LITERATURE REVIEW

1. E. Bhargavi and G.V. Rama Rao (2015), discussed about comparative parametric
study of steel bridge trusses by applying external prestressing. The aim of the present
analytical work is to know the effect of Pre-stressing on the member forces,
deflections and total weight of steel of a statically determinate three types of trusses
such as Pratt type(Type A), Warren truss(Type B), Lattice Truss(Type C). Pre-
stressing technique has been adopted to upgrade the performance of the truss. The
truss is pre-stressed with high tensile steel cable and the profile of the cable is straight.
The truss is analysed for member forces and deflections using STAAD PRO Software.
This paper concluded that from the obtained analytical results, it is seen that there is a
noticeable improvement in the performance of the structure. Member forces have
been reduced significantly in the entire truss members and there is a reduction in
deflection at the centre and material requirement after pre-stressing.
2. Liang Xiao (2015), investigated on large span steel truss bridge finite element
simulation to investigate the boundary conditions In this paper, through theoretical
analysis and finite element software simulation, illustrates the principle of three kinds
of boundary selection, And according to the viewpoint of stress nephogram real
simulation presents a recommended boundary conditions which formed at both ends
simply supported constraints. This paper concluded that through the above analysis,
Three gusset plate boundary conditions (the left consolidation, the center for
consolidation, chord simply supported on both ends) simulated maximum Mises stress
both appear in the left vertical webs, and their values were about 234 MPa.
3. Ruly Irawan, Henricus Priyosulistyo (2014) at all, investigated about evaluation of
forces on a steel truss structure using modified resonance frequency. This research
shows that the error of estimation of member forces in the compression and tension
members using modified natural frequency and rotational spring parameter by linear
regression method varies from 0.26% to 1.99% and 0.2% to 2.41% respectively. The
value of rotational spring parameters indicates that the members have semi rigid
behaviour and closer to fixed rather than pinned conditions.
4. Akihiro MANDA and Shunichi NAKAMURA (2010), discussed about progressive
collapse analysis of steel truss bridges. Progressive collapse analysis is carried out for
the three continuous steel truss bridges using large deformation elastic plastic
analysis. It is intended to clarify how the live load intensity and distribution affect
structural safety and ductility for these two truss bridges. Although the collapse
process is different depending on live load distribution and length of the spans, the
steel truss bridge collapses due to plastic buckling or elastic buckling. It is found that
ductility of Model Bridge-B with a span ratio of 1:1.3:1 is larger than that of Model
Bridge-A with a span ration of 1:2:1.
 INTRODUCTION
1) TRUSS BRIDGES:
1.1 INTRODUCTION:
A truss bridge is a bridge whose load-bearing superstructure is composed of a truss, a
structure of connected elements usually forming triangular units. The connected
elements (typically straight) may be stressed from tension, compression, or sometimes
both in response to dynamic loads. Truss bridges are one of the oldest types of modern
bridges. The basic types of truss bridges shown in this article have simple designs
which could be easily analysed by 19th- and early 20th-century engineers. A truss
bridge is economical to construct because it uses materials efficiently.

Figure 1: Integral members of a truss bridge.

1.2 TYPES OF TRUSSES:

Trusses are triangular frame works in which the members are subjected to essentially
axial forces due to externally applied load. They may be plane trusses [Fig. 1(a)],
wherein the external load and the members lie in the same plane or space trusses [Fig.
1(b)], in which members are oriented in three dimensions in space and loads may also
act in any direction. Trusses are frequently used to span long lengths in the place of
Steel members subjected to axial forces are generally more efficient than members in
flexure since the cross section is nearly uniformly stressed. Trusses, consisting of
essentially axially loaded members, thus are very efficient in resisting external loads.
They are extensively used, especially to span large gaps. Since truss systems
consume relatively less material and more labour to fabricate, compared to other
systems, they are particularly suited in the Indian context. Trusses are used in roofs of
single storey industrial buildings, long span floors and roofs of multi-storey buildings,
to resist gravity loads [Figs. 1(a) and 1(b)].
 Figure 2: Types of trusses
Trusses are also used in multi-storey buildings and walls and horizontal planes of
industrial buildings to resist lateral loads and give lateral stability [Figs. 1(c) and
1(d)]. Trusses are used in long span bridges to carry gravity loads and lateral loads
[Fig. 1(e)].
1.3 CONFIGURATION OF TRUSSES:
i. Pitched Roof Trusses:
Most common types of roof trusses are pitched roof trusses wherein the top chord
is provided with a slope in order to facilitate natural drainage of rainwater and
clearance of dust/snow accumulation. These trusses have a greater depth at the
mid-span. Due to thiseven though the overall bending effect is larger at mid-span,
the chord member and web member stresses are smaller closer to the mid-span
and larger closer to the supports. The typical span to maximum depth ratios of
pitched roof trusses are in the range of 4 to 8, the larger ratio being economical in
longer spans. Pitched roof trusses may have different configurations. In Pratt
trusses [Fig. 2(a)] web members are arranged in such a way that under gravity
load the longer diagonal members are under tension and the shorter vertical
members experience compression. This allows for efficient design, since the short
members are under compression. However, the wind uplift may cause reversal of
stresses in these members and nullify this benefit. The converse of the Pratt is the
Howe truss [Fig. 2(b)]. This is commonly used in light roofing so that the longer
diagonals experience tension under reversal of stresses due to wind load.

