Low Floor Light Rail Vehicles-Structural and Crashworthiness Requirements

LOW-FLOOR
LIGHT RAIL VEHICLES

LOW-FLOOR LIGHT RAIL VEHICLES
Structural and Crashworthiness Requirements of

Light Rail Vehicles with Low-Floor Extension
RADOVAN SARUNAC
NICHOLAS ZEOLLA
Booz Allen Hamilton
T his study explores the option of integrating a low-floor extension (LFE) module into the
existing light rail vehicle to augment current service with low-floor boarding and
simultaneously to increase passenger capacity. In an effort to minimize the overall weight of the
LFE, this study examines two feasible approaches for specifying the longitudinal strength of the
LFE that are consistent with current industry practice. The two alternatives considered are the
“strength-based approach” and the “energy-based approach.” The de facto industry practice is a
strength-based approach where the buff strength is specified to meet two times the AWO weight
of the vehicle, commonly referred to as the “2-g buff load.” The feasibility of an alternative
approach, using crash energy management (CEM) principles to limit the longitudinal loads in a
controlled manner, was also examined in detail. In a collision, the CEM zone would generate
predetermined peak and average reaction loads, thereby limiting the longitudinal load transmitted
to the LFE, while still meeting minimum static strength values. These zones would be designed
to deform or crush in a controlled manner, thereby absorbing collision energy and minimizing
the potential for deformation and acceleration in the passenger compartment.
Since structural strength is fundamental to passenger safety, the new structure of the LFE
module and any modifications made to the existing vehicle must meet the same structural
standards used originally or must provide an equivalent level of performance in a collision. For
this study, collision performance is being measured only in terms of peak and average
accelerations and deformation.
BACKGROUND
Currently, light rail vehicles (LRVs) comply with the federally mandated Americans with
Disabilities Act requirements by providing access to disabled patrons through the use of wayside
high blocks combined with onboard bridge plates. This set-up has limited access to the leading
doors on each train for passengers using the high block, which is typically used by passengers
who are disabled or those with luggage or strollers.
The Authority wanted to investigate the possibility of replacing the existing method of
providing access and increasing system capacity at a lower cost. The concept under consideration
is the addition of an low-floor extension (LFE) to the center of existing LRVs. The driving force
for investigating this concept is that the “non-powered LFEs may be added to the entire fleet at a
significantly lower cost, as compared to procuring a fleet of new low-floor LRVs.”
89
90 Transportation Research Circular E-C058: 9th National Light Rail Transit Conference
LFE CONCEPT
The conceptual LFE modification to the existing LRV may include the insertion of a 30-ft-long
low-floor section between the two existing car bodies. The low-floor section is constructed of
steel in a similar manner as the existing car. For the purposes of this study, the existing
propulsion capacity was assessed only for an LFE constructed from steel similar to that used in
existing LRVs. As the propulsion capacity remains unchanged, the heavier steel LFE was used in
all calculations. Two articulation joints identical in design to the existing joint are located at each
end of the LFE body. Conventional un-powered trucks identical to the existing center truck
design support each articulation joint. Doors located on each side of the car body at the middle
provide easy entrance and exit from the car at curb level. The LFE section provides 20 seats,
standing area for 43 passengers at crush load, and air comfort through an independent heating,
ventilation, and air conditioning system.
The LFE concept developed employs the existing propulsion equipment with no changes
to the principal physical components such as traction motors, gear units, propulsion inverter, and
high-voltage systems. As a consequence, the maximum acceleration rate is reduced.
Among other issues, structural strength and crashworthiness are two performance
parameters being reviewed because the available designs for the LFE section do not meet the
same structural requirements as those designed for the existing vehicles. Since structural strength
is fundamental to passenger safety during a collision—for example, it acts to prevent the collapse
and gross penetration of occupied volumes—the new structure of the LFE module and any
modifications made to the existing vehicle must, at minimum, meet the same structural standards
used in the existing vehicle so that the LFE structure provides an equivalent level of performance
in foreseeable collision scenarios. An LFE design with less strength than the existing passenger
compartment would be at risk of sustaining a large part of the deformation in a collision,
potentially endangering passengers.
LFE STRUCTURAL ISSUES
To determine the technical feasibility of implementing the LFE from a structural perspective, the
primary structural characteristics of the LFE LRV, namely the car body compression strength
and crush strength, were studied. The key issue is in determining the load values to use in the
design of the LFE such that adequate structure strength is provided without excessive weight.
One of the main parameters considered in the design of railway structures is the longitudinal
yield strength or buff strength of the vehicle. This load requirement is important because when
vehicles are properly designed to this load, they are protected against large-scale deformation,
penetration, and damage during a variety of collision events ranging from hard couplings to end-
to-end collisions with other rail vehicles, and, to some degree, collisions with highway vehicles.
Typical industry practice for LRVs has been to design the vehicle structure to resist a
static load equal to two times the empty weight of a ready-to-run vehicle applied at the ends of
the vehicle without permanent deformation (yield or buckling), often called the “2-g buff load”
criterion.
Sarunac and Zeolla 91
STRUCTURAL STRENGTH OVERVIEW
As the LFE concept has been popular in Europe exclusively, the structural characteristics and
crashworthiness of the LFE need to be reviewed thoroughly to account for the differing design
philosophies between Europe and North America. In North America, traditionally a 2-g load
(two times vehicle ready-to-run weight) has been required, in some cases by law (California
CPUC).
In Europe, the compression strength is typically specified according to the vehicle type,
category, or rail service and not necessarily according to the tare or ready-to-run weight only.
For instance, in Europe, the typical compression strength of an LFE LRV is in the range of
40,000-100,000 lbs; however, the requirements for North American LRVs are in the range of
140,000-200,000 lbs. Caution must be used when comparing requirements for European LRVs
with their North American counterparts, particularly for structures and crashworthiness, because
the operating speeds in North America tend to be higher, even though the classification or label
of “LRV” is used in both regions.
CRASHWORTHINESS—STANDARDS AND INDUSTRY PRACTICE
In the past, the crashworthiness of LRVs was mostly defined by their buff load. A high buff load
was seen as the most straightforward path to ensuring that the vehicle structure did not suffer
large-scale collapse in collision, hence providing protection to passengers. Traditionally a 2-g
buff load has been required for LRVs in locations such as Baltimore, Maryland; San Francisco,
Santa Clara, San Diego, and Los Angeles, California; Denver, Colorado; St. Louis, Missouri;
Pittsburgh, Pennsylvania; and Boston, Massachusetts. The exception is New Jersey, where 1.0 to
1.1 g was specified. The Parsons Brinckerhoff study for NJ Transit’s LRV revealed that the
working team could not determine the technical rationale behind the 2-g buff load. The 1-g
requirement could not be explained convincingly either. Furthermore, specifying a minimum
buff load does not necessarily limit the peak loads and accelerations experienced during a
collision because the collapse load of the structure is often higher than the static buff load.
Today, the industry is considering an alternative approach. The strength-based
crashworthiness approach is augmented with energy-based crashworthiness requirements; the
ultimate goal is to reduce passenger deceleration rates during a collision while controlling the
absorption of collision energy to minimize loss of space in the occupied volume of the vehicle.
The strength-based philosophy is currently the most widely used approach to ensure some
level of rail vehicle crashworthiness. Under this approach, vehicle specifications typically limit
the various strength requirements of the structure. The most significant strength requirement is
the buff strength. Ultimately, the strength requirement does not consider the energy-absorbing
capability of the vehicle structure. If the structure were designed not to crush by utilizing very
high strength, then the load and decelerations would also reach very high values during an
accident. Depending on the design, collision energy might be dissipated as fracture, derailment,
or override—all of which are generally uncontrolled and undesirable outcomes.
The energy-based crashworthiness (or CEM) philosophy is based on providing protection
against one or more specific collision scenarios. These scenarios are used to determine, by
considering the physics of collisions, an estimate of the energy that must be absorbed and the
structural components that must be used to absorb the energy. Selection of specific collision
scenarios is the most difficult step in this approach because there must be a rationale underlying
selecting parameters such as speed, consist configuration, and impact angle. The selected
parameters must be representative of the most likely collision in a given system. Such a scenario
can be derived from the review of accident data or from a careful review of proposed operations
and the types and likelihood of various accidents. A common accident scenario for passenger
train operations is the collision of a moving train with a stationary train. The moving train is
often considered to be traveling at a speed substantially lower than its maximum operating speed
because the collision event is envisioned to occur as a train is entering the station or has its
brakes applied in anticipation of the collision. Once the accident scenario is selected, it is
necessary to decide how the energy will be dissipated between various vehicles in the train and
the individual ends of a vehicle. A very conservative approach is to require the impacted ends of
the lead vehicles to absorb all of the collision energy in its couplers and end structure. However,
optimized designs tend to distribute the collision energy along the entire train consist. Collision
dynamics computer models can be used to obtain estimates of the distribution of collision energy
in the train consist.
The energy-based approach also requires carefully designed structural features in order to
ensure that a survivable volume remains after collision energy is absorbed. Limiting
accelerations, which can result in secondary impact injuries to passengers, is also a main goal of
this approach. To ensure that passenger areas maintain a survivable volume, the passenger
compartment structure should have greater strength than the crushable zones, and the amount of
deformation or crush required to absorb specified collision energy must be limited. An equally
important consideration is the secondary impact injury to passengers. Secondary impact injuries
can be controlled by limiting the strength requirement for the crush zone (the maximum force
required to initiate controlled deformation), thereby limiting the maximum passenger
accelerations.
The use of CEM principles is also being considered for new standards currently
developed by industry committees. The American Society of Mechanical Engineering Rail
Transit-1 (ASME RT-1) standards committee is in the process of developing a standard for
structural requirements for LRVs, which includes CEM specifications similar to European
standards, although important revisions are being made to meet North American standards. The
standard will include the same general approach for crashworthiness as that discussed above—
absorption of collision energy in a controlled manner at a predefined location on the vehicle
structure. Typically, collapsible structural elements are located either at the end(s) of the car or
within the end structure. Depending on the type of vehicle and the collision scenario considered,
the energy absorption structure could have a static “end” strength lower than the typical buff
strength (2-g) value in order to minimize the average deceleration in the passenger compartment.
To guard against loss of occupied volume, passenger compartments will be specified to have a
higher strength than the crush zone to sustain loads from the collapsing structure.
An example of a crash energy design requirement in the proposed RT-1 standard is that
each car end should be designed to absorb 350 kJ of energy with a collapsing distance of 20 to
28 in. for a collision between two LRVs: one moving at 12 mph, the other parked with its brakes
off. [Development on the RT-1 standard continues; discussions during the American Society of
Mechanical Engineers (ASME) meeting in March 2002 will likely result in further revisions to
the requirements.] Another example is New Jersey (Hudson–Bergen) where 308 kJ was required.
It should be noted that these requirements are specified for new vehicle procurements.
LRV–LFE DESIGN ALTERNATIVES
This study was focused primarily on the overload conditions resulting from vehicle collisions in
order to assess the buff strength requirements for the LFE module. From the structural
perspective, the design of LFE can be executed using the following criteria for the longitudinal
strength of the vehicle:
• Design the LFE to follow the traditional 2-g criteria based on using the new total
weight of a vehicle with an LFE.
• Design the LFE to match the existing LRV buff load and collapse load; for example,
use the strength-based LRV design philosophy.
• Design the LFE using a strength-based approach in conjunction with the CEM
applied to the existing vehicle structures to definitively limit the longitudinal loads during a
collision.
These design approaches are not new to LRV designs and have all been applied on
various light rail projects throughout the United States, with the most typical approach being the
2-g criteria. (Examples include San Francisco Municipal Rail LRVII; Pittsburgh Stage II; Boston
#8; and Southern New Jersey Light Rail, to name a few.)
Industry Practice—Traditional 2-g Criteria
The most straightforward approach for sizing the buff strength of the LFE would be to maintain
the 2-g ratio based on the overall weight of the modified vehicle. This is the strength-based
approach, which has the benefit of simplicity and service history. However, when compared with
the other approaches, the 2-g criteria could result in added weight from the structure, which is
undesirable. If the 2-g ratio is strictly maintained, the empty weight of a vehicle with a 30-foot-
long LFE is estimated to be in the range of 145,000 to 147,000 lbs. Therefore, the buff load of
the LFE would need to be approximately 294,000 lbs plus some additional factor to ensure that
the LFE would not collapse. In addition, the 2-g approach, again if strictly followed, would
require some structural modifications to the existing cars, which were designed for a buff load of
200,000 lbs, although some excess capacity might already exist in the structure. The
disadvantages of this approach are excessive weight from the LFE and the possible modifications
to the existing vehicle. The increased vehicle buff strength, as discussed before, does not
necessarily provide for a crashworthy design because the higher longitudinal body stiffness can
result in higher passenger decelerations in a collision.
Strength-Based Design: Match the Existing Buff Strength
Another option for the LFE design could be to follow a second strength-based design
philosophy—that is, to match the existing buff load (for instance, 200,000 lbs). The main
advantage of this approach is that the existing cars remain unchanged or may need fewer
modifications than with the 2-g design. The weight of the structures would also be lower as
compared with the traditional 2-g approach. Using the LFE AWO weight of 33,000 lbs, the “g”
ratio of the LFE LRV would be approximately 1.4 g; the traditional industry practice of 2-g
would not be achieved. Therefore, the performance of the structures during a collision must be
carefully examined to ensure that the LFE module will not suffer loss of occupied volume during
a collision. Hence, the difference between the buff strength, which is typically specified as a
minimum value, and the actual collapse strength of an existing vehicle must also be examined
relative to the design of the LFE.
One disadvantage of this approach that must be considered is legality. If some of the
revised strength requirements are lower than those traditionally used—for example, lower than
the 2-g industry practice without any other reinforcements or changes—there is potential
exposure to a legal argument in case of injury or fatality. The inevitable argument would be that,
if the LRV had been designed to “industry standards,” then this scenario would have never
happened.
Strength-Based Approach with Crash Energy Design Principles
This approach is essentially a variation of the strength-based approach as described above. In this
case, it is assumed that the existing structure requires modification in order to achieve the desired
force level and deformation response during collision. The foundation of the CEM concept is the
definition of crush zones where impact energy is absorbed in a controlled manner. The structure
in the CEM zone is designed based on vehicle weight, estimated velocity at impact (i.e., kinetic
energy), available energy absorption via the coupler, available space, and passenger deceleration
limits. By limiting the peak buff load generated during the collision event via the CEM zones,
the overall weight of the structures may be reduced as compared with the strength-based (2-g)
approach. Deceleration levels could also be lower with CEM, thereby reducing the potential for
injury to passengers. The CEM zone must be optimized such that the desired amount of energy
absorption is achieved without creating forces that exceed the strength and space limitations of
the vehicle structure.
APPROACH AND METHODOLOGY
The use of CEM principles includes the important step of defining the collision scenario used to
size the energy absorption structures. For typical LRV specifications written to date, the collision
scenario is based on LRVs colliding with other LRVs on the system. [Examples of where CEM
concepts using pre-defined collisions have been utilized include Southern New Jersey Light Rail,
Hudson–Bergen Light Rail, John F. Kennedy (JFK) Airport Access, and the ASME RT-1 Light
Rail specification, which is currently under development.] Collision scenarios are typically
defined by specifying the following:
• The track geometry and conditions (typically dry, level tangent track);
• The couplers and anticlimbers fully engaged;
• The number of vehicles in the consists involved (typically the maximum consist size
for both trains);
• The vehicle weights, including passenger loads; and
• The impact speed, also called closing speed (the maximum value often used is 15
mph for systems with a 55 to 60 mph maximum design speed).
The values noted are examples. These variables must be reviewed for each system when
defining a scenario so that the CEM zone is custom-designed for the intended service. When
defining the design collision scenario, it is particularly important to consider the system speed
(since the collision energy to be absorbed by the structure increases with the square of velocity).
Once the collision scenario is defined and the resulting collision energy is determined, the CEM
zones can be designed.
Again, the force level and available space must be considered so that the smallest
possible deformation is achieved without unduly high accelerations. The desired deceleration
rate, location, available space, and strength of the existing structure are all key variables in
determining the size and force level of the energy absorption devices. In general, the longer the
absorption device, the lower the force level (deceleration rate) can be to absorb a given amount
of energy. As part of this study, the use of the energy-based design approach will be considered
in an effort to limit the peak longitudinal loading on the vehicle, and hence the LFE.
Approach
A variety of collision scenarios were analyzed using dynamic motion analyses to estimate
acceleration (or deceleration) levels in each vehicle and overall deformation. The analyses use
lumped mass-spring models to estimate the interaction and response of vehicles.
These analyses are first-order approximations of the collision events. The response of the
structure on a detailed level is not obtained from this approach. Rather, the global structural
characteristics needed to achieve a desired response (such as a limited deceleration rate, limited
peak force, etc.) can be ascertained from the results. Collision scenarios similar to those
discussed previously will be defined and simulated. The analysis is conducted using the
following steps:
• Define the vehicle characteristics to use as model input data;

