Bridge Design Practice 4 • October 2022

CHAPTER 4
STRUCTURAL MODELING AND ANALYSIS
TABLE OF CONTENTS
4.1 INTRODUCTION .......................................................................................... 4-3
4.2 STRUCTURAL MODELING ......................................................................... 4-3
4.2.1 General ................................................................................................... 4-3
4.2.2 Structural Modeling Guidelines ............................................................... 4-7
4.2.3 Material Modeling Guidelines.................................................................. 4-9
4.2.4 Types of Bridge Models ........................................................................ 4-10
4.2.5 Slab-Beam Bridges ............................................................................... 4-11
4.2.6 Abutments............................................................................................. 4-17
4.2.7 Foundation ............................................................................................ 4-18
4.2.8 Examples .............................................................................................. 4-20
4.3 STRUCTURAL ANALYSIS......................................................................... 4-29
4.3.1 General ................................................................................................. 4-30
4.3.2 Analysis Methods .................................................................................. 4-30
4.4 BRIDGE EXAMPLES – 3-D VEHICLE LIVE LOAD ANAYSIS ................... 4-39
4.4.1 Background........................................................................................... 4-39
4.4.2 Moving Load Cases .............................................................................. 4-40
4.4.3 Live Load Distribution For One And Two-Cell Box Girders Example .... 4-42
NOTATION ............................................................................................................. 4-55
REFERENCES ....................................................................................................... 4-57

Chapter 4 – Structural Modeling and Analysis 4-1

© 2022 California Department of Transportation. ALL RIGHTS reserved.
 Bridge Design Practice 4 • October 2022

This page is intentionally left blank.

4-2 Chapter 4 – Structural Modeling and Analysis

© 2022 California Department of Transportation. ALL RIGHTS reserved.
 Bridge Design Practice 4 • October 2022

4.1 INTRODUCTION
Structural analysis is a process to analyze a structural system to predict its responses
and behaviors by using physical laws and mathematical equations. The main objective of
structural analysis is to determine internal forces, stresses, and deformations of structures
under various load effects.

Structural modeling is a tool to establish three mathematical models, including (1) a

structural model consisting of three basic components: structural members or
components, joints (nodes, connecting edges or surfaces), and boundary conditions
(supports and foundations); (2) a material model; and (3) a load model.

This chapter summarizes the guidelines and principles for structural analysis and
modeling used for bridge structures.

4.2 STRUCTURAL MODELING

4.2.1 General
For designing a new structure, connection details and support conditions should be
designed as close to the computational models as possible. For evaluating an existing
structure, the structural model should be as close to the actual as-built structural
conditions as possible. The correct choice of modeling and analysis tools/methods
depends on:

a) Importance of the structure

b) Purpose of structural analysis
c) Required level of response accuracy

This section will present modeling guidelines and techniques for bridge structures.

4.2.1.1 Types of Elements

Different types of elements may be used in bridge models to obtain expected

characteristic responses of a structure system. Elements can be categorized based on
their principal structural actions.

Chapter 4 – Structural Modeling and Analysis 4-3

© 2022 California Department of Transportation. ALL RIGHTS reserved.
 Bridge Design Practice 4 • October 2022

a) Truss Element

A truss (bar) element is a two-force member subject to axial loads either tension
or compression. It is used to model truss structures or pin-jointed frames. The
only degree of freedom for a truss element is the axial displacement at each
node. The cross-sectional dimensions and material properties of each element
are usually assumed constant along its length. The element may be
interconnected in a two-dimensional (2-D) or three-dimensional (3-D)
configuration.

b) Beam Element

A beam element is a member subject to lateral loads and moments. It is used

to model members in which one dimension (the length) is significantly greater
than the other two dimensions and only the stress in the direction along the
axis of the beam is significant. A 3-D beam element has six degrees of freedom
(DOF) at each node including translations and rotations. A 3-D beam element
under pure bending has only four degrees of freedom.

c) Frame Element

A frame element is a member subject to lateral loads, axial loads, and

moments. It is used to model framed structures since it possesses the
properties of both truss and beam elements and is also called a beam-column
element. A 3-D frame formulation includes the effects of biaxial bending,
torsion, axial deformation, and biaxial shear deformations. A frame element is
modeled as a straight line connecting two joints. Each element has its own local
coordinate system for defining section properties and loads.

d) Plate Element

A plate element is a 2-D solid element that acts like a flat plate. It is used to
model the bending deformation of plate structures and the resulting forces such
as shear forces and moments. There are two out-of-plane rotations and the
normal displacement as DOF. The element can model the two normal
moments and the cross moment in the plane of the element. The plate element
is a special case of a shell element without membrane loadings.

e) Shell Element

A shell element (Figure 4.2-1) is a 3-D solid element (one dimension is very
small compared with the other two dimensions) subject to plate bending, shear
and membrane loadings. A shell element may have either a quadrilateral shape
or a triangular shape. Shell element internal forces are reported at the element
mid-surface in force per unit length and are reported both at the top and bottom
of the element in force per unit area. It is primarily used to determine local stress
levels in cellular superstructure or in cellular piers and caissons. It is generally
recommended to analyze the full behavior unless the entire structure is planar

4-4 Chapter 4 – Structural Modeling and Analysis

© 2022 California Department of Transportation. ALL RIGHTS reserved.
 Bridge Design Practice 4 • October 2022

and is adequately restrained.

Figure 4.2-1 Shell and Solid Elements

f) Plane Element
The plane element is a 2-D solid, with translational DOF, capable of supporting
forces but not moments. One can use either plane stress elements or plane
strain elements. Plane stress element is used to model thin plate that is free to
move in the direction normal to the plane of the plate. Plane strain element is
used to model a thin cut section of a very long solid structure, such as walls.
Plain strain element is not allowed to move in the normal direction of the
element’s plane.

g) Solid Element
A solid element is an eight-node element as shown in Figure 4.2-1 for modeling
three-dimensional structures and solids. It is based upon an isoparametric
formulation that includes nine optional incompatible bending modes. Solid
elements are used in evaluation of principal stress states in joint regions or
complex geometries (CSI, 2021).

h) The NlLink Element

A NlLink element (CSI, 2021) is an element with structural nonlinearities. A

NlLink element may be either a one-joint grounded spring or a two-joint link and
is assumed to be composed of six separate springs, one for each deformational
degrees of freedom including axial, shear, torsion, and pure bending. Nonlinear
behavior is exhibited during nonlinear time-history analyses or nonlinear static
analyses.

4.2.1.2 Types of Boundary Elements

Selecting the proper boundary conditions has an important role in structural analysis.
Effective modeling of support conditions at bearings and expansion joints requires a

Chapter 4 – Structural Modeling and Analysis 4-5

© 2022 California Department of Transportation. ALL RIGHTS reserved.
 Bridge Design Practice 4 • October 2022

careful consideration of continuity of each translational and rotational component of

displacement. For a static analysis, it is common to use a simpler assumption for supports
(i.e., fixed, pinned, roller) without considering the soil/foundation system stiffness.
However, for dynamic analysis, representing the soil/foundation stiffness is essential. In
most cases choosing a [6×6] stiffness matrix is adequate.

For specific projects, the nonlinear modeling of the system can be achieved by using
nonlinear spring/damper. Some finite element programs, such as ADINA (2021), have
more capabilities for modeling the boundary conditions than others.

4.2.1.3 Types of Materials

Different types of materials are used for bridge structure members such as concrete,
steel, prestressing steel, etc. For concrete structures, see Article 5.4 and for steel
structures see Article 6.4 of the AASHTO LRFD Bridge Design Specifications (2017) with
California Amendments (2019a) (AASHTO-CA BDS-8).

The material properties used for an elastic analysis usually are: modulus of elasticity,
shear modulus, Poisson’s ratio, coefficient of thermal expansion, mass density, and
weight density.

The material properties used for an inelastic analysis usually are: modulus of elasticity,
shear modulus, yield strength and strain, ultimate tensile strength and strain,
hardening/softening parameters, Poisson’s ratio, coefficient of thermal expansion, mass
density, and weight density.

One should pay attention to the units used for material properties.

4.2.1.4 Types of Loads

There are two types of loads in a bridge design:

Permanent Loads: Loads and forces that are assumed to be either constant upon
completion of construction or varying only over a long time interval (Article 3.2). Such
loads include the self-weight of structure elements, wearing surface, curbs, parapets and
railings, utilities, locked-in forces, secondary forces from post-tensioning, force effect due
to shrinkage and due to creep, and pressure from retained earth (Article 3.3.2).

Transient Loads: Loads and forces that can vary over a short time interval relative to the
lifetime of the structure (Article 3.2). Such loads include gravity loads due to vehicular and
pedestrian traffic, lateral loads due to wind and water, ice flows, force effects due to
temperature gradient and uniform temperature, and force effects due to settlement and
earthquakes (Article 3.3.2).

Loads are discussed in Chapter 3 in detail.

