Steel Bridge Design Handbook

APPENDIX
Design Example 4: Three-Span
Continuous Straight Composite
Steel Tub Girder Bridge
February 2022
 © AISC 2022

by

American Institute of Steel Construction

All rights reserved. This book or any part thereof must not be
reproduced in any form without the written permission of the publisher.
The AISC and NSBA logos are registered trademarks of AISC.

The information presented in this publication has been prepared following recognized principles of design
and construction. While it is believed to be accurate, this information should not be used or relied upon
for any specific application without competent professional examination and verification of its accuracy,
suitability and applicability by a licensed engineer or architect. The publication of this information is not a
representation or warranty on the part of the American Institute of Steel Construction, its officers, agents,
employees or committee members, or of any other person named herein, that this information is suitable for
any general or particular use, or of freedom from infringement of any patent or patents. All representations
or warranties, express or implied, other than as stated above, are specifically disclaimed. Anyone making
use of the information presented in this publication assumes all liability arising from such use.

Caution must be exercised when relying upon standards and guidelines developed by other bodies and
incorporated by reference herein since such material may be modified or amended from time to time sub-
sequent to the printing of this edition. The American Institute of Steel Construction bears no responsibility
for such material other than to refer to it and incorporate it by reference at the time of the initial publication
of this edition.

Printed in the United States of America

Foreword
The Steel Bridge Design Handbook covers a full range of topics and design examples to provide bridge
engineers with the information needed to make knowledgeable decisions regarding the selection, design,
fabrication, and construction of steel bridges. The Handbook has a long history, dating back to the 1970s
in various forms and publications. The more recent editions of the Handbook were developed and
maintained by the Federal Highway Administration (FHWA) Office of Bridges and Structures as FHWA
Report No. FHWA-IF-12-052 published in November 2012, and FHWA Report No. FHWA-HIF-16-002
published in December 2015. The previous development and maintenance of the Handbook by the FHWA,
their consultants, and their technical reviewers is gratefully appreciated and acknowledged.
This current edition of the Handbook is maintained by the National Steel Bridge Alliance (NSBA), a
division of the American Institute of Steel Construction (AISC). This Handbook, published in 2021, has
been updated and revised to be consistent with the 9th edition of the AASHTO LRFD Bridge Design
Specifications which was released in 2020. The updates and revisions to various chapters and design
examples have been performed, as noted, by HDR, M.A. Grubb & Associates, Don White, Ph.D., and
NSBA. Furthermore, the updates and revisions have been reviewed independently by Francesco Russo,
Ph.D., P.E., Brandon Chavel, Ph.D., P.E., and NSBA.
The Handbook consists of 19 chapters and 6 design examples. The chapters and design examples of the
Handbook are published separately for ease of use, and available for free download at the NSBA website,
www.aisc.org/nsba.
The users of the Steel Bridge Design Handbook are encouraged to submit ideas and suggestions for
enhancements that can be implemented in future editions to the NSBA and AISC at [email protected].
 TECHNICAL REPORT DOCUMENTATION PAGE

1. Title and Subtitle 2. Report Date

Steel Bridge Design Handbook, Appendix February 2022
Design Example 4: Three-Span Continuous
Straight Composite Steel Tub Girder Bridge
3. Original Author(s) 4. Revision Author(s)
Brandon Chavel, Ph.D., PE (HDR) and James Michael A. Grubb, P.E (M.A. Grubb &
Carnahan, PE (HDR) Associates, LLC)

5. Sponsoring Agency Name and Address 6. Revision Performing Organization Name

National Steel Bridge Alliance, a division of the and Address
American Institute of Steel Construction HDR, Inc.
130 E. Randolph, Suite 2000 301 Grant Street, Suite 1700
Chicago, IL 60601 Pittsburgh, PA 15219

7. Supplementary Notes
The previous edition of this Handbook was published as FHWA-HIF-16-002 and was developed to be
current with the 7th edition of the AASHTO LRFD Bridge Design Specifications. This edition of the
Handbook was updated to be current with the 9th edition of the AASHTO LRFD Bridge Design
Specifications, released in 2020.

8. Abstract
Tub girders are often selected over I-girders because of their pleasing appearance offering a smooth,
uninterrupted, cross section. Bracing, web stiffeners, utilities, and other structural and nonstructural
components are typically hidden from view within the steel tub girder resulting in the tub girder’s clean
appearance. Additionally, steel tub girder bridges offer advantages over other superstructure types in
terms of span range, stiffness, durability, and future maintenance.

This design example demonstrates the design of a tangent three-span continuous composite tub girder
bridge with a span arrangement of 187.5 ft - 275.0 ft - 187.5 ft. This example illustrates the flexural
design of a section in positive flexure, the flexural design of a section in negative flexure, the shear
design of the web, an evaluation of using a stiffened versus an unstiffened bottom flange in the
negative flexure region, as well as discussions related to top flange lateral bracing design and bearing
design.

9. Keywords 10. AISC Publication No.

Steel Tub Girder, Top-Flange Lateral Bracing, St. B955-22
Venant Torsional Shear Flow, Diaphragms,
Flange Longitudinal Stiffeners
 Steel Bridge Design Handbook
Design Example 4: Three-Span Continuous Straight
Composite Steel Tub Girder Bridge
Table of Contents
1.0 INTRODUCTION ................................................................................................................. 1
2.0 OVERVIEW OF LRFD ARTICLE 6.11 ............................................................................... 4
3.0 DESIGN PARAMETERS ..................................................................................................... 6
4.0 STEEL FRAMING ................................................................................................................ 8
4.1 Span Arrangement ........................................................................................................... 8
4.2 Bridge Cross-section ........................................................................................................ 8
4.3 Intermediate Cross-frames ............................................................................................... 9
4.4 Top Lateral Bracing ....................................................................................................... 11
4.5 Support Diaphragms ...................................................................................................... 12
4.6 Length of Field Sections ................................................................................................ 13
5.0 PRELIMINARY GIRDER PROPORTIONS ...................................................................... 16
5.1 Girder Depth .................................................................................................................. 16
5.2 Cross-section Proportions .............................................................................................. 16
5.3 Special Restrictions for use of Live Load Distribution Factors ..................................... 21
6.0 LOADS ................................................................................................................................ 23
6.1 Dead Loads .................................................................................................................... 23
6.2 Live Loads ..................................................................................................................... 24
6.2.1 Design Vehicular Live Load (Article 3.6.1.2) ...................................................... 25
6.2.2 Fatigue Live Load (Article 3.6.1.4) ...................................................................... 25
6.2.3 Construction Live Load ........................................................................................ 25
6.3 Load Combinations ........................................................................................................ 25
7.0 STRUCTURAL ANALYSIS............................................................................................... 29
7.1 Live Load Distribution Factors (Article 4.6.2.2) ........................................................... 29
7.1.1 Dynamic Load Allowance (Article 3.6.2.1).......................................................... 30
7.2 Analysis Results ............................................................................................................. 31
7.2.1 Optional Live Load Deflection Evaluation (Article 3.6.1.3.2) ............................. 32

i
8.0 SAMPLE CALCULATIONS .............................................................................................. 38
8.1 Section Properties .......................................................................................................... 38
8.1.1 Section 2-2: Maximum Positive Moment in Center Span .................................... 39
8.1.1.1 Effective Width of Concrete Deck (Article 6.10.1.1.1e) .................... 39
8.1.1.2 Elastic Section Properties for Section 2-2 .......................................... 40
8.1.1.3 Plastic Moment Capacity for Section 2-2 ........................................... 42
8.1.1.4 Yield Moment for Section 2-2 ............................................................ 43
8.1.2 Section 2-1: Maximum Negative Moment at Interior Support ............................. 44
8.1.2.1 Effective Width of Concrete Deck (Article 6.10.1.1.1e) .................... 44
8.1.2.2 Minimum Negative Flexure Concrete Deck Reinforcement .............. 45
8.1.2.3 Elastic Section Properties for Section 2-1 .......................................... 46
8.2 Girder Constructability Check: Section 2-2 (Positive Moment, Span 2)....................... 48
8.2.1 Top Flange Lateral Bending due to Horizontal Component of Web Shear .......... 49
8.2.2 Top Flange Lateral Bending due to Deck Overhang Loads ................................. 50
8.2.3 Top Flange Lateral Bending Amplification .......................................................... 53
8.2.4 Flexure (Article 6.11.3.2)...................................................................................... 55
8.2.4.1 Top Flange - Local Buckling Resistance (Article 6.10.8.2.2) ............ 57
8.2.4.2 Top Flange - Lateral Torsional Buckling Resistance ......................... 58
8.2.4.3 Web Bend-Buckling Resistance (Article 6.10.1.9) ............................. 58
8.2.4.4 Top Flange Constructability Checks ................................................... 59
8.2.4.5 Bottom Flange Constructability Checks ............................................. 60
8.2.5 Shear (Article 6.10.3.3) ......................................................................................... 61
8.2.6 Concrete Deck (Article 6.10.3.2.4) ....................................................................... 61
8.3 Girder Service Limit State Check: Section 2-2 (Positive Moment, Span 2) ................. 61
8.3.1 Permanent Deformations (Article 6.10.4.2) .......................................................... 62
8.4 Girder Fatigue and Fracture Limit State Check: Section 2-2 (Span 2) .......................... 63
8.4.1 Fatigue (Article 6.10.5) ......................................................................................... 63
8.4.1.1 Fatigue in Bottom Flange.................................................................... 64
8.4.2 Fracture (Article 6.6.2) ......................................................................................... 65
8.5 Girder Strength Limit State Check: Section 2-2 (Span 2) ............................................. 66
8.5.1 Flexure (Article 6.11.6.2)...................................................................................... 66

ii
 8.5.1.1 Nominal Flexural Resistance (Article 6.11.7.1.2) .............................. 68
8.5.1.2 Shear (Article 6.11.6.3) ....................................................................... 69
8.5.1.3 End Panel Shear (Article 6.10.9.3.3) .................................................. 71
8.5.1.4 Interior Panel Shear (Article 6.10.9.3.2) ............................................. 71
8.6 Girder Constructability Check: Section 2-1 (Interior Pier Location) ............................ 72
8.6.1 Flexure (Article 6.11.3.2)...................................................................................... 72
8.6.1.1 Top Flange Stress due to Lateral Bending .......................................... 73
8.6.1.2 Top Flange Constructability Check .................................................... 73
8.6.1.3 Bottom Flange - Flexural Resistance in Compression - Stiffened
Flange (Article 6.11.8.2.3) .................................................................................... 74
8.6.1.4 Web Bend-Buckling (Article 6.10.1.9) ............................................... 77
8.6.2 Shear (Article 6.11.3.3) ......................................................................................... 78
8.7 Girder Strength Limit State Check: Section 2-1 (Negative Moment at Interior Pier
Location) ................................................................................................................................. 80
8.7.1 Flexure (Article 6.11.6.2)...................................................................................... 80
8.7.1.1 Bottom Flange - Flexural Resistance in Compression - Stiffened
Flange (Article 6.11.8.2.3) .................................................................................... 81
8.7.1.2 Top Flange - Flexural Resistance in Tension (Article 6.11.8.3) ......... 86
8.7.2 Shear (Article 6.11.6.3) ......................................................................................... 86
8.7.2.1 Interior Panel (Article 6.10.9.3.2) ....................................................... 87
8.8 Girder Service Limit State Check: Section 2-1 (Interior Pier) ...................................... 89
8.8.1 Permanent Deformations (Article 6.10.4.2) .......................................................... 89
8.8.2 Concrete Deck (Article 6.10.1.7) .......................................................................... 92
8.9 Girder Fatigue and Fracture Limit State Check: Section 2-1 (Negative Moment at
Interior Pier Location)............................................................................................................. 93
8.9.1 Fatigue (Article 6.11.5) ......................................................................................... 93
8.9.2 Fracture (Article 6.6.2) ......................................................................................... 93
8.10 Girder Check: Section 1-2 and 1-3 ................................................................................ 93
8.10.1 Comparison of Unstiffened and Stiffened Bottom Flange in End Spans ............. 93
8.10.1.1 Option A - Unstiffened Flange ........................................................... 93
8.10.1.2 Option B - Stiffened Flange ................................................................ 95

iii
 8.10.1.3 Summary of Unstiffened Flange versus Stiffened Flange ................ 101
8.11 Top Flange Lateral Bracing ......................................................................................... 101
8.11.1 Diagonal Bracing Members ................................................................................ 102
8.11.2 Top Lateral Strut ................................................................................................. 104
8.11.3 Detailing .............................................................................................................. 105
8.12 Bearings ....................................................................................................................... 106
8.13 Design Example Summary .......................................................................................... 107
8.13.1 Maximum Positive Moment Region, Span 2 (Section 2-2) ................................ 107
8.13.2 Interior Pier Section, Maximum Negative Moment (Section 2-1) ..................... 107
9.0 REFERENCES .................................................................................................................. 109

iv
 List of Figures

Figure 1 Typical Bridge Cross-Section.......................................................................................... 9

Figure 2 Sketch of the Framing Plan ........................................................................................... 15
Figure 3 Sketch of the Girder Elevation ...................................................................................... 20
Figure 4 Center-to-Center Flange Distance ................................................................................. 21
Figure 5 Maximum Web Inclination............................................................................................ 22
Figure 6 Dead- and Live-Load Moment Envelopes .................................................................... 33
Figure 7 Dead- and Live-Load Shear Envelopes ......................................................................... 34
Figure 8 Fatigue Live Load Moments ......................................................................................... 35
Figure 9 Fatigue Live Load Shears .............................................................................................. 36
Figure 10 Sketch of Section 2-2................................................................................................... 39
Figure 11 Moment of Inertia of an Inclined Web ........................................................................ 40
Figure 12 Sketch Showing Section 2-1 ......................................................................................... 44
Figure 13 Deck Overhang Bracket Loading ................................................................................ 51
Figure 14 Suggested Bottom Flange Bolted Splice Detail at a Flange Stiffener Termination .... 86
Figure 15 Option B in Elevation, Stiffened Bottom Flange ........................................................ 96
Figure 16 Option B, Longitudinal Flange Stiffener Termination ................................................ 99

v
 List of Tables

Table 1 Construction Live Load (CLL) Moments and Shears .................................................... 37

Table 2 Section 2-2: Steel Section Properties ............................................................................... 41
Table 3 Section 2-2: Composite (3n) Section Properties .............................................................. 41
Table 4 Section 2-2: Composite (n) Section Properties ................................................................ 42
Table 5 Section 2-1: Steel Section Properties ............................................................................... 46
Table 6 Section 2-1: Composite Section Properties with Longitudinal Steel Reinforcement ...... 47
Table 7 Section 2-1: Composite (3n) Section Properties .............................................................. 47
Table 8 Section 2-1: Composite (n) Section Properties ................................................................ 48

vi
1.0 INTRODUCTION

Steel boxes may either be tub sections or closed-box sections, with either inclined or vertical webs.
Most composite box girders built in the U.S. are tub girders having a solid bottom flange, two solid
webs, and an open top with two separate top flanges on each web connected with top lateral bracing
to form a pseudo-box to resist the torsion prior to hardening of the concrete deck. Narrow
noncomposite closed steel boxes are often employed as straddle beams to provide support and
necessary underclearance.

Tub girders are sometimes selected over I-girders where aesthetic considerations are a significant
factor because of their pleasing appearance offering a smooth, uninterrupted, cross-section.
Bracing, web stiffeners, utilities, and other structural and nonstructural components are typically
hidden from view within the steel tub girder, resulting in the tub girder’s clean appearance.
Additionally, steel tub-girder bridges offer some distinct advantages over other superstructure
types in terms of span range, stiffness, durability, and future maintenance.

Steel tub girders can potentially be more economical than steel plate I-girders in long-span
applications due to the increased bending strength offered by their wide bottom flanges, and
because they require less field work due to the handling of fewer pieces. Steel tub girders can also
be suitable in short span ranges as well, especially when aesthetic preferences preclude the use of
other structure types. However, tub girders should be no less than 5 feet deep to allow access for
inspection, thus limiting the efficiency of conventional steel tub girders in short-span applications.

Tub girders provide a more efficient cross-section for resisting torsion than I-girders. The
increased torsional resistance of a closed composite steel tub girder results in an improved lateral
distribution of loads. Tub girders offer some distinct advantages over I-girders for horizontally
curved bridges since the torsional stiffness of a tub girder is much larger than the torsional stiffness
of an I-girder. The high torsional resistance of individual tub-girder sections permits the tub girder
to carry more of the load applied to it rather than shifting the load to the adjacent tub girder with
greater radius, as is the case for torsionally weaker I-girders. The tendency to share gravity loads
more uniformly reduces the relatively large deflection of the girder on the outside of the curve.
Also, less material needs to be added to tub girders to resist the torsional effects. Torsion in tub
sections is resisted mainly by St. Venant torsional shear flow, rather than by warping torsion
(which is the primary torsional response mechanism in I-shaped girders). Thus, warping shear and
normal stresses due to warping torsion are typically quite small and AASHTO LRFD BDS Article
C6.11.1.1 recommends that these stresses be neglected.. However, warping associated with
distortion of the cross-section should be considered when evaluating the fatigue performance of
tub girders in certain cases, as discussed further in Section 8.4.1 of this design example.

The exterior surfaces of tub girders are less susceptible to corrosion since there are fewer details
for debris to accumulate, in comparison to an I-girder structure. For tub girders, stiffeners and most
diaphragms are located within the tub girder and are protected from the environment. Additionally,
the interior surface of the tub girder is protected from the environment, further reducing the
likelihood of deterioration. Tub-girder bridges tend to be easier to inspect and maintain since much
of the inspection can occur from inside the tub girder, with the tub serving as a protected walkway.

1
Erection costs for tub girders may be lower than that of I-girders because the erection of a single
tub girder, in a single lift, is equivalent to the placement and connection of two I-girders. However,
a single tub girder will typically require the use of a larger crane than an I-girder of the same
length. Tub girders are also inherently more stable during erection, due to the presence of lateral
bracing between the top flanges. Overall, the erection of a tub girder bridge may be completed in
less time than that of an I-girder counterpart because there are fewer pieces to erect, a fewer number
of external cross-frames or diaphragms to be placed in the field, and subsequently fewer field
connections to be made. This is a significant factor to consider when available time for bridge
erection is limited by schedule or site access.

In many instances, these advantages are not well reflected in engineering cost estimates based
solely on quantity take-offs. Consequently, tub girder bridges have historically been considered
more economical than I-girder bridges only if they have resulted in a reduction in the total number
of webs in cross section, particularly for straight bridges. This is, in part, due to the cross-sectional
restrictions placed on the use of approximate live load distribution factors for straight tub girders
specified in Article 6.11.2.3 of the AASHTO LRFD Bridge Design Specifications [1], referred to
hereafter as the AASHTO LRFD BDS, when the design is based on line girder analysis. For the
AASHTO live load distribution factors to be applicable, the tub girder cross-sectional dimensions
must satisfy limits that may make a tub girder cross-section less competitive than a comparable I-
girder cross-section.

However, these cross-sectional restrictions do not apply when a refined analysis is employed, thus
allowing the designer to explore additional, and perhaps, more economical design options. Also,
if a particular fabricator has the experience and is equipped to produce tub girders efficiently, the
competitiveness of tub girders in a particular application can be enhanced. Therefore, the
comparative economies of I- and tub girder systems should be evaluated on a case-by-case basis,
and the comparisons should reflect the appropriate costs of shipping, erection, and future
inspection and maintenance, as well as fabrication. A more in-depth discussion on the relative
advantages of steel tub girders and on steel tub girder design and construction may be found in the
NSBA publication Practical Steel Tub Girder Design [2], which is available on the NSBA website
(www.aisc.org/nsba).

This design example demonstrates the design of a tangent three-span continuous composite tub-
girder bridge with a span arrangement of 187.5 feet ― 275.0 feet ― 187.5 feet. This example will
illustrate the flexural design of a section in positive flexure, the flexural design of a section in
negative flexure, the shear design of the web, the evaluation of using a stiffened versus an
unstiffened bottom flange in the negative flexure region, as well as discussions related to top flange
lateral bracing and bearing design. Since the cross-sectional restrictions on the use of approximate
live-load distribution factors for straight tub girders are satisfied in this design example (Section
5.3), all flexural moments and shears in this example are determined using a line girder analysis.
However, as mentioned above, more economical design options may be possible by going outside
of these cross-sectional restrictions and utilizing a more refined analysis.

The bridge cross-section consists of two trapezoidal tub girders with top flanges spaced at 11.5
feet on centers, 12.0 feet between the centerline of adjacent top tub flanges, and 4.0-foot overhangs
for a deck width of 43.0 feet out-to-out. For the sake of brevity, only the Strength I, Service II, and

2
Fatigue load combinations (Section 6.3) are examined for dead- and live-load force effects in this
design example. The effects of wind loads, design permit loads, and other loads (braking forces,
seismic forces, etc.) are not considered. It is recommended that the reader refer to NSBA’s Steel
Bridge Design Handbook: Design Example 1: Three-Span Continuous Straight Composite Steel
I-Girder Bridge [3] for information regarding additional load combination cases and design for
wind-load force effects both during construction and in the final constructed condition.

3
2.0 OVERVIEW OF LRFD ARTICLE 6.11

The design of composite tub girder flexural members is contained within Article 6.11 of the 9th
Edition of the AASHTO LRFD BDS. The provisions of Article 6.11 have been organized to
correspond more closely to the general flow of the calculations necessary for the design of
composite tub girder flexural members. Most of the provisions are written such that they are largely
self-contained, however, to avoid repetition, some portions of Article 6.11 refer to provisions
contained in Article 6.10 for the design of I-section flexural members when applicable. The
provisions of Article 6.11 are organized as follows:

6.11.1 General
6.11.2 Cross-Section Proportion Limits
6.11.3 Constructability
6.11.4 Service Limit State
6.11.5 Fatigue and Fracture Limit State
6.11.6 Strength Limit State
6.11.7 Flexural Resistance - Sections in Positive Flexure
6.11.8 Flexural Resistance - Sections in Negative Flexure
6.11.9 Shear Resistance
6.11.10 Shear Connectors
6.11.11 Stiffeners

It should be noted that Article 6.11, and specifically Article 6.11.6.2, does not permit the use of
Appendices A6 and B6 because the applicability of these provisions to tub girders has not been
demonstrated; however, Appendices C6 and D6 are generally applicable. Flow charts for flexural
design of steel I-girders, along with an outline giving the basic steps for steel-bridge superstructure
design, are provided in Appendix C6. Appendix C6 may also prove to be a useful reference for tub
girder design. Fundamental calculations for flexural members are contained in Appendix D6.

Example calculations demonstrating the provisions of Article 6.10, pertaining to straight and
horizontally curved I-girder design and straight rolled-beam design, are provided in NSBA’s Steel
Bridge Design Handbook Design Examples 1, 2A, 2B, and 3 [3-6]. This design example will
demonstrate the application of the provisions of Article 6.11 of the AASHTO LRFD BDS as they
relate to straight tub girder design. NSBA’s Steel Bridge Design Handbook: Design Example 5:
Three-Span Continuous Horizontally Curved Composite Steel Tub-Girder Bridge [7] demonstrates
the application of these provisions to a horizontally curved tub girder design.

The provisions of Articles 6.10 and 6.11 provide a unified approach for consideration of major-
axis bending and flange lateral bending for both straight and horizontally curved bridges. Even for
straight tub-girder bridges, the top flange can be subjected to significant lateral bending stresses
during construction. Bottom flange lateral bending stresses tend to be quite small, since (as
explained earlier) torsion in a tub girder is carried primarily by St. Venant torsional shear flow,
rather than by warping torsion. Top flange lateral bending is caused by the outward thrust induced
by the inclination of the webs, by wind loads, by eccentric loading of temporary support brackets
for deck overhangs, and by forces introduced by the lateral bracing system.

4
In addition to checking that the design provides adequate strength, the constructability provisions
of Article 6.11.3 verify that nominal yielding does not occur and that there is no reliance on post-
buckling resistance for main load-carrying members during critical stages of construction. The
AASHTO LRFD BDS specifies that for critical stages of construction, both compression and
tension flanges must be investigated, and the effects of lateral bending in the top flanges should be
considered. For noncomposite top flanges in compression, constructability design checks verify
that the maximum combined stress in the flanges will not exceed the minimum specified yield
strength, the flanges have sufficient strength to resist lateral torsional and flange local buckling,
and that theoretical web-bend buckling and web shear buckling will not occur during construction.
For noncomposite bottom flanges in compression during critical stages of construction, local
buckling of the flange is checked in addition to the web-bend buckling and shear buckling
resistance. For noncomposite top and bottom flanges in tension, constructability design checks
verify that the maximum combined stress will not exceed the minimum specified yield strength of
the flanges during construction. At the strength limit state, the top flanges are continuously braced
by the hardened concrete deck and flange lateral bending stresses along with lateral torsional and
flange local buckling of the flanges is not a concern. Also, due to the inherent torsional stiffness
and strength of the closed section represented by the tub girder with the hardened composite
concrete deck, global lateral torsional buckling of the composite tub girder is also not a concern.