• Define the collision scenarios;
• Build the model and run the analyses;
• Collect the data generated from the simulation and examine the results; and
• Formulate conclusions.
The simulation is based on the dynamic analysis of mass-spring systems. A commercially

available computer program, Working Model 2D, was used to formulate the analysis and solve
the resulting system of equations. (Working Model 2D has been used by Booz Allen on the JFK
Airport Access project, on the Southern New Jersey Light Rail project, and for the development
of the ASME RT-1 standard to simulate vehicle collisions.) Working Model 2D can simulate the
static and dynamic (including impulse) response of any combination of masses, springs,
dampers, pulleys, and gears under external loads such as forces, friction, and torque. All of the
elements and parameters in the program can be customized to simulate any desired scenario,
including non-linear or time-varying inputs such as a spring with non-linear stiffness or a
pulsating force.
In the simulation, vehicle consists were placed in motion at the envisioned collision
velocity and collided with standing vehicles that had their brakes on. Vehicle configurations
included in the study were
• Existing vehicle with no LFE;

• Existing vehicle with an LFE and CEM zones (strength-based approach with CEM);
and
• Existing vehicle with an LFE both designed to 2-g based on the new total weight.
These configurations were selected because they represent the possible range of vehicle
strengths. The strength-based approach was not modeled because it is essentially a variation of
the CEM approach.
These configurations were analyzed under four different scenarios, each scenario having
three closing speeds. The collision scenarios included single- and multi-vehicle consists,
modified vehicles with LFEs, and the existing vehicles.
For each analysis conducted, the vehicles were modeled as a rigid block or series of
blocks connected by springs with stiffness values selected to simulate the crush strength of the
vehicle structure or couplers. The stiffness and crush strength of the structures were represented
as spring elements located between the respective mass elements. Where possible, a single spring
was used to represent components that act in series, for example, couplers between vehicles in a
train. Couplers at the front of the colliding trains have been modeled individually to explicitly
determine deflection at the impact point. Connections such as articulation joints are not modeled
directly; sections of a complete vehicle are connected directly by spring elements simulating the
stiffness of the structure so that load transfer between sections (the end car and LFE) is properly
simulated. A two-step approach was used for this study:
• Simulation 1 included a number of analyses to determine the worst-case combination

of speed and number of vehicles in terms of both deformation and accelerations. Particular
attention will be given to deformation because the available space on the existing cars is a clear
constraint. For this simulation, the mass of the LFE was lumped into the mass of the end cars and
the LFE structure was simulated by the springs between the end cars. Hence, acceleration values
will be average values since the LFE mass is not discretely modeled.
• Simulation 2 was conducted using refined models so that the worst-case scenario
determined from Simulation 1 can be examined in more detail. The refined model included
additional mass-spring elements to depict the LFE unit separately.
For this study, the collision performance has been measured only in terms of peak and
average accelerations and deformation.
Methodology and Definition of Vehicle Model
The vehicle characteristics in the models were selected using data and structural analysis
information from the procurement project records. The data collected from previous evaluations
or studies of the LFE concept were also used. Where noted, estimates were made for any missing
details based on established industry standards and past experience. The main input data for the
vehicle model are weight, coupler force data, and structural stiffness (force deformation
characteristics).
For the purposes of this study, it is assumed that the collapse load will be higher than the
buff load and will remain constant as the end car structure is crushed. Although this assumption
is highly dependent on the structural geometry of the under-frame, the intent of the study is not
to evaluate a specific design, but rather to compare vehicle design philosophies and how they
might influence the design loads for the LFE. In order to limit the number of permutations in the
analysis, the LFE buff strength was selected based on a reasonable maximum value given
consideration for weight limitations.
The weight for the LFE using CEM on the existing vehicle was estimated using the unit
weight of the existing vehicle: 1,100 lbs/ft. This value is assumed to include the weight from
structures added for the CEM zone. The unit weight estimate is somewhat conservative because
it includes the weight of subsystems, such as propulsion inverters, that will not be present on the
LFE. For this study, it has been assumed that the collapse strength of the LFE is equal to the buff
strength, which is a very conservative assumption. In reality, collapse strength or ultimate
strength is typically higher than the static buff strength.
Additionally, models with CEM zones require the maximum length of crush to be
specifically defined in order to properly simulate the difference between CEM and the passenger
compartment collapse. The following lengths were used:
• Crush zone length for Simulation 1 is 21 in. (this value was selected as a base point).
The zone is assumed to begin at the back of the anticlimber, extending 21 in. into the cab area.
• Crush zone length for Simulation 2 is 30 in., extending from the back of the
anticlimber inwards.
The springs that represent the passenger compartments and the LFE structure have been
modeled with sufficient length to capture the full dynamic response of the system; therefore, the
maximum crush can be estimated. In other words, a crush length was specified for CEM zones
only.
Generation of Characteristic Force Deflection Curve for the Vehicles
The individual structural characteristics and the crush zone lengths noted in the previous section
were combined in the models to represent the response of the structure under different collision
scenarios. The simulation depicts the response of the vehicle as a whole and of the separate
“cars” or modules (such as the leading car or LFE). To achieve this, the idealized force-
deflection curves were generated using the strength characteristics. The curves represent the
average force generated as the coupler and structure collapse over distance. These curves are
idealized to simplify the calculation; they are estimations with sufficient accuracy for a
preliminary study. In reality, a rail car structure collapsing under dynamic loading displays
highly varying force deflection curves with many peaks, although an average level is typically
maintained.
The force versus deflection curves used in the simulations represent
• The existing vehicle, including the couplers;

• The end cars of a vehicle with CEM zones designed to 200,000 lbs buff strength;
• The end cars of a vehicle designed to a 2-g (based on total weight of end cars and the
LFE) buff strength;
• The LFE module designed to the 200,000 lbs buff strength; and
• The LFE designed to 2-g (based on total weight).
Once all properties (i.e., velocity and static/dynamic frictions) of the mass and spring
elements are defined and connected, the model is ready to be analyzed.
Analysis Assumptions and Collision Scenarios
The analysis is based on assumptions that are commonly used in recent vehicle specifications
(for example, Southern New Jersey Light Rail and JFK Airport Access), which include
crashworthiness and CEM requirements. The assumptions are:
• Both consists are on level tangent track;

• The standing vehicle(s) is fully braked and the coefficient of friction between the
wheel and the rails is equal to 0.3;
• The couplers fully engage and absorb energy during all collisions;
• The moving vehicle(s) is not in braking mode—for example, it maintains a constant
velocity up to impact;
• The maximum number of vehicles in any consist is two; and
• Collisions with vehicles of dissimilar strength may occur.
The following collision scenarios were analyzed as part of Simulation 1:
• Scenario 1: One existing vehicle colliding into two vehicles with an LFE and CEM.
• Scenario 2: One vehicle with an LFE and CEM running into two existing vehicles.
• Scenario 3: Two existing vehicles running into two existing vehicles.
• Scenario 4: Two vehicles with LFEs and CEM running into two identical vehicles.
Each of these scenarios was analyzed at 5, 15, and 17.5 mph closing speeds. The first two
speeds were selected based on industry practice. Typical LRV specifications with
crashworthiness requirements call for a maximum collision design speed of 15 mph. The 17.5
mph was selected because it is approximately half the estimated average system speed. It has
been included for reference only.
RESULTS
The following table shows the results obtained for the simulations conducted. In Simulation 1,
several different scenarios were analyzed to determine the worst-case scenario. Typically, the
scenario with the highest speed and largest consist proves to be the worst case for consists of the
same vehicle strength, although this may not always be the case when vehicles of mixed strength
are considered.
Results of Simulation 1: Determining Worst-Case Scenario—Simplified Model
In Simulation 1, all of the scenarios noted previously were analyzed. The results of these
scenarios, shown in Table 1, indicate that, in collision Scenario 4, two modified vehicles (with
LFE and CEM zones) colliding with the same two vehicles consist is the worst case, assuming
TABLE 1 Summary of Results for Collision Scenario 4 (Simplified Model)
SIMULATION 1
Scenario 4: Two Vehicles with LFEs and CEM Colliding into Two Similar Vehicles
A) Speed = 5 mph
Passenger
Total Coupler CEM Zone
Compartment Max. Accel (+)/ Avg. Accel (+)/
Vehicle Deformation Deformation
Deformation Decel. (-) (g) Decel. (-) (g)
(in.) (in.)
(in.)
Vehicle 2 0.39 --- --- -0.51 -0.51
Vehicle 1 9.8 --- --- -0.51 -0.51
Vehicle 3 12 0.5 --- 0.51 0.51
Vehicle 4 0.49 --- --- 0.51 0.51
B) Speed = 15 mph
Passenger
(in.) (in.)
(in.)
Vehicle 2 10.6 0 --- -0.51 -0.51
Vehicle 1 12 21 4 -1.18 -1.18
Vehicle 3 12 21 2 1.18 1.18
Vehicle 4 11 0 --- 0.51 0.51
C) Speed = 17.5 mph

