The ALJOIN project was a European Union funded research project running from 2002-2005 that aimed to improve the crashworthiness of welded joints in aluminium rail vehicles. It brought together experts from industry and academia in welding, materials, and modelling to address the issue of "weld unzipping" identified as contributing to a fatal rail accident in 1999. The project objectives were to test aluminium alloy welds under static and dynamic loads, investigate alternative welding techniques, develop constitutive modelling of materials, and simulate crashworthiness through numerical modelling to solve the problem of welds catastrophically failing during accidents.

1. Case
study
-­‐
Project
“ALJOIN”
Crashworthiness
of
joints
in
aluminium
rail
vehicles|
George
Kotsikos
NewRail
|
Newcastle
University
|
UK.

2. Content
•
•
•
•
Background
Development
of
the
ALJOIN
Project
Project
descrip@on
Lessons
learnt

3. Background
• Aluminium
alloys:
Lightweight,
corrosion
resistant,
weldable.
• Earliest
applica@ons
of
aluminium
in
rail
coach
design
in
1935.
• First
aluminium
monocoque
bodyshells
were
designed
in
the
70s
(APT,
TGV)
• Large
closed
cell
extrusions.
– Design
versa@lity
– Superior
surface
finish
– Superior
collapse
strength
and
impact
resistance
in
the
longitudinal
direc@on

4. Background
The Ladbroke Grove Accident
5th October 1999, 21 fatalities and 400 injured

5. Background
‘…the
aluminium
extrusions
had
fractured
along
the
weld
lines
and
there
was
a
lack
of
plas6c
deforma6on
(…)
the
structure
appeared
to
have
failed
along
the
welds
rather
than
deforming
in
a
controlled
manner’
Cullen
Report
The
catastrophic
failure
of
welds
in
this
manner
is
a
phenomenon
known
as
‘weld-­‐
unzipping’.

6. Background
•
•
Weld
unzipping
-­‐
dynamic
duc@le
tearing
of
the
weld
metal
or
heat
affected
zone
is
known
to
materials
engineers.
The
process
is
controlled
by:
–
–
–
–
–
•
•
Geometry
of
applied
stresses/crack
trajectory
Plas@c
deforma@on
at
the
crack
@p
Material
composi@on
Impuri@es
Microstructure
Fusion
welding
can
be
cri@cal
in
localising
failure
as
it
affects
microstructure,
mechanical
proper@es
and
can
introduce
defects.
Aluminium
alloys
are
sensi@ve
to
heat
input
introduced
by
the
fusion
welding
processes.
– Proof
strength
of
6005A
HAZ
~
50%
parent
plate

7.
DEFINITION
OF
OBJECTIVES
AND
PARTNER
SELECTION

8. Project
development
• A
detailed
research
programme
of
work
was
necessary
to
provide
a
solu@on
to
eliminate
weld
unzipping
in
aluminium
rail
vehicles.
• How?
(defini@on
of
intermediate
objec@ves)
– Provide
physical
evidence
of
the
energy
absorp@on
capability
of
aluminium
alloy
welds
by
tes@ng;
– Assess
the
adequacy
or
inadequacy
of
current
design
and
construc@on
prac@ces
of
aluminium
alloy
welds
in
the
context
of
crashworthiness
– Inves@gate
alterna@ve
welding
techniques
and/or
joint
designs
for
improved
impact
performance
of
aluminium
alloy
joints;
– Development
of
the
material
cons@tu@ve
modelling
for
the
parent
material
and
the
welds;
– Numerical
modelling
of
simple
joints
subjected
to
quasi-­‐sta@c
and
dynamic
loads;
– Develop
numerical
modelling
techniques
for
simula@on
of
crashworthiness
of
full
rail
vehicles

9. Project
development
• Partner
selec@on
– Can
it
be
done
by
a
single
organisa@on?
NO
– Wide
ranging
exper@se
required?
YES
– Is
industry/end
user
involvement
important?
YES
• Rail
coach
manufacturer(s)
• Materials
supplier(s)
– Welding/joining
specialists
required?
– Modelling
experts
required?
– Academic
exper@se
required?
– Specialist
test
facili@es
required
YES
YES
YES
YES