4-6 Chapter 4 – Structural Modeling and Analysis

© 2022 California Department of Transportation. ALL RIGHTS reserved.
 Bridge Design Practice 4 • October 2022

4.2.1.5 Modeling Discretization

Formulation of a mathematical model using discrete mathematical elements and their

connections and interactions to capture the prototype behavior is called discretization.
For this purpose:

a) Joints/Nodes are used to discretize elements and primary locations in the

structure at which displacements are of interest.
b) Elements are connected to each other at joints.
c) Masses, inertia, and loads are applied to elements and then transferred to
joints.

Figure 4.2-2 shows a typical model discretization for a bridge bent.

Figure 4.2-2 Model Discretization for Monolithic Connection

4.2.2 Structural Modeling Guidelines

4.2.2.1 Lumped-Parameter Models (LPMs)

• Mass, stiffness, and damping of structure components are usually combined

and lumped at discrete locations. It requires significant experience to formulate
equivalent force-deformation with only a few elements to represent structure
response.
• For a cast-in-place prestressed (CIP/PS) concrete box girder superstructure, a
beam element located at the center of gravity of the box girder can be used.
For non-box girder structures, a detailed model will be needed to evaluate the
responses of each separate girder.

Chapter 4 – Structural Modeling and Analysis 4-7

© 2022 California Department of Transportation. ALL RIGHTS reserved.
 Bridge Design Practice 4 • October 2022

4.2.2.2 Structural Component Models (SCMs) - Common Caltrans Practice

• It is based on idealized structural subsystems/elements to resemble geometry

of the structure. Structure response is given as an element force-deformations
relationship.
• Gross moment of inertia is typically used for non-seismic analysis of concrete
columns.
• Effective moment of inertia can be used under loads, such as prestressing and
thermal effects. Effective moment of inertia falls in the range between gross
and cracked moment of inertia. To calculate effective moment of inertia, see
Article 5.6.3.5.2.
• Cracked moment of inertia is obtained using section moment - curvature
analysis (e.g. CSiBridge (CSI, 2021) Section Designer), which is the moment
of inertia corresponding to the first yield curvature. For seismic analysis, refer
to Caltrans Seismic Design Criteria (SDC) Article 3.4 “Effective Section
Properties” (Caltrans, 2019c) and Caltrans Seismic Design Specifications for
Steel Bridges (SDSSB) Article 3.2.5 “Effective Section Properties” (Caltrans,
2016).

4.2.2.3 Finite Element Models

• A bridge structure is discretized with finite-size elements. Element

characteristics are derived from the constituent structural materials.

Figure 4.2-3 shows the levels of modeling for seismic analysis of bridge structures.

Figure 4.2-3 Levels of Modeling for Seismic Analysis of Bridge

(Priestley, et al 1996)

4-8 Chapter 4 – Structural Modeling and Analysis

© 2022 California Department of Transportation. ALL RIGHTS reserved.
 Bridge Design Practice 4 • October 2022

The importance of the structure, experience of the designer and the level of needed
accuracy affects type of model, location of joints and elements within the selected model,
and number of elements/joints to describe geometry of the structure. For example, a
horizontally curved structure would be defined better by shell elements in comparison
with straight elements. The other factors to be considered are:

a) Structural boundaries - e.g., corners

b) Changes in material properties
c) Changes in element sectional properties
d) Support locations
e) Points of application of concentrated loads - Frame elements can have in-span
loads

4.2.3 Material Modeling Guidelines

Material models should be selected based on a material’s deformation under external
loads. A material is called elastic, when it returns to its original shape upon release of
applied loads. Otherwise it is called an inelastic material.

For an elastic body, the current state of stress depends only on the current state of
deformation while, in an inelastic body, residual deformation and stresses remain in the
body even when all external loads are removed.

The elastic material may show linear or nonlinear behavior. For linear elastic materials,
stresses are linearly proportional to strains (σ = Eє) as described by Hooke’s Law. The
Hooke’s Law is applicable for both homogeneous and isotropic materials.

• Homogeneous means that the material properties are independent of the

coordinates.
• Isotropic means that the material properties are independent of the rotation of
the axes at any point in the body or structure. Only two elastic constants
(modulus of elasticity E and Poisson’s ratio ν) are needed for linear elastic
materials.

For a simple linear spring, the constitutive law is given as: FS = ky where y is the relative
extension or compression of the spring, while FS and k represent the force in the spring
and the spring stiffness, respectively. Stiffness is the property of an element which is
defined as force per unit displacement.

For a nonlinear analysis, nonlinear stress-strain relationships of structural materials

should be incorporated.

• For unconfined concrete a general stress-strain relationship proposed by

Hognestad is widely used. For confined concrete, generally Mander’s model is
used (Akkari and Duan, 2014).

Chapter 4 – Structural Modeling and Analysis 4-9

© 2022 California Department of Transportation. ALL RIGHTS reserved.
 Bridge Design Practice 4 • October 2022

• For structural steel and reinforcing steel, the stress-strain curve usually
includes four segments: elastic, perfectly plastic, strain-hardening, and
softening region (Caltrans, 2016).
• For prestressing steel, an idealized nonlinear stress-strain model may be used.

4.2.4 Types of Bridge Models

4.2.4.1 Global Bridge Models

A global bridge model includes the entire bridge with all frames and connecting structures.
It can capture effects due to irregular geometry such as curves in plane and elevation,
effects of highly skewed supports, contribution of ramp structures, frames interaction,
expansion joints, etc. It is primarily used in seismic design to verify design parameters for
the individual frame. The global model may be in question because of spatially varying
ground motions for large, multi-span, and multi-frame bridges under seismic loading. In
this case, a detailed discretization and modeling force-deformation of an individual
element is needed.

4.2.4.2 Individual Frame and Continuous Global Models

The individual frame (i.e., discrete tension) and continuous global (i.e., global
compression) models are used to capture nonlinear responses for bridges with expansion
joints to model the non-linearity of the hinges with cable restrainers. Maximum response
quantities from the two models are used for seismic design.

a) Individual Frame Model

An individual frame model is used to capture out-of-phase frame movement.
The individual model allows relative longitudinal movement between adjacent
frames by releasing the longitudinal force in the rigid hinge elements and
abutment joints and activating the cable restrainer elements. The cable
restrainer unit is modeled as an individual truss element with equivalent spring
stiffness for longitudinal movement connecting across expansion joints.

b) Continuous Global Model

A continuous global model is used to capture in-phase frame movement. The
continuous global model locks the longitudinal force and allows only moment
about the vertical and horizontal centerline at an expansion joint to be released.
All expansion joints are rigidly connected in the longitudinal direction to capture
effects of joint closing-abutment mobilized.

4.2.4.3 Frame Models

A frame model is a portion of structure between the expansion joints. It is powerful to

assess the true dynamic response of the bridge since dynamic response of stand-alone

4-10 Chapter 4 – Structural Modeling and Analysis

© 2022 California Department of Transportation. ALL RIGHTS reserved.
 Bridge Design Practice 4 • October 2022

bridge frames can be assessed with reasonable accuracy as an upper bound response
to the whole structure system. Seismic characteristics of individual frame responses are
controlled by mass of superstructure and stiffness of individual frames. Transverse stand-
alone frame models shall assume lumped mass at the columns. Hinge spans shall be
modeled as rigid elements with half of their mass lumped at the adjacent column (SDC
Figure 4.3.2.1-1, Caltrans, 2019c). Effects from the adjacent frames can be obtained by
including boundary frames in the model.

4.2.4.4 Bent Models

A transverse model of the bent cap and columns is needed to obtain maximum moments
and shears along the bent cap. The dimension of the bent cap should be considered
along the skew.

An individual bent model should include foundation flexibility effects and can be combined
in a frame model simply by geometric constraints. Different ground motions can be input
for individual bents. The high in-plane stiffness of bridge superstructures allows rigid body
movement assumption which simplifies the combination of individual bent models.

4.2.5 Slab-Beam Bridges

4.2.5.1 Superstructures

For modeling girder system bridges, either a spine model or a grillage model should be
used.

Figure 4.2-4 Superstructure Models (Priestley, et al 1996)

Chapter 4 – Structural Modeling and Analysis 4-11

© 2022 California Department of Transportation. ALL RIGHTS reserved.
 Bridge Design Practice 4 • October 2022

a) Spine Model
Spine models with beam elements are usually used for ordinary bridges. The
beam element considers six DOF at both ends of the element and is modeled
at their neutral axis.
 The effective stiffness of the element may vary depending on the
structure type.
 Use SDC V2.0 to define effective flexural stiffness EIeff for reinforced
concrete box girders and pre-stressed box girders as follows:
- For reinforced concrete (RC) box girder, (0.5~0.75) EIg
- For prestressed concrete (PS) box girder, 1.0 EIg and for tension
it considers Ig, where Ig is the gross section moment of inertia.
 The torsional stiffness for concrete superstructures can be taken as: GJ
for un-cracked section and 0.5 GJ for cracked section.
 Use SDSSB (Caltrans, 2016) to define effective flexural stiffness, EIeff,
for steel members.
 A spine model can’t capture the superstructure carrying a wide roadway
or high-skewed bridges. In these cases, use a grillage model.
b) Grillage Models/3-D Finite Element Model
Grillage models are used for modeling steel composite deck superstructures
and complicated structures where superstructures can’t be considered rigid
such as very long and narrow bridges for example, interchange connectors.