Figure 3: Pitched roof trusses

 Fink trusses [Fig. 2(c)] are used for longer spans having high pitch roof, since the
web members in such truss are sub-divided to obtain shorter members.
Fan trusses [Fig. 2(d)] are used when the rafter members of the roof trusses have
to be sub-divided into odd number of panels. A combination of fink and fan [Fig.
2(e)] can also be used to some advantage in some specific situations requiring
appropriate number of panels.
Mansard trusses [Fig. 2(f)] are variation of fink trusses, which have shorter
leading diagonals even in very long span trusses, unlike the fink and fan type
trusses.
The economical span lengths of the pitched roof trusses, excluding the Mansard
trusses, range from 6 m to 12 m. The Mansard trusses can be used in the span
ranges of 12 m to 30 m.
ii. Parallel chord trusses:
The parallel chord trusses are used to support North Light roof trusses in industrial
buildings as well as in intermediate span bridges. Parallel chord trusses are also
used as pre-fabricated floor joists, beams and girders in multi-storey buildings
[Fig. 3(a)]. Warren configuration is frequently used [Figs. 3(b)] in the case of
parallel chord trusses. The advantage of parallel chord trusses is that they use
webs of the same lengths and thus reduce fabrication costs for very long spans.
Modified Warren is used with additional verticals, introduced in order to reduce
the unsupported length of compression chord members. The saw tooth north light
roofing systems use parallel chord lattice girders [Fig.3(c)] to support the north
light trusses and transfer the load to the end columns.

Figure 4: Parallel chord trusses

The economical span to depth ratio of the parallel chord trusses is in the range of
12 to 24. The total span is subdivided into a number of panels such that the
individual panel lengths are appropriate (6m to 9 m) for the stringer beams,
transferring the carriage way load to the nodes of the trusses and the inclination of
the web members are around 45 degrees. In the case of very deep and very
 shallow trusses it may become necessary to use K and diamond patterns for web
members to achieve appropriate inclination of the web members [Figs. 3(d), 3(e)].
iii. Trapezoidal Trusses:
In case of very long span length pitched roof, trusses having trapezoidal
configuration, with depth at the ends are used [Fig. 4(a)]. This configuration
reduces the axial forces in the chord members adjacent to the supports. The
secondary bending effects in these members are also reduced. The trapezoidal
configurations [Fig. 4(b)] having the sloping bottom chord can be economical in
very long span trusses (spans > 30 m), since they tend to reduce the web member
length and the chord members tend to have nearly constant forces over the span
length. It has been found that bottom chord slope equal to nearly half as much as
the rafter slope tends to give close to optimum design.

Figure 5: Trapezoidal trusses

1.4 TRUSS MEMBERS:
The members of trusses are made of either rolled steel sections or built-up sections
depending upon the span length, intensity of loading, etc. Rolled steel angles, tee
sections, hollow circular and rectangular structural tubes are used in the case of roof
trusses in industrial buildings [Fig. 5(a)]. In long span roof trusses and short span
bridges heavier rolled steel sections, such as channels, I sections are used [Fig. 5(b)].
Members built-up using I sections, channels, angles and plates are used in the case of
long span bridge trusses [Fig. 5(c)]
Access to surface, for inspection, cleaning and repainting during service, are
important considerations in the choice of the built-up member configuration. Surfaces
exposed to the environments, but not accessible for maintenance are vulnerable to
severe corrosion during life, thus reducing the durability of the structure. In highly
corrosive environments fully closed welded box sections, and circular hollow sections
are used to reduce the maintenance cost and improve the durability of the structure.
 Figure 6: Cross Sections of Truss Members
1.5 DESIGN CONSIDERATIONS:
I. Preliminary Design:
For the preliminary design of a composite truss the following data is needed:

The maximum bending moments and shear forces in the member

(a) at the construction stage (Ms, Vs),
(b) at the factored load acting at the limit state of collapse of the
composite section (Mc, Vcs).
the concrete slab (regular or composite) sizes and
the truss spacing.

