup movement during the pour, but the
designer should be aware of possible variations in the actually constructed dimensions.
Page 128
Several conditions can aggravate the chances for lateral tendon breakout. These include:

1. Reduction of cover over the duct or rebar that can affect resistance to breakout.
2. Excessive wobble of the ducts that can result in either reduced resistance to
breakout or locally elevated lateral forces.
3. Out-of-plane forces in a vertically curved tendon due to wedging of the stand
4. Pressure from grout leakage due to poor quality duct (excessive flexibility),
damaged duct, or improperly sealed duct
5. Distortion of empty ducts acted on by adjacent stressed ducts
6. Local curvature in ducts near anchorage zones or blisters

The specified load and resistance factors (1.2 and 0.75) are based on the assumption that
construction tolerances are reasonably well controlled. If this may not be the case, the designer
may wish to consider one of the following three options.

1. Use higher load factors and/or lower resistance factors. Some engineers familiar
with the potential problems have recommended o factors be reduced from 0.75 to
0.55 for local lateral shear failure. Load factors could also be raised above 1.2 to
say 1.5.
2. Use dimensions that include an allowance for misplacement of the duct, rebar, or
forms. As suggested above, critical dimensions could be reduced by 0.5” or even
1.0”
3. When in doubt, provide web and duct tie reinforcement

Tendon breakout failures can be expensive to repair. Although the recommended design
specifications should provide an adequate factor of safety in most cases, the designer is
ultimately responsible for assessing the likely conditions in the field.

Page 129
6. CONCLUSIONS

Several critical issues relative to the design of concrete box girder bridges were identified at the
beginning of the project. Methods exercised in this study that were intended to address these
issues consisted of a survey of the state of practice, a review of published literature, analytical
studies of global and local response, discussions of the experienced research team that attempted
to reach a consensus on critical design requirements, and a review by an advisory panel with
expertise in this area of study.
Several conclusions were drawn from the research conducted in this project. In many cases these
conclusions have translated into recommended AASHTO LRFD Specification provisions as
presented in Appendix A. Analysis guidelines were also developed to assist designers in
performing response analysis. These are provided in Appendix C. Other conclusions found that
current design practice was adequate and did not require a change. The following paragraphs
discuss conclusions relative to the critical issues defined at the beginning of the project.
- Applicability –Curved concrete box girder bridges are used throughout the United States.
Most modern bridges of this type are prestressed. A review of the state of practice in the United
States found that both single and multi-cell box girder bridges are widely used. The predominate
construction type in some west coast states is multi-cell box girder bridges cast-in-place on
falsework. This type is also widely used throughout the United States. Single cell box girder
bridges are also common but tend to dominate the type of box girder construction used on the
east coast. East coast construction also uses more pre-cast segmental construction than is used on
the west coast where cast-in-place construction is more dominate, even when segmental methods
are used. Some states do not use this type of bridge on a regular basis. Curved spread box beams
are an emerging structure type but are not widely used at this time.
Both single and multi-cell concrete box girder bridges are covered by this project. They may be
cast-in-place or pre-cast and may be constructed segmentally or on falsework.
- Appropriate levels of analysis and design – Selecting the type of global analysis that
should be used for curved concrete box girder bridges is one of the most important issues
addressed by this project. Published research shows that these types of bridges are most
accurately analyzed using three-dimensional finite element or similar techniques. Unfortunately,
these analysis methods are tedious and in general not practical for production design work. Also,
in many cases, more simplified analysis methods will produce acceptable results. To determine
the range of applicability of various analysis methods, a detailed global analysis study was
undertaken
The first step was to identify a more simplified three-dimensional analysis approach that would
yield results comparable to the more detailed finite element technique. This was accomplished
with the grillage analogy approach. In this method, the bridge is simulated as a grillage of beam
elements in the longitudinal and transverse direction. Guidelines for preparing the computer
model, performing the analysis and interpreting results were developed and are included in
Appendix C. From the designer’s point of view this analysis method has advantages over the
finite element approach. Besides being a smaller and less computationally intense analytical
model, the grillage analogy produces results in terms of the structural members commonly
considered by the designer. This makes it easier to design these elements, whereas the finite
Page 130
element approach would involve considerable post processing of analytical results to accomplish
the same goal.
Secondly, the limits of applicability of three analytical approaches were assessed. The three
methods considered were:
1. Plane Frame Analysis – This allows the bridge to be analyzed as if it were straight.
2. Spine beam Analysis – This is a space frame analysis in which the superstructure is
modeled as a series of straight, chorded beam elements located along the centerline of the
superstructure.
3. Full Three-dimensional Analysis – This includes several different sophisticated
approaches that include the grillage analogy described above as well as the finite element
and other sophisticated approaches.
Extensive parameter studies were performed that included the effects of structural framing
(simply supported or continuous), span lengths, radius of curvature, cross-section (includes
bridge width), and bearing configuration on the response of bridges. These studies included both
grillage analysis and spine beam analysis for which plane frame analysis constituted the case of a
bridge with a very large radius. These studies showed that the radius to span length ratio as
represented by the central angle between two adjacent supports was the dominant parameter that
determined the accuracy of the various analysis methods. The span length to width ratio (aspect
ratio) of the superstructure also had a minor effect. Based on these parameter studies, the
following limits for the various types of analysis are recommended:
1. For central angles less than or equal to 12 degrees – Plane frame analysis is acceptable
2. For central angles between 12 and 46 degrees and an aspect ratio above 2.0 – spine beam
analysis is required.
3. For central angles between 12 and 46 degrees and an aspect ratio less than 2.0 –
sophisticated 3-D analysis is required.
4. For central angles greater than 46 degrees – sophisticated 3-D analysis is required.
5. For all bridges with otherwise unusual plan geometry - sophisticated 3-D analysis is
recommended.
- Section properties and member stiffness – The section properties and member stiffnesses
that should be used in the spine beam analysis and the grillage analogy analysis are critical and
are discussed in the analysis guidelines presented in Appendix C. For the spine beam analysis the
cross-sectional area and the three rotational moments of inertia are important. In the case of the
grillage analogy, all six section properties of each beam member are required. Special formulae
for some of these section properties are used to simulate various aspects of the behavior of a
curved concrete box girder bridge. This in turn requires special interpretation of some of the
results.
- Critical position of live loads – The number and position of the live load lanes in the
transverse direction as well as their position along the longitudinal axis of the bridge is critical
for curved concrete box girder bridges. Given the number of possible load positions, it will be
desirable to use the live load generating capabilities of sophisticated commercially available
software to rigorously envelope the live load response. This can be a daunting task if such
software is not used. Fortunately, the whole-width design approach as described in the LRFD
Page 131
specifications was shown to yield conservative results when used in conjunction with the plane
or spine beam approaches. This will greatly simplify the effort of the designer in determining
live load response. When using this approach, it is important to distinguish between the
longitudinal response along each of the webs and the affect of torsion across the whole section.
When torsional response is being assessed by the whole-width design approach, the number of
live load lanes should be reduced to the actual number of lanes that can fit on the cross-section
and adjusted by the multiple presence factor and dynamic load factor (for truck loading only).
When a three-dimensional model of the bridge (either a spine beam or grillage analogy model) is
being used it is important to consider the transverse position of prestress tendons. Note that the
length of the various tendons will have an effect on friction losses and that tendons will also
produce a transverse response in the bridge superstructure.
Vehicular impact may be assessed using the method prescribed in the LRFD Bridge Design
Specifications. Other load conditions such as centrifugal forces, breaking or acceleration forces,
wind, etc. should be determined according to the LRFD Specifications and then applied to the
spine beam model. If the plane frame approach is being used then these loads may be analyzed in
the same manner as if the bridge were straight. The effect of bridge superelevation can usually be
ignored.
- Torsion design – The design of concrete box sections for torsion is covered in the current
AASHTO LRFD Specifications. However, some clarification of these requirements is in order.
These were discussed in the review of published literature included in Chapter 3.
Torsion demands usually translate to additional lateral shear demands in the webs of concrete
box girders. These may be determined from both the spine beam and grillage analogy methods.
In the case of the spine beam (i.e. spine beam) analysis, the torsion demands are taken directly
from the torsion forces generated in the spine beam. These forces must be transformed into shear
flow around the perimeter of the box section. This shear flow will increase the effective shear in
one web while decreasing it in another. Webs should be designed for the combined flexural and
torsional shear.
In the case of the grillage analogy, the effects of torsion on web shear are partially accounted for
because each web is explicitly included in the analytical model. However, because of the way
torsional stiffness of the superstructure is distributed to the individual longitudinal members of
the grillage model, the total effect of torsion on the entire cross-section is not completely
accounted for by the longitudinal member shear demands. To correct for this deficiency it is
necessary to consider the torsional forces in each of the longitudinal members at a given
longitudinal location in the grillage model and apply the sum of these torsions to the entire cross-
section to obtain residual shear flow about the perimeter of the section. This is done in a manner
similar to that used for the spine beam. When this residual shear flow is combined with the
flexural shear in the extreme longitudinal members, the correct demands to be used for web shear
design are obtained.
The procedures to be used for torsion design for both the spine beam and grillage analogy
analysis methods are illustrated in the example problem included in Appendix B.
- Tendon Breakout – Extensive analytical studies were performed to investigate lateral
bursting stresses in curved concrete box girder bridges with internal prestress tendons. The first
step in these studies was to verify that the nonlinear finite element models used could accurately
predict lateral tendon breakout behavior observed in experimental studies performed at the
Page 132
University of Texas. The results of the analyses compared well with the experimental results so
that there is confidence that both the experimental and analytical results have yielded accurate
results.
Based on parameter studies conducted using these verified nonlinear finite element techniques,
and the results of the University of Texas studies, modifications to the specifications for
considering in-plane force are recommended. These include:
1. A method for assessing the local lateral shear resistance to pullout – These provisions
are the recommendations from the University of Texas, which were further verified by
the nonlinear finite element parameter studies conducted as part of this study. They also
include provisions for considering the effect of construction tolerances, which have been
shown by past failures to have a significant effect on web performance and are discussed
in Chapter 5.
2. A method for checking flexural cracking of the unreinforced concrete cover over the
inside of the prestress tendons – This is a new requirement that applies only to tendons
that are vertically stacked. It is included to prevent maintenance, architectural and
structural problems that can arise due to longitudinal cracking of the web. The results are
used to determine the need for web and duct tie reinforcement. Vertical duct stacks are
limited to three tendons high and concrete cover over the inside of the ducts should be
maximized. Generic web and duct tie reinforcement details are included in the
commentary.
3. A method for calculating the regional transverse bending moments within a web – These
moments result from the regional transverse bending of a web between the top and
bottom slab of the bridge due to lateral prestress forces. When combined with global
forces such as flexural shear and torsion, regional transverse bending can result in the
need for more stirrup reinforcement in the webs. Regional transverse bending also
exacerbates flexural cracking of the concrete cover as described in item 2 above.
- Consideration of stresses at critical locations – There are several critical stresses that
should be considered in the design of curved concrete box girder bridges. These include:
1. Axial stresses in the top and bottom slabs and the webs – These stresses result from
vertical flexure of the bridge between supports and the primary and secondary effects of
longitudinal prestressing. Regional transverse bending of the superstructure may also
occur and should be considered when determining these stresses. Because the web
lengths vary in a curved bridge, moments and flexural shears in each web may also vary.
This effect is best captured in the grillage analogy approach. To best capture it with the
spine beam approach, prestress tendons should be located at their correct transverse
positions with respect to the bridge centerline.
2. Shear stresses in the webs – These stresses result from the flexural and torsional behavior
of the superstructure. Torsion results in shear flow around the perimeter of the cross-
section that should be combined with the flexural shear. In continuous superstructures or
between the joints in pre-cast superstructures, these shear forces result in diagonal tension
stresses that can combine with the flexural tensile stresses resulting from regional
transverse bending. Stirrup design may be accomplished by combining the reinforcing
requirements for each of these actions. At the joints in pre-cast bridges, the shear is
carried by a shear friction mechanism.
Page 133
3. Transverse stresses in the cross-section – These stresses can generally be determined
using the same methods employed for a straight bridge. They govern the design of the
deck and soffit. The transverse deck and soffit reinforcing must also participate in
carrying the shear flow generated by torsion, but because concrete is often sufficient for
this purpose, this is often not a significant consideration in design.
4. Flexural and lateral shear stresses in the vicinity of prestress tendons – Complex stresses
are developed in the webs of curved concrete box girders due to the lateral forces
developed by the curvature of prestress tendons. Simplified methods for assessing these
effects have been developed and are included in the recommended LRFD specifications
and commentary.
Because design for the above forces is often optimized, it is prudent to evaluate them at several
longitudinal locations along the length of the bridge. Prestress forces and path location, web and
slab thicknesses, and the size and spacing of stirrups can be designed accordingly.
- Bearing load and bearing movement considerations – Both the spine beam and grillage
analogy methods of analysis will accurately predict elastic bearing forces if used according to the
criteria outline in the proposed AASHTO LRFD Bridge Design Specifications and Commentary
and the Analysis Guidelines included in Appendix C. Because of the curvature of the structure,
the bearing forces at any longitudinal position along the bridge will vary across the width of the
bridge.
In addition to this, both field experience and time dependent analysis show that the bearing
forces will change over time. The extent of this change is not accurately determined by currently
available time dependant software because of the treatment of torsion creep in these programs,
but software that takes into account axial creep is thought to give conservative results. In lieu of
a time dependent analysis, elastically determined abutment dead load torsions should be
increased by 20%. It is recommended that bearing force capacities be designed to accommodate
both initial and long term conditions.
Methods for addressing bearing design when bearing forces are excessive (i.e. either too high or
too low) may include, but not be limited to, one or more of the following:
1. Size individual bearings to accommodate the calculated range of bearing forces.
2. Specially design bearings so that they will not be displaced if the applied load goes into
tension or very low compression.
3. Provide ballast in the superstructure to assure the envelope of bearing forces are within an
acceptable range.
4. Reshore the structure at its bearing locations prior to setting the bearings and then release
the shoring after the bearings are set.
5. Use an outrigger diaphragm to increase the eccentricity of the individual bearings
6. Place the bearing group eccentric to the centerline of the superstructure in order to make
the individual bearing forces more equal.
7. Select bridge framing to better control bearing forces. Balancing the center and end span
lengths can mitigate bearing problems.
Considering the curvature of long bridges in a spine beam analysis can mitigate excessive design
movements at the bearings due to temperature change and possibly eliminate the need for interior
Page 134
expansion joints. Care should be taken that bearing travel is through the center of movement so
that binding of the shear keys does not occur. Prestress shortening may occur along a slightly
different orientation.
- Diaphragms – Current AASHTO LRFD provisions require that diaphragms be used for
curved box girder bridges with a radius of less than 800 feet, but also allows that they be omitted
if justified by analysis or tests. Analytical grillage analogy and finite element studies performed
as part of this project demonstrated that interior diaphragms have a minimal effect on the global
response of a curved concrete box girder bridge with a 400 ft radius and 300 ft span lengths.
Therefore, it is proposed that the requirement that interior diaphragms be included in bridges
with a radius less than or equal to 800ft. be eliminated. It is recommended that end diaphragms
still be used at all supports.
- Post tensioning sequence – Because the curvature of tendons can increase the transverse
bending of the super structure and result in tensions on the inside of the curve and compression
on the outside of the curve, it is recommended that at least one tendon on the inside of the curve
be stressed first.
With respect to varying the final distribution of prestress forces across the width of the bridge,
there does not seem to be any significant advantage in doing this. Although webs to the inside of
the curve are shorter and thus theoretically subject to less dead load and live load bending forces,
decreasing prestress forces for this web will be overcome by the transverse bending of the bridge
that will put the inside web in tension. Thus it is thought to be important to model tendons in
their correct transverse position for analysis, but a relatively even distribution of prestress forces
is desirable. It is theoretically more important for the designer to consider the incidental
distribution of prestress forces as allowed by some construction specifications.
- Skew effects – Analytical studies were performed to consider the effect of skew at the
abutments on the overall response of the bridge. It is commonly known that skew will affect the
shears in the web near the obtuse corner of a skewed abutment support. The point of the study
was to determine if bridge curvature altered the relationship between the relative response of a
skewed and non-skewed abutment. Two skew cases were studied. One case was where the skew
occurred at only one of the abutments and other case where both abutments were skewed but in
opposite directions. The second case is the likely orientation of a curved bridge that crosses over
an obstruction that is linear in orientation. In both cases it was found that the relationship
between the response of a non-skewed support and the skewed support followed the same
relationship as for a straight bridge. Thus it was concluded that existing skew correction factors
are applicable to curved concrete box girder bridges analyzed by the spine beam method.
The affect of skew on interior supports was not studied nor was the effects of different abutment
skew configurations. In all cases a grillage analogy analysis would capture any effect of skew.
This method should be used to analyze any curved concrete box girder bridges with large skew
angles at the interior supports or abutment skew configurations that vary significantly from those
studied.
- Lateral restraint issues – Horizontal curvature may result in force demands in the lateral
direction at the supports if lateral restraint is present and is modeled as rigid elements for
computer analysis. Such may be the case for supports consisting of integrally cast abutments or
piers. In these cases lateral restraint should be modeled as the stiffness of the restraining element
under consideration. In the case of bearings, however, steel or concrete shear keys usually
provide lateral restraint. These are usually provided with a small transverse gap to prevent
Page 135
binding. This gap is large enough to prevent lateral restraint and for gravity load purposes should
be modeled as a lateral force release. The key to properly considering lateral restraint is to
accurately model the actual condition and to use either spine beam or grillage analogy analysis.
- Thermal effects - Thermal movement and prestress shortening will result in movements in
different directions at the expansion joints in curved structures. This difference should be
reflected in a properly conducted spine beam or grillage analysis. Bearings should be capable of
travel through the center of movement although normal gaps provided in shear keys will allow
for slight variations in movement.
- Time dependent effects – Because of the interaction between bending moment and torsion
in curved bridges, consideration of time dependent effects is important. However, rigorous three-
dimensional analysis to determine the time dependent effects of torsion is not present in
commercially available software. This requires a reliance on the time dependant software that is
available. Fortunately, torsion creep is expected to mitigate the effect that has been observed at
the bearings, and thus this software will tend to yield conservative results for bearing force
redistribution. It can be used for design purposes until improved software is developed.
Vertical construction cambers can utilize the results from currently available time dependent
software. In fact many design engineers interviewed claim to have had good results from two-
dimensional time dependent analysis. However the bending of tall columns and twisting of the
superstructure had to be approximated using elastic three-dimensional spine beam analysis.
In the case of curved bridges, horizontal cambers may be required for segmental construction. A
curved concrete box girder bridge with relatively tall piers in California that was constructed by
the segmental cantilever method required a horizontal camber of approximately three inches at
the pier. In other words the pier had to be constructed 3” out of plumb.
- Construction methods - The impact of construction methods on the behavior of curved box
girder bridges is critical. However, the analysis methods studied are applicable to staged
construction analysis as well as cast-in-place on falsework construction. The same parameters
can be used to select the most appropriate analysis method except that time-dependant analysis
should be used. Commercially available software does not consider torsion creep but should
yield generally conservative results and is adequate for design until more sophisticated software
is developed.
- External post tensioning deviators – The use of pre-cast construction results in less weight
and quicker on site assembly and is thus increasing in popularity. Deviation blocks or saddles for
external prestress tendons in curved pre-cast concrete box girder bridges may be designed in the
same manor as for straight bridges using strut-and-tie methods or as recommended by an
experimental study at the University of Texas (Beaupre, et.al., 1988). For LRFD design of
deviators, a load factor of 1.7 should be used for the prestress deviator force and capacity
reduction (o) factors should be 0.9 for direct tension and flexure and 0.85 for shear. It is
recommended that reinforcing bar sizes in deviation saddles be limited to #5’s to insure the
proper development of this reinforcement.
It is recommended that deviation saddles in tightly curved bridges be continuous across the
bottom soffit. Another consideration for curved bridges is that straight segments of tendons
cannot rub against the interior of the webs. If necessary, the designer should include extra
deviation blocks or saddles to prevent this from happening. A deviation saddle design example,
which is reproduced from the University of Texas report, is included in Appendix B.
Page 136
- Friction loss/wobble - The current formulae for determining prestress losses due to friction
and wobble are applicable to curved bridges if the three dimensional effects of angle change and
tendon length are considered. It is necessary to explicitly consider the difference in tendon length
in individual webs and thus prestress tendons should be modeled in their actual transverse
location in a three-dimensional spine beam or grillage analogy analysis. Friction losses should be
based on a tendon curved in space when a curved bridge is being designed using two-
dimensional analysis techniques.
- Web and flange thickness limits - It is recommended that webs and flanges be designed
based on structural and constructability considerations. No minimum thickness requirements are
recommended by this study.


Page 137
7. REFERENCES:

Al-Rifaie, W. N., and Evans, H. R., 1979,. An Approximate Method for the Analysis of Box
Girder Bridges that are Curved in Plan, Proc., Int. Association of Bridges and Structural
Engineering, Int. Association for Bridge and Structural Engineering (IABSE), pp 1–21.
AASHTO, 1996, Standard Specifications for Highway Bridges, 16th Edition with Interims,
American Association of State Highway and Transportation Officials, Washington, D.C.
AASHTO, 1999, Guide Specifications for Design and Construction of Segmental Concrete
Bridges, 2
nd
Edition with Interims, American Association of State Highway and
Transportation Officials, Washington, D.C.
AASHTO, 2003, Guide Specifications for Horizontally Curved Steel Girder Highway
Bridges Bridges, American Association of State Highway and Transportation Officials,
Washington, D.C.
Abendroth, R. E., Klaiber, F. W., and Shafer, M. W., 1995, Diaphragm Effectiveness in
Prestressed-Concrete Girder Bridges, Journal of Structural Engineering, Vol. 121, No. 9, pp.
1362-1369
American Association of State Highway and Transportation Officials (AASHTO), 2003,
Guide Specifications for Horizontally Curved Highway Bridges, Washington, D.C.
American Association of State Highway and Transportation Officials (AASHTO), 2004,
AASHTO LRFD Bridge Design Specifications, 2
nd
Edition with Interims, American
Association of State Highway and Transportation Officials, Washington, D.C.
Aneja, I. K., and Roll, F., 1971, A Model Analysis of Curved Box-Beam Highway Bridges, J.
Struct. Div., 97(12), pp 2861–2878.
ASCE Committee on Construction Equipment and Techniques, 1989, Concrete Bridge
Design and Construction in the United Kingdom, Journal of Construction Engineering and
Management, Vol. 115, No. 4, pp. 618-635
ASCE Committee on Construction Equipment and Techniques, 1989, Concrete Bridge
Design and Construction in the United Kingdom, Journal of Construction Engineering and
Management, Vol. 115, No. 4, pp. 618-635
Aslam, M., and Godden, W. G. ,1973,. Model Studies of Curved Boxgirder Bridges, Rep. No.
UC/SESM 73-5, Dept. of Civil Engineering, California Univ., Berkeley, Calif.
Aslam, M., and Godden, W. G. ,1975,. Model Studies of Multicell Curved Box-Girder
Bridge, J. Eng. Mech. Div., 101(3), 207–222.
Bazant, Z. P., and El Nimeiri, M. ,1974,. Stiffness Method for Curved Box Girders at Initial
Stress, J. Struct. Div., 100(10), 2071–2090.
Beaupre, R.J., Powell, L.C., Breen, J.E., and Kreger, M.E., 1988, Deviation Saddle Behavior
and Design for Externally Post-Tensioned Bridges, Research Report 365-2. Center for
Transportation Research, University of Texas at Austin, Austin, Texas.
Page 138
Branco, F. A., and Martins, L. ,1984,. Temperature Distribution in Curved Concrete Box
Girder Bridges, Proc., Int. Conf. on Computer-Aided Analysis and Design of Concrete
Structures, Pineridge, Swansea, U.K. 1213–1223.
Bridge Design System (BDS), 1986, A Computer Program for Analysis and Design of Multi-
Cell Box Girder Bridges, ECC, 1986
Buragohain, D. N., and Agrawal, B. L. ,1973,. Analysis of Curved Box Girder Bridges, J.
Struct. Div., 99(5), 799–819.
Caltrans, 1996, Bridge Memo to Designers Manual, Memo 11-31, California Department of
Transportation, Sacramento, California
Canadian Standards Association (CSA), 1988 (Updated 2000), Design of Highway Bridges,
CAN/CSA-S6-88, Rexdale, Ont.
Cazaly, and Huggins, 1964, Design Handbook, Canadian Prestressed Concrete Institute,
Canada
Chang, S. T., and Gang, J. Z. ,1990,. Analysis of Cantilever Decks of Thin-Walled Box
Girder Bridges, J. Struct. Eng., 116(9), 2410–2418.
Chang, S. T., and Zheng, F. Z. ,1987,. Negative Shear Lag in Cantilever Box Girder with
Constant Depth, J. Struct. Eng., 113(1), 20–35.
Choudhury, D., and Scordelis, A. C., 1988, Structural Analysis and Response of Curved
Prestressed Concrete Box Girder Bridges, Transportation Research Record 1180,
Transportation Research Board, Washington, D. C., 72-86.
Chu, K. J., and Pinjarkar, S. G., 1971, Analysis of Horizontally Curved Box-Girder Bridges,
Journal of the Structural Division, ASCE, 97(10), 2481-2501.
Collins, M. P., and Mitchell, D., 1980, Shear and Torsion Design of Prestressed and Non-
Prestressed Concrete Beams, Journal of the Prestressed Concrete Institute, Vol. 25, No. 5, pp
32-100.
Computers and Structures, Inc., 1998, “SAP2000 – Integrated Finite Element Analysis and
Design of Structures,” CSI, Berkeley, California, October 1998.
Computers and Structures, Inc., 1998, SAP2000 – Integrated Finite Element Analysis and
Design of Structures, CSI, Berkeley, California, October 1998
Cordtz, K., 2004, Design of Curved Post-Tensioned Bridges for Lateral Prestress Forces,
David Evans Associates, Roseville, California
Danesi, R. F., and Edwards, A. D., 1982, Bending, Torsion, and Distorsion of Prestressed
Concrete Box Beams of Deformable Cross Section: A Comparison of Experimental and
Theoretical Results, Proceedings Institution of Civil Engineers, London, 73(Part 1), 789-810.
Danesi, R. F., and Edwards, A. D., 1983, The Behavior up-to-Failure of Prestressed
Concrete Box Beams of Deformable Cross Section Subjected to Eccentric Loads,
Proceedings Institution of Civil Engineers, London, 75(Part 2), 49-75
Evans, H. R., and Al-Rifaie, W. N. ,1975,. An Experimental and Theoretical Investigation of
the Behavior of Box Girders Curved in Plan, Proc., Inst. Civ. Eng., 59(2), 323–352.
Page 139
Fam, A. R. M., and Turkstra, C. J., 1976, Model Study of Horizontally Curved Box Girder,
Journal of the Structural Division, ASCE, Vol. 102, No. 5, pp 1097-1108.
Fu, C. C., and Tang, Y., 2001, Torsional Analysis for Prestressed Concrete Multiple Cell
Box, ASCE Journal of Engineering Mechanics, 127(1).
Fu, C. C., and Yang, D., 1996, Design of Concrete Bridges with Multiple Box Cells due to
Torsion Using Softened Truss Model, ACI Structural Journal, 93(6), 696-702.
Goodall, J. K., 1971, Torsional Stiffness of Multicellular Box Sections, Proceedings,
Conference on Development of Bridge Design and Construction, Lockwood, London.
Grant, C., 1993, Shear Flow in Multicell Sections, Proceedings, Institution of Mechanical
Engineers, London, 207(4), 247-253.
Hasebe, K., Usuki, S., and Horie, Y. ,1985,. Shear Lag Analysis and Effective Width of
Curved Girder Bridges, J. Eng. Mech., 111(1), 87–92.
Heins, C. P., Bonakdarpour, B., and Bell, L. C. ,1972,. Multi-Cell Curved Girder Model
Studies, J. Struct. Div., 98(4), 831–843.
Hsu, T. T. C., 1997, ACI Shear and Torsion Provisions for Prestressed Hollow Girders, ACI
Structural Journal, Vol. 94, No. 6, pp 787-799.
Ibrahim, A. M. M., Soin, P., Masroor, T., and Quiogue, J., 2005, Torsional Analysis and
Design of Curved Bridges with Single Columns - LFD vs. LRFD Approach, Paper presented
at the Western Bridge Engineers Conference, Portland, OR.
Lim, P. T., Kilford, J. T., and Moffatt, K. R. ,1971,. Finite Element Analysis of Curved Box
Girder Bridges, Developments in Bridge Design and Construction, U.K., 264–286.
Lopez, A., and Aparico, A.C., 1989, Nonlinear Behavior of Curved Prestressed Box-Girder
Bridges, IABSE Periodica, Zurich, 132(1), 13-28.
Maisal, B. I., and Roll, F., 1974, Methods for Analysis and Design of Concrete Box Beam
with Side Cantilevers, Cement and Concrete Association, London.
Menn, C., 1990, Curved Girder Bridges, Chapter 7.6, Prestressed Concrete Bridges, ISBN 3-
7643-2414-7, Birkhauser Verlag, Basel.
Meyer, C., 1970, Analysis and Design of Curved Box Girder Bridges, Structural Engineering
and Structural Mechanics Report No. UC SESM 70-22, University of California, Berkeley,
December 1970
Meyer, C., and Scordelis, A. C., 1971, Analysis of Curved Folded Plate Structures, Journal
of the Structural Division, Vol. 97, No. 10, pp. 2459-2480
Nakai, H., and Heins, C. P., 1977, Analysis Criteria for Curved Bridges, Jounal of the
Structural Division, ASCE, Vol. 103, No. 7, pp 1419-1427.
Okeil, Ayman M., and El-Tawil, Sherif, 2004, Warping Stresses in Curved Box Girder
Bridges: Case Studies, Journal of Bridge Engineering, Vol. 9, No. 5, ASCE, September 1,
2004.
Oleinik, J. C., and Heins, C. P., 1975, Diaphragms for Curved Box Beam Bridges, Journal
Structural Engineering, 101(10), 2161-2178.
Page 140
Ontario Ministry of Transportation and Communication (OMTC), 1983 (Updated 1998),
Ontario Highway Bridge Design Code, OHBDC, 2nd Edition, Downviview, Ontario.
Ozakea, M., and Tavsi, N., 2003, Analysis and Shape Optimization of Variable Thickness
Box Girder Bridges in Curved Platform, Electronic Journal of Structural Engineering
International, 3(2003), Queensland, Australia.
Perry, S. H., Waldron, P., and Pinkney, M.W., 1985, Design and Construction of Model
Prestressed Concrete Bifurcated Box-Girder Bridges, Proceedings Institution of Civil
Engineers, London, 2(79), 439-454.
Pinkney, M. W., Perry, S. H., and Waldron, P., 1985, Elastic Analysis and Experimental
Behavior to Collapse of 1:12 Scale Prestressed Concrete Bifurcated Bridges, Proceedings
Institution Civil Engineers, London, 2(79), 454-481.
Podolny, W. Jr., 1985, The Cause of Cracking in Post-Tensioned Concrete Box Girder
Bridge and Retrofit Procedures, PCI Journal, March-April 1985, Precast Concrete Institute,
Chicago, Illinois
Rabizadeh, R. O., and Shore, S., 1975, Dynamic Analysis of Curved Box-Girder Bridges,
Journal Structural Division, ASCE, 110(9), 1899-1912.
Rahai, A., 1996, Nonlinear analysis of Box-Girder Bridges Under Thernal Loads, Journal of
Science and Technology, London, 8(32), 206-217.
Rahal, K. N., and Collins, M. P., 1996, Simple Model for Predicting Torsional Strength of
Reinforced and Prestressed Concrete Sections, ACI Structural Journal, Vol. 93, No. 6, pp
658-666.
Rahal, K. N., and Collins, M. P., 2003, Experimental Evaluation of ACI and AASHTO-LRFD
Design Provisions for Combined Shear and Torsion, ACI Structural Journal, Vol. 100, No. 3,
pp 277-282..
Rasmussen, L. J., 1996, Plastic Behavior of Deformable Reinfoced Concrete Box Sections
Under Eccentric Load, PhD Thesis, Department of Civil Engineering, University of
Queensland, Brisbane, Australia.
Rasmussen, L. J., and Baker, G., 1998, Stress Resultant Plasticity Model for the Analysis of
Reinforced Concrete Shell and Box Structures, Proceedings Institution of Civil Engineers,
London, 128(2), 177-187.
Rasmussen, L. J., and Baker, G., 1999, Large Scale Experimental Investigation of
Deformable RC Box Sections, Journal Structural Division, ASCE, 125(3), 227-235.
Reilly, R. J., 1972, Stiffness Analysis of Grids Including Warping, Journal of the Structural
Division, Vol. 98, No. 7, pp. 1511-1523
Rogowsky, D.M. and Marti, P., 1991, Detailing for Post-Tensioning, VSL International Ltd.,
Bern, Switzerland
Schlaich, J., and Scheef, H., 1982, Concrete Box Girder Bridges, ISBN 3 85748 031 9,
Switzerland.
Scordelis, A. C., Elfgren, L. G., and Larsen, P. K., 1977, Ultimate Strength of Curved RC
Box Girder Bridge, Journal of the Structural Division, Vol. 103, No. 8, pp. 1525-1542
Page 141
Seible, F., R. Dameron, and B. Hansen, 2003, Structural Evaluation of the 405-55 HOV
Connector and the Curved Girder Cracking/Spalling Problems, StD&A, San Diego,
California
Sennah, K. M., and Kennedy, J. B., 2001, State-of-the-Art in Curved Box-Girder Bridges,
Journal of Bridge Engineering, ASCE, 6(3), 159-167.
Sennah, K. M., and Kennedy, J. B., 2002, Literature Review in Analysis of Curved Box-
Girder Bridges, Journal of Bridge Engineering, ASCE, 7(2), 134-143.
Song, S.T., Y.H. Chai, and S.E. Hida, 2001, Live Load Distribution in Multi-Cell Box-Girder
Bridges and its Comparison with the AASHTO LRFD Bridge Design Specifications, Final
Report to Caltrans for Contract Number 59A0148, July 2001
Strasky, J., 2001, Influence of Prestressing in Curved Members, Betonve Mosty, Report
TK21, Prauge, Czech Republic.
Trikha, D. N., and Edwards, A. D., 1972, Analysis of Concrete Box Girders Before and After
Cracking, Proceedings, Institution of Civil Engineers, Part 2, London, 59, 743-761.
Tung, D.H.H., 1967, Analysis of Curved Twin Box Girder Bridge, presented at the May 8-12,
1967, ASCE National Meeting of Structural Engineering held at Seattle, Washington
Turkstra, C. J., and Fam, A. R. M., 1978, Behavior Study of Curved Box Bridges, Journal of
Structural Division, ASCE, 104(3), 453-462.
Van Landuyt, D. W., 1991, The Effect of Duct Arrangement on Breakout of Internal Post-
Tensioning Tendons in Horizontally Curved Concrete Box Girder Webs, Theses, University
of Texas at Austin
Van Landuyt, D., and J.E. Breen, 1997, Tendon Breakout Failures in Bridges, Concrete
International, American Concrete Institute, Farmington Hills, MI, November 1997
Vecchio, F. J., and Collins, F. P., 1986, The Modified Compression Field Theory for
Reinforced Concrete Elements Subjected to Shear, ACI Journal, Proceedings Vol. 83, No. 2,
pp 219-231.
Zhang, L., and Huang, J. ,1989,. Three-Dimensional Analysis of Curved P. C. Box Girder
Bridges, Proc., 6th Conf. on Computing in Civil Engineering, Atlanta, 867–874.
Zokaie, T., K.D. Mish, and R.A. Imbsen, 1993, Distribution of Wheel Loads on Highway
Bridges, Phase 3, Final Report to NCHRP 12-26 (2), December 1993
Zokaie, T., T.A. Osterkamp, and R.A. Imbsen, 1991, Distribution of Wheel Loads on
Highway Bridges, Final Report to NCHRP 12-26 (1), 1991

NCHRP Project 12-71 Design Specifications and Commentary for Horizontally Curved Concrete Box-Girder Highway Bridges Final Report TABLE OF CONTENTS 1. INTRODUCTION 2. STATE-OF-PRACTICE REVIEW a. Domestic Practice b. Foreign Practice c. Field Problems 3. PUBLISHED LITERATURE REVIEW a. Codes and Design Standards b. Design Philosophy c. Response of Curved Concrete Box Girder Bridges i. Global Analysis ii. Laboratory Experiments d. Design Issues i. Bearings ii. Diaphragms iii. Flexure and Shear iv. Torsion v. Wheel Load Distribution vi. Tendon Breakout and Deviation Saddles vii. Time Dependency viii. Vehicular Impact ix. Seismic Response x. Design Optimization xi. Detailing 4. GLOBAL RESPONSE ANALYSIS STUDIES a. Objective b. Model Verification c. Parameter Study d. Special Studies 5. LOCAL RESPONSE ANALYSIS STUDIES a. Local Analysis Validation b. Local Analysis of Multi-cell Box Girders c. Local Analysis of Single-cell Box Girders d. Conclusions from Local Analysis
Page i

1 3 3 6 8 11 11 19 22 22 31 32 32 33 33 37 39 39 45 45 46 46 47 49 49 49 51 65 71 72 88 111 122

Bridge Model Scattergrams – Spine vs. Grillage E-3 b. Grillage Analogy Analysis C-4 D. Effect of Diaphragms – Ratio Plots E55 f. Three Dimensional Spine Beam Analysis C-1 b. Multi-Cell Box Girder F-63 Page ii .Comprehensive Problem (Spine and Grillage) B-5 b. Grillage E27 c. Example B-2 . PROPOSED LRFD DESIGN SPECIFICATIONS A-1 B. Effect of Hinges – Grillage Analysis E57 g. Example B-5 – Global plus Regional Combination (Poldony) B-147 f. Example B-4 – Global plus Regional Combination (Menn) B-137 e. REFERENCES APPENDICES 129 137 A. DETAILED LOCAL RESPONSE ANALYSIS RESULTS F-1 a.Deviation Saddle Design B-149 C. Single Cell Box Girder F-1 b. Effect of Diaphragms – Grillage Analysis E51 e. Bridge Model Ratio Plots – Spine vs. CONCLUSIONS 7. Effect of Skews E63 F. DESIGN EXAMPLES B-1 a.Tendon Confinement B-121 c. GLOBAL ANALYSIS GUIDELINES C-1 a. Example B-6 . STATE OF PRACTICE SUMMARY FOR UNITED STATES D-1 E.Straight to Curve (spine and grillage) E39 d.Tendon Confinement B-131 d. Example B-1 . DETAILED GLOBAL ANALYSIS RESULTS E-1 a. Bridge Model Ratio Plots . Example B-3 .6. Effect of Hinges – Ratio Plots E61 h.

because concrete can be easily molded into the required shape. bridges on curved alignments were rare. This is important. times have changed and modern highway bridges and traffic separation structures are commonly built on a horizontal curve. However. The concrete box girder. it is ideal for curved construction. A common application of curved structures is in freeway interchanges where connector ramps or “fly-overs” carry traffic from one freeway to another at relatively high speed. and improved structural forms that lend themselves to curved construction. For these reasons. because curvature induces high torsion forces. Also. geometric constraints of the urban environment. Single-cell Box Girder Multi-cell Box Girder Spread Box Beams Figure 1-1 Types of Cross Sections Page 1 . only the first two types were considered in this study. INTRODUCTION At one time. Because only a very few spread box beam bridges use curved beams. This has come about because of higher traffic volumes and speeds. is one such structural form.NCHRP 12-71 Design Specifications and Commentary for Horizontally Curved Concrete Box-Girder Highway Bridges Final Report 1. Cross sections of curved box girders may consist of single cell. particularly post-tensioned prestressed concrete that can span large distances. multi-cell or spread box beams as shown in Figure 1-1. The cross section of these structures is inherently strong in torsion. prestressed concrete box girders have become the structure type of choice in many jurisdictions.

are beyond the scope of this project. There is a need to collect and analyze this information with the goal of evaluating its merit for nationwide design rules. There has been a significant body of research and development relative to the design and analysis of curved prestressed concrete box girder bridges. That is the purpose of this project. Page 2 . Live load distribution is often addressed using the whole-width design approach. while important in and of itself.It has become common practice to analyze and design these structures as if they were straight. this design approach has been used successfully. Although most issues relative to design of curved concrete box girders have been studied to some degree. Because it is likely that the use of curved structures is going to increase. span lengths and depths. detailed results from both global and local response analysis studies are presented. etc. While much of this work has been conducted domestically. It is evident that additional physical testing of existing structures or laboratory experimentation. it is possible to do most of this through analytical studies. are often handled using specific design rules and details that have been developed over the years. it is evident that better guidelines are required for their design. Some of this information has found its way into design specifications. Local problems. and that the geometries of some of these structures will continue to push the envelope with respect to the degree of curvature. With modern computer programs and analytical models calibrated to existing physical and experimental results.. and are not necessary to accomplish this project’s intended goals. such as the lateral forces induced by curved prestressing ducts. but some recent cases of poor bridge performance have made it clear that this approach has its limitations. there is a need to fill in the gaps in our understanding. the amount of required prestressing force. Example problems are also presented that illustrate the application of these recommendations. For the most part. This report is intended to present the results of a review of the literature and the state-of-practice with respect to curved concrete box girder bridges. Final recommendations are presented in the final chapter and are implemented in the form of recommended changes to the AASHTO LRFD Bridge Design Specifications and Commentary and in analysis guidelines for these types of bridges. while much of it has not. there is also a significant body of work conducted by other countries that also needs to be considered. In addition.

telephone interviews were conducted with representatives from key state Departments of Transportation (DOT’s). Colorado. With the exception of California and Washington. Domestic Practice To obtain a better understanding of current US practice. Curved precast girder systems may also increase. Lateral displacement of columns has also been observed. Northern and eastern states. 2. more skewed supports. A few of these are discussed below. where weather conditions cause more rapid deck deterioration. Tennessee.g. 20% to 30%) have skewed abutments. spliced “U” girders with a cast-in-place deck. 6. Florida. Cast-in-place construction is most popular in the West. variable overhangs). STATE-OF-PRACTICE REVIEW a. Idaho. Many states have experienced some problems with the performance of curved box girder bridges. Texas. and more segmental construction are expected. Other states were contacted but elected not to participate. Some states have special design rules. Most curved box girders are relatively narrow continuous ramp structures. Span lengths are usually less than 160 feet but approximately 20% exceed this limit. particularly in Colorado where good success has been experienced with this type of construction 5. Page 3 . when present. The use of pre-cast box-beams in most other states is limited to straight girders.e. is provided by a curved deck slab (i. Based on our interviews current US practice can be summarized as follows: 1. A few are single span and a significant number of all structures (approx. of the column group has caused the transverse shear keys to engage. Torsion and flexural shear cracking seems to be rare and not necessarily limited to curved bridges. Many other states are tending toward segmental construction (cantilever and span-by-span using both pre-cast and castin-place concrete) in order to avoid conflict with traffic. Hawaii. but California has had some problems on skewed multicolumn bents where movement about the c. Nevada. curved. most states interviewed have a fairly small inventory of curved concrete box girder bridges (i. tend to avoid prestressed boxes due to the need to provide for future deck replacement. Washington and Wisconsin. Cracking along the tendon and tendon breakout problems are absent or minimal where sufficient space is provided between the ducts. More curved alignments. Curvature. 4.2. but not many as a percentage of the total. 3. New York. longer spans. Colorado has used pre-cast. In some cases bearing uplift at the abutments has been observed to occur over time and are thought to be due to the time dependant behavior of concrete. The trend for the future appears to be dictated by the requirements of a built-up urban environment. The states surveyed included California. Many states either already use AASHTO LRFD (2004) or are in the process of adopting it. A few bearing failures have occurred but have been avoided in states that avoid bearings altogether or use conservative bearing designs. Oregon. A small percentage of structures have skewed multi-column bents.e. <1% each of reinforced and prestressed) although many see the number increasing in the future. Unexpected vertical or horizontal displacements of the superstructure are rare.