5
3.0 DESIGN PARAMETERS

The following data apply to this example design:

Specifications: 2020 AASHTO LRFD Bridge Design Specifications [1], Customary

U.S. Units, Ninth Edition
Structural Steel: ASTM A709, Grade 50W uncoated weathering steel with Fy = 50 ksi,
Fu = 70 ksi
Concrete: f’c = 4.0 ksi,  = 150 pcf
Slab Reinforcing Steel: ASTM A615, Grade 60 with Fy = 60 ksi

Permanent steel stay-in-place deck forms are used between the girders; the forms are assumed to
weigh 15.0 psf, since it is assumed concrete will be in the flutes of the deck forms. In this example,
the steel stay-in-place deck forms are used between the top flanges of individual tub girders, and
between the top flanges of adjacent girders. The tub girders in this example are composite
throughout the entire span, including regions of negative flexure.

An allowance for a future wearing surface of 25.0 psf is incorporated in the design. Also, an
allowance for temporary construction loading of 10.0 psf is applied to the noncomposite structure
during construction.

For the fatigue design, the Average Daily Truck Traffic (ADTT) in one direction, considering the
expected growth in traffic volume over the 75-year fatigue design life, is assumed to be 2,000
trucks/day.

Composite tub girder bridges fabricated using uncoated weathering steel have performed
successfully without any interior corrosion protection. However, the interior of tub girders should
always be coated in a light color to aid visibility during girder inspection. Without Owner-agency
direction towards a specific coating and preparation, the girder interior should receive a light brush
blast and be painted with a white or light-colored coating capable of telegraphing cracks in the
steel section. Specified interior coatings should be tolerant of minimal surface preparation. At the
Engineer’s discretion, for painted tub girders, an allowance may be made for the weight of the
paint as discussed in Article C6.11.3.1.

Provisions for adequate draining and ventilation of the interior of the tub are essential. As
suggested in the NSBA Publication Practical Steel Tub Girder Design [2], bottom flange drain
holes should be 1 ½ inches in diameter and spaced along the low side of the bottom flange every
50 feet and be placed 4 inches away from the web plate. Access holes must be provided to allow
for periodic structural inspection of the interior of the tub. The access holes should provide easy
access for authorized inspectors. Solid doors can be used to close the access holes, however, they
should be light in weight, and they should be hinged and locked, but not bolted. Alternatively,
wire-mesh screens can be placed over access holes. Wire mesh should be 10-gage to withstand
welding and blasting and have a weave of approximately ½ inch by ½ inch. Wire-mesh screens
should always be used over the bottom flange drain holes to prevent entry of wildlife and insects.

Additional detailing guidelines can be found at www.aisc.org/nsba, which is the NSBA website,
with particular attention given to the AASHTO/NSBA Steel Bridge Collaboration document G1.4,

6
Guidelines for Design Details [8]. Three other detailing references offering guidance are the Texas
Steel Quality Council’s Preferred Practices for Steel Bridge Design, Fabrication, and Erection
[9], the Mid-Atlantic States Structural Committee for Economic Fabrication (SCEF) Standards,
and the AASHTO/NSBA Steel Bridge Collaboration document G12.1, Guidelines to Design for
Constructability and Fabrication [10].

7
4.0 STEEL FRAMING

4.1 Span Arrangement

Careful consideration to the layout of the steel framing is an important part of the design process
and involves evaluating alternative span arrangements based on the superstructure and substructure
cost to arrive at the most economical solution. Often, site-specific features will influence the span
arrangement required. However, in the absence of these issues, choosing a balanced span
arrangement for continuous steel bridges (end spans approximately 80% of the length of the center
spans) will provide an efficient design. The span arrangement for the example bridge has spans of
187.5 feet ― 275.0 feet ― 187.5 feet. It is evident that this is not an ideal balanced span
arrangement; however, the span arrangement is chosen to illustrate some concepts generally not
found in an ideal span arrangement. Refer to NSBA’s Steel Bridge Handbook Design: Example 1:
Three-Span Continuous Straight Composite Steel I-Girder Bridge [3] for further discussion on
span arrangement considerations.

4.2 Bridge Cross-section

When developing the bridge cross-section, the designer will evaluate the number of girder lines
required, relative to the overall cost. Specifically, the total cost of the superstructure is a function
of steel quantity, details, and erection costs. Developing an efficient bridge cross-section should
also consider the provision of an efficient deck design, which is generally influenced by girder
spacing and overhang dimensions. Specifically, with the exception of an empirical deck design,
girder spacing significantly affects the design moments in the deck slab. Larger deck overhangs
result in a greater load on the exterior web of the tub girder. Larger overhangs will increase the
bending moment in the deck, caused by the cantilever action of the overhang, resulting in
additional deck slab reinforcing for the overhang region of the deck.

In addition, wider deck spans between top flanges can become problematic for several reasons.
Some owners have economical deck detail standards that may not be suited, or even permitted, for
wider deck spans. At the same time, wider deck spans are progressively more difficult to form and
construct.

Special attention should be paid to the design of decks for steel tub girder bridges in the area near
the girder top flanges between adjacent girders. The inherent torsional stiffness of tub girders can
produce a situation where the deck is subjected to a racking effect when there is differential vertical
displacement between adjacent girders. This phenomenon is illustrated in Figure C9.7.2.4-1 of the
AASHTO LRFD BDS [1]. This effect is not directly addressed in the empirical deck design method
(as noted in the AASHTO LRFD BDS Commentary C9.7.4.2). When the traditional deck design
method is used, the effects of this phenomenon should be addressed either by approximate
calculation methods (when a line girder analysis method is being used) or by evaluating deck
stresses (when a refined analysis model is being used).

If empirical live load distribution factors are to be employed, the final cross-section must meet the
requirements of Article 6.11.2.3, which states that the deck overhang should not exceed 60 percent
of the distance between centers of the top flanges of adjacent tub girders, or 6.0 feet. Also, the

8
distance center-to-center of adjacent tub girders is not to be greater than 120 percent nor less than
80 percent of the top flange center-to-center distance of a single tub girder.

The example bridge cross-section consists of two trapezoidal tub girders with top flanges spaced
at 11.5 feet on centers, 12.0 feet between the centerline of adjacent top flanges with 4.0-foot deck
overhangs and an out-to-out deck width of 43.0 feet. The deck overhangs are 33 percent of the
adjacent tub girder spacing. The 40.0-foot roadway width can accommodate up to three 12-foot-
wide design traffic lanes. The total thickness of the cast-in-place concrete deck is 9.5 inches,
including a 0.5-inch thick integral wearing surface. The concrete deck haunch is 3.5 inches deep
measured from the top of the web to the bottom of the deck. The width of the deck haunch is
assumed to be 18.0 inches. Deck parapets are each assumed to weigh 520 pounds per linear foot.
The typical cross-section is shown in Figure 1.

Figure 1 Typical Bridge Cross-Section

4.3 Intermediate Cross-frames

Internal intermediate cross-frames are provided in tub girders to control cross-sectional distortion.
Cross-sectional distortion results due to the St. Venant torsion shear flow changing direction at the
corners of the tub. Cross-sectional distortion introduces additional stresses in the tub girder and,
therefore, should be minimized. The distortion stresses basically occur because the section is not
perfectly round. The shear flow must change direction at the corners, which tends to distort the
cross-section. Adequate internal cross-bracing usually controls the magnitude of these stresses in
tub girders of typical proportion such that they are not critical to the ultimate resistance of the tub
section at the strength limit state. As a minimum, internal cross-frames should be placed at points
of maximum moment within a span and at points adjacent to field splices in straight bridges.

9
Spacing of internal cross-frames, considered during development of the framing plan, should be
influenced by factors such as the angle and length of the lateral bracing members.

Most internal cross-frames in modern tub girder bridges are K-frames, often without a bottom
strut, which allow for better access during construction and inspection. Slenderness requirements
(KL/r) generally govern the design of cross-frame members, however handling and strength
requirements should always considered. When refined analysis methods are used and the cross-
frame members are included in the structural model to determine force effects, the cross-frame
members are to be designed for the calculated force effects. Consideration should be given to the
cross-frame member forces during construction. When simplified analysis methods are used, such
cross-frame forces due to dead and live loads are typically difficult to calculate. Therefore, the
cross-frame members should at least be designed to transfer wind loads and carry any construction
loads due to deck overhang brackets, in addition to satisfying slenderness requirements.

External intermediate cross-frames may be incorporated to control the differential displacements

and rotations between individual tub girders during deck placement. In a finished bridge, when the
tub girders are fully closed and the concrete deck effectively attaches the girders together,
transverse rotation is expected to be small and external cross-frames are not necessarily required.
However, during construction the rotational rigidity of the tub girder is not nearly as large and,
since the two top flanges of a single tub girder are spaced apart but rotate together, the resulting
differential deflections may be large even with a small girder rotation. Helwig et al. [11] present
an approximate method for estimating these differential deflections that can be very helpful in
evaluating the possible need for external intermediate cross-frames early in the design process.

External intermediate cross-frames typically utilize a K-frame configuration, with the depth
closely matching the girder depth for efficiency and simplification of supporting details. Solid web
(plate girder) diaphragms have been successfully used as well. At locations of external
intermediate cross-frames, there should be bracing inside the tub girder to receive the forces of the
external bracing. In some cases, for aesthetic reasons, it may be desirable to remove the external
intermediate cross-frames after the deck has hardened. However, extreme care should be taken in
evaluating the effects that the removal of external intermediate cross-frames has on the structure.
The NSBA Publication Practical Steel Tub Girder Design [2] offers further discussion on this
topic.

Based on the preceding considerations, the cross-frame spacings shown on the framing plan in
Figure 2 were chosen for this example. The internal cross-frames are uniformly spaced in the end-
span and center-span field sections. However, this is not the case for the two field sections at the
interior supports. Due to the lack of symmetry in the interior-support field sections, the internal
cross-frame spacing in the end-span region differs from the internal cross-frame spacing in the
center-span region. Internal cross-frame spacing in the center-span positive flexure region is 31′-
9″; however, to reduce the unbraced length of the top flange so as to increase the lateral torsional
buckling resistance for noncomposite loading, a top strut is located in the center of each internal
cross-frame bay.

10
4.4 Top Lateral Bracing

Lateral bracing between common top flanges of a tub girder is required to provide proper shear
flow in the individual tub girders. Without lateral bracing, the section acts as an open section and
is much less stable under torsional loading. The bracing acts to enhance the global lateral torsional
buckling stability of the section. Top lateral bracing raises the shear center to the inside of the tub
section resulting in a pseudo-box section and significantly increasing the torsional stiffness. A
single tub girder with a properly designed top flange lateral bracing system has substantially
greater stability than an equivalent pair of I-shaped girders without lateral bracing, even if the pair
of I-shaped girders has the same net major-axis bending section modulus as the single tub girder.

AASHTO LRFD BDS Article C6.7.5.3 recommends that a full-length lateral bracing system be
provided within straight tub sections utilized on spans greater than about 150 feet; a full-length
system is provided in this design example. For horizontally curved tub girders, a full-length lateral
bracing system must always be provided according to Article 6.7.5.3. A full-length lateral bracing
system is particularly important when the torques on the noncomposite section are large; e.g., in
tub-section members on which the deck weight is applied unsymmetrically, or in members resting
on skewed supports. A full-length lateral bracing system can also help limit distortions that may
result from temperature changes occurring prior to deck placement, and resist the torsion and twist
resulting from any eccentric loads that may act on the steel section during construction, including
the effects of deck overhang brackets. If a full-length lateral bracing system is not provided for
straight tub girders with spans less than about 150 feet, the designer is advised to perform a full
investigation of girder stability during all stages of erection and deck placement, considering
various possible lift points and lifting scenarios, various interim erection conditions, and various
loading effects (including self-weight, wet concrete deck weight, construction dead and live loads,
wind loads, etc.). Top lateral bracing should always be continuous across field-splice locations.
Otherwise, large flange lateral bending stresses might occur in the top flanges of the tub where the
bracing is discontinued.

Top lateral bracing is designed to resist the shear flow in the pseudo-box section resulting from
any factored torsion acting on the steel section before the deck has hardened. Forces in the bracing
due to flexure of the tub during construction should also be considered since these members act
with the tub in flexure. In the absence of a refined analysis, design equations have been developed
by Fan and Helwig [12] to evaluate the bracing member forces due to tub girder bending. Top
lateral bracing members are also subject to wind-load forces acting on the noncomposite tub
section during construction.

Top flange lateral bracing systems for steel tub girders typically take the form of a truss system in
the horizontal plane between the tub girder top flanges, with transverse struts and diagonals.
Various studies and guides have examined different options for the truss system. AASHTO LRFD
BDS Article C6.7.5.3 notes that single-diagonal systems are preferred over X-type systems because
they involve fewer members and fewer connections, facilitating fabrication and assembly. The
Commentary also discusses the various merits of Warren-type vs. Pratt-type truss systems. This
design example utilizes a Warren-type truss system. NSBA’s Steel Bridge Design Handbook:
Bracing System Design [13] discusses issues related to the selection of the lateral bracing
configuration in greater detail.

11
Single angle or structural tee (WT) sections are most commonly used as top flange lateral bracing
members; full rolled beam (W) sections have occasionally been used when forces are large.
Whenever possible, direct attachment of the top flange lateral bracing members to the tub girder
top flanges via bolted connections is preferred to avoid the extra fabrication and assembly costs
associated with using gusset plates; in some cases, a slight increase in tub girder top-flange width
to accommodate the bolted connections can be more economical than a “least weight” design that
would require gusset plates to accommodate the connections to the flanges. Although not checked
in this example, wherever the bracing members are bolted to a top flange subject to tension,
AASHTO LRFD BDS Equation 6.10.1.8-1 must be satisfied at cross-sections of flexural members
containing holes in the tension flange at the strength limit state and when checking constructability.

As noted previously, the internal intermediate cross-frames in tub girders are typically integrated
with the top flange lateral bracing system by using some or all of the top flange lateral bracing
transverse struts as the top chord members of the internal cross-frames. Section 8.11 illustrates the
design of a diagonal and a top-strut lateral member in the top lateral bracing system from this
example.

4.5 Support Diaphragms

Internal diaphragms at points of support are typically full-depth, solid web sections with a top
flange. These diaphragms are subjected to bending moments which result from the shear forces in
the inclined girder webs. If a single bearing is used at the support that does not approach the full
width of the tub girder bottom flange, bending of the internal diaphragm over the bearing will
result, causing tensile stresses in the top flange of the diaphragm and compressive stresses in the
bottom flange of the tub girder. Additionally, a torsional moment reaction in the tub girder at the
support will induce a shear flow along the circumference of the internal diaphragm. To provide
the necessary force transfer between the tub girder and the internal diaphragms, the internal
diaphragms should be connected to the web and top flanges of the tub girder.

Inspection access through the internal diaphragms at interior supports must be provided with
access holes at least 18 inches wide and 24 inches high; however, if feasible, a larger hole at least
36.0 in. deep is preferable. In addition to restraining distortion of the box section, the internal
diaphragms at supports also transfer load from the girder webs to the bearing(s). If a single
centered bearing is used, the diaphragm must be stout enough to resist the reaction and transfer
the load around any access hole. Bearing stiffeners are usually attached to the diaphragms. If a
single centered bearing is employed, two stiffeners are generally used. A bearing stiffener on
each side of the access hole generally removes the shear from the diaphragm before it is engaged
by the hole. Torsion generally causes a different magnitude of shear in the webs of the box on
the two sides of the diaphragm. Reinforcement around the hole may be required, particularly if
the access hole requires a large portion of the diaphragm or if a single bearing is located under
the diaphragm. Auxiliary stiffeners on the diaphragm or webs may be employed to spread out
the reaction.

As discussed in AASHTO LRFD BDS Article C6.7.4.3, external plate diaphragms with aspect
ratios, or ratios of length to depth, less than 4.0 and internal plate diaphragms act as deep beams

12
and should be evaluated by considering principal stresses rather than by simple beam theory.
Fatigue-sensitive details on these diaphragms and at the connection of the diaphragms to the
flanges should be investigated by considering the principal tensile stresses.

Similar to internal diaphragms, external diaphragms are typically full-depth plate girder sections,
with top and bottom flanges. As acknowledged in the NSBA publication Practical Steel Tub
Girder Design [2], the behavior of an external diaphragm at a point of support is highly dependent
on the bearing arrangement at that location. If dual bearings used at each girder sufficiently prevent
transverse rotation, external diaphragms at the point of support should theoretically be stress free.
The force couple behavior of a dual bearing system resists the torsion that would otherwise be
resisted by the external diaphragm and, in turn, minimizes the bending moments applied to the
external diaphragm.

If a single bearing under each tub girder is employed, torsional moments must be resisted by the
external diaphragm through vertical bending. In a single bearing arrangement, the internal
diaphragms of adjacent girders function with the external diaphragms to form a system (or beam)
which resists the girder torsional moments. The total torque is resisted by differential reactions at
the bearings of adjacent girders. The diaphragms then are subjected to bending and shear forces.
Torsional moments resisted by the external diaphragm often require the use of a moment
connection to the tub girder in which the flanges and webs of the external diaphragm are connected.
The largest torsional moment will typically occur during the construction stage and can be quite
large, particularly in horizontally curved structures. Torsional moments in straight bridges are
typically smaller but should still be considered in the design.

4.6 Length of Field Sections

The lengths of field sections are generally dictated by shipping (weight and length) restrictions.
Generally, the weight of a single shipping piece is restricted to 200,000 lbs, while the piece length
is limited to a maximum of 140 feet, with an ideal piece length of 120 feet. However, shipping
requirements are typically dictated by state or local authorities, in which additional restrictions
may be placed on piece weight and length. Handling issues during erection and in the fabrication
shop also need to be considered in the determination of field section lengths, as they may govern
the length of field sections. Therefore, the Engineer should consult with contractors and fabricators
regarding any specific restrictions that might influence the field-section lengths.

Field-section lengths should also be determined with consideration given to the number of field
splices required, as well as the locations of the field splices. It is desirable to locate field splices as
close as possible to dead-load inflection points so as to reduce the forces that must be carried by
the field splice. Field splices located in higher moment regions can become quite large, with cost
increasing proportionally to their size. The Engineer must determine what the most cost-
competitive solution is for the given span arrangement. For complex and longer span bridges, the
fabricator’s input can be helpful in reaching an economical solution.

Due to the span arrangement for this example, and the desire to limit field section lengths to 130.0
feet, field splices are not located ideally at dead-load inflection points. Five (5) field sections are
used in each line of girders (Figure 2). For this layout, an end span field section weighs

13
approximately 107,000 lbs, an interior support field section weights approximately 165,000 lbs,
and the center span field section has a weight of approximately 101,000 lbs. Field sections in this
length and weight range can generally be fabricated, shipped, and erected without significant
issues.

14
Figure 2 Sketch of the Framing Plan

15
5.0 PRELIMINARY GIRDER PROPORTIONS

5.1 Girder Depth

Proper proportioning of tub girders involves a study of various girder depths versus girder weight
to arrive at the least weight solution that meets all performance and handling requirements. The
overall weight of the tub girder can vary dramatically based on web depth. Therefore, selection of
the proper girder depth is an extremely important consideration affecting the economy of steel-
girder design. The NSBA Publication, Practical Steel Tub Girder Design [2] points out that a
traditional rule of thumb for steel tub girder bridge depths is L/25, however designers should not
be reluctant to exceed this ratio. Tangent steel tub girders have approached L/35 while meeting all
code requirements for strength and deflection. Article 2.5.2.6.3 provides suggested minimum span-
to-depth ratios for I-girders but does not specifically address tub girder sections. The suggested
minimum depth of the steel section in a composite I-girder, in a continuous span, is given as
0.027L, where L is the span length in feet. This criterion may be applied to determine a starting
depth of the tub girder for the depth studies. Using the longest span of 275.0 ft, the suggested
minimum depth of the steel section is:

0.027(275.0) = 7.425 ft = 89.1 in.

Considering an approximate thickness for the top and bottom flange will lead to a vertical web
depth of approximately 86.5 inches. A preliminary web depth study was performed to determine
an appropriate optimal web depth based on minimum steel weight. This study considered various
web depths and associated flange sizes that satisfied the design requirements in a preliminary
sense. The optimal web depth was chosen from the preliminary design that resulted in the least
amount of steel girder weight. The optimal vertical depth for this study was found to be 84.5 inches.
Therefore, a vertical web depth of 84.4 inches was used which resulted in a web plate size of 87.0
inches, assuming a maximum web inclination of 1:4. This, in turn, resulted in a bottom flange
width of 98.5 inches.

Tub girders typically employ inclined webs, as they are advantageous in reducing the width of the
bottom flange. Article 6.11.2.1 specifies that the web inclination should not exceed 1:4
(horizontal:vertical). Because progressively deeper webs may result in a narrower and potentially
thicker bottom flange plate (at location of maximum flexure), it is necessary for the Engineer to
explore a wide range of web depths and web spacing options in conjunction with the bottom flange
design requirements to determine the optimal solution.

5.2 Cross-section Proportions

Proportion limits for webs of tub girders are specified in Article 6.11.2.1. Provisions for webs with
and without longitudinal stiffeners are presented. For this example, a longitudinally stiffened web
is not anticipated. Therefore, the web plate must be proportioned such that the web plate thickness
(tw) meets the following requirement:

D
 150 Eq. (6.11.2.1.2-1)
tw

16
For inclined webs, Article 6.11.2.1.1 states that the distance along the web is to be used for D in
all design checks.

Rearranging:

D 87
( t w )min. = = = 0.58 in.
150 150

Therefore, utilizing 1/16-inch increments for web plate thickness, select an initial web thickness
of 0.625 inches.

Cross-section proportion limits for top flanges of tub girders are specified in Article 6.11.2.2. The
minimum width of the top flanges is specified as:

D
b f Eq. (6.11.2.2-2)
6

( bf )min. = D/6 = 87/6 = 14.5 in.

Article C6.10.2.2 suggests the following additional guideline for the minimum top-flange width
for each individual unspliced girder field section, btfs, to be used in conjunction with flange
proportion limit specified above. This guideline is intended to provide more stable field pieces for
lifting, erection, and shipping without the need for special stiffening trusses or falsework.

L fs
b tfs  Eq. (C6.10.2.2-1)
85

Lfs is the length of the unspliced individual girder field section under consideration in feet.

Eq. C6.10.2.2-1 is intending primarily for application to unspliced I-girder field sections.
However, as discussed in Article C6.11.3.2, in cases where a full-length lateral bracing system is
not employed within a straight tub girder, which is not the standard practice and should only be
considered if the spans are less than about 150 feet (Section 4.4), the minimum width of the top
flanges within each individual unspliced field section should satisfy the preceding guideline. In this
case, Lfs is to be taken as the larger of the distances along the field section between panels of lateral
bracing or between a panel of lateral bracing and the end of the piece. For cases where a full-length
lateral bracing system is employed, which is the typical case (and the case in this example), Eq.
C6.10.2.2-1 need not be considered for top flanges of tub sections.

In this case, a minimum top flange width of 18 inches is advantageous to connect the top flange
lateral bracing directly to the top flange. Therefore, a minimum top flange width of 18 inches will
be provided so that the flange will be wide enough to accommodate the bolted lateral bracing
connections.

17
The minimum thickness of the top flanges is specified as:

tf ≥ 1.1tw Eq. (6.11.2.2-3)

or:

( t f )min. = 1.1t w = 1.1(0.625) = 0.6875 in.

However, the AASHTO/NSBA Steel Bridge Collaboration document G12.1 Guidelines to Design
for Constructability and Fabrication [8] recommends a minimum flange thickness of 0.75 inches
to enhance girder stability during handling and erection. Therefore, use (tf)min = 0.75 inches.

Additionally, the top flanges must satisfy the following ratio:

b
f  12.0 Eq. (6.11.2.2-1)
2t
f

Therefore, checking the minimum size top flanges:

18
= 12.0 ok
2(0.75)

Top-flange thicknesses of 1.0 inch will be used in regions of positive flexure so that the
noncomposite section will have a greater likelihood of satisfying the constructability checks.

The AASHTO LRFD BDS currently imposes no limitation on the b/t ratio of bottom flanges of
composite tub girders in tension. Past and current industry guidance has suggested “rules of
thumb” for the maximum b/t ratio ranging from as slender as 120 to as stocky as 80. White et al.
(2019) [14] developed guidance (described below) which has been adopted in the AASHTO LRFD
BDS for noncomposite steel box girder members, and which White et al. suggested should also be
considered for composite steel tub girder bottom flanges. These limits are intended to address
several fabrication concerns, including waviness and warping effects during welding of the bottom
flange to the webs. Furthermore, the Engineer should be aware that it is possible that the bottom
flange in tension in the final condition may be in compression during lifting of the tub girder during
erection, possibly causing buckling of the slender bottom flange.

Article 6.12.2.2.2b suggests a limit on the b/t ratio, based on the inside width of the flanges, of 90
for longitudinally unstiffened compression and tension flanges in noncomposite box-section
members to address similar concerns. Compression flanges exceeding this value must include
longitudinal stiffeners. Tension flanges in these members with a b/t ratio exceeding 130 must
include longitudinal stiffeners to prevent noticeable out-of-plane deflections of the flange under
self-weight or under self-weight with a small concentrated transverse load. Unless otherwise
specified by the Owner, a minimum thickness of 0.5 inches is also specified for compression and
tension flanges in these members to limit potential local deformation or distortion of box section
flanges during fabrication, transportation, erection, and service conditions. Additional information

18
on these limits may be found in White et al. (2019) [14]. Additional discussion concerning this
issue can also be found in the NSBA publication Practical Steel Tub Girder Design [2].