The following are the steps in the preliminary design:

1. Decide on the depth of the truss girder.

The span to depth ratio of a simply supported composite truss is normally

15 to 20.
2. Develop the web member layout, usually using Warren configuration.

Use a slope of 30 degrees to horizontal to increase the opening and reduce

the number of connections.
3. Design the top chord member.

Force in the top chord member at the construction load, R, is calculated

from the corresponding moment, Mst, and the lever arm between the chord
members.
Size of the member is based on the member strength as governed by
lateral buckling between the lateral supports to the top chord until the
concrete hardens.
A minimum width of 120 mm for the top chord is usually acceptable to
support the decking in a stable manner during erection.
Minimum of 8 mm thickness of the leg of the compression chord is
required to weld the stud through the deck on to the leg.
 Vertical leg of the member should be adequate to directly weld the web
members.
Otherwise gusset may be required.

Figure 7: Moment Capacity of Steel and Composite Trusses

1. Design the bottom chord member.
Calculate the tension in the bottom chord, R b, at the factored load moment
using the following equation.
Rb = Mc/(Dt+ Ds - 0.5 XcXb)

Where Xc= (Ds Dp) Rb/Rc, , Dp = Depth of the profile, Rb Rt, R are the forces in the
bottom chord, top chord of steel truss and the force in concrete slab, respectively. Area of the
bottom chord and the bottom chord member shape may be designed based on this force, Rb,,
2. Check the slab capacity for the compression force at the limit state of collapse.
Considering the yield strength of the member.
The slab capacity is given by

where fck = cube strength of concrete and beff is the effective width of the concrete slab acting
integral with the truss.
3. Design the web member.
The maximum force in the web member is calculated by setting the
vertical component of the member force equal to the maximum shear force
in the truss.
The web member is designed to carry the force considering its yield
strength in tension and buckling strength in compression.

2) STEEL BRIDGE:
I.1 INTRODUCTION:
The main advantages of structural steel over other construction materials are its strength and
ductility. It has a higher strength to cost ratio in tension and a slightly lower strength to cost
ratio in compression when compared with concrete. The stiffness to weight ratio of steel is
much higher than that of concrete. Thus, structural steel is an efficient and economic material
in bridges. Structural steel has been the natural solution for long span bridges since 1890,
when the Firth of Forth cantilever bridge, the world's major steel bridge at that time was
completed.
Steel is indeed suitable for most span ranges, but particularly for longer spans. Howrah
Bridge, also known as Rabindra Setu, is to be looked at as an early classical steel bridge in
India. This cantilever bridge was built in 1943. It is 97 m high and 705 m long. This
engineering marvel is still serving the nation, deriding all the myths that people have about
steel.
 Figure 8: Howrah Bridge
1.1 Steel used in bridges:
Steel used for bridges may be grouped into the following three categories:
(i) Carbon steel: This is the cheapest steel available for structural users where stiffness is
more important than the strength. Indian steels have yield stress values up to 250 N/mm2 and
can be easily welded. The steel conforming to IS: 2062 - 1969, the American ASTM A36, the
British grades 40 and Euronorm 25 grades 235 and 275 steels belong to this category.
(ii) High strength steels: They derive their higher strength and other required properties
from the addition of alloying elements. The steel conforming to IS: 961 - 1975, British grade
50, American ASTM A572 and Euronorm 155 grade 360 steels belong to this category.
Another variety of steel in this category is produced with enhanced resistance to atmospheric
corrosion. These are called 'weathering' steels in Europe, in America they conform to ASTM
A588 and have various trade names like ' cor-ten'.
(iii) Heat-treated carbon steels: These are steels with the highest strength. They derive
their enhanced strength from some form of heat-treatment after rolling namely normalisation
or quenching and tempering.
The physical properties of structural steel such as strength, ductility, brittle fracture, weld
ability, weather resistance etc., are important factors for its use in bridge construction. These
properties depend on the alloying elements, the amount of carbon, cooling rate of the steel
and the mechanical deformation of the steel. The detailed discussion of physical properties of
structural steel is presented in earlier chapter.
1.3 Classification of steel bridges:
.
Steel bridges are classified according to

the type of traffic carried

the type of main structural system
the position of the carriage way relative to the main structural system