Almost all states use AASHTO wheel load distribution. it is possible that spacing requirements could be relaxed at these locations. Most states have access to computer programs that are capable of such an analysis. It appears that duct ties and duct spacing requirements have been successful in preventing tendon breakout failures. Many states use an 800 ft radius as the trigger where designers should consider 3-D analysis. no stacks). It should be noted that the current AASHTO LRFD specifications have duct spacing requirements that are similar to Oregon’s. The number of stacked ducts is reduced to two for duct diameters exceeding 4”. it should be possible to Page 4 . These guidelines dealt with the need for special detailing in curved webs including criteria for when these details are not needed. California recently experienced a tendon breakout failure on the 405/55 interchange (Seible. 2003). Because the action of the deck and soffit slabs tend to prevent breakout failures at points of maximum tendon eccentricity in box girder structures. Several states had requirements for providing space between tendons. No state had specific guidelines for varying the prestress force in the individual webs of curved box girder bridges although at least two states recognize that stresses can vary transversely across the section and encourage designers to take the initiative to specify varying prestress force. Therefore. Texas also indicated that they have similar duct spacing requirements. These are the location of most breakout failures. excessive duct spacing requirements can present problems at midspan and over the bents in continuous concrete box girder superstructures due to the reduction in prestressing eccentricity and the corresponding increase in prestress forces that results. This is more conservative than AASHTO and most other states. tendon ties were not required in the 405/55 structure. thus mitigating the severity of stress distribution across the section and the need to vary prestress force. Prior to that failure they had published guidelines for designers related to the design of curved post tensioned bridges (Caltrans. Because the 405/55 was a long span structure there were several tendons that needed to be stacked resulting in radial forces being applied over a relatively wide area of essentially unreinforced cover concrete. Some other states indicated that they used the Caltrans tendon tie details to prevent tendon breakout failures.Most states have no special rules for when a 3 dimensional analysis such as a grillage or finite element analysis should be performed and leave it to the discretion of the designer. It should be noted that the horizontal curvature of the tendons produce additional tension on the girder toward the inside of the curve. This applies to all ducts (i. that allowed up to six tendons 4-1/2” or less in diameter to be stacked on top of one another without any space in between. This memo indicated that because of its relatively large radius. California commonly uses the whole-width design approach. Breakout failures have not occurred when these ties are present. Caltrans indicates that they currently have no special policy for prevention of tendon breakout failures except that designers are to provide tendon ties under certain circumstances. Oregon requires that no more than three 4” diameter or less ducts be stacked without a space of 1-1/2” between a subsequent stack of ducts. but Colorado reports no breakout failures due to web curvature. Duct spacing requirements do not affect tendon eccentricities where the ducts lie near the mid height of the webs.. Colorado requires a duct spacing of 44% of the duct diameter or 1-1/2” minimum.e. 1996). not specifically related to curved bridges. These are also areas where actual duct curvatures may be amplified due to the horizontal deviation of tendons to accommodate end anchorage systems. et. The problem resulted because of a separate Caltrans standard plan. It is thought that this was the primary cause of the failure. However. al.

Several states have standard details for concrete box girder bridges. Colorado expressed a need for better guidelines for combining shear and flexural stresses. Specific guidelines on how to determine the number and spacing of interior diaphragms are not provided. which are temporarily braced for torsion during the placement of the cast-in-place deck. The original design required building the superstructure on falsework set above final grade. One point of interest is the combination of global shear and regional transverse bending stresses in the webs of curved box girder bridges. and regional transverse bending stresses are first realized. lowering it to its final grade position and then casting a monolithic bent cap to connect the bridge to the column. pre-cast prestressed “U” sections set at final grade spanned over the required falsework opening. The requirements for the number and spacing of interior diaphragms vary among the states. and when the radius falls below 400 feet the maximum diaphragm spacing is 40 feet. Other states have no specific guidelines and leave it to the designer to determine how these stresses should be combined. Most states interviewed did not have specific guidelines for the design of bearings in curved box girder bridges. was developed by NRV for use by a contractor on a bridge originally designed as a cast-in-place multi-cell box. By the time falsework has been released. but the code implies that their number and spacing are to be determined by the designer and depends on the radius and the dimensions of the cross section. Diaphragms are not required in curved bridges with a radius of 800 feet or greater. The AASHTO LRFD Bridge Design Specifications state that diaphragms are required in curved concrete box girder bridges with a radius of 800 feet or less. This approach overcame falsework clearance problems but presented some challenges for the contractor. the prestress force is reduced due to relaxation and is not as critical for regional transverse bending.refine guidelines regarding duct spacing to both facilitate prestressing economies and prevent breakout failures. Caltrans. The cast-in-place pours were monolithic with the columns. A cast-in-place deck was then poured Page 5 . The current AASHTO Standard Specifications for Highway Bridges (AASHTO. For a radius between 400 and 800 feet the maximum diaphragm spacing shall not exceed 80 feet. Most of these deal with prestress duct patterns and web reinforcing. Cast-in-place bottom soffit and stem pours were made on either end on the “U” girders. although straight. Design for torsion in most states follows the AASHTO requirements. Colorado also uses pre-cast “U” girders. Colorado projects that they will use this form more often in the future. Some states expressed a preference for certain types of bearings and others try to avoid the use of bearings in curved box girder bridges. In the contractor redesign. the bridge is on falsework and experiences no flexural or torsion shear stress. but this might be because they have only designed large radius bridges. These bridges often consist of several pre-cast segments spliced together to form a span. which uses mostly cast-in-place bridges constructed on falsework. 1996) has specific requirements for the number and spacing of interior diaphragms in concrete box girder bridges and several states use these or similar requirements. It has the advantage of potentially eliminating falsework over traffic A similar “U” section. does not combine these stresses. Some of these were discussed above. The reasoning is that when the bridge is stressed. Colorado has standards for curved “U” girders with a cast in place deck that they have used with good success. At least one state said they ignored torsion design.

the contractor elected to use curved pre-cast beams as opposed to straight beams with a curved cast-in-place deck. The textbook discussed above (Menn. Many of the provisions developed in this early code have influenced the development of the current Canadian Standards Association design specifications (2000) and the updated version of the Ontario Code (1998). One each for loadings.). Girder erection required pick points to be located so that the girders would not “roll” and a temporary tie down system at the ends of the girders during the deck pour. trade journals and from personal experience. Menn. Despite these examples. Although curved box girders are not specifically addressed in these codes. Many of these are on horizontally curved alignments. The Ontario Highway Bridge Design Code (1983) was an early attempt to codify design and analysis requirements for these types of bridges. most countries and most designers base their designs on the first principles of structural analysis and design. and Concrete and Prestressed Concrete) are relatively brief documents compared to the current AASHTO LRFD. In Switzerland the structural code is split into separate booklets. etc. wood. such as the orthotropic plate analogy and the grillage analogy methods may have some applicability to curved structures. its results were representative of the bridges being built today. The two most applicable to the NCHRP 12-71 project (Loadings. Canada has been very active in developing design specifications that address the behavior of concrete box girder bridges. they are generally brief and bridge designers rely on traditional text books such as those by Menn (1990). Instead. b. does provide some general guidance on the design of horizontally curved beams and skewed bridges. As a cost savings measure. Page 6 . In Europe. Menn. as in Colorado. While the survey was not exhaustive. This approach has some advantages and. etc. Many of these bridges were built using segmental construction techniques. in Zurich and Lausanne. In general the Concrete and Prestressed Concrete booklet does not or only briefly deals with special structural configurations such as horizontal curvature. students at the two Universities. could be used on curved structures. many of the analysis techniques identified for box girders. preparing them for the professional situation in their own country. The Oyster Point Off-ramp in California also had a span consisting of curved pre-cast girders made continuous with a cast-in-place multi-cell box section. This span in question crossed over railroad tracks where falsework was not allowed. Although design codes are used.and continuity prestressing used to tie the entire structure together. Schlaich and Scheef (1980) or Strasky (2001). Foreign Practice Concrete box girder bridges are used around the world. specific course work published by their professors at their university (Leonhardt. who was a Professor for many years. A partial survey of recently constructed curved concrete box girder bridges outside the United States was conducted by reviewing material published in engineering magazines. study structural design in a practical manner. or personal experience. Strasky. steel. 1990) is very similar to what students will encounter when at the university. curved box beams appear to be relatively rare. concrete and prestressed concrete.

The AASHTO design specifications (not necessarily LRFD) are widely used by many other countries around the world.Swiss bridges on a curved alignment with a large radius are often designed without considering the curvature.. Japan. etc. Trikha. except for the bearing design. 1972) it appears the British have been quite active in researching the behavior of curved concrete box girder bridges. et. the DIN code for Germany. et. Taiwan.) and for the most part practice still follows the traditional codes of the countries involved. except that bridge bearings are designed for eccentric loads taking into account the curve of the superstructures. they rely on the engineer to apply first principals in selecting analysis techniques and design details. However. et.g. Rasmussan. 1985. Maisal.S.. et. Pickney. 1984.. design codes. engineers with experience designing bridges in France have been contacted.. but as is typical of most European practice. et.. 1998. 1982 & 1983.. 1971. Typically these structures utilize external tendons with deviator blocks. their design specifications are less prescriptive than the U. and engineers like Fritz Leonhardt have been considered pioneers for this type of construction. frequently builds curved concrete box girder bridges. but only with general instructions found in most textbooks. al.. et. al. Lim. Germany has been a leader in the design of concrete box girder bridges. Korea and Philippines) are widely used. Chile. al. 1974. Argentina.S. Brazil. al. Perry. and designers rely on their own experience as well as other published material to analyze and design these bridges.. Many design firms use methods they have developed over the years involving graphs and influence surfaces. Bridges that have alignments with slight curvature are generally designed as straight bridges without consideration for the curve.S. As of today. 1993. et. Curved beams are not covered directly although there is a section on torsion.S. al. U.. 1985. Malasia and Hong Kong) and AASHTO (Thailand. which has its own code. It is our current understanding that the French favor pre-cast segmental construction. al. In Asia the British BS5400 (India. Danesi. Page 7 . As in most European countries. 1975. al. The structural code in Brazil is quite brief and all encompassing. 1971. They have their own DIN code. Based on the published literature (Branco. and Mexico as well as other Latin American Countries use computers and similar programs as the U. Grant. et. The new “Eurocode” is intended to supersede the codes of the major European Countries. al. al. 1996. It is much more concise than the current U. BS5400 for England. Germans tend to favor cast-in-place construction. It has been developed over a number of years and is currently in use. Rahai. Evans. The British use their own code (BS5400) for bridge design.. Goodall. this code has appendices which direct the designer to special provisions by individual countries (e. et.

have been found not to be important in these failures. These stresses can either add to or subtract from the stresses developed from horizontal alignment of the bridge depending on the direction of the tendon flare. Global actions. none was placed. Page 8 . This problem was corrected at the abutments by retrofitting the bridge with prestress bar tie-downs and eliminating the bearings. First. the failure could have been prevented. Abutment bearing failure that is progressing with time was experienced on the I-5 NB to Hwy 217 NB ramp in Oregon. This essentially converted the seat abutments to end diaphragm abutments. The resulting large number of ducts required (5 per stem) for the increased prestress force were placed one on top of the other without any space in between. This particular design pushed the envelope for Caltrans design requirements to prevent a breakout failure. This theory is corroborated somewhat by the results observed in the time dependant analyses of similar structures. When the falsework was being removed. The combination of proportionately larger radial prestress forces applied to a deeper web exacerbated regional concrete stresses. and it is generally agreed that have the Caltrans lateral tie detail been used. Another recent bearing failure occurred on the bridge at Wildcat Road in Shasta County. Many such failures have occurred even in straight bridges where local curvatures of prestressing ducts occur near the prestress anchorages.000). The outside elastomeric bearing was overloaded and was destroyed and the bearing at the inside of the curve began to lift off. When these stresses were combined with the local stresses generated in the concrete cover over the stacked ducts. although theoretically a factor. Several members of the project team were involved in the investigation of the 405/55 HOV Connector OC failure.000. Over time the entire load at these bearings has shifted to only one of the bearings while the other has experienced uplift. although it is thought that currently commercially available software will tend to over predict the problem because torsion creep is not considered. This is a single span curved prestressed concrete box girder bridge that was under construction. Prestress breakout failures have been linked to a combination of the regional action of the web acting as a beam between the top and bottom deck and the local slab action of the concrete cover over the prestressing tendons. Also. Fortunately the relatively short bridge length and the fact that most of the prestress shortening had already taken place made this possible. because prestress radial forces per duct were under the limit beyond which Caltrans Memo 11-31 required tie reinforcement. the Kapiolani Interchange in 1981. concrete cracking and spalling occurred. The single cell box girder is supported on two large bearings at the south abutment. there were several other geometric characteristics of this structure that led to the failure. because the structure had fairly long spans. and the 405/55 HOV Connector OC in 2002. California.c. Repair costs for some of these structures were significant (the Kapiolani Interchange was $4. the structure depth was relatively large as were the prestressing forces in each web. These include the Las Lomas Bridge in 1978. the bearings at the abutments began to fail. Field Problems Several failures of stem concrete due to the radial forces developed by curved prestress tendons have occurred over the years. These problems are thought to be due to the time-dependent behavior of concrete. Although the curvature of this structure was less than many.

These relatively narrow multi-span ramp structures are experiencing on-going deflections and lateral movement. Significant cracking of the superstructure was also observed. The wider piers will allow bearings to be placed eccentric to the centerline of the superstructure and hopefully stabilize the situation. The web was retrofitted with external prestressing tendons that will correct the problem. A recent problem with two box girder bridges in Coahuila. continuous concrete box girder bridges supported on single bearings at each non-integral single column. It is not clear what is causing this behavior but it is fairly certain that curvature is a factor. A lightweight overlay is also being considered to completely restore the superelevation. These bridges. Figure 2-1 Uplift at Edge of Bearing Page 9 . This repair was deemed to be preferable to adding extra mild reinforcement within a web overlay. are cast-inplace. The bearings have experienced uplift due to rotation of the superstructure as shown in Figure 2-1. The bridge owner is attempting to correct the problem by increasing the size of the piers in the transverse direction as shown in Figure 2-2 and jacking and shimming the superstructure back to its original as-built position. which are relatively new. post-tensioned. Mexico that is apparently due to the curvature of the structures has developed.Stirrups in the outside web were also inadequate to resist the combined effects of torsion and flexural shear in this structure. The movement was severe enough to remove the superelevation that had been placed in the bridge at the time of construction.

Figure 2-2 Construction of Widened Column Page 10 .

Some of the documents most important to this project are discussed below. Analysis requirements are based on the classification of a bridge as “regular” or “irregular.4 “Acceptable Methods of Structural Analysis.6. Models intending to quantify torsional warping and/or transverse frame action should be fully three-dimensional. Strut-and-tie models are specified for regions near discontinuities.” except the yield line method.6 and 5.3. 3rd Edition with Interims.2. PUBLISHED LITERATURE REVIEW There is a large body of published literature related to curved concrete box girder bridges.2. AASHTO LRFD Bridge Design Specifications. AASHTO. AASHTO LRFD Bridge Design Specifications provide comprehensive specifications with commentary for the design of highway bridges. torsionally stiff single-girder superstructures for global force effects as a curved spine beam.2 “Single-Girder Torsionally Stiff Superstructures” allows for the analysis of horizontally curved.” specifies the requirements for duct material and curvature. Article 4. Article 4.1. a. Article 5.4. A number of sections are applicable to the design issues associated with curved concrete box girders.6 “Tendons with Angle Points or Curves” cross references Articles 5.8 “Shear and Torsion” specifies comprehensive design procedures for flexural shear and torsion.10.4.  American Association of State Highway and Transportation Officials (AASHTO).6. Codes and Design Standards Currently there is no national code within the U.4 “Cellular and Box Bridges” allows for the refined analysis of cellular bridges by any of the methods specified in Article 4. Page 11 . Article 4. Selected specifications articles are summarized as follows: Article 4. Article 5.9.” The classification of a curved bridge include the maximum subtended angle and whether the spans are continuous or are multiple simple-spans. specifically developed for the design of curved concrete box-girder bridges.6 “Ducts.4 for duct curvature and stress concentration considerations. unless concerns about force effects dictate otherwise.6. A summary of our review of Codes and Design Standards are summarized below. which accounts for the two dimensions seen in plan view and for the modeling of boundary conditions.C. The modified compression field theory is specified for flexural regions.3 “Multicell Concrete Box Girders” allows for the design of horizontally curved cast-in-place multicell box girders as single-spine beams with straight segments. Article 5.3.4.S. Washington.7. D.3 “Multispan Bridges” specifies the minimum requirements for the seismic analysis of multispan bridges. 2004.1. Alternative design procedures are permitted for segmental bridges. for central angles up to 34o within one span.1. respectively.

Article 5. acting in flexure. Where the factored in-plane deviation force exceeds the lateral shear resistance of the concrete cover.8. additional cover and/or confinement reinforcement shall be provided.2.3 “Effects of Curved Tendons” specifies that where tendons are placed in curved webs.4.Article 5.6 contains the flexural shear and torsion provisions specific to segmental posttensioned concrete box girder bridges.4.6 shall be used for segmental post-tensioned concrete box girder bridges in lieu of Article 5. skew.6 is a permissible or mandatory procedure for segmental bridges. torsion.3. Article 5. The commentary includes additional discussion pertinent to curved bridges.3 states that Article 5. It needs to be clarified whether Article 5.8.8.3.2. Intermediate diaphragms shall be used in curved box girders with an inside radius of less than 800-feet. tieback reinforcement is required. Where stacked ducts are used in curved girders.8.5.3 and is applicable in most cases. shall be investigated but no specific methodology or stress requirements are provided.1 “In-Plane Force Effects” defines the in-plane deviation force effects due to the change in direction of the tendon as Fu-in = Pu/R where Pu is the factored tendon force and R is the radius of curvature of the tendon.3 and 5.2.8. outlines the basic steps of designing a box section for the combined actions of flexural shear. Article 5.6 states that Article 5.8. a detailed review of the specifications was performed.4.10. and moment. Article 5. Intermediate diaphragms may be used in beams in curved systems or where necessary to provide torsional resistance. A sectional model using the modified compression field theory with a variable angle truss model is the basis of Article 5.8.10.2 “Friction” specifies the friction loss due to curvature and includes specific requirements for determination of the total three-dimensional angle change as typically found in curved girders with draped tendons.3.8.6 may be used for segmental post-tensioned concrete box girder bridges while Article 5. For curved girders.8. It includes the requirement to consider the effects of curvature.13.8.2 “Diaphragms” requires the use of diaphragms at abutments.6 that are dependent on construction method and structure type. piers. Additionally.10. Article 5. Article 14. the global flexural effects of out-of-plane forces shall be investigated.8.4. Diaphragms may be omitted where tests or structural analysis show them to be unnecessary.10.3. and includes a discussion (in italics) where further guidance is required in interpreting or applying the LRFD specification. Specific requirements for local lateral shear on the unreinforced concrete cover are given and neglect any increase in lateral shear capacity for widely spaced tendons. A conservative expression for the concrete contribution along with a 45O degree truss model is assumed.6.3 and 5.1 “General” specifies the movement requirements for joints and bearings. the moment resistance of the concrete cover. and hinge joints.3 “Curved Post-Tensioning Ducts” specifies the clear distance between curved ducts as required for tendon confinement as specified in Article 5. rotations. General Comment – There is a minor conflict between Articles 5.3 but not less than that required for straight ducts. DESIGN METHODS Two basic design methods are specified in Articles 5.9. With respect to torsion design. Article 5. The following briefly describes the design methods.8. and support restraint. Commentary Article Page 12 .

3 and.6.2.8. Step 2 – Determine the Cross Section Parameters: Acp – total area enclosed by outside perimeter of concrete cross section (Article 5.2. torsion. and moment.25f’cbvdv + Vp (Equation 5.) (Article 5.3-2) (Comment – It may be prudent to reduce Vn by the torsional shear when web crushing governs the capacity.1-5) (Comment – LRFD Article 5.8.8.) Ao – area enclosed by the shear flow path (in. Maximum torsion and concurrent actions Perform steps 3 through 7 separately for each the above cases and any additional cases that may potentially govern the design.1) ph – perimeter of the centerline of the closed transverse torsion reinforcement (in.2.) (Article 5.1 does not address the case when the thickness of the flange of a non-segmental box section is less than the effective web width.8.8. therefore.3 may be applied in lieu of the provisions of Article 5.8. “For types of construction other than segmental box girders.1 states.8.8. This is addressed in Article 12.8.8.9) Step 3 – Check the Web Width: Verify the effective web width is adequate to prevent web crushing: Vu Vn Vn 0.8.C5.) (Article 5.2.6.2.8. the provisions of Article 5.” It appears that the word may should be replaced by shall unless the intent is to permit Article 5.) (Commentary 5.6.10 of the Segmental Guide Specification (Reference 2) and LRFD Article 5.8.2) (Article 5.2.1) Acp2 / pc 2 Aobv (Equation 5.6 as an alternative design method for bridge types other than segmental.1) ds – the length of the torsional shear flow path on the exterior web (in. Consider the concurrent actions on the section.8.1) bv – effective web width (in.8.2.9) dv – effective shear depth (in. As a minimum. does not cover the steps for a segmental post-tensioned concrete box girder bridge.3 for segmental bridges. Maximum flexural shear and concurrent actions 2.8.2.2.2. Step 1 – Determine the Controlling Load Cases Determine the controlling load cases for the applicable strength limit states.1) pc – the length of the outside perimeter of concrete cross section (Article 5. consider the following two cases: 1. It is based on the provisions of Article 5. Note that provisions for considering the combined flexural shear and torsional Page 13 .3. DESIGN STEPS (GENERAL SECTIONAL MODEL) The following outlines the basic steps of designing the exterior web of a box section for the combined actions of flexural shear.

3. Collins and Mitchell.5 Vu  V p cot  Aps f po   d  v  (Equation 5.2.8.1 of the Segmental Guide Specification and LRFD Equation 5.) Increase web width if Equation 5. acting on the exterior web where the flexural shear and the torsional shear are additive to calculate the shear stress as follows: vu = (Vu – Vp) / (bvdv) (Equation 5.8. It appears that equivalent factored shear stress due to torsion in calculating x.8. Article 5.5 N u  0.8.5.8. Step 4 – Calculate Shear Stress: Check if torsion must be considered: Tu > 0.5-3 for segmental bridges.3.25Tcr. Vu. determined in Step 4.8. Reference 3.8.4. calculate the equivalent factored shear force. (Comment – This simply adds the tensions due to flexural shear and due to torsion for the exterior web where the flexural shear and torsional shear are additive and appears to be conservative. acting on the web where the flexural shear and the torsional shear are additive as follows: Vu = Vu (flexure) + Tuds / (2Ao) (Equation 5.8.) General Comment – The fifth paragraph of Commentary C5.9-1) (Comment – Note that vu includes the effects of flexural shear and torsional shear.2.8.3. There is a mention of a stress limit for the principal tension at the neutral axis of the section but the specific code section was not referenced. force should not be considered in determining the required shear capacity.125 f ' c 2 Acp pc 1 f pc 0.1-4) For Tu > 0.2. Vn. pages 399-400.4.2.6.3.1-3) Tcr  0.25Tcr (Equation 5. As the principal stresses are checked using service load.125 f c' (Equation 5. Using the equivalent factored shear force.) Step 5 – Calculate vu / f’c and x and Find  and : Calculate vu / f’c using vu calculated in Step 4. use an equivalent longitudinal Page 14 .3.shear are required by Article 12.  Mu    0.2-1.1-7) (Comment – Note that Vu is determined for a single girder and Tu is acting on the total crosssection. It appears that the intent of the equivalent factored shear force is to consider the increased shear force and resulting shear and Vc.1 regarding the equivalent factored shear force would benefit from additional clarification.8. Vu is not applicable. or the required tensile capacity specified in Article 5. Vu. Vu.8. This accounts for the increased shear stress due to torsion and will be used to determine  and  from Table 5.5 provides limits on the principal tensile stress in the webs of segmental concrete bridges at the Service III limit state and during construction.2.2-1) using the Calculate x   2  E s As  E p A ps  equivalent factored shear force.3-2 is not satisfied.

8. determined in Step 4 multiplied by the total number of webs. Step 6 – Determine Required Spacing of Stirrups: The amount of transverse reinforcement required for shear is found from: Vu  Vn (Comment – Clarify that the factored flexural shear.3-1) Av / s = Vs / (fydvcot) The amount of transverse reinforcement required for torsion is found from: Tu Tn Where: Tn = (2AoAtfycot) / s (Equation C5.2-1. (Comment – Whole-width design is not specifically addressed and would benefit from clarification in this area. therefore the maximum spacing of the stirrups. Vu. is found from: Page 15 . not the equivalent factored shear force. Using the calculated values of vu / f’c and x.3.6-2) At / s = Tn / (2Aofycot) For the exterior web of a box section. contribute to Av and At.3-3) Vs = (Avfyd vcot) / s (Equation C5.tension for combined flexural shear and torsion equal to the square root of the sum of the squares of the individually calculated tensions for flexural shear and for torsion.3.) If whole-width design is used.4. In this case.3. the shear force is conservatively taken as the equivalent factored shear force. Smax. as the additional longitudinal reinforcement for torsion is determined separately.3.8. find  and  from Table 5.4.2-1. not the equivalent factored shear force.0316 f c' bvd v (Equation 5.8.5-1.3-1) Where: Vc = 0.8. (Comment – Clarify that the flexural shear. is used for Vu.) The longitudinal reinforcement required for torsion.3. in addition to that required for flexure. is given by: Smax = Astirrups / [(Av / s)flexural shear + (Av / s)torsion] Step 7 – Check the Longitudinal Reinforcement: The required tensile capacity of the longitudinal reinforcement on the flexural tension side of the member is found from Equation 5. the combined area of both stirrup legs in the web.) This accounts for the increased longitudinal tensile force due to torsion and will be used to determine  and  from Table 5. Astirrups. as the transverse reinforcement for torsion is determined separately. forces acting on the total section are applied. is used for Vu.3.3.) Vn = Vc + Vs + Vp (Equation 5.8.8.8.

applicable to a segmental post-tensioned concrete box girder bridge.8.3 would benefit from a clarification that Al is also in addition to the required tensile capacity from Equation 5.8.5-1 when Equation 5. The area of longitudinal torsion reinforcement in the flexural compression zone should be permitted to be reduced similar to Segmental Guide Specification Article 12.3-2) Comments: Article 5. DESIGN STEPS (SEGMENTAL BOX GIRDER) The following outlines the basic steps of designing the exterior web of a box section for the combined actions of flexural shear.2.6.8. Article 5. Consider the concurrent actions on the section. The shear component of the primary effective longitudinal prestress force.3. The distribution of Al within the cross section needs to be clarified.3-2 and 5.4.8. Prestressing steel should also be permitted to satisfy Equation 5.4. Maximum flexural shear and concurrent actions 2. “Torsional Reinforcement.14.8.3.4-3 are essentially identical to the equation in Article 12.3. Vp.0.6. and during construction). In step 4. Maximum torsion and concurrent actions Perform steps 3 through 6 separately for each of the above Strength cases and any additional Strength cases that may govern the design.3.9.8. and moment.2-1 at the Service III limit state and Table 5. It is based on the provisions of Article 5.3.3.8.6.4-4 for segmental bridges. The component of inclined flexural compression or tension. “Notation.6.5-1 exceeds the longitudinal reinforcement required for flexure.6.6.9 and LRFD Equation 5. In accordance with Article 5.8. principal stresses at the neutral axis of segmental bridges shall not exceed the tensile stress limits of Table 5.6. Service III.8.2. check the principal stresses separately for each of the above Service III and construction cases and any additional Service III and construction cases that may govern the design.3. shall be added as a load effect with a load factor of 1. in variable depth members shall be considered when determining the design factored shear force.” define it as the area of longitudinal torsion reinforcement in the exterior web of the box girder.3.3. LRFD Equations 5. Step 1 – Determine the Controlling Load Cases The design for flexural shear and torsion in segmental bridges shall be performed at the strength limit state per Article 5.Al = Tn ph / (2 Ao fy) (Equation 5. torsion.” and Article 5.8. consider the following two cases for each of the limit states: 1.6 and. therefore. Step 2 – Determine the Cross Section Parameters: Page 16 .8.3-1 during construction Determine the controlling load cases for each of the three applicable limit states separately (Strength.8 of the Segmental Guide Specification. which specifies that Al shall be distributed around the perimeter of the closed stirrups.2.4 for segmental bridges.6. which appears to be incorrect. As a minimum.5.3.3-2 similar to Article 12. in the direction of the applied shear.6.8 of the Segmental Guide Specification and LRFD Commentary Article C5.3.8.8.

6.8.5-3) Vu / (bvdv) + Tu / (2Aobe) (Comment – It appears that Vu and Tu should be replaced by Vn and Tn.8.5) Step 3 – Check if Torsion Must be Considered: Check if torsion must be considered: Tu > 1 / 3Tcr (Equation 5.8.6.3-1) Tcr = 0.3.8.3) be may be taken as Acp / pc (Article 5.(in.8.) (Article 5.1.6.6.1b of the Segmental Guide Specification.8.4) bv – effective web width (in.2) (Article 5. in this equation to be consistent with Article 12.8. Page 17 .6.8.5) de – effective depth from extreme compression fiber to the centroid of the tensile force in the tensile reinforcement (in.3) Step 4 – Check the Web Width: Verify the effective web width is adequate to prevent web crushing: Vu Vn Vn 0.) be shall be adjusted to account for the presence of ducts as specified in Article 5.8.8.4) dv – effective shear depth (in.) (Article 5.8.6. Consider the compressive stress due to vertical tendons in the webs.6.6.6.379f’c0.0 (Equation 5.5.3) Acp – area enclosed by outside perimeter of concrete cross section (in.8.) (Article 5. Check the allowable principal tensile stress for Service Limit State III and during construction in accordance with Article 5.3-3) fpc = Unfactored compressive stress in concrete after prestress losses have occurred either at the centroid of the cross-section resisting transient loads or at the junction of the web and flange when the centroid lies in the flange.3-2) f pc 0.2) (Article 5.8. respectively.6.5-2 or Equation 5.6.6.) (Article 5.5bvdv (Equation 5.8.8.5-3 is not satisfied.3) ph – perimeter of the centerline of the closed transverse torsion reinforcement (in.3) pc – the outside perimeter of concrete cross section (in.) Increase web width if either Equation 5.Ao – area enclosed by the shear flow path (in.6.6.6. (Article 5.0632K K = 1 f c' 2Aobe (Equation 5.5-2) 0.8.3) be – effective width of shear flow path. but not exceeding the minimum thickness of the webs or flanges comprising the closed box section.6.0632 f c' 2.8.) (Article 5.6. (Article 5. Increase the web width or the vertical prestressing force in the web if the allowable principal stresses are exceeded.5 (Equation 5.474f’c0.6.8.8.

5-5) Av / s = Vs / (fydv) The amount of transverse reinforcement required for torsion is found from: Tu Tn Where: Tn = 2AoAvfy / s (Equation C5.4 contains a conflicting statement that Al shall be distributed around the perimeter of the closed stirrups in accordance with Article 5.8.Step 5 – Determine Required Spacing of Stirrups: The amount of transverse reinforcement required for shear is found from: Vu Vn Vn = Vc + Vs (Equation 5. therefore the maximum spacing of the stirrups.2.8.8.0632K f c' bvdv (Equation 5. shall satisfy: Al = Tuph / (2Aofy) (Equation 5.8.3.8.4) Page 18 .8.8. which appears to be incorrect.6. contribute to the transverse hoop reinforcement for flexural shear and torsion.8 of the Segmental Guide Specification which specifies that Al shall be distributed around the perimeter of the closed stirrups.6. the design yield strength for flexural shear and torsion design shall be taken in accordance with Article 5.8.5-1) Where: Vc = 0.6.8. the beneficial effect of longitudinal prestressing is taken into account by considering it equivalent to an area of reinforcing steel with a yield force equal to the prestressing force.3. “Torsional Reinforcement.6.6.6.6 which appears to be correct. Smax.4-3 is essentially identical to the equation in Article 12. LRFD Equation 5. Article 5.8.6. Article 5.4-3) Comment: The distribution of Al within the cross section needs to be clarified.8. Astirrups.” and Article 5. In determining the required amount of longitudinal reinforcement.6.4-2) Av / s = Tn / (2Aofy) For the exterior web of a box section. is given by: Smax = Astirrups / [(Av / s)flexural shear + (Av / s)torsion] When vertical tendons are provided in the web. (Commentary Article C5.8.5-4) Vs = Avfyd v / s (Equation 5. the combined area of both stirrup legs in the web. “Notation. Step 6 – Check the Longitudinal Reinforcement: The minimum additional longitudinal reinforcement required for torsion (in addition to that required for other concurrent actions).6.4.6.” define it as the area of longitudinal torsion reinforcement in the exterior web of the box girder.

