Precast Balanced Cantilever Bridge
Precast Balanced Cantilever Bridge
Precast Balanced Cantilever Bridge
@Seismicisolation
Design Example
Prepared by
Teddy S. Theryo, PE
Major Bridge Service Center
Prepared for
American Segmental Bridge Institute
January, 2005
@Seismicisolation
@Seismicisolation
NOTICE OF DISCLAIMER
This document is a draft effort at developing a design example of a precast balanced cantilever
bridge design based on the third edition of the AASHTO LRFD Bridge Design Specifications,
2004. In its current form, it is not intended to represent a definitive reference for the design of
either a segmental bridge or for the application of the LRFD Specifications to segmental bridge
design. Additionally, the design steps shown in this example problem do not represent all of the
steps that are required for the complete design of a segmental box girder bridge.
@Seismicisolation
@Seismicisolation
TABLE OF CONTENTS
1. Introduction ..................................................................................................................1-1
2. Design Criteria ............................................................................................................2-1
3. Span Configuration and Typical Sections.......................................................3-1
4. Erection Scheme ........................................................................................................4-1
5. Deck Design
5.1 Introduction...............................................................................................................5-1
5.2 Design Approach......................................................................................................5-1
5.3 Live Load Analysis ...................................................................................................5-4
5.4 Post-Tensioning Tendon Layout ............................................................................5-15
5.5 Summary of Design Forces ...................................................................................5-17
5.6 Service Limit State Design .....................................................................................5-17
5.7 Ultimate Flexural Strength Check ..........................................................................5-20
5.8 Ultimate Shear Strength Check .............................................................................5-20
6. Longitudinal Design
6.1 Design Methodology ................................................................................................6-1
6.2 Tendon Layout/Envelope .........................................................................................6-2
6.3 LRFD Live Load .......................................................................................................6-7
6.4 Shear Lag Effect ....................................................................................................6-14
6.5 Temperature Load..................................................................................................6-19
6.6 Time Dependent Effect ..........................................................................................6-24
6.7 Secondary Forces ..................................................................................................6-27
6.8 Summary of Design Forces ...................................................................................6-31
6.9 Service Limit State Design .....................................................................................6-39
6.10 Principal Tension Stress Check .............................................................................6-52
6.11 Flexural Strength Check.........................................................................................6-56
6.12 Shear and Torsion Design .....................................................................................6-59
7. Construction Stage Analysis
7.1 Stability during Construction ....................................................................................7-1
7.2 Erection Tendons ...................................................................................................7-10
8. Detailing
8.1 Combined Transverse Bending and Longitudinal Design .......................................8-1
8.2 Shear Key Design ....................................................................................................8-1
9. Discussion and Recommendations for Improvement
9.1 Discussion ................................................................................................................9-1
9.2 Longitudinal Design..................................................................................................9-1
9.3 Transverse Design ...................................................................................................9-2
10. Acknowledgment .....................................................................................................10-1
11. Reference .................................................................................................................11-1
1. INTRODUCTION
The AASHTO-PCI-ASBI Joint Committee was formed in Chicago, Illinois in October, 1994. The
main goal of the committee was to develop a set of standard box sections for precast segmental
grade separation bridges which would cover bridges of short to medium span ranges
(approximately 200’-0” maximum span). The present practice in the industry shows that only
sufficiently large projects can be competitively built using the precast segmental erection method
due to the high cost of setting up a casting yard which is generally used only for one particular
project. However, by using standardized cross sections, it is expected that precasters in the
concrete industry could afford to build their own casting cells due to potential for repetitive work
from contractors. In addition, the standard cross sections would be useful for structural
engineers as an initial section for conceptual design and preliminary design stages.
Development of a family of standardized segmental box girder sections in metric units was
completed and published by PCI / ASBI in 1998. The next step of the Committee’s work is to
evaluate the proposed standard sections through the creation of design examples. Three
separate design examples were created: span-by-span erection with all external tendons, span-
by-span erection with both external and internal tendons, and balanced cantilever erection. In
additional to design issues, detailing will be discussed. The design examples will be done in
accordance with the current AASHTO LRFD Bridge Design Specifications, Third Edition, 2004.
This design example has been extensively used over the years in the annual ASBI “Design and
Construction of Segmental Concrete Bridges” Seminar since its publication in 1996.
The following design report will cover only precast segmental balanced cantilever construction.
The design example is a five-span precast segmental bridge with three 200’-0” interior spans and
two 150’-0” end spans. The bridge will be supported on bearings, all of which are sliding
bearings except for fixed bearings at Pier 4. The width of the bridge deck is 43’-0” which will
accommodate two lanes of traffic plus inside and outside shoulders of an interstate ramp. The
Type 2700-2 AASHTO-PCI-ASBI Standard Section was selected for this design example. The
section depth is 9’-0” with a maximum span-to-depth ratio of 22. This report will also make a brief
comparative study between AASHTO LRFD and LFD Design Specifications Load Combinations,
including shear design.
The longitudinal analysis of the bridge will be performed using the Proprietary TANGO Program
which enables the effects of stage-by-stage construction and time dependent analysis to be
considered in the design. The transverse design will be accomplished with the aid of Proprietary
GT-STRUDL and BDAC Programs.
2. DESIGN CRITERIA
B. Design Loadings:
1. Load Modifier:
A load modifier of 1.0 will be used for all limit states based on redundant members with the
possibility of non-ductile components and connections, assuming an operational importance
factor of 1.0 for all components.
2. Dead Load:
Unit Weight of Reinforced Concrete (DC): 0.150 KCF (23.5 KN/m3)
Unit Weight of Post-Tensioned Concrete (DC): 0.155 KCF (24.3 KN/m3)
Wearing Surface (DW): 0.015 KSF (0.72 KN/m2)
Traffic Barriers (DC): 0.421 KLF (6.14 KN/m each)
Weight of Blisters (DC): 1 KIP each (4.4 KN each)
3. Live Load:
Vehicle: HL-93 (3 design lanes) using multiple presence factors and dynamic load allowance,
as appropriate.
4. Wind Loads:
Design in accordance with LRFD Article 3.8.
5. Thermal Forces:
Seasonal Variation:
Mean Temperature: 70o F (21oC)
Thermal Coefficient: 6.5 x 10-6 o F (10.8 x 10-6 per oC)
Temperature Rise: 30o F (17o C)
Temperature Fall: 45o F (25o C)
Differential Temperature:
Longitudinal:
Non-linear temperature gradient as per LRFD Article 3.12.3 using a plain concrete
surface for Solar Radiation Zone 3.
Transverse:
Reversible linear gradient of 10o F (6o C) between inside and outside of box girders.
7. Earthquake:
Seismic Zone 1
Acceleration Coefficient: 0.06
Soil Type II
8. Construction Loads:
Construction loads are in accordance with LRFD Article 5.14.2.3. using the appropriate
construction load combinations and allowable stresses. Load factor for temperature gradient
during construction γTG = 0.0.
C. MATERIALS:
1. Concrete:
28 day Cylinder Compressive Strength: 6.0 KSI (42 Mpa)
Modulus of Elasticity: 4933 KSI (34,000 Mpa)
Allowable Stresses: As per LRFD Article 5.9.4.
Superstructure concrete cover for main
reinforcing, plastic (PE) ducts, and hardware:
Top riding surface 2 Inches (50 mm)
Exterior and interior 2 Inches (50 mm)
Concrete cover to plastic ducts shall not be less than one-half the diameter of the duct.
2. Reinforcing Steel:
Yield Strength: 60 KSI (400 Mpa)
Modulus of Elasticity: 29,000 KSI (200,000 Mpa)
3. Prestressing Steel:
Strand tendons shall consist of low-relaxation steel.
Material Properties:
Ultimate Tensile Strength (fpu): 270 KSI (1860 Mpa)
Yield Strength (fpy): 243 KSI (1674 Mpa)
Apparent Modulus of Elasticity: 28,500 KSI (197,000 Mpa)
Friction Coefficient: 0.23 per RAD
Wobble Coefficient: 0.00020 per ft (0.00066 per m)
Anchor Set: 3/8 “ (10 mm)
Allowable Stresses:
Jacking Force: 0.80 fpu
At anchorages After Anchoring 0.70 fpu
At other locations After Anchoring 0.74 fpu
At Service Limit State After Losses 0.80 fpy
Bar tendons shall consist of high strength threaded bars.
