Structural Crashworthiness of Railway Vehicles

1
Manuel S. Pereira

1
IDMEC, Department of Mechanical Engineering, Instituto Superior Técnico, Technical University of
Lisbon, Avenida Rovisco Pais 1049-001 Lisbon, PORTUGAL, [email protected]

Abstract

As a form of transport, rail has always maintained lower accident levels than its counterparts, the car and
the airplane. The railway industry, however strives for still better safety measures to be implemented – a
fact which intensifies with each accident that occurs. In the last years the continuous developments in
technology, computer modeling capabilities and know-how have opened up various research activities in
the passive safety area, also known as crashworthiness.

Whereas the objective of active safety systems (such as signalling and automatic train protection
systems) is to avoid accidents, passive safety only comes into effect in the event of collision and its
objective is to significantly reduce the severity of accidents.

Rail vehicles can be designed to behave in ways that minimize the injuries of passengers and crew during
collisions. Crashworthy vehicles contain in-built design features, which are not relevant in normal train
operations, but protect the train occupants should an accident occur. A methodology has been developed
within the projects TRAINCOL, SAFETRAIN and SAFETRAM for improved passive safety in railway
transportation. This methodology includes:

i) A review of past accidents, identification of reference collision scenarios and evaluation of their
consequences. Risk assessment for improved passive safety was also considered;

ii) The establishment of a set of reference collision scenarios for main line train and city tram
operations. Together with statistical and risk analysis, design feasibility studies played a major role in
the definition of the reference scenarios;

iii) The development of a general framework for structural crashworthiness design train and in city and
periurban tram vehicles and

iv) Guidelines for design validation procedures through modeling, component and real size dynamic
testing.

Improved passenger and crew safety issues have been systematically addressed. Passenger and crew
accommodations on a number of railways within Europe have been reviewed from the standpoint of
interior safety. Conclusions about typical acceleration pulses and appropriate values of injury criteria for
the railway industry have also been established.

One of the most important achievements of these projects was then to demonstrate the feasibility of
optimized carbody structures to present an improved safety level to occupants, within acceptable cost
and masses for the defined construction solutions. A European Standard is now being completed
providing the framework for determining the crash conditions that railway vehicle bodies should be
designed to sustain based on the most common accidents and associated risks. It also defines suitable
passive safety features to meet the requirements.
Introduction

Occupant safety is dependent on the configuration and severity of the accident, as well as the degree of
crashworthiness engineered in the overall vehicle design. Train vehicle occupant survivability in a given
crash scenario is a function of the kinematic behavior of the entire train set, the integrity and collapse
characteristics of the structure of each vehicle and the overall interior configuration of a compartment and
occupant/surfaces contact characteristics.

Train crash events can be basically depicted into two phases: In a first phase, normally referred to as
primary collision, the initial kinetic energy is progressively dissipated by means of plastic structural
deformation resulting from the crash generated impact loads. In this phase occupant compartment
integrity and acceptable vehicle acceleration levels are the most important design requirements to be
considered. In a second phase, normally referred to as secondary collision, the occupant will be subject
to a great variety of potentially harmful occupant/interior or occupant/occupant contacts. Design
requirements must involve the aspects of interior layouts, acceptable severity levels and biomechanical
response to vehicle crash pulses. The friendliness of the compartment interior is obviously a major design
issue.

The energy generated in train collisions has to be dissipated by plastic crushing of designated structural
arrangements developing forces, which in turn cause decelerations, which are directly responsible for the
severity of secondary collisions. These designated structural arrangements are normally located at the
extremities of the vehicles and are designed to absorb maximum possible energy and collapse in a
controlled manner preserving the occupant compartment integrity. The crushable zones must be confined
to reasonable lengths and the range of crush loads are to be consistent with passenger compartment
buckling loads and static specified buffer loads [9].

Train collision are always assumed to occur under fully plastic impact conditions, which means that after
collision both sets have the same velocity, which corresponds to a zero coefficient of restitution.

For single vehicle collisions, such as the case of many tram configurations, the starting point for design
requires the knowledge of the design collision scenario speed and the maximum allowed acceleration
experienced by the occupant compartment. Closed form solutions are available for determining the
required crushing length. Forces can subsequently be determined to specify the design requirements, in
terms of a force-displacement curve for the extremities.