American Association of State Highway and Transportation Officials. Some of the details of this specification were discussed above. It also includes a construction specification and design examples for both an “I” girder and box girder bridge. the material is presented in a manner that stresses fundamental principles.Subject to the minimum reinforcement requirements of Article 5. Washington. It is based on recent research in the United States and abroad. deflections.4-4)  AASHTO. While much of the content is based on Swiss standards and practices. Prestressed Concrete Bridges.8. consulting engineers. and curved bridges Page 19 . skewed. C. 1999. The content is based on the author’s direct experience gained from the design and construction of bridges in Switzerland. This is a recently published AASHTO Guide Specification for curved steel bridges including box girders.8.. Guide Specifications for Design and Construction of Segmental Concrete Bridges. and constructability.6. This publication. splices and connections. 2nd Edition with Interims. A few of these publications are discussed below. This specification discusses design philosophy and limit states and includes provisions for loads. 2003. This second edition of the guide specifications for segmental concrete bridges was prepared for use in conjunction with the Standard Specifications for Highway Bridges (non-LRFD).6. 1990. which are significant departures from previous design and construction provisions. Birkhauser Verlag. D. Washington. design of flanges.6. It has been suggested that the design specification contained therein be used as a model for the NCHRP 12-71 project. Design Philosophy A number of books and papers have been written about the design of concrete box girder bridges. and subsequent interim revisions to these specifications. academicians. structural analysis. ISBN 3-7643-2414-7. webs. shear connectors. and suppliers. This book covers straight.  Menn.9defy) (Equation 5. Guide Specifications for Horizontally Curved Steel Girder Highway Bridges. the area of additional torsion reinforcement in the flexural compression zone may be reduced by an amount equal to: Mu / (0. American Association of State Highway and Transportation Officials.  AASHTO.C. Basel This book provides engineers with a comprehensive overview of the fundamental principles governing the design and construction of concrete bridges. embodies several concepts. The committee included representatives of state departments of transportation. b. Many of these discuss the effect of horizontal curvature on the behavior of these bridges. contractors. which was developed by a broad-based committee organized by the American Segmental Bridge Institute.C. bearings. D. the Federal Highway Administration.

“Detailing. Article 5. Article 7.” discusses the function.6. a condition that occurs in horizontally curved post-tensioned box girder bridges. Article 4.3.3. but greatly simplifies the computational effort. which can increase flexural stresses.” provides guidance on developing analytical models that can be applied to curved box girder bridges. Expansion bearings must be able to accommodate both temperature and prestress shortening displacements which will be in different directions. must be considered in determining the required prestressing force. and Prestressing Concept and Tendon Layout.” discusses the qualitative difference in superstructure displacements due to temperature and shrinkage vs. Transformation of Torque into Torsional Sectional Forces.consisting of both open sections and closed box sections. longitudinal prestressing in horizontally curved bridges.3(b) “Influence of Girder Curvature. techniques for developing grillage models of multiple cell box girders are presented. and design of internal diaphragms. necessity. The method does not satisfy compatibility equations exactly. Diaphragms are recommended at abutments and piers. Page 20 . “Curved Girder Bridges. An example of the regional transverse bending moment in the web of a horizontally curved member is presented. Shear-regional transverse bending interaction diagrams are presented. Article 5.6.4. An example is given.1.4 “Structural Models for Bridge Superstructures. The book presents a simplified method for analyzing curved bridges iteratively.” presents a rational method for the design of webs subject to combined shear and regional transverse bending. Simple vector diagrams are presented to illustrate how torsional section forces are developed by a variation in the direction of longitudinal bending moments due to the curvature of the superstructure and resisted by shear flow around the perimeter of the box section. The use of intermediate diaphragms is usually not necessary in straight and lightly curved box girder bridges.4 “Diaphragms. The article includes subsections on Conceptual Design.” is devoted entirely to horizontally curved bridges. Analysis. The book addresses a number of issues that are applicable to horizontally curved box girder bridges. Prestressing.2(b) “Web Design for Shear and Transverse Bending. Prestressing can produce longitudinal and transverse bending and shear forces as well as torque in curved box girder bridges. The discussion on the conceptual design of curved box girders points out the role of torsion in design and how at ultimate loads torsion and bending moment can be redistributed. For example. The effect of torsion on bearing forces may require the bearings to be placed away from the webs. Prestressing can also be used to enhance torsional and transverse bending resistance although this is often avoided for economic reasons. The method is based on Swiss practice and neglects the concrete contribution to the shear capacity. Torsion. Article 6. The deviation force per unit length is determined as the tendon force divided by the radius of curvature of the tendon. Article 5.” discusses the deviation forces generated by curved post-tensioning tendons.1.

(6) cross-bracing requirements. Switzerland This publication is the outcome of a comprehensive survey of concrete box girder bridges. The views of the supervising engineer. This paper shows the different approaches on a number of factors. The “Design” section covers several aspects of curved bridges. Subjects discussed in this review include (1) different box-girder bridge configurations. and Kennedy.. contractor. However these steps have also brought new problems in their wake for the engineer. (12) ultimate load carrying capacity. M. Vol. The paper is divided into three parts as follows: 1. (4) load distribution. 115.  Schlaich. (13) buckling of individual components forming the box cross section. and “Dimensioning and Detailing. Concrete Bridge Design and Construction in the United Kingdom. (8) thermal effects. The “Structural Analysis” section discusses the mutual influence of the longitudinal bending moments and torsional moments for horizontally curved bridges. and supervisor. 6(3). (3) dynamic response. 3. Alternative substructure configurations are discussed for curved bridges. ISBN 3 85748 031 9. (9) vibration characteristics. and curved bridges. 2001. and increased mechanization and better planning in the construction of these bridges have brought this about. 159-167 The objective of this paper is to provide highlights of the most important references related to the development of current guide specifications for the design of straight and curved box-girder bridges.. A simple table of equations Page 21 . skew. and (15) curvature limitations provided by the codes for treating a curved bridge as a straight one. (2) construction issues. J.. As such it provided an excellent bibliography from which to identify other papers that were reviewed in detail. 1989. State-of-the-Art in Curved Box-Girder Bridges. 618-635 The design and construction of concrete bridges in the United Kingdom has changed rapidly during recent decades. 2. The design of bridges in classes for span and type with reference to the pertinent factor for that design. The publication is divided into three main parts. Zurich. Journal of Bridge Engineering. “Structural Analysis”. 4. The contractor’s approach to construction that illustrates the need for flexibility in the construction method in order to meet contract deadlines. K. (2) load distribution. Sennah. H. B. ASCE. (10) impact factors.  ASCE Committee on Construction Equipment and Techniques. (3) deck design.” A comprehensive reference list is included. (14) fatigue. (11) seismic response. “Design”. and his means of achieving a balance between the designer’s intentions and the contractor’s proposals. pp. and (4) ultimate load response of box-girder bridges. J. (7) end diaphragms. Concrete Box Girder Bridges. 1982. Journal of Construction Engineering and Management. Better analytical methods. and Scheef.. The literature survey presented herein encompasses (1) the construction phase. (5) deflection and camber. The publication addresses straight. No. International Association for Bridge and Structural Engineering. Recommendations are given for when the longitudinal bending moments can be determined as for a straight bridge and then combined with the torsional effects without considering the coupling of the two effects upon each other.

Association of Bridges and Structural Engineering. Association for Bridge and Structural Engineering (IABSE).  Al-Rifaie. Int. is adapted for curved box girders and is useful for preliminary design of these structures. R. The “Dimensioning and Detailing” section covers several aspects of curved bridges. These refinements are also being considered for curved box-girders. This approach results in substantial computational savings over the finite element method and is well suited to preliminary design studies. W. Several different span-to-radius ratios and loading conditions were considered. c. The method as developed does not accurately account for shear lag in the deck. Nodes on the frame are assumed to be fixed against translation but free to rotate. The influence of horizontal curvature on the movements at bearings is also discussed. A number of analytical techniques have been studied. good correlation was found for critical stresses although some discrepancy was found for non-critical stresses. The plates are only capable of resisting in-plane forces. The method was checked against finite element results for model bridges representing both concrete and steel box girders. which was originally developed for straight box girders. N. 1979. single cell box girders. Page 22 . In this procedure. The following paragraphs discuss some of the papers and reports that were reviewed. In general. The final step is to apply a “sway correction” procedure that will make the displacements of the nodes in each of the transverse nodal sections compatible with the deflections of the nodes along the edges of the longitudinal plates. This method. An Approximate Method for the Analysis of Box Girder Bridges that are Curved in Plan. Response of Curved Concrete Box Girder Bridges i. Proc. The method developed is applicable to simple span.. but many others are suitable for production design practice. An approximate method for analyzing curved box girder bridges using the nodal section method is described. Each frame is analyzed and the reactions at the nodes determined. torsion. Many of these are relatively complex. Our current feeling is that a properly applied grillage analogy method provides good results and may be most suitable for analyzing bridges with significant curvature.based on classical curved beam theory is presented for several different loading conditions of a curved single span bridge with fixed supports. H. Int. Studies have extended this method to multi-cell box girders and have developed methods to account for shear lag in straight box girders.. and regional transverse bending is addressed including a rational method of designing the web reinforcement for combined shear and regional transverse bending.. and Evans. Global Analysis Most published research seems to be directed toward the global response of box girder bridges. the transverse nodal section is idealized as a plane frame. pp 1–21. The reactions are then applied to longitudinal plates that represent the components of the box girder. Dimensioning and reinforcement of the web for flexural shear.

Numerical examples of deformation of horizontally curved bridge girders. Stiffness Method for Curved Box Girders at Initial Stress. The two types of circumferential strip elements are obtained by using a modified finite difference discretization in the transverse direction. 799-819.. in general.. 1973. A.. and Agrawal. L. M. No.. For nonlinear material Page 23 . D. Washington. pp. It is based on thin-wall beam elements that include the modes of longitudinal warping and of transverse distortion of the cross section. validate the formulation and indicate good agreement with solutions by other methods. The method is based on harmonic analysis in the circumferential direction. The cross section of the element is a rectangular single-cell box with side cantilevers. E. and the tool (software) to implement this method is not readily available. 1974. and an interior mode is used. Results of two examples obtained by the method are compared with available solutions. Transportation Research Record 1180. pp. is considered. The applicability of this paper to NCHRP 12-71 is limited because it is only applicable to simply supported bridges. B.  Buragohain D. and of lateral buckling of box arches. The use of minimum energy principles yields two types of element matrices that are assembled to form the overall stiffness matrix of the structure following stiffness matrix procedures.. C. 1988. as well as straight girders. consisting of posttensioned bonded tendons in the longitudinal direction. Vol. Prestressing. arbitrary cross sections. This method is most applicable to steel box girders and is of little use to our project. 72-86. and it is found that the well-known spurious shear stiffness in very slender beams is eliminated by virtue of the fact that the interpolation polynomials for transverse displacements and for longitudinal displacements (due to rotations and warping) are linear and quadratic. N. including transverse distortion and longitudinal warping of the cross section. be segments of conical frustra. Analysis of Curved Box Girder Bridges. are considered at each of the three element nodes. Structural Analysis and Response of Curved Prestressed Concrete Box Girder Bridges. which is numerical in general. and Nimeiri. Deformations due to shear forces and transverse bi-moment are included. Vol. constant curvature and pinned supports at both ends. Journal of the Structural Division. Z. No.. The total potential energy of the structure is discretized into energy due to extension and bending and energy due to shear and twisting. The element is treated as a mapped image of one parent unit element and the stiffness matrix is integrated in three dimensions. C. A curved nonprismatic thin-walled box beam element is used to model the bridges. 2071-2090 A sophisticated numerical method of analysis of the global behavior of long curved or straight single-cell girders with or without initial stress is presented. 100. and Scordelis. 5. Journal of the Structural Division. The method described is applicable to orthotropic material properties. Eight displacement degrees of freedom. A discrete strip energy method is presented for the analysis of curved box girder bridges of arbitrary cross section and various forms of curved folded plate structures simply supported at the two ends and composed of elements that may.. A numerical finite element analysis method for linear-elastic analysis and nonlinear material analysis of curved prestressed concrete box girder bridges is demonstrated through two examples. 99. Transportation Research Board.  Choudhury. D. but could be carried out explicitly in special cases. Bazant. respectively. 10.

October 1998. and prestressing steel are modeled. 1971. The authors claim that this analysis is much more accurate than other methods of analysis. but the results of the analysis can serve as a comparison case for measuring the accuracy of other methods. live load... 1998. “SAP2000 – Integrated Finite Element Analysis and Design of Structures. In the second example. and Pinjarkar. Journal of the Structural Division. Stiffness coefficients of sector plates are presented herein whereas stiffness coefficients of shell elements are based on Hoff’s solution of Donnell’s equations. H. No. The shear and the transverse flexural responses of the box beam cross section are modeled using trilinear constitutive relationships based on cracking.  Bridge Design System (BDS). California. The program provides the user with options to create. SAP2000 features integrated modules for design of both steel and reinforced concrete structures. and prestressing load cases are analyzed. analyze and design structural models. This reference constitutes the concrete structure portion of the SAP-2000 Manual.” CSI. Inc. reinforcing steel. This modeling technique is significant for the NCHRP 12-71 project as it is commonly used in practice and its limits of applicability need to be investigated. Although a comparison is made between the results of a curved twin box girder bridge obtained by the proposed method and another approximate analytical method (Tung.  Computers and Structures. modify. The FEM analysis tool itself is not readily available and thus is of limited use. pp. ECC. 1986. particularly those with respect to the effect of radius of curvature. Berkeley. K. 1967). with emphasis on design code check analysis. Results of a sample bridge analysis are shown with stresses and deflections reported for a simply supported multi-cell concrete bridge. if more detail on the presented example can be obtained. Dead load. Although the finite element formulation might be detailed and comprehensive and conducive to studying the ultimate behavior of concrete box girder bridges. 10. A Computer Program for Analysis and Design of MultiCell Box Girder Bridges. overload behavior and ultimate strength of a three-span curved prestressed concrete box girder bridge under increasing vehicular load are investigated. Bridges are modeled as a plane frames ignoring all horizontal curve effects. 2481-2501 A finite element method for the analysis of simply supported curved girder bridges with horizontal sector plates and vertical cylindrical shell elements is outlined. Some interesting results. were obtained. no comparisons with other (simpler) analysis methods are given.analysis. 1986 The described software program is the most commonly used software for design of multi-cell box girder bridges. Vol. The first example demonstrates the versatility of the numerical method in determining the linear-elastic distribution of forces in a three-span prestressed box girder bridge of curved plan geometry and variable cross section. yielding. Analysis of Horizontally Curved Box Girder Bridges. The program is structured to support a variety of design Page 24 . 97.  Chu. its applicability to the planned global elastic analysis studies is limited due to its complex nature. The different response characteristics of the bridge induced by different transverse locations of the overload vehicle are presented. and ultimate stages. S. the uniaxial stress-strain curves of concrete. G.

C. column and beam design procedures. Other than this. very useful for global analysis. 2001. the presented paper. in its current form. of course. Using the Softened Truss Model. Reinforcing steel and prestress strand are stressed uniaxially according to an assumed multi-linear stress strain relationship. these design code checks are of limited usefulness for the specialized needs of curved concrete box girder bridge “local” or sectional (“regional”) analysis. Torsional Analysis for Prestressed Concrete Multiple Cell Box. Given that the design code check features of the program focus on frame analysis. Page 25 . This paper describes an analytical study of the ultimate strength of horizontally curved reinforced and prestressed concrete box girder bridges. this document is useful as a summary (and side-by-side comparison) of various design codes for concrete columns and beams. and other special consideration required by the code. Y. The program currently supports a number of foreign and domestic design codes. The analysis was performed using materially nonlinear plane stress finite elements (i. the authors present the formulation for calculation of torsional effects in a multi-cell reinforced and prestressed concrete box girder bridge.e. C. Each of six subsequent chapters gives a detailed description of a specific code of practice as interpreted by and implemented in SAP2000. pp 45-51. 1989. Classical matrix analysis techniques are used to perform the analysis. A. 132(1). it is of limited direct utility to the NCHRP 12-71 project. 13-28. Section warping is not considered.  Lopez. panels) that exhibit membrane action. 1986).C. But the program is. however. and Aparico. A 5-span bridge was selected to study the difference between linear and nonlinear response. is too complex to be used in practical design situations or for parametric studies. IABSE Periodica. and vary through the thickness.. and Tang. This paper asserts that because concrete box girder sections are not made of thin webs and flanges. Based on these studies the following conclusions were drawn. A.codes for the automated design and check of concrete frame members. Live loads were located at various transverse positions and the behavior observed as the intensity of these loads was increased.. ASCE Journal of Engineering Mechanics. Each chapter describes the design loading combination. 127(1). This chapter describes the common terminology of concrete design as implemented in SAP2000. Aside from the obvious usage as a SAP user reference. causing the effective stiffness of the member to be less than that observed at low values of load (torque). This research may be of some value to NCHRP 12-71 if the methodology can be simplified and used as the basis of simplified methods for calculating torsional effects. However.. Zurich. It should also be noted that there is no coverage of design or analysis of prestressing in this document.. Nonlinear Behavior of Curved Prestressed Box-Girder Bridges.  Fu. Panel behavior is based on the evolutive truss analogy with peak stress reduction (Vecchio and Collins. The formulation is coded in a computer program and the results from an example problem are presented. the stress distribution in these components are not constant. The material is assumed to have a variable modulus of elasticity that is strain dependent. Chapter II outlines various aspects of the concrete design procedures of the SAP2000 program.

The following criteria are proposed for design of curved prestressed box girder bridges. refined curved beam theories. 5. 3 and 4 cell concrete box-girder bridges. and the finite strip method analysis of curved folded plates. The girder on the inside of the curve is stiffer than the girder on the outside of the curve and will attract more load. The CURSTR program was used to study wheel load distribution in 1. warping is generally small and ordinary curved beam theory can be used successfully. The first program. plate and grillage analysis methods. University of California. Analysis and Design of Curved Box Girder Bridges. Transverse prestressing significantly affects post-cracking response. The structural response is highly nonlinear at ultimate loads. 3. 3. finite element analysis. UC SESM 70-22. Because concrete box sections have relatively thick walls. Elastic models cannot accurately (and will often non-conservatively) predict the ultimate limit state. uses the finite element method.. CURSTR. These include straight beam approximation. the following observations were made: 1. The report outlines the methods developed over the years for analyzing curved bridges. Page 26 . Several parameter studies were conducted with different curvatures. 2. Berkeley. C. cracked flexural and torsional section properties should be used to determine demands and plastic sectional analysis should be used to determine capacities. Refined curved beam theories are required to analyze thinwalled box sections that can experience warping of the cross-section in the transverse direction. 1. Structural Engineering and Structural Mechanics Report No.  Meyer. The solution methodology requires that loadings be applied in the form of Fourier series. The type of failure depends on the loading case considered. When using linear analysis to determine the factor of safety against failure. Both programs yield essentially the same results. The form and degree of internal force redistribution at ultimate loads depends on the loading case considered. Ultimate internal force response can be evaluated accurately using plastic sectional analysis. December 1970 The history of curved bridges and the highway geometric requirements of these structures are discussed. FINPLA2. 1970. uses the finite strip method of curved folded plates. span lengths. The response of the bridge under service loads can be accurately predicted using elastic models. curved beam theory. With respect to single cell boxes. Internal forces are redistributed due to progressive cracking and structural coupling between bending and torsion. 4. 6.1. Two methods of analysis are developed in the form of computer programs. deck widths. depth to span ratios and loading configurations. The second program. 2. 2.

the overall approach may be utilized or modified to be applicable to concrete bridges and NCHRP 12-71. H. and the induced stresses and deformation. or single cell box). For two-cell boxes: 1. Cell width accelerates the curvature effects. The influence of depth to span ratio is also similar. 3.2. No. spread box. C. Design criteria for curved bridges have been formulated using these equations along with parametric studies. The range for the parameter . The moments in the middle girder and the girder on the inside of the curve increase with curvature. However. (i. 3. Load distribution is generally worse in continuous bridges. approximate methods are justified and even preferred in most cases. 7.e. The moment in the girder on the outside of the curve decreases with curvature up to a certain level and then starts to decrease.  Nakai. The influence of span length on load distribution is similar to straight girders. 6. This behavior is independent of span.. 4. A girder moment distribution factor is developed:   1 B L  2. 5. The equations consider the type of supporting element. and deflection of curved and straight bridges. The girders may be assigned moments proportional to their moments of inertia. pp 1419-1427. P.1  600  R  Bridges with curvatures radii large than 1000 ft. bending and torsional stiffness and central angles. It appears that these equations are specific to steel girders and warping torsion is a part of this methodology. Analysis Criteria for Curved Bridges. which “relates the cross-sectional geometry Page 27 . especially for flowcharting the decision path for analysis. cell width. The girder on the outside of the curve has a larger statical moment because of its longer span. stress. The response of three and four cell box-girders exhibit similar characteristics to one and two cell boxes. 4. Journal of the Structural Division.. 2. With respect to negative moments over a “fixed” support: 1. For design. The paper provides equations for moment. 2. The combination of items 1 and 3 result in nearly equal moments in the two girders. Load distribution improves with an increase in curvature. The influence of span on load distribution is small. and depth to span ratio. open girder. The paper reports on a series stiffness equations and limiting angle equations developed for determination of need for analysis of a bridge as a curved structure. and Heins. 103. ASCE. Vol. 1977. may be considered straight for analysis purposes.

The curvilinear nature of box girder bridges along with their complex deformation patterns and stress fields have led designers to adopt approximate and conservative methods for their analyses and design. K. multiple-spine. and monobox girders. The expressions are functions of the girder types. This paper presents highlights of references pertaining to straight and curved box girder bridges in the form of single-cell.”  Sennah. The Canadian Highway Bridge Design Code (CHBDC. Folded Plate Method 4. B. webs. 2000) recommends limiting this method to straight bridges with multi-spine cross-sections. The evaluation of gives the following criteria: “when is less than 0. Grillage-Analogy Method 3. Data for multicell girders are not available from this paper. which provide information to the designer in determining the need for a curved girder analysis. Recent literature on straight and curved box girder bridges has dealt with analytical formulations to better understand the behavior of these complex structural systems..” The relationship between and the central angle were also found for the three bridge types studied. and multicell cross sections. In the grillage analogy method. Finite-Element Method The orthotropic plate method lumps the stiffness of the deck. Conclusions made in the paper are as follows: A series of stiffness parameter equations and limiting angle equations have been presented. soffit and diaphragms into an equivalent orthotropic plate. bending and torsional stiffness. Literature Review in Analysis of Curved BoxGirder Bridges. 2002. and (2) experimental studies on the elastic response of box girder bridges.4 evaluation of stresses due to pure torsion may be omitted. 7(2). The bounds for the torsional stiffness parameter are derived and are dependent on “the central angle. The elastic analysis techniques that are discussed include: 1. the multi-cellular structure is idealized as a grillage of beams. M. which “reflects the bending and torsional stiffness of the girders. is determined for multiple I. ASCE. Few authors have undertaken experimental studies to investigate the accuracy of existing methods. Orthotropic Plate Theory Method 2. J. and EI. and Kennedy. the torsional rigidity of the crosssection. twin box.” which is the total horizontal angle the girder passes through between supports. and central angles. The CHBDC does not recommend that this method be used for sections with less than three cells or box beams. Finite-Strip Method 5. Journal of Bridge Engineering. When is greater or equal to 10 evaluation of stresses due to warping may be omitted. The deflection ratio is primarily dependent on . This method requires special attention to the modeling of shear lag and the Page 28 .and the spacing between the outside girders.. 134-143. Parameter studies indicated that acceptable results are given for bridge with three or more spines. The literature survey presented herein deals with: (1) elastic analysis.

the finite element method has become the method of choice for complex structural problems. particularly those obtained using the finite element method of analysis. number of diaphragms. In conclusion. J. The folded plate method uses plates to represent the deck. and Fam. Implications for preliminary design are presented for both concrete and composite concrete/steel sections Page 29 . length. The plates are connected along their longitudinal edges and loads are applied as harmonic load functions. Even this adaptation does not account for all warping and bending stresses. webs and soffit of box girders. Vlasov first developed an adaptation of Saint Venant theory to thin-walled sections. Many researchers have applied this technique to the analysis of curved box girder bridges. A. Many computer programs have been developed specifically for box girder bridges..  Turkstra. Behavior Study of Curved Box Bridges. including dynamics. 453-462 A numerical analysis of a number of single-cell curved box-girder sections with variable curvature. The finite-strip method has been widely researched. though more difficult to apply. 3. creep. With the advent of powerful personal computers and computer programs. The effects of these parameters on longitudinal stresses are considered. shrinkage. distortion and bending deformations of the individual wall elements of the box. etc. Several researchers have attempted to overcome this difficulty by developing special elements or using special substructuring techniques. Its drawback is that it is limited to simply supported bridges with line supports and thus not applicable as a general use analysis tool for production design. 104. A problem that occurs is that a large number of flat plate elements are required to properly model the curved elements of a curved bridge. temperature. No. this method yields results that compare well with finite element techniques. accounts for all relevant behavior in curved box girder bridges and yields the most reliable analysis results. When modeling is properly done. C. these experiments have shown good agreement with analytical results. pp. Curved box girder structures cannot be accurately analyzed using the classical curved beam theory developed by Saint-Venant because it does not account for warping. The method is time consuming and only applicable to restrictive support conditions. It is essentially a special case of the finiteelement method but requires considerably less computational effort because a limited number of finite strips connected along their length are used. Journal of the Structural Division. and loading was performed. but most of these are not commercially available. based on selected numerical results. The versatility of this method has allowed researchers to investigate several aspects of bridge behavior.. the finite-element method. Considerable research effort has been expended over the years to develop computational techniques to over come shortcomings in the present theory.torsional stiffness of closed cells. M. Diaphragms are not modeled. R. Several laboratory experiments involving model box girder bridges have been conducted over the years. In general. web spacing. 1978. Vol.

The bridge used in the example above was noncomposite so that warping would be significant.. 97. September 1.  Okeil.. Stiffness Analysis of Grids Including Warping. found that in all cases the effect of warping stress was insignificant. Bimoment and warping torsion are obtained for grillage structures. These studies. 1511-1523 Two methods of including warping effects in the stiffness method of analysis are presented. and Scordelis. No. they are even less vulnerable to warping. Vol. Vol. (Since concrete box girders will have thicker webs and soffits. and El-Tawil. This is not a serious practical problem. When I w is small relative to Ix (approximately pL > 5 for each element) computational errors grow. ASCE. 2459-2480 A finite strip method of analysis is presented which can be used to analyze curved folded plate structures simply supported at the two ends and composed of elements that may in general be segments of conical frustra. A description of a general computer program developed for the analysis and the results of several examples are also given. 1972. In the limiting case where Iw = 0. Ayman M. Analysis of Curved Folded Plate Structures. as good solutions can be obtained using an ordinary grillage analysis neglecting warping for structures where warping stiffness is small and p is large. No. These analytical studies were conducted using the computer program ABAQUS and a special 7-degree of freedom beam-column element that can account for warping. Reilly. The 1997 AASHTO curved girder specifications places limitations on the span to radius ratio for designing the bridges as straight. 1971. Sherif.. J. The direct stiffness method is used to assemble the structure stiffness matrix and to determine displacements and element stresses. 7. 9. Journal of the Structural Division. with the loadings expressed by Fourier series. 98. No. These ratios were found to be conservative by a factor of 2 or more when it comes to the need to consider warping. 10. pp. because the stiffness matrix tends to become singular as the elements on the main diagonal associated with warping approach zero. pp. which were designed to investigate warping related stresses in these bridges. Results of computer programs based upon these methods are shown to agree closely with published solutions for straight beams. and it is likely that the effects of warping can be ignored in almost all of these bridges. Composite bridges seem to fall near the borderline where warping can be neglected. R.. 2004. and on a finite element stiffness analysis in the transverse direction. Journal of Bridge Engineering. Journal of the Structural Division.) Page 30 . 2004. Vol. Warping Stresses in Curved Box Girder Bridges: Case Studies. a curved beam and a curved highway bridge. the warping effects disappear and leave only the familiar GIx/L.  Meyer. A. The method is based on a harmonic analysis in the circumferential direction. This paper discusses a number of case studies that were performed on 18 actual composite steelconcrete box girder bridges. 5. C. C. Method B seems to be superior to Method A for cases where the warping constant is not large.

12. and had a radius of curvature of 282 ft (86 m). M. No. The quantities measured were: (1) Boundary reactions.. with or without a midspan radial diaphragm. Laboratory Experiments Most. The comparison shows a good agreement between the shapes but not the magnitudes of the stress plots obtained experimentally and theoretically. G. The model was extensively instrumented with rosette strain gages at three cross sections. Page 31 .. Model Studies of Multicell Curved Box-Girder Bridges. Vol. A large-scale test of a concrete structure was performed at the University of California at Berkeley during the 1970’s. and a close agreement is found between the test and analytical results.  Fam.ii.47 m) deep. and Roll. 45 in. The prototype bridge was a four-cell reinforced concrete design. Experimental data for three lane-loading conditions was obtained.. and Turkstra C. (10. K. No. pp. 101. Static loads were applied at midspan to cause a complex pattern of membrane and bending stresses with the effects of diaphragms clearly evident. (1.  Aneja. A.140 mm). Model Analysis of Curved Box-Beam Highway Bridge. although not all. pp. J. and 30 in. (1. and (3) deflections at selected points.. and Godden. Vol. An approximate theoretical analysis of the model was obtained by using finite element method. pp.31 m) wide. No. 33 ft 10 in. 5. W. It showed that finite element models with curved shell elements provide better predictions than those with straight plate elements. Journal of the Engineering Mechanics Division. The data were reduced by computer.  Aslam. 1975. 1976. 2861-2878 Fabrication. 207-222 A model study on the static response of curved box-girder bridges is presented. A 1/29 scale aluminum model was studied for spans of 60 in. and deflections were plotted by Calcomp plotter. Journal of the Structural Division. Experimental data at the three gage sections for each load condition is also given. Based on the model data. F. 102. 97. 1097-1108 This paper describes an experimental study of a single span horizontally curved plexiglass boxgirder beam with diaphragms and flange overhangs. 1971. (1. I. and distribution graphs of tangential plate forces.. 3. Model Study of Horizontally Curved Box Girder. Experimental results in typical cases are shown graphically and compared to the results of a special purpose finite element program developed especially for curved box analysis. Vol. these tests have shown that refined analytical techniques predict structural behavior quite well. The following paragraphs discuss published papers and reports on these tests in greater detail. In general. laboratory experiments related to curved concrete box girder bridges have been conducted on small scale plexiglass or metal models of these bridges. (2) strains at a radial section close to midspan. preparation and instrumentation of a plexiglass model of a horizontally curved boxbeam highway bridge are described. some general observations are made regarding the behavior of curved box-girder bridges. (760 mm).M. R. Journal of the Structural Division. radial-bending moments. A typical comparison between the experimental and theoretical stress distribution across the mid-span gage section for one of the loading conditions is shown graphically. 4 ft 10 in.500 mm).

The model. 1975.82 scale replica of a prototype. This issue is also discussed in some textbooks (Menn. An algorithm is presented to deal with the torsional problem for reinforced concrete and prestressed concrete box girder bridge superstructures with multiple cell sections. al.. 831-843 The behavior of a single two-span.5 times the nominal design stress. 98. et. I-type. 103. A. al.7 m) wide. B. Good agreement was obtained between experimental and analytical results. This analytical technique had previously been applied to only open cross-sectional. Journal of the Structural Division. The loading to failure and the ultimate strength behavior is examined. The resulting experimental deflection. no published research exclusively addressing this issue were found. P. In this theory. Results are compared with previous theoretical and experimental work for single cell cases.. 1972. deflections. The model was tested under various static concentrated loads.. et. Bearings Although several bearing failures consisting of uplift. No. The radius of curvature was 100 ft (30. Elfgren. Scordelis.. Multicell Curved Girder Model Studies. However. Bonakdarpour. d. 1977. Until now the theory has been applied only to the case of pure torsion with a single cell section. Experimental and theoretical results are considered for reactions. Choudhury. 1525-1542 Results obtained in a study of a large-scale curved two-span four-cell reinforced concrete box girder bridge model are presented. Vol.. 1977. and strain data for some loadings are reported. bridge systems.  Heins.. No. Vol. al.. rotation. L. Journal of the Structural Division. pp. (Aslam. the concrete torsional problem is solved by combining equilibrium and compatibility conditions and constitutive laws of materials. et. L. Effects of single and multicell torsional properties are examined. and moments due to conditioning overloads producing stress values as high as 2.. 2002). pp. C. and Larsen. Design Issues i. 8. 1988. steel and concrete strains. This represents the sharpest curvature normally used for bridges in the California highway system. C. 4. G. Results indicate that single cell properties can be applied in the analysis. Ultimate Strength of Curved RC Box Girder Bridge. and Bell.  Scordelis. P. P. K. 1990). and warping effects may be neglected. had overall plan dimensions of 72 ft (21 m) long by 12 ft (3. which was a 1:2.. Conclusions appropriate for the design of this type of bridge are given. et. because an accurate 3-D analysis will account for differences in bearing forces and displacements. al. three-cell plexiglass model is predicted by the Slope Deflection Fourier Series Technique. C.. several references that deal with global analysis and laboratory experimentation deal with this issue. overload or binding have been experienced in curved box girder bridges. Page 32 .5 m). Sennah. The excellent live-load overload capacity of the bridge is evaluated and comparisons are made with the similar behavior of an earlier tested straight bridge model.This program used the softened truss model theory applied to a prestressed concrete multiple cell box.

121. 1986). normal warping. the testing of a fullscale. al. warping torsion. simple-span.. the effectiveness of the deck and soffit slabs in resisting flexural Page 33 . C. M.. J.. pure torsion. The bending and torsional distortions as well as cross-sectional distortions can then be determined throughout the curved box girder.. pp. These forces. The following two papers discuss research directed toward determining the number and spacing of interior diaphragms in box girder bridges. The research included a comprehensive literature review. a survey of design agencies. Journal of Structural Engineering. 10. 2161-2178 A finite difference procedure is used to determine the response of a single span curved single box beam bridge with any number of interior diaphragms. Construction details at the girder supports produced significant rotational-end restraint for both vertical and horizontal loading. the mechanism for resisting flexural and shear stresses in box girders is important. Diaphragm Effectiveness in Prestressed-Concrete Girder Bridges. W. Vol. 1995. 101. R. P/C girder-bridge model with eight intermediate diaphragm configurations. pp. in addition to distortional functions. and distribute transverse load. The mechanisms of shear resistance and its interaction with flexural stresses in reinforced and prestressed concrete have been well researched (Marti. Klaiber.  Abendroth. and the finite-element analyses of the bridge model assuming both pinned. Vol. Both the vertical and horizontal load distributions were found to be affected by the girder-end restraint. facilitate wheel load distribution. 1975. while the horizontal load distribution was a function of the intermediate diaphragm type and location. Also. No. Journal of the Structural Division. The effects of this type of loading on P/C bridge behavior were investigated for various types and locations of intermediate diaphragms. et.ii. F. 9. The vertical load distribution was determined to be essentially independent of the type and location of the intermediate diaphragms. A study of the data has resulted in a series of empirical design equations. The forces that are determined include bending moment and flexural shear.  Oleinik. 1999 and Vecchio. and Shafer. Flexure and Flexural shear Beyond the issue of global analysis.and fixed-end conditions. 1362-1369 Each year many prestressed-concrete (P/C) girder bridges are damaged by overheight vehicles or vehicles transporting overheight loads. A fabricated intermediate structural steel diaphragm was determined to provide essentially the same type of response to lateral and vertical loads that was provided by the reinforced-concrete intermediate diaphragms presently used by the Iowa Department of Transportation. and bi-moment. yield resulting normal bending. The technique is then used to determine the dead load and live load response of a series of typical curved box beams. and Heins. and normal distortional stresses. C. Diaphragms Diaphragms help prevent excessive distortions of the cross section. No. W. E. Diaphragms for Curved Box Beam Bridges. P. as well as a model without diaphragms.. iii.

This model was consistent with the results from the minimum potential energy method. A cross sectional analysis of shear stress in the flange is taken at several locations. Some of these are discussed below. This analysis showed that negative shear lag occurred only with the first load case of a distributed load. Positive shear lag is the phenomenon in which. Several published papers and reports have dealt with these issues. Near the fixed end where shear lag is greatest. With a finite element model analysis. and Zheng.. Finally. 1987. Negative shear lag is dependent not only on the load case but also the boundary conditions. following Reissner’s procedure with slight modifications. conclusions are drawn with regard to further study and research.” J. a distributed load. Shear Lag and negative shear lag effect in cantilever box girders are analyzed through a variation approach and finite element techniques. As the ratio increases. the bending stress away from the webs is greater than the stress near the webs. This includes the phenomenon commonly known as shear lag. The ratio of the length of the cantilever to the width of the box girder affects the amount of moment caused by shear. and a combination of a downward point load and an upward distributed force. three load cases were considered.. not only is positive shear lag more severe compared to a point load. the relevant conclusions are: Page 34 . a point load. T. negative shear lag occurs. This is a result of negative shear lag. In short. the conclusions in this paper. This paper addresses the classical phenomenon of shear lag in box girders. shows that the additional moment created by flange shear deformation plays an important role in both positive and negative shear lag. only positive shear lag is created. 113(1). 20–35. but negative shear lag is also present. The theoretical results obtained are compared with a plexiglass model test. both positive and negative shear lag decreases. the bending stress near the web is much larger than the stress away from the webs. flange longitudinal stresses near the web are larger than away from the web. Struct. Z. When there is a uniformly distributed load along the full span of the cantilever box girder however. S. The region of the cantilever affected by negative shear lag is from the free end to more than ¾ L from free end. This paper is only indirectly applicable to this project because the paper does not deal specifically with curved girders.  Chang. and the theoretical solutions are noteworthy. At a cross section where negative shear lag is significant.. F. Expressions are derived to determine the region of negative shear lag effect with the interrelation of span/width parameters involved. Eng. Negative shear lag affects a larger region than positive shear lag. However. away from the support. Actual testing using a plexiglass model confirmed the theoretical results. But for a cantilever box girder with constant depth under a uniform load. “Negative Shear Lag in Cantilever Box Girder with Constant Depth. the bending stress in the deck near the webs is smaller than the stress away from the webs. and draws attention to distinguishing between positive and negative shear lag. since shear lag effects are an important consideration in developing analysis and design strategies. When a uniform load is applied. Using the principle of minimum potential energy. near the support of a cantilever. For a single point load at the free end of the cantilever.compressive forces has been studied.