These are briefly discussed in this section.

a. Classification based on type of traffic carried:
Bridges are classified as
Highway or road bridges
Railway or rail bridges
Road - cum - rail bridges
b. Classification based on the main structural system:
Many different types of structural systems are used in bridges depending upon the
span, carriageway width and types of traffic. Classification, according to makeup of
main load carrying system, is as follows:
I. Girder bridges - Flexure or bending between vertical supports is the main
structural action in this type. Girder bridges may be either solid web girders or
truss girders or box girders. Plate girder bridges are adopted for simply
supported spans less than 50 m and box girders for continuous spans upto
250m. Cross sections of a typical plate girder and box girder bridges are
shown in Fig. 9 (a) and Fig 9 (b) respectively. Truss bridges [See Fig.9 (c)] are
suitable for the span range of 30 m to 375 m. Cantilever bridges have been
built with success with main spans of 300 m to 550 m. In the next chapter
girder bridges are discussed in detail. They may be further, sub-divided into
simple spans, continuous spans and suspended-and-cantilevered spans, as
illustrated in Fig.9

Figure 9 (a): Plate girder bridge section

Figure 9 (b): Box girder bridge section

Figure 9 (c): Some of the trusses used in steel bridges

 Figure 9: Typical girder bridges
II. Rigid frame bridges - In this type, the longitudinal girders are made
structurally continuous with the vertical or inclined supporting member by
means of moment carrying joints [Fig.7.4]. Flexure with some axial force is
the main forces in the members in this type. Rigid frame bridges are suitable
in the span range of 25 m to 200 m.

Figure 10: Typical arch bridges

III. Cable stayed bridges Cables in the vertical or near vertical planes support
the main longitudinal girders. These cables are hung from one or more tall
towers, and are usually anchored at the bottom to the girders. Cable stayed
bridges are economical when the span is about 150 m to 700 m. Layout of
cable stayed bridges are shown in Fig. 7.6.

Figure 11: Layout of cable stayed bridges

IV. Suspension bridges - The bridge deck is suspended from cables stretched
over the gap to be bridged, anchored to the ground at two ends and passing
over tall towers erected at or near the two edges of the gap. Currently, the
 suspension bridge is best solution for long span bridges. Fig. 7.7 shows a
typical suspension bridge.

Figure 12: Suspension bridge

c. Classification based on the position of carriageway:

The bridges may be of the "deck type", "through type" or "semi-through type". These
are described below with respect to truss bridges:
(i) Deck Type Bridge: The carriageway rests on the top of the main load carrying
members. In the deck type plate girder bridge, the roadway or railway is
placed on the top flanges. In the deck type truss girder bridge, the roadway or
railway is placed at the top chord level as shown in Fig. 12 (a).

