UIC CODE 720

2nd edition, March 2005

Original R

Laying and Maintenance of CWR Track

Pose et maintenance de la voie en longs rails souds
Verlegung und Instandhaltung von lckenlosem Gleis
Leaflet to be classified in Volume :
VII - Way and Works

Application :
With effect from 1 January 2005
All members of the International Union of Railways
Improved knowledge of forces in CWR track (including switches)

Record of updates
1st edition, January 1986 First issue

2nd edition, March 2005 Overhaul of Leaflet

Adaptation to the editors guide M1 - New lay-out

The person responsible for this leaflet is named in the UIC Code

720
R
Contents

Summary ..............................................................................................................................1

1- Introduction ................................................................................................................. 2

1.1 - Purpose ................................................................................................................ 2

1.2 - Scope ................................................................................................................... 2
1.3 - Definition............................................................................................................... 3

2- Key parameters ........................................................................................................... 5

2.1 - Rails...................................................................................................................... 5
2.2 - Fastenings ............................................................................................................ 5
2.3 - Type and spacing of sleepers............................................................................... 5
2.4 - Ballast-bed dimensions......................................................................................... 5
2.5 - Alignment.............................................................................................................. 6
2.6 - Track geometry..................................................................................................... 6
2.7 - Important rail temperatures................................................................................... 6
2.8 - Additional temperature characteristics.................................................................. 7

3- Additional parameters ................................................................................................ 9

3.1 - Subgrade .............................................................................................................. 9

3.2 - Vehicle parameters............................................................................................... 9
3.3 - Maintenance work................................................................................................. 9
3.4 - Bridges ............................................................................................................... 10
3.5 - Tunnels, gradients and braking .......................................................................... 10
3.6 - Switches ............................................................................................................. 11
3.7 - Variation of gauge .............................................................................................. 11

4- Installation of CWR track.......................................................................................... 12

4.1 - Preconditions ...................................................................................................... 12

4.2 - Initial temperature within the installation temperature range .............................. 12
4.3 - Initial temperature below the installation temperature range .............................. 12
4.4 - Initial temperature above the installation temperature range.............................. 13

720
R
 4.5 - Measures to eliminate friction ............................................................................. 13
4.6 - Stressing records................................................................................................ 13

5- Maintenance .............................................................................................................. 14

5.1 - General principles............................................................................................... 14

5.2 - Range of temperatures within which work may be undertaken .......................... 14
5.3 - Measures to ensure track safety......................................................................... 15

6- Elimination of problems ........................................................................................... 17

6.1 - Rail breaks.......................................................................................................... 17

6.2 - Track buckling and distortion caused by heat..................................................... 18

Appendix A - Safety ........................................................................................................... 20

A.1 - Safety fundamentals / philosophy....................................................................... 20
A.2 - Safety concepts and criteria ............................................................................... 21
A.3 - Safety Evaluation and Safety Limit Charts ......................................................... 23
A.4 - CWR Buckling Safety Based on Risk Methodology -
Extension to risk-based safety assessment........................................................ 24
A.5 - Special Track Equipment: Switches, Turnouts and Approaches to Bridges....... 25
A.6 - Safety Criteria for High-Speed Applications: Track Lateral Shift ........................ 25
A.7 - Figures................................................................................................................ 28

Appendix B - Case studies................................................................................................ 34

B.1 - Use of CWERRI and CWR-BUCKLE Models ..................................................... 34
B.2 - Example 1: High-speed and main-line tracks ..................................................... 36
B.3 - Example 2: Secondary and freight-line tracks .................................................... 39
B.4 - Application of results........................................................................................... 42
B.5 - General remarks ................................................................................................. 42

Bibliography .......................................................................................................................43

List of figures .....................................................................................................................45

720
R
Summary

This leaflet contains guidelines for the use, installation and inspection of ballasted track with
continuous welded rails (CWR), and also lists the associated safety criteria. It is based on the practical
experience of several railways and a large number of tests and theoretical studies carried out over a
period of more than 40 years. It also includes the results of studies undertaken by ERRI Committee
D 202.

The leaflet has been enhanced by inclusion of a point on safety aspects and a point on case studies
conducted using the CWERRI and CWR-BUCKLE models. A new safety criterion has been formulated
on the basis of maximum and minimum buckling temperatures (Tb,max and Tb,min). The parameters
were selected to reflect a maximum temperature which corresponds to 50% of the buckling energy
level. In addition, a concept of buckling safety based on risk assessment methodology has been
introduced.

The case studies include examples of high-speed, main-line passenger tracks on both concrete and
wooden sleepers, together with secondary and freight tracks.

Permissible rail temperatures for the various scenarios studied are presented graphically.

