Bridge Basics

Because of the wide range of structural possibilities, this Spotter's Guide shows only the most
common fixed (non-movable) bridge types. Other types are listed in the Bridge
Terminology page. The drawings are not to scale. Additional related info is found on the
other Terminology pages which are linked to the left.

The four main factors are used in describing a bridge. By combining these terms one may give a
general description of most bridge types.

 span (simple, continuous, cantilever),

 material (stone, concrete, metal, etc.),

 placement of the travel surface in relation to the structure (deck, pony, through),

 form (beam, arch, truss, etc.).

The three basic types of spans are shown below. Any of these spans may be constructed using
beams, girders or trusses. Arch bridges are either simple or continuous (hinged). A cantilever bridge
may also include a suspended span.

Examples of the three common travel surface configurations are shown in the Truss type drawings
below. In a Deck configuration, traffic travels on top of the main structure; in a Pony configuration,
traffic travels between parallel superstructures which are not cross-braced at the top; in
a Through configuration, traffic travels through the superstructure (usually a truss) which is cross-
braced above and below the traffic.
Beam and Girder types
Simple deck beam bridges are usually metal or reinforced concrete. Other beam and girder types
are constructed of metal. The end section of the two deck configuration shows the cross-bracing
commonly used between beams. The pony end section shows knee braces which prevent deflection
where the girders and deck meet.

One method of increasing a girder's load capacity while minimizing its web depth is to add
haunches at the supported ends. Usually the center section is a standard shape with parallel
flanges; curved or angled flanged ends are riveted or bolted using splice plates. Because of the
restrictions incurred in transporting large beams to the construction site, shorter, more manageable
lengths are often joined on-site using splice plates.

Many modern bridges use new designs developed using computer stress analysis. The rigid
frame type has superstructure and substructure which are integrated. Commonly, the legs or the
intersection of the leg and deck are a single piece which is riveted to other sections.

Orthotropic beams are modular shapes which resist stress in multiple directions at once. They
vary in cross-section and may be open or closed shapes.
Arch types
There are several ways to classify arch bridges. The placement of the deck in relation to the
superstructure provides the descriptive terms used in all bridges: deck, pony, and through.

Also the type of connections used at the supports and the midpoint of the arch may be used - -
counting the number of hinges which allow the structure to respond to varying stresses and loads.
A through arch is shown, but this applies to all type of arch bridges.

Another method of classification is found in the configuration of the arch. Examples of solid-
ribbed, brace-ribbed (trussed arch) and spandrel-braced arches are shown. A solid-ribbed arch
is commonly constructed using curved girder sections. A brace-ribbed arch has a curved through
truss rising above the deck. A spandrel-braced arch or open spandrel deck arch carries the deck on
top of the arch.

Some metal bridges which appear to be open spandrel deck arch are, in fact, cantilever; these rely
on diagonal bracing. A true arch bridge relies on vertical members to transmit the load which is
carried by the arch.
The tied arch (bowstring) type is commonly used for suspension bridges; the arch may be trussed
or solid. The trusses which comprise the arch will vary in configuration, but commonly use Pratt or
Warren webbing. While a typical arch bridge passes its load to bearings at its abutment; a tied arch
resists spreading (drift) at its bearings by using the deck as a tie piece.

Masonry bridges, constructed in stone and concrete, may have open or closed spandrels A closed
spandrel is usually filled with rubble and faced with dressed stone or concrete. Occasionally,
reinforced concrete is used in building pony arch types.

Truss - simple types

A truss is a structure made of many smaller parts. Once constructed of wooden timbers, and later
including iron tension members, most truss bridges are built of metal. Types of truss bridges are
also identified by the terms deck, pony and through which describe the placement of the travel
surface in relation to the superstructure (see drawings above). The king post truss is the simplest
type; the queen post truss adds a horizontal top chord to achieve a longer span, but the center
panel tends to be less rigid due to its lack of diagonal bracing.

Covered bridge types (truss)

Covered bridges are typically wooden truss structures. The enclosing roof protected the timbers
from weathering and extended the life of the bridge.

One of the more common methods used for achieving longer spans was the multiple kingpost
truss. A simple, wooden, kingpost truss forms the center and panels are added symmetrically. With
the use of iron in bridge construction, the Howe truss - - in its simplest form - - appears to be a
type of multiple kingpost truss.