Figure 12: Typical deck, through and semi-through type truss bridges
(ii) Through Type Bridge The carriageway rests at the bottom level of the main
load carrying members [Fig. 12 (b)]. In the through type plate girder bridge,
the roadway or railway is placed at the level of bottom flanges. In the through
type truss girder bridge, the roadway or railway is placed at the bottom chord
level. The bracing of the top flange or lateral support of the top chord under
compression is also required.
(iii) Semi through Type Bridge - The deck lies in between the top and the bottom
of the main load carrying members. The bracing of the top flange or top chord
under compression is not done and part of the load carrying system project
above the floor level as shown in Fig. 12 (c). The lateral restraint in the system
is obtained usually by the U-frame action of the verticals and cross beam
acting together.
3) FINITE ELEMENT METHOD CONCEPT:
TRUSS STRUCTURES:
Truss structures constitute a special class of structures in which individual straight members
are connected at joints. The members are assumed to be connected to the joints in a manner
that permit rotation, and thereby it follows from equilibrium considerations, to be detailed in
the following, that the individual structural members act as bars, i.e. structural members that
can only carry an axial force in either tension or compression. Often the joints do not really
permit free rotation, and the assumption of a truss structure then is an approximation. Even if
this is the case the layout of a truss structure implies that it can carry its loads under the
assumption that the individual members act as bars supporting only an axial force. This
greatly simplies the analysis of the forces in the structure by hand calculation and
undoubtedly contributed to their popularity e.g. for bridges, towers, pavilions etc. up to the
middle of the twentieth century.
Counting joints and bars:
Some typical planar trusses are shown in Figure 13. It is seen that they are formed by
triangles, and this suggests that they are statically determinate, when supported appropriately
by three independent reaction components. It is now demonstrated by a common method for
planar trusses, that they are indeed statically determinate. The method leads to a necessary
relation between the number of joints and the number of bars. However, and probably equally
important, it identies a rational way of thinking about a truss structure, in which a process is
constructed by which the structure is extended joint by joint, simulating an actual
construction of the truss from bar elements connected by joints. For a planar truss the
hypothetical construction process starts from a simple triangle, and in order to be specic this
triangle is supported by a xed and a movable support as shown in Fig. 14a. Equilibrium of
the nodes can be established by two projection equations for the unsupported node, and a
vertical projection equation of the forces on the node with the moveable support. This gives
three equations, corresponding to the three bar forces to be determined. Thus, the initial
triangle is statically determinate.

Figure 13: a) V-truss, b) N-truss, c) Roof truss, d) K-truss.

 Figure 14: Construction of plane truss girders by triangles.
The process is continued by attaching a new joint by two new bars as illustrated in Fig. 14. If
the bars are not parallel, they will uniquely determine the position of the new joint, and two
projection equations for the forces on the new joint will determine the bar forces. This step, in
which a new joint is added and fastened by two new bars, can be continued as illustrated in
the gure. The process denes a simple relation between the number of bars b and the
number of joints j in a statically determinate planar truss:
b =2j 3
This relation is easily veried by observing that it is correct for the original triangle with
b=3andj = 3, and that inclusion of one new joint leads to two additional bars. The relation
between the number of bars and the number of joints is necessary, but clearly not sucient.
This becomes obvious e.g. by considering removing one of the diagonals and re-joining it as
a diagonal crossing the remaining diagonal. Hereby part of the truss becomes kinematically
indeterminate, implying a mechanism, while the additional bar in the remaining structure
makes this part statically indeterminate. Thus, the process, in which a gradual construction of
the truss by statically determinate steps is imagined, is probably more valuable than the
formula, if left alone.

Figure 15: Change of supports after constructing the truss.

At rst sight it may appear that the process is dependent on the supports being applied to the
initial triangle. However, this is not the case. The result is independent of the specic support
conditions as long as they provide three independent reaction components. After completing
the truss structure, the supports can be moved as illustrated in Fig. 15.

Figure16: Construction of space truss by addition of tetrahedra.

The results for planar trusses are easily extended to space trusses as illustrated in Fig. 16. The
starting point is a tetrahedron (pyramid), formed by 4 joints and 6 connecting bars. The
tetrahedron is supported by 6 independent reaction components. This leaves 4 3 6 = 6
equilibrium conditions from the 4 nodes for determination of the 6 bar forces. The process is
continued in steps consisting in the addition of 1 new joint connected by 3 new bars. The
three bars keep the joint xed in space, and the three force projection equations associated
with equilibrium of the new joint determine the three new bar forces. The gure shows the
two rst steps in this process leading to a truss girder of a type typically used for building
cranes. This leads to the following relation between the number of bars b and the number of
joints j of a statically determinate space truss:
b =3j 6.
Also in this case the relation is necessary but not sucient, and the imaginary process of
constructing the space truss constitutes an important part.
Method of joints:
The magnitude of the forces in the bars of a statically determinate truss structure can be
determined by the method of joints. The idea of the method of joints is to consider each joint
as separated from the rest of the truss structure by the introduction of a virtual section. The
parts on the two sides of the section will exchange identical but opposite forces, and by
introducing the section and identifying these forces explicitly, they can be analysed by the
equilibrium equations. The principle is illustrated in its simplest form in Fig. 17. The left part
of the gure shows a joint C in a planar truss loaded by the vertical force P and connected to
the rest of the truss by the two bars AC and BC. A section is now introduced, separating the
joint from the rest of the structure. The forces NAC and NBC, by which the bars act on the joint,
are indicated as acting on the joint together with the load P. Thus, the joint C is acted on by
three forces. The forces in the bars are considered as positive, when representing tension in
the bar. Thus, the eect on the joint is a force directed away from the joint. By the law of
action and reaction equal but opposite forces act on the bars. As seen, these forces represent
tension in the bars. It is noted that the forces N AC and NBC are uniquely dened as being
positive in tension. A representation in terms of vectors is less direct, as it would require
identication of the part on which the force acts.