If it is desired to exceed the suggested b/t limit of 90 for tension flanges, the Engineer should
consult with fabricators to verify that a tub girder with the selected bottom flange thickness in
regions of positive flexure can be fabricated without causing any significant handling and/or
distortion concerns without providing any flange longitudinal stiffeners. For this example, tension
flanges in regions of positive flexure with a thickness of 13/16 inches and a maximum b/t ratio
(based on the inside width of the flanges) of approximately 117 are utilized. This represents a
significant reduction in the b/t ratios from the original design for this example, which was
completed before the preceding guidance was available. In an actual design, consideration should
probably be given to using a somewhat lower b/t for these flanges.

Bottom flange extensions of 1-3/8 inches (measured from the centerline of the webs) were assumed
in this design example for welding access. It should be noted that the AASHTO/NSBA Steel
Bridge Collaboration document G1.4, Guidelines for Design Details [8] (Page 116) suggests
preferred bottom flange extensions of 1-1/2 inches.

Based on the above minimum proportions, the trial girder shown in Figure 3 is suggested.

19
Figure 3 Sketch of the Girder Elevation

20
5.3 Special Restrictions for use of Live Load Distribution Factors

Special consideration must be given to preliminary proportions for straight tub girder bridges that
will employ use of the live load distribution factors presented in Article 4.6.2.2.2b. Specifically,
cross-sections of straight bridges consisting of two or more single-cell tub girders must satisfy the
geometric restrictions specified in Article 6.11.2.3.

In particular:

• Bearing lines are not to be skewed.

• The distance center-to-center (a) of the top flanges of adjacent tubes, taken at mid-span,
must satisfy:

Figure 4 Center-to-Center Flange Distance

Note: For nonparallel tub girders, in addition to mid-span requirements, Article 6.11.2.3
imposes additional geometric restrictions at the supports.

• The distance center-to-center (w) of the top flanges of individual tub girders must be the
same.

• The inclination of the web must not exceed 1 (horizontal) to 4 (vertical) to a plane normal
to the bottom flange, as shown in Figure 5.

• The overhang of the concrete deck, including the curb and parapet cannot exceed 60
percent of the average distance between the centers of the top flanges of adjacent tub
girders, a, or 6.0 feet.

21
 Figure 5 Maximum Web Inclination

For this example, there are no skewed supports and the distance center-to-center (w) of the top
flanges of the individual tub girders is a constant 11.5 feet:

0.8 (11.5) = 9.2 ft ≤ a = 12 ft ≤ 1.2 (11.5) = 13.8 ft

The inclination of the web is 1 (horizontal) to 4 (vertical) in this example, therefore satisfying the
previously mentioned requirement.

The cantilever deck overhang used in this example is 4.0 feet, therefore less than 0.60(12.0) = 7.2
feet and 6.0 feet.

The requirements of Article 6.11.2.3 are satisfied for this example; therefore, live load flexural
moments and shears for this example may be computed in accordance with Article 4.6.2.2.2b.

22
6.0 LOADS

6.1 Dead Loads

As defined in Article 3.5.1, dead loads are permanent loads that include the weight of all
components of the structure, appurtenances and utilities attached to the structure, earth cover,
wearing surfaces, future overlays and planned widenings.

The component dead load (DC) consists of all the structure dead load except for non-integral
wearing surfaces, if anticipated, and any specified utility loads. For composite steel-girder design,
DC is further divided into:

• Noncomposite dead load (DC1) is the portion of loading resisted by the noncomposite
section. DC1 represents the permanent component load that is applied before the concrete
deck has hardened or is made composite.
• Composite dead load (DC2) is the portion of loading resisted by the long-term composite
section. DC2 represents the permanent component load that is applied after the concrete
deck has hardened or is made composite.

For this example, the dead load component (DC1) is calculated as follows:

9.5
Concrete deck = ( 43.0 )( 0.150 ) = 5.106 kips/ft
12

 1  13.0 + 10  18 2  
Concrete deck overhang tapers = 2   − 9.5  4.0 −  ( 0.150 ) = 0.162 kips/ft
12  2  12  

18 ( 3.50-0.875 ) 
Concrete deck haunches = 4  ( 0.150 ) = 0.197 kips/ft
 144 
(The minimum top flange thickness and associated width are used in the above computation.)

  18  
Stay-in-place forms =  2 (11.5) + 12 − 3    (0.015) = 0.457 kips/ft = 0.457 kips/ft
  12  

Steel girder self-weight

(based on preliminary sizing and confirmed during subsequent analysis) = 1.635 kips/ft

Cross-frames and details = 0.110 kips/ft

DC1 load total (per 2 girders) = 7.908 kips/ft

Therefore, the distributed DC1 load per a girder is:

DC1 load per girder = 7.908 kips/ft  2 girders = 3.954 kips/ft per girder

23
Unless otherwise stipulated by the Owner, it is sometimes assumed, in accordance with Article
4.6.2.2.1, that composite dead loads are supported equally by all girders of straight, non-skewed
bridges with typical deck overhangs and girders of similar stiffness. In most modern designs, large
and heavy concrete barriers are often placed at the outer edges of the concrete deck. When refined
methods of analysis are employed, the self-weight of the concrete barriers (the DC2 loads in this
case) should be applied at their actual locations at the outer edges of the deck. In I-girder bridges,
this results in the exterior girders carrying a larger percentage of these loads. Some Owner-
agencies prescribe the use of different, semi-arbitrary percentages for distribution of the barrier
weight to the exterior girder and to the adjacent interior girders, while others continue to distribute
the barrier weight equally among all girders. In this example, with a pair of torsionally stiff tub
girders and a relatively narrow total cross-section width, the weight of each concrete barrier is
assumed to be distributed equally to each girder in the cross-section; a uniform distribution
assumption would be likely less valid for the design of a much wider bridge with more girders in
the cross-section.

For this example, the composite section dead load (DC2) will consist of the self-weight of the
concrete barrier only. Therefore:

DC2 load per girder = 0.520 kips/ft per girder

The component dead load (DW) consists of the dead load of any non-integral wearing surfaces
and any utilities. DW is also assumed to be equally distributed to all girders. For this example, a
future wearing surface is anticipated but no utilities are included. Therefore:

DW load per girder = [(0.025) x 40]  2 girders = 0.500 kips/ft per girder

For computing flexural stresses from the composite dead loads, DC2 and DW, the stiffness of the
long-term composite section in regions of positive flexure is calculated by transforming the
concrete deck using a modular ratio of 3n (Article 6.10.1.1.1b). In regions of negative flexure, the
long-term composite section is assumed to consist of the steel section plus the longitudinal
reinforcement within the effective width of the concrete deck (Article 6.10.1.1.1c).

6.2 Live Loads

Live loads are assumed to consist of gravity loads (vehicular live loads, rail transit loads and
pedestrian loads), the dynamic load allowance, centrifugal forces, braking forces and vehicular
collision forces. Live loads illustrated in this example include the HL-93 vehicular live load and a
fatigue load, with the appropriate dynamic load allowance included.

Live loads are treated as transient loads applied to the short-term composite section. For computing
flexural stresses from transient loading, the short-term composite section in regions of positive
flexure is calculated by transforming the concrete deck using a modular ratio of n (Article
6.10.1.1.1b). In regions of negative flexure, the short-term composite section is assumed to consist
of the steel section plus the longitudinal reinforcement within the effective width of the concrete
deck (Article 6.10.1.1.1c), except as permitted otherwise for the fatigue and service limit states
(see Articles 6.6.1.2.1 and 6.10.4.2.1).

24
When computing longitudinal flexural stresses in the concrete deck (see Article 6.10.1.1.1d), due
to permanent and transient loads, the short-term composite section should be used.

6.2.1 Design Vehicular Live Load (Article 3.6.1.2)

The design vehicular live load is designated as HL-93 and consists of a combination of the
following placed within each design lane:

• a design truck or design tandem, and

• a design lane load.

The design vehicular live load is discussed in more detail in NSBA’s Steel Bridge Design
Handbook: Design Example 1: Three-Span Continuous Straight Composite Steel I-Girder
Bridge[3].

6.2.2 Fatigue Live Load (Article 3.6.1.4)

The vehicular live load for checking fatigue consists of a single design truck (without the lane
load) with a constant rear-axle spacing of 30 feet (Article 3.6.1.4.1).

The fatigue live load is discussed in more detail in NSBA’s Steel Bridge Design Handbook: Design
Example 1: Three-Span Continuous Straight Composite Steel I-Girder Bridge [3].

6.2.3 Construction Live Load

A construction live load (CLL) should also be considered in evaluating the adequacy of the
superstructure during construction. The construction live load is intended to account for all
miscellaneous construction equipment that cannot be easily quantified at the time of design.
Typically, a load of 10 psf over the width of the bridge is assumed for the construction live loading.
A CLL of 10 psf is applied in this example, resulting in:

CLL load per girder = [(0.010) x 43]  2 girders = 0.215 kips/ft per girder

6.3 Load Combinations

Limit states are defined in the LRFD specifications to satisfy basic design objectives; that is, to
achieve safety, serviceability, and constructability. A detailed discussion of these limit states is
provided in NSBA’s Steel Bridge Design Handbook: Design Example 1: Three-Span Continuous
Straight Composite Steel I-Girder Bridge [3]. For each limit state, the following basic equation
(Article 1.3.2.1) must be satisfied:

ΣiγiQi ≤ Rn = Rr Eq. (1.3.2.1-1)

where: i = load modifier related to ductility, redundancy and operational importance

25
 i = load factor, a statistically based multiplier applied to force effects
 = resistance factor, a statistically based multiplier applied to nominal resistance
Qi = force effect
Rn = nominal resistance
Rr = factored resistance

The load factors are specified in Tables 3.4.1-1 and 3.4.1-2 of the specifications. For steel
structures, the resistance factors are specified in Article 6.5.4.2.

In the AASHTO LRFD BDS, redundancy, ductility, and operational importance are considered
more explicitly in the design. Ductility and redundancy relate directly to the strength of the bridge,
while the operational importance relates directly to the consequences of the bridge being out of
service. For loads for which a maximum value of i is appropriate:

i = DR I  0.95 Eq. (1.3.2.1-2)

where: D = ductility factor specified in Article 1.3.3

R = redundancy factor specified in Article 1.3.4
I = operational importance factor specified in Article 1.3.5

For loads for which a minimum value of i is appropriate:

1
i =  1.0 Eq. (1.3.2.1-3)
D R I

Eq. (1.3.2.1-3) is only applicable for the calculation of the load modifier when dead- and live-load
force effects are of opposite sign and the minimum load factor specified in Table 3.4.1-2 is applied
to the dead-load force effects (e.g., when investigating for uplift at a support or when designing
bolted field splices located near points of permanent load contraflexure); otherwise, Eq. (1.3.2.1-
2) is to be used.

For typical bridges for which additional ductility-enhancing measures have not been provided
beyond those required by the specifications, and/or for which exceptional levels of redundancy are
not provided, the D and R factors have default values of 1.0 specified at the strength limit state.
Note that some owner-agencies specify redundancy factors greater than 1.0 for certain types of
steel tub girder bridges depending on the number of girders in the cross-section and the existence
and number of external intermediate diaphragms. The value of the load modifier for operational
importance I should be chosen with input from the Owner-agency. In the absence of such input,
the load modifier for operational importance at the strength limit state should be taken as 1.0. At
all other limit states, all three  factors must be taken equal to 1.0. For this example, i will be
taken equal to 1.0 at all limit states.

For this example, it has been assumed that the Strength I load combination governs for the strength
limit state, so only Strength I loads are checked in the sample calculations for the strength limit
state included herein. In some design instances, other load cases may be critical, but for this

26
example, these other load cases are assumed not to apply. The Service II and Fatigue load
combinations will be investigated for permanent deflection checks at the service limit state and
checks of selected welded details at the fatigue limit state, respectively. Refer to Design Example
1 of the NSBA Steel Bridge Design Handbook for further detail on all the load combinations
specified in Table 3.4.1-1.

Strength I: 1.25DC + 1.5DW + 1.75(LL+IM)

Service II: 1.0DC + 1.0DW + 1.3(LL+IM)

And for the fatigue limit state:

Fatigue I: 1.75(LL+IM), or

Fatigue II: 0.80(LL+IM)

where LL is the fatigue load specified in Article 3.6.1.4.1.

When evaluating the strength of the structure for the maximum force effects during construction,
the load factor for construction loads, for equipment and for dynamic effects (i.e., temporary dead
and/or live loads that act on the structure only during construction) is not to be taken less than 1.5
in the Strength I load combination (Article 3.4.2.1). Also, the load factors for the weight of the
structure and appurtenances, DC and DW, are not to be taken less than 1.25 when evaluating the
construction condition.

Article 3.4.2.1 further states that unless otherwise specified by the Owner, primary steel
superstructure components are to be investigated for maximum force effects during construction
for an additional load combination consisting of the applicable DC loads and any construction
loads that are applied to the fully erected steelwork. For this additional load combination, the load
factor for DC and construction loads including dynamic effects (if applicable) is not to be taken
less than 1.4. For steel superstructures, the use of higher-strength steels, composite construction,
and limit-states design approaches in which smaller factors are applied to dead load force effects
than in previous service-load design approaches, have generally resulted in lighter members
overall. To provide adequate stability and strength of primary steel superstructure components
during construction, an additional strength limit state load combination is specified for the
investigation of loads applied to the fully erected steelwork (i.e., for investigation of the deck
placement sequence and deck overhang effects).

Construction: Strength I: η x [1.25(D) + 1.5(C)]

Special Load Combination: η x [1.4(D + C)]
where:

D = Dead load
C = Construction loads

27
In this design example, for brevity, only the first of these load combinations is
considered/illustrated in the constructability checks. Wind load effects during construction and in
the final constructed condition are also not considered herein. Refer to NSBA’s Steel Bridge
Design Handbook: Design Example 1: Three-Span Continuous Straight Composite Steel I-Girder
Bridge [3] for an illustration of these checks.

It should be noted that when one force effect decreases another effect, minimum load factors are
to be applied to the load reducing the total effect at the strength limit state. Minimum load factors
for permanent dead loads are specified in Table 3.4.1-2. For example, for the strength limit state
when the permanent load vertical bending moment is positive, but the governing live load vertical
bending moment is negative, the Strength I Load Combination would be: 0.90DC + 0.65DW +
1.75(LL+IM). It is important that these minimum load combinations are considered as appropriate,
especially for structures that do not have an ideal balanced span arrangement.

28
7.0 STRUCTURAL ANALYSIS

Structural analysis is covered in Section 4 of the AASHTO LRFD BDS. Both approximate and
refined methods of analysis are discussed in the Specifications. Refined methods of analysis are
given greater coverage in the AASHTO LRFD BDS than they have been in the past recognizing the
technological advancements that have been made to allow for easier and more efficient application
of these methods. For this example, approximate methods of analysis (discussed below) are
utilized to determine the lateral live load distribution to the individual girders, and the girder
moments and shears are determined from a line-girder analysis.

7.1 Live Load Distribution Factors (Article 4.6.2.2)

Live loads are distributed to the individual girders according to the approximate methods specified
in Article 4.6.2.2. For cross-sections with concrete decks on multiple steel tub girders, each tub
may be assumed to carry the following number of lanes (Table 4.6.2.2.2b-1):

N L 0.425
0.05 + 0.85 +
Nb NL

where: NL = number of design lanes

Nb = number of girders in the cross-section

NL
and: 0.5   1.5
Nb

For this example:

NL 3
= = 1.5 ok
Nb 2

As the ratio of NL/Nb increases beyond the upper limit of 1.5 and fewer girders per lane are used,
the effects of torsion increase and a more refined analysis is required. Where there are no depth or
deflection limitations, the most efficient designs are those having the largest ratios of NL/Nb, or
the fewest practical number of tubs per design lane. Such designs will also require the least number
of pieces to be fabricated, shipped, and erected.

As specified in Article 6.11.2.3, there are some restrictions to the use of the above equation for
live load distribution. The satisfaction of the Article 6.11.2.3 requirements was demonstrated
previously in Section 5.3 of this example.

Also, it should be noted that shear connectors must be provided in the negative flexure regions, in
accordance with Article 6.11.10. Prototype bridges studied in the original development of the
preceding live load distribution factor for straight tub girders utilized shear connectors throughout
the negative flexure regions.

29
Distribution Factor for Three Lanes (for Strength and Service Limit State)

For the strength and service limit states, the lateral live load distribution factor for determining
bending moment and shear in each tub girder in this example is computed as follows:

 3  0.425
0.05 + 0.85   + = 1.467 lanes
2 3

Distribution Factor for Single Lane (for Fatigue Limit State)

When checking the fatigue limit state, the fatigue vehicle is placed in a single lane. Therefore, the
distribution factor for one design lane loaded is used when computing stress and shear ranges due
to the fatigue load, as specified in Article 3.6.1.4.3b.

 1  0.425
0.05 + 0.85   + = 0.900 lanes
2 1

According to Article C4.6.2.2.2b, multiple presence factors, specified in Table 3.6.1.1.2-1, are not
applicable to the preceding equation. Multiple presence factors have already been considered in
the development of the current equation.

According to Article 3.6.1.1.2, multiple presence factors should not be applied when checking the
fatigue limit state. However, the specified multiple presence factor of 1.2 for one-lane loaded
(Table 3.6.1.1.2-1) is assumed not to apply to the distribution factor equation for tub girders. Thus,
the preceding value of the distribution factor is not divided by 1.2.

7.1.1 Dynamic Load Allowance (Article 3.6.2.1)

The dynamic load allowance (IM) is an increment applied to the static wheel load to account for
wheel-load impact from moving vehicles.

For the strength and service limit states and live-load deflection checks:

IM = 33% (Table 3.6.2.1-1)

Therefore, the factor applied to the static load is to be taken as:

IM 33
Factor = 1 + = 1+ = 1.33
100 100

This factor is applied only to the design truck or tandem portion of the HL-93 design live load, or
to the truck-train portion of the special negative-moment loading (Section 7.2).

30
For the fatigue limit state checks:

IM = 15% (Table 3.6.2.1-1)

15
Factor = 1 + = 1.15
100

This factor is applied to the fatigue live load.

7.2 Analysis Results

The analysis results for a single girder are shown in the following figures. As specified in Article
6.10.1.5, the following stiffness properties were used in the analysis: 1) for loads applied to the
noncomposite section, the stiffness properties of the steel section alone, 2) for permanent loads
applied to the composite section, the stiffness properties of the long-term composite section
assuming the concrete deck to be effective over the entire span length, and 3) for transient loads
applied to the composite section, the stiffness properties of the short-term composite section
assuming the concrete deck to be effective over the entire span length. Note that for a continuous
span with a nonprismatic member, changes to the stiffness of individual sections can have a
significant effect on the analysis results. Thus, for such a span, whenever plate sizes for a particular
section are revised, it is always desirable to perform a new analysis.

NOTE: The analysis results shown herein apply to an example girder designed using earlier
versions of the AASHTO LRFD (i.e., prior to the 8th Edition). Revisions to some of the plate sizes
in this example design were necessary to provide a more reasonable b/t ratio for the bottom
(tension) flange in regions of positive flexure, and to reduce the performance ratio (i.e., demand-
to-capacity ratio) for the stiffened bottom flange in compression over the interior piers below 1.0
at the strength limit state. While it is nearly always desirable to perform a new analysis whenever
plate sizes are revised, the effect on the analysis results in this case was felt to be relatively minor
and so new analyses were not performed. The primary intent of this example is to illustrate the
proper application of the AASHTO LRFD provisions to the design of a straight continuous steel
tub-girder bridge. However, this also illustrates that a designer should always be aware of
specification changes and how they may affect a design and perhaps future load ratings.

In the first series of plots (Figure 6 and Figure 7), moment and shear envelopes due to the
unfactored dead and live loads are given. Live-load moments in regions of positive flexure and in
regions of negative flexure outside points of permanent-load contraflexure are due to the HL-93
loading (design tandem or design truck with the variable axle spacing combined with the design
lane load; whichever governs). Live-load moments in regions of negative flexure between points
of permanent-load contraflexure are the larger of the moments caused by the HL-93 loading or a
special negative-moment loading (90 percent of the effect of the truck-train specified in Article
3.6.1.3.1 combined with 90 percent of the effect of the design lane load). Live-load shears are due
to the HL-93 loading only. However, it should be noted that interior-pier reactions are to be
calculated based on the larger of the shears caused by the HL-93 loading or the special negative-
moment loading. The indicated live-load moment and shear values include the appropriate lateral
distribution factor and dynamic load allowance for the strength limit state, computed earlier

31
(Sections 7.1 and 7.1.1). DC1 is the component dead load acting on the noncomposite section and
DC2 is the component dead load acting on the long-term composite section. DW is the wearing
surface load also acting on the long-term composite section.

The second series of plots (Figure 8 and Figure 9) shows the moment and shear envelopes due to
the unfactored fatigue load specified in Article 3.6.1.4.1. The appropriate lateral distribution factor
and reduced dynamic load allowance for the fatigue limit state are included in the indicated values
(Sections 7.1 and 7.1.1).

The unfactored moments and shears resulting from the application of the construction live load
(CLL) are presented in Table 1.

7.2.1 Optional Live Load Deflection Evaluation (Article 3.6.1.3.2)

The AASHTO LRFD BDS contains provisions for optional live load deflection criteria, to be
invoked at the discretion of the Owner-agency.

The vehicular live load for checking the optional live load deflection criterion specified in Article
3.6.1.3.2 is taken as the larger of:

• The design truck alone.

• The design lane load plus 25 percent of the design truck.

These loadings are used to produce apparent live load deflections similar to those produced by
older traditional AASHTO HS20 design live loadings. It is assumed in the live load deflection
check that all design lanes are loaded and that all supporting components are assumed to deflect
equally (Article 2.5.2.6.2). For composite design, Article 2.5.2.6.2 also permits the stiffness of the
design cross-section used for the determination of the deflection to include the entire width of the
roadway and the structurally continuous portions of any railings, sidewalks, and barriers. Concrete
barriers and sidewalks, and even railings, often contribute to the stiffness of composite
superstructures at service load levels. However, inclusion of concrete items other than the deck
complicates the calculation of the composite stiffness of the superstructure and is virtually never
considered for routine bridges. Barriers are generally located at the edges of the deck, where they
tend to stiffen and draw load to the exterior girders. Thus, any beneficial stiffening of the system
tends to be counterbalanced by unequal distribution of the loading among the girders and the
associated reduction in computed deflections resulting from consideration of the barriers tends to
be negligible. Live load deflection is checked using the live load portion of the SERVICE I load
combination (Table 3.4.1-1), including the appropriate dynamic load allowance.

Because live load deflection is not anticipated to be of significant concern for this example, the
stiffness of the barriers is not included for simplicity. For this example, the maximum live load
deflection was found to occur in the center span and is:

(LL+IM) center span = 3.32 in.

32
In the absence of specific criteria, the live load deflection limits of Article 2.5.2.6.2 may be used.
Note that for steel tub girders, the provisions of Article 6.11.4 apply regarding control of permanent
deflection at the service limit state.

Figure 6 Dead- and Live-Load Moment Envelopes

33
Figure 7 Dead- and Live-Load Shear Envelopes

34
Figure 8 Fatigue Live Load Moments

35
Figure 9 Fatigue Live Load Shears

36
Table 1 Construction Live Load (CLL) Moments and Shears

37
8.0 SAMPLE CALCULATIONS

Sample calculations for two critical sections in the example bridge follow. Section 2-2 (refer to
Figure 3) represents the section of maximum positive flexure in the center span (Span 2), and
Section 2-1 represents the section at each interior pier. The calculations illustrate the application
of some of the more significant design provisions contained in Article 6.11. The calculations
include checks to be made at the service, fatigue, and strength limit states. Detailed constructability
checks are also illustrated. Web-stiffener design and the design of the stud shear connectors are
not illustrated in this example. The application of the provisions for the design of those elements
is illustrated in Design Example 1 of the NSBA Steel Bridge Design Handbook and would be
performed similarly for this example.

The calculations herein make use of the moment and shear envelopes shown in Figure 6 through
Figure 9 and the section properties calculated below. In the calculation of the major-axis bending
stresses throughout the sample calculations, compressive stresses are always shown as negative
values and tensile stresses are always shown as positive values. This convention is followed
regardless of the expected sign of the calculation result, in which the sign of the major-axis bending
moment is maintained.

8.1 Section Properties

The calculation of the section properties for Sections 2-2 and 2-1 is illustrated below. In computing
the composite section properties, the structural slab thickness, or total thickness minus the
thickness of the integral wearing surface, is used.

Compute the modular ratio, n (Article 6.10.1.1.1b):

E
n= Eq. (6.10.1.1.1b-1)
Ec

where Ec is the modulus of elasticity of the concrete determined as specified in Article 5.4.2.4. A
unit weight of 0.150 kcf is used for the concrete in the calculation of the modular ratio. The
correction factor for source of aggregate, K1, is taken as 1.0. The traditional equation for Ec for
normal-weight concrete given in Article C5.4.2.4 is used in this example.

E c = 33, 000 K1 w c1.5 f'c Eq. (C5.4.2.4-2)

E c = 33, 000 (1.0) (0.150)1.5 4.0 = 3,834 ksi

29,000
n= = 7.56, use 8.0
3,834

Therefore, n = 8 will be used in all subsequent computations.