For train collisions involving several vehicles the problem is considerable more complex. The energy
distribution along train sets is strongly dependent on:
• Train configuration, namely the number of vehicles
• Mass distribution – masses of individual vehicles along the train set
• Collision speed
• Type of scenario.

The characteristics of structural crashworthy vehicles are expressed in terms of force-displacement

curves with different levels of plastic loads. Different force-displacement curves are used for inter-rake
and inter-trailer zones and are referred to as EE and IE ends, respectively.

SAFETRAIN project [2], [3] and [11] concentrated in the study of regional, intercity and high speed train
sets configurations. SAFETRAM project [12] dealt with the specificities of city trams and periurban trams.

Based on results of these projects, methodologies for improved passive safety, including the
specification, design, testing and validation procedures for crashworthy rail vehicles, is outlined. Specific
recommendations are proposed for passive safety development, minimisation of loss of survival space,
train energy management issues, vehicle design, minimisation of severity of occupant injuries, methods of
validation of crashworthy designs, static and dynamic testing and structural and passenger/crew
numerical simulations.
Review of accidents and choice of representative collision scenarios

The principal aim of this task is to define the representative collision scenarios, their characteristics and
parameters to be applied for modelling, design and test of crashworthy structures and other protection
means (anti-overriding devices, obstacle deflectors, interior design). The main characteristics of collision
accidents in Europe have been gathered into a representative database [8]. The quantity of data collected
corresponds to approximately 60% of total European railway production (in terms of train kilometres,
passengers transported, passenger kilometres and length of lines in service). The database forms a
statistical population of collision accidents over the five years 1991 to 1995.

The analysis showed that building practicable levels of energy absorption into passenger rolling stock
would improve the passive safety of vehicles in head on collisions, rear on collisions, collisions with cars
and lorries on level crossings and collisions with buffer stops.

Risk analysis on the city tram accident data, collected through an inquiry was performed [5]. The purpose
of the LRV statistics study was to identify relevant collision scenarios including an evaluation of their
consequences, in terms of material damage and injuries and fatalities as applied to city tram operations.
As there are few periurban trams in service and periurban traffic is relatively recent, all analysis refers to
accident data involving only regional traffic. Risk and statistical analysis was conducted on the German
and French accident data of ERRI B205 database, involving 248 and 329 accidents, respectively [10].

Based on the findings of the statistical analyses described herein the scenarios and train and tram
configurations, presented in table1, have been selected.

Rollling Collision scenarios Configurations

stock
Trains S1) Train vs. train train types were identified in single (multiple) units.
S2) Train vs. buffer stops Type A – Main line train – mass 340 t – 8(16) vehicle
S3) Train vs. deformable obstacle In level crossing train sets
different statistical impact velocities were defined from Type B – Main line train – mass 412 t – 8(16) vehicle
the B205 data base analysis. Each velocity covers, train sets
respectively, 70, 80 and 90 % of accidents occurred. Type C – Regional train - mass 129 t – 3(6) vehicle train
For scenario 1 : V70% = 30 km/h, V 80% = 40 km/h, V 90% = sets
55 km/h. Type D – Motor coach, single vehicle – mass 50 t 1(2)
For scenario 3 : V 70% = 80 km/h, V80% = 100 km/h, V 90% vehicle train sets
= 120 km/h.
City C1) Emergency braking. Not applicable in this Multiple-articulated tram configurations
trams analysis Masses of 20, 35 and 45 t.
C2) Collision with an identical city tram. 20 km/h A crushing length of 0.5 m is assumed and based on the
C3) Collision against a light rigid truck (3t). 30 km/h assumptions of a buffer load of 200kN
Compressive force reaches a maximum of 470kN
C4) Collision against a 55 t periurban tram with
appropriate compatibility. 12 km/h
Peri P1) Collision against a rail vehicle (80 t freight wagon The periurban tram is also multiple-articulated
with UIC 526-1 buffers). 25 km/h Mass of 55 t.
urban
trams P2) Collision against a regional train (129 t train). 23 A crushing length of 0.7 m is assumed and based on the
km/h. assumptions of a buffer load of 600kN
P3) Collision against an identical periurban tram. 36 Compressive force reaches a maximum of 1400kN.
km/h
P4) Collision against a 16.5 t rigid truck on level
crossing. 40 km/h