Although sagging moment is only local. This paper addresses the cantilever decks (“wings”) of single-cell box girder bridges. so it should not be ignored. The paper reports on a spline finite strip approach used to analyze the cantilever decks. J. the relative additional stress induced by this effect is often considerably greater. Effects of distortion of thin-walled box sections are taken into account by treating the cantilever deck as a slab with horizontally distributed spring supports along the cantilever root. A plexiglass model of a single-cell box beam was evaluated. many point loads at the same longitudinal location can cause significant sagging moment. 4. Tests showed that the moment along the cantilever root approaches zero as the longitudinal distance from the point load increases. The spline finite strip analysis was shown to be in close agreement with actual model test results. The sagging Page 35 . It should never be believed that in all cases only positive shear lag is produced. Struct. and Gang. Cantilever decks of thin-walled box girder bridges can be treated as cantilever slabs with horizontally distributed spring support along the cantilever root. it was observed that it is reasonable to treat the cantilever decks as cantilever slabs with horizontally distributed spring support along the cantilever root with the spring constant K depending on dimensions and material properties of the deck. taking into account the influence of local distortion of the box section. based on the simplified idealization of the cantilever decks of thin-walled box girder bridges. although the negative shear lag yield in the region of the bending stress is small.. In cantilever box girders. . but negative shear lag occurs only under uniform load. Simplified solutions are also given for the distribution of transverse moment along the cantilever root. The smaller the ratio. Z..1. are in close agreement with test results. “Analysis of Cantilever Decks of Thin-Walled Box Girder Bridges.  Chang. Spline finite strip results. It cannot be neglected. But it does present some useful qualitative information about cantilever deck evaluation. 2410 –2418.” J. S. Conclusions made in this paper are as follows: 1. The results based on the spline finite strip method are compared with those of the model test. T. Perspex model tests were conducted in the model structural laboratory at Tong Ji University. and does not make any distinction about the effects of horizontal curves. Negative shear lag is also dependent upon the ratio of L/b. As a point load moved transversely across the box girder. where b is the net width of the box section. 3.1990. There is also sagging in the cantilever around the point load. The maximum moment at the root is when the point load is at the end of the cantilever and approximately zero when the load is not on the cantilever. Positive shear lag may occur under both point and uniform load. From this analysis. the bending stress and membrane stress at the root of the overhang of the deck were obtained. 2. in general. 116(9). the more severe are the effects of positive and negative shear lag. Eng. Negative shear lag depends upon the boundary condition of displacement as well as on the external force applied to the girder. 2.

in the case of a uniformly distributed loading.1. In cantilever slabs of infinite length and large cantilever length. and d) box cross section vs. Eng. The parametric study involves calculating the effective width ratio for simply supported girders and comparing results of. there is no specific information pertaining to curved box-girders. and the inside and outside effective width ratio agree with the values of the straight beams.. When b/L = 0. effective width ratio for straight girders. the effective widths remain fairly constant for b/L = 0. 0 is half of the effective width for a straight beam and 2b = width of the flange.2.2 with a point load however. For a box girder under a distributed load.” J. the values are maximum at the mid span and decrease towards the supports. The thin-walled curved girders used in this investigation are based on box and channel cross sections.. irrespective of cross-sectional shapes or types of load distributions. Usuki. The methodology is formulated by substituting the flange stress derived from present theory into the equation of effective width definition for the curved girders. b) present theory vs. and Horie. channel cross section. respectively. Other than the general information provided for evaluation of cantilever decks of straight boxgirder bridges. a) point load at mid span vs. 3. The inner and outer effective widths are denoted as 1 and 2. S. 111(1). The required information in formulating the effective width rule for design of curved girder bridges is provided.. The actual longitudinal stress distributions for the curved girders are evolved from the present theory for shear lag in order to determine the effective width. “Shear Lag Analysis and Effective Width of Curved Girder Bridges. Mech. The values of the effective width obtained by the present theory are compared with those of the straight girder bridges. K. the values of the effective width ratio are minimum at the center of the span length and increase rapidly towards the supports. the effective width also remains fairly constant for R/L > 4.1 when R/L > 2 and only decreases slightly for b/L=0. 1985. This paper develops and guidelines and graphs for estimating the effective width of curved girder bridges. Page 36 . Numerical examples are shown for several problems to investigate the effect on effective width of curved girder bridges.. it is OK to say the values of effective width of curved girder bridges are the same (approximately) as the values of the straight girder bridges. In the case of a concentrated loading. and are analyzed for a uniform lateral load and for a concentrated load. c) inner and outer effective width ratio for curved girders vs. The authors analyze the inner and outer effective widths at mid span relative to the curvature R/L.1. This is a very important conclusion for the NCHRP 12-71 project. the effective widths deviate significantly for small R/L. The term effective width ratio is defined as the ratio of an effective width to the actual breadth of the flange. However.moment at the cantilever root can be obtained in conjunction with information tabulated in the article. The values of effective width ratio of the present theory are lower than the folded plate theory curves. folded plate theory. 87–92.  Hasebe. the curved beam has a radius of curvature R = 4L and b/L = 0. uniformly distributed load. Y. for a curved girder. In the test. The sagging moment may be taken from information also provided in the article. sagging moment cannot be ignored. With a point load and b/L = 0. According to the results.

For a curved girder, as b/L increases, the difference between the inner and outer effective widths becomes significant, especially for box girders. For small values of b/L however, those two values are almost identical. One needs to be careful when R/L is small, b/h is small, or b/L is large, especially under concentrated loads. Otherwise, it is reasonable to assume the effective width of a curved girder is equal to that of a straight girder for practical applications. iv. Torsion Torsion design is currently addressed by the AASHTO LRFD code and in the case of box girders is integrated with shear design. Recent studies (Rahal, et. al., 2003) indicate that this approach yields good results when the correct value of  is used. A number of other papers on the subject of torsional resistance have been reviewed and it is concluded that the current design methods are acceptable. Only minor clarifications of the current specifications and guidelines for applying them to box girders are required. Following is a summary of the papers that were reviewed on torsion design.  Collins, M., and Mitchell, D., 1980, Shear and Torsion Design of Prestressed and NonPrestressed Concrete Beams, Journal of the Prestressed Concrete Institute, V. 25, No. 5 This document presents design proposals for flexural shear and torsion of prestressed and nonprestressed concrete beams based on the compression field theory. A rational method is proposed which addresses members subject to flexural shear, torsion, combined flexural shear and flexure, and combined torsion, flexural shear and flexure. Early design procedures using truss models are also presented. The compression field theory, which is a development of the traditional truss model for flexural shear and torsion, considers in addition to the truss equilibrium conditions, geometric compatibility conditions and material stress-strain relationships. The compression field theory is capable of predicting the failure load as well as the complete load-deformation response. Measured and predicted response of numerous members is presented. The proposed design recommendations are provided in code format. Comparisons to the provisions of the ACI 318-77 and CEB codes are provided. Numerous worked examples are provided that demonstrate the proposed method.  Rahal, K. N., and Collins, M. P., 1996, Simple Model for Predicting Torsional Strength of Reinforced and Prestressed Concrete Sections, ACI Structural Journal, Vol. 93, No. 6, pp 658666. A noniterative method for calculating the ultimate torsional strength and the corresponding deformations of reinforced and concentrically prestressed concrete sections is presented. This method is based on the truss model. It avoids the need for iterations by making simplifying assumptions about the thickness of the concrete diagonal, the softening of the concrete due to diagonal cracking, and the principal compressive strains at ultimate conditions. A simple check on the spalling of the concrete cover is implemented. The calculated torsional capacities of 86 beams are compared with the experimental results and very good agreement is obtained.
Page 37

 Rahal, K. N., and Collins, M. P., 2003, Experimental Evaluation of ACI and AASHTO-LRFD Design Provisions for Combined Shear and Torsion, ACI Structural Journal, Vol. 100, No. 3, pp 277-282. The experimental results from four large nonprestressed specimens loaded in combined flexural shear and torsion are used to evaluate the torsion design procedures of ACI 318-02 and AASHTO-LRFD. Both sets of procedures calculate the required amounts of hoop reinforcement for torsion based on a space truss model with compression diagonals inclined at an angle of  to the longitudinal axis. It is shown that the ACI provisions give very conservative results if the recommended value of 45 degrees is used for . If the lower limit of 30 degrees is used, however, some unconservative results are possible. The AASHTO-LRFD provisions predicted values of  of approximately 36 degrees for these specimens and gave accurate estimates of the strengths.  Fu, Chung C, and Yang, Dailli, 1996, Designs of Concrete Bridges with Multiple Box Cells Due to Torsion Using Softened Truss Model, ACI Structural Journal, Vol. 93, No. 6, pp 696-702. Using a softened truss model, this paper presents a method for torsional design of multicell concrete bridges. Earlier researchers have successfully shown the development for solving single cell torsion by combining the equalibrium, compatibility, and the softened constitutive laws of concrete. By solving the simultaneous equations based on the membrane analogy, multicontinuous or separate cells can be solved.  Hsu, T.T.C., 1997, ACI Shear and Torsion Provisions for Prestressed Hollow Girders, ACI Structural Journal, Vol 94, No. 6, pp 787-799. New torsion design provisions have been proposed for the 1995 ACI Building Code. As compared to the 1989 provisions, these generalized 1995 provisions have three advantages: First, they are applicable to closed cross sections of arbitrary shapes. Second, they are applicable to prestressed concrete. Third, they are considerably simplified by deleting the “torsional concrete contribution” and its interaction with flexural shear. These new provisions are suitable for application to concrete guideways and bridges, because these large structures are always prestressed and are often chosen to have hollow box sections of various shapes. This paper discussed the background of the new code provisions, suggests modifications to code formulas, and illustrates the application of the code provisions to prestressed hollow girders by way of a guideway example.

Page 38

v. Wheel Load Distribution Wheel load distribution has been the subject of many of the analytical studies cited earlier. Other studies are discussed in the following paragraphs.  Song, S.T., Y.H. Chai, and S.E. Hida, 2001, Live Load Distribution in Multi-Cell Box-Girder Bridges and its Comparison with the AASHTO LRFD Bridge Design Specifications, Final Report to Caltrans for Contract Number 59A0148, July 2001

This report presents the results of a series of analyses done on box girder bridges with normal and skewed supports and straight and curved geometry. The analysis models included 3-D finite element as well as grillage, and single line of elements for superstructure. The goal of the project was to evaluate the LRFD live load distribution factors for box girder bridges. However, the modeling techniques used and verified for this project can be used as guidelines for NCHRP 1271.  Zokaie, T., K.D. Mish, and R.A. Imbsen, 1993, Distribution of Wheel Loads on Highway Bridges, Phase 3, Final Report to NCHRP 12-26 (2), December 1993

This report presents a computer program (LDFac) that was developed for modeling bridge superstructure with straight or skewed supports and obtaining live load distribution factors. Although this computer program did not consider curved geometry specifically, the modeling process and load placement guidelines may be used for analysis of curved bridges as well. One of the key issues discussed in this report is the modeling of distortion of box girders in a grillage analysis via an equivalent shear deformation parameter.  Zokaie, T., T.A. Osterkamp, and R.A. Imbsen, 1991, Distribution of Wheel Loads on Highway Bridges, Final Report to NCHRP 12-26 (1), 1991

This report presents a series guidelines for analysis of various bridge types. The guidelines include calculation of equivalent section property parameters to be used in plate and grillage analyses, as well as guidelines for setting the boundary conditions. Although the research did not specifically consider curved bridge geometry, many of the guidelines for modeling and analysis using common analysis tools is applicable to modeling that will be needed in NCHRP 12-71 global analysis studies. vi. Tendon Breakout and Deviation Saddles Prestress tendon breakout in curved bridges has occurred on a number of bridges over the years. It is evident from observing the reasons for these failures, that they can be prevented through close attention to details such as tendon spacing and tie back reinforcement. The following paragraphs summarize the references reviewed on the subject.

Page 39

M.. Texas. Deviation Saddle Behavior and Design for Externally Post-Tensioned Bridges. This report is the second in a series outlining a major study of the behavior of post-tensioned concrete box girder bridges with post-tensioning tendons external to the concrete section.J. Detailed instrumentation led to a very good understanding of the behavior of the various patterns of reinforcement in the deviators. Center for Transportation Research. Figure 3-1 . Beaupre. The lateral prestress force. This assumes the web to act as a simple beam spanning between the top and bottom slabs with a concentrated load. and Kreger. The results provide the basis for deviator design recommendations and several examples are presented to illustrate the practical use of these recommendations. California Department of Transportation. Page 40 . Bridge Memo to Designers Manual. R. 1988.  Caltrans.. Graphs are provided to check webs for containment of tendons and adequate stirrup reinforcement to resist regional transverse bending. University of Texas at Austin. Sacramento. California Memo 11-31 addresses the design of curved post-tensioned concrete box girders for lateral prestress forces. F. acting at midheight of the web. The regional transverse bending moment in the web is taken as Mu = 0.Caltrans Detail A A review of the 405/55 failure has led to the identification of several issues related to the Caltrans Memo to Designers 11-31.0. Austin. The report evaluates the results with respect to both simplified conventional analysis methods and strut-and-tie models.C. It presents the results of an experimental program in which ten very accurately sealed reinforced concrete models of typical tendon deviators were tested. several arrangements of tendon inclinations. The models included two very different patterns of detailing. The force effects considered are tendon confinement and web regional transverse bending. and both normal and epoxy coated reinforcement.E. These are discussed below. The design of stirrup reinforcement does not combine regional transverse bending and shear requirements. Research Report 365-2. Breen. L. is determined by dividing the jacking force (Pj) per girder by the horizontal radius (R) of the web.E. The load factor is taken as 1. F. J. The resulting simple beam moment is reduced 20% for continuity.20 F hc where hc is the clear distance between the top and bottom slabs.. Powell. A standard detail for tendon confinement (see Figure 3-1) is required for all webs with a Pj / R > 100 kN per m or a horizontal radius (R) of 250 m or less. Memo 11-31 Curved Post-Tensioned Bridges. 1996..

Page 41 .

The document provides a discussion of the actions on curved posttensioned girders. W. The article discusses the pullout of horizontal curved tendons that occurred on several cast-inplace post-tensioned concrete box girder bridges.. The document provides interim recommendations for the tensile stresses in the cover concrete. The Cause of Cracking in Post-Tensioned Concrete Box Girder Bridges and Retrofit Procedures. California This document presents the internal guidelines of David Evans and Associates. thin concrete cover over the tendons. lack of quality workmanship to meet the tolerances necessary for problem free structures.. improper or inappropriate construction techniques. K. Design of Curved Post-Tensioned Bridges for Lateral Prestress Forces. insufficient attention to stresses developed by curvature of tendons. the cover concrete is the only element restraining the lateral prestress force. 1985. Local slab action of the concrete cover over the tendons has been identified as the major cause of failure in several curved post-tensioned bridges that did not have duct or web ties. Specific requirements for local lateral shear are given in AASHTO LRFD 5. The primary focus is on the regional beam action of the webs and local slab action of the cover concrete over the tendons. the document recommends the use of a rational method for design such as a strut and tie model. Cracking is attributed in a broad sense to the following factors: inadequate flexural and shear capacity. March-April 1985 This article discusses the types of problems that lead to cracking in post-tensioned concrete box girder bridges and have been encountered in both Europe and the United States.3.1 “In-Plane Force Effects”. The cover concrete acts as a plain concrete beam to restrain the lateral prestress force. and the bundling of a number of large sized tendons close together. nonconsideration of thermal stresses.. The document provides interim recommendations for the design of the stirrups.4. Inc.10. Roseville. Journal of the Prestressed Concrete Institute. No specific design methodology for these combined actions is given by AASHTO. In two of these structures. Jr. Cordtz. for the design of horizontally curved post-tensioned concrete box girders for lateral prestress forces. Numerous worked examples are also provided. 2004. there was a combination of relatively sharp horizontal curvature. and recommends design procedures that reflect current best practice. No specific design methodology for local flexure is given by AASHTO.. Where duct ties are required. identifies those actions not completely addressed in current design codes and guidelines.  Podolny. Podolny divides the analysis of the failures of these bridges in three separate actions: Page 42 . Inc. David Evans and Associates. A flowchart is presented that outlines the recommended procedures. For a web without duct or web ties. and understrength materials. The local slab is subject to lateral shear and bending from the lateral prestress force. The vertical reinforcement in the web is subjected to combined global shear and regional transverse bending due to regional beam action. It is noted that in general the cause of cracking can be attributed to the superposition of stresses of multiple effects.

2001. if corrected would have prevented the damage. and then documents a step-by-step analysis of the 405-55 bridge. as suggested by Podolny. 2. The steps followed include global. This will spread out the radial forces and result in lower stresses tending to rip the strand out of the side of the web. Care is also required for tendons in the soffit of segmentally constructed bridges (straight or curved) where tendons are anchored in blisters along the length of the bridge and deviated across the width of the soffit resulting in transverse tensile stresses that can fail the soffit. California. The report provides background on the observed cracking and spalling (caused by horizontal breakout of web tendons in the multi-cell boxes). but concludes that the one cause which..  Seible. StD&A. Report TK21. it is recommended that prestress tendons be separated vertically where they pass near the middle of the web. The influence of prestressing in curved members is discussed. and points to some potential shortcomings (under the right set of unusual circumstances) of the Memo.. Local slab action of the concrete cover over the tendons. Prauge. Dameron. J. including illustrations) project report on results of structural evaluation of the 405-55 HOV Connector’s curved girder cracking/spalling problems.1. This document is a detailed (70 pages. and could by itself have overstressed some of the stirrups. Regional beam action of each web supported at the top and bottom flanges as a beam. Such a failure occurred in Page 43 . R. The global or overall girder action of the bridge together with it’s supporting piers and abutments. especially at the “regional” and “local” level for girder cross-section and web evaluation.. Structural Evaluation of the 405-55 HOV Connector and the Curved Girder Cracking/Spalling Problems.  Strasky. The regional and local analysis utilize detailed finite element modeling with cracking concrete constitutive models. While this report is very structure-specific and contains detailed project information that should not be directly referenced in a set of design Specifications. but various hand calculation methods for evaluating regional and local effects are also described and demonstrated as a check on the detailed analysis. F. It appears that for both of these bridges local slab action of the concrete cover over the tendons was the primary cause of the failure. San Diego. 3. and Hansen. summary of occurrences of similar problems at Las Lomas (California) and Kapiolani (Hawaii). Betonve Mosty. regional. The report then lists eight contributory causes for the damage that occurred.. Influence of Prestressing in Curved Members. Czech Republic. but the regional beam action could have been a contributory cause. 2003. The global action had a very small effect on these failures. To mitigate the effect of radial prestressing forces in the webs of box girders. it does provide a useful reference for summarizing the various analyses (both hand calculation and finite element) appropriate to this class of problems. was the omission of duct ties from the design. and local analysis. The report also provides in depth summary of Caltrans’s Memo to Designers (MTD) 11-31. B.

Van Landuyt. 1997. Concrete International. thin concrete cover over the tendons. Local slab action of the concrete cover over the tendons. D. and Breen. In both of these structures. and the AASHTO Load and Resistance Factor Design Bridge Design Specifications for tendon confinement. 2. American Concrete Institute. November 1997 The article discusses the pullout failure of horizontal curved tendons that occurred on several cast-in-place post-tensioned concrete box girder bridges. The article discusses the theory of transverse stresses in curved box girder cross sections due to post-tensioning including “distributed radial force arch action. The load path is envisioned primarily as a vertical one with the lateral prestress load carried through flexural bending of the web until cracking. there was a combination of relatively sharp horizontal curvature. the Texas Department of Transportation. Farmington Hills. MI.. A recommendation for a design methodology for the flexure of the concrete cover is not proposed due to the lack of understanding of its behavior. Page 44 . The article discusses the current design philosophy of the California Department of Transportation. J. the AASHTO Guide Specifications for the Design and Construction of Segmental Concrete Bridges. 1991) that discusses this research in greater detail was also reviewed. Tendon Breakout Failures in Bridges.. The global or overall girder action of the bridge together with it’s supporting piers and abutments. Regional beam action of each web supported at the top and bottom flanges as a beam. Both closely and widely spaced ducts (duct spacing less than or greater than one duct diameter respectively) were tested and different lateral shear failure modes were observed. the stiffness was reduced and the load was carried primarily through longitudinal arching until a local lateral shearing failure occurred.a bridge constructed with a gantry in Austria. Recommendations for lateral shear capacity are proposed that are more conservative than the AASHTO LRFD Specifications and yielded a consistently narrow range of factors of safety for the four test specimens (1. 3. and the bundling of a number of large sized tendons close together. The vertical curvature of these tendons in a haunched bridge can also present a problem. A test program was carried out by the authors on webs without tendon confinement reinforcement. It appears that for both of these bridges local slab action of the concrete cover over the tendons was the primary cause of the failure.” Horizontal curved post-tensioned bridges are subjected to the following three separate actions: 1.99 to 2. Once cracking occurred.34). The test specimens did not fail in regional transverse bending of the web due to the formation a second load path after formation of flexural cracks in the web. A University of Texas theses (Van Landuyt.

The mass condensation technique is used to reduce the number of coupled differential equations obtained from the finite element method. Vehicular Impact Vehicular impact in curved bridges is different than in straight bridges. R. 6) optimal seismic design of structures. 5) effect of automation on construction changes.vii. A vehicle. having component in the radial and transverse directions and moving with constant angular velocities on circumferential paths of the bridge. 9. The resulting differential equations are solved by the linear acceleration method.. Time Dependency The redistribution of stresses due to creep and shrinkage may be important in curved concrete bridges. viii. The proceedings cover five major areas of concern: 1) Computing in construction. 1993. Time-Dependent Analysis of Nonprismatic Curved PC Box Girder Bridges.. 101. A paper further exploring this issue is described below. June 7-9. Liu. Vol. Conference Proceeding Paper. The effect of centrifugal forces is considered and the effect of damping of the bridge is neglected in the analysis. Page 45 . L. 2) geographic information systems. pp. This issue is discussed in at least one of the references previously described (Menn. pp. O.  Zhang. as the applied time varying forcing function. and Huang. Dynamic Analysis of Curved Box-Girder Bridges. L. 2) artificial intelligence in highway CAD systems. 4) automated systems for construction bidding. and 7) CAD instruction for civil engineering students. S. 3) expert systems and artificial intelligence. 4) computing in structures. California. Within these broad topics. 1990). A paper addressing this subject is discussed below. Journal of the Structural Division.. 1899-1912 The finite element technique is used for the forced vibration analysis of horizontally curved box girder bridges.  Rabizadeh. A number of bridges with practical geometries are analyzed and impact factors are calculated. Computing in Civil and Building Engineering. and Shore. is simulated by two sets of concentrated forces. The book also presents several papers discussing different aspects of multimedia information systems and geographic information systems. bottom flanges and webs. 1993. and 5) computing in transportation. 1975. Annular plates and cylindrical shell elements are used to discretize the slab. Rectangular plate elements and pin-jointed bar elements are used for diaphragms discretization.. 1703-1710 This proceeding consists of papers presented at the Fifth International Conference on Computing in Civil and Building Engineering held in Anaheim. No. subjects such as: 1) Computer analysis of cable-stayed bridges. M.

. Portland. Electronic Journal of Structural Engineering International. It is concluded that the finite strip method offers an accurate and inexpensive tool for the optimization of box girder bridges having regular prismatic-type geometry with diaphragm ends and in curved planform.. 2005. automatic mesh generation. M. 2003..  Ozakea. This paper deals with the development of reliable and efficient computational tools to analyze and find optimum shapes of box girder bridges in curved planform in which the strain energy or the weight of the structure is minimized subject to certain constrains. Caltrans practice is to design the superstructure in bridges with monolithic columns to remain elastic. J. An automated analysis and optimization procedure is adopted which integrates finite strip analysis. at least one paper reviewed addressed this issue. Most of these deal with substructure response and are beyond the scope of this study. The finite strip method is used to determine the stresses and displacements based on Mindlin-Reissner shell theory. In this case the limitation on seismically induced torsional forces resulting from column yielding causes a redistribution of superstructure forces in the box girder. LRFD Approach. This study is described below. This unpublished study compared torsional superstructure design of a curved concrete box girder bridge subjected to seismic loading using the Load Factor Design approach currently used by Caltrans and the AASHTO LRFD method. Page 46 . but also by column plastic hinging of the single column bent during an earthquake. Paper presented at the Western Bridge Engineers Conference. Seismic Response Several papers and reports have addressed the seismic response of curved box girder bridges. Australia. Although design optimization is generally not the subject of design specifications. However. 3(2003). T. Analysis and Shape Optimization of Variable Thickness Box Girder Bridges in Curved Platform. and Tavsi.  Ibrahim. parametric cubic spline geometry definition.. A.. Torsional Analysis and Design of Curved Bridges with Single Columns . M. M. Design Optimization Several combinations of slab and web width can be selected to resist the applied loads. at least one presentation (no paper available) has implications for superstructure design. Soin. Masroor. and Quiogue. Queensland.ix. N. P.. sensitivity analysis and mathematical programming methods. x.LFD vs. Torsion is induced in the superstructure not only by curvature. OR.

Switzerland Detailing for Post-Tensioning includes discussions and examples demonstrating the forces that are produced by post-tensioning. particularly as it applies to box girders is well established and further refinement of these methods is beyond the scope of this project. Although some research work has been performed on skewed bridges. Article 4. and minimum tangent length requirements.xi.  Rogowsky. bearings and interior diaphragms. research can be broadly divided between steel and concrete bridge. If grillage analysis methods can be shown to produce reliable results. Sophisticated elastic analysis techniques such as finite element methods have been shown to produce excellent results that compare well with physical testing. those in anchorage zones and regions of tendon curvature. Part of our goal is to develop design procedures to handle these issues.M. Bern. Concrete bridges have been found to be stiff enough so that torsion warping can generally be ignored. Detailing The detailing of prestressed concrete in bridges is addressed by several agencies that commonly use this structural form. 1991. minimum radius requirements. It also seems that there are several potential configurations of curved box girder bridges that need further study from the designer’s point of view.. a large prestresser with international experience. P. and Marti. Torsion design. Conventional reinforced and prestress concrete design methods can be used for curved concrete box girder design provided accurate global demands can be established. Page 47 . in particular. The goal is to identify the simplest methods that can be safely used. Methods for preventing tendon breakout in thin curved webs include adequate lateral shear capacity of the concrete cover (adequate cover) or providing tieback reinforcement. Considerable work has been performed over the years in these areas. Detailing for Post-Tensioning. D. and as a verification tool for even less sophisticated analysis methods. With respect to global response analysis. most of it has not found its way into design specifications. is discussed below. SUMMARY: A considerable body of research has been conducted on box girder bridges.4 “Tendon Curvature Effects” deals with special issues associated with curved tendons including: in-plane deviation forces. VSL International Ltd. Much of this is useful to this project. Emphasis is placed on the use of strut-and-tie models to determine the tensile reinforcement requirements. than they can be used both in design. It is necessary for our project to explore the accuracy of less sophisticated methods such as grillage analysis.. A general reference published by VSL. out-of-plane bundle flattening forces. It is therefore not necessary to do any more sophisticated research on this subject. The radial force generated by a curved tendon is given as P/R where P is the tendon force and R is the radius of curvature of the tendon.

Page 48 .The local behavior of prestess tendons in curved concrete box girder bridges is an issue to be addressed by this project. Although excellent research has been conducted at the University of Texas )Van Landuyt. 1991) this needs to be studied further using available analytical techniques.

pier connection (bearing vs. and find approximate modeling methods to obtain accurate results for design purposes. monolithic). the magnitude of analysis cases desirable for parametric studies in this project is such that a more simplified model is desirable. cross sectional geometry and load patterns among other parameters. Page 49 . grillage models were used throughout the rest of the study as the basis of comparisons and such results are deemed to be accurate for all practical purposes. The structural analysis required to capture such effects. The curvature results in off-center placement of loads and subsequently such loads induce torsion into the superstructure. which can cause an increase the normal stresses on the outside edges of the bridge. The results for superstructure dead load and a concentrated midspan load obtained from the two models (grillage and finite element) for a two-cell box are compared in Table 4-1. which can result in significant tension on the inside of the curve and compression on the outside edge. available guidelines for grillage analysis were used. In order to make sure that the grillage and finite element models produced similar results. These guidelines may be used for design purposes when the bridge configuration requires it. The objective of the global analysis study in this research is to quantify the effect of increased shear force and normal stresses in the cross section. In particular. identify common trends. Given that the parametric studies are based on bridges with radial supports. Post-tensioned bridges also have an additional equivalent transverse load. The torsion in turn causes the shear stresses on the outside of the curve to increase. As a result. shear forces and normal stresses due to dead loads. and can result in higher tension and/or compression stresses. The magnitudes of such effects depend on radius of curvature. b Model Verification: Three-dimensional finite element analysis using plate and shell elements is accepted as the most accurate level of analysis available for box girder bridges. a Objective: The global behavior of a curved bridge can be distinctly different from a straight bridge. A set of special studies is also performed to review the effect of diaphragms. The models are shown in Figures 4-1 and 4-2. a detailed model of a three span bridge on a tight curve (400 ft radius) was created and results were compared for various load effects.4. Additional detailed results are presented in Appendix E. GLOBAL RESPONSE ANALYSIS STUDIES This chapter is a summary of the work performed for global response analysis and the results and conclusions obtained from this study. and post-tensioning are studied to obtain analysis modeling limitations and develop empirical adjustment factors for simplified analysis. the curved geometry of the bridge will result in development of transverse moments. Also. However. Guidelines for performing an analysis with a grillage analogy are included in Appendix C. span configuration. and skewed abutments on curved bridges. live loads. in most cases. These results indicate a very close comparison. is beyond the scope of day-to-day normal bridge design activities.

Figure 4-1 – Finite Element Model of the Bridge used for Model Verification Figure 4-2 – Grillage Model of the Bridge used for Model Verification Page 50 .

The study included the following cases: Four bridge cross sections shapes were considered:     Single-cell box based on a typical cast-in-place cross section Single-cell box based on a typical pre-cast cross section Two-cell box based on a typical cast-in-place cross section Five-cell box based on a typical cast-in-place cross section The typical cross sections are shown in figures 4-3.87 -4. 4-6.40 -3. without jeopardizing the global behavior of the bridge.31 -3.91 Gross Negative Bent 3 Moments (ft.91 1.80 1.88 0. 4-8 and 4-10.96 1. Note that the section geometry remains the same for various bridge span lengths with the exception of the overall section depth.86 0.19 0.01 1.Table 4-1 – Comparison of Results.02 1.26 -0.25 -0.96 1.02 -0.97 0. The cross section properties used for the analytical models are shown in Tables 4-2 through 4-5 Page 51 .23 -0.23 0.86 c Parameter Studies Analysis Cases – In order to study the effect of various bridge parameters on the response of curved bridges. kips) Left Girder Center Girder Right Girder -91261 -27469 -38265 -25528 -94623 -27309 -39494 -27820 0.10 Gross Midspan Span 2 Moments (ft. in order to simplify the analysis cases and automate the model generation.04 0.97 0. 4-7 and 4-9.92 -3095 -1007 -1342 -746 -3235 -845 -1526 -864 0. FEM – Two Cell Box Action Location Grillage DL FEM Grillage/FEM Grillage LL2C FEM Grillage/FEM Midspan Span 2 Deflections (inches) Left Right -4. the cross section was idealized as shown in Figures 4-4. and loading.01 0. 4-5. a parametric study was performed which was focused on the variation of span configuration and length.99 1.07 3791 1197 1566 1028 3829 1190 1504 1135 0. Also. bridge cross-section geometry. Grillage vs.99 1. kips) Left Girder Center Girder Right Girder 52103 14016 21523 16565 53523 16290 21735 15499 0.

Figure 4-3 – Typical Single-Cell Cast-in-Place Cross Section Figure 4-4 – Idealized Single-Cell Cast-in-Place Cross Section Page 52 .

Figure 4-5 – Typical Single-Cell Pre-cast Cross Section Figure 4-6 – Idealized Single-Cell Pre-cast Cross Section Page 53 .

Figure 4-7 – Typical Two-Cell Cast-In-Place Cross Section Figure 4-8 – Idealized Two-Cell Cast-In-Place Cross Section Page 54 .

five typical bridge models were considered:      Single Span – 200 ft long Single Span – 100 ft long Three Span – 200’-300’-200’ Three Span – 150’-200’-150’ Three Span – 75’-100’-75’ Each of the three span bridges was analyzed with two types of piers to identify the effect of softer vs. and also Page 55 .e. medium and long bridges respectively. 7’x7’ and 8’x8’ for the short. 600’. stiffer transverse pier stiffness. they have a moment of inertia of 5000 ft4.. Structural Analysis – Each bridge configuration was modeled as a spine model (in which one line of elements was used for superstructure. 800’. resulting in 192 bridge configurations. The softer columns are 6’x6’.Figure 4-9 – Typical Five-Cell Cast-In-Place Cross Section Figure 4-10 – Idealized Five-Cell Cast-In-Place Cross Section For each cross-section type. i. The stiffer columns are 6’x18’ which can also be considered similar in behavior to a multi-column pier. All column heights were assumed to be 50ft to the point of fixity at the base. and 1000’. 400’. Each of the above 32 bridges was configured as a straight bridge and as curved bridges with radii of 200’. located along the centerline of the bridge). It was assumed that the pier and abutment diaphragms are relatively stiff.

Each span element is limited to a central angle of 3. Note that each model uses several elements per span in the longitudinal direction of the bridge.5 degrees as recommended in Appendix C. Typical spine and grillage modeling techniques are shown in figures 4-11 and 4-12. Figure 4-11 – Typical Curved Line (Spine Beam) Bridge Model (showing 3-span unit) Figure 4-12 – Typical Curved Grillage Bridge Model (showing 3-span unit with 3 webs) Page 56 .as a grillage analogy (grid) model (where each web line was modeled as a line of elements along its centerline and transverse elements along radial lines were used to connect them in the transverse direction).

Furthermore.08 1.24 66.00 1. Note that the transverse element properties may be different for each element based on its width and therefore is shown here for a unit width.05 286.50 197.00 1.60 36.366.25 9.29 10. except that “Izz” is multiplied by L3.488.44 2. Table 4-2 – Section Properties for Two-cell Curved Line (Spine) Bridge Model 200 ft. Likewise. Note that the members on the outside of the curve are longer than the ones on the inside.60 15.0 4.20 1. The actual member property used in analysis was calculated by multiplying the values shown here by the average width of each element.133.73 56. and a set of properties for the transverse element.0 2.00 9. An example of these for the two-cell cross section is shown in Table 4-2. Multi-Span 75/100/75 ft.00 2.73 56.326.559. Single Span 100 ft.60 30.61 365. Multi-Span CG (ft) Area (ft2) Avy (ft2) Avz (ft2) Iyy (ft4) Izz (ft4) Jxx (ft4) Depth (ft) 4.037.80 7283.46 87.24 9.267.5 The section properties for the grillage model included a set of properties for the exterior girder. Single Span 200/300/200 ft.60 13. An example of these properties for the two-cell section is shown in Table 4-3.53 81. Multi-Span 150/200/150 ft.The section properties for the spine model were based on the entire section.028.01 65. the widths of transverse members along the outside of the curve are larger than those of the members along the inside.0 2.60 27.08 5.544.73 56. the transverse element shear area is calculated via a special formula to account for the warping on the cells.00 256.0 5.06 78.213.23 56.30 12.74 4.73 56.23 7. Page 57 . a set of properties for the interior girder.82 10.

63 34.50 1.87 4.50 216.87 9.50 30.87 12.63 0.01 1.11 95. Multi-Span 75/100/75 ft.00 73.01 1.87 9.43 519. Multi-Span 4.28 208.00 386.63 0.21 0.50 26.14 101.50 56.35 422. Multi-Span 6.97 95.58 2.00 301.77 3.00 106.13 399.03 18.87 4.Table 4-3 – Section Properties for Two-cell Grillage Bridge Model EXTERIOR GIRDER CG (ft) Area (ft2) Avy (ft2) Avz (ft2) Iyy (ft4) Izz (ft4) Jxx (ft4) INTERIOR GIRDER CG (ft) Area (ft2) Avy (ft2) Avz (ft2) Iyy (ft4) Izz (ft4) Jxx (ft4) Transverse Element CG (ft) Area (ft2) Avy (ft2) Avz (ft2) Iyy (ft4) Izz (ft4) Jxx (ft4) 200 ft. Multi-Span 150/200/150 ft.14 68.19 121.94 200 ft.14 54. Multi-Span 6.34 737.60 737.00 823.94 200 ft.75 219.05 Page 58 .63 7. Single Span 2.63 0.00 33.00 1.01 1.50 82.74 399.63 27.58 5 31.00 541.03 18. Single Span 100 ft.78 24.00 594.87 10. Multi-Span 2. Multi-Span 4.32 0. Single Span 200/300/200 ft.04 399.47 75/100/75 ft. Single Span 5.67 18.41 75/100/75 ft.38 0.03 18.76 519. Single Span 200/300/200 ft.14 11.53 18.58 5 1.17 18.69 150/200/150 ft.23 25. Multi-Span 2.01 1.67 18.45 207.63 5. Single Span 2.94 0.41 1.77 150/200/150 ft.59 399.06 20.13 27.69 100 ft. Multi-Span 4.69 100 ft.85 19.50 1.63 0.02 121.61 0.63 50.25 1.31 422.87 10.01 1.25 26.44 398.67 18.67 18.14 14.00 425.19 215.87 12.87 5.25 200/300/200 ft.03 18.63 0.87 5.

. iii) on center of bridge. An example of such graphs is shown in Figure 4-13. additional load cases were studied by loading only the inside and only the outside webs.33 7' x 7' Column 49 49 49 200.33 341.e. This loading captures the effect of concentrated axle loads and may magnify its effect on curved bridges to some extent.The bent cap and abutment diaphragms were assumed to be relatively rigid. This load was applied at the middle of the bridge and transversely was located at various positions: i) on outside web. the axial forces along the prestress path are applied at each end of each element which in effect is the same as modeling the prestress tendon as a series of straight lines. The prestress effects are modeled as equivalent loads at nodes. Midspan Normal Stress at bottom outside corner of cross section. at middle of center span Midspan Normal Stress at bottom inside corner of cross section. the results from this case were studied in more detail when developing guidelines for design.. and iv) on all webs.08 800.e. ii) on inside web. Post-tensioning was also applied along all webs of the section. at middle of center span The graphical review of results included scatter-grams of each response quantity from spine and grillage models.33 1365. Table 4-4 – Section Properties for cap and column elements Cap Section 108 108 108 5000 5000 5000 6' x 18' Column 108 108 108 324 2916 1296 8' x 8' Column 64 64 64 341. therefore. live load and Posttensioning loads.        Midspan Deflection at middle of center span Midspan Rotation at middle of center span Midspan Longitudinal Bending Moment at middle of center span Midspan Transverse Bending Moment at middle of center span Midspan End Shear at first abutment in single span and at start of middle span in 3-span case. i. Page 59 . equally distributed over the bridge width. In case of grillage models. Results Review – The following results were obtained for each load case and compared from spine and grillage models.33 6' x 6' Column 36 36 36 108 108 432 Area (ft2) Avy (ft ) Avz (ft2) Iyy (ft4) Izz (ft4) Jxx (ft4) 2 Loads – Each bridge configuration was subjected to dead load (self weight). therefore it is justified as a simplification for a conservative upper bound. i. Maximum stresses occur when the entire bridge width is loaded. Cap and column element properties are shown in Table 4-4.08 200. A concentrated load of 100 kips was used to simulate the live load.