38
8.1.1 Section 2-2: Maximum Positive Moment in Center Span

Section 2-2 located at the center of Span 2, as shown in Figure 10. For this section, the longitudinal
reinforcement is conservatively neglected in computing the composite section properties as is
typically assumed in design.

Figure 10 Sketch of Section 2-2

8.1.1.1 Effective Width of Concrete Deck (Article 6.10.1.1.1e)

As specified in Article 6.10.1.1.1e, the effective flange width is to be determined as specified in

Article 4.6.2.6. The individual webs of the tub girder must be initially considered separately since
one web is an exterior web and the other is an interior web. According to Article 4.6.2.6, for an
exterior web, the effective flange width may be taken as one-half the effective width of the adjacent
interior girder, plus the full width of the overhang.

For an interior web, the effective flange width may be taken as one-half the distance to the adjacent
girder’s nearest web plus one-half the distance to the adjacent web of the same girder.

39
For an interior web in regions of positive flexure, beff is the least of:

144.0 138.0
beff_int_web = + = 141.0 in.
2 2

For an exterior web, beff is the least of:

138.0
beff_ext_web = + 48.0 = 117.0 in.
2

The total effective flange width for the tub girder is calculated as:

beff = 141.0 + 117.0 = 258.0 in.

8.1.1.2 Elastic Section Properties for Section 2-2

The moment of inertia of a single inclined web, Iow, with respect to a horizontal axis at mid-depth
of the web (Figure 11) is computed as:

S2
Iow = Iw
S2 + 1

where: S = web slope with respect to the horizontal = 4.00

Iw = moment of inertia with respect to an axis normal to the web

 4.02  1
= 2  ( 0.625)(87.0 ) = 32,280 in.4
3
Iow
 4.0 + 1  12

Figure 11 Moment of Inertia of an Inclined Web

40
In the calculation of the section properties, d is measured vertically from a horizontal axis through
the mid-depth of the web to the centroid of each element of the tub girder.

Table 2 Section 2-2: Steel Section Properties

Component A d Ad Ad2 Io I
2 Top Flanges 1" x 18" 36.00 42.70 1,537 65,638 3.00 65,641
2 Webs 5/8" x 87" 108.75 64,560 64,560
Bottom Flange 13/16" x 98½" 80.03 -42.61 -3,410 145,303 4.40 145,307
Σ 224.78 -1,873 275,508
-8.33(1,873) = -15,602
INA = 259,906 in.4
−1,873
ds = = −8.33 in.
224.78

d Top of Steel = 43.20 + 8.33 = 51.53 in. dBot of Steel = 43.01 − 8.33 = 34.68 in.

259,906 259,906
STop of Steel = = 5, 044 in.3 SBot of Steel = = 7, 494 in.3
51.53 34.68

Table 3 Section 2-2: Composite (3n) Section Properties

Component A d Ad Ad2 Io I
Steel Section 224.78 -1,873 275,508
Concrete Slab 9ʺ x 258ʺ/24 96.75 50.20 4,857 243,814 653.1 244,467
Σ 321.53 2,984 519,975
-9.28(2,984) = -27,692
INA = 492,283 in.4
2,984
d3n = = 9.28 in.
321.53

d Top of Steel = 43.20 − 9.28 = 33.92 in. d Bot of Steel = 43.01 + 9.28 = 52.29 in.

492, 283 492, 283

STop of Steel = = 14,513 in.3 SBot of Steel = = 9, 414 in.3
33.92 52.29

41
 Table 4 Section 2-2: Composite (n) Section Properties

Component A d Ad Ad2 Io I
Steel Section 224.78 -1,873 275,508
Concrete Slab 9ʺ x 258ʺ/8 290.25 50.20 14,571 731,442 1,959 733,401
Σ 515.03 12,698 1,008,909
-24.65(12,698) = -313,006
INA = 695,903 in.4
12, 698
dn = = 24.65 in.
515.03

d Top of Steel = 43.20 − 24.65 = 18.55 in. dBot of Steel = 43.01 + 24.65 = 67.66 in.

695,903 695,903
STop of Steel = = 37,515 in.3 SBot of Steel = = 10, 285 in.3
18.55 67.66

*Note that the above computations for composite section properties consider the height of the
concrete haunch but neglect the area of the concrete haunch. Including or excluding the concrete
haunch area for section resistance is generally an Owner-agency preference. It has not been
included in this example for simplicity.

8.1.1.3 Plastic Moment Capacity for Section 2-2

Determine the plastic-moment, Mp, of the composite section using the equations provided in
Appendix D6 (Article D6.1). The longitudinal deck reinforcement is conservatively neglected. Mp
is calculated for the tub girder as follows:

Pt = Fytbttt = (50)(98.50)(0.8125) = 4,002 kips

Pw = 2FywDtw = (2)(50)(87.00)(0.625) = 5,438 kips

Pc = 2Fycbctc = (2)(50)(18.0)(1.0) = 1,800 kips

Ps = 0.85fc′beffts = (0.85)(4.0)(258.0)(9.0) = 7,895 kips

Is Pt + Pw ≥Pc + Ps ?

Pt + Pw = 9,440 kips < < Pc + Ps = 9,695 kips; Therefore, PNA is in the top flange, use
Case II in Table D6-1.

t c  Pw + Pt − Ps 
y=  + 1
2 Pc 

42
 1.0  5, 438 + 4, 002 − 7,895 
y= + 1
2  1,800 
= 0.93 in. downward from the top of the top flange

Pc  2
Mp = y + ( t c − y )  +  Ps d s + Pw d w + Pt d t 
2

2t c  

Calculate the distances from the PNA to the centroid of each element:

d t = 84.4 + 1.0 − 0.93 + 0.4063 = 84.88 in.

84.4
dw = + 1.0 − 0.93 = 42.27 in.
2

9.0
ds = + 3.5 − 1.0 + 0.93 = 7.93 in.
2

Calculate Mp:

 1,800 
Mp =   ( 0.93) + (1.0 − 0.93)  +
2 2

 2 (1.0 ) 
( 7,895 )( 7.93) + ( 5,438 )( 42.27 ) + ( 4,002 )(84.88 ) 

Mp = 632,944 kip-in

Mp = 52,745 kip-ft

8.1.1.4 Yield Moment for Section 2-2

Calculate the yield moment My of the composite section using the equations provided in Appendix
D6 (Article D6.2.2). My is taken as the sum of the factored moments at the strength limit state
applied separately to the steel, long-term, and short-term composite sections to cause first yield in
either steel flange. Flange lateral bending is to be disregarded in the calculation.

M D1 M D2 M AD
Fy = + + Eq. (D6.2.2-1)
SNC SLT SST

where MD1, MD2, and MAD are the moments applied to the steel, long-term and short-term
composite sections, respectively, factored by  and the corresponding load factors.

43
Solve for MAD (bottom flange governs by inspection):

1.25 (10,110 )(12 ) 1.25 (1,594 )(12 ) + 1.50 (1,532 )(12 ) M AD 

50 = 1.0  + + 
 7,494 9,414 10,285 

MAD = 249,871kip-in = 20,823kip-ft

My = MD1 + MD2 + MAD Eq. (D6.2.2-2)

My = 1.0[1.25(10,110) + 1.25(1,594) + 1.50(1,532) + 20,823]

My = 37,751 kip-ft

8.1.2 Section 2-1: Maximum Negative Moment at Interior Support

Section 2-1 is at the interior support and is shown in Figure 12.

Figure 12 Sketch Showing Section 2-1

8.1.2.1 Effective Width of Concrete Deck (Article 6.10.1.1.1e)

The effective flange width for Section 2-1 is calculated using the procedures discussed previously
for Section 2-2.

44
For an interior web in regions of negative flexure, beff is the least of:

144.0 138.0
beff_int_web = + = 141.0 in.
2 2

For an exterior web, beff is the least of:

138.0
beff_ext_web = + 48.0 = 117.0 in.
2

The total effective flange width for the tub girder is calculated as:

beff = 141.0 + 117.0 = 258.0 in.

8.1.2.2 Minimum Negative Flexure Concrete Deck Reinforcement

To control concrete deck cracking in regions of negative flexure, Article 6.10.1.7 specifies that the
total cross-sectional area of the longitudinal reinforcement must not be less than 1 percent of the
total cross-sectional area of the deck. The minimum longitudinal reinforcement must be provided
wherever the longitudinal tensile stress in the concrete deck due to either the factored construction
loads or Load Combination Service II exceeds fr, where fr is the modulus of rupture of the concrete
determined as specified in Article 6.10.1.7 and  is taken as 0.90. Article 6.10.1.7 further specifies
that the reinforcement is to have a specified minimum yield strength not less than 60 ksi and the
size should not exceed No. 6 bars. The reinforcement should be placed in two layers uniformly
distributed across the deck width, and two-thirds should be placed in the top layer. The individual
bars should be spaced at intervals not exceeding 12 inches.

Article 6.10.1.1.1c states that for calculating stresses in composite sections subjected to negative
flexure at the strength limit state, the composite section for both short-term and long-term moments
is to consist of the steel section and the longitudinal reinforcement within the effective width of
the concrete deck. Referring to the cross-section shown in Figure 1:

Adeck = (entire width of 9-inch-thick deck) + (triangular portion of overhang)

 1 3.5 + 0.5  30 2  
( 43.0 ) + 2  
9.0
A deck =  4.0 −   = 33.17 ft = 4,777 in.
2 2

12 12  2  12  

0.01(4,777) = 47.77 in.2

47.77
= 1.11 in.2 ft = 0.093 in.2 in.
43.0

0.093(258.0) = 23.99 in.2

45
For the top layer, alternate #5 bars @ 12 inches and #6 bars @ 12 inches, and in the bottom layer
use #4 bars @ 6 inches. Therefore, the total area of steel in the given effective width of concrete
deck is:

 258.0 
AS = ( 0.31 + 0.44 + 0.40 )   = 24.73 in.  23.99 in.
2 2

 12 

0.31 + 0.44 2
Also, two-thirds of the reinforcement is in the top layer: = 0.65 
1.15 3

For the purposes of this example, the longitudinal reinforcement in the two layers is assumed to
be combined into a single layer placed at the centroid of the two layers (with each layer also
including the assumed transverse deck reinforcement). From separate calculations, the centroid of
the two layers is computed to be 4.54 inches from the bottom of the concrete deck.

For members with shear connectors provided throughout their entire length that also satisfy the
minimum reinforcement requirements of Article 6.10.1.7, dead load and live load stresses and live
load stress ranges for the fatigue and service limit state design at all sections in the member due to
loads applied to the composite section may be computed using the short-term or long-term
composite section, as appropriate, assuming the concrete deck to be effective for both positive and
negative flexure (Articles 6.6.1.2.1 and 6.10.4.2.1). Article 6.10.4.2.1 also requires that the
maximum longitudinal tensile stress in the concrete deck at the service limit state (i.e., due to the
Service II loads) not exceed two times the modulus of rupture of the concrete, fr. Therefore, section
properties for the short-term and long-term composite section, including the concrete deck but
neglecting the longitudinal reinforcement, are also calculated.

8.1.2.3 Elastic Section Properties for Section 2-1

Calculations for the elastic section properties of Section 2-1 are shown in Table 5 through Table
8. The section properties include the contribution of the bottom flange longitudinal stiffener (WT
12 x 42).

Table 5 Section 2-1: Steel Section Properties

Component A d Ad Ad2 Io I
2 Top Flanges 2¾" x 30" 165.00 43.58 7,191 313,371 104.0 313,475
2 Webs 5/8" x 87" 108.75 64,560 64,560
Bottom Flange 1¾" x 98½" 172.38 -43.08 -7,426 319,918 43.99 319,962
Stiffener WT12x42 12.40 -33.07 -410.1 13,561 166.0 13,727
Σ 458.53 -645.1 711,724
-1.41(645.1) = -910
INA = 710,814 in.4

46
 −645.1
ds = = −1.41 in.
458.53

d Top of Steel = 44.95 + 1.41 = 46.36 in. dBot of Steel = 43.95 − 1.41 = 42.54 in.

710,814 710,814
STop of Steel = = 15,332 in.3 SBot of Steel = = 16, 709 in.3
46.36 42.54

Table 6 Section 2-1: Composite Section Properties with Longitudinal Steel Reinforcement

Component A d Ad Ad2 Io I
Steel Section 458.53 -645.1 711,724
Long. Reinforcement 24.73 50.24 1,242 62,420 62,420
Σ 483.26 596.9 774,144
-1.24(596.9) = -740.2
INA = 773,404 in.4
596.9
d reinf = = 1.24 in.
483.26

d Top of Steel = 44.95 − 1.24 = 43.71 in. dBot of Steel = 43.95 + 1.24 = 45.19 in.

773, 404 773, 404

STop of Steel = = 17, 694 in.3 SBot of Steel = = 17,114 in.3
43.71 45.19

Table 7 Section 2-1: Composite (3n) Section Properties

Component A d Ad Ad2 Io I
Steel Section 458.53 -645.1 711,724
Concrete Slab 9” x 258”/24 96.75 50.2 4,857 243,814 653 244,467
Σ 555.28 4,212 956,191
-7.59(4,212) = -31,969
INA = 924,222 in.4
4, 212
d3n = = 7.59 in.
555.28

d Top of Steel = 44.95 − 7.59 = 37.36 in. dBot of Steel = 43.95 + 7.59 = 51.54 in.

924, 222 924, 222

STop of Steel = = 24, 738 in.3 SBot of Steel = = 17,932 in.3
37.36 51.54

47
 Table 8 Section 2-1: Composite (n) Section Properties

Component A d Ad Ad2 Io I
Steel Section 458.53 -645.1 711,724
Concrete Slab 9” x 258”/8 290.3 50.2 14,573 731,568 1,959 733,527
Σ 748.83 13,928 1,445,251
-18.60(13,928) = -259,061
INA = 1,186,190 in.4
13,928
dn = = 18.60 in.
748.83

d Top of Steel = 44.95 − 18.60 = 26.35 in. dBot of Steel = 43.95 + 18.60 = 62.55 in.

1,186,190 1,186,190
STop of Steel = = 45, 017 in.3 SBot of Steel = = 18,964 in.3
26.35 62.55

8.2 Girder Constructability Check: Section 2-2 (Positive Moment, Span 2)

Article 6.11.3 directs the Engineer to Article 6.10.3 for the constructability checks of tub girders.
For critical stages of construction, the provisions of Articles 6.10.3.2.1 through 6.10.3.2.3 are to
be applied to the top flanges of the tub girder. The noncomposite bottom tub flange in compression
or tension is to satisfy the requirements specified in Article 6.11.3.2. Web shear is to be checked
in accordance with Article 6.10.3.3 with the shear taken along the slope of the web in accordance
with the provisions of Article 6.11.9.

For this example, a deck placement sequence is not investigated. The demonstration of the
constructability checks for a deck placement sequence is provided in Design Example 1 of the
NSBA Steel Bridge Design Handbook. In the absence of a deck placement sequence, the weight
of the concrete deck is assumed placed in a single stage in this example. Although not illustrated
in this example, Article 6.10.3.4.1 requires that sections in positive flexure that are composite in
the final condition, but are noncomposite during construction, be investigated for flexure according
to the provisions of Article 6.10.3.2 during the various stages of the deck placement. Furthermore,
wind loads during construction are not considered in this example (refer again to Design Example
1 for an illustration of these checks).

Calculate the factored maximum flexural stresses in the flanges of the steel section resulting from
the application of steel self-weight and the assumed single placement of the concrete deck (DC1).
As specified in Article 6.10.1.6, for design checks where the flexural resistance is based on lateral
torsional buckling, fbu is to be determined as the largest value of the compressive stress throughout
the unbraced length in the flange under consideration, calculated without consideration of flange
lateral bending. For design checks where the flexural resistance is based on yielding, flange local
buckling or web bend buckling, fbu may be determined as the stress at the section under
consideration. From Figure 2, brace points adjacent to Section 2-2 are located at intervals of 15.875
feet, and the largest stress occurs within this unbraced length. As mentioned in Article C6.11.3.2,
top lateral bracing attached to the flanges at points where only struts exist between the flanges may
be considered as brace points at the discretion of the Engineer. In the case of this design example,

48
which features a full-length top flange lateral bracing system, it is reasonable to consider both the
struts with internal cross-frames and the alternating struts without internal cross-frames as brace
points for the top flanges. As discussed previously, the  factor is taken equal to 1.0 in this
example. Therefore,

For Strength I:

  M DC1
General: (f bu )DC1 =
Snc
1.0(1.25)(10,110)(12)
Top flange: (f bu ) DC1 = = −30.07 ksi
5,044
1.0(1.25)(10,110)(12)
Bot. flange: (f bu ) DC1 = = 20.24 ksi
7,494

In addition to the steel, permanent metal deck forms, and concrete self-weight loads, it is prudent
to assume a construction live loading (CLL) on the structure during placement of the concrete
deck, as discussed in Section 6.2.3. In the Strength I load combination; a load factor of 1.5 is
applied to all construction loads, in accordance with Article 3.4.2. Therefore,

1.0(1.5)(553)(12)
Top flange: (f bu )CLL = = −1.97 ksi
5,044
1.0(1.5)(553)(12)
Bot. flange: (f bu )CLL = = 1.33 ksi
7,494

Top flange: fbu = -30.07 + (-1.97) = -32.04ksi

Bot. flange: fbu = 20.24+1.33 = 21.57ksi

Although not included in this example in the interest of brevity, the special load combination
specified in Article 3.4.2.1 must also be considered in the design checks for the deck placement
sequence (see Section 6.3).

8.2.1 Top Flange Lateral Bending due to Horizontal Component of Web Shear

The change in the horizontal component of the web shear in the inclined web along the span acts
as a lateral force in the flanges of the tub girder. Under initial noncomposite dead load DC1, the
lateral force due to shear is assumed to be distributed to the top flanges of the open tub girder.
Recent research has suggested that the top and bottom flanges do not equally resist the lateral force
due to the horizontal component of the web shear, as has been generally assumed in past practice
Fan and Helwig [12] suggest that, with the exception of the girder self-weight, the entire lateral
force should be assumed to act on the top flanges. To simplify the calculations for this example, it
will conservatively be assumed that the entire DC1 horizontal component of web shear is applied
to the top flanges. The change in vertical shear force, equal to the lateral load on the top flanges,

49
is constant and is equal to the change in DC1 shear force in the girder measured at adjacent supports
divided by the span length.

The change in DC1 girder shear over the length of the Span 2 is:

 −540 + 540 
ΔVV =  = 3.93 kip/ft
275

The shear force used above is the total for the girder (2 webs). Therefore, the horizontal component
of the web shear per top flange is:

1 1
ΔVH = ΔVV tan(θ WEB ) = (3.93)(0.25) = 0.49 kip/ft
2 2

Assuming the flange is continuous and that the adjacent unbraced lengths are approximately equal,
the lateral bending moment due to a statically equivalent uniformly distributed lateral load may be
estimated as follows, similar to Equation C6.10.3.4.1-1, where s is the brace spacing:

ΔVHs 2 (0.49)(15.875)2
M LAT = = = 10.29 kip-ft
12 12

The section modulus of the 1.0 in. x 18 in. top flange about a vertical axis through the web is:

(1.0)(18) 2
Sf = = 54.00 in.3
6

The Strength I lateral bending stress due to the horizontal component of web shear, including the
dead load factor of 1.25, is then computed as:

M LAT 12(1.25)(10.29)
f LAT = = = 2.86 ksi
Sf 54.00

8.2.2 Top Flange Lateral Bending due to Deck Overhang Loads

Assume the deck overhang bracket configuration shown in Figure 13 with the bracket extending
to the bottom flange:

50
 Figure 13 Deck Overhang Bracket Loading

Although the brackets are typically spaced at 3 to 4 feet along the exterior girder, all bracket loads
except for the finishing machine load are assumed to be applied uniformly. For this example, the
bracket is assumed to extend near the edge of the deck overhang. Therefore, it is assumed that half
the deck overhang weight is placed on the exterior girder web and half the weight is placed on the
overhang brackets. Conservatively, one-half the deck haunch weight will be included in the total
overhang weight. Therefore:

Deck Overhang Weight:

 18/12  9.5   13 − 1.0  18/12  1  18/12   ( 3.5 + 0.5 )  

P = 0.5*150  4 −  +  + 4−    = 290 lbs/ft
 2  12   12  2  2  2  12  

Construction loads, or dead loads and temporary loads that act on the overhang only during
construction, are assumed as follows:

Overhang deck forms: P = 40 lbs/ft

Screed rail: P = 85 lbs/ft
Railing: P = 25 lbs/ft
Walkway: P = 125 lbs/ft
Finishing machine: P = 3000 lbs

The force imposed by the weight of the finishing machine is estimated as one-half of the total
finishing machine truss weight, plus additional load to account for the weight of the engine, drum
and operator assumed to be located on one side of the truss.

51
The lateral force on the top flange, due to the vertical load on the overhang brackets, is computed
by summation of the moments about the web-bottom flange junction (Figure 13):

FLAT (84.40) - P(69.25) = 0

FLAT = (0.819) P

In the absence of a more refined analysis, the equations given in Article C6.10.3.4.1 may be used
to estimate the maximum flange lateral bending moments in the discretely braced compression
flange due to the lateral bracket forces. Assuming the flange is continuous with the adjacent
unbraced lengths and that the adjacent unbraced lengths are approximately equal, the lateral
bending moment due to a statically equivalent uniformly distributed lateral bracket force may be
estimated as:

F L2b
M = Eq. (C6.10.3.4.1-1)
12

The lateral bending moment due to a statically equivalent concentrated lateral bracket force
assumed placed at the middle of the unbraced length may be estimated as:

P Lb
M = Eq. (C6.10.3.4.1-2)
8

In the Strength I load combination, a load factor of 1.5 is applied to all construction loads (Article
3.4.2.1). The lateral bending stress in the bottom flange will be small as compared to the top flange;
therefore, bottom flange calculations are not shown for this particular example.

For Strength I:

Dead loads: P = 1.0 1.25(290) + 1.5(40 + 85 + 25 + 125)  = 775.0 lbs/ft

F = F = (0.819)P = (0.819)(775.0) = 635 lbs/ft

F L2b 0.635 (15.875 )

2

M = = = 13.33 kip − ft
12 12

M 13.33(12)
Top flange: f = = = 2.96 ksi
S (1.0)(18) 2 6

Finishing machine: P = 1.0 1.5 ( 3,000 )  = 4,500 lbs

F = P = (0.819)P = (0.819)(4,500) = 3,686 lbs

52
 P L b 3.686 (15.875 )
M = = = 7.31 kip − ft
8 8

M 7.31(12)
Top flange: f = = = 1.62 ksi
S (1.0)(18) 2 6

Deck Overhang Total:

f = 2.96 +1.62 = 4.58 ksi

8.2.3 Top Flange Lateral Bending Amplification

As specified in Article 6.10.1.6, for design checks where the flexural resistance is based on lateral
torsional buckling, the stress, f, is to be determined as the largest value of the stress due to lateral
bending throughout the unbraced length in the flange under consideration. For design checks
where the flexural resistance is based on yielding or flange local buckling, f may be determined
as the stress at the section under consideration. For simplicity in this example, the largest value of
f within the unbraced length will conservatively be used in all design checks. f is to be taken as
positive in sign in all resistance equations. The unbraced length, Lb, for Section 2-2 is equal to
15.875 feet (Figure 2).

According to Article 6.10.1.6, lateral bending stresses determined from a first-order analysis may
be used in discretely braced compression flanges for which:

Cb R b
L b  1.2L p Eq. (6.10.1.6-2)
f bu Fyc

Lp is the limiting unbraced length specified in Article 6.10.8.2.3 determined as:

E
L p = 1.0rt Eq. (6.10.8.2.3-4)
Fyc

where rt is the effective radius of gyration for lateral torsional buckling specified in Article
6.10.8.2.3 determined as:

b fc
rt = Eq. (6.10.8.2.3-9)
 1 Dc t w 
12 1 + 
 3 b fc t fc 

53
For the steel section, the depth of the web in compression in the elastic range, Dc, at Section 2-2 is
computed along the web as follows:

Note that for the steel section only: dTOP OF STEEL = 51.53 in.

S2 + 1
Dc = (d TOP OF STEEL − t f )
S2

42 + 1
Dc = (51.53 − 1.0)
42

Dc = 52.09in.

It should be noted that values of Dc and D are taken as distances along the web, in accordance with
Article 6.11.2.1.1. Therefore,

18
rt = = 4.10 in.
 1 52.09(0.625) 
12 1 + 
 3 18(1.0) 

1.0(4.10) 29,000
Lp = = 8.23 ft
12 50

Cb is the moment gradient modifier specified in Article 6.10.8.2.3, and may conservatively be
taken equal to 1.0. According to Article 6.10.1.10.2, the web load-shedding factor, Rb, is to be
taken equal to 1.0 when checking constructability. Finally, fbu is the largest value of the factored
compressive stress throughout the unbraced length in the flange under consideration, calculated
without consideration of flange lateral bending. In this case, use fbu = -33.90 ksi, as computed
earlier for the Strength I load combination. Therefore:

1.0(1.0)
1.2 ( 8.23) = 12.34 ft  L b = 15.875 ft
−32.04 50

Because the Equation 6.10.1.6-2 is not satisfied, Article 6.10.1.6 requires that second-order elastic
compression-flange lateral bending stresses be determined. The second-order compression-flange
lateral bending stresses may be determined by amplifying first-order values (i.e. f1) as follows:

 
 0.85 
f = f 1  f 1 Eq. (6.10.1.6-4)
 1 − f bu 
 
 Fcr 

54
or:

f = (AF)f 1  f 1

where AF is the amplification factor and Fcr is the elastic lateral torsional buckling stress for the
flange under consideration specified in Article 6.10.8.2.3 determined as:

Cb R b π 2 E
Fcr = 2 Eq. (6.10.8.2.3-8)
 Lb 
 
 rt 

1.0(1.0)π 2 (29,000)
Fcr = 2
= 132.6 ksi
 15.875(12) 
 
 4.10 

The amplification factor is then determined as follows:

0.85
AF = = 1.12  1.0 ok
 −32.04 
1 − 
 132.6 

The above equation for the amplification factor conservatively assumes an elastic effective length
factor for lateral torsional buckling equal to 1.0.