Table 1: Collision scenarios and train and tram configurations

Design feasibility and optimization studies for train set configurations
Geometrical characteristic of the extremities were fixed identical for the different trains:
Driver’s cabin:
Total crush length available is 1900 mm,
Tolerated crushing length of the driver’s cabin: 1000 mm,
Component (non structural devices) crushing length in front of the cabin: 800 mm
Coupler crushing length 100 mm.
Intertrailer:
Inter-trailer length: 800 mm
Coupler crushing length: 200 mm
Component (non structural devices) crushing length at the inter-trailer: 200 mm
Crushing length of the structure at the inter-trailer: 500 mm

Safety objectives were established in terms of mean accelerations below 5g in the passenger
compartment and no permanent deformation should occur in the passenger area.
Optimisation procedures have been applied to trains type A, B, C and D. In all cases and for all collision
scenarios the proposed crashworthy train design with the optimised plastic load levels respect the
acceleration criteria and the limits in the levels of deformation. The extremities so determined should
ensure no deformation in passenger areas for a collision speed of 55 km/h in scenario 1 (train vs. train)
and 100 km/h in scenario 3 (train vs. lorry in level crossing).

The obtained optimal force-displacement curves result in decelerations in the passenger areas, which are
below 5g. The resulting force-displacement curves for the HE and LE ends are represented in figure 1.

Fig. 1: HE and LE end characteristics for train C

Table 2 summarizes the general requirements in terms of energy absorption for the most critical
scenarios:

Scenario 1 Scenario 3 M=16.5 t

Train
Type Mass V (km/h) E1 (MJ) E2 (MJ) V (km/h) E1 (MJ) E2 (MJ)
A 340 t 55 3.6 2.7 96 5.0 0.7
B 412 t 55 2.95 2.8 96.5 4.9 0.7
C 129 t 55 2.3 1.4 98 4.6 0.6
Notes: Scenario 1 and Scenario 3 – Reference Accident Scenarios
E1 – Energy absorbed in the front end
E2 – Energy absorbed in the inter-trailer areas

Table 2: Energy levels for train collisions

Design feasibility and optimization studies for tram configurations

For all the scenarios, the allowable consequences of the accidents are:
• Plastic deformation only in localized crash-zones, not affecting the survival space of the
occupants.
• Minor injuries for the tram occupants. As reference, a 5g maximum deceleration of the car body is
assumed.

Using the SELFA plots [6] methodology the analysis of design feasibility involves the following steps:
1. From the energy-force curves energy values associated with the maximum force levels are found
in the appropriate curve for the available crush length.
2. Equivalent masses can also be found depending on the masses of the city or periurban tram and
the colliding partner.
3. Using the energy levels and the equivalent mass previously calculated the maximum allowable
speed is then obtained.

Table 3 summarises the results for design feasibility.

City Trams (35 t) Periurban trams (55 t)
Force Energy Meq Speed Force Energy Meq Speed
Scenario Scenario
(kN) (MJ) (t) (km/h) (kN) (MJ) (t) (km/h)
City Tram (2) 470 .290 35 20 80 t wagon (6) 1400 .8 65.2 25
3 t truck (1) 470 .145 5.5 30 129 t train (7) 1400 .9 77.1 23
Periurban (3) 470 .145 42.8 12 Periurban (5) 1400 1.6 55.0 36
16.5 lorry (4) 1400 .8 25.4 40
st
The numbers in bracket refer to numbered points in the energy-mass plots in the 1 quadrant of figure 1.

Table 3: Energy absorption characteristics and allowable speeds for trams and periurban trams

Overriding studies

A review of accident data within Europe and the USA has highlighted that end-on collisions (head to head
or rear-on collisions) are responsible for most of the serious and fatal injuries to passengers and crew in
train accidents. Within these reviews, vehicle overriding has been identified as one of the most important
factors in determining the number and severity of injuries; it has been demonstrated that the casualty
rate, especially fatality rate increases if overriding occurs. It is evident therefore that override prevention
will significantly improve passenger and crew safety on the European rail network. Before this can be
achieved, however, a thorough understanding of the causes of overriding is necessary.

The study of European accidents produced the following general conclusions.