Passenger
(in.) (in.)
(in.)
Vehicle 2 12 2 --- -1.35 -1.35
Vehicle 1 12 21 15.6 -1.20 -1.20
Vehicle 3 12 21 4 1.20 1.20
Vehicle 4 12 2 --- 1.35 1.35
the entire fleet is fitted with LFEs (mixed scenarios—i.e. vehicles of different strength colliding
together—were also considered and will be discussed in a later section).
At 15 mph, the coupler and the CEM zone are expended completely and deformation has
begun in the passenger compartment on vehicles 1 and 3, which are at the impact interface. This
result indicates that a 21-in. CEM zone at 200,000 lbs does not have sufficient energy absorption
to prevent deformation of the passenger compartment at 15 mph. In Simulation 2, the CEM zone
was increased to 30 in., but the force level stayed the same. This was examined closer in the
refined analysis (Simulation 2) with multi-spring models. The average accelerations are about
the same for each of the respective vehicles in the train consists.
The main goal of this simulation was to confirm that the two-cars-into-two-cars scenario is
the worst case, which has been proven. Thus, Scenario 4 can be used in the refined analysis without
the need to check all other scenarios.
Results of Simulation 2: Refined Analysis for Scenario 4
Using the three-block model refinement and collision Scenario 4, another analysis was conducted for
5, 15, and 17.5 mph speeds. Note that the length of the CEM zone is increased to 30 in. (versus 21 in.
from Simulation 1). This was done to allow for sufficient deflection so that the maximum length
required in the CEM zone could be determined. As in Simulation 1, the crush force of the CEM zone
was 200,000 lbs. All of the parameters and assumptions noted previously were also retained. The
impact occurs at the same interface as Simulation 1, between vehicles 1 and 3. The results for this
simulation (at 15 mph) are presented below.
Vehicles 1 and 2—Moving Consist
• Coupler stroke is used entirely at both ends of vehicle 1, one end located at the impact
interface and the other coupled with vehicle 2.
• The entire coupler stroke is expended on vehicle 2, indicating that the collision energy
has been distributed in part along the length of the vehicle through the couplers. Although the
deformation-to-crush zones show that the majority of the energy was absorbed at the impact ends.
• The CEM zone crushed on vehicle 1 (at end car 1A only) was over 28 in. No damage
occurred in the passenger compartment. This indicates that, in principle, a 30-in. crush zone with an
average force of 200,000 lbs is sufficient to absorb the collision energy generated from this scenario.
• The LFE deformed approximately 1 in. in total (the sum of the deflections at both ends).
Hence, the LFE had just begun plastic deformation in this model, but extensive collapse (loss of
occupied volume) did not occur. Peak deceleration was 1.72 g.
• Since the LFE did not significantly crush, the CEM zone in this model was effective in
limiting peak loading.
• Peak deceleration levels occurred in vehicle 1 at 2.21 g. This deceleration level is not
excessively high and it occurred over a very short duration. The average accelerations were lower.
Vehicles 3 and 4—Standing Consist (Brakes On)
• The coupler stroke was used entirely at the impact interface of vehicle 3. At the coupling
interface, between vehicle 3 and vehicle 4 only, 11.3 in. was used.
• The CEM zone on vehicle 3 was crushed almost 19 in. Peak acceleration was 2.7 g. The
length of crush was not the same as vehicle 1 because of the brake force (simulated as a friction
force). It is expected that one vehicle would sustain more damage than another when the standing
vehicle has its brakes applied.
• As in vehicle 3, the LFE deformed approximately 1 in. in total (the sum of the
deformations at both ends). Hence, the LFE has likely just begun plastic deformation, but extensive
collapse (loss of occupied volume) did not occur. The maximum deceleration was recorded as 2.47 g.
Comparison of the CEM Versus the 2-g Approach—Scenario 4.1
An additional scenario similar to Scenario 4 was analyzed using vehicles designed to the 2-g
approach. As previously noted, the 2-g strength ratio is based on the assumed weight of the LFE
and the existing vehicle. The results of this scenario were compared with Scenario 4 to determine
the differences between using CEM and the 2-g approach:
• In comparing the 2-g results with the CEM design, it can be seen that the CEM zones
limit the peak and average accelerations in passenger compartments significantly.
• The strength-based approach with a CEM zone was successful in limiting the loading
transferred to the LFE body because the acceleration levels are lower at the LFE.
• The strength-based approach with a CEM zone also reduces the acceleration levels in
the second vehicle in the train consist.
• The deformation is higher in the ends of the vehicle with CEM, although this is
expected. The intent of the CEM zone is to deform in a controlled manner (i.e., at a specific
force) over a predetermined distance.
• The second vehicle in both consists sustained structural damage, indicating higher
energy levels between coupled vehicles. This did not occur in the CEM case due to the lower
weight of the vehicles and the energy absorption at the impact interface.
• The deformation level in the LFE is very low for both the CEM and 2-g vehicle.
Hence, loss of occupied volume did not occur. However, the overall weight of vehicle with an
LFE and CEM is estimated to be less than the 2-g approach. This weight savings may be
significant when considering propulsion and braking requirements. Additionally, the lighter LFE
may have lower energy and maintenance costs.
To simplify the comparison, Table 2 shows only the maximum values for the deformation
and acceleration along the vehicles in the moving consist from Scenarios 4 and 4.1.
TABLE 2 Comparison of Results between the CEM and 2-g Approach for Collision
Scenarios 4 and 4.1 (2 vehicles into 2 vehicles at 15 mph)
1st Vehicle 2nd Vehicle

CEM 2-g CEM 2-g
Max. Deformation (in.)
Coupler 12 12 12 12
End Car Structure
28.4 11.3 0 1.3
(CEM or Under-frame)
LFE Structure
0.5 0.35 0.173 0.3
(Max. value at any end)
Accelerations Peak, g
End Cars 2.2 4.6 0.85 4.7
LFE 1.72 3.7 0.51 4.0
Accelerations Avg., g
End Cars 1.0 1.5 0.51 2.1
LFE 1.1 1.5 0.51 1.7
SUMMARY AND CONCLUSION
A number of collision scenarios were analyzed via first-order lumped mass-spring dynamic
models to examine the feasibility of implementing an LFE module designed to a longitudinal
compression specification not based on the 2-g de facto industry standard. In the strictest
interpretation of the 2-g approach, both the existing vehicles and the LFE would need to meet
this requirement based on their total combined weight. This would result in a buff-strength value
for the LFE and end cars that is higher than that of the existing LRVs. Instead, the use of a
strength-based approach with CEM principles to limit peak longitudinal loading in the vehicles,
and hence the LFE, was examined with the goal of minimizing the added weight from an LFE.
Using available data from the records, estimates were made for the weight and strength values of
the LFE and for the size and strength of the CEM zone. The maximum longitudinal strength of
the LFE, when fitted to a vehicle with CEM, was assumed to be 200,000 lbs.
The dynamic analysis was conducted to examine the acceleration and deformation of
various simulated structural elements in the vehicle under different collision scenarios, with
speeds up to 15 miles per hour. Analyses with speeds up to 17.5 miles per hour (half the average
system speed) have been included for reference.
Based on the data available, the assumptions noted, and the analysis results presented in
this report, the following is concluded:
• There is a risk that an LRV LFE designed to the traditional 2-g criterion would not
meet the desired structural weight limits needed to reinforce existing structure due to additional
weight of LFE.
• It is technically feasible to implement an LFE designed to a specific longitudinal
strength value ( i.e., using a strength-based approach), instead of applying the 2-g buff load
criteria as calculated from the new total weight. The longitudinal strength value for the LFE
could be established in two ways.
Option 1 is to maintain a minimum static strength (200,000 lbs for instance). Determine
the load generated from the existing end car structure in a collision under a rational collision
scenario and design the LFE structure to resist this load without loss of occupied volume.
However, due to the structural design of the existing end structures, the end sill and center sill
could be very stiff and may generate very high loads before collapsing in a vehicle-to-vehicle
collision. The end structures must be capable of progressive energy absorption in order to avoid
high load fluctuation and excessive or uncontrolled deformation. It is likely that the significant
changes in the existing structure would be required to achieve this.
Option 2 is to augment the end car structure, in accordance with CEM principles, by
adding the energy absorption elements designed to mitigate a rational collision scenario. As in
Option 1, the minimum static strength must at least match that of the existing structure. The
results of the analysis in this study demonstrate how adding CEM can be beneficial in limiting
longitudinal loads transferred through a vehicle during the collision scenarios noted. The analysis
shows that for a collision of two vehicles into two vehicles at 15 mph, (a) for vehicles with an
LFE and CEM zones having an average crush strength of 200,000 lbs, the collision energy could
be absorbed with approximately 28 in. of crush at the front of the vehicles involved; (b) as with
the CEM zone, the LFE module, having 200,000 lbs of crush strength, sustained approximately 1
in. of total deformation under this scenario; and (c) for vehicles with an LFE designed to the 2-g
approach, approximately 12 to 13 in. of deformation occurs at the ends of each vehicle, which is
less than the CEM approach. However, the peak acceleration levels are much higher.
For either option, the collapse strength of the structures must be
examined in detail. Passenger compartments of the existing vehicle and the LFE must be
capable of sustaining the peak crush loads from either the CEM zones or the existing end
structure without collapsing. Both options require a significant engineering effort, including
detailed linear and non-linear finite element analyses of the structures in order to determine the
loads generated during a collision. The optimum location for the CEM zone would be forward of
the cab area, configured with minimal effect on visibility. One possible concept could be two
tube-like elements with a plunger arrangement that compresses a crushable element designed to
achieve the desired force level. Tubes approximately 25 to 28 in. long could be integrated into
the end sill structure.
The CEM concept also has technical risks and difficulties that must be considered:
• There are significant engineering costs and a long lead-time for complex analyses and
testing.
• The CEM elements are custom designs that could prove difficult to develop.
• There are significant labor costs to retrofit the existing structure.
• Depending on the design, the weight savings may be lower than anticipated. It is
estimated that the weight of the LFE–LRV combination designed using the 2-g approach can be
approximately 5,000 lbs greater than the CEM approach.
The use of couplers with the highest possible energy absorption level significantly aid in
mitigating a collision, regardless of the design approach selected.
Finally, as part of the specification development for a CEM zone and the LFE, many
other structural aspects of the vehicle must be carefully reviewed to determine if strength levels
are adequate and consistent with the CEM design.
Light Rail Vehicles’ Low-Floor Center Section

G. JAMES MORGAN
LTK Engineering Services
T he issue addressed is how to increase capacity and update passenger accessibility to keep pace
with the “low-floor” trend in the transit industry, without making obsolete an existing fleet.
The benefits of adding a new, low-floor center section to an existing light rail vehicle is
discussed, and various issues and characteristics that must be considered and resolved before
proceeding are identified, including, but not limited to: size of vehicle, aesthetic appearance,
performance, propulsion and braking systems, all other onboard systems, structural strength, size
and length of station platforms, interface with station platforms, and maintenance shop and other
wayside facilities.
It is concluded that with good base line conditions, the addition of a low-floor center
section to an existing vehicle allows a transit agency to upgrade capacity, facilitate passenger
access to the rail vehicles, and transition toward the “low-floor” trend in the transit industry
without having to either sell or scrap a relatively new, existing fleet or be faced with operating a
mixed fleet of existing high-floor vehicles and new low-floor vehicles.
INTRODUCTION
From their inception, rail transit systems have continually expanded, improved service, enhanced
availability, and increased the level of comfort for passengers. One of the latest developments in
this continuing evolution is the ability to provide direct, level boarding for light rail vehicles
(LRVs). Previously, this type of passenger accessibility was an exclusive characteristic of heavy
rail systems using high platforms. High platforms are inherently incompatible in city streets with
mixed pedestrian and automobile traffic. The new designs of LRVs address and partially mitigate
this problem with floor levels that are approximately 14 in. above the top of rail (TOR). The low-
floor rail vehicle, combined with a matching station platform, can be accepted into the streets of
a city with relatively minor inconveniences. The low-floor, level boarding mode of operation
greatly enhances accessibility for all passengers, reduces station dwell times and facilitates
overall operation of the system. Also, level boarding accessibility allows the elimination of the
special equipment and facilities, both on the vehicles and at the stations that are required for
compliance with Americans with Disabilities Act (ADA).
DEFINITION OF THE PROBLEM
Providing level boarding for all passengers with the low-floor vehicle configuration is a very
attractive feature. For a new light rail transit system there is almost no decision to make. The
design of the new stations, maintenance facility, and other wayside facilities can be matched to
the dimensions and configuration of the rail vehicles and the overall low-floor, level boarding
104
Morgan 105
concept can be readily incorporated into a new system. For an existing light rail system that is
either expanding or upgrading, the decision to change to level boarding configuration is much
more difficult. For an existing system there are two choices:
• Continue to operate and expand or upgrade the system, “as is;” or