Stephen H. Long (1784-1864) was one of the U.S. Army Topographical Engineers sent to explore
and map the United States as it expanded westward. While working for the Baltimore and Ohio
Railroad, he developed the X truss in 1830 with further improvements patented in 1835 and 1837.
The wooden truss was also known as the Long truss and he is cited as the first American to use
mathematical calculations in truss design.

Theodore Burr built a bridge spanning the Hudson River at Waterford, NY in 1804. By adding a arch
segments to a multiple kingpost truss, the Burr arch truss was able to attain longer spans. His
truss design, patented in 1817, is not a true arch as it relies on the interaction of the arch segments
with the truss members to carry the load. There were many of this type in the Pittsburgh area and
they continue to be one of the most common type of covered bridges. Many later covered bridge
truss types used an added arch based on the success of the Burr truss.

The Town lattice truss was patented in 1820 by Ithiel Town. The lattice is constructed of planks
rather than the heavy timbers required in kingpost and queenpost designs. It was easy to construct,
if tedious. Reportedly, Mr. Town licensed his design at one dollar per foot - - or two dollars per foot
for those found not under license. The second Ft. Wayne railroad bridge over the Allegheny River
was an unusual instance of a Town lattice constructed in iron.

Herman Haupt designed and patented his truss configuration in 1839. He was in engineering
management for several railroads including the Pennsylvania Railroad (1848) and drafted as
superintendent of military railroads for the Union Army during the Civil War. The Haupt
truss concentrates much of its compressive forces through the end panels and onto the
abutments.
Other bridge designers were busy in the Midwest. An OhioDOT web page cites examples of designs
used for some covered bridges in that state. Robert W. Smith of Tipp City, OH, received patents in
1867 and 1869 for his designs. Three variations of the Smith truss are still standing in Ohio
covered bridges.

Reuben L. Partridge received a patent for his truss design which is appears to be a modification of
the Smith truss. Four of the five Partridge truss bridges near his home in Marysville, Union
County, OH, are still in use.

Horace Childs' design of 1846 was a multiple king post with the addition of iron rods. The Childs
truss was used exclusively by Ohio bridge builder Everett Sherman after 1883.

Truss - Pratt variations

The Pratt truss is a very common type, but has many variations. Originally designed by Thomas
and Caleb Pratt in 1844, the Pratt truss successfully made the transition from wood designs to
metal. The basic identifying features are the diagonal web members which form a V-shape. The
center section commonly has crossing diagonal members. Additional counter braces may be used
and can make identification more difficult, however the Pratt and its variations are the most
common type of all trusses.

Charles H. Parker modified the Pratt truss to create a "camelback" truss having a top chord which
does not stay parallel with the bottom chord. This creates a lighter structure without losing
strength; there is less dead load at the ends and more strength concentrated in the center. It is
somewhat more complicated to build since the web members vary in length from one panel to the
next.

When additional smaller members are added to a Pratt truss, the various subdivided types have
been given names from the railroad companies which most commonly used each type, although
both were developed by engineers of the Pennsylvania Railroad in the 1870s.
The Whipple truss was developed by Squire Whipple as stronger version of the Pratt truss.
Patented in 1847, it was also known as the "Double-intersection Pratt" because the diagonal tension
members cross two panels, while those on the Pratt cross one. The Indiana Historical Bureau notes
one bridge as being a "Triple Whipple" -- possibly the only one -- built with the thought that if two
are better than one, three must be stronger yet.

The Whipple truss was most commonly used in the trapezoidal form -- straight top and bottom
chords -- although bowstring Whipple trusses were also built.

The Whipple truss gained immediate popularity with the railroads as it was stronger and more rigid
than the Pratt. It was less common for highway use, but a few wrought iron examples survive. They
were usually built where the span required was longer than was practical with a Pratt truss.

Further developments of the subdivided variations of the Pratt, including the Pennsylvania and
Baltimore trusses, led to the decline of the Whipple truss.

Truss - Warren variations

A Warren truss, patented by James Warren and Willoughby Monzoni of Great Britain in 1848, can
be identified by the presence of many equilateral or isoceles triangles formed by the web members
which connect the top and bottom chords. These triangles may also be further subdivided. Warren
truss may also be found in covered bridge designs.
Truss - other types
The other truss types shown are less common on modern bridges.

A Howe truss at first appears similar to a Pratt truss, but the Howe diagonal web members are
inclined toward the center of the span to form A-shapes. The vertical members are in tension while
the diagonal members are in compression, exactly opposite the structure of a Pratt truss. Patented
in 1840 by William Howe, this design was common on early railroads. The three drawings show
various levels of detail. The thicker lines represent wood braces; the thinner lines are iron tension
rods. The Howe truss was patented as an improvement to the Long truss which is discussed with
covered bridge types.

