|
Dynamic Bridge Analysis |
ASSESSMENT OF ACTUAL IMPACT FACTORS FOR FLORIDA BRIDGES
|
The State of Florida has large numbers of lakes and rivers. Consequently, an efficient transportation
system that we developed in the state is characterized by a significant number of bridges. The Florida Department of
Transportation maintains over 6,200 bridges. Modern design techniques and high performance materials used result in more
economical and lighter bridges, which exhibit significant dynamic response caused by heavy traffic. The major goal of
this research is to assess the actual dynamic impact factor due to dynamic truck . The Maintenance Office of the
Florida Department of Transportation (FDOT) is interested in the assessment of actualoadingl dynamic loading and impact factors
for Florida bridges. This information is critical for the decision making process of routing oversized and overweight
vehicles across the state. A two-prong approach is used in a current research project sponsored by FDOT. In the first
stage of the project, a Finite Element (FE) model of a selected bridge was developed using blueprints provided by the
FDOT Structures Maintenance Office. The deck, railings, girders, and diaphragms were modeled with 3-D solid elements
while the rebars and prestressed strands were represented by beam elements. Initial elongation was applied to the
stands to induce prestressing. Figure 1 depicts the FE model of the girderand one bridge span.
|
 
|
Figure 1. FE bridge modeling
|
A full-scale dynamic loading of the same bridge was conducted by FDOT in the second stage of the project. The test
provided extensive data on displacement, velocities, accelerations and strains of selected points of this bridge.
This data is used for verification of the FE model of the bridge.
|
TRUCK MODELING - REVERSE ENGINEERING
|
Analysis of dynamic interaction between trucks and bridges requires the development of an FE model of a truck and
that of a bridge. A reverse engineering process was used to aid in this effort. A FARO digitizing arm (a counter
balanced, temperature compensated, six degree of freedom measurement arm) and numerous software tools (e.g. AnthroCAM,
AutoCAD and PATRAN) was successfully implemented during reverse engineering process. It resulted in data acquisition
of the major geometric entities of the selected truck including reference points and polylines, which were mapped
into a computer model. The numerical model is shown in Figure 2. It well represents the overall dimensions of the truck,
its weight, and location of its center of gravity. A pneumatic wheel model consists of a tread, sidewall, rim, and a
drum with an airbag constitutive model applied. The suspension system is also well represented in the FE model.
|
|
Figure 2. FE truck model
|
ANALYSIS OF DYNAMIC TRUCK - BRIDGE INTERACTION
|
Traditional bridge analysis is based on several fundamental simplifications of geometry, material models, boundary conditions,
and loading. Bridge live load is considered as one of the most questionable idealizations. Interaction between a vehicle and
bridge structure is usually represented by concentrated and uniformly distributed static loads. Dynamic effects of the actual
live loads are considered only by scaling static loads by estimated impact factors. The magnitude of the dynamic load allowance
(impact factor) is estimated without reference to the actual dynamic behavior of the bridge, truck impact, deck surface
imperfections and dynamic characteristics of the vehicles.Non-linear, explicit FE analysis enables for more advanced numerical,
3-D dynamic analysis of bridge structures. The detailed 3-D models of the bridge and the truck are now used for advanced analysis
of dynamics of the entire truck - bridge system. Advanced material models for steel and concrete, rebar option for modeling of
reinforcement, application of different types of constraints, and damping options allow for an accurate description of the actual
bridge response due to dynamic loadings. Application of these models in FE analysis provides an efficient way for studying dynamic,
vehicle-bridge interaction in time as a function of road and bridge surface imperfections. An example of such simulations and results
is shown in Figure 3. Such computational regime requires a modern computing environment capable of providing research results in
a relatively short period of time. The Crashworthiness and Impact Analysis Lab (CIAL) developed technology and is well equipped
to conduct such research. Several Silicon Graphics Octane workstation with single 300 MHz processors are used with UNIX operating
system for FE model development (pre-processing) and post-processing. Major computations are performed on a fast Origin 2000
Silicon Graphics server with 16 R10000 250 MHz parallel processors. Due to computational complexity of contact - impact problems,
typical computer runs take from 5 to 7 days.
|
|
Figure 3. Dynamic two span bridge analysis
|
| Back to Top |
|
Dynamic Loading of Bridges Due to Approach Depression |
This project is a continuation of our previous study commissioned by the Florida Department of Transportation titled:
"Analytical and Experimental Evaluation of Existing Florida DOT Bridges" in which the bridge-vehicle interaction was
thoroughly investigated with numerical and experimental methods. It was found in that study that the approach depression
due to soil settlement, road imperfection, as well as threshold would significantly increase the dynamic response of
highway bridges subjected to moving heavy trucks. Figure 1 shows the profile of the approach on US 90, before a bridge
#500133 near Chattahoochee, Florida. This phenomenon drew much attention from the FDOT Structures Laboratory. The current
project consists of full scale field testing complemented by advanced computational mechanics study.
|
|
Figure 1. Profile of the approach before bridge #500133
|
FIELD TESTING
|
The information regarding the most common and extreme cases of bridge approach depression and bridge deck imperfections in
Florida will be collected during this study. FDOT Materials Office will be contacted with a request to use a specialized
van for fast road surface scanning. Once the testing crane is selected, its suspension will be examined and the suspension
parameters will be determined through a series of static and dynamic excitation processes. Figure 2 presents a potential
candidate (LTM 1080 crane) for field testing with the total weight of 400 kN.