Material Properties:
Ultimate Tensile Strength (fpu): 150 KSI (1035 Mpa)
Yield Strength (fpy): 120 KSI (828 Mpa)
Modulus of Elasticity: 30,000 KSI (207,000 Mpa)
Friction Coefficient: 0.30 per RAD
Wobble Coefficient: 0.00020 per ft (0.00066 per m)
Anchor Set: 0.0625 inches (1.6 mm)
Allowable Stresses:
Permanent Bars:
Jacking Force: 0.75 fpu
At Anchorages After Anchoring: 0.66 fpu
At Service Limit State After Losses: 0.80 fpy
Temporary Bars for Reuse:
D. Design Method:
All applicable limit states (Strength, Extreme Event, Service, and Fatigue) will be satisfied in
accordance with the LRFD Specifications.
The structure is a five-span bridge with span configuration of 150’, 200’, 200’, 200’, 150’,
producing a total length of 900 feet. The bridge carries two 12’-0” lanes of traffic in one direction
with a left shoulder width of 6’-0” and a right shoulder width of 10’-0”. Expansion bearings are
placed at all piers except Pier 4 which is fixed.
The typical section selected is the AASHTO-PCI-ASBI Segmental Box Girder Standard Type
2700-2, a single-cell concrete box girder with 43’-0” wide deck and 9’-0” in depth. Cantilevered
overhangs are 10’-4.5” each. Minimum top slab thickness is 9”. The thickness of the bottom slab
is 18” for three segments on both sides of each pier and 9” thick elsewhere. The thickness of the
webs is 16”, which are sloped at 2.5:1.
The top slab can accommodate 12 tendons in each half of the box girder, for a total of 24
tendons in the top slab. The bottom slab can accommodate 6 tendons in each half of the box, for
a total of 12 tendons in the bottom slab. Additional tendons may still be accommodated either in
the top or bottom slab.
Ic
ρ=
Acy tyb
where,
Ic = Moment of inertia of the section
Ac = Area of the section
yt = Distance from the top fiber to the center of gravity of the section
yb = Distance from the bottom fiber to the center of gravity of the section
The efficiency of the cross-section, ρ, is 0.6 which is considered to be high. For the sake of
comparison, the flat slab is the most inefficient section with a ρ value of 0.33.
This design example utilizes a 12’-0” typical segment length, resulting in a maximum segment
weight of 72.5 tons for the thin bottom slab segment and 80 tons for the thick bottom slab
segment.
4. ERECTION SCHEME
The structure is erected using the precast balanced cantilever method of construction, where
individual segments are placed successively on alternating sides of the cantilever. A segment is
attached at either end of the cantilever by use of temporary post-tensioning bars after epoxy has
been applied to the interface with the previously erected segment. In this example, temporary
post-tensioning bars will be left in the segments and grouted afterward. Temporary post-
tensioned bars may also be re-used. Cantilever tendons are then stressed, and the process is
repeated for the entire cantilever.
2. 180 Erect span 1 segments on falsework, cast CIP closure, and stress
span top and bottom tendons.
4. 200 Cast span 2 closure, and stress span top and bottom tendons.
6. 220 Cast span 3 closure, and stress span top and bottom tendons.
8. 240 Cast span 4 closure, and stress span top and bottom tendons.
9. 250 Erect span 5 segments on falsework, cast CIP closure, and stress
span top and bottom tendons
10. 300 Cast barriers, Install expansion joints, and place overlay if applicable
12. 500 Total forces and deformations after creep and shrinkage at day 500
13. 1000 Total forces and deformations after creep and shrinkage at day 1000
14. 2000 Total forces and deformations after creep and shrinkage at day 2000
15. 4000 Total forces and deformations after creep and shrinkage at day 4000
16. 10000 Total forces and deformations after creep and shrinkage at day 10000
5.1 Introduction
The top deck of a box girder is subjected to complex external forces, static and dynamic loads,
thermal gradients, and creep and shrinkage effects. Proper consideration should be given to
these effects to prevent cracking and deterioration. De-icing chemicals and freeze-thaw action
should also be considered in design to counteract degradation.
Deck replacement is not only costly, but results in inconvenience to the traveling public. For
segmental bridge superstructures, deck replacement is not practical and almost impossible to do
without closing the entire bridge. Therefore, when designing decks for segmental bridges, it is
always good strategy to be conservative and allow for reserved capacity.
Studies have shown that transverse post-tensioning of top decks improves long-term deck
durability and results in low life cycle cost (See Reference 12). It is recommended that for all
post-tensioned box girders the top deck be transversely post-tensioned, even for short
overhangs. For bridges not subjected to freeze-thaw action and de-icing chemicals, at least the
deck should be partially prestressed. The top deck should be designed using elastic methods
and then checked for ultimate limit states, not the other way around.
In general it is standard practice to select a minimum top deck thickness of eight inches, although
AASHTO-PCI-ASBI Standard Sections Committee recommends a minimum deck thickness of
nine inches.
To correctly represent the final system of the box girder, one would need to do a three
dimensional analysis and incorporate all loads the box is subjected to along with proper boundary
conditions. Due to complexity of this type of analysis, in particular the application of prestressing
to three dimensional systems, this is seldom done. In lieu of this complex analysis, it is common
practice to model the box as a 2-D (two dimensional) plane frame of unit length, as shown in
Figure 5.2-1. If the thicknesses of the web and bottom slab vary along the length of the bridge,
several 2-D frames may have to be analyzed in order to obtain a more representative
interpretation of these varying cross-sectional properties. The 2-D frame model allows for load
distribution to the webs and slab members relative to their stiffness.
A typical 2-D frame model is assumed to be supported at the lower end of the webs as shown in
Figure 5.2-1. While it could be argued that different boundary conditions exist for this model, this
simplified assumption produces reasonable results.
The design loads considered in transverse design include, but are not limited to:
Secondary forces of post-tensioning shall be included in ultimate limit state load combinations
with a load factor of 1.0.
In addition to service and strength limit state load combinations, the deck design should be
checked for construction load combinations, such as segment lifting, construction equipment, and
segment stacking (see LRFD Article 5.14.2).
FIG. 5.2-1
When a static concentrated load is applied on a deck, the deck will deflect transversely as well as
longitudinally, similar to a two-way slab. The load distribution becomes more complex when
multiple point loads are applied to the deck, such as a truck load. Since the structural model is
simplified to a 2-D frame model, as stated in Section 5.2, it is important to obtain the resulting 3-D
forces to the 2-D model.
Commonly, there are two ways of handling live load distributions in the transverse direction:
1. In the past, influence surfaces from Pucher or Homberg Charts have been extensively used in
box girder transverse design. These charts are based on elastic theory of plates
(homogeneous and isotropic). Some charts are valid for constant depth plate thickness and
some for variable depth plate thickness with a parabolic soffit. Depending on the boundary
conditions of the selected plate, the dimensionless charts provide bending moments per unit
length at the fixed end and mid span only. The Fixed End Moments (FEM) are then applied
as external forces to the 2D frame. The bending moments between supports are
approximated by interpolation. The method has limitations for haunched deck slabs, regarding
the support depth over mid span depth ratio. This method is approximate and can be useful
for preliminary design.
2. A more accurate method is based on a partial 3-D (three dimensional) finite element model of
the box girder. The term “partial” implies that the entire bridge superstructure need not be
modeled; rather it should be interpreted as a partial length of the box that will be long enough
to include three dimensional effects. From this model, influence lines can be generated at any
section of interest. The influence lines should be generated using a line load consisting of front
and rear wheels of a design truck. Since general finite element programs are readily available
presently, it is recommended that this method be used for final design.
It should be noted that theoretically, a continuous vehicle barrier could be incorporated into
this model to further distribute live load longitudinally. However, due to discontinuities of the
barrier and uncertain future quality, this edge stiffening effect is neglected and not
recommended.
In this design example, the second method was implemented for analysis. Keep in mind the live
load configuration should be strategically placed in order to produce the worst condition (see
Figures 5.3-1 to 5.3-3). Listed below are some common points where stresses are checked:
• Maximum positive and negative bending moments at the center line between two webs
• Maximum negative bending moment in the top deck at the interior face of the webs
• Maximum negative and positive bending moments in the webs and bottom slab
• Maximum negative moment in the deck overhang where the taper begins
See Figures 5.3-4 to 5.3-8 for influence lines corresponding to these locations.