925 0.945 0.9 0 1 2 3 4 5 6 7 Radius (ft) a) Midspan Outside Girder Shear Ratio b) Longitudinal Stress Ratio Figure 4-14 – Line Graphs of Shear and Stress Ratios of Results from Spine and Grillage Models Summary of Results – Numerous scatter-grams were plotted for the various items described on page 58 and bridge types shown in Figures 4-4 thru 4-10. with values furthest away from 1. Figure 4-14 shows the plots with the most divergent results (i.915 0.935 0.94 0. Midspan Outside Corner Stress (ksi) 0.92 0.00) between the spine and grillage models.Two-Cell CIP Dead Load: Midspan Moment -70000 -60000 -50000 -40000 -30000 -20000 -10000 0 0 -10000 -20000 -30000 -40000 -50000 -60000 -70000 Grid Model Figure 4-13 – Scatter-gram Comparison of Results from Spine and Grillage Models The ratio of stresses and shear forces from the spine model to those of the grillage model were also obtained to review the effect of radius of curvature and modeling technique.905 0. As a result. the outside web shear force and longitudinal stress are the main effects that need attention in design. An example of such graphs is shown in Figure 4-14a and Figure 4-14b.91 0. To better quantify these results and the effects of curvature on various Page 60 .e.93 0.

1 1.06 1. Figures 4-15 and 4-17 show the ratios of Curve to Straight Bridge spine models. points designated as PC1_6x6_sp3m_dl are for a single cell pre-cast bridge (PC1) with a 6 x 6 pier (6x6). The ratios of “Spine to Grillage Model” shown in Figures 4-16 and 4-18 reveal if a curve spine model can be used in cases of tighter curvatures and the limit of this type of model to obtain accurate results for design purposes. These results are the primary source of the recommendations for design at the end of this study.5 2 CIP1_6x18_sp3l_dl CIP1_6x18_sp3m_dl CIP1_6x18_sp3s_dl CIP1_6x6_sp3l_dl CIP1_6x6_sp3m_dl CIP1_6x6_sp3s_dl PC1_6x18_sp3l_dl PC1_6x18_sp3m_dl PC1_6x18_sp3s_dl PC1_6x6_sp3l_dl PC1_6x6_sp3m_dl PC1_6x6_sp3s_dl CIP2_6x18_sp3l_dl CIP2_6x18_sp3m_dl CIP2_6x18_sp3s_dl CIP2_7x7_sp3l_dl CIP2_7x7_sp3m_dl CIP2_7x7_sp3s_dl CIP5_6x18_sp3l_dl CIP5_6x18_sp3m_dl CIP5_6x18_sp3s_dl CIP5_8x8_sp3l_dl CIP5_8x8_sp3m_dl CIP5_8x8_sp3s_dl Figure 4-15 – Ratio of Outside Web Dead Load Shear Forces in Curved (Vr) to Straight (Vs) Bridges where Length Equals Middle Span Length Figures 4-15 thru 4-19 are for the four bridge cross-sections (cast in place single cell (CIP1).12 1.bridge types. Therefore. The stresses are also obtained from the combined effect of axial force. and cast in place five cell (CIP5) with different pier configurations and span lengths. precast single cell (PC1).16 1.04 1. Average ratios and standard deviations are also calculated for each group of bridges with same cross-section type and loading. This figure reveals the limit of radius of curvature beyond which the curvature effects may be ignored altogether and the bridge may be analyzed as if it was straight.08 1. cast in place two cell (CIP2). shear forces are obtained by considering the shear flow from torsion in addition to the vertical shear. Note that in the case of the spine models. ratios of results for various bridge models are combined and plotted as shown in Figures 4-15 through 4-18. Page 61 .5 1 Length/Radius (L/R) 1. All 3-Span V r/V s Dead Load 1.18 1. and longitudinal and transverse moments.14 1. 3 span configuration of medium length spans (sp3m) and dead load response (dl).02 1 0 0.

98 0.5 1 Length/Radius (L/R) 1.02 1 0.06 1.All 3-Span Vline /Vgrid Dead Load Total 1.5 2 CIP1_6x18_sp3l_dl CIP1_6x18_sp3m_dl CIP1_6x18_sp3s_dl CIP1_6x6_sp3l_dl CIP1_6x6_sp3m_dl CIP1_6x6_sp3s_dl PC1_6x18_sp3l_dl PC1_6x18_sp3m_dl PC1_6x18_sp3s_dl PC1_6x6_sp3l_dl PC1_6x6_sp3m_dl PC1_6x6_sp3s_dl CIP2_6x18_sp3l_dl CIP2_6x18_sp3m_dl CIP2_6x18_sp3s_dl CIP2_7x7_sp3l_dl CIP2_7x7_sp3m_dl CIP2_7x7_sp3s_dl CIP5_6x18_sp3l_dl CIP5_6x18_sp3m_dl CIP5_6x18_sp3s_dl CIP5_8x8_sp3l_dl CIP5_8x8_sp3m_dl CIP5_8x8_sp3s_dl Figure 4-16 – Ratio of Outside Web Dead Load Shear Forces from Spine to Grillage Model Page 62 .96 0.08 1.9 0 0.04 1.94 0.92 0.

02 1 0.1 CIP1_6x18_sp3l_dl CIP1_6x18_sp3m_dl CIP1_6x18_sp3s_dl CIP1_6x6_sp3l_dl CIP1_6x6_sp3m_dl 1.4 1.04 1.2 0.06 1.08 CIP1_6x6_sp3s_dl PC1_6x18_sp3l_dl PC1_6x18_sp3m_dl PC1_6x18_sp3s_dl PC1_6x6_sp3l_dl PC1_6x6_sp3m_dl PC1_6x6_sp3s_dl CIP2_6x18_sp3l_dl CIP2_6x18_sp3m_dl CIP2_6x18_sp3s_dl CIP2_7x7_sp3l_dl CIP2_7x7_sp3m_dl CIP2_7x7_sp3s_dl CIP5_6x18_sp3l_dl CIP5_6x18_sp3m_dl CIP5_6x18_sp3s_dl CIP5_8x8_sp3l_dl 1.98 0.4 0.2 1.6 0.96 0 0.8 1 1.All 3-Span fr/fs Dead Load 1.6 Length/Radius (L/R) CIP5_8x8_sp3m_dl CIP5_8x8_sp3s_dl Figure 4-17 – Ratio of Outside Corner Dead Load Longitudinal Stress in Curved to Straight Bridge Page 63 .

40 1.00 0.20 1.8 shall be analyzed with more detailed analysis models such as grillage or finite element model.80 1.2 and an L/R larger than 0. Figures 4-15 and 4-17 reveal results are within 4% of a curved spine model when this is done. Bearing forces (i.60 0.97 0.94 0. short span 5-cell sections) Bridges with L/R larger than 0. Figure 4-18 shows that longitudinal stresses will be unconservative by less than 4% up to this limit unless the bridge has a low span length to width ratio (i. Bridges with L/R less than 0. Conclusions of the Parametric Study – The above results have been studied and the following general conclusions are drawn:   Bridges with L/R less than 0.8 L/R limit. Figures 4-16 and 4-18 reveal spine beam analysis will generally become increasingly unconservative beyond the 0.8 may be modeled with single girder spine model using a curved (spine) model and lateral effects shall be included in the analysis.e.2 also require detailed analysis as revealed by the unconservative results for tightly curved short span 5-cell bridges in Figures 4-16 and 4-18. Curved bridges with length/width ratios of less than 0. Page 64    .95 0.20 0.2 may be designed as if they were straight. Shear at the abutments) obtained from a spine model must consider the effect of torsion.00 1. Bearings must be designed considering the curvature effect.96 0.60 CIP1_6x18_sp3l_dl CIP1_6x18_sp3m_dl CIP1_6x18_sp3s_dl CIP1_6x6_sp3l_dl CIP1_6x6_sp3m_dl CIP1_6x6_sp3s_dl PC1_6x18_sp3l_dl PC1_6x18_sp3m_dl PC1_6x18_sp3s_dl PC1_6x6_sp3l_dl PC1_6x6_sp3m_dl PC1_6x6_sp3s_dl CIP2_6x18_sp3l_dl CIP2_6x18_sp3m_dl CIP2_6x18_sp3s_dl CIP2_7x7_sp3l_dl CIP2_7x7_sp3m_dl CIP2_7x7_sp3s_dl CIP5_6x18_sp3l_dl CIP5_6x18_sp3m_dl CIP5_6x18_sp3s_dl CIP5_8x8_sp3l_dl CIP5_8x8_sp3m_dl CIP5_8x8_sp3s_dl Length/Radius (L/R) Figure 4-18 – Ratio of Outside Corner Dead Load Longitudinal Stress from Spine to Grillage Model Detailed results from the parameter studies are presented in Appendix E.All 3-Span fline/f grid Dead Load 1 0.99 0.e.98 0.40 0.

live load and post-tensioning. Figure 4-19 – Scatter-gram Comparison of Live Load Results from Grillage Models With and Without Interior Diaphragm Page 65 . and therefore may be eliminated altogether. These results were compared using scatter-grams and ratios (line diagrams) similar to the results shown in Figures 4-19 and 4-20. The results from each model with and without diaphragms were compared for dead load. bearing conditions. a set of special studies were performed to get a better understanding of the effect of diaphragms. Diaphragm – All five-cell grillage bridge models that were used in parametric study were also modified to have a stiff diaphragm in the center of each span.d Special Studies In addition to the above parametric study. The overall conclusion from these results is that interior diaphragms have minimal effect on the shear and stress responses. skewed abutments and long term creep when combined with curved geometry.

in general. The results were studied using scatter-grams and ratio (line) diagrams comparing the results from spine to grillage models. the same guidelines for modeling limitations are equally applicable to integral and non-integral conditions. the two-cell finite element model used in the model verification studies was modified to include interior diaphragms. Page 66 . Bearings at the Bents – The purpose of this study is to determine if bridges with integral and non-integral bents respond differently to curve effects. It was found that. i. These diaphragms were located on one side of the midpoint of the bridge. free to move in transverse or longitudinal directions. the diaphragms were shown to have very little effect on the global response of the bridge.e. These diaphragms were placed at the center of one of the end spans and at one of the third points in the center span.Figure 4-20 – Line Graph of ratio of Results from Grillage Models With and Without Interior Diaphragm To verify the conclusions relative to interior diaphragms. All three-span five-cell spine and grillage bridge models that were used in the parametric study were also modified to have a point bearing support at the piers. Therefore. which was otherwise symmetrical. Therefore. Results were compared on both sides of the bridge and found to be nearly symmetrical. the final conclusion is that as long as the support conditions are modeled correctly.. magnitudes of curve to straight results are in the same order as the integral bridges.

The simple span bridge results were compared for dead load and live load and the three span results were compared for dead load responses.49156869 -190. These corrections will be necessary when the bridge is designed using a spine beam analysis. see Figure 4-21. The obtuse and acute abutment shear values were compared in straight and skewed bridges with and without curved alignment.3547734 -782.96775213 Straight Radial Skew-Left Skew-Both 400’ Radius Radial Skew-Left Skew-Both .0092205 -392. any consideration taken for straight bridges can be equally valid for curved bridges.35 1.65 0. The abutments are supported on rollers.89 1. Skew at One Abutment Only Skew at Both Abutments Figure 4-21 – Skew Configurations Studied Table 4-5 . The final outcome of these results is that the curved alignment does not aggravate the effect of skewed abutments and therefore.66 Ratio 1 0. The results are shown in Tables 4-5 through 4-7.279 -183.78 -58.Dead Load Shear Results for 200 ft/Single Span Skew Obtuse (Element 1) Ratio Value 1 -389.14 Page 67 Acute (Element 21) Ratio Value 1 -387.84 0.23 Acute Value -189. Skew can be automatically accounted for in a grillage analysis approach as shown in Figures 4-21.94 Obtuse (Element 1) Ratio Value 1 -577.30708715 0.0627282 -614.55387184 -605 1.Skewed Abutments – A two-cell single span (200 ft long) and a two-cell three span bridge (200ft-300ft-200ft) were modified to have 30 degree skew support at the left support and another case with 30 degree skew support at both ends.88239996 -342.91 1.

In addition.84 1.02 Obtuse (Element 1) Ratio Value 1 -370 1.079 1.81 -135.96775213 Acute Ratio 1 -1.66539668 0.66 Value -202.0432248 0.91 1. The same structure used in the comprehensive example problem was analyzed over a period of 10 years.35 -197.89 1.007 12.22672973 -453.27388599 -378. In all cases the abutments were fixed against torsion.15 -85. four models were evaluated.279 -183.618 1.79 1. 200/300/200 ft Multi-Span Skew Straight Radial Skew-Left Skew-Both 400’ Radius Radial Skew-Left Skew-Both Obtuse (Element 1) Ratio Value 1 -297.35 1.89 Acute Value -296.67166172 Long Term Creep – The effect of the time dependant properties of concrete (principally creep) on the response of curved bridges was investigated using the LARSA 4D computer program.59 -197.81 Ratio 1 0.22389189 -452.Table 4-6 – Live Load Shear Results for 200 ft/Single Span Skew Obtuse (Element 1) Ratio Value 1 -577.42309295 -116. The bridge was modeled as a 3-D spine beam.78 -58.66644189 Acute Ratio 1 0.3547734 -782.04240486 Acute Value -189.66 Value -12.76052303 Straight Radial Skew-Left Skew-Both 400’ Radius Radial Skew-Left Skew-Both Table 4-7 – Dead Load Shear Results for 400 ft Radius.0627282 -614.67166172 0.1316 Ratio 1 0.30708715 0. the model was modified by changing the curve radius in two cases and changing the length of the end span in another case.14 Obtuse (Element 1) Ratio Value 1 -81. Therefore.526 -9. The end span and radii of these bridges are shown in Table 4-8 Table 4-8 – Models of Two-cell Bridge Used to Study the Effect of Creep Model Number 1 2 3 4 Radii (ft) 400 800 1600 400 Page 68 End Span Length (ft) 200 200 200 140 .2 -135. which can consider both the three-dimensional geometry of the bridge plus the time dependant behavior of concrete.27129645 -378.

The major concern.9 32. The above study assumed bridges constructed on falsework. The forces in outside bearings will increase and inside bearings will decrease. however. it is recommended that dead load torsion reactions at the abutment from a three-dimensional spine beam analysis be increased by approximately 20% for the final condition and bearings or bearing systems be designed to envelope both the initial and final conditions. Therefore. was abutment bearing reactions. as would modeling the flexibility of bearing systems subjected to torsion loads from the superstructure.7 25. it would appear to be safe to analyze curved concrete box girder bridges using commercially available software. Table 4-9 – Time Dependence of Abutment Torsion Moment Model Number 1 2 3 4 Torsion Moment 7 Days (ft-kips) -8826 -7357 -6395 -4012 Torsion Moment 3600 Days (ft-kips) -10034 -9262 -8500 -5842 13. it is difficult to make an accurate assessment of time dependent bearing forces.The long-term deflection of models 1 through 3 did not appear to be affected by the radius of the bridge.6 % Difference The following conclusions can be drawn from this limited study:  The torsion reaction and thus the relative bearing forces in continuous curved concrete box girder bridges vary over time and are dependent on both the radius of the curve and the relative length of the end span with respect to the center span. This is a crude recommendation. Given the many possible bridgeframing configurations. as is the case with most other commercially available software. which in turn will affect the bearing reactions. The LARSA 4D program does not consider torsion creep. Torsion creep is expected to mitigate long-term changes in bearing forces. In the absence of such an analysis. methods used for adjusting cambers for straight bridges would appear to be applicable to curved bridges analyzed as three-dimensional spine beams.9 45. These will change over time.   Given these conclusions. Segmentally constructed bridges are expected to behave differently. This should definitely be done for segmentally constructed bridges. This is manifested by the change in the torsion reaction at the abutment. Table 4-9 summarizes the dead load and prestress results from the four models investigated. but given the mitigating factors not considered Page 69 . It is recommended that the vertical flexibility of bearings be considered in these analyses.

in this limited study. increasing that moment by 20% and recalculating the new bearing forces. it should provide a reasonable hedge against bearing failure. In the case of a grillage analysis. the same adjustment can be made by resolving bearing reactions into a torsional moment. Page 70 .

This is sometimes referred to as Lateral Tendon Breakout (LTB) Global or overall girder action of the bridge together with its supporting piers and abutments is covered in Chapter 4. Local action of the concrete cover over the tendons. For a web without duct or web ties. and/or local lateral shear/breakout failure adjacent to the ducts. Note that the system is in static equilibrium and the support reactions will be zero. 2. that have been considered in the analysis work for this project. Figure 5-1 Types of Actions Considered 1. Global or overall girder action of the bridge together with its supporting piers and abutments. The regional moments can be determined from a 2-D frame analysis of the cross section. The prestress lateral force is determined individually for each web with due consideration for the allowable variation in prestress force between webs. Regional beam action of each web supported at the top and bottom flanges as a beam. The compressive reactive forces on the concrete are applied as distributed loads on the webs and as concentrated loads at the centerlines of the slabs. LOCAL RESPONSE ANALYSIS STUDIES There are three types of actions.5. as shown in Figure 5-1. the cover concrete is the only element restraining the lateral prestress Page 71 . Local slab action of the concrete cover over the tendons has been identified as a major cause of failure in several curved post-tensioned bridges that did not have duct or web ties. This chapter focuses on “regional” and “local” action. 3. Regional beam action considers each web as a beam supported at the top and bottom flanges.

Finite element (FE) models were developed to perform parameter studies. F. Detailed 3-D. and prescribe a methodology to prevent such damage. 1991). a Local Analysis Validation / Demonstration Case (UT Test Case) The local analyses were begun with a validation case simulating Specimen “BC” from prior research conducted at the University of Texas (Van Landuyt. The case studied is a 1/3 scale representation of the configuration of Las Lomas. and these in combination with strain contour and crack pattern plots show the evolution of damage with increasing force. nonlinear analyses were performed using ABAQUS Version 6.5 ‘damaged plasticity’ concrete cracking model. investigate capacities and damage modes. Mn jd V LOCAL MOMENT C T LOCAL SLAB ML L F dmin W = F/L d min C T V Mn V ~ F/2 T ~ Mn /(jd) ML ~ FL/12 Figure 5-2 Regional and Local Actions on a Web Detailed local analysis models were used to evaluate the local stresses resulting from longitudinal tendons generating transverse forces on curved webs. The web is subject to regional transverse bending which results in tensile stresses on the local slab. The cover concrete acts as a plain concrete beam to restrain the lateral prestress force. It should be emphasized that the focus of the FE models was on the comparison of parameter sensitivities rather than on the prediction of “post failure” behavior.force. Comparison of such plots amongst different geometries and reinforcing schemes provide quantitative and qualitative parameter sensitivity evaluation. Page 72 . Such models provided tendon horizontal force plots versus deformation. The “local slab” is subject to lateral shear and bending from the lateral prestress force. which is a well-known bridge that failed in lateral tendon breakout. as shown in Figure 5-2.

anchorage zone failure may have resulted or the strands would have been loaded to an unsafe level. At Las Lomas. The curved region was the most difficult part of the model to construct and was kept as short as possible. The previously determined cell width mandated that the companion web radius be 22’. Duct arrangement controlled the design of the curve radius. If the curve was not sharp enough.3 inches was scaled to 6mm Page 73 . This was more than a third larger than the anticipated failure load. etc. 1991.2 k/ft was required for all four tendons. (The actual cantilever length has little effect on web behavior. as it would permit a total Fr of 20. The top slab thickness at Las Lomas varied transversely and the bottom varied longitudinally.8k / ft where Fr is the lateral (radially oriented relative to the curve) prestress force. A 5’ curve was more than twice the clear height of the web and was thought to be sufficient to allow regional transverse bending. scaled and rounded off to an integral number of inches.125  3. Equivalent spacing of #6’s at 21. The curve length of 5’ was chosen for the 18’ radius. This did not significantly affect results (it was assumed by the researchers that gravity loads are unimportant to breakout). A jacking force of 372 kips could be delivered from the loading apparatus. 75 ksi bars from Sweden already available in the lab. The model top slab therefore represents the bottom slab at Las Lomas. Therefore a total Fr of 15. The stirrup spacing needed to be adjusted to reflect the imprecise scaling of stirrup sizes. Van Landuyt.6 k/ft. Las Lomas was reinforced with GR40 #5 stirrups spaced at 15”.” by D. Fr  2 5000 2 121. the cantilevers were shortened slightly from a scaled value of 1’-10” to 1’-4”. The closest match was 6mm.W. No standard bars match this on a 1/3 scale. The capacity to resist lateral shear failure was calculated for each tendon assuming two failure planes would form and that the maximum concrete strength would be 5000 psi. The model was constructed with a centerline web-to-web distance of 4’ so that the radius of each web was a whole number.) The dimensions of the web were considered important. A 5’ straight transition zone on each end insulated the curve from the complex stresses at anchorages. To save on materials and labor. The box cross-section was a scaled version of Las Lomas with changes made for simplifying construction (Figure 5-3). This scaled to 3’-8”. Average thicknesses were calculated.Test Model and Test Conduct The following test model and test conduct description comes from the test report and thesis: “The Effect of Duct Arrangement on Breakout of Internal Post-tensioning Tendons in Horizontally Curved Concrete Box Girder Webs. An 18’ radius for the inside of web was selected. the centerline distance between interior and exterior webs was 11’. To avoid having to build cantilever forms. The exact scaled values of 4” for the thickness and 3’ for the overall height were maintained (Figure 5-4). The web radii were chosen to be small enough so that the tendon breakout in the web would occur before failure of any other part of the girder or testing apparatus. the girder was built and tested in an inverted position. This is nearly equivalent to a #2 bar.

75”. The nearest available duct size (1. Specimen 1.0DC follows the Texas State Department of Highway and Public Transportation design of the San Antonio “Y” project with an arrangement that maintained a clear spacing between ducts equal to the diameter of the duct.75”) was used. The scaled vertical spacing was 1. a vertical stacked bundle would have been 17½”. All ducts were centered on the web vertical axis. ½”Ø. That same type of load distribution could be approximated with a minimum of three ½” diameter strands per duct (Figure 5-5). Page 74 . Post-tensioning was applied to the specimens with 7-wire. The ducts at Las Lomas were nearly filled to capacity with strands.bars at 7”. A four-tendon bundle and three other promising arrangements were tested. Galvanized. 270 ksi. Since there were four ducts. All test concrete strengths were much greater than the 28-day design strength of 3500 psi. low relaxation strands. Stirrup spacing was reduced in the anchorage zones to 5½”. 6mm diameter hot rolled bars were used for all reinforcement. eliminates the single large discontinuity found at Las Lomas. All ducts are centered on the web vertical axis. Only duct positioning varied from web to web. should have been used. more importantly. The spacing was not increased to account for the greater yield strength of the Swedish bar. Scaling required 1½” ducts for the model.035”. It was necessary to consider how the strand or strands that would constitute a tendon would bear on a duct. all other details remained the same. ducts of approximately 4½” O.60” and the gauge was 0. The slab and web concrete had higher overall strengths and faster strength gains as is typical of concrete containing super-plasticizers.75fpu to develop a maximum force of 372 kips. This arrangement is conservative and would be considered an upper limit beyond which further spacing of ducts would provide no benefit. The vertical bundle height at Las Lomas was approximately 16”. Test data provided with the strand showed an actual yield of 276 ksi and ultimate of 289. Tensile tests on bars conducted at the lab showed average yield strength of 75 ksi. The relative horizontal offsets between ducts in a zigzag pattern can change from bridge to bridge depending on the clear distance between the stirrup legs and the diameter of the ducts.75”. A straight vertical formation was used in lieu of the zigzag because it was considered more universal. corrugated. A slight modification was made for the model (Figure 5-4). No large difference in behavior between the two arrangements was anticipated. The web concrete strength was 5300 psi. The inside diameter was 1. a total of twelve strands could be safely stressed to 0. but the major manufacturers of post-tensioning duct do apparently not make this size. This meant that almost the entire 180 degrees of the duct on the inside of the curve was in contact with the strands. It is believed that this allows for better consolidation and.D. folded metal ducts were used in all instances. Duct size at Las Lomas was not given although based on the maximum number of strands (28). The outside ridge-tooutside ridge dimension was 1. A loading system was developed to apply gradually increasing load simultaneously to all tendons with an equal force in each tendon.5 ksi. Specimen BC is the duct arrangement similar to the one at Las Lomas.

This solution procedure is often referred to as Full-Newton.0DC. The description begins with the letters WS to signify that deflections of the web relative to the slab are being measured. Ideally. causes “longitudinal” compression prestress in the concrete in a manner similar to the actual test specimen. The FE model and boundary conditions are illustrated in Figure 5-10. A small mirror glued to the specimen provided a smooth surface on which the potentiometer stem could bear. The wedge slice model configuration and “symmetry” boundary conditions allowed horizontal (radial) displacement to occur naturally in the FE model. and this in turn. The web-slab potentiometer nomenclature is given in Figure 5-6. The number of the stirrup nearest the potentiometer follows these letters. as was shown in Figure 5-3. U-shaped frames were mounted to the web face on the outside of the curve (Figure 5-6). displacements are magnified by 50. Also potentiometers attached to the back face were protected from exploding concrete. which could be mounted on either web.07 inches) is shown in Figure 5-11. Figure 5-7 shows a photo of the test specimen and Figure 5-8 shows a sketch of how the concrete cracked and failed during the test. where upon the next load increment. The analysis was run out to large web displacements and significant damage (failure).0DC was developed for a typical slice of the test model in the curved region. Mounting the potentiometers on the outside face of the curve meant that deflections should not be influenced by local beam action. the frame should have been mounted on the slabs. What is precisely mathematically represented with such boundary conditions is a wedge-slice of a complete circle. This permitted the construction of only one type of frame. A single potentiometer was mounted at the mid-height of the frame (which is also the c. but the boundary conditions are also reasonably accurate for a section of a curve. well beyond the displacements plotted in Figure 5-9. of the tendon group). A typical deformed shape from the analysis (at the 1. This agrees with test observations. Page 75 . based on the material damage. regional beam behavior should be solely responsible for measured deflections. It can be noted that the amount of deflection and the sharpness of the flexural curvature is significantly more severe in Girder-BC than in Girder-1.. the structure stiffness is updated. The cross-section and duct geometry (as was shown in Figure 5-4. i.e. The actual attachment points were about 2” below the top slab and 2” above the bottom slab. Web delaminations were measured by wires/potentiometers placed in tubes cast through the webs above and below the duct group. Figure 5-9 shows a comparison of tendon horizontal force versus deflections for the four different webs tested. which has occurred (concrete cracking/crushing and steel yielding).0 DC web displacement of 0. Sudden movements in these measurements were good indicators of imminent failure. This shows how a 3-dimensional slice was modeled with horizontal tendon loads applied directly to the inner surfaces of the ducts.g. 5-5) were modeled per the test configuration and dimensions. However the deflection anticipated in the first 2” is negligible. The tendon forces were applied in incremental fashion with enough equilibrium iterations and small load increments to achieve solution convergence at every increment.Regional beam behavior was monitored by deflection of the web relative to the top and bottom slabs. In the illustration. Finite Element (FE) Model and Analysis A FE model of Girder BC and 1.

Figure 5-14 shows the same deformed shape as Figure 5-13. the stress reduces to nearly zero. which occurs at strain of approximately 1. but are easier to see in cross-section. that the FE representation and modeling strategy is appropriate for moving on to the prototype models. Using this information. the analysis is simulating the response (initial stiffness and stiffness degradation with accumulating damage) and the lateral force capacity reasonably well. In making this conclusion.0DC girder webs.00E-4.Figure 5-12 shows the same deformed shape. It is concluded from these comparisons. but in an end view at a lower displacement. In reinforced concrete analysis.0DC. Comparing to Figure 5-8 shows similar trends in the location of the cracking as compared to the test. so a zero or small tensile stress may be displayed in a zone that is already highly damaged. and much more severe cracking at the BC configuration than at the 1. because after concrete cracks. Here the crack zones are also fully formed. Another significant analysis versus test comparison is shown in Figure 5-15: the lateral force versus deflection at the mid-height of the duct bank relative to the slab. Page 76 . but the cracks shown in Figure 5-8 were the primary failure planes. these are one of the most effective ways to show damage and deformation distributions in the concrete and rebar. is used to estimate the extent of cracks. and judgment from experience with many similar analyses. The comparison shows that for both the BC and 1. but with contours of maximum principal strain. Figure 5-13 shows similar strain contours. But maximum principal strains (generally. and from experience with many similar reinforced concrete analyses. it is assumed that the test also showed other small cracks in the vicinity of the ducts. but viewed in conjunction with principal strains. Figure 5-12 shows widespread cracking in the vicinity of the ducts. Concrete stress contours are generally not helpful. actual cracks implied by the analysis are drawn onto the figure.50E-4 to 2. the maximum tension found in any orientation for a given point in the continuum) can indicate concrete cracking.

T.Figure 5-3 Plan and End View of U. 1991) Page 77 . Girder Test Specimen (from Van Landuyt.

Figure 5-4 Girder #1 Cross Section in Curved Region (from Van Landuyt. 1991) Page 78 .

Figure 5-5 Strand Positions in Curve (from Van Landuyt. 1991) Figure 5-6 Web Potentiometer Configuration (from Van Landuyt. 1991) Page 79 .

1991) Page 80 .Figure 5-7 Photo of 1.0 DC and BC (from Van Landuyt.

1991) Page 81 .Figure 5-8 Specimen Failure Plans (form Van Landuyt.

1991) Page 82 .Figure 5-9 Comparison of Web Deflections (from Van Landuyt.

Figure 5-10 Finite Element Model and Boundary Conditions Page 83 .

Page 84 .Figure 5-11 Deformed Shape (Displacements x 50) at 1.0 DC Deflection of 0.07 inches.

Page 85 .

Figure 5-13 Contours for Max Principle Strains at BC Deflection of 0. X 25) Page 86 .03 inches (displ x 25) Figure 5-14 Estimated Cracking in the Webs Based on Strains (Displ.

Figure 5-15 Force vs. Deflection (FE model compared with U.T. Test) Page 87 .

by using this model prototype. In configuration “Type 4”.) The tendon duct arrangements and local reinforcements are shown in Figure 5-18. also ran with vertical exterior webs) (prototypical pre-cast box with super-elevation and inclined webs) The prototype geometry for the multi-cell box girder was super-elevated since this is how curved girders often occur in practice. Model Prototype: Three-Cell CIP Box Girder A multi-cell.. cast-in-place box configuration was used as shown in Figure 5-16. Also. Pt/R. especially assumed tensile strength A last variable. the only effect of superelevation was a small difference in the end conditions of the regional moment calculation in the sloping exterior webs. analyses were run with (“4b”) and without the center web ties (“4a”). there was no apparent effect on the behavior of vertical webs caused by super-elevation so it was not included in the single cell section that was studied. etc.b Local Analysis of Multi-cell Box Girders For the NCHRP 12-71 project study. o o o o o o o o Web depth Web thickness Web slope Cover thickness Number and configuration of tendon ducts Number and configuration of duct ties Stirrups Concrete material properties. Webs A and D have different slopes and can be compared to B and C. The basic model (shown in Figure 5-17) uses 3d elements and a slice. This refers to the two rebar ties in the middle of the group of ducts. and another have a different thickness. and inclined exterior webs. In the multi-cell studies conducted. where the stirrup is ‘smeared’). So each analysis includes the full range of Pt/R up to failure.e. The stirrups (and other rebar) are modeled explicitly. separation of the ducts by 1-1/2” was found to provide increased resistance to lateral tendon Page 88 . (unlike in two dimensions. The analysis matrix to study the parameters for the multi-cell series is shown in Table 5-1. geometry variabilities were introduced directly into the models – i. there are two basic local model types: o Multi-cell box o Single-cell box (three cell box with super-elevation / vertical interior webs. Though not studied in depth. the effects of web slope are included and can be compared. one web can have one thickness. This allows introduction of the out-of-page compression due to prestress. with out-of-page thickness equal to one stirrup spacing. which are vertical. The variables that may significantly influence local behavior are listed below. is evaluated by the analyses producing Pt/R versus deformation curves. In configuration Type 3. (Two additional cases were later added with vertical-web exterior webs to provide additional comparisons. And Webs A and D demonstrate the differences related to web sloping away from the radius of curvature versus sloping toward the radius of curvature. Using this model framework.

The rebar was Grade 60.25” 102” 72. for typical designs. which produced reasonable correlation between analysis and test.86” 47. 5 fc' is considered a reasonable average. Multicell Girder Cross-section Geometry Page 89 . This is also the stirrup spacing for the baseline model.86” I-405 . but on average is equal to 1.030.5 ksi. Further separation may provide even higher resistance. but it was the opinion of the project team that enforcing even larger separations between the ducts begins to pose a very significant limitation on the effectiveness of the prestressing.25” 70.5 ft. so the sector width varies slightly from the inside of the curve (Web D) to the outside (Web A). The model is a sector slice taken from a curve. and Young’s Modulus = 4. Designers do need to be allowed some flexibility in duct placement in order to achieve a range of vertical positions of prestressing within the webs.000 psi.breakout. but most results fall within the range 4 fc' to 6 fc' .83” 72. This bridge example is assumed to be on an 800-foot curve radius.55 HOV Connector O/C General Cross-Section (inches) Figure 5-16.75” 275.5” & Var 70. The boundary conditions for the models were the same as for the UT Test simulation.83” 102” 47. The dimension longitudinally varies between inside and outside edge. The concrete properties were fc’ = 5. Plots of material stress-strain curves for the concrete and steel are shown in Figure 519. Tensile strengths (ft) for concrete when tested in direct uniaxial tension can show large variations.75” 10% & Var 9” 13” & Var 12” & Var 8. 275.

Figure 5-17. All Duct Diameters are 4-1/2” @ all sections 12” T yp Inside of Curve. T @ all yp sec tions # 4 @ 12”. Tendon Duct and Local Reinforcement for Multi-Cell Box Local Analysis Prototypes (Note that this figure is an idealization of bar placements that are implemented in the FE analysis) Page 90 . Alternative S ides for 135o hooks 3” C lr To Du c t Typ 3” C lr To Du c t Typ 2” c lr 2” c lr Duct T yp 3” Clr To Du c t 3” C lr To Du c t Case B Only # 4 12” Stirrups T yp T ype 1 T ype 2 T ype 3 T ype 4 Figure 5-18. Example Finite Element Mesh for Multi-Cell Local Analysis Prototype.

Figure 5-19. Stress vs Strain Curves Concrete and Steel in Local FE Analysis Page 91 .

D-12 10 14 Duct/Tie Config. (x fc`) 4 4 4 4 4 4 4 4 4 4 4 4 4 4 6 4 4 2 4 4 4 4 4 4 4 4 4 4 4 4 4 4 4 2 4 6 4 4 4 4 4 4 4 4 4 4 4 4 Analysis # 1M "baseline" 2M Model Type Multi-cell Web # A. D B C A. D B C A. Pos. D B C A. midheight midheight midheight midheight midheight midheight midheight midheight midheight 1/4 height 1/4 height bottom midheight midheight midheight midheight midheight midheight midheight midheight midheight midheight midheight midheight 1/4 height 1/4 height bottom midheight midheight midheight midheight midheight midheight Midheight midheight midheight midheight midheight midheight 1/4 height 1/4 height bottom midheight midheight midheight midheight midheight midheight Concr. D B C A. D-12 10 14 12 12 12 12 12 12 12 12 12 12 12 12 12 12 12 12 12 12 12 12 12 A-10. D B C A. D B C A. D B C A. D B C A.Table 5-1 – Parameters for Multi-cell Girder Local Response Analysis Bundle Vert. 1 1 1 2a 2a 2a 2b 2a 2a 2a 2a 2a 2b 2b 2a 2a 2a 2a 2a 2a 2a 2a 2b 2a 3a 2a 2a 3a 3a 3a 4a 4a 4b 4a 4a 4b 4a 4a 4a 4b 4b 4b 2a 2a 2a 4a 4a 4b Stirrup Spacing (in. D B C A. D B C A. D B C A. D B C A. D B C A. D B C A. D B C Web Thickness 12 12 12 12 12 12 A-10. D B C A.) 18 18 18 18 18 18 18 18 18 18 18 18 18 18 18 12 12 18 27 27 18 18 18 18 18 18 18 27 27 18 18 18 18 18 18 18 27 27 18 18 18 18 18 18 18 18 18 18 Cover Thickness 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 3 2 2 2 4 2 2 2 3 3 2 3 3 3 3 3 3 3 3 2 3 3 3 3 3 3 3 3 3 Multi-cell 3M Multi-cell 4M Multi-cell 5M Multi-cell 6M Multi-cell 7M Multi-cell 8M Multi-cell 9M Multi-cell 10M Multi-cell 11M Multi-cell 12M Multi-cell 13M Multi-cell 14M Multi-cell 2m-Vert Multi-cell 11M-Vert Multi-cell Page 92 . Str. Tens. D-12 10 14 12 12 12 12 12 12 12 12 12 12 12 12 A-10.

125%. 150% Pc Pc refers to a lateral force applied to the web that will cause theoretical web failure calculated using conventional means. 125%. The displacements were measured at the “outside curve” edge of the webs.e. the local analyses required a decision as to loading of the individual webs. 100%. Although the four individual webs in each multi-cell model tend to act independently. These allow for qualitative and quantitative assessment and comparison of the cases analyzed. 100%. This creates a framework for comparing the results of the detailed FE analysis to a baseline capacity. 125%. An initial study using the baseline geometry (Model 1M) and all-elastic material properties showed that web mid-height deflections varied as shown when equal loads (1. In planning this loading it was found that the interaction of the webs with their end conditions (i.Multi-Cell Models – Analysis Results The 16 different multi-cell girder local models were developed.000 lbs per web) were applied to the webs: Figure 5-20. and removing various safety factors.      Lateral Force vs. Deformed shape of Typical Cross-section with the Same Force Applied to Each Web Page 93 . 100%. The results are shown in this chapter through the following plots and tables. Deflection of Web Quarter-height Maximum Principal Strain Contours in Concrete at 75%.. 150% Pc Distortions (change in web width) 3 locations along duct bank 75%. 150% Pc Strains in stirrup rebar at 3 locations along duct bank at 75%. Deflection of Web Mid-height Lateral Force vs. the stiffness characteristics of the top and bottom slabs) was important to how the webs behaved. and analyses were completed.

as shown below. Deformed Shape/Strain Contour When the Same Force is Applied to Each Web These figures show how the exterior web ends are freer to rotate than the interior webs.Figure 5-21. Tendon Forces Applied to a Beam Model of Typical Cross-Section Page 94 . the ratios of mid-height moment to applied tendon force (P) would be h/8 versus h/4 or a ratio of one-half. For the two extremes of fixed-fixed versus pinned-pinned. Figure 5-22. But the web end conditions are neither fixed-fixed nor pinned-pinned. One way to quantify these differences associated with end effects was to apply the tendon forces to a beam model.

