Title: Rail metallurgy and inservice performance
1Rail metallurgyand in-service performance
International Union of RailwaysTehran - December
2003
- Daniel Boulanger, Corus Rail
2Rail functions
- Wheel support and guidance
SAFETY
QUALITY
SAVING
3Traffic effects
- Static loads
- Vehicle weight
- Quasi-static loads
- Centrifugal force, centering force, cross winds
- Dynamic loads
- Track irregularities geometry, stiffness,
discontinuities (welds, joints), rail running
surface (corrugation, ballast prints), vehicle
defects (wheel flats, vibrations, hunting) - Contact stresses
- Acceleration and braking areas
- Thermal stresses
RAIL
Internal Fatigue Wear Plastic Flow Rolling
Contact Fatigue
BREAKS
Risks
DERAILMENTS
4Corus Rail500 000 tons of rails per year
80 profiles and 25 steel grades
5Track specificities
80 profiles and 25 steel grades
6Rail characteristics
- Qualifying tests
- Fracture toughness
- Fatigue crack growth rate
- Fatigue test
- Residual stress
- Variation of centre line running surface hardness
of HT rails - Tensile strength and elongation
- Segregation
- Acceptance tests
- Chemical composition
- Microstructure
- Decarburisation
- Oxide cleanness
- Sulfur prints
- Hardness
- Tensile tests
- Dimension tolerances
- Straightness, surface flatness and twist
- Internal quality
- Surface quality
7Rail abilities
1960
1980
8Internal fatigue
RAIL PRODUCTION
Basic Oxygen Converter Low aluminium killing
process Vacuum degassing Continuous
casting Ultrasonic inspection
Gauge corner shelling
Tache ovale
9Corus Rail steel production
10Ultrasonic inspection
- European project
- ³70 Head
- ³ 60 Web
- Foot
- Calibration
- Holes Æ 2 mm
11Wear and plastic flow
RAIL PRODUCTION
Heat treatment
12Wear vs. hardness for perlitics
13Rail failure evolution
RAIL PRODUCTION
INTERNAL FATIGUE
Basic Oxygen Converter Low aluminium killing
process Vacuum degassing Continuous
casting Ultrasonic inspection Heat treatment
NEW GRADES
RAIL CONTACT FATIGUE
NEW MAINTENANCE
WEAR
1980 / 1990
14Rolling Contact Fatigue
- Surface damage is caused by high tangential loads
or / and high local friction coefficient - Squat and head check are initiated on the running
surface of the rail in shear mode then develop in
a sublayer according to the local stress field - The material of the rail running surface presents
a high degree of cold hardening, together with
plastic flow
15Squat
16Head check
17Steel grades vs. wear
721R Draft
350LHT
350HT
260
260Mn
320Cr
18Main Corus Rail steel grades
Bainitics
Tramway
Arema
UIC
Euronorm
19Steel grades vs. RCF?
?
MHH
B320
B360
350LHT
ROLLING CONTACT FATIGUE
20RCF test areas
- Squat
- SNCF, Reims region (Oeilly), tangent track,
references 700 900A, 1993 - Studied parameters steel hardness,
decarburisation, grinding - Head checking
- SNCF, Bordeaux region (Saint Benoit), radius 985
m, reference 900A, 1997 - Studied parameters steel hardness, wear
- SNCF, Montauban region (Dieupentale), radius 1140
m , reference 900A, 1998 - Studied parameters steel structure (perlitic vs
bainitic) - SBB, Basel region (Frick), radius 449 m,
reference 350LHT, 1999 - Studied parameters steel structure (perlitic vs
bainitic) - ProRail, Utrecht region (Bunnik), radius 2275 m,
reference 260Mn, 2000 - Studied parameters steel hardness, wear
21SNCF - Oeilly
22White layer on standard grades
900A / 9
- The white layer causes the cracking
- The cracks sink into the base metal as soon as
the critical thickness of the white layer is
reached
900A / 9
23 Development of white layer
DISSOLUTION OF THE CEMENTITE LAMELLA
900A
No differentiation between 700, 900A and 900A
highly decarburised rails
DEVELOPMENT OF A STRUCTURE WITH NIDDLES
24White layer on HH grades
900A HH / 56
- In spite of a fatigue limit that is higher than
that of the grade 900A, the treated grades react
far less well on the track - After the cracking of the hard white layer, the
base metal is stressed on severe cuts to which it
is less resistant than the grades with a lower
yield strength
25SNCF - Saint Benoit
26"Hairy" head checking for MHH
Initial hardness
- MHH
- Immediate slight cracks
- No development afterwards
- Very low cold hardening
27Profile evolution
800
900A
900X
Profile vs. position
MHH
Plastic flow
Profile vs. grade
28No magic natural wear
Deep penetrant tests
29ProRail - Bunnik
30Head checkingaspect and dimensions
31Corus Rail MHHUltra fine pearlite
No transition zone
x8000
- Accurate heating of the whole section by
induction - Air cooling of the whole head
32High hardness and ductilityOptimised residual
stresses
33The Corus Rail bainitic grades
- Low Carbon base, more ductile and readily
weldable steel - Radically different microstructure embrittling
carbide phase replaced with soft and ductile
austenite - As-rolled, i.e. no heat treatment required
B 320 bainite residual austenite
34SNCF - Dieupentale
35Hardness and wear evolution
36Head checking on 900A
37Good weldability of the bainitics
- Expansion coefficient
- 11,8.10-6 m/mC at 20C
- 12,2.10-6 m/mC at 60C
- Resistivity
- 39 µW.cm (25 for 900A)
38SBB - Frick
39Head checking on 350LHT
40No more wear for bainitic
Mixed welds
41Conclusion
- Changes in the steel manufacturing process and in
the rail inspections have eliminated
manufacturing defects as a major cause of
replacement - Heat treated rails are available to address the
wear issue - RCF is now the principal cause of rail failure
- In alignment,
- limited hardness grades and smooth grinding can
prevent the white layer formation and avoid the
crack propagation in the base material - In curve,
- the low carbon bainitic steels show very
encouraging results, but adjustments are
essential to carry the experiment further - MHH gives the best results against wear and RCF.
The stress distribution limits the crack
propagation and makes it possible to grind less
often