In the AASHTO Standard Specifications (LFD), only the effect of a design truck (or tandem) is to
be considered for transverse design. However, the current 3rd Edition of LRFD requires the
design truck and lane load to be combined to achieve maximum effects. In combination with this,
if one truck controls, a multi-lane increase of 1.2 is to be applied. Due to these new
requirements, LRFD will produce more conservative results when compared to the Standard
Specification. Although impact and multi-lane factors have not been included, a live load moment
envelope is given in Figure 5.3-9 to show the difference in codes.
In recent AASHTO T-5 and T-10 Committee meetings held in Orlando in June, 2004, revisions
have been proposed for transverse deck design. In particular, the elimination of multi-lane
factors, Service Limit State III, as well as lane load elimination have been proposed. Hence, only
the design truck (or tandem) will be used to calculate maximum effects. Service Limit State III
with a factor of 0.8 for live load will no longer be used for transverse deck design. Rather, it will
be eliminated and Service Limit State I with a live load factor of 1.0 for both tension and
compression will be checked. These revisions will produce results similar to that of the Standard
Specifications and also have positive impacts on ultimate limit states.
For this design example, all limit states have been checked incorporating the proposed T-5 and
T-10 Committee revisions for transverse deck design.
Please note that although the above deck design revisions have been proposed and approved by
AASHTO T-5 and T-10 committees, they cannot be adopted until they are officially published in
the AASHTO LRFD Bridge Design Specifications, 3rd Edition Interim Specifications.
FIG. 5.3-1
FIG. 5.3-2
FIG. 5.3-3
-20.0
-10.0
-5.0
Moment (kip-ft/ft)
0.0
5.0
10.0
-24.0 -20.0 -16.0 -12.0 -8.0 -4.0 0.0 4.0 8.0 12.0 16.0 20.0 24.0
Distance from Centerline Box Grider (ft)
@Seismicisolation
HL-93 Truck Only - Maximum Values Truck Only - Minimum Values Truck & Lane - Maximum Values
@Seismicisolation
HL-93 Truck & Lane - Minimum Values HS25 Only - Maximum Values HS25 Only - Minimum Values
FIG. 5.3-9
Post-tensioning in the transverse direction typically consists of three to four 0.5” or 0.6” diameter
strands per tendon passing through the top slab and anchored at the face of the overhang on
each side of the box girder. These tendons are usually housed in flat ducts due to the thin top
slab. To efficiently utilize the tendon, it should be suitably profiled for maximum structural
efficiency.
A typical tendon is generally anchored at mid-height of the slab at wing tips and then gradually
rises to a level above the neutral axis of the deck over the webs. This helps the tendon resist the
negative moments at the webs. The tendon then gradually drops to a level below the neutral axis
of the top slab near the centerline of the box girder in order to resist the positive bending in that
region. The tendon path used for this example is shown in Figure 5.4-1.
Longitudinally, the tendon spacing is determined using the appropriate service and strength limit
state checks. The maximum spacing of tendons is typically restricted to 4 feet in effort to limit
shear lag effects between anchorages. If maximum tendon spacing is not addressed, zones near
outside edges of the slab may be without effective prestressing.
FIG. 5.4-1
The design forces obtained from the two-dimensional frame analysis and three-dimensional live
load influence lines are combined in a spreadsheet using the LRFD Service Limit State and
Strength Limit State combinations. The maximum tensile and compressive stresses at each
predetermined section in the top slab are summarized and compared to the LRFD allowable
stresses. In this example, the prestressing force is estimated in preliminary hand calculations,
and then analyzed in a 2-D time dependent run using the BDAC program. All other loads are
incorporated into the 2-D model, except live loads. The results are then compiled in a
spreadsheet to check stresses. By varying the prestressing force, the combined stresses of
service limit states are calculated. Using the selected tendon forces per unit length, the size and
spacing of transverse tendons in the segment are determined.
The LRFD Strength Limit States are also tabulated in a spreadsheet and an envelope of
maximum and minimum values is determined for each chosen section. The values in this
moment envelope can then be compared to the calculated bending capacities for each of the
corresponding transverse components.
As stated in Section 5.3, only Service Limit State I will be checked with a live load factor of 1.0 for
tension as well as compression. Also, a linear temperature gradient of 10 degrees Fahrenheit
between interior and exterior surfaces of the box will be used in Service Limit State I. The current
LRFD specification does not specify this loading, leaving it up to the owner or designer to
establish if it should be included on a project-by-project basis. This example is based on a load
factor of 0.5 for transverse temperature gradient when accompanying live load. Also, in addition
to Service Limit State I, LRFD requires a check for service load stresses due to dead load and full
temperature gradient. This limit state can often govern at locations where live load influences are
small.
To show a comparison of the new proposed Service Limit State I verses the current LRFD
Service Limit State III and Standard Specification, a graph of stresses is given in Figure 5.6-1.
Since the box is symmetrical, minimum and maximum stresses for the top of the deck have been
shown on one side and bottom deck stresses on the other. After examination of this figure, it can
be seen that stresses resulting from the Committee T-5 and T-10 proposal closely follow those
from the Standard Specification. The slight difference is due to the 1.33 impact factor from LRFD
compared to 1.3 for the Standard Specification. It can also be seen that the stresses produced
from the current LRFD specification are similar to those produced from a Standard Specification
HS25 loading.
In addition to service limit states under maximum loading, temporary stresses such as those prior
to barrier placement and vehicular traffic should be checked to ensure allowable stresses are not
exceeded during the construction process.
For purposes of the transverse design, Strength Limit State IV is the same as Strength Limit
State I without live load, with 25 percent more self-weight. This loading does not govern in this
example.
For temperature gradient load factors, LRFD Specifications suggest determining a load factor on
a project specific basis, with a recommendation of 0.0 for most instances. Since these loads are
a result of restrained deformations, the loads should disappear if the reinforcement begins to
yield at ultimate. In addition, the Segmental Guide Specifications does not include this
component in ultimate load combinations. For these reasons, the temperature gradient was not
used in the strength limit state combinations.
The LRFD specifications require minimum reinforcement equal to that required to resist 1.2 times
the cracking moment. This requirement governed only for the bottom slab (soffit) design. To
satisfy the minimum steel requirement, the transverse bar spacing in the bottom soffit was
decreased from 12 inches to 8 inches, which represents an increase in reinforcement of 50
percent.
Also under ultimate flexure, the amount of web steel reinforcing required for transverse bending
should be calculated. This should be combined in an appropriate manner with reinforcing
required for longitudinal shear.
Strength I
Traditionally, shear behavior has been ignored in the design of concrete decks for AASHTO
bridges. Box girder decks are similar in this sense, but can often have large construction loads
placed on them. In these special cases, both one-way and two-way action shear should be
investigated.
6. LONGITUDINAL DESIGN
This structure is erected using the precast balanced cantilever method of construction. Due to
changes in the statical system during erection, as cantilevers are made continuous through cast-
in-place closure joints, it is necessary to analyze the structure for time-dependent effects. Time
dependent analysis is a function of the segment casting date, times that the segments are
incorporated into the structure, as well as dates associated with changes in the structural system
throughout the construction process.
Time dependent properties of concrete are established based on environmental humidity and
dimensions of the cross-section, and can be adjusted for concrete composition (e.g. limestone
aggregate), rate of hardening, and ambient temperature. Section properties shall be determined
for each segment considering effects of shear lag in the top and bottom slab.
The above information is entered into time dependent analysis software such as TANGO, among
others. A stage-by-stage analysis is performed using an assumed post-tensioning layout while
carefully modeling appropriate boundary conditions for each step of the construction process.
After the construction has been modeled, the structure is stepped through time to day 4000 or
day 10000 to allow all time dependent effects to occur. It is also essential in statically
indeterminate structures to sum up all locked-in forces that result from various stages of
structural systems until day 10,000. Additional loads are placed on the structure such as live
load, temperature gradient, and support settlement, as appropriate, and analyzed for initial (at
end of construction) and final conditions at day 10,000.
An approximate tendon layout can be based on preliminary calculations for construction loading
of a typical cantilever. Span continuity tendons can be estimated by preliminary design based on
final structure approximate creep and shrinkage effects using load factor dead and live load
combinations. The assumed layout can then easily be modified during final design to satisfy all
applicable LRFD Limit State Load Combinations.