• Convert to low-floor, level boarding configuration and modify all existing stations
and facilities as required.
From a maintenance and operations point of view, continuing to operate “as is” is the
path of least resistance. The agency simply buys more vehicles with the same dimensions and
characteristics and continues to operate them in the same manner, using the same stations and
maintenance facilities. However, the general ridership, ADA community, local and federal
government agencies, combined with the transit agency’s inherent desire to improve the system
may influence the decision to convert to low-floor, level boarding configuration. Regardless of
these influences, the transit agency must consider the financial impact of the modifications and
changes that will be necessary to convert to level boarding operation and the disruption caused
by the implementation of these changes. Converting to low-floor, level boarding configuration is
difficult and expensive. Changing an existing, high-floor system to the low-floor, level boarding
mode will require
• A new fleet of low-floor rail vehicles;

• Modifications to all existing passenger stations to provide level boarding onto the
new low-floor vehicles;
• Modifications to maintenance facilities to provide for full maintenance access to the
equipment installed on the roof of the new low-floor cars; and
• Modifications to any other wayside facilities that are affected by the change in
characteristics of the low-floor vehicle.
Additionally, the disposition of the existing fleet of rail vehicles has to be determined. If
the vehicles have reached the end of their useful or economic life they can be retired (scrapped),
and new vehicles can be purchased as part of the upgrade. Selling the existing fleet of vehicles to
another property—while sounding like an easy solution—requires a willing buyer with a
compatible system and available funds. With other transit systems also considering and trying to
convert to low-floor operation, the sale of a used, high-floor vehicle becomes improbable.
Somehow the existing high-floor and new low-floor vehicles have to be integrated
together while still providing the desired, full-time, level boarding accessibility for the
passengers. The most direct solution is to establish operational procedures for a mixed fleet of
vehicles, requiring that all single car trains and at least one car in any multicar train is a low-floor
vehicle. Obviously, these operational restrictions can be implemented, but there will be an
adjustment period for the passengers and a continuing complication in the operational and
maintenance procedures, to maintain the criteria.
If the decision is to continue “as is,” the course of action to expand is clear: follow the
general criteria and guidelines that have been previously established and used for construction of
the existing system.
106 Transportation Research Circular E-C058:9th National Light Rail Transit Conference
The premise addressed in this paper is that an existing transit agency makes a decision to
convert to an overall low-floor, level boarding mode of operation as part of an upgrade or
expansion of the system.
The apparent course of action is to define and procure a new low-floor vehicle, modify
the stations and facilities to match the new vehicle and establish procedures to operate a mixed
fleet of high-floor and low-floor vehicles. However, depending on the characteristics of the
existing fleet, another option is available for converting to low-floor, level boarding operation.
Insert a new, low-floor center section into the existing, articulated LRV.
The following discussion will explore and analyze some of the more significant issues
that must be addressed for each aspect of the system that will be affected by the low-floor, level
boarding configuration and compare the differences between the low-floor vehicle and insertion
of a low-floor section into an existing vehicle.
REAL WORLD SCENARIO
Many agencies have discussed the possibilities of incorporating a new center section into an
existing vehicle configuration. Until now in North America, this has been an intellectual
exercise. This paper will discuss the design, review, engineering process, and decisions leading
up to the implementation of an actual test program of a low-floor center section by the Dallas
Area Rapid Transit (DART).
DART is the newest and fastest growing, fully operational, high speed (65 mph), LRT
system in the United States:
• Design and construction of the DART Starter System was initiated in 1990.
• DART issued the first contract for the procurement of LRVs in November 1991.
• Revenue service on the Starter System opened in June 1996 with 21 mi of track and
40 LRVs.
• In November 1997, DART issued a contract modification to procure 34 additional
LRVs to supplement the existing fleet and prepare for Build-Out Phase I of the North East
Corridor.
• In May 1998, DART issued a new contract for the procurement of 21 LRVs to
support service on the Build-Out Phase I of the North Central Corridor.
• At the end of 2002, DART completed the 23-mi, Build-Out Phase I expansion and is
now operating a 44-mi system with 95 LRVs.
In less than 10 years DART has doubled the size of the system and the fleet of rail
vehicles. Continuing this rapid expansion, DART has initiated design and construction for Build-
Out Phase II, which is scheduled for completion by 2012. Build-Out Phase II will again double
the size of the system by adding approximately 44 mi of double track and increasing the fleet by
at least 100 LRVs.
As part of the Phase II expansion, DART has decided to change to the low-floor, level
boarding, mode of operation.
The current fleet of DART LRVs consists of “high-floor” vehicles. Passengers board
from an 8 in. platform, step up 8 in. to the vehicle, and then there are three more 8 in. steps up to
the passenger floor that is 40 in. above TOR. ADA compliance is provided with high blocks
located at the end of the platform combined with a trap/bridge arrangement on the vehicle. The
Morgan 107
DART LRVs are all relatively new and hence were not considered for retirement (scrapping).
Selling the vehicles to another property was recognized as an unrealistic course of action. To
convert to level boarding operation DART either has to procure new low-floor cars and operate a
mixed fleet of high-floor and low-floor vehicles or insert a low-floor center section into the
existing vehicles. (As of August 2003, DART has not made a final decision with respect to the
type of vehicle that will be used for future expansion fleets. The following information is a
summary of activities to date. DART’s final decision will be based on this information, any new
factors that may be identified, discussion with DART member cities, local community and
special interest groups, and coordination with FTA.)
The (strong) possibility of a conversion to low-floor, level boarding operation was
recognized very early in the Build-Out Phase II program. Preliminary, internal reviews and
discussions were initiated to determine if a typical low-floor vehicle could be procured to meet
DART’s criteria of appearance, operational requirements and performance characteristics.
Alternative courses of action were explored as well. Some of the salient characteristics that
DART did not want to loose in the conversion to low-floor, level boarding operation were
• “Signature” appearance of the DART LRV,

• Acceleration and braking performance (3.0 mphps), and
• 65 mph operating speed of the rail vehicles.
Vehicles
DART has an existing fleet of 95 high-floor LRVs and is in the process of exercising an option
for 20 additional high-floor vehicles, all from the same manufacturer and of the same design.
The overall age of the fleet is relatively low. The 40 oldest vehicles started service in 1995–1996.
The 55 newer vehicles were put into service in 2000–2001. The 20 additional LRVs are
scheduled for delivery in 2005. In general, the fleet is considered to be reliable and is well
accepted by the public. The specified design life of the vehicles is 30 years. As discussed
previously, the ideas of retiring or selling the fleet were discarded early in the studies. Therefore,
the existing fleet of vehicles has to be incorporated into the new level boarding mode of
operation, until they reach the end of their economic life. There are two ways of accomplishing
this:
• Operate a mixed fleet of vehicles consisting of the existing fleet of DART LRVs
combined with new low-floor vehicles; or
• Insert a low-floor center section into the existing vehicles and procure additional
vehicles to meet the desired fleet size.
Low-Floor Light Rail Vehicles
Low-floor light rail vehicles (LFLRVs) are considered to be established, proven technology.
They are in successful operation on numerous properties in Europe and in North America. At the
present time there is no LFLRV in service that is designed for 65-mph operation. With the
exception of Metro Rail in Houston, Texas, all of the LFLRVs in operation are designed for 55-
mph operation. The LFLRV that is being procured for Houston, Texas, is specified to have an
operating speed of 66 mph. Dynamic testing of this vehicle is scheduled to begin in 2003.
Because of the established, acceptable, history of operation, a detailed review of the