Therefore, the total flange stress due to lateral bending, including the amplification factor is:

flat = (AF)[(flat)WEB SHEAR + (f1)OVERHANG] = (1.12)[2.86 + 4.58] = 8.33 ksi

Note that first or second-order flange lateral bending stresses, as applicable, are limited to a
maximum value of 0.6Fyf according to Eq. (6.10.1.6-1).

(0.6)Fyf = (0.6)(50) = 30 ksi > 8.33ksi ok (Ratio = 0.278)

8.2.4 Flexure (Article 6.11.3.2)

Article 6.11.3.2 directs the engineer to the provisions of Article 6.10.3.2 for the flange
constructability checks. Article 6.10.3.2.1 requires that discretely braced flanges in compression
satisfy the following requirements, except that for slender-web sections, Eq. (6.10.3.2.1-1) need
not be checked when f is equal to zero.

f bu + f  f R h Fyc Eq. (6.10.3.2.1-1)

55
 1
f bu + f  f Fnc Eq. (6.10.3.2.1-2)
3

fbu  f Fcrw Eq. (6.10.3.2.1-3)

Article 6.11.3.2 requires that the noncomposite tub flange (bottom flange) in tension satisfy:

f bu  f R h Fyf Δ Eq. (6.11.3.2-3)

where: f = resistance factor for flexure = 1.0 (Article 6.5.4.2)

Rh = hybrid factor specified in Article 6.10.1.10.1 (= 1.0 at homogeneous Section 2-2)
Fcrw= nominal elastic bend-buckling resistance for webs determined as specified in
Article 6.10.1.9
Fnc = nominal flexural resistance of the compression flange determined as specified in
Article 6.10.8.2 (i.e., local or lateral torsional buckling resistance, as applicable).
The provisions of Article A6.3.3 are not to be used to determine the lateral torsional
buckling resistance of top flanges of tub girders, per Article 6.11.3.2.
 = a factor dependent on the St. Venant torsional shear stress in the bottom flange. St.
Venant torsional shear stress will be addressed later in this example.

First, determine if the noncomposite Section 2-2 is a compact or noncompact web section
according to Eq. (6.10.6.2.3-1), or alternatively, see Table C6.10.1.10.2-2:

2Dc
  rw
tw Eq. (6.10.6.2.3-1)

where:

E  5.0  E E
4.6   rw =  3.1 +   5.7
Fyc  a wc  Fyc Fyc Eq. (6.10.6.2.3-3)

2Dc t w
a wc =
bfc t fc Eq. (6.10.6.2.3-4)

2Dc 2(52.09)
= = 166.7
tw 0.625

E 29, 000
4.6 = 4.6 = 111
Fyc 50

E 29, 000
5.7 = 5.7 = 137
Fyc 50

56
 2(52.09)(0.625)
a wc = = 3.62
18(1.0)

 5.0  29, 000

111   rw =  3.1 +  = 107.9  137
 3.62  50

2Dc
 rw = 111  = 166.7
tw

Therefore, the noncomposite Section 2-2 is a slender-web section. As a result, for the top flanges,
Eq. (6.10.3.2.1-1) must be checked since f is not zero.

8.2.4.1 Top Flange - Local Buckling Resistance (Article 6.10.8.2.2)

Determine the slenderness ratio of the top flanges:

b fc
f = Eq. (6.10.8.2.2-3)
2t fc

18
λf = = 9.00
2 (1.0 )

Determine the limiting slenderness ratio for a compact flange (alternatively see Table C6.10.8.2.2-
1):

E
 pf = 0.38 Eq. (6.10.8.2.2-4)
Fyc

29, 000
 pf = 0.38 = 9.15
50

Since f < pf, the top flanges are compact. Therefore:

Fnc = R b R h Fyc Eq. (6.10.8.2.2-1)

As specified in Article C6.10.3.2.1, when computing Fnc for constructability, the web load-
shedding factor Rb is to be taken equal to 1.0 because the flange stress is always limited to the web
bend-buckling stress according to Eq. (6.10.3.2.1-3) (see Article C6.10.3.2.1). Therefore,

Fnc = (1.0)(1.0)(50) = 50.00 ksi

57
8.2.4.2 Top Flange - Lateral Torsional Buckling Resistance

The limiting unbraced length, Lp, was computed earlier to be 8.23 feet. The effective radius of
gyration for lateral torsional buckling, rt, for the noncomposite Section 2-2 was also computed
earlier to be 4.10 inches (Section 8.2.3).

Determine the limiting unbraced length, Lr:

E
L r = π rt Eq. (6.10.8.2.3-5)
Fyr

where Fyr is the compression flange stress at the onset on nominal yielding, including residual
stress effects, and is to be taken as the smaller of 0.7Fyc and Fyw, but not less than 0.5Fyc. Since Fyc
and Fyw are both equal to 50 ksi,

Fyr = 0.7(50) = 35 ksi

π (4.10) 29,000
Lr = = 30.90 ft
12 35

Since Lp = 8.23feet < Lb = 15.875 feet < Lr = 30.90feet,

  F  L − L p  
Fnc = Cb 1 − 1 − yr  b   R b R h Fyc  R b R h Fyc Eq. (6.10.8.2.3-2)
  R h Fyc 
 L r − L p  

As discussed previously, the moment-gradient modifier, Cb, is taken equal to 1.0. Therefore,

  35.0   15.875 − 8.23  

Fnc = 1.0 1 − 1 −    (1.0 ) (1.0)(50) = 44.94 ksi  1.0(1.0)(50) = 50 ksi ok
  1.0(50)   30.90 − 8.23 

8.2.4.3 Web Bend-Buckling Resistance (Article 6.10.1.9)

Determine the nominal elastic web bend-buckling resistance at Section 2-2 according to the
provisions of Article 6.10.1.9.1 as follows:

0.9Ek  Fyw 
Fcrw = 2
 min  R h Fyc ,  Eq. (6.10.1.9.1-1)
D  0.7 
 
 tw 

where:

58
 9
k= Eq. (6.10.1.9.1-2)
( Dc D)
2

9
k= = 25.11
( 52.09 87.0 )
2

Therefore,

0.9(29,000)(25.11)
Fcrw = 2
= 33.82 ksi  R h Fyc = 50 ksi ok
 87.0 
 
 0.625 

8.2.4.4 Top Flange Constructability Checks

Now that all the required information has been assembled, check the requirements of Article
6.10.3.2.1:

For yielding:

f bu + f  f R h Fyc Eq. (6.10.3.2.1-1)

f bu + f = −32.04 ksi + 8.33 ksi = 40.37 ksi
f R h Fyc = 1.0(1.0)(50.00) = 50.00 ksi
40.37 ksi  50.00 ksi ok ( Ratio = 0.807 )
1
f bu + f  f Fnc Eq. (6.10.3.2.1-2)
3

For local buckling:

1 8.33
f bu + f = −32.04 ksi + ksi = 34.82 ksi
3 3
f Fnc = 1.0(50.00) = 50.00 ksi
34.82 ksi  50.00 ksi ok (Ratio = 0.696)

59
For lateral torsional buckling:

1 8.33
f bu + f = −32.04 ksi + ksi = 34.82 ksi
3 3
f Fnc = 1.0(44.94) = 44.94 ksi
34.82 ksi  44.94 ksi ok (Ratio = 0.775)

For web bend buckling:

fbu  f Fcrw Eq. (6.10.3.2.1-3)

f Fcrw = 1.0(33.82) = 33.82 ksi
−32.04 ksi  33.82 ksi ok (Ratio = 0.947)

8.2.4.5 Bottom Flange Constructability Checks

Noncomposite tub flanges (bottom flanges) in tension, must satisfy the following constructability
requirement:

f bu  f R h Fyf Δ Eq. (6.11.3.2-3)

where:
2
f 
Δ = 1− 3 v  Eq. (6.11.3.2-4)
F
 yf 

The term fv is the factored St. Venant torsional shear stress in the flange at the section under
consideration. However, in accordance with Article C6.11.2.3, if the provisions of Article 6.11.2.3
are satisfied, shear due to St. Venant torsion and secondary distortional bending stress effects may
be neglected if the width of the tub flange does not exceed one-fifth the effective span defined in
Article 6.11.1.1. For continuous spans, the effective span length is to be taken as the distance
between points of permanent load contraflexure, or between a simple support and a point of
permanent load contraflexure, as applicable. Therefore, Span 2 has an effective span length of 145
feet. One-fifth of the effective span length is equal to 29 feet, which is much greater than the
bottom flange width of 8.208 feet. Therefore, the St. Venant torsional shear stresses can be
neglected for this case (fv = 0), and:

2
 0 
Δ = 1 − 3   = 1.0
 50 

Consideration of St. Venant torsional shear stresses is illustrated in the horizontally curved tub
girder design example NSBA’s Steel Bridge Design Handbook: Design Example 5: Three-Span
Continuous Horizontally Curved Composite Steel Tub-Girder Bridge [7] .

60
The longitudinal flange stress, calculated previously, is:

f bu = 21.57 ksi
f R h Fyf Δ = 1.0(1.0)(50)(1.0) = 50.00 ksi
21.57 ksi  50.00 ksi ok (Ratio = 0.431)

Although the checks are illustrated here for completeness, the bottom flange will typically not
control at the positive moment location.

8.2.5 Shear (Article 6.10.3.3)

Article 6.10.3.3 requires that interior panels of stiffened webs satisfy the following requirement:

Vu  v Vcr Eq. (6.10.3.3-1)

where: v = resistance factor for shear = 1.0 (Article 6.5.4.2)

Vu = shear in the web at the section under consideration due to the factored permanent
loads and factored construction loads applied to the noncomposite section
Vcr = shear-yield or shear-buckling resistance determined from Eq. (6.10.9.3.3-1)

Only the interior panels of stiffened webs are checked because the shear resistance of the end panel
of stiffened webs and the shear resistance of unstiffened webs is already limited to the shear yield-
or shear-buckling resistance at the strength limit state.

For this example, the web is unstiffened in the positive flexure regions. Therefore, the
constructability check for shear is not required at this section. This check is demonstrated,
however, for the stiffened web at Section 2-2.

8.2.6 Concrete Deck (Article 6.10.3.2.4)

Generally, the entire deck is not placed in a single pour. Typically, for continuous span bridges,
the positive flexure regions are placed first. Thus, positive flexure regions may become composite
prior to casting the other ections of the bridge. As the deck placement operation progresses, tensile
stresses can develop in previously cast regions that will exceed the allowable rupture strength (fr)
of the hardened concrete deck. When cracking is predicted, longitudinal deck reinforcing as
specified in Article 6.10.1.7 is required to control the cracking. Otherwise, alternative deck casting
sequences may be employed to minimize the anticipated stresses to acceptable levels. This check
is illustrated in Design Example 1 of the NSBA Steel Bridge Design Handbook.

8.3 Girder Service Limit State Check: Section 2-2 (Positive Moment, Span 2)

Article 6.11.4 directs the Engineer to Article 6.10.4, which contains provisions related to the
control of elastic and permanent deformations at the service limit state. For the sake of brevity,
only the calculations pertaining to permanent deformations will be presented for this example.

61
8.3.1 Permanent Deformations (Article 6.10.4.2)

Article 6.10.4.2 contains criteria intended to control permanent deformations that would impair
rideability. As specified in Article 6.10.4.2.1, these checks are to be made under the Service II load
combination.

Article 6.10.4.2.2 requires that flanges of composite sections satisfy the following requirements:

Top flange of composite sections: f f  0.95R h Fyf Eq. (6.10.4.2.2-1)

f
Bottom flange of composite sections: f f +  0.95R h Fyf Eq. (6.10.4.2.2-2)
2

The term ff is the flange stress at the section under consideration due to the Service II loads
calculated without consideration of flange lateral bending. The f term, the flange lateral bending
stress, in Eq. (6.10.4.2.2-2) is to be taken equal to zero for tub girders, in accordance with Article
6.11.4. A resistance factor is not included in these equations because Article 1.3.2.1 specifies that
the resistance factor be taken equal to 1.0 at the service limit state.

With the exception of composite sections in positive flexure in which the web satisfies the
requirement of Articles 6.11.2.1.2 (i.e., D/tw  150), web bend-buckling of all sections under the
Service II load combination is to be checked as follows:

fc  Fcrw Eq. (6.10.4.2.2-4)

The term fc is the compression-flange stress at the section under consideration due to the Service
II loads calculated without consideration of flange lateral bending, and Fcrw is the nominal elastic
bend-buckling resistance for webs determined as specified in Article 6.10.1.9. Because Section 2-
2 is a composite section subject to positive flexure satisfying Article 6.11.2.1.2, Eq. (6.10.4.2.2-4)
need not be checked. An explanation as to why these particular sections are exempt from the above
web bend-buckling check is given in Article C6.10.1.9.1.

It should be noted that in accordance with Article 6.11.4, redistribution of negative moment due
to the Service II loads at the interior-pier sections in continuous-span flexural members using the
procedures specified in Appendix B6 is not to be applied to tub girder sections. The applicability
of the Appendix B6 provisions to tub girder sections has not been demonstrated, hence the
procedures are not permitted for the design of tub girder sections.

Check the flange stresses due to the Service II loads at Section 2-2.  is specified to always equal
1.0 at the service limit state (Article 1.3.2):

0.95R h Fyf = 0.95(1.0)(50) = 47.50 ksi

Top flange: f f  0.95R h Fyf Eq. (6.10.4.2.2-1)

62
 1.0(10,110) 1.0(1,594 + 1,532) 1.3(9,396) 
f f = 1.0  + + 12 = −30.54 ksi
 5,044 14,513 37,515 
−30.54 ksi  47.50 ksi ok (Ratio = 0.643)

f
Bottom flange: f f +  0.95R h Fyf Eq. (6.10.4.2.2-2)
2

1.0(10,110) 1.0(1,594 + 1,532) 1.3(9,396)  0

f f = 1.0  + +  12 + = 34.43 ksi
 7,494 9,414 10,285  2
34.43 ksi  47.50 ksi ok (Ratio = 0.725)

8.4 Girder Fatigue and Fracture Limit State Check: Section 2-2 (Span 2)

8.4.1 Fatigue (Article 6.10.5)

Article 6.11.5 directs the Engineer to Article 6.10.5, where details on tub girder section flexural
members must be investigated for fatigue as specified in Article 6.6.1. Either the Fatigue I or
Fatigue II load combination specified in Table 3.4.1-1, along with the fatigue live load specified
in Article 3.6.1.4, are to be employed for checking load-induced fatigue in tub girder sections. The
Fatigue I load combination is to be used in combination with design checks for infinite fatigue life.
The Fatigue II load combination is to be used in combination with design checks for finite fatigue
life.

One additional fatigue limit state requirement specified for certain tub-girder sections is related to
longitudinal warping and transverse bending stresses due to cross-section distortion. When tub
girders are subjected to eccentric loads (i.e., torsion), their cross-sections become distorted,
resulting in secondary bending stresses. Loading the opposite side of the bridge will produce a
stress reversal, and possible fatigue concerns for sections subjected to significant torsion.
Therefore, according to Article 6.11.5, longitudinal warping stresses and transverse bending
stresses due to cross-section distortion are to be considered for:

• Single tub girders in straight or horizontally curved bridges;

• Multiple tub girders in straight bridges that do not satisfy requirements of Article
6.11.2.3;

• Multiple tub girders in horizontally curved bridges; or

• Any single or multiple tub girder with a bottom flange that is not fully effective
according to the provisions of Article 6.11.1.1.

In cases where these stresses are to be considered, the stress range due to longitudinal warping
resulting from cross-section distortion is to be considered in checking the fatigue resistance of the
base metal at all details in the tub girder according to the provisions of Article 6.6.1. Longitudinal
warping stresses are considered additive to the longitudinal major-axis bending stresses.

63
The transverse bending stress range (i.e., the out-of-plane bending stress range due to box cross-
section distortion) is to be considered separately in evaluating the fatigue resistance of the base
metal adjacent to flange-to-web fillet welds and adjacent to the termination of fillet welds
connecting transverse elements to the webs and tub flanges. Possible steps for reducing the
transverse bending stress range, where necessary, are discussed further in Article C6.11.5. Where
force effects in the cross-frames or diaphragms are computed from a refined analysis, stress ranges
for checking load-induced fatigue and torque ranges for checking fatigue due to cross-section
distortion in cross-frame and diaphragm members, the single fatigue truck should be positioned as
specified in Article 3.6.1.4.3a, with the truck confined to a single transverse position during each
passage of the truck along the bridge (per Article C6.6.1.2.1). Transverse bending and longitudinal
warping stress ranges due to cross-section distortion can be determined using the Beam-on-Elastic
Foundation (BEF) analogy, as discussed in Article C6.11.1.1. These calculations are illustrated in
NSBA’s Steel Bridge Design Handbook: Design Example 5: Three-Span Continuous Horizontally
Curved Composite Steel Tub-Girder Bridge [7].

The tub girders in this design example do not fall under any of the categories listed above; hence,
longitudinal warping and transverse bending stresses need not be considered.

In addition to checking fatigue of the base metal at the transverse element welded connections,
there is a special fatigue requirement for the tub girder webs, with transverse stiffeners, that must
be satisfied in accordance with Article 6.10.5.3. The satisfaction of Article 6.10.5.3 is intended to
eliminate significant elastic flexing of the web due to shear, such that the member is assumed able
to sustain an infinite number of smaller loadings without fatigue cracking due to this effect. For
Article 6.10.5.3, the factored fatigue load is to be taken as the Fatigue I load combination specified
in Table 3.4.1-1, with the fatigue live load taken as specified in Article 3.6.1.4. For sections with
inclined webs, the factored shear is to be determined using Eq. (6.11.9-1). This check is not
illustrated in this design example; refer to NSBA’s Steel Bridge Design Handbook: Design
Example 1: Three-Span Continuous Straight Composite Steel I-Girder Bridge [3] for an
illustration of this check.

8.4.1.1 Fatigue in Bottom Flange

At Section 2-2, it is necessary to check the base metal at the interior cross-frame connection plate
welds to the bottom flange of the tub girder for fatigue. This detail corresponds to Condition 4.1
in Table 6.6.1.2.3-1 and is classified as a Category C′ fatigue detail. Only the bottom flange is
checked herein, as a net tensile stress is not induced in the top flange by the fatigue loading at this
location.

According to Eq. (6.6.1.2.2-1), the factored fatigue stress range, (Δf), must not exceed the nominal
fatigue resistance, (ΔF)n. In accordance with Article C6.6.1.2.2, the resistance factor, , and the
load modifier, , are taken as 1.0 for the fatigue limit state.

 (f )  (F)n Eq. (6.6.1.2.2-1)

From Table 6.6.1.2.3-2, the 75-year (ADTT)SL equivalent to infinite fatigue life for a Category C′
fatigue detail is 975 trucks per day. For the fatigue design in this design example, the Average
Daily Truck Traffic (ADTT) in one direction, considering the expected growth in traffic volume
over the 75-year fatigue design life, is assumed to be 2,000 trucks/day. Since the (ADTT)SL =

64
2,000 trucks per day x 0.8 = 1,600 trucks per day (refer to Article 3.6.1.4.2) exceeds the limit of
975 trucks per day equivalent to infinite life, the detail must be checked for infinite fatigue life
using the Fatigue I load combination. Per Article 6.6.1.2.5, the nominal fatigue resistance for
infinite fatigue life is equal to the constant-amplitude fatigue threshold:

(F)n = (F)TH Eq. (6.6.1.2.5-1)

where (ΔF)TH is the constant-amplitude fatigue threshold and is taken from Table 6.6.1.2.5-3. For
a Category C′ fatigue detail, (ΔF)TH = 12.0 ksi, and therefore:

(ΔF)n = 12.0 ksi

As shown in Figure 8, the unfactored negative and positive moments due to fatigue at Section 2-2
are -396 kip-ft and 2,558 kip-ft, respectively. As shown in Table 4, the short-term composite
section properties (n = 8) used to compute the stress at the bottom of the web (top of the bottom
flange, where the weld in question is located) are:

INA(n) = 695,903 in.4

dBOT OF WEB = dBOT OF STEEL – tf_BOT FLANGE = 67.66 in. – 0.8125 in. = 66.85 in.

Therefore, the unfactored stress range at the bottom of the web due to vertical loads only is:

 ( −396 + 2,558 ) (12 )( 66.85 ) 

f range _ vert =   = 3.41 ksi
 695,903 
 
Per Table 3.4.1-1, the load factor, , for the Fatigue I load combination is 1.75. The total factored
stress range at the edge of the connection is therefore:

 ( f ) = (1.75)(3.41) = 5.97 ksi

Checking Eq. (6.6.1.2.2-1),

 (f ) = 5.97 ksi  ( F ) = 12.00 ksi OK ( Ratio = 0.498 )

8.4.2 Fracture (Article 6.6.2)

As specified in Article 6.10.5.2, fracture toughness requirements in the contract drawings must be
in conformance with the provisions of Article 6.6.2.1. Material for main load-carrying components
subject to tensile stress under the Strength I load combination is assumed for this example to be
ordered to meet the appropriate Charpy V-notch fracture toughness requirements (Table C6.6.2.1-
1) specified for Temperature Zone 2 (Table 6.6.2.1-2).

Article 6.6.2.2 provides provisions for Fracture-Critical Members (FCMs). A FCM is defined as a
steel primary member or portion thereof subject to tension whose failure would probably cause a

65
portion of or the entire bridge to collapse. Article 6.6.2.2 specifies that the Engineer is to have the
responsibility for identifying and designating on the contract plans which primary members or
portions thereof are fracture-critical members (FCMs). The tension components of tub girders in
single- and twin-tub girder systems have typically been designated as FCMs.

The designation of a particular member, or member component, as a FCM entails additional and
more stringent fabrication requirements given in Clause 12 of the AASHTO/AWS D1.5M/D1.5
Bridge Welding Code (D1.5) [15], and hands-on inspections every two years. The additional
fabrication requirements are an initial cost premium in the design of new bridges that has been
proven to be effective in preventing fracture. However, the hands-on inspection requirements give
rise to considerably larger expenses that take place throughout the service life of the bridge, which
involve risks to the safety of the inspectors and bridge users.

Article 6.6.2.2 further indicates that a primary member or portion thereof subject to tension, for
which the redundancy is not known by engineering judgment, but which is demonstrated to have
redundancy in the presence of a simulated fracture in that member through the use of a refined
analysis, is to be designated as a System Redundant Member (SRM) in the contract documents.
SRMs are to be fabricated in accordance with Clause 12 of D1.5 and are to have routine inspections
performed but need not be subject to the hands-on in-service inspection requirements.

One acceptable detailed finite element analysis and evaluation procedure for classification of
SRMs [16] is provided in the AASHTO Guide Specifications for Analysis and Identification of
Fracture Critical Members and System Redundant Members [17]. The Guide Specification is
intended to provide Engineers and Owners with an analytical framework to evaluate the
redundancy of typical steel bridges and designate primary steel members as FCMs or SRMs. This
framework is composed of the finite element analysis procedure, techniques, and inputs needed to
create a reliable model of the steel bridge; as well as the minimum required primary steel member
failure scenarios, load combinations, and performance criteria used to evaluate the redundancy of
a steel bridge. Connor et al. (2020) [18] also provides a suggested alternative simplified approach
for classifying SRMs in continuous composite twin tub-girder bridges.

8.5 Girder Strength Limit State Check: Section 2-2 (Span 2)

8.5.1 Flexure (Article 6.11.6.2)

Determine if Section 2-2 qualifies as a compact section. According to Article 6.11.6.2.2, composite
sections in positive flexure qualify as compact when:

1) the specified minimum yield strengths of the flanges and web do not exceed 70 ksi;

2) the web satisfies the requirement of Article 6.11.2.1.2 such that longitudinal stiffeners are
not required (i.e., D/tw  150);

3) the section is part of a bridge that satisfies the requirements of Article 6.11.2.3 (i.e., the
special restrictions for use of the live load distribution factors – Section 5.3);

66
 4) the bottom flange is fully effective as specified in Article 6.11.1.1 (i.e., the bottom flange bf
is less than one-fifth of the effective span); and

5) the section satisfies the following web-slenderness limit:

2Dcp E
 3.76 Eq. (6.11.6.2.2-1)
tw Fyc

where Dcp is the depth of the web in compression at the plastic moment determined as specified in
Article D6.3.2.