Preliminary design for this example indicated the need for twelve cantilever tendons and five
bottom continuity tendons per web. Based on previous experience, two four strand continuity
tendons were added in the top slab across the closure pour to control stresses resulting from
temperature gradients. Final design resulted in an increase of one cantilever tendon and one
bottom span continuity tendon at interior spans only.
The tendons used are based on a twelve-strand system using 0.6” (15.24 mm) diameter strands.
Only eleven strands were used for bottom continuity tendons to provide space for 5%
contingency post-tensioning as required for internal tendons. One out of twelve strands will
provide approximately 8% of the contingency post-tensioning if needed. An empty duct was
provided for the cantilever tendons combined with an anchorage on the last segment of the
cantilever in order to allow for contingency post-tensioning. This empty duct should be grouted if
no contingency tendons are required.
Provisions are also made for future post-tensioning by addition of anchorages and deviation
points for external tendons (inside the box section), which can be used for adjustment of
deflections or for other unforeseen conditions. Provisional post-tensioning ducts and anchorages
are covered under Article 5.14.2.3.8 of AASHTO LRFD Bridge Design Specifications.
LRFD live load (HL-93) consists of a single design truck per lane or tandem combined with a
uniformly distributed lane load. For negative moments only, a second truck is added and the total
effect is reduced by 10%. The second truck is required only between points of uniform load
contraflexure, and should leave a space of at least 50 feet (15 meters) between trucks measured
between the rear axle of the leading truck and the front axle of the trailing truck. A fatigue truck
is also specified but was not considered for this example.
A dynamic load allowance (impact) of 33 percent is added to the design truck, but is not required
for design lane load. Multiple presence factors range from 1.2 for a single lane to 0.85 for three
lanes and 0.65 for more than three lanes. This example is based on 3 lanes, and has a multiple
presence factor of 0.85 (the current AASHTO Standard Specifications would dictate an impact of
15% and a multiple presence factor of 0.90).
For comparison purpose, HS20-44 and HS25-44 AASHTO loadings were run in addition to the
HL-93 LRFD loading. After impact and multiple presence factors are included, results for this
example show that live load moments are increased by approximately 30% for negative moment
and approximately 50% for positive moment when compared to HS 20-44 live load. Live load
shears are increased by approximately 40% when compared to HS 20-44 live load. The HS25-
44 loading increases the HS20-44 results by 25%, thus narrowing the difference, but HL-93
results remain slightly higher.
The AASHTO Guide Specifications for Design and Construction of Segmental Concrete Bridges,
First Edition adopted shear lag provisions of DIN 1075 (German Concrete Code) using linear
transition of effective flanges. However, in the second edition, shear lag provision changed to a
step function between span and support regions. In contrast to this change, the AASHTO LRFD
Bridge Design Specifications, Third Edition adopted shear lag provisions similar to DIN 1075, as
shown in Article 4.6.2.6.2. The difference between the two methods is insignificant, but the LRFD
shear lag provision is considered more accurate.
When determining section properties, it is commonly assumed that shear lag applies to moment
of inertia and location of the neutral axis of the section. However, cross-sectional area remains
based on the full cross-section, so as to not overestimate the “P/A” component of post-tensioning
stresses.
Shear lag is a function of the structural system at the time under consideration. If software
permits, section properties can be changed in the construction model to approximate true statical
conditions at all intermediate steps. This additional accuracy may not be warranted for all
designs, but could be evaluated on a case-by-case basis.
The following shear lag effect calculation is in accordance with article 4.6.2.6 of AASHTO LRFD
Bridge Design Specifications, Third Edition 2004.
I. COMPLETED STRUCTURE
END SPAN
INNER SPAN
II . DURING CONSTRUCTION
CANTILEVER.
Both uniform temperature and temperature gradient shall be included in service limit state load
combinations. Temperature gradient may be reduced by 50% if live load is present in service
load combinations. For segmental bridge design only, a special load combination (LRFD
equation 3.4.1-2) for service shall be checked. This load combination has no live load; therefore
100% of the temperature gradient shall be included. In general, this load combination controls for
segmental concrete bridges where live load force effects are small. In this example, such an
area occurs at closure pours in the top of the box. Please note, for uniform temperature use a
load factor of 1.0 when checking stresses, and 1.2 for structural deformations.
Temperature gradient shall not be included in strength limit state load combinations, while
uniform temperature shall be included. Two load factors are assigned to uniform temperature in
strength limit states. A factor of 0.5 shall be used for strength capacity calculations and 1.2 for
structural deformations.
Creep and shrinkage of concrete, including relaxation of prestressing steel are commonly
referred to as time dependent longterm effects. These effects are important factors that demand
consideration in design of segmental bridges. Non-linear time dependent deformations will result
in force redistribution due to changes in statical system during the course of the construction, and
continue through day 10,000 when longterm effects are considered diminished.
The redistribution of sectional forces due to change in statical system and creep effect can be
estimated by Dischinger’s equation.
Where:
φ = Creep coefficient
The above equation can be re-written to obtain M due to creep effects: Mcr = (1-e φ) (MII – MI)
Steel relaxation is the loss of tension in prestressing steel under constant length and temperature
over a period of time. To prevent excessive relaxation loss in segmental bridges, low relaxation
strand shall be used. The low relaxation strands shall meet the ASTM Standard requirement that
relaxation loss after 1000 hours under 70˚ F shall be no more than 2.5% when initially stressed to
70% G.U.T.S. (Guaranteed Ultimate Tensile Strength) and not more than 3.5% when stressed to
80% G.U.T.S.
Although AASHTO LRFD Bridge Design Specifications allow creep and shrinkage effects to be
evaluated using the provisions of CEB-FIP Model Code or ACI 209, for segmental bridge design,
the CEB-FIP Mode Code provisions are commonly used. This design example utilizes the CEB-
FIP Model Code 1990.
Secondary forces are internal forces generated as a result of applied deformations or imposed
loads to statically indeterminate systems.
Listed below are several recognized secondary forces in segmental bridge design:
All of the above secondary forces shall be included in service limit state load combinations
without exception. However, inclusion of different types of secondary forces in strength limit state
load combinations may differ from code to code.
For instance, in the AASHTO LRFD Bridge Design Specifications, Third Edition, 2004, the
secondary forces due to prestressing and erection loads (locked-in forces) are lumped together
as “EL” with a permanent load factor γp equal to 1.0 for all strength limit state load combinations.
On the other hand, in the AASHTO Guide Specification for Design and Construction of
Segmental Concrete Bridges, Second Edition, the erection loads (locked-in forces) are lumped
together with permanent dead loads, receiving a factor higher than 1.0. Under this assumption,
since temporary loads are added during construction and then removed, only the effects due to
permanent load will receive a load factor higher than 1.0.
The combination of prestressing and construction process secondary forces under “EL” as shown
in AASHTO LRFD Bridge Design Specifications, Third Edition serves little merit in segmental
bridge design. The author of this example recommends that the secondary forces due to
prestressing and erection loads be separated and applied in accordance with AASHTO Guide
Specification for Design and Construction of Segmental Concrete Bridges, Second Edition. For
purposes of service limit state combinations, the separation of prestressing secondary forces and
locked-in forces will make no difference in stresses. However, for ultimate limit state load
combinations, a difference will occur. In most segmental software, dead loads are not
distinguished from locked-in forces. Due to many construction stages during the erection
process, it is possible to accumulate large quantities of dead load cases and locked-in force load
cases. Once completing the construction process, backtracking to separate dead load cases
from locked-in load cases creates complex book-keeping, and serves little benefit to end results.
Secondary forces due to temperature gradient are not included in strength limit state load
combinations, while support settlement secondary forces are to be considered on a project
specific basis.
Uniform temperature secondary forces, including creep and shrinkage effects, are included in
strength limit state load combinations with load factor of 0.5.
As mentioned previously, a comparison of service live load forces was conducted for AASHTO
LRFD and the AASHTO Standard Specification. This was done to get an idea of how much
larger forces will be for the HL-93 loading. At maximum locations, the differences in positive and
negative moments were 50% and 30% respectively. The difference in shear was 40%.
Even though these numbers represent large differences, for the span lengths under consideration
live load only constitutes approximately 25% of the total factored load. This occurrence
combined with lower ultimate load factors used by AASHTO LRFD will bring the ultimate limit
states for the two codes very close to one another.
The results of the different load combination envelopes can be observed in Figures 6.8-1 to 6.8-
7. It is interesting to note that the negative bending moments of the three groups only differ by
5%, with the largest value coming from the AASHTO Standard Specification HS25-44 loading.