Figure 17: Node C with load P and bar forces NAC and NBC
Equilibrium of the joint C requires that two force projection equations are satised. Vertical
projection gives
By taking a vertical projection, the force N AC in the horizontal bar AC does not contribute to
the equilibrium equation can be determined by projection on the direction orthogonal to BC.
The present case is simple due to the angle 45. The remaining bar force N AC, and gives NAC=P
directly. In many cases it will be more convenient to use a horizontal projection, whereby

Thus, there is compression in the inclined bar BC, while the horizontal bar AC is in tension to
ensure horizontal equilibrium. In this simple illustration there were only two bar forces, and
thus they could be determined directly by the two equilibrium equations available for the
planar joint C. Most joints in truss structures are connected by more bars than there are
equilibrium equations available for the particular joint. The bar forces can therefore only be
determined sequentially, if the joints are considered in a certain order. This is illustrated in the
following example.

Planar truss structures:

Many truss structures can conceptually be broken down into planar parts, and this section
illustrates the calculation of bar forces for some simple planar trusses.
Example: V-truss by the method of joints. Figure 2.14 shows a simply supported
V-truss, loaded by a concentrated force at B. The support conditions permit the vertical
reaction components RA and RC plus the horizontal reaction RA. They are determined from the
equilibrium conditions for the total truss structures. The reactions follow from horizontal
projection, moment about C and moment about A as

Control by vertical projection gives RA+ RC= P, corresponding to the load P.