1 720
R
1 - Introduction

1.1 - Purpose

This leaflet contains guidelines for the installation and use of ballasted track incorporating continuous
welded rails (CWR), together with details of inspection methods and safety criteria. It is based on the
practical experience of several railways and a large number of tests and theoretical studies
undertaken over more than 40 years. It also includes the results of international work on this topic
carried out by ERRI Committee D 202. This leaflet replaces UIC Leaflet 720, 1st edition, dated 1986
on the same topic.

1.2 - Scope

1.2.1 - Use of CWR track

CWR is now in widespread use on most railways, since, although the construction costs of CWR are
similar to those of jointed track, CWR has important technical, ecological and economic advantages,
in particular, a significant reduction in maintenance costs.

CWR is also a prerequisite for high-speed services.

1.2.2 - Advantages of CWR track

The main advantages of CWR track are:

- reduced maintenance costs,

- fewer rail defects and failures,

- reduced wear on vehicles,

- reduced noise and vibration emissions,

- better ride comfort,

- lower energy costs for traction,

- easier mechanisation of track installation and maintenance.

All these features translate into reduced life cycle costs compared to jointed track.

1.2.3 - Consequences

The use of CWR requires precautions in the design of the track structure, and in its installation,
maintenance and inspection. It also requires an understanding of the effects of temperature, rolling-
stock and track features, on track stability and rail failures.

2 720
R
1.3 - Definition

1.3.1 - CWR track

CWR track is track in which the free expansion and contraction of the rails caused by temperature
changes are constrained, by artificially introducing a controlled expansion into the rails prior to
installation of the track fastenings.

Consequently high longitudinal forces are induced in the rails as a result of temperature changes.

Resistance to longitudinal movement of the rails, under the action of these forces, is secured through
the clamping action of the rail fastenings and the restraining effect of the boxing ballast.

Total longitudinal force P in the track, either tensile or compressive, is given by:

P = 2EAT

where: = coefficient of expansion of steel

E = Young's modulus of steel
A = rail cross-sectional area
T = change in temperature from the "stress-free temperature"

The longitudinal stresses in each rail are given by:

= ET

The change in temperature from the "stress-free" condition can reach values as high as 40-45C
resulting in rail forces of 800 kN and rail stresses of up to 100 N/mm2 in UIC 60 rails.

1.3.2 - Aim of the leaflet

The aim of the leaflet is to explain the parameters affecting track stability and to set out the
methodology for ensuring the stability and safety of CWR track.

3 720
R
1.3.3 - Structure of the leaflet

The leaflet is structured around the following points:

2. Key parameters Effect of track structure, layout, geometry and temperature on

track stability.
3. Additional parameters Effect of rolling stock, bridges, switches and crossings,
substructure, etc.
4. Installation of CWR track Requirements for installation.
5. Maintenance Requirements for maintenance, inspection and repair.
6. Elimination of problems Track buckling and distortion.
Appendix A. Safety Methodology for risk assessment and establishment of safety
criteria.
Appendix B. Case studies Case studies using mathematical model CWERRI: examples
and application of results.

4 720
R
2 - Key parameters

2.1 - Rails

All flat-bottom rail profiles are suitable for CWR track.

It is recommended that intermediate section rails are introduced where it is required to join rails having
cross sectional areas which differ by more than 22%.

New rails should not be drilled.

It is recommended that serviceable rails undergo ultrasonic examination prior to CWR conversion.

Existing fishbolt holes should preferably be removed prior to CWR conversion but this is not essential,
particularly in lower-speed and lower-tonnage track categories.

2.2 - Fastenings

Only fastening systems which provide adequate guaranteed longitudinal and torsional resistance to
rail movement relative to the sleepers, shall be used in CWR.

The end of a CWR section of track over which rail movement can occur is called "the breathing length".

Expansion switches can be fitted at the end of a section of CWR track where it abuts an adjoining
section of jointed track or at a specific location such as a bridge, etc.

2.3 - Type and spacing of sleepers

Concrete, wooden or steel sleepers are suitable for use in CWR. However, the type and shape of the
sleeper can have a significant effect on the lateral resistance and stability of the track.

The mass of the sleeper, the surface area of the sleeper end and the sleeper spacing, have a major
effect on the lateral resistance. Steel sleepers require a particular tamping technique to ensure that
the ballast penetrates the hollow underside of the sleeper completely.

2.4 - Ballast-bed dimensions

Satisfactory longitudinal and lateral resistance of the track depends on the mass, particle size, shape,
cleanliness and compaction of the ballast.

The particle size shall be within the range 22,4/63 mm and the individual particles shall be sharp-
edged and uniformly distributed.

The recommended minimum depth of ballast below the sleeper soffit is 300 mm, and not less than
200 mm.

The recommended minimum width of ballast shoulder, i.e. the horizontal distance between the end of
the sleeper and the slope of the ballast, is 400 mm for speeds up to 160 km/h and 500 mm for higher
speeds. The top of the ballast shoulder shall be at least level with the top of the sleeper.