|
|
Figure 2. LTM 1080 crane as a candidate for field testing
|
The proposed bridge for testing is the same one which was tested before. Some strain gauges installed on the bridge at that
time may still be functional and useful. Others will be replaced by the FDOT Structures Lab staff. The dynamic response of
the bridge will be recorded when the crane crosses the bridge span during a full scale loading experiment.
|
NUMERICAL SIMULATION
|
Detailed finite element models of the bridge and the heavy crane used for the testing will be developed, and the approach
depression and threshold will be represented in the FE models.
|
Figure 3 shows the FE model of bridge #500133 which was developed before by CIAL. This model includes all five structural
components: the slab, beams, bridge barriers, diaphragms, and neoprene pads. The total number of elements for one span exceeds
200,000.
|
|
Figure 3. FE model of bridge #500133
|
These models will be validated by the experimental data and further used to evaluate other bridge-vehicle configurations.
Experimental data and FE analysis results will be used to comprehensively investigate the dynamic response of bridges subjected
to moving cranes under the condition of approach depression, and evaluate the corresponding impact factors. This research is
specifically focused on Florida DOT needs. The results of the project will serve as a basis for highway bridge design and
maintenance.
|
| Back to Top |
|
Crashworthiness of Transit Buses |
REVERSE ENGINEERING
|
A typical engineering process begins with an engineering concept, which often follows with an analysis and design. Desirable
material properties are studied and selected materials are used for part production. Finally all required parts are assembled
in a final product or a system. Although nowadays the process is frequently aided with computers to reduce possible errors,
the final product is not always free of flaws, which could be introduced at any stage of the engineering process. A REVERSE
ENGINEERING process is therefore needed to:
1. Identify all potential problems in all stages of the product manufacturing
process, and
2. Assess an impact of different engineering options/design and manufacturing imperfections on the performance
of the final product.
Reverse engineering begins with a final product, which is disassembled into individual parts. The parts
are taped, scanned, digitized, and mapped into computers for further analysis of the entire design and manufacturing process.
Reverse engineering assists engineers in further improvement of their product.
|
|
Figure 1. Ford Eldorado Aerotech transit bus
|
REVERSE ENGINEERING LABORATORY
|
Reverse Engineering Laboratory (REL) was established in November 2001. REL recently conducted a study sponsored by the FDOT
titled: "Crashworthiness and Safety of Public Transit Buses". The major tasks of this project were the development and validation
of a high fidelity finite element model of Ford Eldorado Aerotech transit bus, which is often used to transport disabled passengers.
The model development requires vehicle teardown and digitization of all major structural parts of the bus. A FARO Arm (a counter-balanced,
temperature compensated, six degree of freedom measurement arm) and numerous software tools (like AnthroCAM, AutoCAD and PATRAN)
were successfully implemented during this stage of the project. The digitizing part of the research resulted in data acquisition
of major geometric entities including reference points and polylines, which are mapped into a computer model. An interim set of
these geometric entities is depicted in figure 2. Geometric data is subsequently used as input at the Crashworthiness and Impact
Analysis Laboratory (see back page of this brochure) for building finite element models.
|
|
Figure 2. Geometric entities of the bus
|
CRASHWORTHINESS AND IMPACT ANALYSIS LABORATORY
|
Geometric data obtained from a digitizing process in the Reverse Engineering Laboratory is used as the first step in building finite
element (FE) models of all major, structural components of the bus. This part of the research is conducted at the Crashworthiness and
Impact Analysis Laboratory (CIAL). PATRAN pre-processor is currently utilized in CIAL as a primary tool to convert geometric entities
into finite element models. Figure 3 shows finite element models of two buses depicting a crash event. The final model of the Ford Eldorado
Aerotech transit bus consists of 73,595 finite elements. In addition to the size of the model, the crashworthiness studies require use of non-linear,
3-D, explicit, dynamic finite element codes capable of handling contact-impact problems with high-strain rate and large non-linear
deformations including plastic deformation and failure. LS-DYNA computer code is a modern computational tool, which is primarily used
for crashworthiness studies at CIAL. Such computational regime requires a modern computing environment capable of providing research
results in a relatively short period of time. CIAL is well equipped to conduct such research. Several Silicon Graphics Octane
workstations with single 300 MHz processors are used in UNIX operating system for FE model development (pre-processing) and
post-processing. Major computations are performed on Origin 2000 Silicon Graphics server with 16 R10000 250 MHz parallel processors.
Due to computational complexity of crashworthiness problems, typical computer runs take from 5 to 7 days.
|
|
Figure3. Finite element models developed from geometric entities
|
Passenger safety and structural integrity of the bus are the primary concerns of this research. Hence, special care is
taken to carefully model the passenger compartment of the bus. This task is computationally challenging since the bus
wall consists of several materials including outer and inner composite layers, structural frame members, honeycomb and
styrofoam insulation. Several tests were performed, including static and dynamic tests of structural samples for
improving the accuracy of the data used for analysis. It is expected that this cutting - edge technology will assist
engineers in design-ing safer and more economical transit buses. Subsequently, research results from this study
may help improving standards similar to FMVSS 220.
|
Back to Top |
|