• If overriding occurs, fatality rate is 3-4 times higher than for similar collisions without overriding.
Most fatalities are in the overridden vehicle.
• Fatalities occur in 70% of all overriding collisions.
• With buffered stock, overriding can occur at closing speeds as low as 15 km/h.
• Overriding may occur at collision energies as low as 2.5MJ but becomes likely at collision
energies greater than 10MJ.
• Collisions involving vehicles with buffers are approximately three times more likely to result in
override than collisions without buffers.
• There is no evidence to suggest vehicle pitch contributes to overriding in passenger vehicles.

2D multibody models have been developed to study overriding effects in train collisions [1] and [4]. These
models include: bogie arrangements with primary and secondary suspensions with rigid stops;
representation of structural and non-structural devices behaviour in the longitudinal direction; flexibility of
passenger cars; contact capabilities between anti-climbers.

A series of scenarios have been considered including train types A, B and C. For each case three
collision speeds have been studied. Favourable comparison of results also indicates that the overriding
phenomena and pitching motions, in general, do not affect the basic mechanisms of energy absorption as
depicted by 1D models.

Major conclusions from this study are:

• Vertical gaps and vertical forces tend to increase for interfaces away from the inter-train areas. •
Maximum values tend to increase with collision speed.
• Maximum values of vertical gaps are around 100-140 mm. However, over riding phenomena in
middle train cars is typically initiated with vertical gaps of around 50 mm.
• Vertical forces have been found to reach values as high as 150 kN. These force levels indicate
that the deformation behaviour of anti-climbers, in the vertical direction, must be subject to careful
design, and provisions must be taken to accommodate some energy absorption with reduced
force levels, in the vertical direction.
• Overall results indicate that the dynamics of overriding is strongly dependent on: Vertical anti-
climber crushing behaviour; Suspension characteristics with emphasis in jounce stop strokes;
Center of mass offsets; Collision speeds.
• And it is less dependent on rain configurations

Overriding in end-on collisions is the single most serious event that can happen as far as safety of
passengers is concerned. Surfaces, which can easily deform or slide over one another, e.g. buffers, are
instrumental in allowing vertical forces to develop which initiate override. The project proposed a draft
specification for anti-climbers for discussion purposes.

A final point that requires consideration is the inter-operability of rolling stock throughout Europe.
Currently there are several designs of anti-climbers, each of which is incompatible with the other. If the
safety of passengers and crew is to be increased by the prevention of overriding, serious consideration
needs to be given to specifying a common design of anti-climber fitted to all vehicles in the same way that
buffer position has been specified in the past.
Design, modelling and testing of rail vehicle ends

In the initial stages of the SAFETRAIN project the technical feasibility of initial designs of new energy
absorption zones located at the train ends (HE) and at the ends of intermediate vehicles of the train (LE)
was appraised. Global design and test specifications have been established for regional type trains. This
case was considered as most challenging, in view of the foreseen small space available for the HE and
LE extremities (1.8m and .7m, respectively). The levels of energy required to be absorbed in these
extremities were 4.6 MJ and 0.7 MJ, respectively. General construction rules were issued concerning
loading gauge, lengths of high and low energy ends, window and door position, and type of coupler,
obstacle deflector and anti-climbers. The design of the HE and LE extremities was carried out completed
and dully validated with appropriate detailed FE model. Components and sub-assemblies were
manufactured and tested dynamically before full scale manufacturing of the HE and LE extremities.

HIGH ENERGY END

In the high-energy end it was originally intended that all deformation would occur forward of the driver’s
accommodation. However, since the design was for regional trains, where the length of trains is limited
by platform length, it was established that part of the driver’s accommodation would be used for structural
energy absorption. A survival space for the driver would be maintained at the rear of the cabin. In order
to demonstrate that there would be sufficient survival space for the driver, a driving console and a sliding
seat were fitted to the structure to be used in the real size dynamic crash test. The console played no
part in the energy absorption. The net outcome of these modifications did not change the force-
displacement and was that the structure would easily meet the 4.6 MJ requirements.

LOW ENERGY END

Although the vehicle end was designed to collapse in a controlled manner and be energy absorbing, the
required 0.7MJ was to be absorbed in phases 1 and 2 only. The general layouts of the manufactured
extremities are shown in figure 2.