Friedrich August von Pauli (1802-1883) published details of his truss design in 1865. Probably the
most famous Pauli truss, better known as the lenticular truss -- named because of the lens
shape, is Pittsburgh's Smithfield Street Bridge. Its opposing arches combine the benefits of a
suspension bridge with those of an arch bridge. But like the willow tree, some of its strength is
expressed in its flexibility which is often noticeable to bridge traffic.

Before the use of computers, the interaction of forces on spans which crossed multiple supports
was difficult to calculate. One solution to the problem was developed by E. M. Wichert of Pittsburgh,
PA, in 1930. By introducing a open, hinged quadrilateral over the intermediate piers, each span
could be calculated independently. The first Wichert truss was the Homestead High Level Bridge
over the Monongahela River in 1937.

The composite cast and wrought iron Bollman truss was common on the Baltimore and Ohio
Railroad. Of the hundred or so following Wendell Bollman's design, the 1869 bridge at Savage, MD,
is perhaps the only intact survivor. Some of the counter bracing inside the panels has been omitted
from the drawing for clarity.

Also somewhat common on early railroads, particularly the B&O, was the Fink truss - - designed
by Albert Fink of Germany in the 1860s.

Cantilever types - truss

A cantilever is a structural member which projects beyond its support and is supported at only one
end. Cantilever bridges are constructed using trusses, beams, or girders. Employing the cantilever
principles allows structures to achieve spans longer than simple spans of the same superstructure
type. They may also include a suspended span which hangs between the ends of opposing
cantilever arms.

Some bridges which appear to be arch type are, in fact, cantilever truss. These may be identified by
the diagonal braces which are used in the open spandrel. A true arch bridge relies on vertical
members to transfer the load to the arch. Pratt and Warren bracing are among the most commonly
used truss types.

The classic cantilever design is the through truss which extends above the deck. Some have trusses
which extend both above and below the deck. The truss configuration will vary.
Suspension types
The longest bridges in the world are suspension bridges or their cousins, the cable-stayed bridge.
The deck is hung from suspenders of wire rope, eyebars or other materials. Materials for the other
parts also vary: piers may be steel or masonry; the deck may be made of girders or trussed. A tied
arch resists spreading (drift) at its bearings by using the deck as a tie piece.

Though Pittsburgh has been a pioneer in bridge design and fabrication, it has had few suspension
bridges. The Pennsylvania Mainline Canal entered the city on John Roebling's first wire-rope
suspension bridge in 1845 (replacing a failing 1829 wooden structure). A similar structure still
stands at Minnisink Ford, NY, crossing the Delaware River. Roebling and his son Washington
Roebling, later famous in building the Brooklyn Bridge, began their work in Saxonburg, PA, north of
Pittsburgh.

Beams and Girders

Girders
R.K. Bharil, in Innovative Bridge Design Handbook (Second Edition), 2022

1 Introduction
Girder bridges are the most natural and simplest form of bridging between two points. Chances are very
high that today, a bridge engineer will learn how to design a bridge girder before any other bridge type.
The use of girders as a natural bridging element is abundantly evident in nature—such as a fallen tree
trunk over a stream or rock formations over eroded soil—providing both people and animals dry and
safe access across an obstacle. The use of girders as a human-made bridging element probably evolved
as an outdoor extension of an indoor dwelling’s floor or roofing system.
The span length and the site conditions often dictate the type of bridge that can be feasible at a given
site. There are physical and economic limitations, and the bridge selection process often starts by
considering a simple culvert, progressing to a slab or girder system, and ultimately evolving to truss and
other more complex systems if and when needed. Figure 15.1 shows the commonly used and
economical span ranges of various bridge types. Keep in mind that there are often exceptions to the
recommended bridge type selection driven by aesthetic preferences, special site conditions,
environmental regulations, political influence, and many other factors.