4.2.5.2 Bents

If the bridge superstructure can be assumed to move as a rigid body under seismic loads,
the analysis can be simplified to modeling bents only. Frame elements, effective bending
stiffness, cap with large torsional and transverse bending stiffness to capture
superstructure, and effective stiffness for outriggers should be considered. Figure 4.2-5
shows single column bent models.

4-12 Chapter 4 – Structural Modeling and Analysis

© 2022 California Department of Transportation. ALL RIGHTS reserved.
 Bridge Design Practice 4 • October 2022

Figure 4.2-5 Single-Column Bent Models (Priestley et al, 1996)

4.2.5.3 Superstructure Bents Connection

In modeling the superstructure to bent connections, two different connections as shown

in Figures 4.2-2 and 4.2-6 may be considered:

a) Monolithic connections for cast-in-place box girders and integral bent cap for
precast girders.
b) Bearing supported connections for precast concrete girders or steel
superstructures on drop cap. Different types of bearings are: PTFE, stainless
steel sliders, rocker bearings and elastomeric bearings. With the bearing-
supported connections, one may use the isolated bearings such as special
seismic bearings and energy-dissipating devices to reduce resonant buildup of
displacement.

In monolithic connections all the degrees of freedom are restrained (three degrees of
translations and three degrees for rotation); however, in bearing supported connections,
only three degrees of translations are restrained but the rotational degrees of freedom
are free.

In the bearing supported structures, the superstructure is not subjected to seismic

moment transferred through the column. However, the design is more sensitive to seismic
displacement than with the monolithic connection.

The energy dissipation devices in the seismic-isolated bridges reduce the seismic
displacement (backwards) significantly in comparison with bearing-supported structures.
The designer should pay attention to the possibility of increased acceleration when using
the bearing-supported connections with or without energy-dissipation devices in soft soils.

Chapter 4 – Structural Modeling and Analysis 4-13

© 2022 California Department of Transportation. ALL RIGHTS reserved.
 Bridge Design Practice 4 • October 2022

Figure 4.2-6 Superstructure-Bent Connection

4.2.5.4 Hinges

Hinges separate frames in long structures to allow for movements due to thermal, initial
pre-stress shortening and creep without introducing large stresses and strains in
members.

A typical hinge should be modeled as 6 DOF, i.e., free to rotate in the longitudinal direction
and pin in the transverse direction to represent shear (Figure 4.2-7).

Linear Elastic Modal Analysis with two different structural models, Tension and
Compression, is used to take care of this analysis issue.

4-14 Chapter 4 – Structural Modeling and Analysis

© 2022 California Department of Transportation. ALL RIGHTS reserved.
 Bridge Design Practice 4 • October 2022

Figure 4.2-7 Span Hinge Force Definitions (Priestley et al, 1996)

4.2.5.5 Substructures

Figures 4.2-8 and 4.2-9 show a multi-column bent model and a foundations spring model
at a bent, respectively. Figure 4.2-10 shows a multi bridge frame model.

a) Column-Pier Sections
 Prismatic or Non-Prismatic
 Shapes Circular, Rectangular, or Hollow-Section

Figure 4.2-8 Multi-Column Bent Model (Priestley et al, 1996)

Chapter 4 – Structural Modeling and Analysis 4-15

© 2022 California Department of Transportation. ALL RIGHTS reserved.
 Bridge Design Practice 4 • October 2022

b) Bent-Foundation Connection
 Pin base: Generally used for multi-column bents.
 Fixed Base: For single column base.

Figure 4.2-9 Foundation Spring Definition at a Bent

4-16 Chapter 4 – Structural Modeling and Analysis

© 2022 California Department of Transportation. ALL RIGHTS reserved.
 Bridge Design Practice 4 • October 2022

Figure 4.2-10 Multi Bridge Frame (Priestley et al, 1996)

4.2.6 Abutments
Abutments can be modeled as pin, roller, or fixed boundary condition. For modeling the
soil-structure interaction, springs can be used. Figure 4.2-11 shows end restraints with
springs to model soil-structure interaction for seat and rigid abutments. Abutment

Chapter 4 – Structural Modeling and Analysis 4-17

© 2022 California Department of Transportation. ALL RIGHTS reserved.
 Bridge Design Practice 4 • October 2022

stiffness, capacities, and damping affect seismic response. SDC V2.0, Section 6.3
discusses the longitudinal and transverse abutment responses in an earthquake. For
modeling gap, backwall and piles effective stiffness is used to simulate their nonlinear
behavior. An iterative procedure should be used to find a convergence between stiffness
and displacement.

Figure 4.2-11 Foundation Spring Definition

4.2.7 Foundation

4.2.7.1 Group Piles

Supports can be modeled using:

• Springs - 6 × 6 stiffness matrix - defined in global/joint local coordinate system.

• Restraints - known displacement, rotation - defined in global DOF.
• Complete pile system with soil springs along with the bridge.

4.2.7.2 Pile shaft

When modeling the pile shaft for non-seismic loading, an equivalent fixity model can be

4-18 Chapter 4 – Structural Modeling and Analysis

© 2022 California Department of Transportation. ALL RIGHTS reserved.
 Bridge Design Practice 4 • October 2022

used (Figure 4.2-12c). For seismic loading, a soil-spring model (Figure 4.2-12b) should
be considered to capture the soil-structure interaction. Programs such as CSiBridge or
ADINA can be used.

Figure 4.2-12 CIDH Pile Shaft Models (Priestley et al, 1996)

4.2.7.3 Spread Footing

Spread footings are usually built on stiff and competent soils, fixed boundary conditions
are assumed for the translational springs, and rotation is considered only when uplift and
rocking of the entire footing are expected.

Chapter 4 – Structural Modeling and Analysis 4-19

© 2022 California Department of Transportation. ALL RIGHTS reserved.
 Bridge Design Practice 4 • October 2022

4.2.8 Examples

4.2.8.1 CTBridge

CTBridge (Caltrans, 2019b) is a finite element analysis and design software using a 3-D
spine model for the bridge structure subjected to non-seismic loads and is comprised of
beam elements positioned in 3 dimensional space. The beam elements are used to
model the bridge superstructure in the longitudinal direction and the bridge bents/piers in
the transverse direction. Its analysis is based on the linear elastic small deformation
theory and its design is based on AASHTO-CA BDS-8 for general application. However,
its design is applicable to concrete bridges and components only.

CTBridge allows user manipulation of various settings such as:

• Number of elements
• Live load step sizes
• Prestress discretization
• P-Jack design limits
• Skewed supports
• Horizontal and vertical curves
• Multi-column bents

For non-skewed bridges, the abutments can be considered pinned or roller. For skewed
bridges, springs should be used at the abutments. The stiffness of the springs should be
based on the stiffness of the bearing pads. If bearing stiffness is not available, slider can
be used instead of pin or roller. For bridges with curved alignments and skewed supports
or straight bridges with skews in excess of 60 degrees, advanced programs having a full
3-D analysis model, such as a grillage or shell model may be required to more accurately
capture the true load and structural responses.

Note that in order to get the result at each 0.1 span length, you should define the offset
from the beginning and the end span, i.e., from the center line of the abutment to the face
of the abutment.

The following structure shown in Figures 4.2-13a to 4.2-13c is used as an example for
CTBridge.

4-20 Chapter 4 – Structural Modeling and Analysis

© 2022 California Department of Transportation. ALL RIGHTS reserved.
 Bridge Design Practice 4 • October 2022

Figure 4.2-13a Elevation View of Example Bridge

Figure 4.2-13b Typical Section View of Example Bridge

Chapter 4 – Structural Modeling and Analysis 4-21

© 2022 California Department of Transportation. ALL RIGHTS reserved.
 Bridge Design Practice 4 • October 2022

Figure 4.2-13c Plan View of Example Bridge

Figure 4.2-14 shows CTBridge model for the example bridge.

Figure 4.2-14 Example Bridge - CTBridge Model

4-22 Chapter 4 – Structural Modeling and Analysis

© 2022 California Department of Transportation. ALL RIGHTS reserved.
 Bridge Design Practice 4 • October 2022

Figure 4.2-15 shows the figure from the CTBridge manual indicating the sign convention
used by the program.

Figure 4.2-15 Sign Convention at CTBridge

Figure 4.2-16 shows two spine models.