The loading is symmetric, and because the horizontal reaction component R A vanishes, so are
the reactions. Thus, the structure and its bar forces are symmetric with respect to a vertical
line through B, and only the right half with the nodes B, C and D need to be considered to
determine all the bar forces. These nodes and the corresponding forces from loads, reactions
and bars are shown in Fig. 2.15. The loading is symmetric, and because the horizontal
reaction component R It is seen from the gure that joint C only contains two unknown
forces. Once they are determined, the joint D only contains two unknown forces. These
forces, and their symmetric counterparts, determine equilibrium at joint B which can be used
to check the previous calculations.
STEELBRIDGE:
1. Finite Element Method
1.1 Basic Concept
The finite element method is also an approximate structural analysis method (93). In this
method, a structure is divided into a series of finite elements. The actual deformation within a
finite element is replaced by approximate displacement functions. Then, the structure is
reassembled by the equilibrium conditions at the nodes. Engineers should realize that using
too few finite elements in bridge analysis may induce a significant error. There are a number
of different applications of the finite element method used in bridge analysis; these include
2D grid analysis methods, plate and eccentric beam analysis methods, generalized grid
analysis methods, and 3D FEM analysis methods. Each is described in subsequent sections of
this document.
1.2 2D Grid Analysis Method
This method is also referred to as plan grid or grillage analysis method. In this method the
structure is divided into plan grid elements with three degrees of freedom at each node
(vertical displacement, rotation angles about the longitudinal and transverse axes, or the first
derivative of the rotational angle about the longitudinal axis, or both). This method is most
often used in steel bridge design and analysis. The aspect ratio of the elements, node spacing,
and other modelling parameters are often set following simplified guidelines such as those
provided in Articles 4.6.3.3.1 and C4.6.3.3.1 of the AASHTO LRFD Bridge Design
Specifications (7).
1.3 Plate and Eccentric Beam Analysis Methods
This is a variant of a 2D grid/grillage analysis model. The deck is modelled using plate or
shell elements, while the girders and cross frames are modelled using beam elements offset
from the plate elements to represent the offset of the neutral axis of the girder or cross frame
from the neutral axis of the deck. This approach is discussed in Article C4.6.3.3.1 of the
AASHTO LRFD Bridge Design Specifications (7). The offset length is typically equal to the
distance between the centroids of the girder and deck sections. This method is somewhat
more refined than the traditional 2D grid method. For this modelling approach, beam element
internal forces obtained from this method need to be eccentrically transformed to obtain the
composite girder internal forces (bending moment and shear) used in the bridge design.
1.4 Generalized Grid Analysis Method
This is a modification of a 2D grid analysis, where more degrees of freedom are modelled.
Some typical enhancements that separate the generalized grid method from the 2D grid
method include modelling of cross frames or diaphragms with consideration of shear
deformation in addition to flexural deformation, modelling of the warping stiffness of open
cross section shapes (such as I shaped girders), modelling of girder supports, lateral bracing,
cross frames or diaphragms at their physical elevation within the structure, or combination
thereof (27, 60). At this time, the Generalized Grid Analysis method is used only for
academic research projects.
1.5 3D FEM Analysis Methods
The category 3D FEM Analysis methods is meant to encompass any analysis/design method
that includes a computerized structural analysis model where the superstructure is modelled
fully in three dimensions, including: modelling of girder flanges using line/beam elements or
plate/shell/solid type elements; modelling of girder webs using plate/shell/solid type
elements; modelling of cross frames or diaphragms using line/beam, truss, or plate/shell/solid
type elements (as appropriate); and modelling of the deck using plate/shell/solid elements.
Though this method is arguably deemed the most accurate analysis method available to
most practicing bridge design engineers, this method is also typically fairly time-consuming
and complicated, and is arguably deemed most appropriate for use for complicated bridges
(e.g., bridges with severe curvature or skew, or both, unusual framing plans, unusual
support/substructure conditions, or other complicating features). 3D analysis methods are also
useful for performing refined local stress analysis of complex structural details. Also, there
are some complicating factors associated with 3D analysis methods. For instance, in a 3D
analysis, girder moments and shears are not directly calculated. Instead, the model reports
stresses in flanges, webs, and deck elements. If the designer wishes to consider girder
moments and shears, some type of conversion/integration of the stresses over the depth of the
girder cross section will be required. This can be a significant undertaking, particularly with
regard to proper proportioning of deck stresses and deck section properties to individual
girders. When and how to use refined 3D FEM analysis for engineering design is a
controversial issue, and such an approach has not been fully incorporated into the AASHTO
specifications to date. The typical AASHTO methodology for design is generally based on
assessment of nominal (average) stresses calculated by simplified methods, such as P/A or
Mc/I, and not localized peak stresses obtained by shell- or solid-based finite element models.
Refined analysis can provide substantially more detailed and accurate information about the
stress state of the structure and allow for more cost effective and reliable design but this often
comes with increased engineering effort and increased potential for error. The results are
often more sensitive to the input parameters and the mathematical assumptions which are
employed by the software. For instance, a given element will have a unique formulation,
interpolation, integration and software implementation, all of which will affect results. The
engineer must understand the assumptions and limitations to ensure correct application.
These results are often difficult for the engineer to verify directly by independent
calculations, and special procedures must often be employed to verify accuracy of modelling.
2. Element Types
There is a wide range of different element types used in general finite element practice. Cook
et al. (39) provides a reasonably comprehensive overview. In short, the following definitions
are used for purposes of discussion in these guidelines:
Truss ElementA 2D or 3D element in which the responses are solely axial
tension/compression along the length of the component. 2D truss elements typically have two
translational degrees of freedom at each node, and 3D truss elements typically have three
translational degrees of freedom at each node.
Beam ElementA 2D or 3D element in which the responses involve both the axial
tension/compression, as in truss elements, as well as structural member flexure, and in the
case of 3D elements, structural member torsion. This type of beam element is sometimes
referred to as a frame element. Typical beam elements have six degrees of freedom (DOFs) at
each node, three translational DOFs, and three rotational DOFs. However, some more
advanced beam elements can include other DOFs to represent the warping of an open thin-
walled cross section (such elements are not commonly available in professional software
applications at the present time). In some circumstances, the term beam element may be
reserved to represent a 2D element that represents only flexural effects. However, a more
general definition involving axial, torsion, and flexure effects is used here.
Plate ElementA 2D element that consists typically of three to nine nodes. The internal
element responses generally consist of moments and shears. The result values are usually per
unit length of the plate. Plate elements can have various combinations of nodal degrees of
freedom.
Shell ElementA 3D element that combines the effects of plate bending as well as
membrane effects. Shell elements can be either flat or curved. Small flat shell elements can
be used to form curved surfaces.
Brick ElementA 3D element supporting three translational degrees of freedom per node.
The number of nodes can range from four to 20 or more. Brick elements generally have three
translational degrees of freedom at each node.
It must be stressed that not all elements in a given category perform equally and that
differences exist depending on the theories and numerical implementation used by the FEM
software developer. Engineers should review the theory manual and verify that the element
selected is appropriate to accurately understand and respond to the demands placed on it.
Engineers need not be mathematicians and computer programmers but they must understand
that all finite element methods have inherent approximations and they must understand how
some FE fundamentals such as element formulation, interpolation, integration, and software
implementation can influence the structural analysis results.
The element formulation refers to the mathematical theory used to define the elements
behaviour. For instance, shell problems generally fall into one of two categories: thin shell
problems or thick shell problems. For a detailed discussion on different shell formulations, as
well as proper integration order for the integration of their stiffness matrices, the reader is
referred to the book by Bathe (20). Thick shell problems assume that the effects of transverse
shear deformation are important to the solution. Thin shell problems, on the other hand,
assume that transverse shear deformation is small enough to be neglected. Thin shell
elements provide solutions to shell problems that are adequately described by classical
(Kirchhoff) plate theory, thick shell elements yield solutions for structures that are best
modelled by shear flexible (Mindlin) plate theory. Mindlin theory-based shell elements are
sometimes used in thin shell analysis because they have a less strict continuity requirement
on element interpolation functions.
The interpolation refers to the displacement functions that are assumed in the element
formulation for describing the deformed shape between the element nodes. It also refers to
the approximation and mathematical simplification of the original shape of the structure
under investigation. In most cases, the interpolation order is either linear or quadratic. Linear
interpolations are arguably the most common type, however quadratic or higher-order
elements are very efficient as the complexity of the domain shape and deformation increases.
Quadratic elements are more accurate on a per-element basis; however their use comes at an
increased computational expense since additional nodes are required to adequately describe
their shape.
The element integration refers to the number of discrete points within each element that are
utilized to calculate the internal strain energy in the deformed configuration, which affects
element stiffness matrices following the energy principle in solid mechanics.. Shell elements
can be either fully integrated or use reduced integration. For full integration, standard Gauss
quadrature is typically employed which results in four integration points for a four-node
quadrilateral and three integration points for a three-node triangular element. For reduced
integration, only a single integration point is used for each of these elements. Reduced
integration elements are attractive because they reduce computational expense while
providing a means for mitigating shear locking effects which become pronounced when shear
deformable shell formulations are used in situations where the through thickness dimension is
small. However, reduced integration elements often exhibit another numerical problem called
hour glassing, in which the element can deform in certain ways with the internal strain energy
remaining zero. Fully-integrated finite elements can in some cases exhibit numerical locking,
which is a phenomenon in which the numerical approximation leads to element responses
that are so over-stiff that the element becomes practically useless in approximating certain
types of response. Some finite element programs have formulations based on the use of
separate interpolation functions for internal stresses, displacements, and other specialized
procedures. These elements are typically aimed at providing improved accuracy while
avoiding spurious zero-energy modes and locking. Once the element type is selected and
verified for use, the engineer should perform a mesh convergence study to ensure that the
model is sufficiently refined to yield accurate results.