Figure 15.1. Common span range, by bridge type. Compiled in part from California Department of Transportation
(2019) and Washington State Department of Transportation (2020).
Originally, the girder selection relied on time-proven depth-to-span ratios that controlled deflections and
served the function of carrying the load. The most commonly used span-to-depth ratios for various
popular bridging elements are described in Figure 15.2. The primary function of these ratios is to control
live load deflections and vibrations; however, modern innovation is constantly pushing these ratios
toward leaner and more efficient systems.
Figure 15.2. Common bridge girder depth-to-span ratios. Compiled in part from California Department of Transportation (2020)
and Washington State Department of Transportation (2020).
As structural analysis methods evolved, moments and shear were added to the beam equation,
and factors of safety were used to guard against uncertainties in building materials and prevalent loads.
As the girder shape evolved from untreated logs, sawn timber, and cut stones to steel, the material’s
properties began to play a greater role in its selection. As analysis and design methodologies progressed,
girder bridges became more complex—from simple rectangular beams to fabricated or rolled shapes,
concrete with steel reinforcement, concrete with prestressing strands, and various other complex
structural systems such as stringer-floor beams and box girders.
Today’s girder bridges consist of the elements described next.
The primary structural elements are as follows:
•
Girders—Transfer load to substructure elements (the primary focus of this chapter)
•
Deck—Provides a riding surface and transfers external loads to stringers or girders
•
Stringers—Transfer load from slab to floor beams (not always present)
•
Floor beams—Transfer load from stringers to girders (not always present)
The secondary structural elements are as follows:
•
Diaphragms—Provide stability to girders during construction (often eliminated due to their
initial cost)
•
Barriers and railings—Serve as a traffic safety element and confine external loads to the
designated riding surface
•
Bearings—Transfer loads to substructure elements while providing for superstructure rotation
and translation
•
Joints—Allow movements of superstructure segments to thermal, shrinkage, and seismic
demands (used sparingly to reduce maintenance costs)
The substructure elements are as follows:
•
Abutments, wing walls, and approach slabs—Connect the bridge structure to the roadway
embankment
•
Pier caps and crossbeams—Transfer loads to columns or piles
•
Piers, bents, and columns—Transfer loads to foundation interface elements
•
Footing and pile caps—Transfer loads to soil/rock strata or other foundation elements
•
Piles, shafts, and caissons—Transfer loads to final soil/rock strata via bearing, friction, or both
Before discussing a design based on geometry, let’s define a bridge girder correctly. Many bridge
inspection manuals define a girder as a longitudinal bridge element that supports the deck slab carrying
external loads and transmits the load to substructure elements such as bearings or abutment/pier caps or
crossbeams. A stringer is defined as a similar longitudinal element that transmits loads to other
superstructure elements (such as a floor beams) and is typically a part of a more elaborate bridge type
such as a truss or a cable-supported system. Other names such as beam and joist are also interchangeably
used but do not necessarily refer to the term girder that defines the bridge. This chapter defines a girder
bridge as a bridge whose primary load-carrying members are girders oriented along the direction of the
traffic.
Due to its inherent simplicity, the girder bridge is the most common form of bridge. The bridging of two
distant points by joining them by a straight line is not only intuitive but also a very efficient form of
overall space planning. For example, with girders, there is little loss of vertical clearance below to
accommodate an arch springing line or deck truss, there are no overhead constraints to accommodate the
lateral bracings of a through truss, and there is no complicated geometry of overhead cables or tied
arches. It requires relatively simple formwork or erection procedures and is often a first choice.
However, a girder bridge eventually loses out to other complex forms as new geometry constraints begin
to play key roles, spans become longer, or construction access becomes difficult. Such limitations are
further given in Figure 15.1 and are described in detail in the following sections.

Underground Construction
Bai Yun, in Underground Engineering, 2019

4.2.2.2 Lattice Girders

Lattice girders (Fig. 4.37) are lightweight triangular steel frames. In crown heading excavation, for
example, immediate support should be applied to the excavation with lattice girders. They are used in
conjunction with shotcrete and act as armor to achieve a good level of support in tunneling. To provide
additional forward support, strata bolts can be inserted through the lattice girders (Tunnel Ausbau
Technik, n.d.).

Figure 4.37. Tunnel lattice girders.

I beam girders (Fig. 4.38) are relative heavy steel frames. They are widely used in China tunneling
projects for weak rock formation. However, the use of I beam girders will have a bad influence on
shotcrete quality due to their geometric shape.

Figure 4.38. I beam girders.