Figure 4.2-16 3-D Frame in CTBridge

Chapter 4 – Structural Modeling and Analysis 4-23

© 2022 California Department of Transportation. ALL RIGHTS reserved.
 Bridge Design Practice 4 • October 2022

4.2.8.2 CSiBridge

CSiBridge is one of the most powerful versions of the well-known Finite Element Analysis
Program SAP series of Structural Analysis Programs, which offers the following features:

• Static and Dynamic Analysis

• Linear and Nonlinear Analysis
• Dynamic Seismic Analysis and Static Pushover Analysis
• Vehicle Live-Load Analysis for Bridges, Moving Loads with 3-D Influence
Surface, Moving Loads with Multi-Step Analysis, Lane Width Effects
• P-Delta Analysis
• Cable Analysis
• Eigen and Ritz Analyses
• Fast Nonlinear Analysis for Dampers
• Energy Method for Drift Control
• Segmental Construction Analysis

The following are the general steps to be defined for analyzing a structure using
CSiBridge:

• Geometry (input nodes coordinates, define members and connections)

• Boundary conditions/joint restraints (fixed, free, roller, pin or partially restrained
with a specified spring constant)
• Material property (elastic modulus, Poisson’s ratio, shear modulus, damping
data, thermal properties, and time-dependent properties such as creep and
shrinkage)
• Loads and load cases
• Stress-strain relationship
• Perform analysis of the model based on analysis cases

CSiBridge templates can be used for generating Bridge Models, Automated Bridge Live
Load Analysis and Design, Bridge Base Isolation, Bridge Construction Sequence
Analysis, Large Deformation Cable Supported Bridge Analysis, and Pushover Analysis.

The user can either model the structure as a Spine Model (Frame) or a 3-D Finite Element
Model.

In this section, we create a CSiBridge model for the Example Bridge using the Bridge
Wizard (BrIM-Bridge Information Modeler). The Bridge Modeler has 13 modeling step
processes of which the 9 major steps needed to define the model are described below:

4-24 Chapter 4 – Structural Modeling and Analysis

© 2022 California Department of Transportation. ALL RIGHTS reserved.
 Bridge Design Practice 4 • October 2022

a) Layout line

The first step in creating a bridge object is to define highway layout lines using
horizontal and vertical curves. Layout lines are used as reference lines for
defining the layout of bridge objects and lanes in terms of stations, bearings
and grades considering super elevations and skews.

b) Deck Section

Various parametric bridge sections (Box Girders & Steel Composites) are
available for use in defining a bridge. See Figure 4.2-17.

User can specify different Cross Sections along Bridge length.

Figure 4.2-17 Various Bridge Sections

c) Abutment Definition

Abutment definitions specify the support conditions at the ends of the bridge.
The user defined support condition allows each six DOF at the abutment to be
specified as fixed, free, or partially restrained with a specified spring constant.

Chapter 4 – Structural Modeling and Analysis 4-25

© 2022 California Department of Transportation. ALL RIGHTS reserved.
 Bridge Design Practice 4 • October 2022

Those six Degrees of Freedom are:

U1- Translation Parallel to Abutment
U2- Translation Normal to Abutment
U3- Translation Vertical
R1- Rotation about Abutment
R2- Rotation about Line Normal to Abutment
R3- Rotation about Vertical

For Academy Bridge consider U2, R1 and R3 DOF directions to have a “Free”
release type and other DOF fixed.

d) Bent Definition

This part specifies the geometry and section properties of bent cap beam and
bent cap columns (single or multiple columns) and base support condition of
the bent columns.

The base support condition for a bent column can be fixed, pinned or user
defined as a specified link/support property which allows six degrees of
freedom.

For the Example Bridge enter the column base supports as pinned. All units
should be kept consistent (kip-ft for this example).

The locations of columns are defined as distance from left end of the cap beam
to the centerline of the column and the column height is the distance from the
mid-cap beam to the bottom of the column.

For defining columns use Bent definition under bridge wizard, then go to
define/show bents and go to Modify/show column data. The base column
supports at top and bottom will be defined here.

e) Diaphragm Definition

Diaphragm definitions specify properties of vertical diaphragms that span

transverse across the bridge. Diaphragms are only applied to area objects and
solid object models and not to spine models. Steel diaphragm properties are
only applicable to steel bridge sections.

f) Hinge Definition

Hinge definitions specify properties of hinges (expansion joints) and

restrainers. After a hinge is defined, it can be assigned to one or more spans
in the bridge object.

A hinge property can be a specified link/support property, or it can be user-

defined spring. The restrainer property can be also a link/support or user
defined restrainer. The user-restrainer is specified by a length, area, and

4-26 Chapter 4 – Structural Modeling and Analysis

© 2022 California Department of Transportation. ALL RIGHTS reserved.
 Bridge Design Practice 4 • October 2022

modulus of elasticity.

g) Parametric Variation Definition

Any parameter used in the parametric definition of the deck section can be
specified to vary such as bridge depth, thickness of the girders and slabs along
the length of the bridge. The variation may be linear, parabolic, or circular.

h) Bridge Object Definition

The main part of the Bridge Modeler is the Bridge Object Definition which
includes defining bridge span, deck section properties assigned to each span,
abutment properties and skews, bent properties and skews, hinge locations are
assigned, super elevations are assigned, and pre-stress tendons are defined.

The user has two tendon modeling options for pre-stress data:

 Model as loads
 Model as elements

Since we calculate the prestress jacking force, Pjack, from CTBridge, use option
a) (Layout line) to input the Tendon Load force. The user can input the Tendon
loss parameters which have two parts:

1) Friction and Anchorage losses (Curvature coefficient, Wobble coefficient

and anchorage setup).
2) Other loss parameters (Elastic shortening stress, Creep stress,
Shrinkage stress and Steel relaxation stress).

When you input values for Friction and Anchorage losses, make sure the
values match your CTBridge which should be based on AASHTO-CA BDS-8
Table 5.9.3.2.2b-1 and there is no need to input other loss parameters. If the
user decides to model tendon as elements, the values for other loss parameters
shall be input; otherwise, leave the default values.

Note:

• If you model the bridge as a Spine Model, only define one single tendon
with total Pjack load. If you model the bridge with shell element, then you
need to specify tendon in each girder and input the Pjack force for each
girder which should be calculated as Total Pjack divided by the number
of the girders.
• Anytime a bridge object definition is modified, the link model shall be
updated for the changes to appear in /CSiBridge model.

Chapter 4 – Structural Modeling and Analysis 4-27

© 2022 California Department of Transportation. ALL RIGHTS reserved.
 Bridge Design Practice 4 • October 2022

i) Update Linked Model

The update linked model command creates the CSiBridge object-based model
from the bridge object definition. Figures 4.2-18 and 4.2-19 show an area object
model and a solid object model, respectively. Note that an existing object will
be deleted after updating the linked model. There are three options in the
Update Linked Model including:

• Update a Spine Model using Frame Objects

• Update as Area Object Model
• Update as Solid Object Model

Figure 4.2-18 Area Object Model

Figure 4.2-19 Solid Object Model

4-28 Chapter 4 – Structural Modeling and Analysis

© 2022 California Department of Transportation. ALL RIGHTS reserved.
 Bridge Design Practice 4 • October 2022

Other analysis steps include:

 Parametric Bridge Modeling

- Layered Shell Element
- Lane Definition Using Highway Layout or Frame Objects
- Automatic Application of Lane Loads to Bridge
- Predefined Vehicle and Train Loads
 Bridge Results & Output
- Influence Lines and Surfaces
- Forces and Stresses Along and Across Bridge
- Displacement Plots
- Graphical and Tabulated Outputs

CSiBridge also has an Advanced Analysis Option that is not discussed in this section
including:

 Segmental Construction
 Effects of Creep, Shrinkage Relaxation
 Pushover Analysis using Fiber Models
 Bridge Base Isolation and Dampers
 Explicitly Model Contact Across Gaps
 Nonlinear Large Displacement Cable Analysis
 Line and Surface Multi-Linear Springs (P-y curves)
 High Frequency Blast Dynamics using Wilson FNA
 Nonlinear Dynamic Analysis & Buckling Analysis
 Multi-Support Seismic Excitation
 Animated Views of Moving Loads

The program also has the feature of automated line constraints that enforce the
displacement compatibility along the common edges of meshes as needed.

4.3 STRUCTURAL ANALYSIS

Structural Analysis provides the numerical mathematical process to extract structure
responses under service and seismic loads in terms of load effects or structural demands
such as member forces and deformations.

Chapter 4 – Structural Modeling and Analysis 4-29

© 2022 California Department of Transportation. ALL RIGHTS reserved.
 Bridge Design Practice 4 • October 2022

4.3.1 General
For any type of structural analysis, the following principles should be considered.

4.3.1.1 Equilibrium

a) Static Equilibrium

In a supported structure system when the external forces are in balance with
the internal forces, or stresses, which exactly counteract the loads (Newton’s
Second Law), the structure is said to be in equilibrium.