145h 0. but removing safety factors.  Mn = 8. Based on the work performed for this project.) Normalizing to the pinned-pinned condition (M=P x h/4) gives ratios of: 0. factors of 0. (7. the webs failed prior to reaching this total load.744 0. Further study of this using fully nonlinear properties showed that once concrete cracking begins to occur.8 for design. as previously mentioned.75 in.588 0.171h h = 92. Pc was calculated as follows.7 k-ft/ft Page 95 .580 0. these can be compared to the Caltrans Memo-to-Designers Formula: Mu = 0. Pc is defined as a “Capacity” calculated using conventional means. Web A 13 k/ft B 17 k/ft C 17 k/ft D 15 k/ft In order to establish a baseline for comparison with design calculations. Caltrans uses a continuity factor of 0. then increase this force for the interior webs.This exercise produced the following ratios of web mid-height moments to applied force. and choosing a prestress force large enough to cause significant damage in all of the parametric models. for this case. Using this as a basis. Of course in some cases. so one finding from the Local Analysis study may be that designers should account for this effect in more detail than to just apply a single formula to calculate mid-height moment demand. led to the following total applied forces in kips/ft.73 ft.6 for interior webs and 0.186h B C D 0. This is analogous to Pj/R. So it was eventually decided to choose a baseline prestress force divided by the four webs. the differences between webs become even larger. For the interior (B or C webs) of the multi-cell geometry prototype.147h 0. The overall distribution of these moments to webs agrees well with damage trends observed in the FE analysis. so to make direct comparison to finite element analyses. Web A 0.684 Considered as coefficients. and reduce this force for the exterior webs.8 (1/4) (Pj/R) hc Where Pj is the tendon force (j is for “jacking force”) Thus.7 for exterior webs are proposed.

6 x 10-4 first visible cracking (micro-cracking occurs at about half of this strain) 2.6 k/ft This is used as “100%” in the results tables and plots. Page 96 .Removing the resistance factor   Mn = 9. typically wide open cracks / sometimes spalling concrete In Table 5-4. The maximum principal strain contours illustrate the general level of damage to the concrete surrounding the tendon ducts (maximum tensile strain regardless of orientation). the “100%” force is less. and “Duct 5”. because the load is applied in the proportions previously listed. The following strain thresholds are important quantities to compare to when viewing these results. It is also recognized that the duct forces are not applied at one point in mid-height.9 / (h/4)] / 0. but are instead. top edge. The results of all the multi-cell analyses are shown in the following tables and figures. “Duct 3”. This decreases the moment to 88.0 x 10-3 first rebar yield 1.). But note that for the exterior webs. and centerline of the duct assembly.6.884 = 10.0 x 10-2 1% strain in rebar.6 / 0. Mo = Mn x 1. The result of all this is to proportion loads so that webs will be “failing” more-or-less simultaneously. This refers to a measure across the bottom edge. E-b-1. So the baseline Pc becomes Pc = [10. delaminations (“distortion”) information is provided at “Duct 1”.9 k-ft/ft The moment-fixity effect is rounded off at 0.125 = 10. therefore maintaining proper flexural stiffnesses in webs and flanges at all loading stages. E-b-2.7 k-ft/ft Applying an over-strength factor for rebar strain hardening (which is included in the FE analysis). 11 = 1. etc. The figures are shown in Appendix E-b and are numbered as (Fig. applied at five points distributed along 18” of the height.4% of that caused by a single point load at midheight.

2091 0.0271 0.3367 2.4100 0.0225 0.6890 0.1074 0.0266 0.0863 0.2460 14.6430 1.2248 0.1285 0.1890 0.5600 0.0520 0.0573 0.0385 0.0702 0.1290 0.0172 0.0409 0.2090 0.2148 0.3756 0.0287 0.0938 0.1682 0.2289 0.0360 0.1900 0.0202 0.3300 0.0922 3.1726 0.0797 0.0575 0.1385 0.Deflections Web A Mid Web B Mid Web C Mid Web D Mid Percent Model # 1m Capacity 75% 100% 125% 150% 75% 100% 125% 150% 75% 100% 125% 150% 75% 100% 125% 150% 75% 100% 125% 150% 75% 100% 125% 150% 75% 100% 125% 150% 75% 100% 125% 150% 75% 100% 125% 150% 75% 100% 125% 150% 75% 100% 125% 150% Quarter Quarter Quarter Quarter 0.0179 0.1450 0.0190 0.0799 0.0197 0.2636 0.0518 0.0683 0.3050 0.0253 0.3171 1.0916 0.3691 13.0220 0.1096 0.0891 0.1974 1.1630 0.0257 0.1371 0.0751 0.1674 0.1366 0.0262 0.0278 0.1563 0.0789 0.1270 0.0202 0.0887 0.5310 1.1935 0.0847 0.1676 0.0230 0.2385 0.0230 0.0687 0.1067 0.4180 0.6354 0.7433 0.0242 0.1778 0.0170 0.0765 0.1048 0.0892 0.0587 0.0213 0.0922 0.7286 0.0986 0.3477 0.4981 0.9112 2.1323 0.6550 0.0199 0.0235 0.3067 0.0254 0.3461 0.0554 0.0269 0.0195 0.2458 0.2249 0.0217 0.1395 0.6840 2.4394 1.3711 2.2544 1.1332 0.1654 0.2601 1.0225 0.2427 0.0178 0.0415 0.0309 0.2020 0.1428 0.1392 3.0545 0.0665 0.9500 0.2803 1.0539 0.3379 0.0363 0.2026 0.0870 0.0254 0.0591 0.2714 0.3047 0.4692 12.0673 0.1702 0.3931 7.1610 0.0660 0.1626 0.1500 0.1325 0.1851 0.0253 0.0216 0.1408 0.0304 0.1553 0.0673 0.0506 0.5231 0.1524 0.1305 0.1983 0.0583 0.0313 0.2610 0.1338 0.6680 9.0627 0.3340 0.1566 0.4859 0.1481 0.0388 0.9963 11.6550 0.0621 0.0990 0.8323 2.0390 0.0206 0.1623 0.1954 0.1115 0.0463 0.7589 2.2141 0.5017 0.0427 0.7176 0.0273 0.4067 0.Table 5-2 .2182 0.3478 0.2103 0.2493 0.0486 0.1240 0.1283 0.0214 0.0511 0.4509 0.0290 0.3033 0.0122 0.1442 0.1295 0.0207 0.7502 2.9640 0.0596 0.4455 0.1390 0.1313 0.0337 0.2393 0.1961 0.1191 0.1350 0.0154 0.2670 0.0251 0.0872 0.2520 9.6878 8.0397 0.0352 0.3821 0.2900 0.6240 0.6490 0.0139 0.0231 0.3083 0.1270 0.0671 0.0610 0.5575 1.8200 0.3130 2m 3m 4m 5m 6m 7m 8m 9m 10m 11m Page 97 .0555 0.0298 0.2731 0.1300 0.0554 0.0962 0.0615 0.1639 0.1024 0.1629 0.3685 0.1697 11.1764 0.0293 0.0491 0.2425 0.0210 0.5099 16.0253 0.1939 0.0686 0.2240 0.0840 0.1841 0.1103 0.1966 0.1825 0.2071 0.0279 0.2603 0.0606 0.2684 0.0183 0.0195 0.3252 0.3520 0.2885 1.0142 0.0225 0.1051 0.0548 0.0386 0.2390 0.6300 0.8020 0.0830 0.1571 0.1981 0.3600 0.3567 0.2538 3.3725 15.1019 0.1098 0.8475 0.7823 11.0480 0.0411 0.9860 0.7760 0.5756 1.0558 0.2972 1.1534 0.0680 0.1160 0.1983 0.3248 11.3781 12.2980 0.3897 0.3123 0.3182 0.0672 0.1755 0.2707 0.0420 0.0924 0.1702 0.8136 2.0309 0.2218 3.7370 0.1082 0.2280 0.0340 0.0566 0.0195 0.1114 0.4180 0.0259 0.2030 0.0156 0.7819 2.1560 0.5420 0.0453 0.7380 11.0262 0.2815 0.3382 0.2109 0.7190 0.0403 0.0690 0.0178 0.4080 0.3314 1.0792 0.7340 0.7348 15.0237 0.5490 0.1867 0.0481 0.3214 1.0625 0.0390 0.

0256 0.0327 0.0180 0.0154 0.0172 0.1118 0.1239 0.0343 0.0288 0.0153 0.1664 0.0474 0.1122 0.1033 0.2772 0.1401 0.2076 0.0348 0.0750 0.0761 0.0181 0.3889 1.0935 0.0320 0.0319 0.1159 0.0285 0.1258 0.0338 0.0977 0.0171 0.1189 0.0245 0.0407 0.1337 13m 14m 2MVert 11MVert Page 98 .0400 0.0748 0.0175 0.8000 2.0161 0.0806 0.0541 0.1365 0.1201 0.0328 0.0479 0.1207 0.0537 0.0153 0.1190 0.0860 0.1026 0.0161 0.0731 0.1831 0.0319 0.0392 0.0493 0.0955 2.0903 0.1949 0.0533 0.4193 1.1346 0.0206 0.1228 0.1937 0.0842 0.5100 0.0154 0.4126 1.1454 0.0146 0.0796 0.0170 0.3656 0.0160 0.0780 0.1089 0.0355 0.0343 0.0182 0.1164 0.1233 0.1688 0.1824 0.2706 0.8586 0.2221 0.0324 0.0255 0.0168 0.0986 0.0169 0.1475 0.7158 1.9897 2.1016 0.9289 0.3779 0.0627 0.0686 0.0489 0.0896 0.7251 0.0775 2.0864 0.0913 0.1059 0.0162 0.0117 0.0254 0.0236 0.1939 0.8840 0.0697 0.1515 0.0991 0.0614 0.1051 0.12m 75% 100% 125% 150% 75% 100% 125% 150% 75% 100% 125% 150% 75% 100% 125% 150% 75% 100% 125% 150% 0.0162 0.1481 0.1917 0.0412 0.0414 0.1332 0.0113 0.0516 0.0665 0.1642 0.2089 0.0594 0.2657 0.0314 0.7535 0.1449 0.6583 0.3100 0.2198 0.7350 1.0212 0.0630 0.0841 0.0372 0.0318 0.0705 0.0398 0.0202 0.0677 0.0610 0.0191 0.9539 2.0748 0.0302 0.0295 0.0355 0.0325 0.0229 2.0190 0.0162 0.

0042 0.0242 0.0071 0.0007 0.0025 0.0172 0.0010 0.0154 0.0158 0.0006 0.0004 0.0010 0.0101 0.0016 0.0789 0.0021 0.0004 0.0020 0.0246 0.0025 0.0002 0.0014 0.0002 0.0017 0.0004 0.0008 0.0012 0.0056 0.0013 0.0235 0.0001 0.0019 0.0008 0.0013 0.0002 0.0010 0.0015 0.0008 0.0006 0.0018 0.0002 0.0008 0.0019 0.0038 0.0178 0.0013 0.0007 0.0028 0.0019 0.0004 0.1241 0.0012 0.0029 0.0002 0.0093 0.0003 0.0664 0.0003 0.0002 0. Duct 5 – top) Web A Percent Model # 1m Capacity 75% 100% 125% 150% 2m 75% 100% 125% 150% 3m 75% 100% 125% 150% 4m 75% 100% 125% 150% 5m 75% 100% 125% 150% 6m 75% 100% 125% 150% 7m 75% 100% 125% 150% 8m 75% 100% 125% 150% 9m 75% 100% 125% 150% 10m 75% 100% 125% 150% 11m 75% 100% 125% 150% 0.0085 0.0004 0.0026 0.0026 0.0003 0.0080 0.0005 0.0004 0.0062 0.0001 0.0001 0.0559 0.0016 0.0003 0.0002 0.0923 0.0001 0.0007 0.0787 0.0002 0.0021 0.0011 0.0002 0.0096 0.0004 0.0197 0.0103 0.0008 0.0026 0.0006 0.0005 0.0004 0.0001 0.0003 0.0002 0.0016 0.0003 0.0002 0.0004 0.0002 0.0005 0.0003 0.0029 0.0019 0.0007 0.0002 0.0002 0.0002 0.0001 0.0001 0.0029 0.0024 0.0005 0.0001 0.0035 0.0019 0.0002 0.0018 0.0012 0.0015 0.0006 0.0009 0.0007 0.0747 0.0001 0.0009 0.0015 0.0019 0.0002 0.0014 0.0003 0.0009 0.0016 0.0008 0.0012 0.0017 0.0008 0.0002 0.0002 0.0004 0.0010 0.0007 0.0002 0.0005 0.0009 0.0011 0.0002 0.0025 0.0010 0.0004 0.0003 0.0002 0.0054 0.0016 0.0108 0.0024 0.0007 0.0016 0.0110 0.0002 0.0002 0.0002 0.0005 0.0002 0.0023 0.0017 0.0018 0.0007 0.0209 0.0091 0.0019 0.0003 0.0004 0.0003 0.0012 0.0009 0.0001 0.0027 0.0008 0.0012 0.0171 0.0005 0.0027 0.0207 0.0005 0.0010 0.0001 0.0010 0.0004 0.0026 0.0244 0.0090 0.0001 0.0009 0.0033 0.0081 0.0048 0.0003 0.0009 0.0003 0.0040 0.0019 0.0017 0.0117 0.0008 0.0001 0.0006 0.0026 0.0005 0.0001 0.0013 0.0002 0.0688 0.0001 0.0007 0.0003 0.0007 0.0002 0.0009 0.0713 0.0013 0.0004 0.0002 0.0008 0.0016 0.0002 0.0005 0.0001 0.0010 0.0017 0.0002 0.0003 0.0001 0.0033 0.0002 0.0003 0.0014 0.0042 0.0099 0.0003 0.0008 0.0002 0.0048 0.0015 0.0001 0.0008 0.0018 0.0003 0.0004 0.0002 0.0016 0.0006 0.0001 0.0008 0.0002 0.0118 0.0104 0.0153 0.0083 0.0023 0.0007 0.0107 0.0010 0.0003 0.0005 0.0007 0.0003 0.0019 0.0068 0.0019 0.0009 0.0086 Duct 1 Duct 3 Duct 5 Duct 1 Web B Duct 3 Duct 5 Duct 1 Web C Duct 3 Duct 5 Duct 1 Web D Duct 3 Duct 5 Page 99 .0006 0.0003 0.0015 0.0013 0.0020 0.0013 0.0013 0.0005 0.0005 0.0019 0.0002 0.0001 0.0065 0.0170 0.0046 0.0016 0.0003 0.0013 0.0027 0.0012 0.0008 0.0002 0.0029 0.0742 0.0005 0.0001 0.0002 0.0004 0.0012 0.0008 0.0018 0.0026 0.0015 0.0008 0.0186 0.0026 0.0005 0.0003 0.0016 0.0007 0.0002 0.0045 0.0024 0.0994 0.0003 0.0003 0.0002 0.0002 0.0008 0.0013 0.0002 0.0019 0.0014 0.0007 0.0041 0.0042 0.0013 0.0015 0.0065 0.0006 0.0057 0.0225 0.0013 0.0008 0.0002 0.0013 0.0006 0.0004 0.0032 0.0002 0.0010 0.0133 0.0010 0.0002 0.0002 0.0003 0.0027 0.0007 0.0017 0.0012 0.0022 0.0019 0.0006 0.0009 0.0018 0.0027 0.0013 0.0002 0.0001 0.0018 0.0876 0.0009 0.0098 0.0012 0.0002 0.0018 0.0002 0.0003 0.1066 0.0001 0.0014 0.0042 0.0013 0.0218 0.0001 0.0003 0.0083 0.0036 0.0025 0.0003 0.0001 0.0001 0.0004 0.0003 0.0592 0.0003 0.0015 0.0006 0.0031 0.0012 0.0002 0.0003 0.0001 0.0006 0.0014 0.0017 0.0001 0.0013 0.0019 0.0756 0.0002 0.0796 0.0022 0.0011 0.0006 0.0004 0.0753 0.0002 0.0006 0.0011 0.0005 0.0004 0.0203 0.0026 0.0002 0.Table 5-3 – Stirrup Strain (%) Adjacent to Mid-height of Ducts (Duct 1 – bottom.0978 0.0009 0.0014 0.0002 0.0001 0.0001 0.0022 0.0023 0.0006 0.0008 0.0003 0.0024 0.0011 0.0002 0.0793 0.0003 0.0013 0.0002 0.0020 0.0008 0.0514 0.0070 0.0105 0.0005 0.0007 0.0164 0.0003 0.0027 0.0010 0.0045 0.0002 0.0005 0.0147 0.0012 0.0002 0.0008 0.0017 0.0021 0.0006 0.0007 0.0008 0.0019 0.0109 0.0004 0.0011 0.0012 0.0012 0.0010 0.0017 0.0010 0.0012 0.0003 0.0015 0.0039 0.0003 0.0040 0.0001 0.0005 0.0006 0.0001 0.0002 0.0003 0.0653 0.0036 0.0022 0.0002 0.0003 0.0012 0.0002 0.0004 0.0006 0.0001 0.0010 0.0023 0.0153 0.0721 0.0003 0.0870 0.0001 0.0001 0.0021 0.0002 0.0001 0.0002 0.0011 0.0005 0.0007 0.0013 0.0070 0.0001 0.0037 0.0708 0.0012 0.0004 0.0003 0.0003 0. Duct 3 – middle.0002 0.0001 0.0008 0.0007 0.0009 0.0003 0.0015 0.0005 0.0007 0.0100 0.0046 0.0020 0.0013 0.0002 0.0002 0.0003 0.

0021 0.0015 0.0013 0.0003 0.0001 0.0017 0.0011 0.0006 0.0004 0.0009 0.0002 0.0001 0.0001 0.0023 0.0002 0.0021 0.0007 0.0017 0.0002 0.0005 0.0001 0.0001 0.0002 0.0003 0.0034 0.0002 0.0006 0.0002 0.0017 0.0007 0.0018 0.0007 0.0001 0.0001 0.0002 0.0006 0.0012 13m 75% 100% 125% 150% 14m 75% 100% 125% 150% 2MVert 75% 100% 125% 150% 11MVert 75% 100% 125% 150% Page 100 .0004 0.0004 0.0015 0.0002 0.0002 0.0001 0.0014 0.0005 0.0002 0.0009 0.0001 0.0113 0.0016 0.0010 0.0010 0.0008 0.0011 0.0014 0.0017 0.0021 0.0001 0.0017 0.0002 0.0001 0.0001 0.0003 0.0003 0.0013 0.0002 0.0010 0.0008 0.0020 0.0009 0.0007 0.0005 0.0007 0.0002 0.0002 0.0006 0.0001 0.0006 0.0008 0.0001 0.0001 0.0002 0.0003 0.0002 0.0019 0.0004 0.0001 0.0007 0.0001 0.0008 0.0032 0.0001 0.0001 0.0009 0.0004 0.0048 0.0002 0.0004 0.0010 0.0063 0.0009 0.0001 0.0002 0.0125 0.0005 0.0006 0.0008 0.0002 0.0019 0.0004 0.0003 0.0011 0.0004 0.0001 0.0001 0.0011 0.0004 0.0001 0.0001 0.0001 0.0001 0.0001 0.0002 0.0006 0.0001 0.0011 0.0001 0.0005 0.0001 0.0008 0.0001 0.0010 0.0013 0.0024 0.0111 0.0011 0.0015 0.0016 0.0011 0.0002 0.0001 0.0003 0.0001 0.0001 0.0001 0.0273 0.0006 0.0003 0.0020 0.0003 0.0009 0.0008 0.0001 0.0008 0.0002 0.0004 0.0045 0.0005 0.0007 0.0002 0.0001 0.0006 0.0002 0.0013 0.0182 0.0143 0.0002 0.0002 0.0001 0.0032 0.0034 0.0022 0.0002 0.0003 0.0011 0.0001 0.0001 0.0002 0.0001 0.0001 0.12m 75% 100% 125% 150% 0.0005 0.0006 0.0007 0.0005 0.0005 0.0003 0.0003 0.0011 0.0002 0.0009 0.0005 0.0081 0.0019 0.0007 0.0006 0.0002 0.0006 0.0001 0.0003 0.0005 0.0012 0.0002 0.0001 0.0001 0.0005 0.0002 0.0006 0.0002 0.0011 0.0013 0.0002 0.0004 0.0009 0.0005 0.0010 0.0014 0.0001 0.0001 0.0026 0.0001 0.0002 0.0003 0.0003 0.0002 0.0004 0.0036 0.0003 0.0002 0.0013 0.0002 0.0011 0.0046 0.0005 0.0005 0.0002 0.0003 0.0008 0.0019 0.0001 0.0004 0.

0146 0.0029 0.0016 0.0020 0.2910 0.0146 0.0087 0.0303 0.1211 1.0752 1.0059 0.0029 0.0254 0.0020 0.0068 0.0158 0.0078 0.0087 0.1211 0.0002 0.0011 0. Duct 5 – top) Web A Percent Model # 1m Capacity 75% 100% 125% 150% 0.0089 0.0024 0.1162 1.0025 0.0068 0.0078 0.0947 1.0013 0.0156 0.0088 0.0020 0.0039 0.0073 0.0321 0.2535 0.0409 0.0137 0.1244 1.0029 0.0014 0.0146 0.6562 0.0034 0.0082 0.0059 0.0021 0.0042 0.0186 0.0020 0.0042 0.0437 0.0047 0.0133 0.0016 0.3154 0.0059 0.0459 0.7197 0.0107 0.1010 0.0039 0.0048 0.0093 0.0029 0.0020 0.0352 0.0197 0.0084 0.0052 0.4268 0.0020 0.0026 0.0078 0.0067 0.0029 0.0117 0.0059 0.0251 0.0049 0.0010 0.0049 0.0059 0.0133 0.0010 0.0479 0.0093 0.0195 0.0056 0.0032 0.0074 0.0109 0.0088 0.0031 0.0447 0.0075 0.0174 0.0215 0.0055 0.0039 0.0020 0.0116 0.0020 0.0225 0.0537 0.0010 0.0078 0.0591 0.0059 0.0014 0.7051 0.1803 0.0029 0.0176 0.0498 0.0049 0.0078 0.0332 0.0234 0.0031 0.0107 0.0443 0.9038 0.0001 0.0021 0.0025 0.9990 0.0068 0.0176 0.0020 0.0098 0.1748 0.0693 0.0438 0.0040 0.0052 0.0029 0.5244 0.0098 0.0283 0.0015 -0.0209 0.0049 0.0021 0.0019 0.0026 0.0130 0.0146 0.0078 0.0020 0.0059 0.0013 0.0041 0.0040 -0.0089 0.0033 0.0006 0.0732 0.0010 0.8164 0.0078 0.0029 0.0244 0.0029 0.0104 0.0801 0.0065 0.0068 0.0195 0.0508 0.0184 0.0013 0.0028 0.0342 0.0059 0.0010 0.0098 0.0234 0.0064 0.0039 0.2105 0.0169 0.0020 0.0049 0.0029 0.7270 0.0081 0.0080 0.0527 0.0625 0.0039 0.0253 0.0312 0.7518 0.0368 0.0068 0.0391 0.0110 0.0029 0.0037 0.0014 0.0112 0.0101 0.Table 5-4 – Distortion (Web Thickness Change) Across the Mid-height of Ducts (Duct 1 – bottom.8428 0.2266 1.0283 0.0044 0.0075 0.0029 0.0034 0.0046 0.0149 0.0034 0.0111 1.0033 0.0088 0.0029 0.0117 0.0020 0.0361 0.0380 0.0221 0.0019 0.0596 0.0014 0.0943 0.0020 0.0039 0.0133 0.0478 0.0016 0.0012 -0.0010 0.9973 0.0078 0.0032 0.0092 0.0031 -0.0022 0.0027 0.0137 0.0020 0.0020 0.0014 0.0020 0.0012 0.1387 0.0049 0.0014 0.0049 0.0020 0.0244 0.0059 0.2803 0.0035 0.0282 0.0049 0.0251 0.0016 0.0029 0.0113 0.0156 0.0186 0.0073 0.1553 0.0293 0.0051 0.0039 0.0078 0.0078 0.0117 0.0021 0.0107 0.9574 0.0022 0.1875 0.0010 0.0105 0.0026 0.7256 0.0174 0.0068 0.0029 0.1843 0.2327 0.0016 0.0303 0.0035 0.1055 0.0140 0.0040 0.0075 -0.1660 0.0029 0.0029 0.0140 0.0130 0.0049 0.0264 0.0107 0.0742 0.0117 0.1494 0.0296 0.0801 0.0089 0.0342 0.1621 0.0023 0.0035 0.0039 0.1113 0.1099 0.0039 0.0036 0.0029 0.0017 0.0098 0.0146 0.0249 0.0117 0.0244 0.8598 Duct1 Duct 3 Duct 5 Duct1 Web B Duct 3 Duct 5 Duct1 Web C Duct 3 Duct 5 Duct1 Web D Duct 3 Duct 5 2m 75% 100% 125% 150% 3m 75% 100% 125% 150% 4m 75% 100% 125% 150% 5m 75% 100% 125% 150% 6m 75% 100% 125% 150% 7m 75% 100% 125% 150% 8m 75% 100% 125% 150% 9m 75% 100% 125% 150% Page 101 .0052 0.0029 0.0170 0.0110 0.0156 0.0046 0.0098 0.0010 0.0186 0.0035 0.0107 0.0059 0.0041 0.0010 0.0056 0.0029 0.0029 0.0029 0.0369 0.0012 0.0007 0.1426 0.0013 0.0032 0.0011 0.0059 0.0117 1.0088 0.1143 1.0244 0.0064 0.0019 0.0037 0.0488 0.6789 0.0352 0.0062 0.3563 0.0547 0.0014 0.0321 0.0918 1.0107 0.0358 0.0264 0.0044 0.0055 0.0167 0.0013 0.0032 0.0039 0.0029 0.0764 0.0057 0.0193 0.0077 0.0087 0.0242 0.0020 0.1504 0.0068 0.0150 0.9424 0.0261 0.0083 0.0014 0.0361 0.0137 0.0127 0.0303 0.0098 0.0029 0.0065 0.0049 0.1789 0.0292 0.0039 0.0098 0.0049 0.0146 0.8594 0.0019 0.0117 0.0117 0.0020 0.0083 0.0011 0.0391 0.0020 0. Duct 3 – middle.0053 0.0049 0.0293 0.0371 0.0046 0.0049 0.0020 0.0045 0.0024 0.0014 0.0010 0.0010 0.0039 0.0413 0.0103 0.0020 0.0791 0.0020 0.0381 0.0018 0.

0014 0.0020 0.0021 0.0019 0.0000 0.0078 0.0488 0.0000 0.0059 0.0086 0.0039 0.0020 0.0011 0.0049 0.0127 0.0074 0.0279 0.0021 0.0020 0.0078 0.0097 0.0050 0.0032 0.0093 0.0064 0.0020 0.0039 0.0039 0.0049 0.0010 0.0062 0.0017 0.0041 0.0010 0.0000 0.0016 0.0068 0.0010 0.0041 0.0020 0.0029 0.0078 0.0107 0.0017 0.0043 0.0049 0.0008 0.0020 0.0038 0.0030 0.0781 0.0010 0.0059 0.0015 0.1211 0.0069 0.0978 0.0072 0.0021 0.0022 0.0038 0.0014 0.0010 0.0093 0.0028 0.1387 0.0071 0.0026 0.0104 0.0068 11m 75% 100% 125% 150% 12m 75% 100% 125% 150% 13m 75% 100% 125% 150% 14m 75% 100% 125% 150% 2MVert 75% 100% 125% 150% 11MVert 75% 100% 125% 150% Page 102 .0117 0.0064 0.0010 -0.0024 0.0078 0.0029 0.0026 0.0117 0.0140 0.0939 0.0049 0.0020 0.0020 0.0020 0.0029 0.0031 0.0010 0.0078 0.0010 0.0088 0.0029 -0.0020 0.0117 0.0033 -0.0014 0.0163 0.1518 0.0020 0.0010 0.0039 0.0020 0.0045 0.0020 0.0014 0.0029 0.0020 0.0020 0.0020 0.0027 0.0095 0.0132 0.0024 0.0000 0.0020 0.0068 0.0068 0.0146 0.0025 0.0018 0.0049 0.0146 0.0168 0.0371 0.0010 0.0000 9.0019 0.0059 0.0560 0.0019 0.0022 0.0008 0.0065 0.0027 0.0020 0.0050 0.0022 0.0015 0.0039 0.0176 0.0010 -0.0090 0.0057 0.0018 0.0029 0.0013 0.0020 0.0041 0.0021 0.0117 0.0020 0.1963 0.0397 0.0059 0.0059 0.0963 0.0039 0.0016 0.0017 0.0046 0.0047 0.0127 0.0033 0.0010 0.0014 0.0020 0.0010 0.0061 0.0088 0.0039 0.0049 0.0010 0.0059 0.0127 0.0065 0.0225 0.0010 0.0018 0.0020 0.0000 0.0015 0.0028 0.0010 0.0264 0.0008 0.0041 0.0020 0.0020 0.0020 -0.0039 0.0038 0.0013 0.0047 0.0060 -0.0012 0.0127 0.0013 8.0010 0.0024 0.0014 0.0059 0.0088 0.0082 0.0010 0.0010 0.0036 0.0013 0.0020 0.0029 0.0302 0.0053 0.0016 0.0020 0.0017 0.0013 0.0029 0.0020 0.0038 0.0010 -0.0364 0.0017 0.0004 0.0029 0.0117 0.0557 0.0039 0.0809 0.0068 0.0010 0.0026 0.0029 0.0035 0.0000 0.0021 0.0019 0.0074 0.0016 0.10m 75% 100% 125% 150% 0.0041 0.0010 0.0021 -0.0268 0.0117 0.0039 0.1160 0.0008 -0.0016 0.0049 0.0020 0.0020 0.0049 0.0000 0.0088 0.0020 0.0010 0.0228 0.0018 0.0010 0.0030 0.0059 0.0109 0.0016 0.0146 0.0060 0.0098 0.0000 0.0015 0.0175 0.0010 0.0078 0.0137 0.0029 0.0977 0.0059 0.0049 0.0030 0.0071 0.0008 0.0015 0.0010 0.0127 0.0042 0.0078 0.0041 0.0020 0.1328 0.0193 0.0041 0.0029 0.0019 0.0049 0.0088 0.0041 0.0035 0.0021 0.0017 0.0000 0.0020 0.0039 0.0093 0.0021 0.0029 0.0071 0.0008 0.0148 0.0078 0.0068 0.0000 0.0029 0.0046 0.0010 0.0018 0.0000 0.0204 0.0029 0.0332 0.0091 0.0023 0.0058 0.0107 0.0039 0.0030 0.0029 0.0233 0.0000 0.0078 0.0022 0.0010 0.0107 0.0000 0.0086 -0.0109 0.0030 0.0039 0.0020 0.0195 0.0020 0.0020 0.

3% can be considered to be heavily cracked. have been summarized in Table 5-5 and 2-6. the lateral shear is divided equally between the top and bottom of the web. and a distortion ratio (average strain through the section) of 0. Visible concrete cracking occurs at strains of approximately 0. for sections without web ties. Significant distortion or delamination (change of width of the webs) would also represent an upper bound on serviceable capacity for webs. note that for Load Factor Design.   Two of the criteria. Stirrup Yield and Web Delamination. When the ducts are located at the midheight. concrete reinforcement is designed to yield at the ultimate member forces.. It was arbitrarily assumed that a crack width of 1/16” is an indicator of such a failure.016%. on average as much as 25% to 40% lower when comparing these cases to other similar cases. the section is at a web splitting or a web lateral shear failure condition. So the “quarterheight” cases can be compared to each other (and the “bottom” cases). and when the web distortion reaches 0.2% strain for Grade 60 steel). and this is a different mechanism than the failure modes observed for tendon ducts at mid-height. this means the web ties have yielded. i. The reason for this is a tendency toward lateral shear failure of the overall web.06 inches.e.Discussion of Results The analysis results have been used to compare the web design parameters. concrete with maximum principal strains of 0. Page 103 .0% will generally show wide-open cracks and potential spalling from the section. the following damage criteria should be considered:  Stirrup rebar strain exceeding yield (i. It was generally observed that when the ducts occur near the bottom of the web (either “quarterheight” or “bottom” as was tested in Configurations 4M and 14M) the force at “failure” is substantially lower than when the ducts are placed at the mid-height. These are the total forces (sum of all tendon ducts in the web) applied when any part of the stirrup reaches yield. For purposes of interpreting and comparing the results. but should not be compared directly to the “mid-height” cases. the bottom of the web carries most of the lateral shear.. For sections with web ties. The first general observation is the comparison of the mid-height cases to the “1/4 height” and “bottom” cases. the delamination is evidence of a local splitting or lateral shear failure within the web. this represents a distortion of 0. but this is not necessarily web failure. Concrete with strains in excess of 1.06 inch.5%. For 12” webs. 0.e. “Pc” only applies to the “mid-height” cases. But when the ducts move down.

305 0.82 12.059 13.53 Web D 122.77 13.96% 200% * 200% * 158.624 21.163 14.613 13.11% 10.153 17.407 1.18 11.365 0.00% 145.250 8.98% 11.93% 129.34 180.36% 128.248 6.602 7.047 % Pc Deflection (in) 0.23% 13.143 10.042 21.613 10.218 16.38% 136.77 13.97% 200% * 140.12% 120.405 9.830 3.458 16.49% 162.67% 144.55 19.05% 143.259 7.90% 142.34 8.17% 200% * 137.06” for 12” Web) Model # % Pc Web A Total Force (K/ft) 13.95% 200% * 200% * 153.513 0.677 18.45% 10.921 18.883 2.5% (0.77 121.75% 153.278 0.15% 165.41% 129.92% 131.24 134.16 Web C 111.200 20.200 Web D Total Force (K/ft) 155.55 140.40% 200% * * Never reached delamination limit.753 14.55% 13.31 15.230 6.183 5.533 0.00 17.337 18.831 15.66 12.139 0.332 0.757 1M 2M 3M 4M 5M 6M 7M 8M 9M 10M 11M 12M 13M 14M 2MVert 11MVert 165.84% 184.59 8.814 21.65% 161.787 0.06 131.57 15.47% 105.440 0.058 16.54% 8.64 15.04% 128.70 17.60% 148.613 16.568 1.59 14.79% 179.33% 173.464 16.111 15.032 1.54% 154.11 12.24% 126.81% 184.87% 145.41 9. so the delamination at 200% of Pc is shown Page 104 .68 .082 11.55% 9.45% 124.041 16.613 16.914 0.372 15.200 14.53% 166.38% 161.55% 200% * 200% * 171.19% Web D 11.429 1.346 0.51% 165.822 16.14% 200% * 200% * 166.62% 200% * 12.65 151.71 124.26 7.13 15.266 Deflection (in) 3.21 16.334 2.170 12.384 0.45% 141.301 0.74 9.74% 15.42 11.38% 134.363 1.266 12.92 160.86% 142.72% 10.Percentages not shown for cases other than ducts placed at midheight Model # 1M 2M 3M 4M 5M 6M 7M 8M 9M 10M 11M 12M 13M 14M 2MVert 11MVert Web B 107.767 18.28 157.69% 167.77% 135.708 10.24% 178.25% 117.235 3.266 13.620 19.96% 11.109 0.604 15.809 10.519 7.42% 142.836 18.85% 118.425 0.697 15.38 13.77 16.59% 11.446 0.85 15.09% 151.156 0.473 0.613 16.59 17.06% 148.670 18.61 10.056 16.200 27.200 21.81 15.06% 123.193 0.95 11.10% 200% * 200% * 200% * 145.601 13.2% Strain) for Stirrups on Inside of Curve Total Force (in % of Pc and in K/ft) Web A Web A 119.33 15.36 7.97 11.21 105.103 0.11% 118.249 0.41% Web C 11.03% 130.288 15.01 8.29% 200% * 162.886 10.77% 200% * 12.87% 200% * 142.613 12.17 8.364 18.08% 159.315 7.762 0.717 9.53 10.Table 5-5 – Lateral Prestress Force at Stirrup Yield (0.277 0.06% 165.341 10.148 0.77 140.93 141.99% 172.575 6.830 11.57 16.160 2.729 21.725 2.58 19.58 17.75 12.181 3.853 0.688 0.266 16.38 17.917 % Pc Deflection (in) 0.200 14.573 0.49% Web B 11.01% 126.613 Web B Total Force (K/ft) 121.50 11.266 18.190 0.321 9.72% 150.05% 12.186 16.200 Web C Total Force (K/ft) 118.27 Table 5-6 – Lateral Prestress Force at Web Delamination of 0.504 10.274 0.28% 200% * 138.66% 11.86 7.51 11.299 21.200 21.58% 200% * 14.300 12.559 % Pc Deflection (in) 0.64% 114.92% 147.58% 184.289 10.79% 152.06 6.563 11.26% 167.830 13.58 12.122 0.48 13.38 15.257 0.97% 148.528 7.810 0.77% 133.47 15.22 15.868 0.939 21.

the sloped webs in this analysis series were found to be significantly weaker (30-40%) than the vertical webs. Stirrups yielded sooner. 12.17 vs. 17. 10. 12.77 vs. 10. For stirrup yield. 3M-A 2M-B vs. Moving the ducts toward the curve outside face within the webs also contributes to resistance against delamination and local lateral shear damage. B – 10”.92 15.81 16. and similarly in Model 11M.20 vs. 10” -31% change between 12” vs. 14” -18% change between 12” vs. 12. Page 105 . and this produces larger mid-height moments. capacity formulae based on regional flexure considerations appear to be appropriate for design.64 vs. 3M-C 12M-A vs. 11M-B Model-Web 2M-B vs.06 13.00 Difference -15% change between 12” vs. 10” Difference -16% change between 12” vs. 3M-B 12M-A* vs. 11M-B Force at stirrup yield (kips/ft) 11.Using these criteria and the results tables and plots has resulted in the following observations: Web Depth This parameter was varied indirectly by subjecting the webs to wide ranges of moments and horizontal shears.61 vs. 10” -27% change between 12” vs.66 vs. Web Thickness Web thickness was varied in Model 3M (A – 10”. Web Slope As described earlier. 10” 12% change between 12” vs. 15. Based on observations of the analysis results. but part of this difference is caused by being exterior webs rather than interior. (kips/ft) 15. Exterior webs have more flexible end conditions at their connection with the top and bottom slab.58 vs. C – 14”). Differences in stirrup yield and especially web delamination were also significantly influenced when the web ties were added because the web ties tended to eliminate the web delamination failure mode. 11M-A 12M-B vs. 21. and concrete damage and web delamination was more extensive with the thinner webs.34 13. Model 11M included web ties.59 13. 11M-A 12M-B* vs. web depth can be adequately accounted for by considering and designing for web moments.06 14.62 * Never reached delamination limit These results demonstrate significant influence on resistance to lateral bending and tendon pullout caused by web thickness. 10” -15% change between 12” vs.34 vs. 10” Force at web delam. 10” -8% change between 12” vs. 3M-B 2M-C vs. The results compared to their respective baselines are shown in Table 5-7: Table 5-7 – Effect of Web Thickness – Thin Webs Model-Web 2M-A vs.