Buckling/Plastic Collapse Behavior and Strength of Plate Girders

Subjected to Combined Bending and Shear Loads
Tetsuya Yao, Masahiko Fujikubo, in Buckling and Ultimate Strength of Ship and Ship-Like Floating
Structures, 2016

7.1 Research on Buckling of Plate Girders in Ship and Ship-Like Floating Structures
Plate girders consisting of web and flange plates have been main longitudinal strength members in
single-hull ship structures. They are also main transverse strength members. The plate girders are
usually subjected to combined shear and bending/compression loads. Basler performed a series of
experiments to investigate into the collapse behavior of plate girders subjected to shear and bending
loads and derived design formulas to evaluate ultimate strength [1–4]. Akita and Fujii [5, 6] and Fujii [7]
revised Basler’s formulas, taking into account of the collapse of flange due to lateral load produced by
the action of tension field in the web panel after shear buckling has occurred.
These formulas are, however, for plate girders with free flanges of a finite width such as deck or bottom
girders or transverse rings in single hull tankers. In 1992, MARPOL convention [8] came into effect and
all tankers had become to have double hull structures. Because of this, flange of bottom girders has
become a part of continuous plating such as inner bottom plating or longitudinal bulkhead. The isolated
free flange and continuous plating may show different buckling behavior as a flange of the girder during
buckling/plastic collapse.
Regarding the collapse behavior and strength of plate girders of which flanges is a part of the continuous
plating, Olaru et al. [9] and Olaru [10] performed a series of nonlinear finite element method (FEM)
analysis to clarify its buckling/plastic collapse behavior and the ultimate strength. The influences
of stiffeners and perforation in the web panel are also examined.

Scantling of Ship's Hulls by Rules

Yong Bai, Wei-Liang Jin, in Marine Structural Design (Second Edition), 2016
Girders
Girders are to comply with the same scantling criteria as stiffeners with respect to the section modulus.
In addition, shear force should be considered, due to the height of the girder (Figure 8.10). The
following equation represents the scantling criterion in terms of the cross-sectional area of the girders.

Figure 8.10. Girder.

(8.16)τ=QA
where τ is the shear stress at the girder end in N/m2 and A is the cross-sectional area at the girder end in
m2. If the load is equally distributed, with each end of the girder carrying half the load, Q will be defined
as
(8.17)Q=0.5·p·b·S
where p and b denote the design pressure acting on the girder (N/m2), and the loading breadth (m). The
girder span is denoted as S (m). Substituting Eqn (8.17) into Eqn (8.16), the following equation is
obtained.
(8.18)τ=QA=0.5·p·b·SA
From Eqn (8.18), the required sectional area is derived as the following
(8.19)A≥0.5·p·b·Sτall
The allowable shear stress τall depends on the girder. In addition, girders are used to satisfy the
requirements of the web plate thickness, the girder web area, and the ratio of the girder flange thickness
to flange width.

Floating Offshore Platform Design

John Halkyard, in Handbook of Offshore Engineering, 2005
Girder Buckling
Girders are connected to the plate and may fail by plate induced failure or flange induced failure. Plate
induced failure results in the girder deflecting away from the plate with yielding in compression at the
connection between the girder and the plate. Flange induced failure is caused by torsional buckling (see
above) and is typically mitigated by the addition of tripping brackets so that the unsupported length does
not exceed the value STO given by (dnv CN 30.1):
(7.104)STO=CEAfσf(Af+Aw/3)
where b= flange width, C= 0.55 for symmetric flanges, = 1.10 for one sided flanges.

Detailed Gate Design☆

Ryszard Daniel, Tim Paulus, in Lock Gates and Other Closures in Hydraulic Projects, 2019