Earlier computations indicated that the plastic neutral axis of the composite section is located in
the top flange. Therefore, according to Article D6.3.2, Dcp is taken equal to zero for this case and
Eq. (6.11.6.2.2-1) is satisfied. Section 2-2 qualifies as a compact section.

Compact sections must satisfy the following ductility requirement specified in Article 6.10.7.3 to
protect the concrete deck from premature crushing:

D p  0.42D t Eq. (6.10.7.3-1)

where Dp is the distance from the top of the concrete deck to the neutral axis of the composite
section at the plastic moment, and Dt is the total depth of the composite section. At Section 2-2:

D p = 9.0 + 3.5 − 1.0 + 0.93 = 12.43 in.

Dt = 0.8125 + 84.4 + 3.5 + 9.0 = 97.71in.

0.42Dt = 0.42(97.71) = 41.04 in.  12.43 in. ok (Ratio = 0.303)

At the strength limit state, compact composite sections in positive flexure must satisfy the
provisions of Article 6.11.7.1. Specifically, the nominal flexural resistance shall satisfy:

Mu  f Mn Eq. (6.11.7.1.1-1)

where: f = resistance factor for flexure = 1.0 (Article 6.5.4.2)

Mn = nominal flexural resistance of the section determined as specified in Article
6.11.7.1.2
Mu = factored bending moment about the major-axis of the cross-section

67
8.5.1.1 Nominal Flexural Resistance (Article 6.11.7.1.2)

The nominal flexural resistance of the section is to be taken as specified in Article 6.10.7.1.2,
except that for continuous spans, the nominal flexural resistance is always to be subject to the
limitation of Eq. (6.10.7.1.2-3) (see below). According to the provisions of Article 6.10.7.1.2, the
nominal flexural resistance of compact composite sections in positive flexure is determined as
follows:

If Dp  0.1Dt, then: Mn = Mp Eq. (6.10.7.1.2-1)

 D 
Otherwise: M n = M p 1.07 − 0.7 p  Eq. (6.10.7.1.2-2)
 Dt 

where Mp is the plastic moment of the composite section determined as specified in Article D6.1.

In continuous spans, the nominal flexural resistance of the section is also limited to the following:

M n = 1.3R h M y Eq. (6.10.7.1.2-3)

where My is the yield moment of the composite section determined as specified in Article D6.2.

For Section 2-2, My and Mp were computed earlier to be 37,751 kip-ft and 52,745 kip-ft,
respectively.

0.1D t = 0.1(97.71) = 9.77 in.  D p = 12.43 in.

Therefore,

  12.43  
M n = 52,745 1.07 − 0.7    = 51,740 kip − ft
  97.71  

Or,

Mn = 1.3(1.0)(37,751) = 49,076 kip − ft (governs)

Therefore: Mn = 49,076 kip-ft

For Strength I:

M u = 1.25 (10,110 + 1,594 ) + 1.5 (1,532 ) + 1.75 ( 9,396 ) = 33,371 kip − ft

f M n = 1.0(49,076) = 49,076 kip − ft
33,371 kip − ft  49,076 kip − ft ok (Ratio = 0.680)

68
8.5.1.2 Shear (Article 6.11.6.3)

Article 6.11.6.3 invokes to the provisions of Article 6.11.9 to determine the shear at the Strength
Limit State. Article 6.11.9 further directs the Engineer to the provisions of Article 6.10.9 for
determining the factored shear resistance of a single web. For the case of inclined webs, D in
Article 6.10.9 is taken as the depth of the web measured along the slope. Inclined webs are to be
designed to resist a shear force taken as:

Vu
Vui = Eq. (6.11.9-1)
cos ( θ )

where Vu is the shear due to factored loads on one inclined web, and  is the angle of inclination
of the web plate.

At the strength limit state, webs must satisfy the following:

Vu  v Vn Eq. (6.10.9.1-1)

where: v = resistance factor for shear = 1.0 (Article 6.5.4.2)

Vn = nominal shear resistance determined as specified in Articles 6.10.9.2 and 6.10.9.3
for unstiffened and stiffened webs, respectively
Vu = Vui = factored shear in a single web at the section under consideration

A flow chart for determining the shear resistance of I-sections is shown in Figure C6.10.9.1-1.
Design Example 1 in the NSBA Steel Bridge Design Handbook presents a complete evaluation of
shear requirements and design of an I-girder section. The shear design for tub girders, other than
that previously presented, follows the same procedure as presented in the Design Example 1.
Therefore, this example will limit discussion to checking on the Strength I Limit State at the girder
end (abutment location). The  factor is again taken equal to 1.0 in this example at the strength
limit state. The unfactored dead load and live load shears are as follows, where the live load shears
are taken as the shear envelope values.

VDC1 = ( 224 ) / 2 = 112 kips / web

VDC2 = ( 31) / 2 = 15.5 kips / web
VDW = ( 30 ) / 2 = 15 kips / web
VLL + I = ( 215 ) / 2 = 107.5 kips / web

A sample calculation of Vui, for a single web, at the abutment is given below:

1.0 1.25(112 + 15.5) + 1.5(15) + 1.75(107.5)

Vui = = 381 kips
  1 
cos  arctan  
  4 

69
The need for and required spacing of transverse stiffeners at this location will now be determined.
First, determine the nominal shear resistance of an unstiffened web according to the provisions of
Article 6.10.9.2. According to Article 6.10.9.2, the nominal shear resistance of an unstiffened web
is limited to the shear-yield or shear-buckling resistance, Vcr, determined as:

Vn = Vcr = CVp Eq. (6.10.9.2-1)

C is the ratio of the shear-buckling resistance to the shear yield strength determined as specified
in Article 6.10.9.3.2 with the shear-buckling coefficient, k, taken equal to 5.0 since the unstiffened
web shear capacity is being calculated.

Since,

Ek 29,000(5.00) D 87.0
1.40 = 1.40 = 75.4  = = 139.2
Fyw 50 t w 0.625

1.57  Ek 
C= 2   Eq. (6.10.9.3.2-6)
 D   Fyw 
 
 tw 

1.57  29,000(5.00) 
C=   = 0.235
(139.2 )  
2
50

Vp is the plastic shear force determined as follows:

Vp = 0.58Fyw Dt w Eq. (6.10.9.2-2)

Vp = 0.58(50)(87.0)(0.625) = 1,577 kips

Therefore,

Vn = Vcr = 0.235(1,577) = 370 kips

v Vn = 1.0(370) = 370 kips

The value of Vui at the end bearing is 381 kips which exceeds the nominal shear resistance of an
unstiffened web, vVn = 370 kips. Therefore, transverse stiffeners are required and the provisions
of Article 6.10.9.3 apply.

70
8.5.1.3 End Panel Shear (Article 6.10.9.3.3)

An end panel is defined as a web panel adjacent to the discontinuous end of a girder. According to
Article 6.10.9.3.3, the nominal shear resistance of a web end panel is limited to the shear-yield or
shear-buckling resistance, Vcr, determined as:

Vn = Vcr = CVp Eq. (6.10.9.3.3-1)

C is the ratio of the shear-yield or shear-buckling buckling resistance to the shear yield strength
determined as specified in Article 6.10.9.3.2. First, compute the shear buckling coefficient, k.
According to Article 6.10.9.3.3, the transverse stiffener spacing for end panels is not to exceed
1.5D = 1.5(87.0) = 130.5 inches. Assume the spacing from the abutment to the first transverse
stiffener is do = 10.75 feet = 129.0 inches.

5
k = 5+ 2
= 7.27 Eq. (6.10.9.3.2-7)
 129.0 
 
 87.0 

Since,

Ek 29,000(7.27) D 87.0
1.40 = 1.40 = 90.91  = = 139.2
Fyw 50 t w 0.625

1.57  29,000(7.27) 
C=   = 0.342 Eq. (6.10.9.3.2-6)
(139.2 )  
2
50

Vp = 0.58Fyw Dt w Eq. (6.10.9.3.3-2)

Vp is the plastic shear force, calculated as follows:

Vp = 0.58(50)(87.0)(0.625) = 1,577 kips

Therefore,

Vn = Vcr = 0.342(1,577) = 539 kips Eq. (6.10.9.3.3-1)

v Vn = 1.0(539) = 539 kips  Vu = 381 kips ok (Ratio = 0.707)

8.5.1.4 Interior Panel Shear (Article 6.10.9.3.2)

Additional web stiffeners are not required beyond the end panel in the positive moment region.
The Strength I factored shear in one web at 10.75 feet from the abutment is 327 kips (i.e. Vui = 327

71
kips). Since the factored shear, Vui, is less than the unstiffened web shear capacity, vVn = 370
kips, no additional transverse stiffeners are required and Article 6.11.6.3 is satisfied through the
remainder of the positive flexure region.

8.6 Girder Constructability Check: Section 2-1 (Interior Pier Location)

8.6.1 Flexure (Article 6.11.3.2)

The bottom flange in regions of negative flexure is to satisfy the requirements of Eqs. (6.11.3.2-1)
and (6.11.3.2-2) for critical stages of construction. Generally, these provisions will not control
because the size of the bottom flange in negative flexure regions is normally governed by the
strength limit state. The maximum negative moment reached during the deck-placement analysis,
plus the moment due to the self-weight, typically do not differ significantly from the calculated
DC1 negative moments assuming a single stage deck placement.

f bu  f Fnc Eq. (6.11.3.2-1)

fbu  f Fcrw Eq. (6.11.3.2-2)

Additionally, the top flanges, which are discretely braced during construction, must satisfy the
requirement specified in Article 6.10.3.2.2.

f bu + f  f R h Fyt Eq. (6.10.3.2.2-1)

As stated previously, the deck-placement sequence and the application of wind loads are not
considered in this example. It is assumed, for this example that the placement of the concrete deck
occurs in a single stage for the purpose of the constructability checks.

Calculate the factored maximum flexural stresses in the flanges of the steel section resulting from
the application of steel self-weight and the assumed full deck-placement (DC1).

For Strength I:

1.0(1.25)(27,012)(12)
Top flange: (f bu ) DC1 = = 26.43 ksi
15,332

1.0(1.25)(27,012)(12)
Bot. flange: (f bu ) DC1 = = −24.25 ksi
16,709

In addition to the applied steel and concrete self-weight loads, it is prudent to assume a construction
live loading (CLL) on the structure during placement of the concrete deck, as discussed Section
6.2.3. In the Strength I load combination, a load factor of 1.5 is applied to all construction loads in
accordance with Article 3.4.2. Therefore,

72
For Strength I:

1.0(1.5)(1,478)(12)
Top flange: (f bu )CLL = = 1.74 ksi
15,332

1.0(1.5)(1,478)(12)
Bot. flange: (f bu )CLL = = −1.59 ksi
16,709

Top flange: f bu = 26.43 + 1.74 = 28.17 ksi

Bot. flange: f bu = -24.25 + ( − 1.59) = -25.84 ksi

Although not included in this example in the interest of brevity, the special load combination
specified in Article 3.4.2.1 must also be considered in the design checks for the deck placement
sequence (see Section 6.3).

8.6.1.1 Top Flange Stress due to Lateral Bending

The change in the horizontal component of the web shear in the inclined web along the span acts
as a lateral force in the flanges of the tub girder, which in turn results in a top flange bending stress.
In addition, the deck overhang bracket will impose lateral forces on the top flange, causing lateral
top flange bending stress. Computation of the lateral bending stress is performed in the same
manner as demonstrated previously for Section 2-2. For the sake of brevity, the calculations will
not be shown, but instead will be summarized.

For Strength I:

flat due to horizontal component of web shear: flat = 0.47 ksi

flat due to cantilever deck overhang bracket: flat = 0.72 ksi

Total top flange flat = 0.47 + 0.72 = 1.19 ksi

8.6.1.2 Top Flange Constructability Check

Checking compliance with Article 6.10.3.2.2:

f bu + f  f R h Fyt Eq. (6.10.3.2.2-1)

For Strength I:

f bu + f = 28.17 ksi + 1.19 ksi = 29.36 ksi

f R h Fyc = 1.0(1.0)(50) = 50.0 ksi
29.36 ksi  50.0 ksi ok ( Ratio = 0.587 )

73
8.6.1.3 Bottom Flange - Flexural Resistance in Compression - Stiffened Flange (Article
6.11.8.2.3)

Calculate the nominal flexural resistance of the bottom flange in compression, Fnc, in accordance
with Article 6.11.8.2. In computing Fnc for constructability, the web load-shedding factor, Rb, is to
be taken as 1.0. The bottom flange is longitudinally stiffened at this location with a single WT12
x 42, placed at the center of the bottom flange, with the stem of the WT welded to the girder bottom
flange. Therefore, Article 6.11.8.2.3 applies.

As specified in Article 6.11.11.2, longitudinal compression flange stiffeners on tub girder bottom
flanges are to be equally spaced across the width of the flange. Furthermore, the yield strength of
the longitudinal stiffeners must not be less than the yield strength of the flanges to which they are
attached.

The projecting width, bl, of the longitudinal flange stiffener must satisfy Eq. (6.11.11.2-1):

E
b  0.48 t s Eq. (6.11.11.2-1)
Fyc

where:

ts = thickness of the projecting longitudinal stiffener element (in.)

In the case of a structural tee, ts is taken as the flange thickness of the structural tee since each half-
flange would buckle similarly to a single plate connected to the web. Furthermore, the projecting
width, bl, of structural tees is to be taken as one-half the width of the tee flange. Therefore,

29,000
bl  0.48(0.770) = 8.90 in.
50

9.02
bl = = 4.51 in.  8.90 in. WT 12x42 flange is OK
2

Determine the slenderness ratio of the bottom flange:

b fc
λf = Eq. (6.11.8.2.2-8)
t fc
where:

bfc = w = larger of the width of the flange between longitudinal flange stiffeners or the distance
from a web to the nearest longitudinal flange stiffener.

74
In this case, since the longitudinal stiffener is at the center of the bottom flange, w is the distance
from the longitudinal stiffener to the inside face of the web.

(95.125) / 2
λf = = 27.18
1.75

Since a single bottom flange stiffener is used, n = 1 and,

1
 8I  3
k =  S3  Eq. (6.11.8.2.3-1)
 wt fc 

and,
1
 I 3
5.34 + 2.84  S 3 
kS =  wt fc   5.34 Eq. (6.11.8.2.3-3)
( n + 1)
2

where: fv = St. Venant torsional shear stress in the flange due to factored loads
n = number of equally spaced longitudinal flange stiffeners
k = plate buckling coefficient for uniform normal stress, 1.0 ≤ k ≤ 4.0
kS = plate buckling coefficient for shear stress
IS = moment of inertia of a single longitudinal flange stiffener about an axis parallel to
the flange and taken at the base of the stiffener

As specified in Article C6.11.11.2, the actual longitudinal flange stiffener moment of inertia, Is,
used in determining the plate-buckling coefficient for uniform normal stress, k, from either Eq.
6.11.8.2.3-1 or Eq. 6.11.8.2.3-2, as applicable, automatically satisfies Eq. 6.11.11.2-2.
Alternatively, for preliminary sizing of the stiffener for example, a value of k can be assumed in
lieu of using Eq. 6.11.8.2.3-1 or Eq. 6.11.8.2.3-2, as applicable, but a range of 2.0 to 4.0 should
generally apply.

Structural tees are efficient shapes for longitudinal stiffeners because they provide a high ratio of
stiffness to cross-sectional area. For the WT12x42 stiffener:

Is = 166 + 12.4 ( 9.13) = 1,200 in.4

2

Therefore,

1
 8(1,200)  3
k = 3 
= 3.35  4.0
 (47.56)(1.75) 

75
 1
 1,200 3
5.34 + 2.84  3 
kS =  (47.56)(1.75)  = 2.52  5.34
(1 + 1)
2

Calculate p,

Ek
λ p = 0.57 Eq. (6.11.8.2.2-9)
Fyc Δ

where:

2
f 
Δ = 1− 3 v  Eq. (6.11.8.2.2-11)
F
 yf 

As stated previously, the St. Venant torsional shear stress, fv, can be assumed to be zero because
the bottom flange width does not exceed one-fifth of the effective span length, and all other
requirements of Article 6.11.2.3 are satisfied (see Article C6.11.2.3).

Therefore, since fv is zero:

2
 0 
Δ = 1 − 3   = 1.0
 50 

(29000)(3.35)
λ p = 0.57 = 25.13 < f = 27.18
(50)(1.0)

Since f exceeds p, it is necessary to calculate r:

Ek
λ r = 0.95 Eq. (6.11.8.2.2-10)
Fyr

and where:

Fyr = ( – 0.4) Fyc ≤ Fyw Eq. (6.11.8.2.2-7)

Fyr = (1.0 – 0.4)(50) = 30 ksi < 50 ksi

76
Therefore, r is calculated as:

(29000)(3.35)
λ r = 0.95 = 54.06 > f = 27.18
(30)

Since p < f < r,

  Δ-0.3   λ f − λ p  
Fcb = R b R h Fyc Δ −  Δ −   Eq. (6.11.8.2.2-3)
  R h   λ r − λ p  

  1.0 - 0.3  27.18 − 25.13  

Fcb = (1.0)(1.0)(50) 1.0 − 1.0 −  
  1.0  54.06 − 25.13  

Fcb = 48.94 ksi

The nominal flexural resistance of the compression flange, Fnc, is calculated as:

2
 f 
Fnc = Fcb 1-  v  Eq. (6.11.8.2.2-1)
 v Fcv 

Since fv =0 ksi, Fnc = Fcb = 48.94 ksi.

For Strength I:

fbu = -25.84 ksi

f Fnc = 1.0(48.94) = 48.94 ksi
|-25.84| ksi < 48.94 ksi ok (Ratio 0.528)

8.6.1.4 Web Bend-Buckling (Article 6.10.1.9)

The web bend-buckling resistance is to be compared with the maximum compressive stress in the
bottom flange. Determine the nominal elastic web bend-buckling resistance at Section 2-2
according to the provisions of Article 6.10.1.9.1 as follows:

0.9Ek  F 
Fcrw = 2
 min  R h Fyc , yw  Eq. (6.10.1.9.1-1)
D  0.7 
 
 tw 

77
where:

9
k= Eq. (6.10.1.9.1-2)
( Dc D)
2

Dc = 42.54 in. −1.75 in. = 40.79 in.

Compute Dc along the inclined web:

42 + 1
Dc = (40.79) = 42.05 in.
42

9
k= = 38.53
( 42.05 87.0 )
2

Therefore,

0.9(29,000)(38.53)
Fcrw = 2
= 51.90 ksi  R h Fyc = 50 ksi
 87.0 
 
 0.625 

Fcrw = 50 ksi

For Strength I:

f bu = −25.84 ksi
f Fcrw = 1.0(50) = 50 ksi
−25.84 ksi  50 ksi ok ( Ratio = 0.517 )
8.6.2 Shear (Article 6.11.3.3)

Article 6.10.3.3 requires that interior panels of stiffened webs satisfy the following requirement:

Vu  v Vcr Eq. (6.10.3.3-1)

where: v = resistance factor for shear = 1.0 (Article 6.5.4.2)

Vu = shear in the web at the section under consideration due to the factored permanent
loads and factored construction loads applied to the noncomposite section
Vcr = shear-buckling resistance determined from Eq. (6.10.9.3.3-1)

In this example, the panel adjacent to Section 2-1 will be checked. The transverse stiffener spacing
in this panel is do = 17.75 feet (Figure 3). The total factored shear will include the contribution of

78
noncomposite dead load (DC1) and the construction live loading (CLL). Note that the shear used
in the following calculation is based on a single web.

For Strength I:

Vu = 1.0(1.25)( − 270) + 1.0(1.5)( −15) = −360 kips

However, it is required that the shear be taken along the inclined web in accordance with Article
6.11.9:

Vu
Vui = Eq. (6.11.9-1)
cos(θ WEB )

−360
Vui = = −371 kip
cos(0.24 rad)

The shear-buckling resistance of the 213-inch panel is determined as:

Vn = Vcr = CVp Eq. (6.10.9.2-1)

C is the ratio of the shear-buckling resistance to the shear yield strength determined as specified
in Article 6.10.9.3.2. First, compute the shear buckling coefficient, k:

5
k = 5+ 2
Eq. (6.10.9.3.2-7)
 do 
 
D

5
k = 5+ 2
= 5.83
 213.0 
 
 87.0 

Since,

Ek 29,000(5.83) D 87.0
1.40 = 1.40 = 81.44  = = 139.2
Fyw 50 t w 0.625

1.57  Ek 
C= 2   Eq. (6.10.9.3.2-6)
 D   Fyw 
 
 tw 

79
 1.57  29,000(5.83) 
C=   = 0.274
(139.2 )  
2
50

Vp is the plastic shear force calculated as follows:

Vp = 0.58Fyw Dt w Eq. (6.10.9.3.2-3)

Vp = 0.58(50)(87.0)(0.625) = 1,577 kips

Therefore,

Vn = Vcr = 0.274(1,577) = 432 kips

v Vcr = 1.0(432) = 432 kips

−371 kips  432 kips ok (Ratio = 0.859)

8.7 Girder Strength Limit State Check: Section 2-1 (Negative Moment at Interior Pier
Location)

8.7.1 Flexure (Article 6.11.6.2)

For composite sections in negative flexure at the strength limit state, Article 6.11.6.2.3 directs the
Engineer to Article 6.11.8. Furthermore, Article 6.11.6.2.3 states the provisions of Appendix A6
are not to be applied, nor is redistribution of negative moment per Appendix B6.

At the strength limit state, bottom flanges of tub girders in compression are to satisfy:

f bu  f Fnc Eq. (6.11.8.1.1-1)

where Fnc is the nominal flexural resistance of the bottom flange determined as specified in Article
6.11.8.2.

At the strength limit state, the top flanges in tension are continuously braced by the deck, and are
to satisfy:

f bu  f Fnt Eq. (6.11.8.1.2-1)

where Fnt is the nominal flexural resistance of the top flanges determined as specified in Article
6.11.8.3.

80
Compute the factored Strength I maximum flange flexural stresses at Section 2-1, calculated
without consideration of flange lateral bending. As discussed previously, the  factor is taken equal
to 1.0 in this example. Therefore:

For Strength I:

Top flange:

1.25( − 27,012) 1.25( − 3,321) 1.5( − 3,193) 1.75( − 12,823) 

f bu = 1.0  + + +  12 = 47.71 ksi
 15,332 17,694 17,694 17,694 

Bottom flange:

1.25( − 27,012) 1.25( − 3,321) 1.5( − 3,193) 1.75( − 12,823) 

f bu = 1.0  + + +  12 = −46.25 ksi
 16,709 17,114 17,114 17,114 

8.7.1.1 Bottom Flange - Flexural Resistance in Compression - Stiffened Flange (Article

6.11.8.2.3)

Calculate the nominal flexural resistance of the bottom flange in compression, Fnc, in accordance
with Article 6.11.8.2. The bottom flange is longitudinally stiffened at this location, with a single
WT12 x 42, placed at the center of the bottom flange.

Determine the slenderness ratio of the bottom flange:

b fc
λf = Eq. (6.11.8.2.2-8)
t fc
where:

bfc = w = larger of the width of the flange between longitudinal flange stiffeners or the distance
from a web to the nearest longitudinal flange stiffener.

In this case, since the longitudinal stiffener is at the center of the bottom flange, w is the distance
from the longitudinal stiffener to the inside face of the web.

(95.125) / 2
λf = = 27.18
1.75

Since a single bottom flange stiffener is used, n = 1 and,

1
 8I  3
k =  S3  Eq. (6.11.8.2.3-1)
 wt fc 

81
and,
1
 I 3
5.34 + 2.84  S 3 
kS =  wt fc   5.34 Eq. (6.11.8.2.3-3)
( n + 1)
2

where: n = number of equally spaced longitudinal flange stiffeners

k = plate buckling coefficient for uniform normal stress, 1.0 ≤ k ≤ 4.0
kS = plate buckling coefficient for shear stress
IS = moment of inertia of a single longitudinal flange stiffener about an axis parallel to
the flange and taken at the base of the stiffener

Structural tees are efficient shapes for longitudinal stiffeners because they provide a high ratio of
stiffness to cross-sectional area. For the WT12x42 stiffener:

Is = 166 + 12.4 ( 9.13) = 1,200 in.4

2

Therefore,
1
 8(1,200)  3
k = 3 
= 3.35  4.0
 (47.56)(1.75) 

1
 1,200 3
5.34 + 2.84  3 
kS =  (47.56)(1.75)  = 2.52  5.34
(1 + 1)
2

Calculate p,

Ek
λ p = 0.57 Eq. (6.11.8.2.2-9)
Fyc Δ

where:

2
f 
Δ = 1− 3 v  Eq. (6.11.8.2.2-11)
F
 yf 

As stated previously, the St. Venant torsional shear stress, fv, can be assumed to be zero because
the bottom flange width does not exceed one-fifth of the effective span length, and all other
requirements of Article 6.11.2.3 are satisfied (see Article C6.11.2.3).