characteristics of a typical LFLRV was not pursued. LFLRVs are considered to be an acceptable
alternative for the DART system and Phase II expansion, with the qualification that an operating
speed of 65 mph can be achieved in normal revenue operation.
Low-floor Center Section
Center section inserts are a newer, less established, technology that have not been employed in
North America. DART undertook an in-depth investigation to determine if the use of a center
section is a viable alternative that could be used as part of the Phase II expansion to both increase
capacity and enhance overall passenger accessibility.
Several transit properties in Europe have inserted center sections into existing vehicles to
increase capacity. In general, all of these applications have been low-speed (35 mph to 45 mph)
applications. With the exception of the prototype operating in Dallas, Texas, a center section
insert has not been used for revenue service in North America. A center section can be installed
in any articulated vehicle. The age of the vehicle and the technology of equipment used on the
vehicle are the determining factors in the decision for installing a center section. If the vehicle is
near the end of its useful life it may not be reasonable to invest a large amount of money in
upgrades. If the propulsion, braking, or auxiliary equipment cannot support the weight increase
and power requirements of the center section, the vehicle would not be acceptable for operations.
Another factor to be considered with the integration of a center section into an existing
articulated vehicle is the overall structural strength. The DART LRV was designed to meet a
“2g” requirement. That is, the buff load (end compression strength) of the car-body structure is
designed to be twice the empty weight of the completed vehicle in the ready to run configuration.
The value of 2g is an historic value selected to represent a car-body strength high enough
to withstand minor to intermediate collisions without basic structural damage, but low enough to
minimize the potential for passenger injury. The car body must demonstrate both 2g end strength
and the ability to crumple from the ends in a controlled manner when collision energy exceeds
the structure’s strength. This “cushions the blow” of the collision, providing a compromise
between two factors that cause injury in collisions: 1) loss of passenger volume (crushing of the
structure), by providing a reasonable level of static strength; and 2) controlled crush behavior
when the car-body strength is exceeded. The ratio of “actual” weight of the DART LRV
(107,000 lb) and “actual” structural buff test load (227,000 lb) is 2.12 g. Compared to similar
articulated LRVs, the DART LRV is stronger. This characteristic makes the addition of the
weight of the center section much more viable.
The DART LRV is both relatively new, and the propulsion, braking, and auxiliary
systems are relatively current technology. These characteristics make the introduction of a low-
floor center section into an existing vehicle—while still maintaining similar performance,
operational characteristics and a top speed of 65 mph—a viable possibility.
The general technical issues of car-body strength, performance, speed, and so forth,
related to the incorporation of a low-floor center section into the DART LRV were reviewed and
the results were found to be favorable. The concept that was further developed was to maximize
the passenger capacity of the low-floor center section while making the minimum number of
changes to the existing vehicle. With no increase in propulsion, the additional weight of the
center section would reduce overall acceleration of the modified vehicle to approximately 2.3
mphps. However, the operating speed of 65 mph would be maintained. The addition of a second
Morgan 109
trailer truck with 4-disc brakes would compensate for the extra weight and maintain the 3.0
mphps deceleration rate. Basically, the concept was to separate an existing LRV at the
articulation joint and insert a low-floor center section and a second trailer truck between the two
existing sections, creating a 3-section, double-articulated, 4-truck, 8-axle, “super” light rail
vehicle (Super-LRV).
Structure
The structure of the DART low-floor center section is designed to be approximately 5% stronger
than the existing LRV structure. This makes the center section uniform and consistent with the
existing car-body structure and the philosophy of crash worthiness (controlled deformation
starting at the ends of the vehicle) that is included in the basic design. The addition of the center
section with an actual weight of 31,000 lb changes the weight/strength ratio of the DART Super-
LRV to 1.64 g. This is equivalent to the older Type I and the new “Type II” low-floor vehicles
currently in service in Portland, Oregon. The actual structural strength, 227,000 lb, and crash-
worthiness characteristics of the DART car body structure remain unchanged.
Dimensions and Performance
The height, width, and appearance of the low-floor center section were matched to the existing
car. The floor height of the low-floor area was selected to be 16 in. above TOR so that it would
match the first step in the stairwells of the existing car. This is 2 in. higher than the typical 14 in.
floor height of a low-floor car. While the low-floor center section could have been built to 14 in.
above TOR, the 16 in. height was selected to eliminate the trip hazard with 14 in. platforms at
stairwells of the existing doors, which are at 16 in. Truck spacing of 31 ft for the center section
was selected to match the existing truck-to-truck distance to maximize capacity without changing
the static and dynamic envelope, with the exception of the increase in overall length from 92 ft, 8
in. to 123 ft, 8 in. The center section was designed with articulation joint assemblies and
electrical boxes for car line and train line signals that would directly match up with the respective
A and B sections of the DART LRV. The second trailer truck and bolster is the same as the
existing trailer truck.
Using the same articulation and center trucks has several advantages and a disadvantage.
The advantages are that no new equipment is added to the vehicle and the use of a traditional
truck configuration, with continuous, straight axles and full size, inboard, disc brakes (as
compared to a typical low-floor center truck with stub axles and wheel or outboard mounted
brake assemblies). The traditional center truck configuration is considered an essential element
for maintaining a 65-mph operating speed and a 3.0-mphps deceleration rate of the Super-LRV.
The disadvantage is that the area of the low-floor portion is limited to the distance between the
center trucks.
Equipment
The capacity of the existing on-board air supply and friction braking systems was determined to
be sufficient to support operation of the friction brakes and air spring on the additional truck and
the door operators on the center section. The capacity of the low-voltage system was determined
to be sufficient to operate all control functions of the center section and the track brakes on the
additional truck. It was determined that an additional heating, ventilation, and air cooling
(HVAC) unit would be required to be installed on the roof of the center section along with a new
10 kW static inverter to provide power for the HVAC unit and interior lightning. To the extent
possible, existing components, equipment, and systems were incorporated into the design of the
center section. The door operators, door panels, trailer truck, brakes, HVAC unit, interior lights,
passenger seats, passenger side windows, portions of the interior lining, and other miscellaneous
parts are identical to those used on the existing car. The operator’s cab and all car and train line
control functions remain unchanged. The only new or different operating component is a 10 kW
static inverter that is used to power the HVAC unit on the center section. Insertion of a
completed low-floor center section into a DART LRV was planned to be a simple process of
unbolting, disconnecting, and separating a LRV, inserting a center section and trailer truck, and
then bolting and connecting everything back together—the ultimate plug and play modification.
Implementation
As with all designs, the Super-LRV met all of the criteria and looked great on paper. The next
logical step was to prove the concept in a real world application. As a joint research and
development effort DART coordinated with Kinkisharyo International, LLC (manufacturer of
the DART LRV), to build a complete, fully operational, low-floor center section, install it in a
DART LRV, and test it on the DART system to confirm the results of the studies and reviews
that had been performed. The basic agreement was that DART would supply equipment and
materials from existing stores for installation on the center section and Kinkisharyo would
provide the raw material, coordination, manufacturing, assembly, testing, and transportation of
the center section. After inserting the center section into a vehicle, DART and Kinkisharyo
would jointly conduct dynamic testing of the Super-LRV to verify the dynamic capabilities and
performance. After this, the Super-LRV would be released for revenue service to verify
operational acceptability and to gauge the reception by the ridership.
Manufacturing and Testing
The low-floor center section was manufactured at the Kinkisharyo manufacturing facility in
Osaka, Japan, in 2001 and delivered to the DART maintenance facility in February 2002. The
center section was installed in LRV 170 in May 2002 creating the first, fully-operational, double-
articulated, 65-mph, Super-LRV. All of the planned static and dynamic tests have been
successfully completed. The initial acceleration rate of the DART Super-LRV is 2.4 mphps,
which is a little higher than expected. The Full Service Braking rate, 3.0 mphps, is the same as
the existing LRV. Operating speed, 65 mph, remains the same. After completion of dynamic
testing the Super-LRV was operated extensively on the DART system as an “out of service”
train. Unofficial reports from DART Operations and Maintenance staff indicate the vehicle
performed well and was well received by all. Currently, Super-LRV 170 is being used in daily
revenue service. The only operational restriction is that the Super-LRV cannot operate as part of
a 3-car consist because the resulting train (two standard DART LRVs and one Super-LRV) is too
long for some of the existing station platforms
Morgan 111
Conclusion
Because of the extensive review that was conducted and actual field-testing of a DART vehicle
on the DART system, the addition of a low-floor center section is considered to be a viable
alternative for further consideration during the Phase II Expansion plans, without qualifications.
Compatibility with the Rest of the System
Parallel with the review and design of the Super-LRV, the operational issues and interfaces with
other systems (electrification, communications, and signals) and wayside facilities (stations and
maintenance shop) were also considered. Each of these areas was carefully reviewed to
determine what requirements needed to be incorporated into the design of the new line sections
in Build-Out Phase II and what modifications needed to be implemented in the existing line
sections.
Wayside Facilities
Length of the station platforms was identified as the primary limiting factor for the use of the
longer vehicle with the center section. A LFLRV of approximately the same size and length as
the current DART LRV will fit into the existing stations and any future plans for capacity
increases.
A Super-LRV, with a center section installed is 123 ft, 8 in. overall. This is 31 ft longer
than an existing LRV (92 ft, 8 in.). The most critical areas of station length are the 4 stations in
the downtown Dallas Central Business District (CBD), because they are constrained by cross
streets. The length of the station platforms that are in use in the CBD can accommodate a 2-car
consist of Super-LRVs without modification. In terms of passenger seating this is equivalent to
the 3-car trains that are being operated during rush hours.
In the future, when increased capacity requires the use of 4-car trains, all of the station
platforms will have to be modified to increase platform length by approximately 100 ft. The
original design of the Starter System included space for this expansion at all stations, including
the four stations in the CBD area. In this event, a 3-car, Super-LRV consist, which has a higher
seating capacity, can be used in place of a normal 4-car LRV train.
Wayside Systems: Signal System, Grade Crossings, Traction Electrification System,

Overhead Catenary System
If a new LFLRV is selected and specified to have the same operating and performance
characteristics as the current LRV, there would be no change in the interfaces with any of the
operating wayside systems; Signals, grade crossings, train electrification system (TES) and
overhead catenary system (OCS).
Although the Super-LRV has a lower acceleration rate, it has the same braking rate as the
current vehicle, so no change would be required for the Signal System block length or the grade
crossing approach circuits. With the Super-LRV, each vehicle has the same propulsion system
and will have the same maximum current limit as the existing vehicle. However, there will be
fewer cars operating for the same number of passenger seats that are in service, resulting in a net
reduction in power consumption (cost) and general wear on the OCS.
Operations
A new LFLRV would be specified to have the same level of performance and speed, as the
existing DART vehicle, and hence would have no impact on the current operation.
Because of the extra weight of the center sections in the Super-LRV, it will have a lower
initial acceleration rate. However, it has been determined that the existing schedules could be
maintained on the existing system. Simulations using the lower acceleration rate of the Super-
LRV indicate that the worst case run from terminal to terminal results in an increase in the
running time of approximately 2 min on one of the lines. Detailed review of the simulations
indicates that the majority of this time is lost in central areas of the system where station spacing
is reduced to approximately 1 mi and even less in the CBD. This run time increase is considered
to be acceptable in the overall schedule of operations and is absorbed in the turn times at the end
of the lines. As the DART system expands and station spacing increases in the outer areas, the
effect of the lower initial acceleration on overall run time is further reduced.
Capacity
A new LFLRV would have equivalent seating capacity as the current vehicle, as shown in Table 1.
As part of the modification to insert a center section into a DART LRV, two double-flip
seats (four seats) that were removed by DART to enhance ADA accessibility will be reinstalled,
and four new single seats will be installed. Combined with the 24 seats in the center section, the
seating capacity of a Super-LRV is 104. A 2-car, Super-LRV consist can be directly substituted
for a 3-car, LRV consist without modification to platform length and provide nearly equivalent
passenger capacity. Table 1 lists a summary comparison of the seating capacity of various train
configurations.
Maintenance
A new LFLRV would have a high percentage of new components, equipment, and systems.
Operations and Maintenance staff would have to be trained on both a new LFLRV and the
existing vehicles. Spare parts would have to be stocked for both types of vehicles.
The introduction of a Super-LRV into the DART system has no change in operations and
effectively no change to the maintenance procedures, spare parts, and training that are required to
keep the Super-LRV in revenue service. With the exception of the new static inverter installed on
the roof of the C-car, all of the operating equipment is the same, with complete
interchangeability, as the equipment used on the current vehicle.
Service and Inspection Facility
The Service and Inspection (S&I)facility will have to be modified regardless of the type of low-
floor vehicle that is selected. For any low-floor vehicle, a second maintenance level would have
to be installed at roof height to provide for full maintenance access to the roof mounted
equipment. The Super-LRV is 31 ft longer, which requires that the maintenance pits be extended,
and the in-floor hoists modified to allow access to the equipment.
Morgan 113
TABLE 1 Comparison of Seating Capacity by Car Consist
DART LRV LFLRV Super-LRV