8.2.4 Vertical Lift Gate and Radial Gate Framing

Horizontal girders form the main structural framing of a typical vertical lift gate. Horizontal girders
carry hydrostatic loading from the skin plate into the ends of the gate through simple span
action. Girder reactions are carried to the end posts or gate support wheels which carry all loads directly
into the supporting concrete monolith structure housing the gate. A framing diagram for a vertical lift
gate from the John Day Lock is shown in Fig. 8.23 that also shows both linear and pointed supports.
This is one of the largest vertical gates in the United States. Linear and pointed supports are discussed
further in Section 8.3. Intercostals, similar to those utilized on miter gates described above, decrease the
skin plate span and distribute loads along the length of the horizontal girders. Intercostals and
diaphragms also transmit the vertical lifting force of the gate along the height of the gate and into the
skin plate.
Fig. 8.23. John Day Navigation Lock vertical lift gate framing system also showing both linear and pointed supports (drawing by
USACE).
Horizontal girders are typically fabricated plate girders welded to the skin plate to achieve the maximum
cross-section required while minimizing total gate weight. To accomplish this, the upstream skin plate
may form a portion or the entire upstream flange of the horizontal girder. Horizontal girder spacing can
be adjusted over the height of the gate in order to minimize total gate weight and increase efficiency.
The deflection of the girder framing system must be checked in order to ensure continuous operation
without binding.
Girder connections must be detailed to coincide with design assumptions. In detailing girder to end post
connections, the effects of thermal stresses on large lock gates should be considered in addition to the
applied hydrostatic loading. Thermal stresses due to water on one side of the gate and sun/shadow on the
adjacent side of the gate as well as seasonal changes can have a global effect on the gate length, which
should be considered to prevent binding and leakage.
For a radial gate, the principal structural elements are the skin plate assembly, horizontal girders, end
frames, and trunnions. See Fig. 8.24 for a typical framing detail [18]. The skin plate assembly, which
forms a cylindrical damming surface, consists of a skin plate stiffened and supported by curved vertical
ribs. Structurally, the skin plate acts compositely with the ribs (usually structural T-sections) to form the
skin plate assembly. The skin plate assembly is supported by the horizontal girders that span the gate
width. The downstream edge of each rib is often attached to the upstream flange of the horizontal
girders. The horizontal girders are supported by the end frames. End frames consist of radial struts or
strut arms and bracing members that coincide at the trunnion which is anchored to the pier through the
trunnion girder. The end frames may be parallel to the face of the pier (support the horizontal girders at
the ends) or inclined to the face of the pier (support the horizontal girders at some distance away from
the end). The trunnion is the hinge upon which the gate rotates.

Fig. 8.24. Typical main radial gate framing components (drawing by USACE).
Radial gate ribs are dimensioned such that the skin plate forms a portion of the effective flange of the
rib. The ribs are designed to span continuously between main radial gate girders (typically 2, 3, or 4
girders). The top section of the skin plate assembly and the bottom section which extends beyond the
bottom girder are designed as a cantilever span beyond the girder with applied water pressure, wave
loads, radial rope pressure, and ice or debris impact loading. The bottom cantilever section will only see
water pressure, as opposed to the upper section which is subjected to hydrostatic, wind, wave, and ice or
debris loading.

Radial gate girders are dimensioned to withstand point loading from the rib skin assembly. See an
example in Fig. 8.25 from a medium-sized radial gate 4.5 m high × 10.6 m long on the Mississippi River
in the United States. Here, the radial gate girders continuously span between the struts without
cantilevers on each end. Skew reinforcements of the end frame sections may be adjusted to decrease the
girder span and subsequent required depth of the girder. The fixity of the girders at both ends results in a
reversal of curvature as the girder spans across the end frame.
Fig. 8.25. Lock 7 Mississippi River radial gate showing girders, skin plate, and struts (USACE).
Radial gate girders are typically fabricated from built up members of a wide flange section. These built
up girders, also called “plate girders”, are dimensioned with consideration for bending loads,
concentrated loads from end frame reactions and drainage requirements including drain holes.
Web stiffeners are required to resist concentrated loads. As the girders lay in a relatively horizontal
plane, both weld access holes and drain holes are required to ensure free drainage of water.

Although radial gate girders are nearly continuously braced on the compression flange where they are
welded or bolted to the skin plate assembly, the vertical lifting forces of the gate must be considered to
prevent lateral torsional buckling. In order to carry vertical lifting forces through the skin plate, into the
ribs and into the girders, the girders must often be vertically braced with bracing typically connecting
the downstream girder flanges (see Fig. 8.9). The end frame of a radial gate is composed typically of
either 2, 3, or 4 struts either built from wide flange sections or built up sections. The end frame struts are
dimensioned for axial and bending loads, the latter in and out of the frame plane. In addition to water
pressure, the end frame must resist trunnion friction moments created at the trunnion pin and thrust
washer for gates with both lateral and skewed end frames. The end frame often forms a truss in the
vertical plane to resist lifting forces such that bracing members between end frame struts ensure that
the dead weight of the gate is distributed among all struts of the end frame. An example of such a design
is shown in drawing (a) of Fig. 8.26. The gate model pictured is of an Italian manufacturer, but it also
reflects many typical American designs.
Fig. 8.26. Typical end frame arrangements of radial gates by European suppliers: (a) truss end arms with cylinder hook-up at the
lower chord, Vortex Hydra, Italy; (b) unbraced frame with cylinder hook-up at the upper strut, HST France; and (c) unbraced struts
with counterweight, DSD Noell, Germany.
In medium and small radial gates, the end frame is often reduced to two struts without additional
members in between. Such solutions, favored particularly by European designers and manufacturers,
often employ box sections instead of wide flange or built up sections in the gate strut arms. Examples
are sector gates shown in the lower drawings of Fig. 8.26; and delivered by manufacturers from
respectively France (b) and Germany (c). Note that the framing of both gates has been divided into
compact subassemblies that are easy to transport to the final location and assembled on site. The gate (b)
has assembly joints in the form of pin-hole connections; the gate (c) has flange joints to be bolted by
high strength bolts.