The positive bending moments of the HL-93 load combination are approximately 7% higher than
the HS20-44 load combination, while the HS25-44 load combination is about 12% higher than the
HS20-44 load combination. The shear forces of the HL-93 load combination are comparable to
the HS20-44 load combination, while the HS25-44 shear force is about 6% higher than the HL-93
and HS20-44 load combinations.
Service limit state design of the superstructure requires a stress check for three load
combinations. These consist of Service Limit State I, Service Limit State III, and a special load
case for segmental bridges. Service Limit State III allows tension to be evaluated using a 0.8 live
load factor, while Service Limit State I checks compression with a 1.0 live load factor. In
combination with these three limit states, a non-linear temperature gradient will be applied. For
Service Limit States I & III, which use maximum live load influence, LRFD recommends a factor
of 0.5 for temperature gradient in lieu of project-specific data. For the special load case applying
to segmental bridges, temperature gradient receives a load factor of 1.0, since live load is not
included. For a description of this load case, see LRFD Equation 3.4.1-2.
It is important to note that although the special load case may not control at locations where large
amounts of post-tensioning are present, it may indeed control at locations where live load effects
are small or at locations outside of the precompressed tensile zone. Such locations for this
example include tension in the top of closure pours and compression in the top of the box at pier
locations. For this example, tendons were added in the top of the box crossing the closure pour
to counteract tension produced by the bottom of the box being warmer than the top.
Results from the service stress load combinations can be referenced in Figures 6.9-1 to 6.9-12.
It can be seen that small amounts of tension exist at nodes 8 and 104 under Service Load Case
III at day 10,000. Due to the conservative boundary conditions assumed while erecting the end
spans, this tension is acceptable.
A principal tensile stress check in shear design is not yet specified by code, but is typically
performed as a method to prevent cracking during service load conditions. Stresses are
calculated using Mohr’s circle to determine principle tension. If the allowable tensile capacity of
the concrete is exceeded, diagonal tension cracks may be anticipated. Typically the maximum
principal tension stress is limited from 3√f’c to 4√f’c (psi). Based on information from AASHTO T-
5 & T-10 Committee meetings held in June, 2004, principle tension stress will be limited to a
value of 3.5√f’c for segmental bridges. It is anticipated that this check will be adopted by
AASHTO LRFD Bridge Design Specifications in the near future. Although it is likely that this
check will only be required at the neutral axis of the web, it is recommended that the top slab and
web interface location be investigated as well. For this example, 3.5√f’c tension will be used as a
maximum allowable value under service loading.
Since principal stress is a function of longitudinal, vertical, and shear stress, it is necessary to
determine concurrent moments for the maximum live load shear. It should be noted that high
principle stresses commonly occur at interior pier locations, and the HL-93 live load moment
corresponding to shear should only use one truck, rather than two as used in calculating negative
moment at interior piers. The live load shall also have a load factor of 0.8 similar to Service III
Limit State or it would be practically impossible to satisfy principal stresses while the extreme
fiber could be in tension.
The maximum principal stresses in this example occurred near the interior piers at the top of the
web for final conditions. From analysis at the critical section, the maximum principle tension
stress was approximately 4.5√f’c; larger than the previously discussed limit. For this particular
example, vertical post-tensioning bars will be used to control the principal tension stress.
Calculations show that (3) 1¼” diameter bars, as shown in Figure 6.10-3, will be needed in each
web to reduce principle tension to an acceptable value. The overstress could also be addressed
by modifying the cross-section (web thickness) or adding more longitudinal compressive stress
(additional strands). The solution presented was deemed acceptable since only a small number
of segments will require vertical post-tensioning. A graph of principle stress prior to addition of
vertical post-tensioned bars can be seen in Figure 6.10-2.
VQ
v=
Ib
where
σx + σy 1
( )
2
f1 = − 4v 2 + σ x − σ y
2 2
v a = fa × ( fa + f)
where
fa = Allowable principal tension
f = Compressive stress at level on web under investigation
Stress (psi)
150
100
50
0
0 150 300 450 600 750 900
Distance Along Bridge (feet)
@Seismicisolation
@Seismicisolation
Principle Stresses (psi) Allowable Tension (3.5 Sq. Roots)
FIGURE 6.10-2
Once service stresses are satisfied in the superstructure, the limit state of flexural strength must
be checked. For most cases with superstructures, Strength Limit State I is the only load
combination that needs to be considered. However, for longer spans where the ratio of dead
load to live load is large, Strength Limit State IV may control. For this example, the magnitudes
of live load force effects are greater than a 25% difference in structural component dead load.
Hence, Strength Limit State IV will not control.
The load factor for support settlement and temperature gradient are not provided by LRFD.
Rather, they are to be determined on a project-specific basis. In lieu of project-specific data,
LRFD recommends using a load factor of 0.0 for temperature gradient. With regard to
temperature gradient, the loads imposed result from restrained deformations and should
disappear if the reinforcement starts to yield at ultimate. Due to this occurrence, temperature
gradient is not considered in strength limit states. Also, support settlements are not considered in
this example.
The LRFD specifications require minimum reinforcement equal to that required to resist 1.2 times
the cracking moment. All sections in the example satisfy this requirement.
In the following pages of this section, example calculations for ultimate flexural capacity are given
for an individual node in the bridge.
A ps f pu 67.704 × 270
c= = = 22.85in
f pu 270
0.85 f c′β 1b + kA ps 0.85 × 6 × 0.75 × 196 + 0.28 × 67.704
dp 102
a = β1b = 0.75 × 22.85 = 17.14in , less than bottom slab thick., mod. comp. block is rectangular
c 22.85
f ps = f pu × (1 − k ) = 270 × (1 − 0.28 ) = 253ksi
dp 102
Find ultimate moment strength:
σ t = 580 psi , compressive stress at top of section due to permanent loads at day 10,000
St = 435740in3
From recent AASHTO T-10 Committee meetings on June 21, 2004 in Orlando, Florida, it was
proposed that for post-tensioned box girder bridges, including segmental bridges, the design
procedure similar to AASHTO Guide Specifications for Design and Construction of Segmental
Concrete Bridges, Article 12.0 may be elected. The current edition of AASHTO LRFD Bridge
Design Specifications, Third Edition uses modified compression field theory for shear and torsion
design. Both shear design methods will be presented in this design example.
Vu ≤ φVn
Vn = Vc + Vs + V p
Where:
Vc = 2 K f c′bw d (lbs)
Where K = 1 + f pc / 2 f c′ ≤ 2.0
bw = effective web width
Av f y d (lbs)
Vs =
s
Where Av = Area of transverse reinforcement within a distance s (in2)
fpc = Compressive stress in concrete after allowance for all prestress losses at the
centroid of cross-section resisting shear (psi)
f’c = Specified concrete strength (psi)
K < 1.0 at any section where stress in the extreme tension fiber due to factored load
Tu ≤ φTn
Tn = 2 Ao At f y / s
Tn p h
Al =
2 Ao f y
Where Tu = Factored torsional moment (in-lb.)
Tn = Nominal torsional resistance (in-lb.)
At = Area of one leg of closed transverse torsion reinforcement within a distance s (in2)
Al = Total additional longitudinal reinforcement required for torsion (in2)
Ao = Area enclosed by shear flow path (in2)
ph = Perimeter of centerline outermost continuous closed transverse reinforcement (in)
be = Minimum effective shear flow web or flange width to resist torsional stresses (in)
The Modified Compression Field Theory (MCFT) was developed by Dr. Michael P. Collins, Dr.
Frank J. Vecchio of University of Toronto and Dr. Denis Mitchell of McGill University in Canada.
The MCFT for shear and torsion design was adopted for the first time by the Ontario Highway
Bridge Design Code in 1991. The 1994 AASHTO LRFD Bridge Design Specifications also
adopted the new method of shear and torsion design in lieu of the traditional ACI empirical
equations. The new method is a simple, unified method which is applicable to both prestressed
and nonprestressed members. Unlike previous empirical methods, MCFT is a rational method
which gives physical significance to the parameters being calculated.
The MCFT is based on variable-angle truss instead of a 45o truss model. Due to this truss
model, the longitudinal reinforcement becomes an important element of shear design. However,
in light of the iterative procedure required in the new design procedure, hand calculation is no
longer practical, and a computer program should be utilized.