Comparison of Webs A to D for the inclined webs show that Web A is generally weaker than D by about 10%, due to orientation of slope relative to the direction of the tendon force. In order to examine the differences between sloped webs and vertical webs more directly, two additional analyses were performed with exterior webs converted to vertical webs. The strain contour and Force versus Deflection plots are included in the Appendix. Models 2M and 11M were chosen for these comparisons because these have the baseline values for all properties, but they investigate duct-tie configurations 2a and 4a/b. Direct comparisons of Force versus Deflection are shown in Figures 5-23a and 5-23b below.
16

14

12

10

8

6

4

2

0 0.00E+00

5.00E-02

1.00E-01

1.50E-01

2.00E-01 Defl (in)

2.50E-01

3.00E-01

3.50E-01

4.00E-01

Model 2M Web A

Model 2M Web D

Model 2MVert Web A

Model 2MVert Web D

Fig. 5-23a. Model 2M and 2MVert Force vs. Deflection Comparison for Webs A and D (quarter height)

Page 106

20

18

16

14 12

10

8

6 4

2

0 0.00E+00

5.00E-02

1.00E-01

1.50E-01

2.00E-01 Defl (in)

2.50E-01

3.00E-01

3.50E-01

4.00E-01

Model 11M Web A

Model 11M Web D

Model 11MVert Web A

Model 11MVert Web D

Fig. 5-23b. Model 11M and 11MVert Force vs. Deflection Comparison for Webs A and D (mid height) These figures show the vertical webs to be stiffer and stronger than the sloping webs, but show that the differences in force capacity (strength) are negligibly small. A comparison of the occurrence of rebar yield and web delamination is shown in Table 5-8: Table 5-8 – Effect of Web Slope – Thin Webs
Model-Web 2M-A vs. 2Mvert-A 2M-D vs. 2Mvert-D 11M-A vs. 11Mvert-A 11M-D vs. 11Mvert-D Model-Web 2M-A vs. 2Mvert-A 2M-D vs. 2Mvert-D 11M-A vs. 11Mvert-A* 11M-D vs. 11Mvert-D* Force at stirrup yield (kips/ft) Difference 11.77 vs. 12.55 7% change with vertical web 11.82 vs. 13.85 17% change with vertical web 10.92 vs. 11.68 7% change with vertical web 11.58 vs. 15.27 32% change with vertical web Force at web delam. (kips/ft) Difference 10.83 vs. 13.82 28% change with vertical web 14.56 vs. 15.67 8% change with vertical web 12.06 vs. 16.61 38% change with vertical web 12.84 vs. 18.27 42% change with vertical web

* Never reached delamination limit

Page 107

Cover Thickness Cover thickness was varied in models 7M, 8M, and 13M. Table 5-9 summarizes sample strength comparisons: Table 5-9 – Effect of Cover Thickness – Thin Webs
Model-Web 8M-A vs. 2M-A 8M-C vs. 2M-C 8M-D vs. 2M-D Model-Web 8M-C vs. 2M-C 8M-D vs. 2M-D Force at stirrup yield (kips/ft) Difference 10.36 vs. 11.77 14% change between 2" vs. 3" 15.41 vs. 11.86 vs. 13.64 11.82 -11% change between 4" vs. 3" -0.3% change between 2" vs. 3"

Qualitative evidence can also be obtained by examining the strain plots in the Appendices for the models where cover thickness was varied.

Force at web delam. (kips/ft) Difference 15.08 vs. 15.06 -0.2% change between 4" vs. 3" 15.34 vs. 14.56 -5% change between 2" vs. 3"

The conclusion reached is that cover thickness influences lateral pullout resistance, but it is not the only driver of pullout resistance. The results of the parameter study are influenced by the fact that when the cover is reduced, for the same overall web thickness, the moment arm for the stirrups is increased, and this is an off-setting influence on pullout resistance. As will be discussed further in the conclusions, it appears appropriate to check cover concrete thickness for resistance to initial cracking, but not to include cover concrete tensile strength in the calculation of regional transverse bending strength. Number and Configuration of Tendon Ducts This was evaluated by comparing Configurations 1, 2, and 3 (from Figure 5-18), which involve comparing Models 1M vs. 2M, 3M-D vs. 2M-D, 5M vs. 2M, and 10M-C vs. 7M-C. Results are shown in Table 5-10. Table 5-10 – Effect of Different Duct Configurations – Thin Webs
Model-Web 1M-A vs. 2M-A 1M-B vs. 2M-B 1M-C vs. 2M-C 3M-D vs. 2M-D 5M-B vs. 2M-B 10M-C vs. 7M-C Model-Web 1M-B vs. 2M-B 1M-C vs. 2M-C 5M-B vs. 2M-B 10M-C* vs. 7M-C
* Never reached delamination limit

Force at stirrup yield (kips/ft) 9.93 vs. 11.77 11.38 vs. 13.66 11.77 vs. 13.64 12.26 vs. 14.11 vs. 17.13 vs. 11.82 13.66 16.31

Difference 19% change - Config. 1 vs. 2A 20% change - Config. 1 vs. 2A 16% change - Config. 1 vs. 2A -4% change - Config. 2B vs. 2A -3% change - Config. 2B vs. 2A -5% change - Config. 3A vs. 2A Difference 18% change - Config. 1 vs. 2A 19% change - Config. 1 vs. 2A -23% change - Config. 2B vs. 2A * change - Config. 3A vs. 2A

Force at web delam. (kips/ft) 12.83 vs. 15.17 12.60 vs. 15.06 19.60 vs. 21.20 vs. 15.17 17.11

Page 108

3A vs.59 vs. 1).22 vs. 12M-C 10M-A vs.Config. 19. 15. In fact. 4B The conclusions from these comparisons are that web/duct ties make a significant contribution to the resistance to lateral tendon breakout. So in general. 14M-B Model-Web 9M-B vs.58 17.61 17. 13M-A-D to 10M-AD. 17. 3A vs. and moved toward the curve’s outside edge of the web.19 85% change . 19.Config.. 4B 15% change .Config. D. Page 109 .2% change .Config.B. 3A vs.B. Table 5-11 – Effect of Duct Tie Arrangements – Thin Webs Model-Web 10M-B vs.Config. This comparison is shown if Table 5-12. 9.e. 3A vs. 13M-A 10M-B vs.06 vs. and comparing 13M-A. as measured by the delamination criteria. the performance significantly improves.B. 14M-B Force at stirrup yield (kips/ft) 17. D to 2M-A. Configuration 3A exceeded 200% Pc.47 Difference 2% change .B. and an additional 4% improvement was demonstrated by spreading the bundles further apart (4. to Configurations 1. D to 12M-A.58 vs. 4B Force at web delam. Number and Configuration of Duct Ties This was evaluated by comparing Configurations 4A. Config.50 12. It should be noted. and due to requirements on location of C. 3. 3A vs. D to 2M-A. D.5” versus 1. that the influence on stirrup yield performance by spreading individual ducts apart was only 5%.B. 12M-C to 12M-B. 4A 17% change . comparing 7M-A. 4A vs. (kips/ft) Difference 10. D to 9M-B. 20.13 vs. D. 12M-C 9M-B vs.94 vs. of the tendon group.Config. when the ducts are spread apart. 4A 14% change . This is covered by comparing Model/Webs 11M-D to 10M-D.5” separation).55 13. 4B.B. 15. 2A vs.17 17. 4A -0. 17.22 vs. 2. 13M-C 10M-D vs. performance further improved. This comparison is shown in Table 5-11. 2A vs. and it is often impractical for designers to spread individual ducts apart due to lack of space in the web.13 vs.Clearly. 17.G. 12M-B 10M-C vs. 3A). however. 2A versus Config.01 17.55 10.Config.97 vs.Config. 4A 2% change . Stirrups Stirrup spacing was evaluated by comparing Model-Webs 6M-A. 4A 14% change .Config. and 14M-B. 12M-B to 10M-B. so the improvement in delamination performance was very large. 4B 9% change . 3A vs. When the individual ducts were separated (i. 13M-D 12M-B vs. 13M-B 10M-C vs. a prudent recommendation is to require a maximum of 3 ducts per bundle. Roughly 20% resistance force improvement was demonstrated by separating the 5-duct bundle into two bundles (Config.

37 vs. Especially Assumed Tensile Strength This was evaluated by comparing Model-Webs 5M-C to 2M-C.(kips/ft) 13. 2M-A 6M-B vs.17 14. Concrete Material Properties.66 9% change with 50% more stirrup steel 10. 15. strength 15. strength Force at web delam. 12M-D to 11MD. 6M-C 12M-B vs.Table 5-12 – Effect of Stirrup Spacing – Thin Webs Model-Web 6M-A vs. (kips/ft) Difference 15. 5M-C 2M-C vs. 2M-D 7M-B vs. stirrups are not a very effective deterrent against this failure mode.06 vs.36 vs. 15.38 -16% change with 33% less stirrup steel -13% change with 33% less stirrup steel -11% change with 33% less stirrup steel -10% change with 33% less stirrup steel -18% change with 33% less stirrup steel Difference 9% change with 50% more stirrup steel 12% change with 50% more stirrup steel -17% change with 33% less stirrup steel -11% change with 33% less stirrup steel Force at web delam.82 13. 13. But if the duct layout and duct ties are properly detailed to eliminate the local pullout failure mode.30 22% change with 50% larger concr. 6M-C Force at stirrup yield (kips/ft) Difference 13. This comparison is shown in Table 5-13.82 10% change with 50% more stirrup steel 16. 2M-B 7M-D vs.66 11. 2M-D Force at stirrup yield (kips/ft) Difference 8. 15. 2M-B 6M-D vs. 12.58 vs. the stirrup spacing does define the web “regional” beam strength.06 vs.56 This indicates that web section strength is significantly influenced by the stirrup spacing only when web/duct tie reinforcement is NOT used or when the web-splitting/lateral shear-failure does not occur. 2M-D 13M-A vs. strength 13.95 17% change with 50% larger concr. 17.64 vs.34 17. 12M-C Model-Web 2M-C vs.57 vs.01 vs.15 -19% change with 50% smaller concr.77 vs.77 34% change with 50% more stirrup steel 12. In other words. strength Page 110 . if the failure mode is tending toward local duct pullout.33 vs.61 vs.92 vs. 5M-C 2M-C vs. 12M-D Model-Web 6M-B vs.17 13. 11.21 17. 12M-B 13M-D vs. 19. 11. 14.17 vs.55 -18% change with 50% smaller concr. 11. 15. Table 5-13 – Effect of Concrete Strengths – Thin Webs Model-Web 2M-C vs. 13. 2M-B 7M-D vs.77 vs. and 12M-C to 12M-B. 13.04 vs.51 vs. 15.56 18. 6M-C to 2M-C. 2M-D 7M-B vs. strength -0. 12M-A 13M-B vs.64 vs. 2M-B 6M-D vs.58 12. 18.2% change with 50% larger concr. 16.

geometry variabilities were introduced directly into the models – i.0” 9’. Thus designers should be directed toward design rules that will ensure good performance regardless of variabilities in concrete tensile strength.5” 21’. This tends to further underscore the importance of providing web/duct tie reinforcement.e.6” T endo ns assum ed 6-31 F 0.. When web/duct tie reinforcement is used. and another have a different thickness. concrete tensile strength has little effect on the section strength. Tendon Duct and Local Reinforcement for the Local Analysis Prototype for a Single-Cell Box Page 111 .0” Figure 5-24. one web can have one thickness. and low reliability. ~ 3. where the stirrup is ‘smeared’). with out-of-page thickness equal to one stirrup spacing. because of the various parameters involved in reinforced concrete design. and with the tendon duct arrangements and local reinforcements that were shown in Figure 5-17 and further illustrated in Figure 5-24.So the web section strength tends to be significantly influenced by the concrete strength only when the section is prone to web-splitting/local-lateral shear-failures. concrete tensile strength has wide variability. Webs 1 and 2 demonstrate the differences related to web sloping away from the radius of curvature versus sloping toward the radius of curvature. i. etc.0” 11’. The duct and tie configurations are referred to as #6 and #6a. (unlike in two dimensions. This allows introduction of the out-of-page compression due to prestress.6” 1 web Duc ts 5”00 # 4 at 12” Typ # 7 at 12” Typ # 6 at 12” 2” Cl 20” 1. when vulnerable duct placement is used or web/duct tie reinforcement is NOT used.. to differentiate from the configurations of the multi-cell.5” Cl 20” 3’-6” 9’. and #6a has the two rebar ties as shown. c Local Analysis of Single-cell Box Girders Model Prototype: Single-Cell CIP Box Girder A single-cell box configuration was used as shown in Figure 5-24. The basic model uses 3d elements and a slice.e. The stirrups (and other rebar) are modeled explicitly.6” 5’-0” 21’. Using this model framework.6” 10’. The only difference between these is that #6 has no duct ties.

6 k-ft/ft As described earlier. For the webs of the single-cell geometry prototype.right. So the baseline Pc becomes Pc = [52. “6a” would include horizontal crossties. (Duct / Tie configuration “6”. The results of the single-cell analyses are shown in the following tables and figures. B-2. but removing resistance factors. the increase to capacity caused by load distribution is small. moment-fixity effects are approximately 0.7 = 31.125 = 52.1 k/ft This is used as “100%” in the results tables and plots.left.Web 1 . Pc was defined as a “Capacity” calculated using conventional means.67 feet Mn = 42.) Similar to the multi-cell analysis series.7 for exterior webs. This provides the best direct comparison to finite element analyses. Page 112 .6 for interior webs and 0.6 / (h/4)] / 0. The maximum principal strain contours illustrate the general level of damage to the concrete surrounding the tendon ducts (maximum tensile strain regardless of orientation).1 k-ft/ft Removing the safety factor   Mn = 46.8 k-ft/ft Applying an over-strength factor for rebar strain hardening (which is included in the FE analysis). h = 9. so no capacity increase is applied for this. Since there are only three ducts distributed vertically.). etc. in order to establish a baseline for comparison with design calculations. Web 2 . The figures are included in Appendix Fa and are numbered (Fig. B-1. Pc was calculated as follows. Mo = Mn x 1.

5"/2" Concr. Deflection of Web Mid-height Lateral Force vs.418 1.5"/2" 1.5"/3" 3:/3.158 0. Table 5-15 – Deflections (inches) Measured at Mid-height of Webs on Backside Web 1 Mid Web 2 Mid Percent Model # Capacity Quarter Quarter 1S 75% 100% 125% 150% 0. 150% Pc Distortions (change in web width) 3 locations along duct bank 75%.5"/2" 1.5"/2" 1. Tens. 150% Pc Detailed results are included in Appendix F-a. Deflection of Web Quarter-height Maximum Principal Strain Contours in Concrete at 75%.5"/2" 1.5"/2" 1.496 Page 113 0.5"/2" 1. (x fc`) 4 4 4 4 4 6 4 2 4 4 4 4 4 4 2 6 4 4 4 4 Analysis # 1S "baseline" 2S 3S 4S 5S 6S 7S 8S 9S 10S Model Type Single-cell Single-cell Single-cell Single-cell Single-cell Single-cell Single-cell Single-cell Single-cell Single-cell Single-Cell Models – Analysis Results The results of the 10 different single-cell girder local models allow for qualitative and quantitative assessment and comparison of the cases analyzed. 125%.* 6 6 6 6 6 6 6 6 6 6 6 6 6a 6a 6a 6a 6a 6a 6a 6a Bundle Vert.) 12 12 12 12 12 12 8 12 18 12 12 12 12 12 12 12 18 8 12 12 Cover Thickness 1.179 0.5"/2" 1.174 0.5" 1.5"/2" 1.5"/2" 1.5"/2" 1.978 2.5"/2" 2.Table 5-14 – Single-cell Box Variations / Parameter Study Web # 1 2 1 2 1 2 1 2 1 2 1 2 1 2 1 2 1 2 1 2 Webties N N N N N N N N N N N N Y Y Y Y Y Y Y N Web Thickness 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 Duct/Tie Config. Pos.5"/2" 1.5"/2" 1.499 0. 100%.903 0. Str.176 0. 100%. 100%. midheight midheight 1/4 height bottom midheight midheight midheight midheight midheight midheight midheight midheight midheight midheight midheight midheight midheight midheight 1/4 height bottom Stirrup Spacing(in.076 2.402 1.5"/2" 1.090 2. 125%.208 2.      Lateral Force vs.5"/2" 1.224 .5"/2" 1.5"/2" 1. 150% Pc Strains in stirrup rebar at 3 locations along duct bank at 75%.360 0. 125%.404 1.

685 1.140 0.345 0.542 0.292 1.961 2.314 0.868 0.409 3.371 3S 4S 5S 6S 7S 8S 9S 10S Page 114 .148 0.358 1.915 2.182 0.156 0.335 0.102 0.707 0.377 0.410 31.906 2.846 1.131 3.570 3.785 5.272 0.168 0.340 4.068 2.372 0.161 0.170 0.256 0.445 0.190 0.274 0.805 0.154 0.050 4.131 0.293 0.056 2.463 0.295 0.169 0.678 1.357 0.600 0.987 2.808 2.246 0.447 0.435 1.526 0.722 1.866 2.463 1.972 2.400 1.146 0.923 2.2S 75% 100% 125% 150% 75% 100% 125% 150% 75% 100% 125% 150% 75% 100% 125% 150% 75% 100% 125% 150% 75% 100% 125% 150% 75% 100% 125% 150% 75% 100% 125% 150% 75% 100% 125% 150% 0.888 2.232 1.251 3.805 4.681 1.286 0.095 2.659 1.274 31.187 27.168 0.835 0.011 2.159 0.239 0.420 1.588 0.877 3.357 0.413 0.298 0.484 36.647 22.212 2.154 0.137 0.192 0.151 0.548 0.871 2.691 0.996 0.127 0.135 0.462 1.701 1.366 0.815 0.335 0.174 0.974 0.405 1.338 0.728 0.320 0.493 19.702 0.778 1.457 1.852 5.764 2.355 0.609 1.140 0.242 6.775 3.204 2.365 1.849 2.472 1.155 0.904 0.301 4.521 26.046 2.341 0.365 6.362 0.307 0.724 1.066 2.462 23.030 0.174 0.192 0.924 3.953 2.918 2.152 0.351 0.288 2.205 0.195 0.

00164 0.02163 0.04636 0.00910 0.00218 0.01638 0.00277 0.03558 0.00345 0.00140 0.03425 0.02560 0.00713 0.00137 0.00851 0.00175 0.00310 0.03653 0.02075 0.02640 0.00335 0.00323 0.00106 0.00329 0.00113 0.00159 0.00112 0.Table 5-16 – Stirrup Strain (%) on Curve Inside .00160 0.00421 0.01066 0.02543 0.04088 0.00134 0.00150 0.00126 0.00311 0.03942 0.00525 0.01271 0.01319 0.03223 0.00109 0.03219 0.00188 0.00177 0.00097 0.00110 0.04032 0.02096 0.00154 0.00807 0.00407 0.01528 0.00176 0.00134 0.00580 0.00049 0.Face Web 1 Percent Model # Capacity Duct 2 Duct 4 Duct 6 Duct 2 Web 2 Duct 4 Duct 6 1S 75% 100% 125% 150% 0.01341 0.01628 0.04969 0.00387 0.00150 0.00149 0.00244 0.01755 0.00187 0.00158 0.00915 0.01851 0.00141 0.01872 0.00724 0.00332 0.01690 0.01529 0.00520 0.00284 0.00243 0.00165 0.01098 0.03987 0.00937 0.06683 0.04032 0.03204 0.00563 0.00150 0.00421 0.00132 0.01953 0.00178 0.00236 0.24986 0.01269 0.00190 0.00680 0.00259 0.05776 0.00257 0.00140 0.00127 0.00339 0.00310 0.00911 0.05699 0.00160 0.00190 0.02766 0.03406 0.04096 0.01838 0.15104 0.00187 0.01017 0.00329 0.00450 0.01506 0.02689 0.00092 0.00361 0.00684 0.00275 0.00215 0.04032 0.00189 0.01506 0.00090 0.00152 0.00452 0.00376 0.00204 0.00172 0.07417 0.01021 0.00215 0.04659 0.02447 0.01499 0.00326 0.21465 0.00415 0.04019 0.00076 0.00222 0.01541 0.04799 0.00059 0.03028 0.00712 0.03334 0.00244 0.01506 0.03765 0.00135 0.04083 0.01612 0.04853 0.00166 0.00162 0.00332 0.00158 0.02955 0.03541 0.04107 0.00902 0.01352 2S 75% 100% 125% 150% 3S 75% 100% 125% 150% 4S 75% 100% 125% 150% 5S 75% 100% 125% 150% 6S 75% 100% 125% 150% 7S 75% 100% 125% 150% 8S 75% 100% 125% 150% Page 115 .00097 0.00759 0.00137 0.00942 0.00200 0.02036 0.01451 0.06873 0.01429 0.10079 0.03034 0.01023 0.00332 0.03238 0.01701 0.00124 0.00642 0.02690 0.00137 0.01750 0.00310 0.02922 0.00173 0.00279 0.00150 0.04288 0.

00380 0.00387 0.02789 0.19842 0.03730 0.23078 0.01708 0.00703 0.02768 0.01851 0.00783 0.02662 0.01958 0.00129 0.00110 0.00207 0.00207 0.03106 0.00178 0.02184 0.01140 0.00841 0.00630 0.02640 0.04844 0.05138 0.00920 0.09676 0.05278 0.00151 0.00750 0.05973 0.00373 0.07189 0.00196 0.01253 0.00072 0.00210 0.02321 0.9S 75% 100% 125% 150% 0.00174 0.06473 0.00698 0.00129 0.00184 0.03636 0.04984 10S 75% 100% 125% 150% Page 116 .09182 0.19175 0.

0503 0.0461 0.2546 0.0213 0.0671 0.0876 0.0058 0.0573 0.0666 0.0303 0.0462 0.0645 0.2152 0.1114 0.0179 0.0069 0.0987 0.0412 0.0455 0.0356 0.2037 0.0405 0.5930 0.3204 0.0735 0.0543 Page 117 .1379 0.0366 0.3115 0.0026 0.0032 0.0936 0.4035 0.0158 0.0190 0.1347 0.0600 0.0246 0.0139 0.0104 0.0339 0.0026 0.2166 0.0813 0.0935 0.1464 0.0212 0.0106 0.0021 0.0094 0.1555 0.0075 0.4472 0.2289 0.0039 0.0077 0.0106 0.0095 0.0971 0.0087 0.0574 0.1594 0.0064 0.0051 0.0939 0.1824 0.0165 0.0085 0.0274 0.0898 0.2584 -0.0057 0.0112 0.0442 0.0133 0.0849 0.3781 0.0141 0.0176 0.0104 0.4056 0.1207 0.1541 0.1446 0.0569 0.4169 0.0031 0.0277 0.0106 0.2876 0.0274 0.0732 0.0247 0.0053 0.0567 0.0273 0.4625 0.0630 0.0126 0.0092 0.0316 0.0253 0.0034 0.0196 0.0442 0.0632 0.2057 0.0261 0.0635 0.1985 0.2093 0.0563 0.0217 0.0298 0.1241 0.0127 0.0302 0.0413 0.0406 0.0922 0.0424 0.0048 0.0223 0.0459 0.0734 0.0217 0.0245 0.0129 0.0425 0.0250 0.1702 0.0862 0.7850 0.0126 0.0658 0.0346 0.0169 0.1441 0.0386 0.0480 0.0227 0.0392 0.0463 0.0207 0.0235 0.0250 0.0328 0.0928 0.0348 0.0137 0.0189 0.1079 0.1254 0.1443 0.1102 0.1398 0.2021 0.0512 0.0095 0.0164 0.0157 0.0933 0.0153 0.0872 0.0496 0.0128 0.0247 0.0052 0.2093 0.0903 0.0779 0.2377 0.1145 0.3469 0.0237 0.4946 0.1959 0.0132 0.0361 0.0173 0.0077 0.0748 0.0053 0.0062 0.0218 0.0673 0.0472 0.2341 0.1084 0.0630 0.0390 0.1995 0.0351 0.0875 0.1122 0.0824 0.0226 0.2064 0.0479 0.0297 0.0345 0.1055 0.0175 0.2166 0.0521 0.Table 5-17 – Distortion (Web Thickness Changes – inches) at Mid-height of Ducts Web A Percent Model # 1S Capacity 75% 100% 125% 150% 2S 75% 100% 125% 150% 3S 75% 100% 125% 150% 4S 75% 100% 125% 150% 5S 75% 100% 125% 150% 6S 75% 100% 125% 150% 7S 75% 100% 125% 150% 8S 75% 100% 125% 150% 9S 75% 100% 125% 150% Duct 2 Duct 4 Duct 6 Duct 2 Web B Duct 4 Duct 6 0.1219 0.0332 0.1356 0.0093 0.0374 0.0278 0.1147 0.0965 0.0108 0.0031 0.0084 0.

For 20” webs.5176 0. for sections without web ties.0600 0.0124 0. the section is at a web splitting or a cover concrete spalling condition. but this is not necessarily web failure. note that for Load Factor Design. Having the double row of tendons was found to concentrate the local damage area within the web.0% will generally show wide-open cracks and potential spalling from the section.0163 0.1321 0.3% can be considered to be heavily cracked.0306 0. Concrete with strains in excess of 1. concrete reinforcement is designed to yield.0072 0.. and there were no surprises as to the performance of the sections. has been summarized in Table 5-18. Visible concrete cracking occurs at strains of approximately 0.3%.016%.0064 0. These are the total forces (sum of all tendon ducts in the web) applied when any part of the stirrup reaches yield.0132 0.1604 0.0250 0. the following damage limit criteria is suggested:  Stirrup rebar strain exceeding yield (i.2% strain for Grade 60 steel). the sections with duct ties performed better than without.2563 0. For sections with web ties.0154 0. this represents a distortion ratio (average strain through the section) of 0.10S 75% 100% 125% 150% 0. it was again assumed that an upper bound on crack width of 1/16” as an indicator of such a failure. concrete with maximum principal strains of 0.0399 0.0376 Discussion of Results Analyses of these models showed similar trends as the multi-cell model analyses. Page 118 .2885 0.0751 0. but having the 20-inch web thickness with local reinforcement is quite adequate to accommodate this. As expected. the delamination is evidence of a local splitting or lateral shear failure within the web.0202 0.1939 0. the analysis results can be used to compare the web design parameters.2918 0. Significant distortion or delamination (change of width of the webs) also represents an upper limit on capacity for webs. 0.0279 0.e. For purposes of interpreting the 3d finite element analysis results.1449 0. Similar to the multi-cell studies.0661 0. yield should be considered an upper bound criteria for unfactored loads. this means the web ties have yielded.0800 0. Stirrup Yield.   One of the criteria.

The slab resistance to these moments is stronger (about 2 times stronger. and this increases capacity rather than decreasing it.5” vs.88 26. 2 were by 50% and 75%. The following summarizes the relevant strength comparisons: Table 5-19 – Effect of.21 28.16 vs. 28. a possible reason for this is that loading the inside radius web (Web 2) creates positive transverse moment in the top slab adjacent to the web (tension on the bottom of the slab).95 29. Cover Thickness Cover thickness was varied in model 6S.87 23.83 28.21 26. and this translates to more strength in the associated web. This is because a) for 2” cover and above.64 26. 25. but the comparisons for stirrup yield are inconclusive. the inside radius web.83 28. cover failure does not control the failure mode. Page 119 . whereas loading Web 1 creates negative moment. where increases for Web 1.05 17.77 24. Web Slope Similar to the multi-cell series.2% Strain) Total Force (K/ft) Web 1 Web 2 25.6 20. and examining the results tables and plots has resulted in the following observations. respectively. 1S-1 6S-2 vs. 2” 0% increase with 3. sloping toward the center of the curve was found to be roughly 10% stronger than the outside radius web. and b) when the cover is less. Cover Thickness – Thick Webs Model-Web 6S-1 vs. based on typical deck reinforcing) in positive moment than in negative moment.18 16.Table 5-18 – Force at Stirrup Yield (0.04 26.16 28.04 vs.78 16.18 28. 2” The local concrete damage was less severe with thicker cover.1 26. sloping away from the center of the curve. 1S-2 Force at Stirrup Yield (kips) 24.12 27. After further study. the “moment arm” between the stirrups is more.09 18.38 Model # 1S 2S 3S 4S 5S 6S 7S 8S 9S 10S Using these criteria.17 31.18 Difference 7% increase with 3” vs.

0. The results are shown in Table 5-21 Page 120 .035”) 26.Number and Configuration of Tendon Ducts The only variations studied in the single cell case was the positioning of the duct group. 28.048” vs. But when the ducts move down. So this is a different mechanism than the failure modes observed for tendon ducts at mid-height. 0. comparing 5S-1 to 1S-1.025”) Model-Web 7S-1 vs. but one that still warrants consideration in design. which had no duct ties.87 vs. with Duct ties (31% less delamination) -5% change with Duct ties (24% less delamination) -2% change with Duct ties (little change in delamination) -4% change with Duct ties (24% less delamination) These comparisons show that for the wider web (20”) and double row of tendon ducts. the lateral shear is divided equally between the top and bottom of the web. Delaminations (width changes in the web) are reduced by 24-31% with duct ties compared to without duct ties. to 6. 0. studied in Configurations 2S and 10S.046”) 29. 1S-2 8S-2 vs.18 (0. i.063”) 27. When the ducts are located at the mid-height.050” vs. 2. But it was again observed that when the ducts occurred near the bottom of the web (either “quarter-height” or “bottom” as was tested in Configurations 4M and 14M) the force at “failure” is substantially lower than when the ducts are placed at the mid-height. 3S-2 9S-1 vs.83 (0. 31. Table 5-20 compares Model/Webs 7S to 1S. Stirrups Stirrup spacing was evaluated by comparing Model-Webs 4S-1 to 1S-1. and 9S-1 to 5S-1.21 vs. The reason for this is a tendency toward lateral shear failure of the overall web. 25.. the ties do not make a significant difference in the force to cause stirrup yield.18 (0.019” vs. the bottom of the web carries most of the lateral shear.60 (0.95 vs. The number and relative position of the ducts to each other was held constant.64 vs. 0. 5S-1 Difference 2% incr. Number and Configuration of Duct Ties This was evaluated by comparing duct tie Configurations 6a. 1S-1 7S-2 vs. and a tendency toward flexural damage in the top slab. and comparing 9S-1. 2 to 7S-1. on average as much as 25% to 40% lower when comparing these cases to other similar cases. thus weakening the whole system. 8S to 3S.024” vs.e. Table 5-20 – Effect of Duct Ties – Thick Webs Force at Stirrup Yield (kips) (and delamination at 100%Pc) 26. which had duct ties. but they do make a large difference in the delamination-damage that can occur within the web. 28.

12 vs. and low reliability.Table 5-21 – Effect of Stirrup Spacing – Thick Webs Model-Web 4S-1 vs. 25. 7S-1 9S-2 vs.87 vs. 26.21 27. and 8S-1. 25.18 26. Of the various parameters involved in reinforced concrete design.83 31. strength 0% change with 50% smaller concrete.64 Difference 2% increase with 50% larger concrete tensile strength 7% decrease with 50% smaller concrete.21 23. 7S-2 Force at stirrup yield (kips) 28.83 29. Especially Assumed Tensile Strength The effect of concrete strength was evaluated by comparing Model-Webs 3S-2 to 1S-2. 28. 26.1 vs. 1S-2 8S-1 vs. strength 5% increase with 50% larger concrete. 1S-1 5S-1 vs. The results are shown in Table 5-22 Table 5-22 – Effect of Material Strength – Thick Webs Model-Web 3S-2 vs. the web section strength tends to be only marginally influenced by the concrete strength. Concrete Material Properties. repeating the trend observed in the multi-cell analysis. Here again.78 vs. 28. strength So. So designers should use design rules that will ensure good performance regardless of variabilities in concrete tensile strength. 7S-2 Force at stirrup yield (kips) 20. 7S-1 8S-2 vs. concrete tensile strength has wide variability. 4S-2 to 1S-2.64 Difference 29% increase with 50% more stirrup steel 21% decrease with 33% less stirrup steel 14% decrease with 50% less stirrup steel 21% increase with 50% more stirrup steel So the web section strength tends to be significantly influenced by the stirrup spacing for this geometry also. perhaps even more so than for the multi-cell geometry. 1S-2 4S-2 vs.18 26. Page 121 .60 vs. 2 to 7S-1. stirrup spacing is a driver of web “regional” beam strength. 2.18 vs.95 vs. When web/duct tie reinforcement is used. concrete tensile strength has less effect on the section strength.17 vs. 26. and mostly this influence occurs when web/duct tie reinforcement is NOT used. 26. 1S-1 9S-1 vs.

depending on the detailing used. Exterior webs have more flexible end conditions at their connection with the top and bottom slab. failures were dominated by local lateral shearing. but much of this difference is caused because these are exterior webs rather than interior ones.. It is believed that a prudent recommendation is to require a maximum of 3 ducts per bundle. Web Slope . and an additional 4% improvement was demonstrated by spreading the bundles further apart (4. but not to include cover concrete tensile strength in the calculation of regional transverse bending strength. with the 20-inch webs and double row of ducts.5” separation). so it is clear that the detailing has very significant influence on resistance to lateral pullout. but this is explained by the fact that for the thicker web.the sloped webs were found to be significantly weaker (roughly 30%) than the vertical webs. Web Thickness – has significant influence on resistance to regional transverse bending and tendon pullout caused by web thickness. stirrup yield was reached at a range from 52% Pc up to 100% Pc). or moving the ducts toward the curve-outside-face of the web reached as high as 185% of Pc. It is believed this is due to the difference in positive bending versus negative bending strength of the top slab. Roughly 20% resistance force improvement was demonstrated by separating the 5-duct bundle into two bundles. The results of the parameter studies are influenced by the fact that when the cover is reduced.e. Lateral force for Web D applies positive moment to the top slab. Summary of Influences from Detailing Parameters Web Depth . for the same overall web thickness.e. The variations in force to cause local duct bank breakout (either local shearing of web delamination) were even larger. performance further Page 122 .. which is an off-setting influence on pullout resistance. it was found that all of the multi-cell box girders achieved this target capacity. For the single-cell example.5” versus 1. the performance significantly improves. Number and Configuration of Tendon Ducts . the finite element analysis showed capacities which were mostly lower than the hand-calculated regional transverse bending capacity (i. while stronger details which utilize spreading apart the ducts.when ducts are spread apart.can be adequately accounted for by considering and designing for web moments. stirrup yield was reached at 107% of Pc). When individual ducts were separated and moved toward the curve’s outside face of the web. but it is not the only driver of pullout resistance. and this produces larger mid-height moments. For stirrup yield. and the positive moment reinforcement is approximately 2 times that of the negative moment reinforcement. developed based on regional transverse bending considerations). The baseline (Model 1M) interior webs achieved it marginally (i. the moment arm for the stirrups is increased. capacity formulae based on regional flexure considerations appear to be appropriate for design. Comparison of Webs A to D for the inclined webs show that Web A is generally weaker than D by about 10%.d Conclusions from Local Analyses General observations on “Capacity” Using the “capacity” definitions described in this section (Pc. Cover Thickness – Inside face duct cover influences lateral pullout resistance. adding duct ties. It appears appropriate to check cover concrete thickness for resistance to initial cracking.

improves. When measured by the delamination/local-lateral shear criteria, Duct Configuration 3A exceeded 200% Pc, so the improvement in delamination performance was very large. It should be noted, however, that it is often impractical for designers to spread individual ducts apart due to lack of space in the web, and due to requirements on location of the C.G. of the tendon group. Number and Configuration of Duct Ties - make a significant contribution to the resistance to lateral tendon breakout. Stirrups –regional transverse bending strength is directly tied to stirrup area, but it controls the design only when web/duct tie reinforcement is NOT used or when the web-splitting/lateral shear-failure does not occur. In other words, if the failure mode is tending toward local duct breakout, stirrups are not a very effective deterrent against this failure mode. But if the duct layout and duct ties are properly detailed to eliminate the local pullout failure mode, the stirrup spacing does define the web “regional” beam strength. Concrete Material Properties, Especially Assumed Tensile Strength - web section strength can be significantly influenced by concrete tensile strength only when the section is prone to web-splitting/local-lateral shear-failures, i.e., when vulnerable duct placement is used or web/duct tie reinforcement is NOT used. When web/duct tie reinforcement is used, concrete tensile strength has little effect on the section strength. Thus designers should be directed toward design rules that will ensure good performance regardless of variabilities in concrete tensile strength. Recommendations for Web Capacity Design Web capacity design for lateral tendon force resistance should be a three-step calculation: Regional flexure check, local-lateral shear/breakout check, and cover concrete cracking check Regional transverse bending The regional mechanism is the web acting as a vertical beam loaded laterally near its center. Fundamentally, the calculation follows the equation: Mu = (Load Factor) (Moment Fixity Factor) (1/4) (Pj/R) hc This equation (a modified version of the Caltrans Equation) and the corresponding stirrup spacing should be evaluated for each web of a box girder separately – not for the total box divided by the number of webs. The radius is different for each web, and it was found that the Moment Fixity Factor is also different. AASHTO LRFD currently applies a load factor of 1.2 to the Pjack tendon force, which is judged to be reasonable. Appropriate moment fixity factors are 0.6 for interior webs and 0.7 for exterior webs. The stirrup sizing and spacing should then be calculated using Ultimate Strength design such that: Mn  Mu
Page 123

However, the Vs stress in the stirrups due to vertical shear in the web should be added to the stress due to flexure in the sizing and spacing of the stirrups. At the midheight of the web, on the inside-curve side of the web, these stresses are directly additive. Local Lateral shear Check The local lateral shear mechanism involves the complex behavior that develops in the concrete and stirrup region immediately adjacent to the duct bank. This may be checked by the following equations developed by the University of Texas (Van Landuyt, 1991): For a strip of web one foot long, the applied lateral shear demand along a plane deff long is: Vd = Pj/R  2 Vc capacity of the cover-beam along this plane may be taken as: Vc = 24 deff Where f c

 0.75 (reduced due to uncertainties in concrete quality within the cover-beam)

When the spacing between ducts is greater than or equal to the duct diameter: d eff  d c  ( Duct Diam.) / 4 + s/2 or deff = tw – (Duct Diam)/2 whichever is least. where: s = space between ducts (assume 0 if s 1.5” or for single ducts) tw = thickness of web When the spacing between ducts is less than the duct diameter or for single ducts: deff = dc + (Duct Diam)/4 where dc = cover over the ducts Figure 5-25 shows what is intended by the above equations for deff.