Since there is
 notranslatory motion, the vector sum of the external forces must
be zero ( ∑ F = 0 ). Since there is no rotation, the sum of the moments of the
 
external forces about any point must be zero ( ∑ M = 0 ).

b) Dynamic Equilibrium

When dynamic effects need to be included, whether for calculating the dynamic
response to a time-varying load or for analyzing the propagation of waves in a
structure, the proper inertia terms shall be considered for analyzing the
dynamic equilibrium:

∑ F = mu
4.3.1.2 Constitutive Laws

The constitutive laws define the relationship between the stress and strain in the material
of which a structure member is made.

4.3.1.3 Compatibility

Compatibility conditions are referred to continuity or consistency conditions on the strains

and the deflections. As a structure deforms under a load, the following principles apply:

a) Two originally separate points do not merge into a single point.

b) Perimeter of a void does not overlap as it deforms.
c) Elements connected together remain connected as the structure deforms.

4.3.2 Analysis Methods

Different types of analysis are discussed in this section.

4-30 Chapter 4 – Structural Modeling and Analysis

© 2022 California Department of Transportation. ALL RIGHTS reserved.
 Bridge Design Practice 4 • October 2022

4.3.2.1 Small Deflection Theory

If the deformation of the structure doesn’t result in a significant change in force effects
due to an increase in the eccentricity of compressive or tensile forces, then such
secondary force effects may be ignored. Small deflection theory is usually adequate for
the analysis of beam-type bridges. Suspension bridges, very flexible cable-stayed
bridges, and some arches rather than tied arches and frames in which flexural moments
are increased by deflection are generally sensitive to deflections. In many cases the
degree of sensitivity can be evaluated by a single-step approximate method, such as the
moment magnification factor method (Article 4.5.3.2.2).

4.3.2.2 Large Deflection Theory

If the deformation of the structure results in a significant change in force effects, the
effects of deformation shall be considered in the equations of equilibrium. The effect of
deformation and out-of-straightness of components shall be included in stability analysis
and large deflection analyses. For slender concrete compressive components, time-
dependent and stress-dependent material characteristics that cause significant changes
in structural geometry shall be considered in the analysis.

Because large deflection analysis is inherently nonlinear, the displacements are not
proportional to applied loads, and superposition cannot be used. Therefore, the order of
load application is very important and should be applied in the order experienced by the
structure, i.e., dead load stages followed by live load stages, etc. If the structure
undergoes nonlinear deformation, the loads should be applied incrementally with
consideration for the changes in stiffness after each increment.

4.3.2.3 Linear Analysis

In the linear relation of stress-strain of a material, Hooke’s law is valid for a small stress-
strain range. For linear elastic analysis, sets of loads acting simultaneously can be
evaluated by superimposing (adding) the forces or displacements at the particular point.

4.3.2.4 Nonlinear Analysis

The objective of the nonlinear analysis is to estimate the maximum load that a structure
can support prior to structural instability or collapse. The maximum load which a structure
can carry safely may be calculated by simply performing an incremental analysis using a
non-linear formulation. In a collapse analysis, the equation of equilibrium is for each load
or time step.

Design based on the assumption of linear stress-strain relation will not always be
conservative due to material or physical nonlinearity. Very flexible bridges, e.g.
suspension and cable-stayed bridges, should be analyzed using nonlinear elastic
methods (Article C4.5.1).

Chapter 4 – Structural Modeling and Analysis 4-31

© 2022 California Department of Transportation. ALL RIGHTS reserved.
 Bridge Design Practice 4 • October 2022

P-Δ effect is an example of physical (geometrical) nonlinearity, where the principle of

superposition doesn’t apply since the beam-column element undergoes large changes in
geometry when loaded.

4.3.2.5 Elastic Analysis

Service and fatigue limit states should be analyzed as fully elastic, as should strength
limit states, except in the case of certain continuous girders where inelastic analysis is
permitted, inelastic redistribution of negative bending moment and stability investigation
(Article C4.5.1).

When modeling the elastic behavior of materials, the stiffness properties of concrete and
composite members shall be based upon cracked and/or uncracked sections consistent
with the anticipated behavior (Article 4.5.2.2). A limited number of analytical studies have
been performed by Caltrans to determine effects of using the gross and cracked moment
of inertia. The specific studies yielded the following findings on prestressed concrete
girders on concrete columns:

1) Using Ig or Icr in the concrete columns does not significantly reduce or increase
the superstructure moment and shear demands for external vertical loads, but
will significantly affect the superstructure moment and shear demands from
thermal and other lateral loads (Article C4.5.2.2).
2) Using Icr in the columns can increase the superstructure deflection and camber
calculations.

Usually an elastic analysis is sufficient for force-based design.

4.3.2.6 Inelastic Analysis

Inelastic analysis should be used for the displacement-based design (Akkari and Duan,
2014).

The extreme event limit states may require collapse investigation based entirely on
inelastic modeling. Where inelastic analysis is used, a preferred design failure mechanism
and its attendant hinge locations shall be determined (Article 4.5.2.3).

4.3.2.7 Static Analysis

Static analysis is mainly used for bridges under dead load, vehicular load, wind load and
thermal effects. The influence of plan geometry has an important role in static analysis
(Article 4.6.1). One should pay attention to the plan aspect ratio and structures curved in
the plan for static analysis.

• Plan Aspect Ratio

If the span length of a superstructure with a torsionally stiff closed crossed
section exceeds 2.5 times its width, the superstructure may be idealized as a

4-32 Chapter 4 – Structural Modeling and Analysis

© 2022 California Department of Transportation. ALL RIGHTS reserved.
 Bridge Design Practice 4 • October 2022

single-spine beam. Simultaneous torsion, moment, shear and reaction forces

and the attendant stresses are to be superimposed as appropriate. In all
equivalent beam idealizations, the eccentricity of loads should be taken with
respect to the centerline of the equivalent beam.
• Structure curved in plan
Horizontally cast-in-place box girders may be designed as a single spine beam
with straight segments, for central angles up to 34° within one span, unless
concerns about other force effects dictate otherwise. For I-girders, since
equilibrium is developed by the transfer of load between the girders, the
analysis should consider the integrated behavior of all structure components.

Small deflection theory is adequate for the analysis of most horizontally curved-girder
bridges. However, curved I-girders are prone to deflect laterally if not sufficiently braced
during erection. This behavior may not be well recognized by the small deflection theory.

4.3.2.8 Equivalent Static Analysis (ESA)

It is used to estimate seismic demands for ordinary bridge structures as specified in

Caltrans SDC (Caltrans, 2019c). A bridge is usually modeled as Single-Degree-of-
Freedom (SDOF) and seismic load is applied as an equivalent static horizontal force. It
is suitable for individual frames with well-balanced spans and stiffness. Caltrans SDC
(Caltrans, 2019c) recommends stand-alone “local” analysis in transverse & longitudinal
directions for demand assessment. Figure 4.3-1 shows a stand-alone model with lumped
masses at columns, rigid body rotation, and half span mass at adjacent columns.

Transverse Stand-Alone Model

Longitudinal Stand-Alone Model

Figure 4.3-1 Stand Alone Model

Chapter 4 – Structural Modeling and Analysis 4-33

© 2022 California Department of Transportation. ALL RIGHTS reserved.
 Bridge Design Practice 4 • October 2022

Types of ESA such as Lollipop Method, Uniform Load Method and Generalized
Coordinate Method can be used.

4.3.2.9 Nonlinear Static Analysis (Pushover Analysis)

Nonlinear Incremental Static Analysis or Pushover Analysis is used to determine the

displacement capacity of a bridge structure.

Horizontal loads are incrementally increased until a structure reaches a collapse condition
or a collapse mechanism. Change in structure stiffness is modeled as member stiffness
due to cracking, plastic hinges, and yielding of soil spring at each step (event).

Analysis Programs are available such as: CSiBridge or ADINA.

Figures 4.3-2 and 4.3-3 show the typical force-displacement curve and moment-curvature
curve for a concrete column.

Figure 4.3-2 Typical Force Displacement Curve for a Concrete Column

a) Pushover Analysis – Requirements

 Linear Elastic Structural Model
 Initial or Gravity loads
 Characterization of all Nonlinear actions - multi-linear force-deformation
relationships (e.g. plastic hinge moment-curvature relationship)
 Limits on strain based on design performance level to compute moment
curvature relationship of nonlinear hinge elements.
 Section Analysis─> Strain─> Curvature
 Double Integration of curvature─> Displacements
 Track design performance level strain limits in structural response

4-34 Chapter 4 – Structural Modeling and Analysis

© 2022 California Department of Transportation. ALL RIGHTS reserved.
 Bridge Design Practice 4 • October 2022

Figure 4.3-3 Typical Moment-Curvature Curve for a Concrete Column

4.3.2.10 General Dynamic Equilibrium Equation

The dynamic equation of motion for a typical SDOF is:

Finput = FI + FD + Fs

where:

FI = mass × acceleration = mü
FD = damping const × velocity = cu̇
FS = stiffness × deformation = ku
Weight
m = mass = ρsV =
g
ρs = material mass density
g = gravitational acceleration (32.2 ft/sec)
V = element volume = A × L
A = element area
L = element length
k = stiffness
c = damping constant = z × ccr
ccr = critical damping = 2 × m × ω
ω = angular frequency
1 EDC
z = damping ratio =
2π ku 2
EDC = Energy dissipated per cycle

Chapter 4 – Structural Modeling and Analysis 4-35

© 2022 California Department of Transportation. ALL RIGHTS reserved.
 Bridge Design Practice 4 • October 2022

u = displacement
u̇ = velocity
ü = acceleration

In addition to earthquakes, wind and moving vehicles can cause dynamic loads on bridge
structures.