(# Seats) (# Seats) (# Seats)
1 Car 72 76 104
2 Car 144 152 208
3 Car 216 228 312
4 Car 288 304 ----
Note: In some hours of intermediate, off-peak operation, when passenger volume is lower, it will be possible to
operate a single Super-LRV in place of a 2-car normal LRV train.
CONCLUSION
A new LFLRV can be incorporated into DART’s plans for system expansion and is considered
to be a readily available alternative for the planning and implementation for Phase II expansion
and future increases in capacity—with the qualification that a 65-mph operational speed can be
provided.
Review of a center section concept and application including actual manufacture and
testing of a low-floor center section shows that the Super-LRV concept is also a viable and now
proven course of action to support the Phase II expansion and future increases in capacity,
without qualification, and has additional benefits when compared to a new LFLRV.
DART BUILD-OUT PHASE II
With the successful completion of the joint development effort of the Super-LRV, DART had
two general vehicle designs, LFLRV or Super-LRV, either of which could be used as the basis
for conversion to low-floor, level boarding operation, expansion of the system and future
increases in capacity for the entire system. Using the two possible vehicle configurations as
starting points, DART next began to review the existing lines and the planned expansions to
determine what features must be designed into the new extensions and facilities and what
modifications would need to be incorporated into the existing lines and facilities in order to
convert to low-floor, level boarding operation.
Currently DART operates two lines with 3-car consists during peak hours. The two lines
start and end in the Northern and Southern urban areas of Dallas and merge together to go
through the CBD creating 4 “branches” (two to the North and two to the South) that feed into the
CBD. Build-Out Phase II will add three more branches (two to the North and one to the South),
also feeding into the CBD.
All of the existing stations can accommodate 3-car trains. When required to increase
capacity the length of all of the existing station platforms can be increased to accommodate 4-car
trains. Based on the enthusiastic public acceptance of the DART Light Rail System and a
ridership in early 2003 of approximately 66,000 passengers per day, DART expects to increase
capacity in the near future. In anticipation of this requirement all of the new stations in Build-Out
Phase II will be built to accommodate 4-car trains. Also, in anticipation of the conversion to level
boarding, the new platforms will be built at 16 in. above TOR to be compatible with either of the
two low-floor vehicle designs.
The new vehicles that DART will purchase as part of the Phase II expansion will have to
be operationally compatible with the existing fleet and will have to serve as the basis for level
boarding operation.
The advantages and disadvantages of several different fleet configurations were reviewed
to determine the best course of action. The existing DART LRV was used as the base line criteria
for comparisons with other fleet configurations. Current plans are that DART will procure 20
LRVs to supplement the existing fleet for a total of 115 LRVs. For Phase II Build-Out an
additional 100 LRVs (estimated) will potentially be procured. This will result in a total fleet size
of 215 cars (95 + 20 + 100 = 215 cars) with 15,480 passenger seats (215 × 72 = 15,480 seats).
The total number of passenger seats is the characteristic that allows direct comparison of the
different fleet configurations. Of the various fleet combinations that were reviewed, three stand
out as being potentially acceptable:
• Base Fleet—Maintain the existing fleet of 115 LRVs (72 seats) and procure 100
additional LRVs (72 seats). Total 215 cars (uniform fleet, high-floor cars) 15,480 seats.
• Option 1—Maintain the existing fleet of 115 high-floor LRVs (72 seats) and procure
a new fleet of approximately 95 new low-floor LRVs (LFLRV) (76 seats). Total 210 cars (mixed
fleet), 15,500 seats.
• Option 2—Procure 114 low-floor center sections to be installed in the existing fleet of
vehicles (104 seats) and procure approximately 47 new LFLRVs (76 seats). Total 162 cars
(mixed fleet), 15,532 seats.
• Option 3—Procure 114 low-floor center sections to be installed in the existing fleet of
vehicles (104 seats) and procure approximately 34 new Super-LRVs that are the same as the
existing fleet with the low-floor center section included (104 seats). Total 149 cars (uniform
fleet), 15,496 seats.
All calculations that resulted in a “fraction” of a vehicle were rounded up to the next
“whole” vehicle with the corresponding number of seats. It should also be noted that DART has
procured one low-floor center section and installed it in LRV 170. In the fleet size calculations
above and the cost estimations following, the numbers have been adjusted to reflect that DART
already has one fully operational, Super-LRV that is ready for service.
Changes to existing wayside facilities and stations to take full advantage of the low-floor
configuration will be required regardless of the course of action. The actual configuration of the
vehicles will determine the modifications necessary at all existing facilities and the design and
construction of all new facilities. A summary discussion of the various elements that will be
affected and a comparison of some of the cost impacts is presented below.
COST—THE DRIVING FACTOR
As a starting point for the Build-Out Phase II studies, cost was used as the primary consideration
in DART’s review and decision-making process. In a special situation another, more important,
factor may override cost as the determining factor. These special situations are addressed and
resolved on a case-by-case basis.
Morgan 115
After the real estate, the primary capital asset of an existing system, and the primary
capital cost for expansion of the system, are the rail vehicles. Additionally the manufacturing
lead time for the new rail vehicles is approximately two years from award of contract. Decisions
with respect to these two major cost drivers, the existing fleet and the new vehicles, must be
established as early as possible in the program, because the modification of all existing facilities
and systems and design of all new facilities and systems will be driven by the basic decisions in
these two areas. The questions addressed were
• What will be done with the fleet of existing vehicles to incorporate them into the low-
floor, level boarding mode of operation?
• What type of new vehicles will be procured to provide service on the expansion
portion of the system?
• What modifications will be required for existing stations and facilities?
Vehicles
Using the fleet sizes noted above, a rough order of magnitude (ROM) cost for the different fleet
configurations with equal passenger seat capacities could be determined. A summary review of
recent LRV procurements indicates that a ROM price for a new, LFLRV is $3 million. The ROM
cost of a low-floor center section is assigned at $1 million. Using these values the ROM fleet
costs for the “base fleet” and the 3 options above are as follows:
• Base Fleet (115 existing LRVs plus 100 new LRVs): $300 million.
• Option 1 (115 existing LRVs plus 95 new LFLRVs): $285 million.
• Option 2 (114 new low-floor center sections plus 47 new LFLRVs): $255 million.
• Option 3 (114 new low-floor center sections plus 34 new Super-LRVs): $250 million.
Station Modifications
To maximize the accessibility benefits of the low-floor vehicles the station platforms must be
designed to be directly compatible with the selected vehicle. All new Phase II station platforms
will be built so that the entire platform will provide level boarding with the vehicle regardless of
the configuration. The configuration of the vehicle determines the extent of the modifications
that are necessary for the existing platforms.
If a LFLRV is selected, the entire length of the platform for each of the existing stations
must be raised 8 in. to guarantee that a passenger that boards at a new station (with full length,
level board, platforms) is not stranded at an existing station. The implications and complications
of this kind of modification are a significant cost factor because of the numerous interfaces with
other wayside facilities (doors, elevators, escalators, stairways, columns, seating, ticket vending
machines, landscaping, etc.). Almost everything on the platform must be removed, the platform
must be raised 8 in., and all equipment and facilities must be reinstalled. There would be a
significant disruption of service as each station is closed for this type of modification.
If a Super-LRV is selected, special use platforms (SUP) can be used to modify the
existing stations. SUPs are essentially “humps” or localized sections of the platforms that are
raised 8 in. so that the door of each low-floor center section lines up with the SUPs when the
train stops. There would be relatively minor interface issues associated with the installation of
SUPs on each station platform. This greatly reduces the length of the disruption at each of the
stations when this modification is made. To minimize conversion time, confusion, and effect on
the passengers during the change over to level boarding operation, temporary or semi-permanent
segments of the SUPs could be prefabricated off site and then placed into position on the
platforms, until the final configuration SUPs have been installed. Regardless of the use of SUPs,
some of the existing platforms will be totally raised 8 in. because of site specific complications
and restrictions. (If SUPs were installed for each door of an LFLRV they would overlap each
other, which is equivalent to raising the entire platform.)
The cost difference between raising all existing platforms 8 in. for use with LFLRVs
versus installing two SUPs on each platform for use with the Super-LRV, for each of the existing
stations, is estimated to be approximately $15 million.
Maintenance Facility Modifications
Regardless of the type of fleet that is selected, modifications to the existing S&I facility will be
required.
If a LFLRV is selected, with the majority of equipment installed on the roof, the inspection and
maintenance areas will have to be modified to provide for full maintenance access to the roof of the
vehicle. Essentially a second floor has to be installed at the roof level of the low-floor vehicles.
Estimated costs for this are approximately $1 million.
The equipment location of the Super-LRV is the same as the existing vehicle; however, the
Super-LRV is 31 ft longer. The maintenance pits will have to be extended and the in-floor hoists will
have to be modified to accommodate the longer vehicles. Estimated costs for modifications to the S&I
for the Super-LRV are approximately $4 million.
The new maintenance facility that is included in Phase II expansion will be built to
accommodate the type(s) of vehicles that are selected.
Fleet Operation
Currently DART operates 1-, 2-, or 3-car consists, depending on the time of day. When system
capacity is increased, 4-car consists will be used. If LFLRVs are used to increase the size of the fleet,
these configurations will continue without change.
If all Super-LRVs are procured the size of the consists at various times of the day will change.
The capacity of two Super-LRVs is equivalent to three LRVs. Therefore, 3-car trains can be replaced
with 2-car Super-LRV trains. In the future, 4-car LRV trains can be replaced with 3-car Super-LRV
trains. Additionally, a portion of 2-car trains used in intermediate, off-peak hours can be replaced with a
single Super-LRV. A single Super-LRV costs the same to operate as a single LRV and costs less to
operate than two LRVs. Two Super-LRVs cost less to operate than three LRVs.
Based on the calculated power consumption and the smaller train consists that are possible with
the Super-LRV, it is estimated that DART could realize energy savings of approximately $500,000 per
year on the existing system if Super-LRVs are used, as compared to the DART LRV. Extending this
projection to include the Phase II expansion, which, approximately, doubles the size of the system, it is
estimated that a savings of $1 million per year could be realized when Phase II is opened for service.
Morgan 117
Fleet Maintenance
Standardization of a fleet of rail vehicles is an intangible but significant benefit to the Maintenance and
Operations Department. Operation of a mixed fleet of low-floor and high-floor vehicles with different
components and parts will be the source of continuing difficulties for the life of the vehicles. All
maintenance staff will be required to be trained on and be familiar with both types of vehicles. The
storerooms will have to stock sufficient quantities of separate spare parts for the different types of
vehicles.
A standardized fleet of Super-LRVs eliminates these difficulties. Another benefit is having
direct interchangeability with all of DART’s existing spare parts, effectively reducing the overall
capital value of spare parts in inventory. Additionally, the overall quantity and value of spare parts will
be reduced, because a fleet of Super-LRVs is approximately 30% smaller than an equivalent fleet of
high and low-floor vehicles. While standardization of the fleet is recognized as a benefit both to
Operations and Maintenance, no cost has been estimated to reflect this (real) value.
Another maintenance benefit associated with Super-LRVs is that there will be a net reduction
in preventative maintenance costs for the fleet. If Super-LRVs are selected there will be 30% fewer
vehicles to be maintained for the same number of passenger seats. Because there is physically more
equipment installed on a Super-LRV (1 HVAC unit, 1 center truck, 1 static inverter) it is estimated
there will be a 20% reduction of the man hours necessary for normal preventative maintenance
activities for the entire fleet of Super-LRVs as compared to an equivalent fleet of mixed high and low-
floor vehicles. Based on current procedures, this represents an approximate savings for scheduled,
preventative maintenance activities of approximately $680,000 per year. In the longer term there will
be a reduction in costs for rebuild or overhaul of equipment because there is less equipment to be
maintained.
SUMMARY
To be able to address the continued enthusiastic reception and demands for expansion of service and a
desire to continue to enhance overall accessibility of the LRT System, DART is in a unique position.
The option of introducing a low-floor vehicle into service, with the associated changes in infrastructure,
maintenance, and operations is available to all transit agencies. Inserting a center section into an
existing vehicle is also possible for all agencies. Inserting a center section into a modern vehicle and
maintaining essentially the same performance and 65-mph operating speed is an option that is only
available to DART. The results of the static and dynamic testing that were performed confirm that the
insertion of a low-floor center section into the DART LRV is a reasonable method of achieving a low-
floor, level boarding configuration and reducing cost, while maintaining
• Schedule performance,
• 65-mph operating speed,
• Standardization of the fleet of vehicles, and
• Signature appearance of the DART rail vehicle.
A summary comparison of the cost impacts of the various aspects of the DART System,
equipment, and facilities identified in the body of the report is presented in Table 2.
TABLE 2 Comparison of Options for DART Fleet Combinations
Base Fleet Option 1 Option 2 Option 3

115 LRV 115 LRV 115 Super-LRV 149 Super-
100 LRV 95 LFLRV 47 LFLRV LRV
Fleet Cost $300 million $285 million $255 million $250 million
Station Mods $25 million $25 million $10 million
Facilities Mods $1 million $5 million $4 million
*Savings Fleet No Change No Change Slight reduction *$1 million per
Power year
*Savings Fleet No Change No Change Slight Reduction *$680,000
Maintenance per year
Fleet Configuration Uniform Mixed Mixed Uniform
(High-floor)
The Base Fleet simply maintains the current configuration with high-floor vehicles. Fleet
Options 1 and 2 both result in the incorporation of a new low-floor vehicle, potentially from a
different manufacturer, into the DART fleet with the associated changes in operations and
maintenance. Option 3 essentially has no changes in the maintenance and operation of the fleet.
All transit systems and applications have different criteria and emphasis on different
individual characteristics. In this comparison the selection of the Super-LRV will achieve the
desired goal of a low-floor, level boarding mode of operation and has a clear cost advantage both
in initial procurement and future operations and maintenance, with no significant operational or
interface disadvantages. In another situation with the emphasis on a different aspect or
requirement, a similar study may result in a different indication.
Integrating New Light Rail Vehicle

Technology into Mature Infrastructure
GAVIN R. FRASER
Booz Allen Hamilton
RICHARD J. LEARY
Massachusetts Bay Transportation Authority
MASSIMO MARIANESCHI
CARLO PELLEGRINI
AnsaldoBreda
T he Massachusetts Bay Transportation Authority operates the oldest light rail system in North
America, known as the Green Line, with sections dating back more than 100 years. In order to
improve accessibility to the Green Line, in 1995, MBTA ordered 100 partial low-floor, double-
articulated light rail vehicles (LRVs) from AnsaldoBreda of Italy. These vehicles will operate in
consist with existing high-floor, single-articulated LRVs. Some of the challenges faced by the
project team are discussed, and some lessons learned that may be of value to other mature light rail
operations contemplating such procurements are highlighted.
Derailments of the leading axle of the center truck of the No. 8 Low-Floor Car have been
the most challenging aspect of the procurement. Following a comprehensive study of the vehicle
dynamics and an investigation of the effects of track quality on derailment performance, several
modifications were implemented to correct the problems. While some of the characteristics of the
Green Line are unique, the important lessons learned can be usefully applied to other procurements
involving the introduction of low-floor LRVs to systems with older infrastructure.
INTRODUCTION
The Massachusetts Bay Transportation Authority (MBTA) operates the most intensive light rail
service in the United States, over some of the oldest infrastructure to be found in the world. This
position as “grandfather” of the light rail industry means that the MBTA and its predecessor
agencies have frequently faced the need to introduce new technology into the existing
infrastructure. For the MBTA light rail system (the Green Line), vehicle technology has been an
area where there has been frequent technological change. The original streetcar lines were
developed for four-wheeled, horse-drawn trolley cars. The MBTA is currently deploying the
eleventh generation of electric trolley vehicles to operate over these same lines.
Boston’s transit providers have never been shy of introducing new technology. From the
early introduction of the first electric streetcars, through the President’s Conference Committee
(PCC) car era, to the (then) advanced Boeing Vertol Standard Light Rail Vehicle (SLRV) in the
119
1970s, Boston has often been at the leading edge of vehicle technology. This trend continues, with the
MBTA becoming the second North American property to order low-floor light rail vehicles (LFLRVs),
and the first “mature” U.S. light rail system to procure these vehicles.
Introducing new technology is not always easy, and the introduction of the low-floor car (LFC)
to Boston has been no exception. As this project has been covered at length in previous papers (1–4),
only a brief summary of the highlights is discussed below.
The procurement of LFCs was driven by two factors: the desire to make the Green Line service
accessible to all, in compliance with the Americans with Disabilities Act (ADA) requirements, and the
need to replace the aging and unreliable Boeing SLRV fleet. After extensive investigations and reviews
of available vehicle designs, the MBTA decided to procure a fleet of partial (70%) LFLRVs. These
vehicles were specified to occupy the same physical envelope as those that they will replace in order to
avoid the need for infrastructure modifications. The new vehicles were also required to be compatible
with the workhorse of the Green Line fleet, the Kinkisharyo-built No. 7 Surface Rail Car (SRC),
permitting operation of mixed consists to ensure there is at least one accessible vehicle per train.
In 1995, after a competitive procurement, MBTA awarded a contract to Breda Costruzioni
Ferroviarie of Italy (now AnsaldoBreda) for design and supply of 100 LFLRVs, to be known as the
No. 8 LFC. The contract also includes requirements for upgrading the 115 No. 7 SRCs to make these
cars compatible with the systems installed on the No. 8 LFC.
Unfortunately, technical problems were encountered that required the fleet to be withdrawn
from revenue service and which significantly delayed the project. These problems illustrate some of the
challenges faced when integrating new vehicle technology into an existing, and aging, light rail system.
SUMMARY OF THE NO. 8 LFC PROJECT
The Vehicle
The No. 8 LFC is a new design (Figure 1), although its design solutions and systems were proven on
other AnsaldoBreda products or on other LFLRVs. The three-section vehicle has an articulated frame
motorized truck at each end, with an independent wheel trailer truck (also an articulated frame) beneath
the center body section. The majority of equipment is roof mounted, including the IGBT propulsion
inverters. Braking is electro-hydraulic, with truck-frame mounted hydraulic pressure control units. Key
parameters are listed in Table 1.
Program Summary
An extensive prototype and development-testing program followed delivery of the first car to Boston in
early 1998. This program focused on areas such as vehicle clearance, dynamic performance, propulsion
and braking integration, and car monitoring systems. The first cars entered revenue service in March
1999. In the fall of that year, concerns with poor braking performance under low adhesion conditions
forced withdrawal from service, and an intensive slide control system investigation program began.
The fleet re-entered service on two further occasions, only to suffer a series of derailments, which
caused further withdrawals. After a very extensive investigation and corrective action program,
revenue service resumed in March 2003.
Fraser, Leary, Marianeschi, and Pellegrini 121
FIGURE 1 Prototype No. 8 LFC at riverside carhouse.
TABLE 1 No. 8 LFC Main Features
Type Bidirectional, double-articulated LRV