82
Therefore, since fv is zero:

2
 0 
Δ = 1 − 3   = 1.0
 50 

(29000)(3.35)
λ p = 0.57 = 25.13 < f = 27.18
(50)(1.0)

Since f exceeds p, it is necessary to calculate r:

Ek
λ r = 0.95 Eq. (6.11.8.2.2-10)
Fyr

where:

Fyr = ( – 0.4) Fyc ≤ Fyw Eq. (6.11.8.2.2-7)

Fyr = (1.0 – 0.4)(50) = 30 ksi < 50 ksi

Therefore, r is calculated as:

(29000)(3.35)
λ r = 0.95 = 54.06 > f = 27.18
(30)

Since p < f < r,

  Δ-0.3   λ f − λ p  
Fcb = R b R h Fyc Δ −  Δ −   Eq. (6.11.8.2.2-3)
  R h   λ r − λ p  

Determine the web load-shedding factor, Rb, in accordance with Article 6.10.1.10.2. As discussed
in Article C6.11.8.2.2, in calculating Rb for a tub section, use one-half of the effective box flange
width in conjunction with one top flange and a single web, where the effective box flange width is
defined in Article 6.11.1.1.

First, compute the depth of the web in compression, Dc, in accordance with Article D6.3.1.
According to Article D6.3.1, at the strength limit state, Dc for composite sections in negative
flexure is to be computed for the section consisting of the steel girder plus the longitudinal
reinforcement as follows:

Dc = 45.19 in. −1.75 in. = 43.44 in.

83
Compute Dc along the inclined web:

42 + 1
Dc = (43.44) = 44.78 in.
42

2Dc 2(44.78)
= = 143.3
tw 0.625

According to the provisions of Article 6.10.1.10.2:

2Dc
If   rw , then R b = 1.0
tw Eq. (6.10.1.10.2-1)

where:

E  5.0  E E
4.6   rw =  3.1 +   5.7
Fyc  a wc  Fyc Fyc Eq. (6.10.1.10.2-5)

2Dc t w
a wc =
bfc t fc Eq. (6.10.1.10.2-8)

E 29, 000
4.6 = 4.6 = 111
Fyc 50

E 29, 000
5.7 = 5.7 = 137
Fyc 50

2(44.78)(0.625)
a wc = = 0.673
(95.125 2)(1.75)

 5.0  29, 000

111   rw =  3.1 +  = 253.6  137
 0.673  50

2Dc
 rw = 137  = 143.3
tw

84
 2Dc
Since > rw, calculate Rb as follows:
tw
 a wc  2Dc 
Rb = 1−   − λ rw   1.0 Eq. (6.10.1.10.2-3)
 1200 + 300a wc  t w 
Therefore,

 0.673   2 ( 44.78) 
R b = 1 −    − 137  = 0.997  1.0
 1200 + 300 ( 0.673)   0.625 

Calculate Fcb,

  Δ-0.3   λ f − λ p  
Fcb = R b R h Fyc Δ −  Δ −   Eq. (6.11.8.2.2-3)
  R h   λ r − λ p  

  1.0 - 0.3  27.18 − 25.13  

Fcb = (0.997)(1.0)(50) 1.0 − 1.0 −  
  1.0  54.06 − 25.13  

Fcb = 48.79 ksi

The nominal flexural resistance of the compression flange, Fnc, is calculated as:

2
 f 
Fnc = Fcb 1-  v  Eq. (6.11.8.2.2-1)
 v Fcv 

Since fv =0 ksi, Fnc = Fcb = 48.79 ksi.

For Strength I:

fbu = -46.25 ksi

f Fnc = 1.0 (48.79) = 48.79 ksi
|-46.25| ksi < 48.79 ksi (Ratio = 0.948)

Longitudinal flange stiffeners are preferably discontinued at field splice locations at the free edge
of the flange where the flange stress is zero, particularly when the span balance is such that the
box flange on the other side of the field splice does not require stiffening, which is not the case in
this design example (see Section 8.10.1). In such cases, the compressive resistance of the
unstiffened box flange on the other side of the splice should always be checked to determine if the
flange is satisfactory without a stiffener or is a slight increase in the flange thickness will suffice
without providing a stiffener (see Section 8.10.1). Figure 14 illustrates a suggested box-flange
bolted splice detail to accommodate a termination of the stiffener at the free edge of the flange.
When the stiffener is terminated as such, fatigue of the base metal at the stiffener-to-flange weld

85
termination need not be checked in regions subject to a net applied tensile stress because the flange
stress is zero at the termination. Otherwise, the base metal at the stiffener termination would need
to be checked as a fatigue Category E or E' detail depending on the stiffener thickness, unless a
transition radius is provided at the termination (refer to Condition 4.3 in Table 6.6.1.2.3-1 and
Section 8.10.1). For further discussion on tub girder bottom flange longitudinal stiffeners, refer to
Section 3.7 of the AASHTO/NSBA Steel Bridge Collaboration document G12.1, Guidelines to
Design for Constructability and Fabrication [10].

BOTTOM BOX FLANGE (PLAN)

LONGITUDINAL
STIFFENER

Figure 14 Suggested Bottom Flange Bolted Splice Detail at a Flange Stiffener Termination

8.7.1.2 Top Flange - Flexural Resistance in Tension (Article 6.11.8.3)

Calculate the nominal flexural resistance of the top flange in tension, Fnt, in accordance with
Article 6.11.8.3.

Fnt = R h Fyt Eq. (6.11.8.3-1)

For a homogeneous girder, Rh is equal to 1.0 (Article 6.10.1.10.1). Therefore,

Fnt = 1.0 ( 50 ) = 50 ksi

For Strength I:
f bu = 47.71 ksi
f Fnt = 1.0(50.00) = 50.0 ksi
47.71 ksi  50.0 ksi ok ( Ratio = 0.954 )
8.7.2 Shear (Article 6.11.6.3)

Article 6.11.6.3 invokes to the provisions of Article 6.11.9 to determine the shear resistance at the
strength limit state. Article 6.11.9 further directs the Engineer to the provisions of Article 6.10.9
for determining the factored shear resistance of a single web. For the case of inclined webs, D, is

86
to be taken as the depth of the web measured along the slope. The factored shear in the inclined
web is to be taken as:

Vu
Vui = Eq. (6.11.9-1)
cos ( θ )

where Vu is the shear due to factored loads on one inclined web, and  is the angle of inclination
of the web plate.

At the strength limit state, webs must satisfy the following:

Vu  v Vn Eq. (6.10.9.1-1)

where: v = resistance factor for shear = 1.0 (Article 6.5.4.2)

Vn = nominal shear resistance determined as specified in Articles 6.10.9.2 and 6.10.9.3
for unstiffened and stiffened webs, respectively
Vu = Vui = factored shear in a single web at the section under consideration

The  factor is again taken equal to 1.0 in this example at the strength limit state. Live-load shears
are taken as the shear envelope values. A sample calculation of Vui, for a single web at the interior
pier is given below for Strength I:

1.0 1.25(270 + 36) + 1.5(35) + 1.75(139) 

Vui = = 698 kips
cos ( 0.24 rad )

It has been previously shown in this example (for the positive moment section) that the shear
capacity of the unstiffened web is:

v Vn = 1.0(370) = 370 kips

The maximum value of Vu in Field Section 2 is 698 kips, which exceeds vVn = 370 kips.
Therefore, transverse stiffeners are required in Field Section 2 and the provisions of Article
6.10.9.3 apply.

8.7.2.1 Interior Panel (Article 6.10.9.3.2)

An interior panel is a web panel not adjacent to the discontinuous end of a girder. Article 6.10.9.1
stipulates that the transverse stiffener spacing for interior panels without a longitudinal stiffener is
not to exceed 3D = 3(87.0) = 261.0 inches. For the first panel to the right of the first interior
support, assume a transverse spacing of do = 17.75 feet = 213.0 inches, which is the distance from
the interior support to the first top lateral strut location in Span 2, and one-half of the internal cross-
frame spacing.

87
For interior panels of girders with the section along the entire panel proportioned such that:

2Dt w
 2.5 Eq. (6.10.9.3.2-1)
( bfc t fc + bft t ft )
the nominal shear resistance is to be taken as the sum of the shear buckling resistance and the
postbuckling resistance due to tension-field action, or:

 
 
 0.87(1 − C) 
Vn = Vp C + Eq. (6.10.9.3.2-2)
2 
  do  
1+  
  D  

Otherwise, the nominal shear resistance is to be taken as the shear resistance determined from Eq.
(6.10.9.3.2-8).

Article 6.11.9 specifies that for bottom flanges of tub girders, bfc or bft, as applicable, is be taken
as one-half of the effective flange width between webs in checking Eq. 6.10.9.3.2-1, where the
effective flange width is to be taken as specified in Article 6.11.1.1, but not to exceed 18tf where
tf is the thickness of the flange.

b fc = 95.125 2 = 47.56 in.  18 (1.75 ) = 31.50 in.  Use b fc = 31.50 in.

For the interior web panel under consideration:

2(87.0)(0.625)
= 0.79  2.5
(31.50)(1.75) + 30(2.75)
Therefore:

5
k = 5+ 2
= 5.83
 213.0 
 
 87.0 

Since,

Ek 29,000(5.83) D 87.0
1.40 = 1.40 = 81.44  = = 139.2
Fyw 50 t w 0.625

1.57  29,000(5.83) 
C=   = 0.274 Eq. (6.10.9.3.2-6)
(139.2 )  
2
50

88
 Vp = 0.58Fyw Dt w Eq. (6.10.9.3.2-3)

Vp = 0.58(50)(87.0)(0.625) = 1,577 kips

Therefore,

 
 
 0.87(1 − 0.274) 
Vn = 1,577 0.274 + = 809 kips
2 
  213.0  
1+  
  87.0  

v Vn = 1.0(809) = 809 kips  Vu = 698 kips ok (Ratio = 0.863)

Separate calculations, similar to these shown above, are used to determine the need for and the
spacing of the transverse stiffeners in the remainder of the negative moment region and will not
be repeated here. The resulting stiffener spacings are shown on the girder elevation in Figure 3.
Note that although larger spacings could have been used in each panel in Field Section 2, the
stiffeners in each panel were located midway between the cross-frame connection plates in each
panel, and at locations of the top lateral struts, for practical reasons to help simplify the detailing.

8.8 Girder Service Limit State Check: Section 2-1 (Interior Pier)

Article 6.11.4 directs the Engineer to Article 6.10.4, which contains provisions related to the
control of permanent deformations at the service limit state.

8.8.1 Permanent Deformations (Article 6.10.4.2)

Article 6.10.4.2 contains criteria intended to control permanent deformations that may impair
rideability. As specified in Article 6.10.4.2.1, these checks are to be made under the Service II load
combination.

Under the load combinations specified in Table 3.4.1-1, Eqs. (6.10.4.2.2-1) and (6.10.4.2.2-2) need
not be checked for composite sections in negative flexure. For sections in negative flexure, these
equations do not control and need not be checked (see Article C6.11.4).

It should be noted, in accordance with Article 6.11.4, that redistribution of negative moment due
to the Service II loads at the interior-pier sections in continuous span flexural members using the
procedures specified in Appendix B6 is not to be applied to tub-girder sections.

Web bend buckling must always be checked, however, at the service limit state under the Service
II load combination for composite sections in negative flexure as follows:

89
 fc  Fcrw Eq. (6.10.4.2.2-4)

where fc is the compression-flange stress at the section under consideration due to the Service II
loads, calculated without consideration of flange lateral bending, and Fcrw is the nominal elastic
bend-buckling resistance for webs determined as specified in Article 6.10.1.9.

Determine the nominal elastic web bend-buckling resistance at Section 2-1 according to the
provisions of Article 6.10.1.9.1 as follows:

0.9Ek  F 
Fcrw =  min  R h Fyc , yw 
2 Eq. (6.10.1.9.1-1)
D  0.7 
 
 tw 

where Fyc is the specified minimum yield strength of the compression flange and,

9
k= Eq. (6.10.1.9.1-2)
( Dc D)
2

According to Article D6.3.1 (Appendix D6), for composite sections in negative flexure at the
service limit state where the concrete deck is considered effective in tension for computing flexural
stresses on the composite section, the depth of the web in compression in the elastic range
measured from the neutral axis down to the top of the bottom flange, Dc, is to be computed as
follows:

 −f c 
Dc =   d − t fc  0 Eq. (D6.3.1-1)
 fc + f t 

where: ft = the sum of the tension-flange stresses caused by the factored Service II loads, in this
case stresses in the top flange, calculated without considering flange lateral bending.
fc = the sum of the compression-flange stresses caused by the factored Service II loads, in
this case stresses in the bottom flange.
d = depth of the steel section.
tfc = thickness of the compression flange, in this case the bottom flange.

Eq. (D6.3.1-1) recognizes the beneficial effect of the dead-load stress on the location of the neutral
axis of the composite section (including the concrete deck) in regions of negative flexure.
Otherwise, Dc is to be computed for the section consisting of the steel girder plus the longitudinal
reinforcement.

In order to consider the deck to be considered effective in negative flexure, Article 6.10.4.2.1
requires that shear connectors be provided throughout the entire length of the tub girder, the
minimum amount of negative flexure concrete deck reinforcement be provided in accordance with
Article 6.10.1.7, and the maximum longitudinal tensile stress in the concrete deck at the section

90
under consideration caused by the Service II loads be smaller than 2fr, where fr is the modulus of
rupture of the concrete specified in Article 6.10.1.7. If these conditions are satisfied, flexural
stresses caused by Service II loads applied to the composite section may be computed using the
short-term or long-term composite section, as appropriate, assuming the concrete to be effective
in tension. Otherwise, the flexural stresses caused by Service II loads applied to the composite
section must be computed using the section consisting of the steel section plus the longitudinal
reinforcement.

The first two of the preceding conditions are satisfied (see Section 8.8.2Check the tensile stress in
the concrete deck due to the Service II load combination at Section 2-1. The longitudinal concrete
deck stress is determined as specified in Article 6.10.1.1.1d; that is, using the short-term modular
ratio n = 8.00. Note that only DC2, DW and LL+IM are assumed to cause stress in the concrete
deck. The calculated stress on the transformed section must be divided by the modular ratio, n =
8, to determine the stress in the concrete deck.

1.0 1.0(-3,321) + 1.0(-3,193) + 1.3( − 12,823)  (12)

f deck = = 1.058 ksi  2f r = 0.960 ksi
32,858(8)

The concrete deck may not be considered effective in tension at the service limit state at Section
2-1.

Therefore, for Service II:

Top flange (tension flange):

1.0( − 27,012) 1.0( − 3,321) 1.0( − 3,193) 1.30( − 12,823) 

f t = 1.0  + + +  12 = 36.86 ksi
 15,332 17,694 17,694 17,694 

Bottom flange (compression flange):

1.0( − 27,012) 1.0( − 3,321) 1.0( − 3,193) 1.30( −12,823) 

f c = 1.0  + + +  12 = −35.66 ksi
 16,709 17,114 17,114 17,114 

Calculate the depth of the web that is in compression, Dc. According to Article D6.3.1 (Appendix
D6), since the deck is not permitted to be considered effective in tension at this section at the
service limit state, Dc is to be computed for the section consisting of the steel girder plus the
longitudinal reinforcement as follows:

Dc = 45.19 in. −1.75 in. = 43.44 in.

91
Calculate Dc along the web:

43.44
Dc = = 44.78 in.
  1 
cos  arctan   
  4 

9 9
k= 2
= 2
= 33.97
 Dc   44.78 
   
 D  87.0 

0.9 ( 29,000 )( 33.97 ) Fyw

Fcrw = 2
= 45.76 ksi  R h Fyc = 50 ksi  = 71.4 ksi
 87.0  0.7
 
 0.625 

−35.66 ksi  45.76 ksi ok (Ratio = 0.779)

8.8.2 Concrete Deck (Article 6.10.1.7)

Article 6.10.1.7 requires the minimum one-percent longitudinal reinforcement in the concrete deck
wherever the longitudinal tensile stress in the deck due to the factored construction loads or due to
the Service II load combination exceeds fr.

Check the tensile stress in the concrete deck due to the Service II load combination at the section
45.0 feet from Pier 1 in Span 2. The longitudinal concrete deck stress is determined as specified in
Article 6.10.1.1.1d; that is, using the short-term modular ratio n = 8.00. Note that only DC2, DW
and LL+IM are assumed to cause stress in the concrete deck.

1.0 1.0(-676) + 1.0(-650) + 1.3( − 4,847)  (12)

f deck = = 0.427 ksi  0.90f r = 0.432 ksi
26,810(8)

Extend the minimum reinforcement to a section 45.0 feet from the Pier 1 in Span 2 for the Service
II loads.

Also, check the tensile stress in the concrete deck due to Service II load combination at the section
115.0 feet from the abutment in Span 1. The longitudinal concrete deck stress is determined as
specified in Article 6.10.1.1.1d; that is, using the short-term modular ratio n = 8.00, and only DC2,
DW and LL+IM are included.

1.0 1.0(122) + 1.0(113) + 1.3( − 5, 703) (12)

f deck = = 0.426 ksi  0.90f r = 0.432 ksi
25,296(8)

92
From Pier 1, extend the minimum reinforcement to a section 115.0 feet from the abutment in Span
1 for the Service II loads. The above locations should also be similarly checked for the factored
construction loads (i.e., the deck placement sequence) with adjustments made as necessary.

8.9 Girder Fatigue and Fracture Limit State Check: Section 2-1 (Negative Moment at
Interior Pier Location)

8.9.1 Fatigue (Article 6.11.5)

Article 6.11.5 directs the Engineer to Article 6.10.5, where details on tub girder section flexural
members must be investigated for fatigue as specified in Article 6.6.1. Either the Fatigue I or
Fatigue II load combination specified in Table 3.4.1-1 and the fatigue live load specified in Article
3.6.1.4 is to be employed for checking load-induced fatigue in tub girder sections. Further
discussion concerning load induced fatigue in tub girders is presented in Section 8.4.1.

The fatigue details employed in this example in the negative moment regions, such as the
connection plate welds to the flanges, satisfy the limit state specified for load induced fatigue in
Article 6.11.5. Furthermore, interior panels of webs with transverse stiffeners satisfy Article
6.10.5.3. The detailed checks are not illustrated in this example; however similar checks are
illustrated in Section 8.4.1 and in Design Example 1 of the NSBA Steel Bridge Design Handbook.

8.9.2 Fracture (Article 6.6.2)

Discussion concerning the fracture limit state for tub girders was previously presented in Section
8.4.2.

8.10 Girder Check: Section 1-2 and 1-3

8.10.1 Comparison of Unstiffened and Stiffened Bottom Flange in End Spans

Because a field section length of 130 feet is required to minimize the number of field sections and
field splices for the given span arrangement, girder Section 1-3 is not located at a point of dead
load contraflexure. Due to the span balance, there is negative bending moment at Section 1-3
causing the bottom flange to be in compression. When proportioning the bottom flange at this
location, two options exist:

• Option A – use a thicker, unstiffened, bottom flange

• Option B – use a longitudinally stiffened bottom flange which will allow a thinner bottom
flange plate to be used.

For comparison, both these options are briefly presented in this section.

8.10.1.1 Option A - Unstiffened Flange

The resistance in compression of a tub girder bottom flange that is unstiffened is limited by the
buckling resistance of the plate, which is a function of the flange slenderness (b/t) ratio. Therefore,
a simple option that may be used to increase the resistance is to increase the thickness of the bottom

93
flange plate. For this particular example, the bottom flange plate thickness is 1.375 inches at
Section 1-3, and this plate is extended to Section 1-2 which is 93.75 feet from the abutment, as
shown previously in Figure 3.

Compute the factored maximum bottom flange flexural stress in Section 1-3 at the field splice
under the Strength I load combination. As discussed previously, the  factor is taken equal to 1.0
in this example. At this location, the unfactored bending moments are approximately as follows:

MDC1 = -4,412 kip-ft

MDC2 = -406 kip-ft
MDW = -390 kip-ft
MLL+I = -7,750 kip-ft

The negative flexure concrete deck reinforcement (Section 8.1.2.2) is assumed extended through
Section 1-3. Therefore, the longitudinal reinforcement is included in the composite section
property calculations. Separate calculations similar to the section property calculations at Section
2-1, but not included herein, show that at Section 1-3:

Steel Section only: SBOT OF STEEL = 11,092 in.3

Steel Section + Long. Reinforcement: SBOT OF STEEL = 12,238in.3

Therefore, for Strength I:

1.25 ( −4,412 ) 1.25 ( −406 ) 1.5 ( −390 ) 1.75 ( −7,750 ) 

f bu = 1.0  + + +  12 = −20.34 ksi
 11,092 12,238 12,238 12,238 

Calculate the nominal flexural resistance of the bottom flange in compression, Fnc, in accordance
with Article 6.11.8.2.2. This calculation is similar to the calculations shown to compute the bottom
flange negative moment flexural resistance at Section 2-1, therefore calculations for Section 1-3
are briefly summarized below.

b fc
λf = Eq. (6.11.8.2.2-8)
t fc

(95.125)
λf = = 69.18
1.375

Compute r,

Ek
λ r = 0.95 Eq. (6.11.8.2.2-10)
Fyr

where k = 4.0 (taken as 4.0 since bottom flange is unstiffened)

94
and where:

Fyr = ( – 0.4)Fyc ≤ Fyw Eq. (6.11.8.2.2-13)

Fyr = (1.0 – 0.4)(50) = 30 ksi < 50 ksi

Therefore, r is calculated as:

(29000)(4.0)
λ r = 0.95 = 59.07
(30)

Since f = 69.18 > r = 59.07, then,

0.9ER b k
Fcb = Eq. (6.11.8.2.2-4)
( λf )
2

where,

Rb = 1.0 (calculated but not shown)

0.9 ( 29,000 )(1.0 )( 4.0 )

Fcb = = 21.81 ksi
( 69.18)
2

The nominal flexural resistance of the compression flange, Fnc, is calculated as:

2
 f 
Fnc = Fcb 1-  v  Eq. (6.11.8.2.2-1)
 v Fcv 

Given the satisfaction of Article 6.11.2.3 requirements, fv =0 ksi and, Fnc = Fcb = 21.81 ksi.

Therefore, for Option A, Strength I:

fbu = -20.34 ksi

f Fnc = 1.0 (21.81) = 21.81 ksi
|-20.34| ksi < 21.81 ksi ok (Ratio = 0.933)

8.10.1.2 Option B - Stiffened Flange

As an alternative to using a thicker bottom flange plate (Option A), the WT12 x 42 bottom flange
longitudinal stiffener can be extended further into the end span, up to 93.75 feet from the end
support, as shown in Figure 15. This will require that the WT12 x 42 stiffener also be spliced at

95
the field splice, and will require careful attention to the detail at the termination of the flange
stiffener in Span 1 (at Section 1-2).

Figure 15 Option B in Elevation, Stiffened Bottom Flange

For this option, the existing 0.8125-inch-thick bottom flange plate, in combination with the WT12
x 42 bottom flange longitudinal stiffener, is used. The longitudinal flange stiffener is extended to
Section 1-2, which is 93.75 feet from the abutment.

The negative flexure concrete deck reinforcement (Section 8.1.2.2) is assumed extended through
Section 1-3. Therefore the longitudinal reinforcement is included in the composite section property
calculations. Separate calculations similar to the section property calculations at Section 2-1, but
not included herein, show that at Section 1-3:

Steel Section only: SBOT OF STEEL = 8,006 in.3

Steel Section + Long. Reinforcement: SBOT OF STEEL = 8,902 in.3

Compute the factored maximum bottom flange flexural stress in Section 1-3 at the field splice
under the Strength I load combination. As discussed previously, the  factor is taken equal to 1.0
in this example. Therefore:

For Strength I:

1.25 ( −4,412 ) 1.25 ( −406 ) 1.5 ( −390 ) 1.75 ( −7,750 ) 

f bu = 1.0  + + +  12 = −28.02 ksi
 8,006 8,902 8,902 8,902 

96
Calculate the nominal flexural resistance of the bottom flange in compression, Fnc, in accordance
with Article 6.11.8.2.2. This calculation is similar to the calculations shown to compute the bottom
flange negative moment flexural resistance at Section 2-1, therefore calculations for Section 1-3
are briefly summarized below.

b fc
λf = Eq. (6.11.8.2.2-8)
t fc

(95.125 / 2)
λf = = 58.54
0.8125

Calculate p,

Ek
λ p = 0.57 Eq. (6.11.8.2.2-9)
Fyc Δ

where,

k = 4.0 (calculated as 7.22, but the limit of 4.0 governs)

 = 1.0 (calculated but not shown, fv = 0.0 ksi)

(29000)(4.0)
λ p = 0.57 = 27.45 < f = 58.54
(50.0)(1.0)

Compute r,

Ek
λ r = 0.95 Eq. (6.11.8.2.2-10)
Fyr

where,

Fyr = 30.0 ksi (calculated but not shown)

(29000)(4.0)
λ r = 0.95 = 59.07
(30.0)

Since p < f < r,

  Δ-0.3   λ f − λ p  
Fcb = R b R h Fyc Δ −  Δ −   Eq. (6.11.8.2.2-3)
  R h   λ r − λ p  

97
where:

Rb = 1.0 (calculated but not shown)

  1.0-0.3  58.54 − 27.45  

Fcb = (1.0) (1.0) (50) 1.0 − 1.0 −  
  1.0  59.07 − 27.45  

Fcb = 35.25 ksi

The nominal flexural resistance of the compression flange, Fnc, is calculated as:

2
 f 
Fnc = Fcb 1-  v  Eq. (6.11.8.2.2-1)
 v Fcv 

Given the satisfaction of Article 6.11.2.3 requirements, fv =0 ksi and, Fnc = Fcb = 35.25 ksi.