In a box girder, the stresses due to shear and torsion will be additive on one side of the web and
will counteract each other on the other side. Therefore, the final transverse web reinforcement
should be based on the summation of reinforcement due to shear and torsion.
Normally, the loading which produces the maximum shear will not be the same loading which
produces the maximum torsion. Therefore, it is conservative to design based on the maximum
shear and maximum torsion. However, it is sufficient to design using the maximum shear with its
associated torsion and the maximum torsion with its associated shear.
For shear design, the following basic relationship must be satisfied at each section:
Vu ≤ φ Vn
where,
Vn = Vc + Vs + V p (LRFD 5.8.3.3-1)
This relationship is similar to the method of shear design prescribed in the AASHTO Standard
Specifications. However, with LRFD, Vc is computed in an entirely different manner. The
equation for Vc is now:
The value of β at a given section must be obtained through an iterative process. The following
two parameters must be computed as part of this process:
Vu − φV p
vu = (LRFD 5.8.2.9-1)
φbv d v
A first trial value of θ is assumed to compute the initial value of εx. Then, knowing v and εx, Table
5.8.3.4.2-1 is used to look up the corresponding values of β and θ. If θ is not within a reasonable
tolerance of the assumed θ, then the current value of θ is used to compute a new εx, and a new
look-up in Table 5.8.3.4.2-1 is performed. When convergence is reached, Vc can be then be
calculated.
Longitudinal Reinforcement
One of the cornerstone principles of modified compression field theory is the recognition that
shear causes tension in longitudinal steel. At each section of the beam not subjected to torsion,
the capacity of the longitudinal reinforcement must be checked for sufficiency. This relationship
is expressed as follows:
⎡M N ⎛V ⎞ ⎤
As f y + A ps f ps ≥ ⎢ u + 0.5 u + ⎜⎜ u − 0.5Vs − V p ⎟⎟ cot θ ⎥ (LRFD 5.8.3.5-1)
⎣ d vφf φc ⎝ φv ⎠ ⎦
Vu − φV p
where: vu =
φbv d v
where:
Step 5: Repeat the calculation from step 2 with the latest θ from step 4 until θ in step 4 matches
close to θ in step 2, then select the new β.
For sections subjected to combined shear and torsion, reference Article 5.8.3.6.2. Strain will
need to be calculated taking into account the combination of these effects. Shear stress,
longitudinal reinforcing, and area of shear reinforcing will also need to be modified.
Vu = 2391kip
φ = 0.90 for shear
Nominal shear resistance:
V n = Vc + V s + V p
or V n = 0.25 f c′bv d v + V p
where:
f c′ = 6ksi , compression strength of concrete
bv = 32in , effective web width
d v = 108 − 6 − 17.1 / 2 = 93.4in = 7.79 ft > Max{0.9(108 − 6),0.72 × 108},effective shear
depth
Vp = 0
Concrete Contribution:
Vc = 0.0316 β f c′bv d v
Vu − φV p 2391
v= = = 0.889ksi
φbv d v 0.90 × 32 × 93.4
v f c′ = 0.889 / 6 = 0.148
Step 2: Calculate the strain in the reinforcement on the flexural tension side of the member:
Assume θ = 27 degrees
Mu
+ 0.5 N u + 0.5Vu cot θ − A ps f po
dv
εx =
2( E s As + E p A ps )
82091
+ 0.5 × 2391 × cot 27 − 67.7 × 189
= 7.79
2(28500 × 67.7)
− 2257 + 1195 × cot 27
= = 0.000023
3858900
Step 3: Find the values of θ and ε x × 1000 in Table 5.8.3.4.2-1 which correspond to
Mu
+ 0.5 N u + 0.5Vu cot θ − A ps f po
dv
εx =
2( E c Ac + E s As + E p A ps )
Vc = 0.0316 β f c′bv d v
= 0.0316 × 2.60 6 × 32 × 93.4 = 602kip
Vs = Vn − Vc = Vu φ − Vc
= 2391 / 0.9 − 602 = 2055kip = 1028kip / web
Av Vs
=
s f y d v cot θ
1028
= = 0.093 in 2 in = 1.12in 2 ft
60 × 93.4 × cot 26.9°
Use double #6 bars at 9” centers per web Av = 1.17 in
2
ft
Longitudinal Reinforcement
For sections not subjected to torsion, longitudinal reinforcement needs to satisfy:
⎡M N ⎛V ⎞ ⎤
As f y + A ps f ps ≥ ⎢ u + 0.5 u + ⎜⎜ u − 0.5V s − V p ⎟⎟ cot θ ⎥ (LRFD 5.8.3.5-1)
⎣ d vφ φ ⎝φ ⎠ ⎦
φ = 0.95 for flexure; (Table 5.5.4.2.2-1)
Mu N ⎛V ⎞
+ 0.5 u + ⎜⎜ u − 0.5Vs − V p ⎟⎟ cot θ
d vφ φ ⎝φ ⎠
82091 ⎛ 2391 ⎞
= +0+⎜ − 0.5 × 2055 − 0 ⎟ × cot 26.9°
7.79 × 0.95 ⎝ 0.90 ⎠
= 14304kip
Therefore, the condition (5.8.3.5-1) is satisfied.
Vu = 1087kip
φ = 0.90
Nominal shear resistance:
V n = Vc + V s + V p
or V n = 0.25 f c′bv d v + V p
where:
f c′ = 6ksi , compression strength of concrete
bv = 32in , effective web width
dv = 108 − 5 − 2.6 / 2 = 101.7in = 8.48 ft > Max{0.9(108 − 5),0.72 × 108} , effective shear
depth;
Vp = 0
Concrete Contribution:
Vc = 0.0316 β f c′bv d v
Transverse reinforcement contribution:
Av f y dv cot θ
Vs =
s
where:
β = factor indicating ability of diagonally cracked concrete to transmit tension
θ = angle of inclination of diagonal compressive stresses
Vu − φV p 1087
v= = = 0.371ksi
φbv d v 0.9 × 32 × 101.7
v f c′ = 0.371 / 6 = 0.062
Step 2: Calculate the strain in the reinforcement on the flexural tension side of the member.
Use Equation 1 since strain will be positive:
Assume θ = 24.3 degrees
Mu
+ 0.5 N u + 0.5Vu cot θ − Aps f po
dv
εx =
2( Es As + E p Aps )
20816
+ 0.5 × 1087 × cot 24.3 − 19.1 × 189
= 8.48
2(19.1× 28500)
− 1154 × 543 cot 24.3
= = 0.0000453
1088472
Step 3: Find the values θ and ε x × 1000 in Table 5.8.3.4.2-1 which correspond to
Vc = 0.0316 β f c′bv dv
= 0.0316 × 3.24 6 × 32 × 101.7 = 816kip
Vs = Vn − Vc = Vu φ − Vc
= 1087 / 0.90 − 816 = 392kip = 196kip / web
Av Vs
=
s f y dv cot θ
196
= = 0.0145in2 in = 0.174 in2 ft
60 ×101.7 × cot 24.3°
s 12
Minimum reinforcing Av = 0.0316 f c′bv = 0.0316 6 × 16 × = 0.248 in 2 ft
fy 60
Longitudinal Reinforcement
For sections not subjected to torsion, longitudinal reinforcement needs to satisfy:
⎡M N ⎛V ⎞ ⎤
As f y + Aps f ps ≥ ⎢ u + 0.5 u + ⎜⎜ u − 0.5Vs − V p ⎟⎟ cot θ ⎥ (LRFD 5.8.3.5-1)
⎣ d vφ φ ⎝φ ⎠ ⎦
φ = 0.95 for flexure; (Table 5.5.4.2.2-1)
Mu N ⎛V ⎞
+ 0.5 u + ⎜⎜ u − 0.5Vs − V p ⎟⎟ cot θ
dvφ φ ⎝φ ⎠
20816 ⎛ 1087 ⎞
= +0+⎜ − 0.5 × 930 − 0 ⎟ × cot 24.3°
8.48 × 0.95 ⎝ 0.90 ⎠
= 4229kip
Therefore, the condition (5.8.3.5-1) is satisfied.