Figure 2: High energy and low energy extremities

MODELLING AND DYNAMIC TEST VALIDATION

The Finite element models are shown in figure 3.The HE end was modelled using ANSYS 5.5.3 to
construct the model and RADIOSS 4.1 to carry out the numerical analysis. The dynamic impact was
modelled by simulating a wagon of mass 46250kg, to which the extremity is attached, impacting a
stationary wagon of mass 44900kg at 20m/s or 72km/hr.

The numerical analysis of the low-energy end was undertaken using PAMCRASH 98 software. The
model was attached to a rigid wall and impacted with a rigid heavy plate (5 m/s for the case of the non
structural device and 4 m/s for the deformable structure). Advantage was taken of symmetry and only
half the structure was modelled. The objective of the dynamic tests was the validation of methods of
controlled deformation and energy absorption for different vehicle types and the preparation of detailed
measurement specifications required for assessment of vehicle crashworthiness. The dynamic test
depicted in figure 3 involves two test vehicles with 45 t each colliding at a speed of 73.5 km/h generating
an energy absorption of 4.6 MJ at the test cabin. The results of the tests have been used for comparison
with numerical simulations and static crush test results [7].

Figure 3: Finite element modelling of high and low energy ends

Both quasi-static and dynamic tests allowed validation of the requirements on energy absorption levels
and on crushing force level of the front end. Force-displacement and energy curves and energy
absorption were in a very good agreement between the numerical simulations and the dynamic tests, as it
can be observed from the plots shown in figure 4. Figure 4 also shows a favourable comparison of the
crushing evolution as obtained from the dynamic and static tests and the dynamic numerical simulation.
 Figure 4: Comparison of tests data: Force-displacement curves and pattern deformations

Design, modelling and testing of tram extremities

Global design and test specifications have been established for city trams and periurban trams.

CITY TRAM
Overriding is prevented as the front end shock absorbers are fitted out with anti climbers made to
withstand vertical force 50 kN. The design is proven for misalignment of up to 50 mm.
With respect to scenario C3, anti-intrusion safety must be accomplished by the design and represents
another requirement for the structure.
Energy absorption: the test specimen has to provide a progressive characteristic for the energy
absorption as displayed. In the first step, reversible devices are used for energies of up to 35 kJ. In this
step the trigger peaks shall not exceed the buffer load of 200 kN. The energy absorbers for the second
step may get deformed irreversibly, but an easy changeability is required. In contrary to frontal collision
scenarios C2 and C4, the carshell structure may be used to absorb some energy in scenario C3 (corner
collision with a truck). But the door openings of the entrance area and the survival space shall be
maintained.
The maximum deceleration levels are generally less than 3g, for each collision scenario, therefore the
design will normally respect the deceleration criteria of a 5g threshold.
Buffing compatibility with respect to collisions with periurban tram must be fulfilled (C4). The height of the
impact zone of the collision partners is schematically represented in the figure.
Driver’s visibility : Any obstacles with a height of 1.2 m above the top of the rail, and with a distance of 0.3
m to the front of the vehicle must be visible for the driver.
The survival space for the driver must be maintained. DIN 5560 [14] is relevant to the specification of the
survival space. Consequently, one of two criteria must be met after the crash. Either the distance
between the driver’s desk and the driver’s seat is 300 mm at a minimum, or the distance between the
driver’s desk or the back wall of the cabin is not less than 750 mm.
The coupler arrangement shall not contribute to the energy absorption. The folded coupler is located
behind the crash zone.
Figure 5 shows the structural arrangement for the city tram test layer.

Figure 5: Design of the city tram test specimen

For the city tram, the tests and the calibrations on the changeable crash modules at the front of the
driver's cabin have been well performed and analyzed involving energy absorption of 135 kJ. Then the
test on the mock-up has presented good results for these elements (protection against overriding and
crushing process). The corresponding absorber tests have been correlated, precisely for the city tram test
design. After the tests, a final correlation has been successfully obtained for the city tram as shown in
figure 6.