Configuration Bo’ 2’ Bo (2 motor trucks, 1 trailer truck)
Low Floor Area Approx. 70%
Track Gauge 4 ft 8.5 in. (1435 mm)
Minimum Curve Radius 42 ft (12.8 m)
Catenary Voltage 620 Vdc nominal
Length over Coupler Faces 74 ft (22. 6m)
Max. Overall Car Width 8 ft 8 in. (2.64 m)
Max. Height, Equipment Included 11 ft 10 in. (3.6 m)
Wheel Diameter (new) 28 in. (711 mm) motor, 26 in. (660 mm) trailer
Floor Height from TOR 14 in. (356 mm) Low Floor, 35 in. (889 mm) High Floor
Side Door Opening Width 50 in. (1270 mm)
Seated Passengers 44
Standing Passengers 77 @ AW2, 154 @ AW3
Empty Weight 86,000 lbs (39,090 kg)
Maximum Speed 55 miles/h (88 km/h)
Maximum Acceleration 2.8 mphps (1.24 m/s2)
Full Service Brake Rate 3.5 mphps (1.55 m/s2)
Emergency Brake Rate 6.0 mphps (2.66 m/s2)
Integration with Existing Vehicles
To operate in compliance with ADA regulations, at least one car per train must be “accessible.”
Given the continuing use of the No. 7 SRC fleet on the Green Line, this meant that the No. 8
LFC was to be operationally compatible with the No. 7 SRCs to operate as a two- or three-car
consist. For successful consist operation, an important goal is to closely match performance of
the two vehicle types in order to avoid uncomfortable and potentially damaging coupler action.
The challenge for the No. 8 LFC designers was to match the performance of the AC drive No. 8
LFC, with its fast acting hydraulic brakes, to the performance of the DC drive No. 7 SRC, with
its air brakes. The different propulsion characteristics and system reaction times of the two
vehicles required extensive fine-tuning of the control systems on the No. 8 LFC.
Slide Control Challenges
The performance demanded of the No. 8 LFC is quite high, particularly in braking. With the
streetcar nature of parts of the Green Line, the ability to brake a train rapidly and stop within
short distances is vital. The No. 7 SRC is equipped with track brakes on all trucks, but space
constraints on the No. 8 LFC prevented installation of track brakes on the center truck, which
placed this heavier car at a stopping-performance disadvantage. This situation was complicated
by the configuration of the center truck, which has four independently braked low-inertia wheels,
the sliding of any one of which could reduce effort on other trucks. By contrast, the No. 7 SRC
has a much simpler system, which did not react to some of the slippery conditions detected by
the No. 8 LFC and, as a result, stopped in a shorter distance.
Extensive testing and software changes eventually resulted in a design that offered
stopping performance equivalent to, or better than, the No. 7 SRC under all foreseeable
conditions (naturally occurring friction coefficients of as low as 0.043 were measured during
night testing along the tree-lined Highland Branch). Comparative testing with application of a
soap solution to the rails was used to confirm the equivalence of the stopping performance under
controlled conditions. It is testament to the ability of the control system that it is able to
overcome disadvantages of weight, low-inertia wheels, and less track brake effort compared with
the No. 7 SRC, and yet still deliver equivalent performance.
DERAILMENT OF INDEPENDENT WHEELED TRUCK
Derailment Incidents
Derailments of the center truck of the No. 8 LFC have been the most challenging aspect of the
project to date. The derailments started at a time when basic qualification testing (including ride
quality and stability) had been completed, and revenue service had commenced. Significant
mileage had been accumulated on the fleet in test and revenue service without incident and
deliveries were starting to ramp up following the suspension of service due to braking problems.
With the good equalizing properties of the articulated frame trucks, derailments were also
unexpected. The derailments started in April 2000, when final preparations were being made for
the fleet to return to revenue service. While the first two derailments were under investigation, two
further derailments occurred, at different locations and with different vehicles. Following an initial
investigation, three further derailments occurred. All derailments were unusual in that they
occurred on main line (rather than yard) track at higher speeds than is expected for derailments on
the Green Line. All incidents involved flange-climbing derailment of the leading axle on the
independent wheel center truck. In all cases, the vehicle was lightly loaded, resulting in the
minimum load on the center truck wheels.
Derailment Investigation Outline
A comprehensive Corrective Action Plan (CAP) was initiated, focusing on three key related system
elements:
• The No. 8 LFC and its dynamic performance;

• The Green Line track condition and future maintenance standards (Table 2); and
• The existing No. 7 SRC and its impact on track condition.
The CAP was developed jointly between the MBTA and its consultants—Booz Allen
Hamilton (coordination), HNTB (track issues) and Transportation Technology Research Center
(TTCI) (vehicle dynamics). The CAP also involved the carbuilder, AnsaldoBreda, supported by
vehicle dynamics experts from the Politecnico di Milano (PdM). The MBTA sought confirmation
of its CAP process by requesting peer review from a panel assembled through the American Public
Transportation Association. Finally, the CAP was submitted to and approved by the MBTA’s State
Safety Oversight body, the Massachusetts Department of Telecommunications and Energy (DTE).
As the nature of the derailment problem and the necessary corrective actions became clear,
the MBTA decided on a phased approach to restoring the No. 8 LFCs to revenue service on a
route-by-route basis. The initial phase, Phase 1, was focused on an interim return to service on one
route [the Commonwealth Avenue (B) line]. This line operates at relatively low speed but presents
many different track geometry and condition challenges, and thus forms an ideal basis for testing
No. 8 LFC operation. Phase 1 was used to establish the baseline requirements for operation of the
cars at limited speeds of up to 35 mph (56 km/h) and included work related to wheel profiles, track
maintenance standards and upgrades, and validation testing. Subsequent and current work, under
Phase 2, will see a return to service on a line-by-line basis and will include the work necessary to
validate safe performance up to the maximum required speed of 55 mph (88km/h).
TABLE 2 MBTA’s New Track Maintenance Standards
Maximum track gage [tangent and curves greater than 1000- 56-⅞ inch (1445 mm)
ft (305-m) radius]
Horizontal alignment, 31-ft (9.4-m) chord ⅝ inch (16 mm)
Runoff over 31 ft (9.4 m) at end of raise 1-¼ inch (32 mm)
Deviation from uniform profile–62-ft (18.8 m) chord 1-⅝ inch (41 mm)
Variation in cross level on spirals over 31-ft (9.4 m) chord ⅞ inch (22 mm)
Deviation from zero cross-level 1-⅛ inch (28.6 mm)
Difference in cross-level over 62 ft (18.8 m) 1-⅝ inch (41 mm)
NOTE: Maintenance standards for unrestricted speed.
Derailment Investigation Approach
The wheel to rail interface was exhaustively investigated, and complex mathematical models
were developed to simulate the behavior of the vehicle on measured Green Line track geometry
and irregularities. The aim of this investigation was to isolate the detailed mechanics of the
derailment, and to identify appropriate corrective action strategies.
Initial work focused on addressing the derailments using the simple Nadal theory as a
safety limit (Figure 2). This limit remains a conservative one; although, for an independent
wheel, more accurate approaches can be employed when full details of the wheel-to-rail interface
are available. Initial efforts were focused on increasing the Nadal limiting value by addressing
the most controllable variable—the effective flange angle. The original wheel profile design
employed a 63-degree flange angle, driven by the fact that all other vehicles on the Green Line
use such a profile. Analysis quickly showed that a change to a more modern 75° flange angle
was feasible and should deliver significant increases in the derailment safety margin by
increasing the Nadal limit. Considered simply, for a constant vertical load, the tolerable lateral
force between wheel and rail could be increased by 53% by changing the flange angle. In order
to confirm the validity of this premise, an extensive investigation was launched.
The conceptual approach followed during the investigation is shown in Figure 3. This
approach was followed to develop, refine, and validate an accurate dynamic model of the vehicle
Vertical Load – V
Nadal’s limit:
L Tan δ − µ
=
V 1 + Tan δ
Lateral
Force – L Where µ is the coefficient of friction
between wheel and rail
Effective Flange Angle δ
FIGURE 2 Nadal’s derailment theory.

Instrumented car Dynamic

(main line Measurements
testing)
Real
environment Refine No Converged Yes
Validated
Model Results Model
Draft Model ?
Software
Track Geometry
Simulation
measurements
FIGURE 3 Outline of derailment investigation approach.
for use as a predictive tool for derailment performance. Two independent mathematical models
were developed (using NUCARSTM by TTCI and ADTres by PdM). These models used some
different approaches and assumptions, but ultimately converged to give comparable results. Both
models were also validated by real world testing, building confidence in their accuracy.
The model was used to develop the new wheel profile for use on the No. 8 LFC. This
new profile was designed to optimize wheel-to-rail contact and increase the effective contact
angle. In conjunction with this investigation, extensive activities were carried out to upgrade the
condition of the Green Line track. Modeling showed that a common defect on the Green Line,
cyclic side wear, or scalloping, was particularly dangerous for the dynamics of the independent
wheel design. In addition, during the investigation of the new profile, some rail wear conditions
were identified that would compromise the effectiveness of the new wheel profile. These
conditions manifested as side wear, with a “lip” of flowed metal formed immediately beneath the
tip of the existing wheel flange. The new wheel profile would interfere with this lip, so an
unprecedented rail side-grinding program was initiated, which removed this condition on the
entire Green Line.
New, viable, track maintenance standards were identified and implemented by the
MBTA. In parallel, through extensive modeling and testing, the maximum line defects
(alignment, gauge, and cross level) acceptable to the No. 8 LFC were identified and compared
with the new maintenance standards.
Test Track Testing
While all this was good in theory, it was agreed that physical testing was necessary to confirm
the true benefits of the wheel profile and track condition changes. The first stage in the process
was to construct a test track, into which predetermined perturbations were installed (Figure 4).
This test track was instrumented with strain gages to measure the actual forces in the rails
due to wheel-to-rail interaction. In addition, a test vehicle was extensively instrumented to
monitor movements and accelerations at various locations throughout the car. One of the key
elements in this was the development by AnsaldoBreda of an instrumented wheelset (Figure 5).
FIGURE 4 Perturbation in test track.
Strain gage array for wheel

to rail vertical forces
Strain gage array for wheel

to rail lateral forces
FIGURE 5 Instrumented axle.