Therefore, for Option B, Strength I:

fbu = -28.02 ksi

f Fnc = 1.0(35.25) = 35.25ksi
|-28.02| ksi < 35.25 ksi ok (Ratio = 0.795)

At the termination of the flange stiffener, the bottom flange is subjected to both tensile and
compressive stresses under the fatigue live load. Since in this option it is necessary to terminate
the flange stiffener beyond the field splice in a region that is subject to a net tensile stress, fatigue
must be investigated at the stiffener termination. Under the condition of welded stiffener
attachments, for base metal at the termination of longitudinal stiffener-to-box flange welds with
no special transition radius provided at the weld termination, the fatigue detail is either Category
E or E' depending on the stiffener thickness (refer to Condition 4.3 in Table 6.6.1.2.3-1). The
thickness of the stiffener web is less than 1.0 inch and so the detail is a fatigue Category E detail.
Separate computations similar to the computations given below indicate that a fatigue Category E
detail is not sufficient at this location.

Therefore, a transition radius will be provided at the stiffener termination with the end weld ground
smooth. A minimum-radius transition of 6 inches will provide the nominal fatigue resistance of a
Category C detail (refer to Condition 4.3 in Table 6.6.1.2.3-1). A continuous fillet weld on both
sides of the stiffener is used to attach the stiffener to the bottom flange. Consideration should be
given to wrapping the weld around the end of the stiffener for sealing. The weld and stiffener
material should then be ground to a smooth contour where the radiused stiffener end becomes
tangent to the flange (Figure 16).

98
 Figure 16 Option B, Longitudinal Flange Stiffener Termination

Since the (ADTT)SL = 2,000 trucks per day x 0.8 = 1,600 trucks per day (refer to Article 3.6.1.4.2)
is less than the 75-year (ADTT)SL Equivalent to Infinite Life of 1,680 trucks per day specified in
Table 6.6.1.3.2-2 for a Category C detail, fatigue of the base metal at the longitudinal flange
stiffener weld termination will be checked for finite life using the Fatigue II load combination
(Table 3.4.1-1). The stress range due to the fatigue live load modified by the corresponding
dynamic load allowance of 15 percent will be used to make this check. The lateral distribution
factor for the fatigue limit state, computed in Section 7.1, is also used.

The provisions of Article 6.6.1.2 apply only to details subject to a net applied tensile stress. In this
example, the effect of the future wearing surface is conservatively ignored when determining if a
detail is subject to a net applied tensile stress. Separate computations indicate that the stiffener
termination at this location is subject to a net tensile stress under the unfactored permanent load
plus the Fatigue I load combination (refer to Article 6.6.1.2.1).

According to Article 6.6.1.2.1, for flexural members with shear connectors provided throughout
their entire length and with concrete deck reinforcement satisfying the provisions of Article
6.10.1.7, flexural stresses and stress ranges applied to the composite section at the fatigue limit
state at all sections in the member may be computed assuming the concrete deck to be effective
for both positive and negative flexure. Shear connectors are assumed along the entire length of the
girder in this example, which is required for composite tub girders according to Article 6.11.10.
Earlier computations were made to verify that the longitudinal concrete deck reinforcement
satisfies the provisions of Article 6.10.1.7 (Section 8.8.2). Therefore, the concrete deck will be
considered effective in computing all stresses and stress ranges applied to the composite section
in the subsequent fatigue calculations.

The stress range, (), at the longitudinal flange stiffener weld termination due to the factored
fatigue load (factored by the specified 0.80 load factor for the Fatigue II load combination) is
computed using the properties of the short-term composite section as follows.

At the stiffener termination, the unfactored fatigue live load bending moments are as follows
(Figure 8):

99
 Positive Flexure: MLL+I = 2,398 kip-ft
Negative Flexure: MLL+I = -1,121 kip-ft

From Table 4, the short-term composite section modulus for the unstiffened bottom flange at the
stiffener termination is:

Composite Section, n=8.0: SBOT OF STEEL = 10,285 in.3

For load-induced fatigue, each detail must satisfy:

γ ( Δf )  ( F )n Eq. (6.6.1.2.2-1)

where:

 = load factor per Table 3.4.1-1 for the appropriate Fatigue Load Combination
(f) = live load stress range due to passage of fatigue truck
(F)n = nominal fatigue resistance per Article 6.6.1.2.5

Therefore, the Fatigue II stress range is computed as:

0.80 ( 2,398 )(12 ) 0.80 −1,121 (12 )

γ ( Δf ) = + = 3.28 ksi
10,285 10,285

Both the resistance factor  and design factor  are specified to be 1.0 at the fatigue limit state
(Article C6.6.1.2.2). The nominal fatigue resistance for the Fatigue II load combination and finite
fatigue life is determined as:

1
 A 3
( F)n =  Eq. (6.6.1.2.5-2)
N

where:

N = (365)(75)n(ADTT)SL Eq. (6.6.1.2.5-3)

The number of stress cycles per truck passage, n, is equal to 1.0 (Table 6.6.1.2.5-2). Therefore,

N = (365)(75)(1.0)(1,600) = 43.8 x 106 cycles

For a Category C detail, the detail category constant A = 44 x 108 ksi3 (Table 6.6.1.2.5-1).
Therefore:
1
 44 x 108  3
( F)n = 6 
= 4.65 ksi
 43.8 x 10 

100
  ( f )  (  F ) n Eq. (6.6.1.2.2-1)

3.28 ksi < 4.65 ksi ok (Ratio = 0.705)

8.10.1.3 Summary of Unstiffened Flange versus Stiffened Flange

To provide the most economical solution for terminating the stiffener beyond a field splice, the
Engineer, preferably in consultation with a fabricator, should evaluate the relative cost to thicken
the bottom flange adjacent to the field splice, terminate the stiffener in the span, or even run the
stiffener the full length of the end span. There are several factors that the Engineer must consider
prior to choosing an option, with regard to the amount of material, fabrication costs, installation
costs, and the performance and long-term serviceability.

For this example, the additional material weight of each of the options that was investigated for
this bottom flange section is as follows:


Option A = ( 490 ) ( 36.25 )
( 98.5)(1.375 - 0.8125)  = 6,834 lbs

 144 

Option B = ( 42 )( 36.25 ) = 1,523 lbs

While Option B saves approximately 5,300 pounds of steel per girder, it represents only a material
savings, which will likely be overcome by the increase in labor costs associated with welding the
stiffener, coping and grinding the stiffener termination, making the fillet welds, and fabrication
and installation of the WT splice. Therefore, Option A was chosen for this design example.

8.11 Top Flange Lateral Bracing

As discussed previously in Section 4.4, top flange lateral bracing increases the torsional stiffness
of tub girder sections during erection, handling, and deck placement. For composite tub girders
closed by the deck slab, the cross-section of the tub is torsionally stiff. However, prior to placement
of the deck slab, the open tub girder is torsionally more flexible and subject to rotation or twist.
The top flange lateral bracing forms a quasi-closed section resisting shear flow from the
noncomposite loading.

The lateral bracing is typically comprised of WT or angle sections and is often configured in a
single diagonal (Warren or Pratt truss) arrangement. The diagonal bracing members commonly
frame into the work point of the girder top flange and internal cross-frame or diaphragm
connection. Alternatively, the length between internal cross-frames can be divided into multiple
lateral bracing panels. Such framing arrangements usually include a single transverse strut at
intermediate brace locations. The plane of the top flange lateral bracing system should be detailed
to be as close as possible to the plane of the girder top flanges so as to increase the torsional
stiffness of the section, while at the same time reducing connection eccentricities and excessive
out-of-plane bending in the web.

101
8.11.1 Diagonal Bracing Members

Diagonal bracing is proportioned to resist tension or compression in combination with flexure as

appropriate, based on connection geometry. Generally, design for compression will govern the
member size. The member must also satisfy slenderness requirements specified in Article 6.9.3,
the minimum thickness requirements of Article 6.7.3, and should satisfy the minimum area
requirement given by Eq. (C6.7.5.3-1).

Preliminary proportions of the diagonal members are determined as follows:

K
For secondary bracing members in compression:  140 (Table 6.6.2.1-1 and Article 6.9.3)
r

The maximum length work point to work point of a diagonal member is 21.15 feet in Span 2 near
the interior supports. This length will be used for design since all diagonal bracing members will
be the same size. For bolted or welded connections at both ends of the member, the effective length
factor K may be taken as 0.750 (Article 4.6.2.5). In this example, a WT sections will be used for
the lateral bracing members. If single-angle sections were to be used, the effective length factor K
should be taken as 1.0.

for
K
 140 ; rmin =
( 0.750 )( 21.15)(12 ) = 1.36 in.
r 140

Calculate the minimum required cross-sectional area, Ad:

Ad  0.03 w Eq. (C6.7.5.3-1)

where:

w = center-to-center distance between the top flanges (in.)

A d  0.03 (138.0 ) = 4.14 in.2

Therefore, select a WT5 x 15:

rmin = ry = 1.37 in.  1.36 in. ok

A d = 4.42 in.2  4.14 in.2 ok

In the noncomposite condition, there are several loading conditions that will generate forces in the
top flange bracing system. As discussed in the NSBA publication Practical Steel Tub Girder
Design [2], torsional moments typically induced by dead loads and construction loads will result
in lateral bracing member forces. These forces can be derived from the St. Venant shear flow at
the girder cross-sections, assuming the horizontal truss acts as an equivalent plate. Where forces

102
in bracing members are not readily available from a refined analysis, the shear flow across the
equivalent plate can be computed from Eq. C6.11.1.1-1, and the resulting shear can then be
resolved into diagonal bracing member forces.

The horizontal component of the web shear in the inclined web along the span also imposes a
lateral force on the top flanges of the tub girder. In the noncomposite condition, the lateral force
due to web shear is assumed to be distributed to the top flanges of the open tub girder. The majority
of this force is resisted directly by the lateral struts of the bracing system and not by the diagonals.
Therefore, the forces in diagonal members resulting from the web shear component are typically
taken as zero.

The lateral bracing members, in conjunction with the tub girder top flanges, form a geometrically
stable horizontal truss. In the noncomposite condition, the horizontal truss is connected to the
girder top flanges in a region of high bending stress, considering that the neutral axis of the
noncomposite section is typically near the mid-height of the steel section. Due to compatibility,
the horizontal truss must experience the same axial strains as the tub girder top flanges that result
from applied bending moments, therefore resulting in axial forces being carried by the bracing
members. In the absence of a refined analysis, design equations have been developed by Fan and
Helwig [12] to evaluate the bracing member forces due to tub girder bending.

Lateral bracing members are also subject to forces due to wind loads acting on the noncomposite
girder prior to deck placement at any point during the construction sequence. The lateral load
resulting from the wind pressure applied to the exposed tub girder area is typically distributed
equally to the top and bottom flanges. In the noncomposite condition, the portion of lateral load
applied to the top flange may then be resolved into bracing member axial forces.

Tee sections subject to axial compression can fail either by flexural buckling about the x-axis or
by torsion combined with flexure about the y-axis, where the y-axis is defined as the axis of
symmetry of the tee section (i.e., a failure mode known as flexural-torsional buckling). Since the
flanges of the tee are typically connected to the girder flange or to a lateral connection plate
(Section 8.11.3), the tee is also subject to a uniform bending moment about the major principal
axis (i.e., the x-axis) due to the eccentricity of the connection at each end of the member. Second-
order effects arise from the additional secondary moment caused by the axial compressive force
acting through the member deflection. Article 4.5.3.2.2b specifies that the single-step adjustment
or moment magnification method may be used to determine the second-order elastic moment in
lieu of a more refined analysis. The flexural resistance of tee-section members is determined as
specified in Article 6.12.2.2.4. The appropriate interaction curves provided in Article 6.9.2.2 can
be conservatively used to check the tee-section member for the effects of combined axial
compression and flexure.

103
8.11.2 Top Lateral Strut

Computations for a top lateral strut in Span 2 will be presented herein. It has been shown previously
that for Span 2, the horizontal component of the unfactored noncomposite (DC1) web shear per
top flange is VH = 0.49 kip/ft. Therefore, the factored Strength I force resisted by the top lateral
strut is:

F = ΔVH d STRUT = 1.25 ( 0.49 )(17.75 ) = 10.87 kips

where dSTRUT is the spacing of the lateral struts in Span 2 (near the interior supports). Note that
the special load combination specified in Article 3.4.2.1 is not checked in this example.

Due to the inclination of the web, the struts are always in tension. Therefore, the member is
designed in accordance with the provisions of Article 6.8.1. A L4 x 4 x ½ will be considered for
the top lateral strut.

According to Article 6.8.2.1, the factored tensile resistance Pr is to be taken as the lesser of:

Pr = y Pny = y Fy Ag Eq. (6.8.2.1-1)

or Pr = u Pnu = u Fu An Rp U Eq. (6.8.2.1-2)

In the preceding equation, Rp is the reduction factor for holes taken equal to 0.90 for bolt holes
punched full size and 1.0 for bolt holes drilled full size or subpunched and reamed to size. For the
purposes of this example, it is assumed that holes are drilled full size; thus, Rp = 1.0.

Also, in Eq. 6.8.2.1-2, the reduction factor U (Article 6.8.2.2) accounts for the effect of shear lag
in the connection. Assuming the top strut will utilize a bolted connection with two fasteners spaced
at 3 inches on center in the direction of applied force U is calculated in accordance with Table
6.8.2.2-1 and Figure C6.8.2.2-1.

x
U = 1− Table (6.8.2.2-1)
L

where:

x = connection eccentricity (in.); for L4 x 4 x ½, this value is 1.18 in.

L= length of connection (in.), and per Figure C6.8.2.2-1, the out-to-out distance of the
bolt holes can be used. Thus, assuming 15/16-inch diameter bolt holes, L = 3 in. +
15/16 in. = 3.9375 in.

1.18
U = 1− = 0.700
3.9375

Values for U are presented in Table 6.8.2.2-1 for other common connection types.

104
Therefore,

Pr = y Fy Ag = (0.95)(50)(3.75) = 178 kips

Pr = u Fu An Rp U = (0.80)(65)(3.28)(1.0)(0.700) = 119.4 kips > 10.87 kips (governs)

where An is based on the use of a 7/8 inch diameter bolt in a standard size (15/16″ diameter) hole.

In addition to tensile resistance, the member must also satisfy the slenderness requirement
specified in Article 6.8.4 for bracing members:

For secondary bracing members in tension:  240

r

The distance between the webs at the top of the tub girder is 138 inches. For an L4 x 4 x ½, rmin =
rz = 0.776 inches.

138
= = 177.8  240 ok
r 0.776

8.11.3 Detailing

Final detailing of lateral bracing and connections must consider long term service and performance
of the structure as well as economy in fabrication and erection. The publication Practical Steel
Tub Girder Design [2], available from NSBA, provides current guidance with regard to design
philosophy and detailing practices for lateral bracing systems.

Whenever possible, the lateral bracing should be connected as close as possible to the horizontal
plane of the tub girder top flanges. Providing bracing connections to the flanges is more
economical than connections to the webs since they involve fewer connection components, and
they are much simpler to fabricate and connect as compared to connections to the tub girder webs.
The connection can be further simplified if gusset plates are eliminated and the bracing members
are connected directly to the tub girder top flanges. Connecting the lateral bracing directly to the
top flanges also provides a direct load path between the bracing member and the tub girder top
flanges, further simplifying the design of the connection and eliminating concerns about out-of-
plane bending of the web. Additionally, inspection of the lateral bracing connection is enhanced
when the bracing is connected to the top flange, because there are fewer components in a top flange
connection as compared to a web connection.

Furthermore, fatigue is an important consideration when selecting the type of connection detail to
use. For example, welded connections to the top flanges, specifically in tension regions, are
typically undesirable, and in some cases forbidden, due to fatigue concerns. Therefore, the use of
gusset plates welded to top flanges is not recommended. A more suitable connection may be to
bolt the gusset plate to the top flange, typically mitigating fatigue concerns. In some cases, where
wide top flanges are used, the lateral bracing may be bolted directly to the top flange, eliminating
the use of gusset plates, and providing a direct load path.

105
Additionally, the block shear rupture resistance of tension members at connections must be
verified in accordance with Article 6.13.4. The lateral bracing members and the gusset plates were
investigated to verify that adequate connection material is provided to develop the factored
resistance of the connection and prevent block shear rupture (calculations not shown).

8.12 Bearings
Common tub girder designs may utilize one or two bearings at the supports. The number of
bearings installed will have a significant effect on the design of the tub girder, as well as the design
of the internal and external diaphragms at the support. Article 6.11.1.2 presents guidance on the
use and design of bearing systems.

At the support, tub girder torsion can be directly resolved into a force couple with the use of two
bearings under each tub girder. The use of two bearings also reduces the design reaction for the
bearing, as compared to the use of a single bearing. Two-bearing arrangements work well for non-
skewed or radial supports but are impractical for supports that are skewed more than a few degrees.
In the case of a skewed support, the tub girder and external diaphragm tend to prevent uniform
bearing contact during construction and deck placement.

If a single bearing is used under each tub girder at the support, contact between the tub girder and
bearing is optimized. Single bearing systems tend to be more forgiving of construction tolerances,
especially for skewed supports. When single bearing systems are used, the external diaphragms at
support lines must be sufficient to resist the torsional moments in the tub girders, as the diaphragm
and adjacent girder form a structural system to counter the torsion in the individual girders. Use of
a single bearing will cause bending of the internal diaphragm, which can be significant in some
cases. When the stresses in the bottom flange of the tub girder, caused by the bending of the internal
diaphragm at interior pier locations, are deemed significant, the Commentary to Article 6.11.8.1.1
provides direction to check the combined stresses in the tub girder bottom flange at the strength
limit state. For tub girders supported on two bearings, the flange stress due to major-axis bending
of the internal diaphragm is typically small and can often be ignored.

Steel-reinforced neoprene pads and pot bearings are the most commonly used bearing types for
tub girders, however, in some cases disc bearings have been successfully used as well. Steel-
reinforced neoprene bearing pads are much more tolerant of construction movements. They also
can be easily inspected, while generally being less expensive than pot bearings. Steel-reinforced
neoprene bearing pads are not as suitable for higher reactions as compared to pot or disc bearings,
and therefore may not be acceptable in some applications.

Girder movement can be accommodated by both steel-reinforced neoprene pads and pot bearings.
Movement in steel-reinforced neoprene pads is accommodated by deformation within the
elastomer. In cases where the magnitude of movement would require a thick and potentially
unstable neoprene pad, a stainless steel/polytetrafluoroethylene (PTFE) sliding surface can be
utilized. A stainless steel/PTFE sliding surface is always required for pot bearings when translation
needs to be accommodated.

Regardless of the bearing type used, consideration should always be given to future jacking of the
structure so that bearings can be repaired or replaced. A detailed design guide for typical bearing

106
types used in steel bridges can be found in the AASHTO/NSBA Steel Bridge Collaboration
publication G9.1 Steel Bridge Bearing Design and Detailing Guidelines [19].

8.13 Design Example Summary

The results for this design example at each limit state are summarized below for the maximum
positive moment and maximum negative moment locations. The results for each limit state are
expressed in terms of a performance ratio, defined as the ratio of a calculated value to the
corresponding resistance.

8.13.1 Maximum Positive Moment Region, Span 2 (Section 2-2)

Constructability
Flexure (Strength I)
Eq. 6.10.1.6-1 – Top Flange 0.278
Eq. 6.10.3.2.1-1 – Top Flange, yielding 0.807
Eq. 6.10.3.2.1-2 – Top Flange, local buckling 0.696
Eq. 6.10.3.2.1-2 – Top Flange, lateral torsional buckling 0.775
Eq. 6.11.3.2-3 – Bottom Flange, yielding 0.431
Eq. 6.10.3.2.1-3 – Web Bend Buckling 0.947

Service Limit State

Permanent Deformations (Service II)
Eq. 6.10.4.2.2-1 – Top Flange 0.643
Eq. 6.10.4.2.2-2 – Bottom Flange 0.725

Fatigue Limit State

Flexure (Fatigue I)
Eq. (6.6.1.2.2-1) – Bottom Flange Connection-Plate Welds 0.498
Strength Limit State (Compact Section)
Ductility Requirement (Eq. 6.10.7.3-1) 0.303
Flexure – Eq. 6.11.7.1.1-1 (Strength I) 0.680
Shear (at abutment) – Eq. 6.10.9.1-1 (Strength I) 0.707

8.13.2 Interior Pier Section, Maximum Negative Moment (Section 2-1)

Constructability
Flexure (Strength I)
Eq. 6.10.3.2.2-1 – Top Flange, yielding 0.587
Eq. 6.11.3.2-1 – Bottom Flange, local buckling 0.528
Eq. 6.11.3.2-2 – Web Bend Buckling 0.517
Shear (Strength I)
Eq. 6.10.3.3-1 0.859

Service Limit State

Web Bend Buckling (Service II) - Eq. 6.10.4.2.2-4 0.779

107
Fatigue Limit State
Flexure (Fatigue I)
Eq. (6.6.1.2.2-1) – Longitudinal Stiffener Termination 0.705

Strength Limit State

Flexure (Strength I)
Bottom Flange – Eq. 6.11.8.1.1-1 0.948
Top Flange – Eq. 6.11.8.1.2-1 0.954
Shear (at interior pier) – Eq. 6.10.9.1-1 (Strength I) 0.863

108
9.0 REFERENCES

1. AASHTO. AASHTO LRFD Bridge Design Specifications, 9th Edition, American Association
of State Highway and Transportation Officials, Washington, DC, 2009.

2. Coletti, D.A., Fan, Z., Holt, J., Vogel, J., Gatti, W. Practical Steel Tub Girder Design, National
Steel Bridge Alliance (NSBA), Chicago, IL, 2005.

3. NSBA. Steel Bridge Design Handbook: Design Example 1: Three-Span Continuous Straight
Composite Steel I-Girder Bridge. National Steel Bridge Alliance, 2022.

4. NSBA. Steel Bridge Design Handbook: Design Example 2A: Two-Span Continuous Straight
Composite Steel I-Girder Bridge. National Steel Bridge Alliance, 2022.

5. NSBA. Steel Bridge Design Handbook: Design Example 2B: Two-Span Continuous Straight
Composite Steel Wide-Flange Beam Bridge. National Steel Bridge Alliance, 2022.

6. NSBA. Steel Bridge Design Handbook: Design Example 3: Three-Span Continuous

Horizontally Curved Composite Steel I-Girder Bridge. National Steel Bridge Alliance, 2022.

7. NSBA. Steel Bridge Design Handbook: Design Example 5: Three-Span Continuous

Horizontally Curved Composite Steel Tub-Girder Bridge. National Steel Bridge Alliance,
2022.

8. AASHTO/NSBA. G1.4: Guidelines for Design Details, 1st Edition, NSBAGDD-1-OL,

American Association of State Highway and Transportation Officials, Washington, DC, 2006.

9. Texas Department of Transportation (TxDOT). Preferred Practices for Steel Bridge Design,
Fabrication, and Erection, Texas Steel Quality Council, 2009.

10. AASHTO/NSBA. G12.1: Guidelines to Design for Constructability and Fabrication, 4th
Edition, NSBAGDC-4-OL, American Association of State Highway and Transportation
Officials, Washington, DC, 2020.

11. Helwig, T., Yura, J., Herman, R., Williamson, E., and Li, D. “Design Guidelines for
Trapezoidal Box Girder Systems,” Report No. FHWA/TX-07/0-4307-1, Center for
Transportation Research, University of Texas at Austin, Austin, TX, 2007.

12. Fan, Z. and Helwig, T. “Behavior of Steel Box Girders with Top Flange Bracing,” Journal of
Structural Engineering, American Society of Civil Engineers, Reston, VA, Vol. 125, No. 8,
pp. 829-837, 2009.

13. NSBA. Steel Bridge Design Handbook: Bracing System Design. National Steel Bridge
Alliance, 2022.

109
14. White, D. W., Grubb, M. A., King, C. M., and Slein, R. “Proposed AASHTO Guidelines for
Bottom Flange Limits of Steel Box Girders”. Final Report for NCHRP Project 20-07, Task
415, National Cooperative Highway Research Program, Transportation Research Board,
Washington, DC, 2019.

15. AASHTO/AWS. Bridge Welding Code, BWC-8 (AASHTO/AWS D1.5M/D1.5:2020).

American Association of State Highway and Transportation Officials and American Welding
Society, Washington, DC, 2020.

16. NCHRP. National Cooperative Highway Research Project Report 883: Fracture-Critical
System Analysis for Steel Bridges. National Cooperative Highway Research Program,
Transportation Research Board, Washington, DC, 2018.

17. AASHTO. Guide Specifications for Analysis and Identification of Fracture Critical Members
and System Redundant Members. First Edition, American Association of State Highway and
Transportation Officials, Washington, DC, 2018.

18. Connor, R. J., Korkmaz, C., Campbell, L. E., Bonachera Martin, F. J., and Lloyd, J. B.. A
Simplified Approach to Design Composite Continuous Twin Tub-Girder Bridges as
Redundant Structures. Purdue University, West Lafayette, IN, 2020.

19. AASHTO/NSBA. G9.1: Steel Bridge Bearing Design and Detailing Guidelines, 1st Edition,
NSBASBB-1-OL, American Association of State Highway and Transportation Officials,
Washington, DC, 2004.

110
 Smarter. Stronger. Steel.
National Steel Bridge Alliance
312.670.2400 | aisc.org/nsba

B955-22