Figure 6: Validation of the city tram test specimen

PERIURBAN TRAM
Overriding is prevented by using anti-climbers withstanding vertical forces of up to 150kN. Vertical offsets
of up to 50 mm are allowed. Compatibility with buffers of freight wagons is assured.
The vehicle shall absorb the energy appropriate to each relevant limiting collision scenario in a controlled
manner. This shall be accomplished in such a way that structural collapse is confined to designated areas
of the structure and the main passenger/crew space is preserved. The static longitudinal strength of the
periurban trams is equal to 600 kN, then the maximum crush force is 1500 kN.
The maximum deceleration levels are generally less than 3g, for each collision scenario, therefore the
design will normally respect the deceleration criteria of 5g.
For the periurban trams, an obstacle deflector equivalent to the classic train deflector is not required.
Nevertheless, a stone deflector equivalent to the trams deflector is required.
Any obstacles with a height of 1.2 m above the top of the rail and with a distance of 0.3 m to the front of
the vehicle must be visible for the driver.
The vehicle shall preserve the survival space and prevent intrusion inside. For the present design the
following criteria was adopted: either the distance between the driver's desk and the driver's seat is 300
mm at a minimum, or the distance between the driver's desk and the back wall of the cabin is not less
than 750 mm.
The coupler arrangement shall not contribute to the energy absorption, but the design must make sure
that the coupler does not disturb the correct functioning of the energy absorbers (in other projects,
manufacturers can propose couplers with contribution of the energy absorption). In case of an impact
against a train with central coupler, the coupler shall be caught to ensure the energy absorption and
prevent intrusion of this last. Space for the installation of a radio system must be provided; therefore the
cabin’s floor level must be specified accordingly.

The test specimen is displayed in figure 7. The requirement of a progressive energy absorption in
designated collapse zones could be met by the energy absorption concept which is composed of three
steps: (1) reversible shock absorber, (2) changeable crash module, (3) main structure. The main structure
provides a margin which is not used in the specified scenarios.

Figure 7: Design of the periurban tram test specimen

The simulation effort represented an important part of the design process. The Periurban cab concept is
very demanding and was based on a full 3D crush kinematics process involving a complex interrelation of
distinct crush processes of A-pillars and the structural devices. Figure 8 compares the crushed test
specimen and the numerical simulation results for the periurban test layer.

Several innovative features included in the present crashworthiness concept provide the necessary
conditions to meet a specific set of challenging design requirements in terms of the variety of colliding
partners and the geometric and loading constraints. Different types of energy absorbers at different
heights are provided to meet the requirements which result from the different collision partners.
 Figure 8: Validation of the city tram test specimen

Improved passenger and crew safety

The design of vehicle interiors to minimise secondary impact injuries is extremely complex. Whilst many
measures, such as preventing a relative velocity build up between the occupant and his surroundings, the
provision of “soft edges” etc are obvious, a means of specifying such features is much more difficult.
Current specifications within Europe tend to provide a list of aspirations rather than specific requirements.
Recent advances in the car industry and in modelling vehicle interiors now allow the rail vehicle interior
designer to provide more specific requirements. An assessment of the known acceleration-time pulses in
both the rail and motor industry was made with the objective of determining a representative pulse to be
used for any future mathematical modelling or sled tests. The study a corridor in which the acceleration
pulse must fall, thereby allowing the exact shape of the pulse to be determined by the vehicle designer.
The speed change associated with the acceleration corridor is 30 km/h, which covers almost 60% of all
head-on collisions, 90% of rear-on collisions and a significant proportion of buffer stop and level crossing
collisions.

Agreed injury criteria are recommended to enable seats and tables, in particular, to be modelled and/or
tested. Finally, drivers are at extreme risk in collisions. Modelling and test work have indicated mitigation
measures, which should significantly enhance their likelihood of survival.

Models representing 5th, 50th and 95th percentile cab occupants have been used to determine the
relative benefits of using a seatbelt, airbag and knee bolster in simulated end-on collisions. Both modern
European regional train cabs and high density UK cabs with very limited driver space have been
modelled. The simulation of passenger impacts with the interior furniture of a railway carriage, specifically
with seats, tables and in some scenarios other passengers was carried out for train interior
configurations. The results highlight the different requirements for passengers according to the different
carriage seat layouts and the sizes of passengers considered in this study.

Figure 9 shows the model used in the analysis of the driver impacts and an example of an occupant
impact in a classical train seating arrangement including seat models with structural flexibility.
 Figure 9: Driver and passenger models for secondary collisions.