5 720
R
The width of the ballast shoulder may, with advantage, also be increased to the higher value in sharp
radius curves or where tilting trains operate.

The lateral resistance of the track can be increased to a limited extent by widening and raising the level
of the ballast at the sleeper ends.

The lateral and longitudinal resistance of the track depends to a large extent on the degree of
compaction of the ballast, being greatest when the ballast is fully compacted and least when the
ballast is freshly tamped (see also point 3.3 - page 9).

The depth of the ballast and the width of the ballast shoulder may, during installation, fall below the
above recommendations temporarily, provided high ambient temperatures can be guaranteed not to
occur, or a temporary speed restriction is imposed. However, the full ballast dimensions should be
obtained prior to commencement of operating at line speed.

2.5 - Alignment

The quality of the alignment has a significant effect on the lateral resistance of CWR track. A greater
degree of misalignment is usually found in curves compared with straight track.

The minimum curve radius in which CWR track may be safely installed depends on the rail section,
the fastenings, the type of sleeper, the ballast profile and the anticipated longitudinal thermal stresses.
CWR may be installed on curves sharper than 300 m radius provided special measures, such as fitting
anchoring devices to the sleepers, are taken.

2.6 - Track geometry

Major errors in alignment can lead to track buckling at high rail temperatures. It is essential therefore
that alignment errors should be minimised by correct maintenance procedures. The magnitude of the
maximum permissible alignment error in any section of CWR track depends on the rail section, sleeper
type, fastening type, ballast condition, stability of the ballast bed and the prevailing temperature.

2.7 - Important rail temperatures

The following rail temperatures must be known prior to the installation of CWR track:

2.7.1 - Mean rail temperature

The mean rail temperature T m is the arithmetic mean between the maximum rail temperature T max
and the minimum rail temperature T min .
T max + T min
T m = -------------------------------
-
2

NB : These temperatures are not generally the same as the maximum and minimum air
temperatures.

6 720
R
2.7.2 - Nominal rail temperature

The nominal rail temperature T s is the optimum temperature at which the track should be fastened
down. It should be several degrees above the mean rail temperature in order to limit the magnitude of
compressive stresses.
T s = T m + X (where X ranges between 0 and 10C)

2.7.3 - Notional temperature

The notional rail temperature is the temperature of the rails prior to the commencement of work.

2.7.4 - Installation temperature range

The installation temperature range is the nominal temperature 3C (for example).

2.7.5 - Installation temperature

The installation temperature T f is the specified equivalent rail temperature at which the fastenings
shall be tensioned. It is, by definition, within the range specified in point 2.7.4.

2.7.6 - Initial temperature

The initial temperature is the rail temperature immediately prior to the commencement of stressing the
rails.

2.7.7 - Stress-free temperature

The stress-free temperature T n is the temperature at which there are no thermal stresses in the rails.

2.7.8 - Maximum temperature

The maximum rail temperature T max is the maximum expected in a particular location.

2.7.9 - Minimum temperature

The minimum rail temperature T min is the minimum expected in a particular location

2.8 - Additional temperature characteristics

The following additional temperature characteristics may be known prior to commencement of work
and may be laid down in rules or calculated from mathematical modelling:

2.8.1 - Maximum Critical Temperature

The maximum critical temperature T b ,max is the predicted rail temperature from mathematical
modelling at which track buckling is almost a certainty.

7 720
R
2.8.2 - Minimum Critical Temperature

The minimum critical temperature T b ,min is the predicted rail temperature from mathematical
modelling below which track buckling is unlikely to occur.

2.8.3 - Allowable Temperature

The allowable temperature T all is the temperature which provides a margin of safety against a
predicted likelihood of track buckling.

2.8.4 - Increment

Increment T is an incremental margin from the allowable temperature, either up or down, which
provides either a lesser or greater margin of safety against buckling.

8 720
R
3 - Additional parameters

3.1 - Subgrade

The characteristics of the subgrade shall be such as to ensure the long-term stability of the CWR track.

On soft ground (e.g. peat), the technical feasibility of using CWR track shall be established and, if
possible, special precautions shall be taken. It is not recommended to install CWR in locations which
are affected by subsidence (e.g. coal mining areas, etc.) since large uncontrolled stresses or
deformations can occur in the track.

The following subgrade characteristics are of particular importance:

- good stability, little likelihood of subsidence, unaffected by frost;

- adequate bearing capacity for the type of track and traffic carried;

- good drainage.

3.2 - Vehicle parameters

The following vehicle characteristics may affect the stability of CWR track:

- Axle load, wheel base and bogie centres may affect both lateral bending of the track and the
precessional uplift wave. The resistance to lateral displacement may be affected by the stiffness
and mass of the track panel.

- The transverse forces ( Y ) generated by the vehicles should be less than the lateral resistance
of the track in the loaded state in order to prevent displacement of the track under the moving train
(for more information, refer to UIC Leaflet 518).