Design Examples (Using AASHTO Segmental Spec modified in accordance with AASHTO
T-5 & T-10 Committee)
Node number: 41 (at critical shear section)
Vu = 2391kip
φ = 0.90 for shear
Concrete Contribution:
Vc = 2 K f c′bw d
K = 1 + f pc / 2 f c′ ≤ 2.0
Note: Tensile stress at the extreme fiber under factored loads with effective prestressing was
Vs = Vn − Vc = Vu φ − Vc
= 2391 / 0.90 − 1011 = 1646kip = 823kip / web
Av V
= s
s f yd
823
= = 0.134in 2 in = 1.61 in 2 ft
60 × 102
Use double #6 bar at 6” centers per web Av = 1.76 in
2
ft
Av f y d
Vs =
s
2 × 1.76 × 60 × 102
Vs = = 1795kip
12
Ultimate shear resistance:
φVn = φ (Vc + Vs + V p )
Vp = 0
Vn = Vc + Vs + V p ≤ 10 f c′b d
thickness or going to deeper section. Note that in the current Guide Specification for Design and
Vu = 1087kip
φ = 0.90 for shear
Concrete Contribution:
Vc = 2 K f c′bw d
K = 1 + f pc / 2 f c′ ≤ 2.0
Note: Tensile stress at the extreme fiber under factored loads with effective prestressing was
Vs = Vn − Vc = Vu φ − Vc
= 1087 / 0.90 − 1021 = 187kip = 93kip / web
Av V
= s
s f yd
93
= = 0.015in 2 in = 0.18 in 2 ft
60 × 103
50bw s 50 × 16 × 12
Minimum reinforcing Av = = = 0.16 in 2 ft
fy 60,000
Minimum reinforcing does not control. However, conservatively use double #5 at 18” centers
Av = 0.413 in 2 ft
Av f y d
Vs =
s
2 × 0.413 × 60 × 103
Vs = = 425kip
12
Ultimate shear resistance:
φVn = φ (Vc + Vs + V p )
Vp = 0
Vn = Vc + Vs + V p ≤ 10 f c′b d
Vn 1446 × 1000
# RootsVn = = = 5.7 f c′ ≤ 10 f c′ , O.K.
f c′b d 6000 × 32 × 103
A stability analysis during construction is one of the design criteria for segmental bridge design.
During the construction of a segmental bridge, the boundary conditions constantly change from
the beginning of construction to the end. At any time during construction, the structure and
foundation must be in a stable state and have ample safety factors against material failure, over-
turning, and buckling. Stability analysis, therefore, becomes an important design issue due to the
lower degree of redundancy and the load imbalance of the structure during this period.
A free cantilever structure is one example that requires a stability check during erection of a
segment. The longer the span length, the larger the unbalanced forces. In many cases,
temporary supports are required to handle the load imbalance during erection. In addition to
balanced cantilever conditions, other partially completed structures may also need to be
investigated.
It is important that the engineer specify on design plans the construction loads that were
assumed during design. The limits of these loads and locations where loads are applied on the
structure should also be shown. Additionally, the engineer’s construction schemes should be
clearly stated, including approximate support reactions due to construction equipment. The
stresses caused by critical construction loads and strengths of the members should also be
checked.
The stability analysis specification was originally covered in article 7.4 of the AASHTO Guide
Specifications for Design and Construction of Segmental Concrete Bridges, Second Edition 1999.
Later, those specifications were adopted by the AASHTO LRFD Bridge Design Specifications,
Third Edition, 2004, under Article 5.14.2.3.
CLL = Distributed construction live load; taken as 0.01ksf of deck area applied to one side
of cantilever and 0.005 ksf on the other side
WUP = Wind uplift on cantilever taken as 0.005 ksf of deck area applied to one side only
T = Thermal loads; the sum of the effects due to uniform temperature variation (TU) and
temperature gradients (TG)
Allowable Tensile
Combination Stress (ksi)
a1 = DC + DIFF + CLL + (CE + IE) 0.19√f’c
a2 = DC + DIFF + CLL + (CE + IE) + OTHER LOADS 0.22√f’c
b1 = DC + U + CLL + (CE + IE) 0.19√f’c
b2 = DC + U + CLL + (CE + IE) + OTHER LOADS 0.22√f’c
c1 = DC + DIFF + 0.7WS + 0.7WUP 0.19√f’c
c2 = DC + DIFF + 0.7WS + 0.7WUP + OTHER LOADS 0.22√f’c
d1 = DC + DIFF + CLL + CE + 0.7WS + WUP + 0.7WE 0.19√f’c
d2 = DC + DIFF + CLL + CE + 0.7WS + WUP + 0.7WE + OTHER LOADS 0.22√f’c
e1 = DC + U + CLL + (CE + IE) + 0.3WS + 0.3WE 0.19√f’c
e2 = DC + U + CLL + (CE + IE) + 0.3WS + 0.3WE + OTHER LOADS 0.22√f’c
f1 = DC + CLL + (CE +IE) + CLE + 0.3WS + 0.3WE 0.19√f’c
f2 = DC + CLL + (CE +IE) + CLE + 0.3WS + 0.3WE + OTHER LOADS 0.22√f’c
Notes: 1. OTHER LOADS = CR + SH + TU + TG + EH + EV + ES + WA
2. Allowable compressive stress in concrete where f’c is the compressive strength at the
time of load application.
3. d: equipment not working
e: normal erection
f: moving equipment
Allowable stress:
= − 0.5 × 6 ksi
= -3 ksi
= 0.19 6
= 0.465 ksi
Since the design example has a 200’-0” typical span, only one balanced cantilever structure will
be considered in the stability analysis during construction.
The load combinations “a” to “f” as specified in the AASHTO LRFD Bridge Design Spec. Table
5.14.2.3.3-1, were computed.
= 5 Kips.
CE + IE = 5 × 1.1 = 5.5kips.
A = 78 × 12 × 0.155 = 145kips.
∑ φFu = DC + CE + A + AI
where:
= 145 kips.
CE = 5 kips.
Although calculations have not been shown in this example, of load cases “a” to “f”, strength limit
state load combination “e” controls.
The epoxy resin is applied to the match cast faces of the joint between two segments before
post-tensioning bars are stressed. Purposes of the epoxy resin are as follows:
2. Hardened epoxy provides a water-tight joint, preventing moisture, water and chlorides
from reaching the tendons.
3. Hardened epoxy helps distribute compressive stresses and shear stresses more
uniformly.
4. Hardened epoxy prevents cementitious grout in the tendon duct from leaking out.
The application of epoxy is normally 1/16” thick applied on both faces of match cast joints.
In accordance with the Article 5.14.2.4.2 of the LRFD Specifications for a Type A joint, the
temporary post-tensioning bars should be designed to provide a minimum stress of 0.03 ksi and
an average stress of 0.04 ksi across the joint until the epoxy has cured. The intention of the
stress limitation is to prevent uneven epoxy thickness across the match-cast joint which could
lead to systematic error in geometry control.
Essentially, there are two load cases that need to be considered when designing temporary post-
tensioning bars:
1. Dead load of the segment plus construction loads and temporary post-tensioning bars.
The erection PT bars should be stressed during the open time of the epoxy
(approximately 45 to 60 minutes). The allowable joint stresses for this load case should
conform to Article 5.14.2.4.2 of the LRFD specifications.
2. Case 1. plus permanent cantilever tendons. Normally, one or two hours after the open
time of the epoxy is completed, the allowable joint stress is zero tension, preferably some
compression.
Ac = 78 sf
Ac eff = 70.38 sf
I = 791.892 ft4
Yt = 3.4 → S t = 232.89 ft 3
Yb = 5.6 Ft → S b = 141.40 ft 3
1 1
M max at the joint = −148 × 12 × − × 0.43 × 12 2
2 2
= −918.96 k − ft
Design Assumptions
Check anchoring forces after anchor set for 1 1/4” dia. PT bars.
(
∆FPF = F pj 1 − e − (κx + µα ) ) (LRFD5.9.5.2.2b-
1)
where:
L = 12 ft (segment length)
κ = 0.0002 per ft
µ = 0.3
α = 0.0
∑P i = 807.84 kips
Ys = 3.65 ft
f =−
∑ P + ∑ Pe + M
i i DL
Ac St St
807.84 807.84 × 0.25 918.96
=− + +
70.38 232.89 232.89
= −11.478 + 0.867 + 3.95
= −6.66ksf = −0.046 ksi
fb = −
∑ P − ∑ Pe − M
i i DL
Ac Sb Sb
807.84 × 0.25 918.96
= −11.478 − −
141.40 141.40
= −11.478 − 1.428 − 6.450
= −19.406 ksf = −0.134 ksi
0.046 + 0.134
Average stress =
2
= 0.09ksi
> 0.04ksi ( LRFD 5.14.2.4.2)
ft = −
∑ P − ∑ Pe
i i
Ac St
1968.96 1968.96 × 2.9
=− −
70.38 232.89
= −27.98 − 24.52
= −52.5 ksf = −0.3646 ksi
fb = −
∑ P + ∑ Pe
i i
Ac Sb
1968.96 × 2.9
= −27.98 +
232.89
= −27.98 + 24.52
= −3.46ksf = −0.024 ksi
SUMMATION OF STRESSES
∑f t = −0.046 − 0.3646
= −0.4106 ksi O.K .