Wind load may induce instability and excessive vibration in long-span bridges. The
interaction between the bridge vibration and wind results in two kinds of forces including
motion-dependent and motion-independent. The motion-dependent force causes
aerodynamic instability with emphasis on the vibration of rigid bodies. For short span
bridges the motion-dependent part is insignificant and there is no concern about
aerodynamic instability. The bridge aerodynamic behavior is controlled by two types of
parameters: structural and aerodynamics. The structure parameters are the bridge layout,
boundary condition, member stiffness, natural modes, and frequencies. The aerodynamic
parameters are wind climate and bridge section shape. The aerodynamic equation of
motion is expressed as:

mu + cu + ku= FUmd + Fmi

where:

FUmd = motion-dependent aerodynamic force vector

Fmi = motion-independent wind force vector

For a detailed analytical solution for effect of wind on long span bridges and cable
vibration, see (Cai, et al., 2014).

4.3.2.11 Free Vibration Analysis

Vehicles such as trucks and trains passing bridges at a certain speed will cause dynamic
effects. The dynamic loads for moving vehicles on bridges are counted for by a dynamic
load allowance, IM. See (Duan, et al., 1999).

Major characteristics of the bridge dynamic response under moving load can be
summarized as follows:

Dynamic load allowance, IM, increases as vehicle speed increases, IM decreases as

bridge span increases.

Under the condition of “Very good” road surface roughness (amplitude of highway profile
curve is less than 0.4 in.) the impact factor is well below the design specifications. But the
impact factor increases tremendously with increasing road surface roughness from “good”
to “poor” (the amplitude of the roadway profile is more than 1.6 in.) beyond the dynamic
load allowance specified in AASHTO-CA BDS-8.

4-36 Chapter 4 – Structural Modeling and Analysis

© 2022 California Department of Transportation. ALL RIGHTS reserved.
 Bridge Design Practice 4 • October 2022

Field tests indicate that in the majority of highway bridges, the dynamic component of the
response does not exceed 25% of the static response to vehicles with the exception of
deck joints. For deck joints, a dynamic load allowance of 75% is considered for all limit
states due to hammer effect, and 15% for fatigue and fracture limit states for members
vulnerable to cyclic loading such as anchor studs (see Article C3.6.2.1).

Dynamic effects due to moving vehicles may be attributed to two sources:

 Hammering effect is the dynamic response of the wheel assembly to riding

surface discontinuities, such as deck joints, cracks, potholes and
delaminations.
 Dynamic response of the bridge as a whole to passing vehicles, which may be
due to long undulations in the roadway pavement, such as those caused by
settlement of fill, or to resonant excitation as a result of similar frequencies of
vibration between bridge and vehicle. (Article C3.6.2.1)

The magnitude of dynamic response depends on the bridge span, stiffness and surface
roughness, and vehicle dynamic characteristics such as moving speed and isolation
systems. There have been two types of analysis methods to investigate the dynamic
response of bridges due to moving load:

 Numerical analysis (Sprung mass model).

 Analytical analysis (Moving load model).

The analytical model greatly simplifies vehicle interaction with bridges. Modeling a bridge
as a plate or beam provides good accuracy if the ratio of live load to self-weight of the
superstructure is less than 0.3.

Free vibration analysis assuming a sinusoidal mode shape can be used for the analysis
of the superstructure and calculating the fundamental frequencies of slab-beam bridges
(Zhang, et al., 2014).

For long span bridges or low speed moving loads, there is little amplification which does
not result in much dynamic responses.

The maximum dynamic response happens when the load frequency is near the bridge
fundamental frequency.

The aspect ratio of the bridge deck plays an important role. When the ratio is less than
4.0, the first mode shape is dominant, and when more than 8.0, other mode shapes are
excited.

Free-Vibration Properties are shown in Figure 4.3-4.

Chapter 4 – Structural Modeling and Analysis 4-37

© 2022 California Department of Transportation. ALL RIGHTS reserved.
 Bridge Design Practice 4 • October 2022

Figure 4.3-4 Natural Period

a) Cycle: When a body vibrates from its initial position to its extreme positive
position in one direction, back to extreme negative position, and back to initial
position (i.e., one revolution of angular displacement of 2 π) (radians)
b) Angular Frequency (ω): If a system is disturbed and allowed to vibrate on its
own, without external forces and damping (free Vibration).

A system having n degrees of freedom will have, in general, n distinct natural frequencies
of vibration.

ω= k /m
2π
=
ω = distance/time
T
ω = 2πf

c) Period (T): Is the time taken to complete one cycle of motion. It is equal to the
time required for a vector to make one complete rotation (one round)
d) Frequency (f): The number of cycles per second, f = 1/T (Hz)

4-38 Chapter 4 – Structural Modeling and Analysis

© 2022 California Department of Transportation. ALL RIGHTS reserved.
 Bridge Design Practice 4 • October 2022

4.4 BRIDGE EXAMPLES – 3-D VEHICLE LIVE LOAD ANALYSIS

4.4.1 Background
The United States has a long history of girder bridges being designed “girder-by-girder”.
That is, the girder is designed for some fraction of a live load lane, depending on girder
spacing and structure type. The method is sometimes referred to as “girder line” or “beam
line” analysis and the fraction of live load lanes used for design is sometimes referred to
as a grid or Load Distribution Factor (LDF).

The approximate methods of live-load distribution in the AASHTO-CA BDS-8 use “girder
load distribution factors” (LDFs) to facilitate beam analysis of multiple vehicular live loads
on a three-dimensional bridge structural system. The formal definition of LDF: “a factor
used to multiply the total longitudinal response of the bridge due to a single longitudinal
lane load in order to determine the maximum response of a single girder” (Barker and
Puckett 2013). A more practical definition for the LDF is the ratio Mrefined / Mbeam or Vrefined
/ Vbeam, where the numerator is the enveloped force effect at one location, and the
denominator is the force effect at the same location in a single girder due to the same
load. The LDFs are in terms of lanes, and the LDFs provided in the AASHTO tables are
multiplied by the single lane demand obtained from a beam line analysis to determine the
amount distributed to the supporting components.

• Although each location within a girder can have a different LDF, the
expressions in the tables of AASHTO-CA BDS-8, Articles 4.6.2.2.2 and
4.6.2.2.3 are based on the critical locations for bending and shear, respectively.
Critical locations refer to the maximum absolute positive moments, negative
moments, and maximum absolute shear. For cast-in-place (CIP) concrete
multicell box girders, the AASHTO tables only apply to typical bridges, which
refer to:
- Girder spacing, S: 7′ < S < 13′
- Span length, L: 60′ < L < 240′ (AASHTO-CA BDS-8 Table
4.6.2.2.2b-1)
- The CA Amendments (Caltrans, 2019a) Table 4.6.2.2.2b-2 provides the
LDF for one cell, and two cell boxes based on:
- Span Length, L: 60' < L < 240'
- Structure Depth, d: 35" < d < 110"
• The use of approximate methods on less-typical structures shall be
investigated. When the structures fall outside of the ranges of applicability
listed for the load distribution factor equations, the approximate methods may
no longer be valid. In some cases, the ranges of applicability may be extended
when the distribution factors are limited to conservative values. Some of the
less-typical structures fall under one of the following cases:
- Two or three-girder beam-slab structures;

Chapter 4 – Structural Modeling and Analysis 4-39

© 2022 California Department of Transportation. ALL RIGHTS reserved.
 Bridge Design Practice 4 • October 2022

- Spans greater than 240 ft in length;

- Structures with extra-wide overhangs (greater than one-half of the girder
spacing or 3 ft).
• Three-dimensional (3-D) finite element analysis (FEA) shall be used to
determine the girder LDFs of these less-typical bridges. The following cases
may also require such refined analysis:
- Skews greater than 45°;
- Structures with masonry sound walls;
- Beam-slab structures with beams of different bending stiffness.
• A moving load analysis on a 3-D FEA model provides accurate load distribution.
However, for the routine design of commonly used bridge superstructure
systems, a 3-D FEA requires the familiarity with sophisticated, usually also
expensive, finite element methods.
FEA software may not be economical due to the additional time required to
build and run the 3-D model, and analyze the results, compared to a simple
FEA program, e.g. Caltrans CTBridge.
• In terms of the reliability of an FEA model, 3-D FEA may not be as reliable as
a simple 2-D FEA model due to the much greater number of details in a 3-D
FEA model. Based on Caltrans experience, a combination of LDF formula with
the in-house 2-D FEA design program, CTBridge, provides sufficient, reliable,
and efficient design procedure and output. The latest version of CTBridge
includes the LDF values for a one- or two-cell box-girder bridge.