- Heating of the rails by, for example, eddy-current brakes may affect the stability of the track.

3.3 - Maintenance work

(See also point 5.3 - page 15).

Consolidation of the ballast will be necessary after tamping in order to restore the full lateral resistance
of the track. This may be achieved naturally by the action of traffic or by other measures e.g. use of a
dynamic track stabiliser.

During the period of reduced lateral resistance of the track, imposition of a temporary speed restriction
may be necessary until the track is sufficiently stable, particularly when it is anticipated that high
temperatures will occur immediately after tamping.

9 720
R
3.4 - Bridges

The following factors will affect the longitudinal forces in CWR track in the vicinity of bridges:

- changes in the track modulus adjacent to the bridge abutments,

- end rotation of the bridge as a result of the load imposed on the structure,

- changes in the length of the structure caused by temperature changes,

- longitudinal displacement of the structure caused by braking and acceleration.

UIC Leaflet 717 and UIC Leaflet 774-3 provide advice on the effects of forces generated within bridge/
abutment transition zones, on the track alignment and the stability of CWR track. The leaflets also
provide advice on the use of CWR track without expansion joints on bridges.

CWR should ideally be installed on bridges when the bridge itself is in a stress-free condition in order
to minimise the effect of expansion of the bridge on the longitudinal forces in the rails.

Switches should not be located over bridge bearings.

Care should be taken to ensure that the presence of check rails or guard rails does not generate
additional longitudinal forces in CWR track.

3.5 - Tunnels, gradients and braking

In locations where large or uneven changes in rail temperature are experienced, longitudinal
movement of the sleepers in the ballast or of the rails relative to the sleepers may alter the original
stress-free temperature of the track.

The effect will depend on the characteristics of the ballast, the type of sleeper or the creep resistance
of the fastenings.

The longitudinal forces generated by vehicles may produce similar effects.

Such effects may occur particularly:

- in the vicinity of tunnel portals and in deep cuttings,

- on rising and falling gradients,

- in long continuous curves,

- on tracks where eddy-current brakes are in use.

In order to minimise the effect on the stress-free temperature and the consequential reduction in
stability of CWR track, the above situations shall be subjected to particularly careful monitoring.
Appropriate additional measures shall be taken to improve the stability of the track where this is
considered necessary.

10 720
R
3.6 - Switches

The presence of additional rails of the diverted track can create increased longitudinal forces in
switches particularly at, and in advance of the switch toe. The lateral resistance at these locations may
be enhanced using anchoring devices or other measures.

3.7 - Variation of gauge

Variations in gauge have only a minor effect on track stability and can usually be ignored in this
context.

11 720
R
4 - Installation of CWR track

4.1 - Preconditions

Whilst it is permissible to pass track maintenance machines or traffic under a speed restriction through
a site during the construction period, when the ballast profile may not be complete, final fixation of the
track and welding of the rails into CWR must not take place until the longitudinal and lateral resistance
of the track frame is assured.

Stressing and final welding of the rails shall only be carried out once the specified vertical profile and
lateral alignment of the track have been achieved and the required ballast profile has been installed.

4.2 - Initial temperature within the installation temperature range

If the initial temperature is within the installation temperature range with the rails in the unstressed
condition (i.e. immediately after placing on the sleepers), the rails can be finally welded and fastened
down without any special measures being required.

4.3 - Initial temperature below the installation temperature range

If the initial temperature is below the installation temperature range, final welding and fastening down
shall only be carried out after the rails have been artificially lengthened by the application of heat or
by the application of tensile force.

The required increase in length l is calculated according to the following formula:

l = l t

where: l = required increase in rail length

= coefficient of expansion of the rail steel (11,5 x 10-6 C)
l = length of rail to be rendered stress-free
t = difference between installation temperature and initial temperature

This formula is only valid if the rails are free to expand. Therefore prior to stressing, the rail fastenings
shall be loosened and appropriate measures taken to eliminate friction between the rails and sleepers.

The required extension shall be calculated and marked on the rail foot and adjacent sleeper.

Check marks shall also be made at appropriate positions within the anchor length against which the
rails are to be stressed.

The rails shall then be lengthened by the application of heat or tensile force until the desired extension
has been achieved, ensuring that the check marks on the anchor length have not moved during this
process.

The final weld shall be made and after completion, the entire rail length shall be fastened down.

12 720
R
Further rail lengths may be subsequently dealt with in the same way.

4.4 - Initial temperature above the installation temperature range

In these circumstances, the rails should be temporarily installed at the initial temperature, and later,
when the appropriate temperature range is achieved, the rails shall be adjusted in accordance with
point 4.2 or point 4.3 - page 12, as appropriate.