Page 124

Figure 5-25 Definition of deff (after Van Landuyt, 1991) There has been discussion within the industry as to selecting deff (some refer to this as the “lateral shearing plane depth). Some say this should be no greater than dc (the cover concrete depth) due to uncertainties in the concrete interaction with the ducts, but the local analyses conducted here allow for the extra width of ¼ of a duct diameter. If this lateral shear is exceeded, the most effective design remedy is the addition of duct-tie reinforcement. Cover Concrete Cracking Check Evaluating the cracking of the cover concrete is a check that is made to insure serviceability since it is recommend that the lateral tendon forces be completely carried by the strength elements of the above two checks. But this serviceability check remains critical to achieving a good design, because significant cover cracks running along the tendons should be avoided for long term structure durability. The flexure on the cover beam involves a complex mechanism because it is uncertain what the level of adhesion is of the cover concrete to the duct bank and to the concrete surrounding the duct bank. Assuming there is no adhesion between the metal duct and the web concrete in the radial direction of the duct, the flexure calculation proceeds as follows. The cover-beam acts as a vertical beam “built-in” or fixed at top and bottom. Thus the following moments are produced:
Page 125

which has come out of the local analysis work.M ends  M center wL 2  (Pj/R / L ) L2 / 12 12 wL2  24 L is the height of the duct bank  and bd c3 12 Mn  Mu Where Mn is defined by an allowable tensile stress for concrete of 5 f c . the moment at the center of the duct banks quickly becomes M center  wl 2 8 So these factors and conservative tensile strengths are judged to be appropriate to prevent this progressive cracking mechanism from occurring.55. While this may appear to be quite conservative in terms of choice of tensile strength and choice of . 1-1/2” for the analyses performed here). and  = 0. and from examination of some local breakout failures in various bridges and test structures. A generic duct tie detail is shown in Figure 5-26 Page 126 . or by a minimum of 1/3 of one duct diameter (for example. The allowable tensile stress should also be reduced by the tensile stress in the concrete at the critical point due to regional transverse bending. Duct ties should be well anchored with hooks around stirrup reinforcement. Sub-bundles should then be separated by either a ducttie rebar. is to limit the number of ducts of a “sub-bundle” to no more than three. Other Local Detailing Guidelines A further guideline. it should be noted that once cracking begins within the interior of the cover concrete near the top and bottom of the duct bank.

in combination with sloped exterior webs. As the stirrup cage gets deeper. can often lead to a substantial reduction in concrete cover between the stirrup cage and the interior face of web. This may be mitigated somewhat by rebar spacer requirements at mid-height of webs to help control stirrup movement during the pour. but the designer should be aware of possible variations in the actually constructed dimensions. Duct float. Dimensional changes of 0.Figure 5-26 – Generic Web and Duct Tie Detail Construction Tolerances Designers should consider the practical aspects of construction tolerances when checking and implementing their designs. and it is not the point of this design recommendation to modify these. and duct float.5” can make considerable difference in the stresses in the web concrete and reinforcing steel. vibration process. During the web and soffit pour. but designers may wish to consider conservatively allowing for field variations in web width and in rebar placement of up to  0.5” when evaluating issues of web regional transverse bending strength. the stirrup cage gets deeper. it becomes more flexible laterally. especially is areas of low lateral shear demand where designers often specify stirrup spacing as large as 24 inches. local breakout resistance. and particularly cover-beam strength. Construction tolerances should be held to industry standards. Page 127 . The following is an example of how design and construction issues can affect conditions for lateral tendon breakout. As a box girder gets deeper. the stirrup cage has been shown to deflect laterally within the web form due to unbalanced concrete placement.

55 for local lateral shear failure.0” 3. Page 128 . 2.Several conditions can aggravate the chances for lateral tendon breakout. rebar. As suggested above.75 to 0. 3. If this may not be the case. critical dimensions could be reduced by 0. Use higher load factors and/or lower resistance factors.5” or even 1. When in doubt. provide web and duct tie reinforcement Tendon breakout failures can be expensive to repair.5. or improperly sealed duct 5.2 and 0.75) are based on the assumption that construction tolerances are reasonably well controlled. the designer is ultimately responsible for assessing the likely conditions in the field. These include: 1. Pressure from grout leakage due to poor quality duct (excessive flexibility). 2. Some engineers familiar with the potential problems have recommended  factors be reduced from 0. Excessive wobble of the ducts that can result in either reduced resistance to breakout or locally elevated lateral forces. Although the recommended design specifications should provide an adequate factor of safety in most cases. Out-of-plane forces in a vertically curved tendon due to wedging of the stand 4. Reduction of cover over the duct or rebar that can affect resistance to breakout. the designer may wish to consider one of the following three options. Distortion of empty ducts acted on by adjacent stressed ducts 6. or forms. Local curvature in ducts near anchorage zones or blisters The specified load and resistance factors (1.2 to say 1. Use dimensions that include an allowance for misplacement of the duct. Load factors could also be raised above 1. 1. damaged duct.

Other conclusions found that current design practice was adequate and did not require a change. whereas the finite Page 129 . CONCLUSIONS Several critical issues relative to the design of concrete box girder bridges were identified at the beginning of the project. The predominate construction type in some west coast states is multi-cell box girder bridges cast-in-place on falsework. Analysis guidelines were also developed to assist designers in performing response analysis. Curved spread box beams are an emerging structure type but are not widely used at this time. the grillage analogy produces results in terms of the structural members commonly considered by the designer. Single cell box girder bridges are also common but tend to dominate the type of box girder construction used on the east coast. and a review by an advisory panel with expertise in this area of study. This type is also widely used throughout the United States. analytical studies of global and local response. The following paragraphs discuss conclusions relative to the critical issues defined at the beginning of the project. in many cases.  Appropriate levels of analysis and design – Selecting the type of global analysis that should be used for curved concrete box girder bridges is one of the most important issues addressed by this project. these analysis methods are tedious and in general not practical for production design work. performing the analysis and interpreting results were developed and are included in Appendix C. the bridge is simulated as a grillage of beam elements in the longitudinal and transverse direction. Some states do not use this type of bridge on a regular basis. From the designer’s point of view this analysis method has advantages over the finite element approach. more simplified analysis methods will produce acceptable results. Several conclusions were drawn from the research conducted in this project.6. Guidelines for preparing the computer model. even when segmental methods are used.  Applicability –Curved concrete box girder bridges are used throughout the United States. Unfortunately. discussions of the experienced research team that attempted to reach a consensus on critical design requirements. This was accomplished with the grillage analogy approach. a review of published literature. Also. They may be cast-in-place or pre-cast and may be constructed segmentally or on falsework. East coast construction also uses more pre-cast segmental construction than is used on the west coast where cast-in-place construction is more dominate. Published research shows that these types of bridges are most accurately analyzed using three-dimensional finite element or similar techniques. A review of the state of practice in the United States found that both single and multi-cell box girder bridges are widely used. This makes it easier to design these elements. These are provided in Appendix C. In many cases these conclusions have translated into recommended AASHTO LRFD Specification provisions as presented in Appendix A. Methods exercised in this study that were intended to address these issues consisted of a survey of the state of practice. a detailed global analysis study was undertaken The first step was to identify a more simplified three-dimensional analysis approach that would yield results comparable to the more detailed finite element technique. Besides being a smaller and less computationally intense analytical model. Both single and multi-cell concrete box girder bridges are covered by this project. Most modern bridges of this type are prestressed. In this method. To determine the range of applicability of various analysis methods.

Extensive parameter studies were performed that included the effects of structural framing (simply supported or continuous). For central angles between 12 and 46 degrees and an aspect ratio less than 2. 4.0 – sophisticated 3-D analysis is required. This in turn requires special interpretation of some of the results. 2. Based on these parameter studies. Full Three-dimensional Analysis – This includes several different sophisticated approaches that include the grillage analogy described above as well as the finite element and other sophisticated approaches. radius of curvature. These studies showed that the radius to span length ratio as represented by the central angle between two adjacent supports was the dominant parameter that determined the accuracy of the various analysis methods. For central angles greater than 46 degrees – sophisticated 3-D analysis is required.element approach would involve considerable post processing of analytical results to accomplish the same goal. all six section properties of each beam member are required.  Section properties and member stiffness – The section properties and member stiffnesses that should be used in the spine beam analysis and the grillage analogy analysis are critical and are discussed in the analysis guidelines presented in Appendix C. This can be a daunting task if such software is not used. 3.sophisticated 3-D analysis is recommended. and bearing configuration on the response of bridges. For central angles less than or equal to 12 degrees – Plane frame analysis is acceptable 2. cross-section (includes bridge width). the limits of applicability of three analytical approaches were assessed. For central angles between 12 and 46 degrees and an aspect ratio above 2. chorded beam elements located along the centerline of the superstructure.0 – spine beam analysis is required. 3. Special formulae for some of these section properties are used to simulate various aspects of the behavior of a curved concrete box girder bridge. These studies included both grillage analysis and spine beam analysis for which plane frame analysis constituted the case of a bridge with a very large radius. Spine beam Analysis – This is a space frame analysis in which the superstructure is modeled as a series of straight. Fortunately. it will be desirable to use the live load generating capabilities of sophisticated commercially available software to rigorously envelope the live load response. The three methods considered were: 1. Plane Frame Analysis – This allows the bridge to be analyzed as if it were straight. the whole-width design approach as described in the LRFD Page 130 . For the spine beam analysis the cross-sectional area and the three rotational moments of inertia are important. 5. In the case of the grillage analogy. span lengths. Secondly. For all bridges with otherwise unusual plan geometry . The span length to width ratio (aspect ratio) of the superstructure also had a minor effect. Given the number of possible load positions.  Critical position of live loads – The number and position of the live load lanes in the transverse direction as well as their position along the longitudinal axis of the bridge is critical for curved concrete box girder bridges. the following limits for the various types of analysis are recommended: 1.

Note that the length of the various tendons will have an effect on friction losses and that tendons will also produce a transverse response in the bridge superstructure. However.specifications was shown to yield conservative results when used in conjunction with the plane or spine beam approaches. the torsion demands are taken directly from the torsion forces generated in the spine beam. If the plane frame approach is being used then these loads may be analyzed in the same manner as if the bridge were straight.  Torsion design – The design of concrete box sections for torsion is covered in the current AASHTO LRFD Specifications. Other load conditions such as centrifugal forces. Vehicular impact may be assessed using the method prescribed in the LRFD Bridge Design Specifications. When using this approach. However. some clarification of these requirements is in order. When this residual shear flow is combined with the flexural shear in the extreme longitudinal members. breaking or acceleration forces. These were discussed in the review of published literature included in Chapter 3. etc. When torsional response is being assessed by the whole-width design approach. Torsion demands usually translate to additional lateral shear demands in the webs of concrete box girders. In the case of the grillage analogy. wind. When a three-dimensional model of the bridge (either a spine beam or grillage analogy model) is being used it is important to consider the transverse position of prestress tendons. should be determined according to the LRFD Specifications and then applied to the spine beam model. it is important to distinguish between the longitudinal response along each of the webs and the affect of torsion across the whole section. The procedures to be used for torsion design for both the spine beam and grillage analogy analysis methods are illustrated in the example problem included in Appendix B.e. To correct for this deficiency it is necessary to consider the torsional forces in each of the longitudinal members at a given longitudinal location in the grillage model and apply the sum of these torsions to the entire crosssection to obtain residual shear flow about the perimeter of the section. the correct demands to be used for web shear design are obtained. These may be determined from both the spine beam and grillage analogy methods. spine beam) analysis. Webs should be designed for the combined flexural and torsional shear. the number of live load lanes should be reduced to the actual number of lanes that can fit on the cross-section and adjusted by the multiple presence factor and dynamic load factor (for truck loading only). This is done in a manner similar to that used for the spine beam. The first step in these studies was to verify that the nonlinear finite element models used could accurately predict lateral tendon breakout behavior observed in experimental studies performed at the Page 131 .  Tendon Breakout – Extensive analytical studies were performed to investigate lateral bursting stresses in curved concrete box girder bridges with internal prestress tendons. the total effect of torsion on the entire cross-section is not completely accounted for by the longitudinal member shear demands. This will greatly simplify the effort of the designer in determining live load response. the effects of torsion on web shear are partially accounted for because each web is explicitly included in the analytical model. In the case of the spine beam (i. The effect of bridge superelevation can usually be ignored. These forces must be transformed into shear flow around the perimeter of the box section. This shear flow will increase the effective shear in one web while decreasing it in another. because of the way torsional stiffness of the superstructure is distributed to the individual longitudinal members of the grillage model.

A method for checking flexural cracking of the unreinforced concrete cover over the inside of the prestress tendons – This is a new requirement that applies only to tendons that are vertically stacked. Stirrup design may be accomplished by combining the reinforcing requirements for each of these actions. At the joints in pre-cast bridges. Shear stresses in the webs – These stresses result from the flexural and torsional behavior of the superstructure.  Consideration of stresses at critical locations – There are several critical stresses that should be considered in the design of curved concrete box girder bridges. It is included to prevent maintenance. A method for assessing the local lateral shear resistance to pullout – These provisions are the recommendations from the University of Texas. Based on parameter studies conducted using these verified nonlinear finite element techniques. Regional transverse bending also exacerbates flexural cracking of the concrete cover as described in item 2 above. To best capture it with the spine beam approach. 2. which have been shown by past failures to have a significant effect on web performance and are discussed in Chapter 5. They also include provisions for considering the effect of construction tolerances. Generic web and duct tie reinforcement details are included in the commentary. Axial stresses in the top and bottom slabs and the webs – These stresses result from vertical flexure of the bridge between supports and the primary and secondary effects of longitudinal prestressing. The results of the analyses compared well with the experimental results so that there is confidence that both the experimental and analytical results have yielded accurate results. This effect is best captured in the grillage analogy approach. Page 132 . regional transverse bending can result in the need for more stirrup reinforcement in the webs. These include: 1. these shear forces result in diagonal tension stresses that can combine with the flexural tensile stresses resulting from regional transverse bending. modifications to the specifications for considering in-plane force are recommended. A method for calculating the regional transverse bending moments within a web – These moments result from the regional transverse bending of a web between the top and bottom slab of the bridge due to lateral prestress forces. 2. 3. Torsion results in shear flow around the perimeter of the crosssection that should be combined with the flexural shear. prestress tendons should be located at their correct transverse positions with respect to the bridge centerline. which were further verified by the nonlinear finite element parameter studies conducted as part of this study. Because the web lengths vary in a curved bridge. These include: 1. Vertical duct stacks are limited to three tendons high and concrete cover over the inside of the ducts should be maximized. moments and flexural shears in each web may also vary.University of Texas. When combined with global forces such as flexural shear and torsion. The results are used to determine the need for web and duct tie reinforcement. and the results of the University of Texas studies. architectural and structural problems that can arise due to longitudinal cracking of the web. In continuous superstructures or between the joints in pre-cast superstructures. Regional transverse bending of the superstructure may also occur and should be considered when determining these stresses. the shear is carried by a shear friction mechanism.

3. 5. 4. web and slab thicknesses. In addition to this. Specially design bearings so that they will not be displaced if the applied load goes into tension or very low compression. Select bridge framing to better control bearing forces. and the size and spacing of stirrups can be designed accordingly. Because of the curvature of the structure. elastically determined abutment dead load torsions should be increased by 20%. but because concrete is often sufficient for this purpose. Reshore the structure at its bearing locations prior to setting the bearings and then release the shoring after the bearings are set. this is often not a significant consideration in design. it is prudent to evaluate them at several longitudinal locations along the length of the bridge. one or more of the following: 1. the bearing forces at any longitudinal position along the bridge will vary across the width of the bridge. Use an outrigger diaphragm to increase the eccentricity of the individual bearings 6. 3. Flexural and lateral shear stresses in the vicinity of prestress tendons – Complex stresses are developed in the webs of curved concrete box girders due to the lateral forces developed by the curvature of prestress tendons. but not be limited to. Provide ballast in the superstructure to assure the envelope of bearing forces are within an acceptable range. The transverse deck and soffit reinforcing must also participate in carrying the shear flow generated by torsion. In lieu of a time dependent analysis. Transverse stresses in the cross-section – These stresses can generally be determined using the same methods employed for a straight bridge. 7. The extent of this change is not accurately determined by currently available time dependant software because of the treatment of torsion creep in these programs.  Bearing load and bearing movement considerations – Both the spine beam and grillage analogy methods of analysis will accurately predict elastic bearing forces if used according to the criteria outline in the proposed AASHTO LRFD Bridge Design Specifications and Commentary and the Analysis Guidelines included in Appendix C. 4. both field experience and time dependent analysis show that the bearing forces will change over time. but software that takes into account axial creep is thought to give conservative results. Methods for addressing bearing design when bearing forces are excessive (i. Because design for the above forces is often optimized. Prestress forces and path location.e. Considering the curvature of long bridges in a spine beam analysis can mitigate excessive design movements at the bearings due to temperature change and possibly eliminate the need for interior Page 133 . either too high or too low) may include. Balancing the center and end span lengths can mitigate bearing problems. 2. They govern the design of the deck and soffit. Simplified methods for assessing these effects have been developed and are included in the recommended LRFD specifications and commentary. Place the bearing group eccentric to the centerline of the superstructure in order to make the individual bearing forces more equal. It is recommended that bearing force capacities be designed to accommodate both initial and long term conditions. Size individual bearings to accommodate the calculated range of bearing forces.

These are usually provided with a small transverse gap to prevent Page 134 . Therefore. It is theoretically more important for the designer to consider the incidental distribution of prestress forces as allowed by some construction specifications. steel or concrete shear keys usually provide lateral restraint. With respect to varying the final distribution of prestress forces across the width of the bridge. The point of the study was to determine if bridge curvature altered the relationship between the relative response of a skewed and non-skewed abutment. but also allows that they be omitted if justified by analysis or tests.  Diaphragms – Current AASHTO LRFD provisions require that diaphragms be used for curved box girder bridges with a radius of less than 800 feet. Two skew cases were studied. The affect of skew on interior supports was not studied nor was the effects of different abutment skew configurations.  Post tensioning sequence – Because the curvature of tendons can increase the transverse bending of the super structure and result in tensions on the inside of the curve and compression on the outside of the curve. Thus it was concluded that existing skew correction factors are applicable to curved concrete box girder bridges analyzed by the spine beam method.  Skew effects – Analytical studies were performed to consider the effect of skew at the abutments on the overall response of the bridge. It is recommended that end diaphragms still be used at all supports. but a relatively even distribution of prestress forces is desirable. One case was where the skew occurred at only one of the abutments and other case where both abutments were skewed but in opposite directions.  Lateral restraint issues – Horizontal curvature may result in force demands in the lateral direction at the supports if lateral restraint is present and is modeled as rigid elements for computer analysis. The second case is the likely orientation of a curved bridge that crosses over an obstruction that is linear in orientation. Analytical grillage analogy and finite element studies performed as part of this project demonstrated that interior diaphragms have a minimal effect on the global response of a curved concrete box girder bridge with a 400 ft radius and 300 ft span lengths.expansion joints. decreasing prestress forces for this web will be overcome by the transverse bending of the bridge that will put the inside web in tension. It is commonly known that skew will affect the shears in the web near the obtuse corner of a skewed abutment support. This method should be used to analyze any curved concrete box girder bridges with large skew angles at the interior supports or abutment skew configurations that vary significantly from those studied. be eliminated. it is recommended that at least one tendon on the inside of the curve be stressed first. Although webs to the inside of the curve are shorter and thus theoretically subject to less dead load and live load bending forces. In both cases it was found that the relationship between the response of a non-skewed support and the skewed support followed the same relationship as for a straight bridge. In the case of bearings. Care should be taken that bearing travel is through the center of movement so that binding of the shear keys does not occur. In all cases a grillage analogy analysis would capture any effect of skew. it is proposed that the requirement that interior diaphragms be included in bridges with a radius less than or equal to 800ft. there does not seem to be any significant advantage in doing this. In these cases lateral restraint should be modeled as the stiffness of the restraining element under consideration. Thus it is thought to be important to model tendons in their correct transverse position for analysis. Prestress shortening may occur along a slightly different orientation. however. Such may be the case for supports consisting of integrally cast abutments or piers.

Commercially available software does not consider torsion creep but should yield generally conservative results and is adequate for design until more sophisticated software is developed.  External post tensioning deviators – The use of pre-cast construction results in less weight and quicker on site assembly and is thus increasing in popularity. It is recommended that reinforcing bar sizes in deviation saddles be limited to #5’s to insure the proper development of this reinforcement. horizontal cambers may be required for segmental construction. This requires a reliance on the time dependant software that is available. A deviation saddle design example. a load factor of 1. Vertical construction cambers can utilize the results from currently available time dependent software. the analysis methods studied are applicable to staged construction analysis as well as cast-in-place on falsework construction. et. Bearings should be capable of travel through the center of movement although normal gaps provided in shear keys will allow for slight variations in movement.7 should be used for the prestress deviator force and capacity reduction () factors should be 0. which is reproduced from the University of Texas report. It can be used for design purposes until improved software is developed. For LRFD design of deviators.binding.The impact of construction methods on the behavior of curved box girder bridges is critical. It is recommended that deviation saddles in tightly curved bridges be continuous across the bottom soffit.9 for direct tension and flexure and 0. This difference should be reflected in a properly conducted spine beam or grillage analysis. This gap is large enough to prevent lateral restraint and for gravity load purposes should be modeled as a lateral force release. Another consideration for curved bridges is that straight segments of tendons cannot rub against the interior of the webs.  Time dependent effects – Because of the interaction between bending moment and torsion in curved bridges.  Thermal effects . However.Thermal movement and prestress shortening will result in movements in different directions at the expansion joints in curved structures. consideration of time dependent effects is important.al. The same parameters can be used to select the most appropriate analysis method except that time-dependant analysis should be used. Deviation blocks or saddles for external prestress tendons in curved pre-cast concrete box girder bridges may be designed in the same manor as for straight bridges using strut-and-tie methods or as recommended by an experimental study at the University of Texas (Beaupre. If necessary. rigorous threedimensional analysis to determine the time dependent effects of torsion is not present in commercially available software.. A curved concrete box girder bridge with relatively tall piers in California that was constructed by the segmental cantilever method required a horizontal camber of approximately three inches at the pier. 1988).  Construction methods . Fortunately. torsion creep is expected to mitigate the effect that has been observed at the bearings. In other words the pier had to be constructed 3” out of plumb. The key to properly considering lateral restraint is to accurately model the actual condition and to use either spine beam or grillage analogy analysis. However. is included in Appendix B. In the case of curved bridges. However the bending of tall columns and twisting of the superstructure had to be approximated using elastic three-dimensional spine beam analysis. In fact many design engineers interviewed claim to have had good results from twodimensional time dependent analysis. the designer should include extra deviation blocks or saddles to prevent this from happening. and thus this software will tend to yield conservative results for bearing force redistribution. Page 135 .85 for shear.

Page 136 . Friction loss/wobble . It is necessary to explicitly consider the difference in tendon length in individual webs and thus prestress tendons should be modeled in their actual transverse location in a three-dimensional spine beam or grillage analogy analysis.The current formulae for determining prestress losses due to friction and wobble are applicable to curved bridges if the three dimensional effects of angle change and tendon length are considered.It is recommended that webs and flanges be designed based on structural and constructability considerations. Friction losses should be based on a tendon curved in space when a curved bridge is being designed using twodimensional analysis techniques. No minimum thickness requirements are recommended by this study.  Web and flange thickness limits .

E. Texas. Stiffness Method for Curved Box Girders at Initial Stress. 1979.. A Model Analysis of Curved Box-Beam Highway Bridges. 9. Bazant. H. and Shafer. No.. Model Studies of Curved Boxgirder Bridges. I. Vol. and Kreger. Eng. F. UC/SESM 73-5. Powell. Washington. An Approximate Method for the Analysis of Box Girder Bridges that are Curved in Plan. Association of Bridges and Structural Engineering.. American Association of State Highway and Transportation Officials. Calif. 16th Edition with Interims. Center for Transportation Research. J. 1989. Breen. W. Mech. California Univ. D. 1999. D. R. Journal of Construction Engineering and Management. Int. 1996. 121. No. of Civil Engineering. 1988.. 1362-1369 American Association of State Highway and Transportation Officials (AASHTO).1974.. and El Nimeiri.. Vol. pp 1–21. AASHTO. W.. Rep. P. J... Klaiber. Guide Specifications for Horizontally Curved Highway Bridges.1973. L. D. American Association of State Highway and Transportation Officials (AASHTO). 207–222. Z. . 2004.C. Beaupre. Dept.C.E. 100(10).C. REFERENCES: Al-Rifaie. and Roll. American Association of State Highway and Transportation Officials. M. Research Report 365-2. . No. 618-635 Aslam. 1971. Concrete Bridge Design and Construction in the United Kingdom. Guide Specifications for Design and Construction of Segmental Concrete Bridges. University of Texas at Austin. Washington. Washington. AASHTO. . Model Studies of Multicell Curved Box-Girder Bridge. K. Austin. J. 618-635 ASCE Committee on Construction Equipment and Techniques. 2071–2090. American Association of State Highway and Transportation Officials. ASCE Committee on Construction Equipment and Techniques. R. pp.. 2003. and Evans.. pp 2861–2878. Journal of Structural Engineering. Int. W. Div. Page 137 .. D.. N. W. Div. Struct. AASHTO. 115. 2nd Edition with Interims. Struct. E. G. and Godden. Diaphragm Effectiveness in Prestressed-Concrete Girder Bridges. Aneja. Deviation Saddle Behavior and Design for Externally Post-Tensioned Bridges. Vol. Washington. 2003... Association for Bridge and Structural Engineering (IABSE). and Godden. No.7. pp. R. M.. 115. American Association of State Highway and Transportation Officials. Berkeley. G..C. 1989.. Abendroth. D. Guide Specifications for Horizontally Curved Steel Girder Highway Bridges Bridges. Journal of Construction Engineering and Management.1975. Concrete Bridge Design and Construction in the United Kingdom. Div. 1995. 4. W.C. AASHTO LRFD Bridge Design Specifications. Aslam.J. pp. 101(3)...C. Proc. Washington.. Standard Specifications for Highway Bridges. 97(12). F. M. 4. J.. M. M.. 2nd Edition with Interims.

F. and Mitchell. J. 116(9). on Computer-Aided Analysis and Design of Concrete Structures.. 25. 1213–1223.. 5. L. Proc. 2410–2418. . 1971. D... . . N. S. and Martins. Negative Shear Lag in Cantilever Box Girder with Constant Depth. and Distorsion of Prestressed Concrete Box Beams of Deformable Cross Section: A Comparison of Experimental and Theoretical Results.Branco.. 789-810. 1986 Buragohain. 20–35. 97(10). and Al-Rifaie. 75(Part 2). 1998. 1980. Eng. 1986. K. L. Vol. Berkeley. 1988. Transportation Research Record 1180. F. Analysis of Cantilever Decks of Thin-Walled Box Girder Bridges... 1996. Rexdale.. R. Caltrans.. Canadian Prestressed Concrete Institute. Canada Chang. October 1998. London. Collins. Choudhury. S. 113(1). Struct.. T. and Agrawal. S. Bending. 2481-2501.. David Evans Associates. Shear and Torsion Design of Prestressed and NonPrestressed Concrete Beams. 72-86. 799–819. ASCE. California Danesi.” CSI. and Edwards.. . N. Inc. Computers and Structures. J. A. Bridge Memo to Designers Manual. Ont. U. 73(Part 1)... Transportation Research Board. SAP2000 – Integrated Finite Element Analysis and Design of Structures. Sacramento. Z. K. Proceedings Institution of Civil Engineers. Eng..1990.. R. and Edwards.1973. Design of Highway Bridges. and Gang. 1988 (Updated 2000). Proc. California Canadian Standards Association (CSA).. California Department of Transportation. Design of Curved Post-Tensioned Bridges for Lateral Prestress Forces. Washington. Struct. D. Design Handbook.K.... A.1975. Analysis of Curved Box Girder Bridges. Inc. 1998. Journal of the Prestressed Concrete Institute. California. Chu. Berkeley. A Computer Program for Analysis and Design of MultiCell Box Girder Bridges. Struct. 2004. CAN/CSA-S6-88. California. Conf. J. C. London. Chang.. The Behavior up-to-Failure of Prestressed Concrete Box Beams of Deformable Cross Section Subjected to Eccentric Loads.. and Zheng. .. Analysis of Horizontally Curved Box-Girder Bridges. 49-75 Evans. 323–352. and Pinjarkar... 1982. H. and Scordelis. J. Danesi. pp 32-100.. Eng. Bridge Design System (BDS). Structural Analysis and Response of Curved Prestressed Concrete Box Girder Bridges. “SAP2000 – Integrated Finite Element Analysis and Design of Structures. Civ. Z. D. R. C. and Huggins.1984.. W. G. 1983. No. F. Memo 11-31. ECC. 1964.. An Experimental and Theoretical Investigation of the Behavior of Box Girders Curved in Plan. 99(5). Computers and Structures. Cazaly. Swansea.. Journal of the Structural Division. D. Proceedings Institution of Civil Engineers. F. A. M. Roseville. Temperature Distribution in Curved Concrete Box Girder Bridges. Inst. Div. Int.. P. T.1987. Torsion. CSI. J. D. Page 138 . 59(2). D. B. October 1998 Cordtz. A. Pineridge.

Heins. Vol. Eng. P. P. No.. Portland. Lopez. A. Div. London. 1989. Diaphragms for Curved Box Beam Bridges. and Roll. P.. Menn.. No. T. Finite Element Analysis of Curved Box Girder Bridges. J. Shear Lag Analysis and Effective Width of Curved Girder Bridges. UC SESM 70-22. T. . 5. Kilford.. C. A.. C... 93(6). C. Soin. Jounal of the Structural Division. C.. Journal of the Structural Division. 2005. Analysis Criteria for Curved Bridges. Y. ASCE. ACI Structural Journal. 7. Hsu. Vol. and Turkstra. 2001. 1977.. 2459-2480 Nakai. B. U. Conference on Development of Bridge Design and Construction. . 101(10). 264–286. and Quiogue... 1976. J.1971. P. M. 1971. C. P. Lim. C.. 97.. and Bell.. A. M. and Tang. LRFD Approach. Journal Structural Engineering. Usuki... K. 831–843. C. . 87–92. September 1. and Scordelis. A. 111(1). 2004. C. Okeil. Basel. J. Design of Concrete Bridges with Multiple Box Cells due to Torsion Using Softened Truss Model. T. 1974. 1996. Masroor. Bonakdarpour.1972. Berkeley. 2161-2178. 1970. No. Hasebe. Institution of Mechanical Engineers. ISBN 37643-2414-7. C. pp. 9. 1993. S.LFD vs. R. Torsional Stiffness of Multicellular Box Sections. Lockwood. 247-253. pp 1097-1108. Torsional Analysis and Design of Curved Bridges with Single Columns . 10. Birkhauser Verlag.K. 207(4).Fam. 5. M. Warping Stresses in Curved Box Girder Bridges: Case Studies. J. and Heins. ACI Structural Journal. and Yang. K. Fu. pp 1419-1427. Paper presented at the Western Bridge Engineers Conference. 1975. Mech.. C. I. and Aparico.. Developments in Bridge Design and Construction. Proceedings.. R. ASCE. C. Structural Engineering and Structural Mechanics Report No. Nonlinear Behavior of Curved Prestressed Box-Girder Bridges. Ayman M. 1997. Ibrahim. No. H.. C.. C. 1971. Page 139 . ASCE.. and Horie. Goodall... ACI Shear and Torsion Provisions for Prestressed Hollow Girders. Maisal. 1990. December 1970 Meyer. B. Analysis of Curved Folded Plate Structures. Vol. 94.. C. Journal of the Structural Division. 696-702. Curved Girder Bridges. Methods for Analysis and Design of Concrete Box Beam with Side Cantilevers. Multi-Cell Curved Girder Model Studies. and Heins. Cement and Concrete Association. 6. Vol. Proceedings. Prestressed Concrete Bridges. K. Shear Flow in Multicell Sections. Fu.6. Meyer. ASCE Journal of Engineering Mechanics. 127(1). Grant. J. Analysis and Design of Curved Box Girder Bridges.. J. D. C. Y. Vol. and El-Tawil. 2004.. 98(4).. T. London. London. 102.1985. L. T. Sherif. pp 787-799... 103.. Struct.C. A. and Moffatt... OR. F. Torsional Analysis for Prestressed Concrete Multiple Cell Box.. Model Study of Horizontally Curved Box Girder.. Chapter 7. IABSE Periodica.. Journal of Bridge Engineering. 132(1). 13-28. J. C. University of California. Oleinik.. Zurich. No..

Proceedings Institution of Civil Engineers. Experimental Evaluation of ACI and AASHTO-LRFD Design Provisions for Combined Shear and Torsion. Design and Construction of Model Prestressed Concrete Bifurcated Box-Girder Bridges. Proceedings Institution of Civil Engineers. Journal of the Structural Division. 1985. Vol. No. 1998.. Precast Concrete Institute. Rahal. 1996. 100. Journal Structural Division. Journal of Science and Technology.. L.. Proceedings Institution Civil Engineers.. No.. R. 93. 2003. 1972. and Collins. and Scheef. 1996. P. 227-235. 3.. 1982. 1996. PCI Journal. Rasmussen. P. Concrete Box Girder Bridges. 1985. D. M. Reilly. The Cause of Cracking in Post-Tensioned Concrete Box Girder Bridge and Retrofit Procedures.. London. 2(79). and Larsen.. ACI Structural Journal... K. 128(2). University of Queensland. Stiffness Analysis of Grids Including Warping. M. Brisbane. No. 206-217. 177-187. VSL International Ltd. A. 8. J. Australia. W. and Marti. ACI Structural Journal. Illinois Rabizadeh.. H.. 7. Pinkney.. P. Elfgren. ASCE. R. Large Scale Experimental Investigation of Deformable RC Box Sections.W. Downviview. M. Dynamic Analysis of Curved Box-Girder Bridges.. pp 658-666. Bern. Rasmussen. Journal Structural Division.... and Shore. Rahal. Ontario. Perry. and Waldron. M. 98. London. P. Plastic Behavior of Deformable Reinfoced Concrete Box Sections Under Eccentric Load. 2(79). Vol.. ISBN 3 85748 031 9. Switzerland. and Pinkney. 1999.. G. W. P. 2003. OHBDC.. H.. Analysis and Shape Optimization of Variable Thickness Box Girder Bridges in Curved Platform. Electronic Journal of Structural Engineering International. L. March-April 1985. and Baker. G. Stress Resultant Plasticity Model for the Analysis of Reinforced Concrete Shell and Box Structures. N.. K. Jr. Queensland. Elastic Analysis and Experimental Behavior to Collapse of 1:12 Scale Prestressed Concrete Bifurcated Bridges. 1525-1542 Page 140 . 1899-1912. 110(9). Scordelis. Podolny. 1991. Ontario Highway Bridge Design Code.. 125(3). Ozakea. O.. C. L. pp 277-282. Rasmussen.. Rahai. Simple Model for Predicting Torsional Strength of Reinforced and Prestressed Concrete Sections. 1975. K. Vol. P. No. M. PhD Thesis. 454-481. H. 2nd Edition. Department of Civil Engineering. J. L. Vol. S. 439-454. 8(32). and Collins. Journal of the Structural Division. pp. A. N. London.M. 6. 1983 (Updated 1998). ASCE.. Perry. Waldron. and Tavsi. N. pp. Australia. 3(2003). Chicago. 1985. Ultimate Strength of Curved RC Box Girder Bridge. 103.. J. London. J. 1977. 1511-1523 Rogowsky..Ontario Ministry of Transportation and Communication (OMTC).. S. Nonlinear analysis of Box-Girder Bridges Under Thernal Loads. Detailing for Post-Tensioning. and Baker.. S. J. G. Switzerland Schlaich.

. 2002.. Literature Review in Analysis of Curved BoxGirder Bridges. State-of-the-Art in Curved Box-Girder Bridges. F. and Collins. R. J. Atlanta. 159-167.A. pp 219-231. 1991. K. 867–874. ACI Journal.. Czech Republic. Hida. 2.D. K. and Edwards. ASCE. London.. Final Report to Caltrans for Contract Number 59A0148. and R. Y. California Sennah. B. . Zokaie. J.T. 1986. Proceedings Vol.. D. 1967.. Institution of Civil Engineers. 2003. B.H. T. Final Report to NCHRP 12-26 (1). Proc. M. Proceedings. S. K.. Distribution of Wheel Loads on Highway Bridges. presented at the May 8-12. J. Song. T. 6th Conf. Zhang. T. Analysis of Concrete Box Girders Before and After Cracking. 1991.. Influence of Prestressing in Curved Members..E. 7(2). Breen.. 83. Chai. 1997..1989. November 1997 Vecchio. ASCE. A. The Modified Compression Field Theory for Reinforced Concrete Elements Subjected to Shear. D. and Kennedy. W. N. Phase 3. Analysis of Curved Twin Box Girder Bridge. on Computing in Civil Engineering. Concrete International. San Diego. Report TK21. Tung. Imbsen. A.Seible. Van Landuyt. Sennah. J. 2001.. Dameron. Part 2. Trikha. 1993. 2001. C.. Theses. ASCE National Meeting of Structural Engineering held at Seattle. Osterkamp. StD&A. 2001. and B. Imbsen..H. D. 59. The Effect of Duct Arrangement on Breakout of Internal PostTensioning Tendons in Horizontally Curved Concrete Box Girder Webs.A. American Concrete Institute. F. Three-Dimensional Analysis of Curved P. No. Behavior Study of Curved Box Bridges. D. 453-462. Structural Evaluation of the 405-55 HOV Connector and the Curved Girder Cracking/Spalling Problems. and Fam. 1991 Page 141 . 6(3). Journal of Structural Division. University of Texas at Austin Van Landuyt. F. 743-761. Final Report to NCHRP 12-26 (2).. Tendon Breakout Failures in Bridges. and Kennedy. 134-143. MI. 1967. Farmington Hills. C.. and J.. Washington Turkstra. J. July 2001 Strasky. ASCE. Distribution of Wheel Loads on Highway Bridges. J. Live Load Distribution in Multi-Cell Box-Girder Bridges and its Comparison with the AASHTO LRFD Bridge Design Specifications. M. M.H. Journal of Bridge Engineering. and S. 104(3)... Hansen. and R. R. L. Betonve Mosty. Journal of Bridge Engineering. 1972. 1978..A.E. December 1993 Zokaie. Prauge. Mish.. Box Girder Bridges. D. and Huang. P.

Sign up to vote on this title
UsefulNot useful