The most significant scenario for CT occupants is the emergency braking (see WP 1), principally for the
standing passengers, resulting from the secondary collision of the passengers with the interior feature of
the vehicle. As the base of the mean deceleration value for the reference collision scenario, the
prescription of 2.73 m/s2 resulting from the German regulation (BO Strab) [14] has been chosen with an
operating speed of 70 km/h. Before the occupant modelling, it was necessary to validate the MADYMO
HIII standing dummy occupant kinematics and injury severity. A standing dummy conversion kit has been
acquired and a sled for specific emergency braking (carriages and platform) has been constructed in
INRETS and validated with tests. Different tests have been carried out, on dummies or volunteers, with
different deceleration levels, surrounding features, dummy orientations and grip systems.

The European Standard: Railway applications — Crashworthiness requirements for railway

vehicle bodies

This standard [15] (prEN 15227:2005) has been prepared by Technical Committee CEN/TC 256 “Railway
Applications”, the secretariat of which is held by DIN. This standard is currently submitted to the CEN
Enquiry. This standard has been prepared under a mandate given to CEN by the European Commission
and the European Free Trade Association, and supports essential requirements of EU Directive(s).

This European standard applies to new designs of locomotives and passenger carrying rolling stock as
defined in categories C-I to C-IV of clause 4 of the static requirements standard. The design of new
vehicles for use in passenger trains is based on operations with compatible rolling stock that also meet
this standard.

The requirements apply to the vehicle body, and to those mechanical elements directly associated with it
that may be used to absorb energy in a collision, such as couplers, buffing systems, anti-climbers and
obstacle deflectors, etc. They do not cover the safety features of doors, windows, system components or
interior features except for specific issues relating to the preservation of survival space.

The requirements do not cover all possible accident scenarios but provide a level of crashworthiness that
will offer an appropriate level of protection in most eventualities, when the active safety measures have
been inadequate. The requirement is to provide a level of protection consistent with the probable collision
risks and this is achieved by addressing the most common types of collision causing injuries and
fatalities.

The proposed standard prescribes the applicable design collision scenarios and gives guidance on
suitable parameter values for use in these scenarios.
The standard identifies common methods of providing passive safety that may be adopted to suit
individual vehicle requirements. The standard also specifies the characteristics of reference obstacle
models for use in the design collision scenarios. Not all vehicles in a train have to incorporate energy
absorption but passenger train configurations formed entirely of new vehicle designs shall, as a whole,
comply with this standard.

The standard also specifies the requirements for demonstrating that the passive safety objectives have
been achieved by comparison with existing proven designs, numerical simulation, component or full-scale
tests, or a combination of all these methods.

To the extent required by this standard the following measures shall be employed to provide protection of
occupants in the event of a collision:
- Reduce the risk of overriding
- Absorb collision energy
- Maintain survival space and structural integrity of the occupied areas
- Limit the deceleration
- Reduce the risk of derailment and limit the consequences of hitting a track obstruction

Conclusions
Passive safety in railway transportation deals with means to provide a safe environment for its occupants
during identified train collision scenarios. Injuries or deaths are caused by severe occupant/interior
contacts, which result mainly from significant damages in the vehicles’ structures or from large
decelerations sustained by a vehicle during the crash events.

A methodology has been developed within the projects TRAINCOL, SAFETRAIN and SAFETRAM for
improved passive safety in railway transportation. This methodology includes: i) A review of past
accidents, identification of reference collision scenarios and evaluation of their consequences. Risk
assessment for improved passive safety was also considered; ii) Development of a general framework for
structural design in train and tram vehicles and iii) Guidelines for design validation procedures through
modeling, static or dynamic testing.

One of the most important objectives of the suite of passive safety European projects was then to
demonstrate, for trains, periurban trams as well as for the city trams, the feasibility of optimised carbody
structures to present an improved safety level to occupants, within acceptable cost and masses for the
defined construction solutions.

New design concepts have been developed by defining technical design requirements to manage the
collision energy. Critical vehicle crashworthy validation procedures were carried out including
mathematical modelling simulation and component and full scale testing.

The friendliness of the compartment interior is now a major design issue and its treatment will complete
the framework of railway passive safety. SAFEINTERIORS, a European project to start now, will present
the European railway industry with a key step towards achieving full interoperability by providing the
scientific and technological basis to implement a consistent methodology to design, test and validate
improved interior solutions, thus reducing the levels of fatalities and injuries in rail accidents.