4.5 - Measures to eliminate friction

Prior to stressing the rails, it is not sufficient simply to loosen the fastenings. The friction between rail
foot and sleeper (or rail pad) must be eliminated by using rollers under the rail foot and additionally on
curves, on the side of the rail head. The rails may also be vibrated using heavy plastic hammers.

4.6 - Stressing records

Records shall be kept of all stressing work undertaken, including on those occasions when the initial
temperature range falls within the installation temperature range.

It is recommended that the stressing records contain the following details:

- location, mileage and direction of traffic,

- statement of completion of track-geometry marking and ballasting,

- details of stressing process including:

date and type of welding process,

stressing method,
installation temperature,
initial temperature,
initial rail length,
calculated rail extension,
name of person responsible for undertaking the work,
name of supervisor.

The records shall be maintained indefinitely or until updated at the next stressing operation on the
same site.

13 720
R
5 - Maintenance

5.1 - General principles

Maintenance operations on CWR which affect rail stresses, track-panel stiffness, creep resistance or
lateral resistance of the track, also impact on track stability. Consequently, operations of this type must
only be undertaken within a limited range of rail temperatures.

In extreme situations, prior to undertaking work which may affect the stresses in the rail or reduce the
lateral resistance of the track, i.e. ballast cleaning, sleeper replacement or significant changes to track
geometry, a decision must be taken whether or not temporarily to convert the track back to jointed
track.

5.2 - Range of temperatures within which work may be undertaken

5.2.1 - Installation temperature

The installation temperature determines the basis of the temperature range within which maintenance
work may safely be undertaken. It is therefore important that this is known for each track section.

5.2.2 - Permissible range

Maintenance work which might affect track stability and rail stresses is only permissible within the
following temperature range:

T f b and T f + a

5.2.3 - Limit values

The recommended values of "a" and "b" lie within the range 5 to 15C and may be altered depending
on the type of operation undertaken, the techniques used, the nature of the track, i.e. type of sleeper
or curvature and the particular maintenance machine in use.

If it is necessary to carry out emergency work outside this temperature range, then special precautions
must be imposed to ensure safety.

5.2.4 - Monitoring

The rail temperature shall be monitored and recorded during such maintenance operations, and the
work must cease and all ballasting and tightening of fastenings must be completed before the limiting
temperatures are reached.

14 720
R
5.3 - Measures to ensure track safety

5.3.1 - Change in lateral resistance

Work affecting the ballast can reduce the lateral resistance of the track to 40% of the value of
consolidated track. Compaction of the ballast at sleeper ends can restore 5% to 10% of this reduction.

The initial lateral resistance is restored under the action of traffic. The use of a dynamic track stabiliser
can restore the equivalent of 75 000 to 100 000 gross tons of traffic (metric tons). A significant
proportion of the lateral resistance can be recovered after 100 000 to 200 000 gross tons of traffic, or
after the equivalent consolidation achieved using a dynamic track stabiliser. The actual value achieved
depends amongst other things on the type of sleepers and the maximum speed and axle-load of trains.
The time taken to achieve full lateral resistance again after maintenance work is called the stabilisation
period.

5.3.2 - Traffic speed during maintenance work

Maintenance work carried out on a track may reduce the ballast level and compaction on that or the
adjacent track, and it may therefore be necessary to impose an appropriate speed restriction. The
severity of the speed restriction will be determined by the extent of the work and the anticipated
temperatures. Imposition of speed restrictions can cause disruption to traffic and their use should be
minimised by avoiding maintenance work on CWR track during periods of extreme temperatures.

5.3.3 - Traffic speed after maintenance work

Following maintenance work, if the rail temperature exceeds T f + d before the stabilisation period has
elapsed, a temporary speed restriction should be imposed. The value of "d" should be fixed by each
railway after consideration of the climatic conditions and the nature of the track condition.

The speed restriction may be increased in stages depending on the sleeper type and traffic.

5.3.4 - Inspection

5.3.4.1 - Methods of inspection

CWR track may be inspected in a variety of ways including by foot patrols, from driving cabs or from
the rear of trains, with track inspection coaches or by local measurement.

5.3.4.2 - Frequency

The frequency of inspection will depend on line speed, the volume of traffic, and the type and condition
of the track.

5.3.4.3 - Inspection details

When inspecting CWR track, particular attention should be paid to:

- the completeness of the ballast,

- the alignment, both vertically and laterally,

- the integrity of the fastenings,

15 720
R
- the completeness of any anti-creep devices or other anchorages,

- the appropriate operating speed.

Particular attention should be paid to track sections which are subjected to different levels of exposure
to sunshine during the day, including in cuttings, in front of tunnel portals and on steep gradients, as
these may be particularly prone to longitudinal movement.

If it is suspected that longitudinal movement of the rail is occurring, measurements should be

instigated.

Particular attention should also be paid to special track features such as:

- switches and crossings,

- level crossings,

- bridges with direct fastenings or slab track,

- bridge bearings or expansion joints.