∑f b = −0.134 − 0.024
= 0.158 ksi O.K .
∑f t = −0.046 − 0.1823
= −0.2283ksi O.K .
∑f b = −0.134 − 0.012
= − 0.146 ksi O.K .
Conclusion:
8. DETAILING
Based on previously determined shear reinforcing and previously determined web reinforcement
required for flexure, the standard practice has been to use the worst case of adding 50% of shear
steel to 100% of the flexural steel, or 100% of the shear steel to 50% of the flexural steel.
A rational approach can also be used, where the compression strut in an equivalent truss model
would be shifted to the extreme edge of the web. This compression would then be eccentric to a
section through the web which would counteract an applied moment. If the applied moment were
to exceed the amount that could be resisted in this manner, additional reinforcing could be
added. This approach has not been shown at this time, but may be included in the future.
There are two types of shear keys in match-cast joints between precast segments:
• Web shear keys - Located on the faces of the webs of precast box girders. Corrugated
multiple shear keys are preferred due to their superior performance.
• Alignment keys - Located in the top and bottom slabs. Alignment keys are not expected to
transfer the major shear forces; rather they facilitate the correct alignment of the two match-
cast segments being erected in vertical and horizontal directions. For a single-cell box,
normally a minimum of three alignment keys are required on the top slab and one on the
bottom slab.
Both shear and alignment keys should not be located in the tendon duct zones.
The design of web shear keys should satisfy two design criteria:
1. Geometric Design: As per LRFD Fig. 5.14.2.4.2-1, the total depth of shear keys shall
extend approximately 75% of the section depth and at least 75% of the web thickness.
2. Shear Strength Design: As per AASHTO Standards Specifications, 17th Edition, 2002,
Article 9.20.1.5, reverse shearing stresses should be considered in shear key design. At
the time of erection, shear stress carried by the shear key should not exceed 2√f’c (psi).
Alternatively, strength of the shear key could also be computed in accordance with article
12.2.21 of AASHTO Guide Specifications for Design and Construction of Segmental
Concrete Bridges, Second Edition, 1999. However, the AASHTO Guide Specificiation
Shear Key Provision was developed for dry joints only.
When designing shear keys, only web shear keys are considered in transferring the shear forces.
However, alignment shear keys help in preventing local relative vertical displacement on the deck
slab between two adjacent precast segments due to concentrated load on one side of the match
cast joint. Therefore, in longer slabs spanning between two webs or longer cantilevers wings, it is
necessary to provide more than one alignment shear key.
1. Geometric consideration.
h = 9 ft
bw = 16 inches
AASHTO LRFD Bridge Design Spec. does not specify any guideline on the strength design of
shear keys. Use AASHTO Standard Specifications, article 9.20.1.5.
. ( VDC + DIFF )
Vu = 11
where: VDC = shear force due to self weight of one typical segment (kips)
= 78 x 12 x 0.155 = 145 kips
DIFF = 2% of VDC
Vn = Vc
Vu φ = Vc
Vc = A k ⋅ v , per key,
v = 0.2 f c′ ( psi )
90.44
Number of male keys required per web = = 13.9 say 14 keys .
6.5
9.1 Discussion
This design example provides an excellent opportunity to review and apply AASHTO LRFD
Design Specifications, Third Edition, 2004 (LRFD) to segmental concrete bridges. At the same
time, comparative studies were made with other current design specifications, namely AASHTO
Guide Specifications for Design and Construction of Segmental Concrete Bridges, Second
Edition, 1999 (Segmental Specification) and AASHTO Standard Specifications for Highway
Bridges, 17th Edition, 2002 (Standard Specification). It was also interesting to study the
difference using the Standard Specification HS20-44 and HS25-44 live loads.
It is interesting to note that in general, the ultimate limit state load combinations from LRFD
verses the Standard Specification produce forces similar for the HL-93 loading and HS25-44
loading.
Perhaps one of more notable comparisons is shear design using LRFD Modified Compression
Field Theory verses the Segmental Specification/AASHTO T-10 proposal. The Vc contribution for
both codes proved to be somewhat similar. However, when comparing amounts of shear steel
required at critical shear locations, the Segmental Specification required approximately 50% more
shear steel than the Modified Compression Field Theory. This is due to the Segmental
Specification assuming compression diagonals at a 45 degree angle of inclination to determine
Vs, while the Modified Compression Field Theory utilizes an angle based on equilibrium which
can be much less than 45 degrees in prestressed components.
1) Locked-in forces contained in the “EL” loading according to the LRFD code should be lumped
with “DC”. Therefore, locked-in forces will receive an identical load factor as “DC”.
2) Post-tensioning secondary forces should be separated from “EL”. These effects shall be
designated as “PS” and given a γ factor of 1.0 for all strength limit states.
3) Revise the limit of Vn from 10√f’c to 12√f’c in the shear design proposal for AASHTO T-10
Committee.
5) Consider specifying an allowable tension of 3√f’c for unreinforced epoxied joints outside the
precompressed tensile zone for the segmental bridge special load case.
The author of this example supports recent AASHTO T-5 and T-10 Committee proposed changes
in reference to transverse design. For transverse design, it seems rational not to superimpose
axle loads with uniform loads, since they cannot occupy the same area coincidentally. Due to
these proposed modifications, this design example indicates a transverse analysis similar to that
produced by the Standard Specifications for service limit states. Under ultimate limit states, the
exclusion of lane load results in moments smaller than that of the Standard Specification (HS20
loading). This is due to small dead load influences in transverse analysis and live load factors of
1.75 verses 2.17. For this reason, it is recommended that a higher load factor be entertained for
live load.
10. ACKNOWLEDGEMENT
First of all, I would like to thank Mr. Cliff Freyermuth, of the American Segmental Bridge Institute
for giving me the opportunity to work on this design example.
I would also like to thank Mr. Philip Rice, Mr. John Chow and Mrs. Jennifer Hill of Parsons
Brinckerhoff’s Office of Professional Practice Technical Training Program for their funding support
of this project, including support from my supervisors, Mr. Vijay Chandra and PB Technical
Director Mr. Juan Murillo.
I am indebted to Mr. Bryce Binney who diligently performed most of the calculations in the design
example. My appreciation is also extended to my colleagues Mr. Victor Ryzhikov and Mr. Antonio
Ledesma who helped me with review of calculations, including Mr. Scott Lawson, Mr. Robert
Logan, Mrs. Cathy Casteleiro and Ms. Kathie Lozo with graphic works, typing and editing. This
document could not be completed without their assistance.
Last but not least, I would like to thank my wife Maria Theryo for her patience and understanding
due to loss of quality family time of many weekends and nights.
11. REFERENCES
1. AASHTO: Guide Specifications for the Design and Construction of Segmental Bridges, Second
Edition; American Association of State Highway and Transportation Officials; Washington, D.C.;
1999.
2. AASHTO: Standard Specifications for Bridge Design; Seventeenth Edition; American Association
of State Highway and Transportation Officials; Washington, D.C.; 2002.
3. AASHTO: LRFD Bridge Design Specifications, Third Edition, American Association of State
Highway and Transportation Officials; Washington, D.C.; 2004.
5. AASHTO-PCI-ASBI Joint Committee: Segmental Box Girder Standards for Span-by-Span and
Balanced Cantilever Construction; March, 1998.
8. Guyon, Y.; Limit-State Design of Prestressed Concrete, Vol. 1, Applied Science Publishers, LTD;
London; 1972.
10. White, David; Epoxy Resins for Segmental Bridge Construction; Sika Corporation; Lyndhurst, NJ;
1995.
11. AASHTO; Bridge and Structures Annual Meeting 2004, T-10 Meeting Agenda, Orlando, June,
2004.
12. Posten, R.W., Carrasquillo, R.L., and Breen, J.E., Durability of Post-Tensioned Bridge Decks, ACI
Material Journal, July – August, 1987.