10. Project
development
• FP5
-­‐
FP5-­‐2002-­‐GROWTH
–
CompeUUve
and
Sustainable
Growth
• Key
AcUon
3,
Land
Transport
and
Marine
Technologies
• Strategic
Objec@ves
– Improved
fuel
efficiency
and
reduced
emissions
-­‐
cukng
CO2
emissions
and
developing
and
valida@ng
zero-­‐emission
vehicles.
– Improved
performance
-­‐
increasing
safety,
reliability,
maintainability,
availability,
operability,
energy
efficiency
and
adaptability.
– Improved
system
compe@@veness
-­‐
reducing
both
@me
to
market
and
development
costs.
• Priority
2:
Technology
integra@on
and
valida@on
– Research
will
focus
on
integra@ng
and
valida@ng
six
technology
plamorms:
• New
land
transport
vehicle
concepts;
enhanced
systems
efficiency,
Advanced
concepts
for
ships
and
vessels;
compe@@ve
shipbuilding,
Enhanced
design
and
manufacturing
for
road
vehicles,
Sustainable
and
modular
trains,
Safe,
efficient
and
environmentally
friendly
vessels
and
plamorms,
Efficient
interoperability
and
transhipment.

11. Project
development
Project
acronym:
ALJOIN
FP:
5
Project
Reference:
G3RD-­‐CT-­‐2002-­‐00829
Call
iden@fier:
FP5-­‐2002-­‐GROWTH
Total
Cost:
€
2,177,806
EU
Contribu@on:
€
1,200,036
Timescale:
08/2002
–
08/2005
Project
Partners
:
Country
• D’Appolonia
SPA
I
• NewRail
–
Newcastle
University
UK
• Bombardier
Transporta@on
F
• Dans@r
DK
• The
Welding
Ins@tute
(TWI)
UK
• Alcan
CH

12.
ALJOIN
PROJECT
-­‐
OVERVIEW

13. ALJOIN
project
overview
• The
project
concept
cons@tuted
of
5
dis@nct
parts:
– Mechanical
characterisa@on
of
joints
(MIG,
Laser
MIG,
FSW,
bonded,
bolted)
– Fracture
mechanics
analysis
– Impact
tests
on
full
size
components
– Solu@on
development
and
valida@on
– Modelling
and
collision
simula@on

14. Mechanical
property
characterisa@on
• Material
supplier
partner
manufactured
full
scale
closed
cell
extrusions
for
assessment.
• Same
extrusions
also
used
for
impact
tests.
• MIG
and
Laser
MIG
welded
extrusions
were
produced
with
two
types
of
filler
wire;
Al-­‐Si
(exis@ng
consumable),
Al-­‐Mg
(proposed
alterna@ve
consumable).
• Bonded
Bolted
and
FSW
extrusions
used
slightly
modified
type
of
extrusions.
MIG
welds
FSW

15. Mechanical
property
characterisa@on
Academic
partner
undertook
material
characterisa@on
work
•
120
300
0.2% PS
100
Parent plate hardness range (HV)
Stress (MPa)
Hardness (HV)
80
60
40
UTS
250
Weld metal
HAZ
200
150
100
HAZ
50
20
6005T6 - 4043 filler (Al-Si)
0
6005T6 - 5356 filler (Al-Mg)
PP
WM Al-Si
WM Al-Mg
HAZ Al-Si
HAZ Al-Mg
0
-2
-1.5
-1
-0.5
0
0.5
1
1.5
2
Normalised distance from weld centreline
•
•
Use
of
Al-­‐Mg
filler
produces
welds
with
improved
mechanical
proper@es
over
Al-­‐Si
filler
Hardness
varia@on
across
the
weld
is
similar
Material
Charpy
impact
energy
(J)
Parent
plate
0.73
Weld
–
(Al-­‐Si
filler)
0.32
Weld
–
(Al-­‐Mg
filler)
0.56

16. Fracture
mechanics
PM
200
180
160
140
J (N/mm)
Sta@c
and
dynamic
J-­‐R
curves
were
obtained
through
SENB
tests
for
the
parent
material
and
weld
region
120
100
80
60
40
20
15mm
5mm
Static
Dynamic
0
0
50mm
1
2
3
Da (mm)
4
5
6

17. Fracture
mechanics
• Fracture
mechanics
tests
using
a
modified
SENB
specimen
design
and
a
CCT
specimen
were
• The
analysis
used
the
Energy
Dissipa6on
Rate
(EDR)
approach.
• The
tests
provided
informa@on
such
as
tearing
resistance
index
(T)
and
cri@cal
CTOA
and
were
intended
to
aid
with
numerical
modelling
of
tearing
process
in
the
welded
aluminium
extrusions.