This innovation used the drop axle arrangement of the center truck as a system for
measuring the actual wheel-to-rail forces, in close to real time. This system proved much quicker
to produce and more cost effective than a more traditional instrumented wheel and formed an
invaluable part of the investigation.
The perturbations introduced into the test track were quite extreme—1⅝-inch (42-mm)
horizontal alignment, ⅝-inch (16-mm) gage, and ¾-inch (19-mm) cross level, all measured over
a 31 ft (9.4 m) chord—to deliberately provoke high L/V ratios at the maximum design speed for
the test track of 25 mph (40 km/h). Under the high friction levels encountered in the summer
months, L/V levels of 0.95 were reached at speeds of up to 28 mph (45 km/h). No derailments
were experienced.
The test track testing program permitted a direct comparison of model predicted, and
actual measured, vehicle reactions to the controlled perturbation inputs. Good correlation was
found between the predicted and measured wheel-to-rail forces and thus L/V ratios. Good
correlation was also achieved with other vehicle parameters such as modes of vibration and
acceleration levels. The program thus achieved its primary objective of validating the
mathematical model(s) of the vehicle and resulted in a believable predictive tool for derailment
behavior.
Main Line Testing
The next step was to use the model to predict vehicle behavior in the real world, represented by
the Commonwealth Avenue (B line) of the Green Line. Detailed Track Geometry Measurement
System (TGMS) data was collected, using a vehicle mounted non-contact (laser) system, to
obtain the most reliable loaded track geometry data. One of the challenges of this process was to
filter the data, as the system was originally designed for the relatively large curve radii and
smooth transitions found on railroads. The tight curves with short or missing spirals found on the
MBTA’s 100-year-old streetcar system often confused the system into reporting major alignment
errors rather than actual curvature. Problems were also encountered with the laser system
detecting wayside features, such as restraining or girder rail and road crossings as rail positions,
and thus reporting gage as tight as 54 in. (1372 mm).
With the TGMS data filtered, the vehicle models were used to predict vehicle
performance over the alignment and to identify problem areas of the track. Certain of these
problem areas were used as focal points for later dynamic testing. The same test vehicle, with its
instrumented axle, was used to conduct tests at various speed increments over the test locations,
in addition to a line speed “minesweep” run of the entire line. The results from this test program
both confirmed vehicle performance and identified a number of locations where corrective action
was required for the track.
Having confirmed that the vehicle performance would be acceptable within the newly
defined track maintenance standards, the track on the B line was upgraded to comply with these
new standards. As a final step, a repeat TGMS run was performed to confirm that the upgrades
had delivered a line that complied with the new standards.
Revenue Service
After a period of test running, and upon receiving CAP approval from the Massachusetts DTE,
on March 22, 2003, revenue service resumed with the No. 8 LFC fleet on the B line. Revenue
operation of the No. 8 LFCs is currently restricted to this route, pending upgrades of other routes
to the new track standards and further investigations and possible changes to raise the current
vehicle maximum speed of 35 mph (56 km/h) to 55 mph (88 km/h).
Summary of Findings
Derailment Causes
The fundamental reason for the derailments was found to be the Green Line track conditions
(type and severity of irregularities) that proved to be critical for the dynamics of the
independently rotating wheel design (Figure 6). In addition, the Green Line fleet original wheel
profile, specifically the flange angle of 63º, resulted in a low L/V limit (under high friction
levels), preventing any additional derailment safety margin. Vertical truck performance was not
found to play a significant role in the derailments, as the articulated frame design provides for
excellent load equalization.
The modeling process identified lateral alignment as the predominant critical track
condition, particularly short wavelength lateral alignment perturbations—magnitude greater than
⅝ in. on a 31-ft wavelength. The problem of lateral perturbations is compounded when these
perturbations are cyclic in nature. Track surveys identified that such cyclic alignment conditions
were to be found especially on the higher speed sections of the Green Line, and it has been
hypothesized that these are the result of interaction between the track structure and older
generations of vehicles. The situation is more severe with combination defects, where
perturbations in alignment are compounded with gage and or cross-level defects.
Conventional Wheelset Independent Wheel
Direction of
Direction of
Travel
Travel
Longitudinal Creep
Force Lateral Contact
Lateral Contact
Force
Force
Resultant Steering
Force Steering force is absent.
FIGURE 6 Comparison of independent and coupled wheel steering.

However, the work to date has confirmed that the No. 8 LFC, with the new wheel profile,
has acceptable margins of safety against derailment at speeds of up to 35 mph (56 km/h) when
operated over track that is maintained according to the MBTA’s newly developed and
implemented track maintenance standards, even in the presence of cyclic alignment defects.
Friction Levels
One of the other contributing factors to the derailments was the magnitude of the coefficient of
friction between the wheel and the rail present on the Green Line, that was greater than expected
in the Northeastern United States. To ensure the correct Nadal limit was used, tribometer
measurements were taken. The results showed levels as high as 0.6 under extreme conditions and
consistent levels of 0.5 over a prolonged period, including overnight. The effect of the
coefficient of friction on the Nadal limit is well known, and the maximum L/V level permissible
under Nadal with a 0.5 coefficient of friction is 0.74. This increases to 1.13 with the new,
steeper, flange angle of the new wheel profile.
Corrective Actions
Wheel Profile
The new 75º wheel profile (Figure 7) designed during the investigation has been implemented, to
all trucks, on all No. 8 LFCs, and the MBTA is investigating changing its existing fleet to this
profile in order to maximize wheel-to-rail compatibility. The MBTA has addressed rail profile
compatibility through the initiation of a rail side-grinding program, and there are plans to
develop a new rail head-grinding profile to optimize compatibility with the new wheel profile.
FIGURE 7 New wheel profile for No. 8 LFC.

Track Condition
Track maintenance standards and procedures were developed and implemented as a result of the
investigation. One aspect of these standards is to adopt the current FRA (5) approach for
alignment measurement, over a 31-ft (9.4-m) chord (older standards required measurement of
alignment only over a 62-ft (18.9-m) chord). Track geometry is therefore now measured over a
31-ft (9.4-m) chord for both curved and tangent track, and new limits have been established for
the various irregularity parameters.
The MBTA currently plans to perform TGMS data collection every three months to
monitor track condition, supplemented by track walking three times per week. In addition, the
rail profile will be optically measured every six months to ensure that an appropriate rail side
contact angle is maintained. This aspect is expected to be particularly critical during the
transition of the vehicle fleet from the existing to the new profile.
Future Efforts
The current derailment investigation (Phase 2) is focused on re-introducing revenue service on

other routes, based upon the results from the Phase 1 investigation. The investigation continues
to investigate possible methods of increasing the safety margin against derailment at higher
speeds. Other efforts under active investigation are the introduction of the new wheel profile to
the No. 7 SRC fleet, changes to the rail profile, and adoption of friction management techniques.
Although flange lubrication should not be depended upon to prevent derailments, it does have a
role as a mitigating method and brings other benefits of reduced wear and noise.
CONCLUSIONS
The introduction of new LRV technology into a system built around very different requirements
many years ago can cause unexpected problems. Problems have been encountered with matching
brake performance with the existing fleet, under extreme rail conditions, and with compatibility
of the new independent wheel center truck design, dictated by the low floor requirement, with
the existing Green Line track infrastructure.
There are valuable lessons to be learned from this project, which are applicable to older
properties that are contemplating introducing new vehicle technology that differs significantly
from that currently in use.
In particular, independent wheel truck designs require careful integration of track
maintenance standards with the vehicle design since such designs are less forgiving of track
irregularities than conventional rigid axle trucks. The track irregularities that appear to cause the
most problems are lateral, whether due to wide gage, side wear, rail deformation, or
misalignment. Reverse curves with short tangent lengths are also particularly challenging for
these trucks, as they are not able to steer themselves correctly through such geometry. It is
important to note that these problems have been the experience with trucks that have good
vertical equalizing properties. Stiffer trucks may also suffer problems due to vertical alignment
and track twist or warp. From the No. 8 LFC experience, the following lessons learned are
offered:
• Ensure that all parties are fully aware of exactly what track conditions (geometry and
quality) and maintenance standards will be maintained (taking into consideration the inherent
characteristics of independent wheel designs). Reliable line geometry and defect measurements
should be taken to avoid any misunderstandings. Such information should be included in the
specification, and must be used as a design input.
• Focus on early identification of any track features that may cause derailment
potential, so that they can be addressed before the vehicles are delivered.
• Perform rigorous and truly representative dynamic modeling of the vehicle design
over the actual track conditions that will be encountered.
• Validate the dynamic model by track testing at the earliest opportunity.
• Recognize that track maintenance standards must be rigorously enforced and may
need to be raised to a higher level.
ACKNOWLEDGMENTS
The authors thank their respective employers for their support during preparation of this paper
and note that any views expressed herein are the authors’ only, and not those of their employers.
The efforts of the following in understanding the derailment mechanisms is recognized: the
TTCI team, most prominently John Elkins and Nicholas Wilson, and the
AnsaldoBreda/Politecnico team, led by Professor Giorgio Diana.
REFERENCES
1. Strong, D., and G. T. Morse. Low Floor Light Rail Vehicles for the MBTA’s Green Line:
History of Development and Progress So Far. Proceedings of the 1996 Rapid Transit
Conference of the American Public Transportation Association, Vol. 2.
2. Fraser, G. R., and G. T. Morse. Low Floor LRVs for Boston’s Green Line: Technical
Features and Challenges. Proceedings of the 1997 IEEE/ASME Joint Railroad Conference.
3. Fraser, G. R., and G. T. Morse. Engineering Solutions for MBTA’s Low Floor LRVs.
Proceedings of the 1997 Rapid Transit Conference of the American Public Transportation
Association, Vol. 2.
4. Fraser, G. R., M. Marianeschi, and G. T. Morse. Boston’s Low-Floor LRV—Program
Update. Proceedings of the 1998 IEEE/ASME Joint Railroad Conference, and Proceedings
of the 1998 Rapid Transit Conference of the American Public Transportation Association.
5. Track Safety Standards. Code of Federal Regulations. 49 CFR Part 213, Subpart A to F, Oct.,
1998.

Low Floor Light Rail Vehicles-Structural and Crashworthiness Requirements

Uploaded by

Copyright:

Available Formats

Low Floor Light Rail Vehicles-Structural and Crashworthiness Requirements

Uploaded by

Document Information

Original Description:

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Low Floor Light Rail Vehicles-Structural and Crashworthiness Requirements

Uploaded by

Copyright:

Available Formats

LOW-FLOOR

LIGHT RAIL VEHICLES

Structural and Crashworthiness Requirements of

LFE STRUCTURAL ISSUES

STRUCTURAL STRENGTH OVERVIEW

CRASHWORTHINESS—STANDARDS AND INDUSTRY PRACTICE

LRV–LFE DESIGN ALTERNATIVES

Industry Practice—Traditional 2-g Criteria

Strength-Based Design: Match the Existing Buff Strength

Strength-Based Approach with Crash Energy Design Principles

APPROACH AND METHODOLOGY

• Define the vehicle characteristics to use as model input data;

The simulation is based on the dynamic analysis of mass-spring systems. A commercially

• Existing vehicle with no LFE;

• Simulation 1 included a number of analyses to determine the worst-case combination

Methodology and Definition of Vehicle Model

Generation of Characteristic Force Deflection Curve for the Vehicles

• The existing vehicle, including the couplers;

Analysis Assumptions and Collision Scenarios

• Both consists are on level tangent track;

The following collision scenarios were analyzed as part of Simulation 1:

Results of Simulation 1: Determining Worst-Case Scenario—Simplified Model

TABLE 1 Summary of Results for Collision Scenario 4 (Simplified Model)

C) Speed = 17.5 mph

Results of Simulation 2: Refined Analysis for Scenario 4

Vehicles 1 and 2—Moving Consist

Vehicles 3 and 4—Standing Consist (Brakes On)

Comparison of the CEM Versus the 2-g Approach—Scenario 4.1

1st Vehicle 2nd Vehicle

SUMMARY AND CONCLUSION

Light Rail Vehicles’ Low-Floor Center Section

DEFINITION OF THE PROBLEM

• Continue to operate and expand or upgrade the system, “as is;” or

• A new fleet of low-floor rail vehicles;

REAL WORLD SCENARIO

• “Signature” appearance of the DART LRV,

Low-Floor Light Rail Vehicles

Because of the established, acceptable, history of operation, a detailed review of the

Low-floor Center Section

Dimensions and Performance

Manufacturing and Testing

Compatibility with the Rest of the System

Wayside Systems: Signal System, Grade Crossings, Traction Electrification System,

Service and Inspection Facility

TABLE 1 Comparison of Seating Capacity by Car Consist

DART LRV LFLRV Super-LRV

DART BUILD-OUT PHASE II

COST—THE DRIVING FACTOR

Maintenance Facility Modifications

TABLE 2 Comparison of Options for DART Fleet Combinations

Base Fleet Option 1 Option 2 Option 3

Integrating New Light Rail Vehicle

SUMMARY OF THE NO. 8 LFC PROJECT

FIGURE 1 Prototype No. 8 LFC at riverside carhouse.

TABLE 1 No. 8 LFC Main Features

Type Bidirectional, double-articulated LRV

Integration with Existing Vehicles

Slide Control Challenges

DERAILMENT OF INDEPENDENT WHEELED TRUCK