4.4.2 Moving Load Cases

In many situations, one- or two-cell box girders are used for the widening of existing
bridges. If they are new bridges, it is also possible that they will be widened in the future.
Both cases imply that the traffic loads may be applied anywhere across the bridge width,
i.e., edge to edge, and this shall be taken into account in the design. This also means that
one wheel line of the truck can be on the new/widened bridge, while the other one is on
the existing bridge. As one can imagine, for certain bridge width, the maximum force effect
may be due to, say, 1.5 or 2.5 lanes. For a particularly narrow bridge, e.g., 6 or 8 ft. bridge,
only one wheel line load can probably be applied.

CSiBridge (CSI, 2021) has the capability to permute all the possible vehicular loading
patterns once a set of lanes is defined. First, the entire bridge response due to a single
lane loaded, without the application of the Multiple Presence Factor (MPF), can be easily
obtained by arbitrarily defining a lane of any width within the bridge. Then, lane
configurations that would generate the maximum shear and moment effects would be
defined and the MPF would be defined. The cases where one lane is loaded are important
for fatigue design; in addition, the cases where one lane is loaded may control over the
cases where two lanes are loaded. Therefore, the cases where one lane is loaded are

4-40 Chapter 4 – Structural Modeling and Analysis

© 2022 California Department of Transportation. ALL RIGHTS reserved.
 Bridge Design Practice 4 • October 2022

separated from the permutation and are defined based on a single lane of the whole
bridge width.

AASHTO standard design vehicular live loads, HL-93, are used as the traffic load for the
CSiBridge analyses of the live load distribution factor (LLDF) (see Articles 3.6.1.2 and
3.6.1.3). Figure 4.4-1 shows the elevation view of the four types of design vehicles per
lane used by CSiBridge, including the details of the axle load and axle spacing. The
transverse spacing of the wheels for the design truck and design tandem is 6 ft. The
transverse width of the design lane load is 10 ft. The extreme force effects, moment, and
shear in girders for this example, at any location of any girder, are the largest from the 4
design vehicles:

• HL-93K: design tandem and design lane load;

• HL-93M: design truck and design lane load;
• HL-93S: 90% of two design trucks and 90% of the design lane load;
• HL-93LB: pair of one design tandem and one design lane load.

Figure 4.4-1 Elevation View of AASHTO Standard HL-93 Vehicular Live Loads
(Caltrans)

Chapter 4 – Structural Modeling and Analysis 4-41

© 2022 California Department of Transportation. ALL RIGHTS reserved.
 Bridge Design Practice 4 • October 2022

Cases (c) and (d) in Figure 4.4-1 are for the maximum negative moment over bent caps.
A dynamic load allowance of 33% is applied and only applied to the design truck and
design tandem in all cases. Multiple Presence Factor as shown in Table 4.4-1 is applied
in accordance with AASHTO-CA BDS-8.

Table 4.4-1 Multiple Presence Factor (MPF)

Number of Loaded Lanes Multiple Presence Factor
1 1.20
2 1.00
3 0.85
>3 0.65

4.4.3 Live Load Distribution For One And Two-Cell Box Girders
Example
Model Bridge in CSiBridge as given data below:

In this example, the method of calculating live load distribution factor is shown for a two-
cell box girder by using a 3-D FEA-CSiBridge model for different lane loading (Figures
4.4-3 to 4.4-6). The bridge data is given as shown below:

• Girder spacing, S: 7′ < S =13′< 13′

• Span length, L: 60′ < L=180′ < 240′
• Structure depth, d: 35″ < d =96″< 110″

Single span, simply supported, 180-foot long, 8-foot depth two-cell Box Girder Bridge with
the following cross section as shown in Figure 4.4-2.

Figure 4.4-2 Live Load Distribution For Two-Cell Box Girders Snapshot

4-42 Chapter 4 – Structural Modeling and Analysis

© 2022 California Department of Transportation. ALL RIGHTS reserved.
 Bridge Design Practice 4 • October 2022

1) Load groups

Load Group 1

Figure 4.4-3 Live Load Distribution For Two-Cell Box Girders Snapshot
In Group 1

Load Group 2

Figure 4.4-4 Live Load Distribution For Two-Cell Box Girders Snapshot
In Group 2

Chapter 4 – Structural Modeling and Analysis 4-43

© 2022 California Department of Transportation. ALL RIGHTS reserved.
 Bridge Design Practice 4 • October 2022

Load Group 3

Figure 4.4-5 Live Load Distribution For Two-Cell Box Girders Snapshot
In Group 3

Load Group 4

Figure 4.4-6 Live Load Distribution For Two-Cell Box Girders Snapshot
In Group 4

In order to calculate the LDF, both spine model and area object model were run for
different lane loading using BrIM.

4-44 Chapter 4 – Structural Modeling and Analysis

© 2022 California Department of Transportation. ALL RIGHTS reserved.
 Bridge Design Practice 4 • October 2022

2) Bridge Modeler (Figure 4.4-7)

• Update Bridge Structural Model as Area Object Model

Figure 4.4-7 Bridge Modeler Snapshot

3) Define Lane (Figure 4.4-8)

Figure 4.4-8 Define Lane Snapshot

Chapter 4 – Structural Modeling and Analysis 4-45

© 2022 California Department of Transportation. ALL RIGHTS reserved.
 Bridge Design Practice 4 • October 2022

• Maximum Lane Load Discretization Lengths:

Along Lane 10 ft
Across Lane 2 ft

4) Define Vehicle (Figure 4.4-9)

Figure 4.4-9 Define Vehicle Snapshot

5) Define Vehicle Classes (Figure 4.4-10)

Figure 4.4-10 Define Vehicle Classes Snapshot

4-46 Chapter 4 – Structural Modeling and Analysis

© 2022 California Department of Transportation. ALL RIGHTS reserved.
 Bridge Design Practice 4 • October 2022

6) Analysis Cases (Figure 4.4-11)

• Group1: 1 Lane loaded

Figure 4.4-11 Analysis Cases Snapshot in One-Lane Loaded

• Group 2: 2 Lanes loaded (Lanes 1, 2 & 3) (Figure 4.4-12)

Figure 4.4-12 Analysis Cases Snapshot in Two-Lane Loaded

Chapter 4 – Structural Modeling and Analysis 4-47

© 2022 California Department of Transportation. ALL RIGHTS reserved.
 Bridge Design Practice 4 • October 2022

• Group 3: 2 Lanes loaded (Lanes 4 & 5)(Figure 4.4-13)

Figure 4.4-13 Analysis Cases Snapshot in Two-Lane Loaded

• Group 4: 3 Lanes loaded (Lanes 1, 2 & 3) (Figure 4.4-14)

Figure 4.4-14 Analysis Cases Snapshot in Three-Lane Loaded

4-48 Chapter 4 – Structural Modeling and Analysis

© 2022 California Department of Transportation. ALL RIGHTS reserved.
 Bridge Design Practice 4 • October 2022

7) Analysis Single Lane Loaded (MPF = 1) with running updated Bridge Structural
Model as Spine Model (Figure 4.4-15)

Figure 4.4-15 Analysis Single Lane Loaded Snapshot

Results:

A) Spine Model
The maximum moments and shears of the entire bridge cross section for one
lane loaded are shown in Figures 4.4-16 and 4.4-17.

The Maximum moment = 6,527 Kips-ft at x = 90 ft

Figure 4.4-16 Maximum Moment Snapshot

Chapter 4 – Structural Modeling and Analysis 4-49

© 2022 California Department of Transportation. ALL RIGHTS reserved.
 Bridge Design Practice 4 • October 2022

The Maximum shear =148 Kips

Figure 4.4-17 Maximum Shear Snapshot

B) Area Model
The maximum moments at exterior and interior girders for one lane loaded are
shown in Figures 4.4-18 to 4.4-20.

Left Exterior Girder, M3 = 2566 Kips-ft at x = 90 ft

Figure 4.4-18 Maximum Moment for One-Lane Loaded at Left Exterior Girder

4-50 Chapter 4 – Structural Modeling and Analysis

© 2022 California Department of Transportation. ALL RIGHTS reserved.
 Bridge Design Practice 4 • October 2022

Interior Girder, M3 = 3410 Kips-ft at x = 90 ft

Figure 4.4-19 Maximum Moment for One-Lane Loaded at Interior Girder

Right Exterior Girder, M3 = 2566 Kips-ft at x = 90 ft

Figure 4.4-20 Maximum Moment for One-Lane Loaded at Right Exterior Girder

Chapter 4 – Structural Modeling and Analysis 4-51

© 2022 California Department of Transportation. ALL RIGHTS reserved.
 Bridge Design Practice 4 • October 2022

The maximum shears at exterior and interior girders for one lane loaded are
shown in Figures 4.4-21 to 4.4-23.