16 720
R
6 - Elimination of problems

6.1 - Rail breaks

Three options are available for the repair of broken rails:

- immediate actions,

- temporary repair,

- permanent repair.

The first two steps may be the same, or only a permanent repair may be undertaken.

6.1.1 - Immediate actions

The immediate actions on finding a broken rail in CWR track are to fit emergency fishplates around
the break and, if necessary, provide support to the rail by means of a block. Depending on the type of
break it may be necessary to impose a temporary speed restriction. In addition the rail fastenings
should be checked and, if necessary, tightened over a distance of at least 30 m. If tightening of the rail
fastenings is not sufficient to restrain the rail, then, in cold climatic conditions, a sufficient number of
anti-creep devices should be placed on each side of the break to limit longitudinal movement. In hot
weather, it may be necessary to cut a joint in the adjacent rail.

6.1.2 - Temporary repair

A temporary repair requires all the work necessary to allow resumption of normal train operations to
be carried out, except for final welding.

It may be necessary to instal a replacement rail fishplated to the existing track. The minimum length
of the replacement rail must be specified by each railway and may depend on line speed. The gap
created by the break shall be measured so that eventually the original length can be restored and the
correct stress condition obtained. It is recommended that one end of the replacement rail shall be
welded in to the existing track, as soon as possible.

6.1.3 - Permanent repair

Different methods can be used to effect a permanent repair but these must always result in a known
stress condition being restored.

6.1.3.1 - Welding existing rail after marking

If the break is clean and the existing rail can be re-welded, a punch mark should be made on the side
of the rail head 1 000 mm each side of the edge of the break, the welding gap cut and the rails
extended to ensure the distance between the two marks is exactly 2 000 mm. The rails should then
be welded using a wide-gap thermit weld if necessary.

17 720
R
6.1.3.2 - Welding replacement rail after marking

If it is decided to replace the broken rail, the gap created by the break must first be measured. Tell-
tale marks should be marked on the rail and sleeper outside the limits of the proposed replacement
rail length. The existing broken rail should then be cut out, and the replacement rail welded in at one
end. The replacement rail should be extended by the length of the original gap created by the break
and welded, using a normal thermit weld.

6.1.3.3 - Welding existing or replacement rail without marking

If for any reason the gap created by the break is not known, then the rails must be stressed in the
normal manner (see point 4 - page 12) irrespective of whether the existing rail or a replacement rail is
used.

6.2 - Track buckling and distortion caused by heat

Track buckling or distortion occurs when the heat-induced compressive forces in the rails are sufficient
to overcome the lateral resistance of the track structure.

The wavelength of a buckle is typically 8 to 20 metres with an amplitude of up to 1 000 mm or more.

The wavelength of a distortion is typically 4 to 8 metres with an amplitude of 20 to 80 mm.

6.2.1 - Operating measures

Following the occurrence of a buckle or distortion, the track must be closed to traffic. On re-opening it
may be necessary to impose a temporary speed restriction because of misalignment of the track or
the existence of track defects caused by the buckle.

6.2.2 - Temporary restoration of the track

The cause of the buckle should be established before remedial work is undertaken.

The track must be restored to its original condition.

It may be necessary to cut both rails, remove excess rail and temporarily fishplate them together.

It may be necessary to free off the fastenings over the affected zone and then re-tighten them.

Any badly distorted rails and damaged sleepers must be replaced.

The affected zone should be liberally ballasted, tamped and re-aligned, and then compacted if
necessary.

A temporary speed restriction should be imposed and the behaviour of the track should be monitored.

18 720
R
6.2.3 - Final restoration of the track

Following a track buckle or distortion, the track shall be re-stressed in the normal way including the
lengths of track abutting the buckle zone.

If necessary during very hot weather when it is considered that there is a risk of further buckling
occurring at the same location, the rails may be stressed to a higher stress-free temperature than
normal, then re-stressed at a later date to the normal stress-free temperature once colder conditions
prevail.

19 720
R
Appendices

Appendix A - Safety

A.1 - Safety fundamentals / philosophy

The basic premise for CWR buckling safety assurance lies in the performance-based requirement
(statement) that "CWR track shall have the buckling strength required to withstand the environmentally
and operationally imposed loads for the range of expected operating conditions". The translation of
this statement into viable and usable safety specifications requires a rational CWR buckling safety
management methodology based on application of the following four key elements:

- CWR track system / component characterisation,

- Conduct buckling / stability analyses,

- Establish and apply safety criteria,

- Perform safety evaluation

The first element requires the track to be defined in terms of its governing parameters, specifically the
rail properties, track curvature, lateral, longitudinal and torsional resistance of the track, permissible
track misalignments, sleeper/ballast friction coefficient, track modulus, and vehicle characteristics and
loads.

The second element requires the application of a dynamic or "quasi-dynamic" buckling analysis/model
which appropriately incorporates all the governing track parameters above, and correctly predicts the
critical buckling forces and temperatures for the determination of the buckling response
characteristics.