Radial gate end frames can also be skinned or wrapped in order to prevent the collection of debris within
the strut arms. The skinning or wrapping of a radial gate end frame and skin plate assembly is often
required for radial gates which are designed for overtopping conditions during period of high water or
flooding. In rare instances, the wrapper plate is assumed to provide for structural stability of the end
frame. In general, the end frame assembly should be dimensioned ignoring the added strength and
stiffness effects of the wrapper plate assembly, unless that plate has sufficient thickness and is designed
to provide stability under all conditions. The reason for dimensioning the gate in this fashion is that it
allows for corrosion of the wrapper plate to occur due to both abrasion from debris and galvanic
corrosion, without a loss of structural stability occurring.
While the conventional radial gate framing is an open, arch-shaped grid, there are tendencies in recent
decades to replace it by large box-shaped sections, in a way similar to gate model (c) from Fig. 8.26.
These tendencies are particularly strong in Europe. As mentioned in the skin plate discussion in Section
8.1.3, there have been a few catastrophic collapse events of radial gates in the recent decades, like the
gate failures of the Wachi Dam near Kyoto, Japan, and Folsom Dam near Sacramento, California [19].
These failures were primarily related to dynamic instabilities of the gates, like vibrations, or to corrosion
issues. To reduce vibration amplitudes, many engineers choose to increase overall stiffness by applying
large box sections or at least double-skinned retaining walls. More examples of such solutions have been
shown in Figs. 3.81, 3.82, 3.85, and 3.98.

Steel–concrete composite plate girders with web openings

N.E. Shanmugam, in Analysis and Design of Plated Structures (Second Edition), 2022

20.2.2 Girders subject to shear and bending

The girders tested under shear and bending showed similar behavior up to collapse. Fig. 20.2 shows the
deformed shape after failure of the girder CPG5. When the load was applied, the concrete slab subject to
tensile stresses started cracking under the load even at the initial stage. The crack width increased with
the loading and the number of cracks increased on the tension side. The web panels buckled after
reaching the elastic buckling load and further loading was resisted by tension field action in the web.
The slab started cracking over the supports near the end bearing stiffener locations and further loading
resulted in widening of cracks. The slab failed suddenly in shear mode when the applied load reached
the ultimate shear strength. The failure mode is similar to the one observed in the girders CPG1 and
CPG2 tested under shear loading. Similar cracking and deformation behavior was observed in girder
CPG6.

Figure 20.2. Deformed shape after failure of the girder CPG5.

The composite girders CPG7 and CPG8 were tested under positive bending. The slab lying in the
compression zone was effectively connected to the compression steel flange by means of shear studs. A
few hairline cracks were noticed at the bottom side of the slab near the midspan at the initial stages of
loading and no significant flexural cracks observed. After reaching the elastic buckling load, the web
buckled in shear, and further increase in load was resisted by the combined tension field mechanism and
axial tension due to bending. A view after failure of the girder CPG7 is shown in Fig. 20.3. The
objective of this investigation is to study the behavior of the middle two panels subject to combined
shear and bending. The end web panels were, therefore, stiffened with horizontal stiffeners. The girder
showed elastic behavior initially, and buckling of web was noticed first in the panel near the midspan
and then in the web panel near the support. As the loading was increased, the concrete slab near the
support gave way and led to total collapse of the girder. The girder CPG8 was tested in the same way as
CPG7. The behavior was similar to CPG7 from the initial stages of loading.
Figure 20.3. View after failure of the girder CPG7.
The tensile force in the steel part strengthened the girder to a certain extent and the load-carrying
capacity increased further after buckling in web panels. Fig. 20.4 shows the view after failure of the
girders CPG10. The girders CPG9 and CPG10 could be assumed as a double cantilever supported at the
midspan. The two girders behaved in a similar manner, and the extreme fibers of the concrete slab were
subjected to tensile cracking from the initial stages of loading. These cracks, formed near the midspan
region, spread over a distance of 800 mm on either side of the load. The top layer reinforcement yielded
at a load of around 40% of the ultimate failure load. Adequate composite action was observed between
the concrete slab and the steel girder. As the load was increased, the compression flange of the girder
CPG9 buckled, and the initial imperfection in the flange enhanced such buckling. In the girder CPG10,
on the other hand, no buckling of flanges was observed although the failure mechanism seemed to be
similar to the other girders.

