Train designers are now using the results of this project on
how train structures perform in crashes in their latest designs. The
project was in part a response to the growing insistence of buyers
on safer trains. Unlike in the car industry, there are no European
safety regulations covering vehicle design in the railway industry.
The partners have developed calculation tools that allow faster preparation
of calculation models than is possible with finite element analysis
packages already on the market. They have also decided to collaborate
further in order to define detailed impact absorption criteria that
could be incorporated in safety standards.
There is little understanding of how the
different parts of train structures actually stand up to crashes,
and there have been few detailed international safety standards
that train operators can specify to the builders of their vehicles.
All countries conform to the recommendations of the International
Union of Railways on crashworthiness, but they adopt different structural
specifications which all pay little attention to the effects of
one train riding over another in a crash, one of the main causes
of passenger injuries. Trains have in most cases been designed to
crumple in a crash, to dissipate the enormous amount of energy released
by the impact. This has meant that trains are usually very unfriendly
places for people in crashes.
Railway companies are, however, increasingly insisting on designs
to high safety standards, and various proposals have emerged in
recent years for how trains should stand up to impacts. For example,
British Rail proposed that the energy of a collision should be absorbed
first by deformation of the end wall of a vehicle before other parts
deform. It also proposed that vehicles would not ride over each
other in a crash and that any collapse would be in areas not likely
to be used by passengers or crew.
The Portuguese trainmaker Sorefame initiated the Brite-Euram project
TRAINCOL to develop a better understanding of the effect of crashes
on vehicle structures, as well as to develop design methodologies
for safer carriages. An important aim of the project was to produce
proposals for better regulations that would minimise passenger injury
The TRAINCOL project involved some of the leading railway engineering
organisations in Europe. The work involved the use of computer aided
design tools, as well as an assessment of how passengers would be
affected if particular designs were involved in crashes. Such work
has been done already for road vehicles such as coaches that, like
trains, are built from sheet metal panels fitted to supporting frameworks.
One of the project partners, Cranfield Impact Centre, has a large
amount of experience in the design and analysis of road coaches,
and was able to transfer that knowledge to its TRAINCOL work. Ensuring
that the computer models were valid required data from real crush
tests. The partners therefore built train parts, sub-assemblies
and complete carriages and carried out static and dynamic tests.
There were full scale and model tests. An important part of the
TRAINCOL project was the static testing of typical end frame components
to determine crush and hinge characteristics of components. The
partners tested six types of carbon steel profiles under different
loading conditions, six substructures with different geometries
and made of different materials under axial crushing, and full scale
end underframes under axial crushing.
There was good correlation between the theory and the test results.
This has allowed the setting up of a database of mechanical characteristics
of profiles for use with different software tools. The partners
also learned more about how well different components and sub-assemblies
dissipate energy through plastic deformation, and how good the results
of reduced scale modelling are.
There were two crash tests of full scale vehicles, one in Portugal
at the facilities of the train operator CP and the other in France
at the facilities of the train company SNCF. The crash tests allowed
the researchers to monitor effects such as displacement and strain,
using techniques such as high-speed filming.
New dimension to design
IST, Instituto Superior Técnico of the Technical University
in Lisbon made a very important contribution to the project. The
Department of Mechanical Engineering has extensive research experience
in crashworthiness analysis and tooling development for aided design
of vehicle structures under impact conditions, along with active
involvement in the field of computational dynamics of rigid and
flexible mechanical systems (FEM and Multi-body dynamics).
The researchers applied a wide range of software tools to their
analyses, from 1D to 3D models. In the case of 1D analysis, the
aim was to minimize peak accelerations in a frontal collision while
allowing deformations that would maintain a minimum survivable passenger
space. Several design proposals have been produced.
The 2D analyses allowed rail vehicle underframes to be modelled
by assemblies of rigid and flexible bodies with degrees of freedom
in longitudinal, transversal and rotational motion. Rigid bodies
and flexible members formed the basis of the 3D models that were
used to assess the overall behaviour of the ends of rail vehicles.
Both the 2D and 3D models were found useful in speeding up the initial
stages of designing structures, but because they are based on a
large amount of simplification they are not suitable for detailed
design work. The partners also developed an algorithm for design
of both linear and non-linear structures, taking into account factors
such as stress, displacement and eigen value.
The module can be used with commercially available finite element
analysis codes as an alternative to existing trial and error systems,
say the Traincol partners. The design sensitivity analysis code
was based on an object oriented approach and implemented in C++
programming language. Tests of the software showed a good correspondence
with various structures, and design improvements were found to be
possible after just a few analysis steps.
To relate train design to passenger safety, the partners developed
a model for what happens to passengers' bodies and their likely
injuries should there be a crash. The model used various types of
seats and dummy passengers. The injury modeling allowed for secondary
impacts, such as passengers hitting the backs of the seats ahead
On track for improvements
Overall the project produced design approaches that can substantially
reduce the amount of time required to design safe structures. The
partners nevertheless recommend that designers carry out detailed
studies of simple elements, including possible tests of components,
to ensure acceptable results. But use of the computer models should
reduce the amount of testing needed for any design.
Some of the partners have decided to work together on developing
crash test methods and design approaches to ensure the safety of
drivers and passengers. The four-year project should begin in 1997.
But the results of the first project are already being used in the
design of trains. Sorefame, for example has used some of the techniques
in the design of trains for the Central Line project in Lisbon.