The third element establishes buckling safety criteria based on an appropriate stability criterion
applied to the analytically determined buckling response curves for the "levels of safety" desired.

The last element requires applying the safety criteria either:

- to perform a buckling safety evaluation and "reserve buckling strength" (margin of safety)
determination directly, as in the US DOT's "CWR-BUCKLE" model (see Bibliography - page 43),
or,

- to develop safety guidelines / specifications in terms of "allowable temperature increase" limits as

functions of track parameters / conditions, and apply these specifications for safety assessments.

In the first case, the safety assessments are conducted through direct application of the safety
analysis model, while in the second, the model is used to establish a set of parametric safety
envelopes, which can be used as safety guidelines.

In the following, it will be assumed that the first two elements of the buckling safety management
methodology are readily available, as in "CWR-BUCKLE", hence the discussion will be limited to the
safety criteria and safety analysis parts. Both deterministic and probabilistic approaches will be
presented.

20 720
R
Appendices

A.2 - Safety concepts and criteria

The general theory of CWR track buckling and safety concepts is available in "Dynamic Bucking of
CWR Track: Theory, Tests and Safety Concepts", "Parametric Analysis and Safety Concepts of CWR
Track Buckling" and "Theory of CWR Track Stability" (see Bibliography - page 43). This theory,
together with the appropriate safety criteria, has been incorporated into a buckling safety analysis
model "CWR-BUCKLE" (see Bibliography - page 43) which has undergone rigorous test validations.
A general purpose analysis code for track and bridge stability and deflection evaluations called
"CWERRI" (see Bibliography - page 43) is also available for buckling analyses, although without
safety criteria. These models predict the buckling response characteristics, based on which safety
concepts can be developed. Consistent with the buckling safety assurance statement in point A.1 -
page 20, the starting point is to require an "allowable or permissible temperature increase", T all , which
should be larger than the anticipated maximum rail temperature increase referenced to the stress-free
temperature, ie:

T all > ( T max T n )

T all can be considered as the "required buckling strength" which is dependent on the governing
track and vehicle parameters discussed above. T all can be derived by applying an appropriate
stability criterion to the analytically-determined buckling response curves, or it can be empirically
determined by dynamic buckling tests.

It should be noted that the stress-free temperature T n is not the installation or fastening temperature,
T f , but the actual value in the service life of CWR which is usually different due to changes induced
by rail / track kinematics and maintenance actions. In the absence of non-destructive techniques to
determine T n , a "safety factor" is usually established to account for the stress-free temperature
variation. The stress-free temperature is then given as:
T n = T f SFTN

where T f is the rail fastening or installation temperature, and SFTN is the stress-free temperature
variation safety factor. For UIC tracks, the recommended values for SFTN are in the range of 5-10C.

A.2.1 - Basis for Determination of Allowable Temperature

The (correctly predicted) buckling response curves provide the basis for the allowable temperature
T all (see Figure 1 - page 28 for illustration). These equilibrium curves are characterised by an upper
and lower temperature (increase) value, and buckling can take place in between these two
temperatures. The allowable temperature can be based on either the Lower Critical Temperature
T b ,min or on a temperature above T b ,min (but less than T b ,max ) depending on the "levels of safety"
desired, i.e.:

- Level 1 Safety: T all = T b ,min

- Level 2 Safety: T all = T b ,min + T

NB : Level 1 tends to be more conservative, i.e. "safer" than Level 2, and the T value is to be
determined based on safety considerations.

21 720
R
Appendices

Level 1 Safety based on T b ,min

The use of the lower critical temperature, T b ,min , as a baseline for allowable temperature is based on
the fact that it "guarantees" (although not absolutely) the safety of CWR track against buckling, since
only above this temperature do there exist possible buckled configurations (i.e. non-zero equilibrium
states). Hence only above this temperature can buckling occur. Research has shown that at this lower
critical temperature the buckling energy (i.e. energy required to displace the track into a buckled
configuration) is "considerable", hence the T all is fairly conservative, thus providing a "high" level of
safety. It should also be noted that since the influences of the more complex train dynamics loads,
such as braking and traction forces, impact loads, and localised thermal load inputs due to wheel
flanging, are not included in the buckling models, this choice for T all provides some added margin of
safety against these dynamic effects (i.e. these dynamic effects are indirectly accounted for by the
large buckling energies required to buckle the track at this temperature). Additionally, it should also be
noted that an implicit assumption in this criterion is that the difference between T b ,max and T b ,min is
larger than 20C. This caveat (or sufficient condition) is required because buckling energy
computations show that buckling energy values at T b ,min decrease substantially as these two values
approach each other, hence they may not be adequate to handle the aforementioned dynamic effects.
It can be shown that most European track conditions and parameters tend to fulfil this requirement of
( T b ,max T b ,min ) > 20C . Note also that as shown in "Theory of CWR track stability" (see
Bibliography - page 43), the choice of T b ,max for T all is prohibitive.