This new interior passive safety framework will provide a systems approach to drastically reduce injuries
and fatalities by combining and exploiting in a cost efficient and optimised manner the already well
matured railway structural crashworthiness (closely linked with primary collisions events), with injury
biomechanics, directly associated with secondary collisions.
Acknowledgements

The work reported herein was carried out in projects TRAINCOL, SAFETRAIN, SAFETRAM, under the
BRITE/EURAM 4th Framework program and the Growth 5th framework program, funded by the European
Commission.

The participation in the SAFETRAIN and SAFETRAM consortia of the following partners is
acknowledged. Bombardier Transportation, Portugal, S.A., ERRI - Foundation European Rail Research
Institute, HOLLAND, SNCF - Direction du Matériel et de la Traction, France, Deutsche Bahn AG
GERMANY, PKP - Railway Scientific and Technical Centre, POLAND, AEA Technology, UNITED
KINGDOM, Alstom – Valenciennes Unit, FRANCE, Alstom De Dietrich, France, DUEWAG , GERMANY,
Bombardier Transportation, GERMANY, Instituto Superior Técnico, PORTUGAL, , LAMIH – UVHC ,
FRANCE, Cranfield Impact Centre, UNITED KINGDOM, FMH - Laboratório de Ergonomia, PORTUGAL,
ANSALDOBREDA S.p.a., ITALY, Alcan Alesa Engineering Ltd., SWITZERLAND, MIRA Ltd, UNITED
KINGDOM, Régie Autonome des Transports Parisiens, France, Technische Universitaet Berlin,
GERMANY, Berliner Verkehrsbetriebe, GERMANY.

The support of Mr. Joost de Bock, European Commission, Directorate-General XII, is greatly
acknowledged.

References
[1] Ambrósio, J., Pereira, M. S., Dias, J. P., “Crashworthiness Analysis and Design Using Rigid-Flexible
Multibody Dynamics With Application to Train Vehicles”, Int. J. of Num. Meth. in Engng., 40, 655-
687, (1997).
[2] Carvalho, A. V., “Managing the Safety Passengers and Crew in Case of Collision, The SAFETRAIN
Project – Train Crashworthiness for Europe”, AIC – 4t h Annual Conference, Railway Safety in
Europe, London, UK, (1998).
[3] Carvalho, A.V., “Safetrain and Safetram Project: Results and Objectives”, In Passive Safety of Rail
Vehicles Symposium”, TUB, Berlin, Germany, Nov, 23, (2000).
[4] Pereira, M., Milho, J., Ambrósio, J.A., “Analysis of Overriding Phenomena in Train Collisions”, In
Passive Safety of Rail Vehicles Symposium, TUB, Berlin, Germany, Nov, 23, (2000).
[5] Pereira, M., “LRV Statistics – Passive Safety in Urban Railway Vehicles. Light Rail Vehicle (LRV)”,
[6] Pereira, M.., Feasibility of Design Collision Scenarios for Passive Safety in Trams, Proceeding of the
3rd International Symposium “Passive Safety of Rail Vehicles. Berlin (2002).
[7] Wolter, W., “SAFETRAIN test results-Appropriate Solution for DB’s Operator Requirements?”. In
Passive Safety of Rail Vehicles Symposium, TUB, Berlin, Germany, Nov, 23, (2000).
[8] “ERRI B 205.1/[RP 1] Analysis of Collision Accidents”, December (1997).
[9] “EN 12663 Railway applications – Structural requirements of railway vehicle bodies”; (2000)
[10] “Statistical and risk analysis of accidents for periurban trams, SAFETRAM Report
06TR_SNCF_WP1/final”, February (2002).
[11] “SAFETRAIN Conclusions and Recommendations. Public Report, September (2001).
[12] “SAFETRAM Final Publishable Report. Public Report, (2005).
[13] BOStrab Verordnung über den Bau und Betrieb der Straßenbahnen (Prescription for design and
operation of tramways); 1987
[14] DIN 5560 Schienenfahrzeuge – Längsfestigkeit der Fahrzeugkästen von Leichttriebwagen (Rail
vehicles – Longitudinal strength of bodies of light rail vehicles); Draft 2001
[15] prEN 15227:2005 (E), Railway applications — Crashworthiness requirements for railway vehicle
bodies.