Sign in to download full-size image

Figure 20.4. View after failure of the girder CPG10.

Fatigue strength of welded components, design improvements

Dieter Radaj Dr-Ing habil, in Design and Analysis of Fatigue Resistant Welded Structures, 1990

5.1.6 Box girders and deck plates

Box girders of varying sizes and applications are shown in Fig.84. The design for vehicle frames (a) can
tolerate the (through-welded) butt welds only, which are relatively favourable as regards the notch
effect, especially with respect to the shear flow by torsional moments superimposed on the bending
moments. The design for statically loaded compressive uprights (b) is completed using fillet welds. In
the design for box girders in arched bridges, the use of double bevel butt welds is preferable, as they are
more favourable as regards notch severity. This design therefore has to be accessible from inside and has
the transverse walls to retain the cross sectional contour. The design for girder bridges with wide spans
(d) is stiffened in small cells to reduce the risk of buckling whilst the cross sectional contour is secured
by diagonal frame members. It is particularly stiff against torsion and therefore stable with respect to
torsion buckling.
84. Box girder subjected to bending, design variants, partially after Neumann and Stüssi and Dubas.
3 10

The deck plate of girder or truss bridges is fitted with longitudinal stiffener ribs or rectangular or
trapezoidal hollow section stiffeners in the direction of traffic to transfer the loading on the deck to the
transverse girders. To avoid fit-up problems and weld crossings, the longitudinal stiffeners are combined
with cutouts as in Fig.85 (also usual in shipbuilding). Trapezoidal hollow section stiffeners for railway
bridges which are subjected to high levels of cyclic load are designed as in Fig.86 to minimise the notch
effect.78

85. Longitudinal stiffeners of bridge deck plate combined with cutouts, after Neumann. 3

Sign in to download full-size image

86. Longitudinal through stiffeners of railway bridges combined with cutouts, after Haibach and Plasil. 78
Structures
Lorraine M. Cahill, in Highway Engineering (Second Edition), 2022

8.3.4.2 Steel girder

Steel girders are often wide-flange designs, designated as W-beams. Often, the largest rolled beams
available have a height of 36 in. If a larger beam is required due to a longer span, appropriate size
girders must be assembled from steel plate.
Consider a rolled beam designated as a W36 × 300. The “W” stands for “wide-flange,” in contrast to a
standard I-beam. The “36” is the nominal height in inches (the actual beam may be slightly taller). The
“300” is for 300 lb/ft. Therefore, if this was the chosen beam, then the contribution to the distributed
load w would be:
(8.11)wsteel girder=300lbft(W36×300)

At the start of the LFD process, the structural engineer uses judgment and experience to select an
appropriate beam. This initial selection then goes through several viability tests at various stages in the
design process, the most important of which is its section modulus compared to the section modulus
required. If the section modulus of the selected beam is greater than the required section modulus, the
design is satisfactory. If it is not, a larger beam will be required, usually with a larger weight per linear
foot. The distributed load w is then recalculated and the process iterated until a satisfactory beam is
selected.

Structures
Thomas H. BrownJr. PhD, PE, in Highway Engineering, 2016

Steel Girder
Steel girders are typically wide-flange designs, designated as W-beams. Typically, the largest rolled
beams available have a height of 36 in. If a larger beam is required due to the longer spans needed in
today’s highway systems, appropriate size girders must be assembled from steel plate.
For discussion, consider a rolled beam designated as a W36×300. The “W” stands for “wide-flange,” in
contrast to a standard I-beam. The “36” is the nominal height in inches (the actual beam may be slightly
taller). The “300” is for 300 lb/ft. Therefore, if this was the chosen beam, then the contribution to the
distributed load w would be:

(8.11) wsteelgirder=300lb/ft(W36×300)

Typically, at the start of the design process, the structural engineer, from experience, selects a beam to
be used. This initial selection then goes through several tests of its applicability at various stages in the
design process, the most important of which is its section modulus compared to the section modulus
required. If the section modulus of the selected beam is greater than the required section modulus, then
the design is satisfactory. However, if not, then a larger beam will be required, usually having a larger
weight per foot. The calculation for the distributed load w must then be redone and the process repeated
until a satisfactory beam is selected.