INNOVATIVE MASS TRANSIT TECHNOLOGY TO BECOME
LIGHTER AND SAFER WITH AID OF ALGOR SOFTWARE
|

Pneumatic propulsion, or forced air, powers
an innovative mass transit system like this one in Porto
Alegre, Brazil. With the help of Algor, Inc. software, engineers
designed a lighter, safer vehicle.
Photo courtesy of PCE Ltda.
|
As city centers worldwide become more crowded, transportation
engineers seek the best way to get people from here to there.
From New York City’s subway to the Bay Area Rapid Transit system
to Japan’s bullet train and the emergence of magnetic levitation,
mass transit technology evolves in search of more comfortable,
faster, environmentally compatible and less-expensive methods.
California-based Aeromovel USA (http://www.aeromovel.com/)
has developed a unique technology that uses air produced by electric
blowers to propel vehicles along elevated guideways. Aeromovel
trains are running in Porto Alegre, Brazil, and Jakarta,
Indonesia.
Brazilian company PCE Ltda. (http://www.pce.com.br/) used
software from Algor, Inc., (www.algor.com) to perform a
linear static stress analysis on Aeromovel’s truck design. PCE
engineers wanted to know whether the truck assembly would withstand
required loading conditions, which include passenger weight, wind,
centrifugal forces and emergency stops.
|

This illustration shows the CAD solid model
of the vehicle's truck assembly. Severo imported the model
into Algor and used Superdraw to create a finite element
model for analysis.
|
Dagoberto Schubert Severo, director of PCE and a mechanical engineer,
used Algor’s linear static stress analysis capability to determine
fatigue and load limits on the vehicle’s truck assembly.
"Algor offers the most extensive range of capabilities at the
best price," Severo said, explaining why he chose Algor. "That,
coupled with the luxury of being able to purchase only those modules
that you need, makes Algor our FEA tool of choice."
Aeromovel provided Severo with a CAD drawing of the truck, created
in Autodesk’s Mechanical Desktop.
From that start, Severo used Superdraw, Algor’s precision finite
element model-building tool, to create three finite element models
of the truck for analysis. He used bricks to model the wheels
and axes, plate elements to model the metal plates making up the
top and bottom parts of the housing and bricks to model connecting
rods attaching the top and bottom parts.
"Unlike other FEA software that I used in the past, with Algor
you can easily create and directly edit the mesh," Severo said.
"Supersurf is another powerful tool to construct plates, because
after I created the surface and the mesh, I could easily generate
different mesh densities."
Severo then assigned material properties to the parts – most
being assigned steel properties. Severo said he used the ASM Metals
Handbook to determine the properties of the steel. He used rubber
in the areas where the connecting rods joined the top and bottom
housing parts and where the housing meets the wheel axes. The
rubber properties were determined through experimentation, he
said. Because of the rubber material’s nonlinear properties, Severo
also could have used Algor’s Accupak/NLM or Accupak/VE analysis
packages. Accupak/NLM performs linear and nonlinear static stress
analysis using linear or nonlinear material models. Even more
versatile is Accupak/VE, which provides the ability to analyze
models at each moment in time while they are in motion.
Severo defined boundary conditions for his analysis at the point
where the wheels contacted the rails and the point on top of the
truck assembly where it was attached to the vehicle.
Among the three models, he spread 11 different loads: vehicle
weight, passenger weight, increased force of vertical acceleration
in vertical curves, increased vertical force due to lateral acceleration
of the centrifugal acceleration in horizontal curves, reaction
from vehicle retention force, wind action, increased lateral force
due to centrifugal acceleration in horizontal curves, friction
braking, deceleration by emergency brakes, torsion loads, superelevation
in horizontal transition curves and suspension failure.
After each analysis, Severo viewed the von Mises stress results
with Superview, Algor’s postprocessing and presentation package
that allows engineers to quickly evaluate and present their work.
The latest Superview version offers such features as dynamic viewing,
Windows-based pull-down menus and floating toolbars.
Based on the results, he changed the thickness or shape of his
elements until he ultimately came up with a truck that was 30
percent (344 kg) lighter than the original design. The new design
also features fewer welds, meaning fewer areas for structural
fatigue, and is fully compliant with National Fire Protection
Association, Americans with Disabilities Act and other U.S. codes
and standards.
Creating a vehicle as light as possible is critical to the success
of the Aeromovel vehicle, Severo said.
"In the Aeromovel vehicle, unlike most transit systems, there
are no onboard motors or traction devices," he said. "So the vehicle
is very light and simple, carrying two to three times more people
per ton of deadweight than most alternatives. The lighter the
vehicle and the truck, the less energy needed to move the vehicle."
When he was through with analysis of the original design, Severo
took his results to the laboratory for validation. Using strain-gages
and load cells, he set up physical tests of the original and altered
designs to determine the correlation between the simulated and
measured analysis. The physical tests proved Algor’s linear static
stress analysis to be accurate within 7.4 percent with regard
to the connection rods and 3.6 percent with the main axle. Severo
said Aeromovel was comfortable with that margin of error because
of the complicated nature of the analysis.
|

This illustration shows the results of Severo's
linear static stress analysis. Severo examined the stresses
on the main axles and the pin in the center that connects
to the Aeromovel vehicle. None of the stresses were significant
enough to initiate yielding.
|
"I am very satisfied with the results of the tests," Severo said.
"Because of the complexity of the model - an assembly with rubber
- I was not anticipating an exact match between the FEA and physical
test results."
Severo said a prototype of the truck assembly – designed according
to the Algor analysis - is under development and should be available
for testing later this year. He also has more plans to work with
Algor software.
"I am involved in several analyses for smelting companies that
produce aluminum, with which I have to perform electric, thermal
and fluid flow analyses," he said. "These are multiphysics analyses
and we are developing programs to interface with Algor using Microsoft's
Visual C++ in order to automate the design process."
|