Shear at Left Exterior girder = 154 Kips at x = 0

Figure 4.4-21 Maximum Shear for One-Lane Loaded at Left Exterior Girder

Shear at Interior Girder = 109 Kips at x = 0

Figure 4.4-22 Maximum Shear for One-Lane Loaded at Interior Girder

4-52 Chapter 4 – Structural Modeling and Analysis

© 2022 California Department of Transportation. ALL RIGHTS reserved.
 Bridge Design Practice 4 • October 2022

Shear at Right Exterior girder = 154 Kips at x = 0

Figure 4.4-23 Maximum Shear for One-Lane Loaded at Right Exterior Girder

C) Actual, Modified LLDF:

Shear (Table 4.4-2)

• Actual LLDF = (VL.Ext. + VInt. +VR.Ext.)/ VSingle lane
• Modified LLDF = (Max (VL.Ext., VInt. ,VR.Ext.)) × 3/ VSingle lane

Table 4.4-2 Live Load Distribution Factor for Shear

Case Cell # VSingle lane VL.Ext. VInt. VR.Ext.
L (ft) LLDFActual LLDFModified
# Type Lanes (Kips) (Kips) (Kips) (Kips)

1 X2C8 180 1 148 154 109 154 2.82 3.12

2 X2C8 180 2 148 172 166 172 3.45 3.49
3 X2C8 180 3 148 154 157 154 3.14 3.18

Moment (Table 4.4-3)

• Actual LLDF = (ML.Ext. + MInt. + MR.Ext.) / MSingle lane
• Modified LLDF = (Max (ML.Ext. , MInt. , MR.Ext.)) × 3 / MSingle lane

Chapter 4 – Structural Modeling and Analysis 4-53

© 2022 California Department of Transportation. ALL RIGHTS reserved.
 Bridge Design Practice 4 • October 2022

Table 4.4-3 Live Load Distribution Factor for Moment

Case Cell # MSingle lane ML.Ext. MInt. MR.Ext.
L (ft) LLDFActual LLDFModified
# Type Lanes (Kips-ft) (Kips-ft) (Kips-ft) (Kips-ft)

1 X2C8 180 1 6527 2566 3410 2566 1.31 1.57

2 X2C8 180 2 6527 4046 5657 4046 2.11 2.60
3 X2C8 180 3 6527 4920 7074 4920 2.59 3.25

Although the CSiBridge analysis provides a more exact distribution of force effects in the
girders, it doesn’t calculate the amounts of prestressing, longitudinal, or shear
reinforcement required on the contract plans. Different two-dimensional tools such as
CTBridge are used for design. The girders are considered individually, or lumped together
into a single-spine model.

Caltrans prefers the latter in the case of post-tensioned box girders because post-
tensioning in one girder affects the adjacent girder.

If the individual demands were simply lumped together and used in two-dimensional
software for design and the girders designed equally, at least one girder would be under-
designed. Hence, the value from the girder with the highest demand is used for all
girders–as shown above, so it is recommended to consider LLDF Modified, as the Live
Load Lanes input for CTBridge.

4-54 Chapter 4 – Structural Modeling and Analysis

© 2022 California Department of Transportation. ALL RIGHTS reserved.
 Bridge Design Practice 4 • October 2022

NOTATION
A = element area
c = damping constant
ccr = critical damping
d = structure depth (in.)
E = Young’s modulus of elasticity (ksi)
EDC = energy dissipated per cycle
F = frequency (Hz)
FD = damping force
FI = inertial force
Fmi = motion-independent wind force vector
FS = spring force
FUmd = motion-dependent aerodynamic force vector
g = gravitational acceleration (32.2 ft/sec)
Icr = cracked moment of inertia
Ieff = effective moment of inertia
Ig = gross moment of inertia
IM = dynamic load allowance
J = St. Venant torsional inertia
k = stiffness
L = span length (ft); element length
LLDFActual = live load distribution factor based on the actual force effect in each
girder
LLDFModified = live load distribution factor based on the girder with the maximum
force effect
m = mass
Mbeam = moment in a single beam from a 1-D analysis
MInt. = moment due to live load at the interior girder (kip-ft)
ML.Ext. = moment due to live load at the left exterior girder (kip-ft)
MR.Ext. = moment due to live load at the right exterior girder (kip-ft)
Mrefined = enveloped moment of a bridge section from a refined analysis
MSingle lane = moment due to live load from a single lane on the cross section (kip-ft)

Chapter 4 – Structural Modeling and Analysis 4-55

© 2022 California Department of Transportation. ALL RIGHTS reserved.
 Bridge Design Practice 4 • October 2022

MPF = multiple presence factor

Pjack = prestress jacking force
R1 = rotation about abutment
R2 = rotation about line normal to abutment
R3 = rotation about vertical
S = center-to-center girder spacing (ft)
T = period (sec)
u = displacement
u̇ = velocity
ü = acceleration
U1 = translation parallel to abutment
U2 = translation normal to abutment
U3 = translation vertical
V = element volume
Vbeam = shear in a single beam from a 1-D analysis
VInt. = shear due to live load at the interior girder (kip-ft)
VL.Ext. = shear due to live load at the left exterior girder (kip-ft)
VR.Ext. = shear due to live load at the right exterior girder (kip-ft)
Vrefined = enveloped shear of a bridge section from a refined analysis
VSingle lane = shear due to live load from a single lane on the cross section (kip-ft)
x = location along the bridge span
z = damping ratio
ε = strain
ρs = material mass density
ν = Poisson’s ratio
σ = stress
ω = angular frequency

4-56 Chapter 4 – Structural Modeling and Analysis

© 2022 California Department of Transportation. ALL RIGHTS reserved.
 Bridge Design Practice 4 • October 2022

REFERENCES
1. AASHTO, (2017). AASHTO LRFD Bridge Design Specifications, 8th Edition,
American Association of State Highway and Transportation Officials, Washington,
D.C.
2. ADINA, (2021). Automatic Dynamic Incremental Nonlinear Analysis, ADINA
9.7.3, ADINA R & D, Inc., Watertown, MA.
3. Akkari, M. and Duan, L. (2014). “Chapter 5: Nonlinear Analysis,” Bridge
Engineering Handbook, 2nd Edition: Seismic Design, Chen, W.F. and Duan, L.,
Ed., CRC Press. Boca Raton, FL.
4. Barker, R. M. and Puckett, J. A., (2013). Design of Highway Bridges: A LRFD
Approach, 3rd Edition, John Wiley & Sons, Inc., New York, NY.
5. Cai, C.S., Zhang, W. and Montens, S., (2014) “Chapter 22: Wind Effects on Long
Span Bridges,” Bridge Engineering Handbook, 2nd Edition: Fundamentals, Chen,
W.F. and Duan, L., Ed., CRC Press. Boca Raton, FL.
6. Caltrans (2016) Caltrans Seismic Design Specifications for Steel Bridges, 2nd
Edition, California Department of Transportation, Sacramento, CA.
7. Caltrans, (2019a). California Amendments to AASHTO LRFD Bridge Design
Specifications - 8th Edition, California Department of Transportation, Sacramento,
CA.
8. Caltrans, (2019b). CTBridge, Caltrans Bridge Analysis and Design v. 1.8.3,
California Department of Transportation, Sacramento, CA.
9. Caltrans, (2019c). Seismic Design Criteria, Version 2.0, California Department of
Transportation, Sacramento, CA.
10. CSI, (2021). CSiBridge, v. 23.3.0, Computers and Structures, Inc., Walnut Creek,
CA.
11. Duan, M., Perdikaris, P.C. and Chen, W. F., (1999). “Chapter 56: Impact Effect of
Moving Vehicle,” Bridge Engineering Handbook, Chen, W.F. and Duan, L., Ed.,
CRC Press. Boca Raton, FL.
12. Priestley, Seible and Calvi (1996). Seismic Design and Retrofit of Bridges, John
Wiley & Sons, Inc., New York, NY.
13. Zhang, W., Vinyagamoorthy, M. and Duan, L., (2014). “Chapter 3: Dynamic,”
Bridge Engineering Handbook, 2nd Edition: Seismic Design, Chen, W.F. and
Duan, L., Ed., CRC Press. Boca Raton, FL.

Chapter 4 – Structural Modeling and Analysis 4-57

© 2022 California Department of Transportation. ALL RIGHTS reserved.
 Bridge Design Practice 4 • October 2022

This page is intentionally left blank.

4-2 Chapter 4 – Structural Modeling and Analysis

© 2022 California Department of Transportation. ALL RIGHTS reserved.