18. Fracture
mechanics
Al-­‐Mg
filler
D=1170
kJ/m2
T=
0.03
Al-­‐Si
filler
D=350
kJ/m2
T=
0.011

19. Full
scale
impact
tests
lever arm
• Dynamic
tear
tests
on
full
scale
welded
extrusions
were
carried
out
by
the
industry
partner’s
facility.
Welded Al extrusion
Clamped to rigid base
air cannon
Test n° 61 of specimen n° 358_0005_099 /8T6/4,2/Mg/1
Date : 05/10/2005
Force 2+3 (N)
Energy absorbed 2+3 (J)
3000
250000
2500
200000
2000
150000
1500
100000
1000
50000
500
0
10
20
30
40
50
-50000
60
70
80
90
0
100
-500
Displacement (mm)
Energy absorbed (J)
3500
300000
Force 2_3 (N)
350000
0
projectile

20. Full
scale
impact
tests
Laser MIG
FSW
MIG
All welds failed by weld unzipping!

21. Solu@on
approach
• What
mauers
when
a
collision
takes
place
is
whether
the
structure
spreads
the
impact
energy
or
concentrates
it
in
a
specific
region
(the
weld
in
this
case).
• Fusion
welding
in
aluminium
alloys
results
in
an
“undermatched
weld”.
• The
impact
energy
in
a
structure
with
a
strength
undermatch,
may
channel
all
the
energy
to
the
weld
region.
This
will
have
to
be
taken
up
by
the
energy
dissipa@on
rate
which
implies
extensive
crack
growth.

22. Solu@on
approach
• Reducing
the
heat
input
– Laser
MIG
– FSW
– Bonded
joints
• Change
joint
design
– Altering
weld
geometry
by
thickening
the
pla@ng
at
the
weld
region

23. Solu@on
approach
FSW
MIG
Laser MIG
Failures away from weld with the exception of Laser MIG welds

24. Modelling
weld
failure
• Detailed
mechanical
property
characterisa@on
was
used
for
modelling
ac@vi@es.
• Detailed
FEA
models
were
prepared
and
validated
against
component
tests.
• Code
used
LS-­‐DYNA
• Failure
criteria
used:
– Maximum
strain
failure
model
– Gurson
-­‐Tvergaard
model

25. Modelling
weld
failure
• Modelling
of
tearing
test
–
Max
strain
failure
model

26. Collision
modelling
• A
Finite
Element
model
of
a
class
165DMU
similar
to
that
involved
in
the
Ladbroke
Grove
accident
in
the
UK
was
created
(undertaken
by
research
ins@tu@on
partner
specialising
in
FEA
modelling)
• Simula@on
of
collision
at
20m/s
(72km/h)
on
a
solid
flat
surface.
• The
simula@on
is
repeated
with
the
new
joint
design
and
consumable

27. Train
collision
simula@on
–
Standard
weld
design

28. Train
collision
simula@on
–
Standard
weld
design

29. Train
collision
simula@on
–
modified
joint
(sec@on
thickening)

30. Train
collision
simula@on
–
modified
joint
(sec@on
thickening)

31. Project
output
• ALJOIN
provided
a
solu@on
to
the
problem
of
“weld
unzipping”
for
welded
aluminium
closed
cell
extrusions.
• Contributed
to
the
development
of
two
industry
standards;
– EN
15085
"Railway
applica@ons
-­‐
Welding
of
railway
vehicles
and
components"
– EN
15227,
“Crashworthiness
requirements
for
railway
vehicle
bodies”
• Contributed
to
the
enhancement
of
safety
for
rail
passengers
and
staff.
• The
solu@on
does
not
introduce
a
significant
economic
penalty
to
industry.
• Results
have
a
Europe
wide
(if
not
global)
impact.
• Contributed
to
the
enhancement
of
knowledge
to
academia,
research
ins@tu@ons
and
industry
partners.

32. Lessons
learnt
Project
idea
Clear
defini@on
of
main
objec@ve
and
expected
outcomes
Well
structured
work
programme
with
clearly
iden@fied
deliverables
Defini@on
of
partners
and
their
roles
–
include
essen@al
end-­‐users
of
the
results
Partnership
should
include
“champions”
to
promote
implementa@on
azer
project
end

33. Lessons
learnt
Project
idea
Clear
definition
of
main
objective
and
expected
outcomes
Well
structured
work
programme
with
clearly
identified
deliverables
Definition
of
partners
and
their
roles
–
include
essential
end-­‐users
of
the
results
Partnership
should
include
“champions”
to
promote
implementation
after
project
end

34.
THANK
YOU