Modelling and Results

Fig. 15 Modelling of Steel truss Bridge

Fig. 16 End restraints

Fig. 17 Von -Mises Stresses

Fig. 18 Total Deformation

Fig. 19 Maximum combined stresses

Fig. 20 Axial force

Fig. 21 Total Bending moment

Fig. 22 Total Shear forces

REFERENCES:
1. E. Bhargavi and G.V. Rama Rao. Comparative Parametric Study of Steel Bridge
Trusses by Applying External Prestressing. Department Of Civil Engineering, Andhra
University, Visakhapatnam, Andhra Pradesh, India. www.ijetmas.com. July 2015.
2. Liang Xiao. Large span steel truss bridge finite element simulation to investigate the
boundary conditions. Chongqing Jiaotong University, Chongqing, China.
International Journal of Technical Research and Applications. May-June 2015.
3. Ruly Irawan, Henricus Priyosulistyo, Bambang Suhendro. Evaluation of forces on a
steel truss structure using modified resonance frequency. Department of Civil
Engineering and Environment, Gadjah Mada University, Yogyakarta, 55281,
Indonesia. www.sciencedirect.com. 2014
4. Akihiro MANDA and Shunichi NAKAMURA. Progressive Collapse Analysis of
Steel Truss Bridges. Proc. Schl. Eng. Tokai Univ., Ser. E. Sep. 27, 2010