NB : For track structures and operating conditions which might require more stringent safety
assurance, such as in heavy-haul freight territories with large train dynamics loads, and for
"weak" track conditions (very low lateral resistance, large misalignments, and small radius
curves) which result in 0 < (Tb,max -Tb,min) < 5C, Tall may need to be "adjusted downward" by
a safety factor T1 to provide the required margin of safety. Based on current research data, a
reasonable value for T1 may be set at 5C, so that for these conditions the recommended
safety criterion is:

T all = T b, min 5C

Level 2 Safety based on T b ,min + T

If a less conservative safety criterion is desired, T all may be dependent on each railway organisation's
ability to maintain tracks within desired tolerances, and to control operating conditions within
"acceptable dynamic load" regimes, T all may be increased to above Level 1 values by a factor T .
The choice or determination of T is not trivial because buckling potential increases rapidly with
temperatures above the T b ,min value. This is based on research which shows that the buckling energy
decreases sharply from a maximum value at T b ,min to zero at T b ,max (refer to Figure 2 - page 29 for
illustration of buckling energy decrease as a function of temperature above T b ,min ).

22 720
R
Appendices

Approach 1 for T determination

In this approach, the buckling energy versus temperature-increase relationship is used as a criterion
for the choice of T , i.e. Level 2 Safety is based on an allowable temperature corresponding to a
temperature at which there exists a finite buckling energy larger than zero, but less than the maximum
value at T b ,min , i.e.:

T all = T ( E max ) , where 0 < 1 , and E max is the buckling energy at T b ,min .

Hence for the special case when = 1 , T all = T ( E max ) = T b ,min , and Level 2 Safety becomes equal
to Level 1 Safety. The value of is to be determined analytically and experimentally, and the Level 2
Safety based on it must be validated through full-scale dynamic buckling tests. Research to date -
"Dynamic Buckling of CWR Track: Theory, Tests and Safety Concepts" (see Bibliography - page 43)
- suggests a value of = 0 ,5 , referred to as the 50% Buckling Energy Level, so that the allowable
temperature for Level 2 Safety is the temperature above Tb,min which corresponds to the 50% Buckling
Energy Temperature (50% BET), or T all = T ( 0 ,5E max ) = 50% BET . Based on this buckling energy
concept then: T = 50% BET-T b ,min .

Approach 2 for T determination

If the CWR-BUCKLE model is not available for the determination of buckling energies, an alternative
definition of T may be based on model prediction of T b ,max and T b ,min , specifically:
T = 0 ,25 ( T b ,max T b ,min ) . This choice is based on ERRI Committee D 202 parametric studies
employing CWERRI and CWR-BUCKLE, which show close equivalence of the two approaches.

CAUTION: When employing any safety criteria based on T b ,max , special attention must be paid to the
accuracy of the buckling model to predict T b ,max and on the extreme sensitivity of T b ,max on the
governing parameters - "Parametric Analysis and Safety concepts of CWR Track Buckling" (see
Bibliography - page 43).

A.2.2 - Applications of Level 1 and Level 2 Safety Limits

NB : Level 1 Safety should be used when 5C (Tb,max - Tb,min) 20C, while Level 2 Safety may
be used when (Tb,max - Tb,min) > 20C. For the special cases when 0 < (Tb,max - Tb,min) < 5C,
the criterion of Tall = Tb,min - 5C is recommended. Track conditions resulting in progressive
buckling, i.e. when the buckling response does not exhibit distinct Tb values, should not be
permitted.

A.3 - Safety Evaluation and Safety Limit Charts

As indicated in point A.1 - page 20, the buckling safety evaluation can be performed either directly, by
a PC-based buckling safety analysis software using the appropriate T all safety criteria, or through
application of safety-limit charts which give the safe allowable temperatures for a set of prescribed
track conditions. This requires testing the buckling model a-priori for the many cases of parametric
variations. From a practical point of view, it is desirable to reduce the number of parameters to the
primary group of:

- rail size,

- lateral resistance,

- misalignment,

23 720
R
Appendices

- curvature,

- fastening torsional resistance,

- rail stress-free temperature.

All other (secondary) parameters may be "suppressed" by setting them at their nominal values as
discussed in "Dynamic Buckling of CWR Track: Theory, Tests and Safety Concepts", "Parametric
Analysis and Safety Concepts of CWR Track Buckling" and "Theory of CWR Track Stability" (see
Bibliography - page 43). The primary parameters have a significant influence on the upper and lower
buckling temperatures, and they are also the parameters to control for buckling safety. For cases when
these secondary parameters are significantly different from the assumed nominal values, then direct
application of the computer model is required for an exact evaluation of the buckling temperatures and
safety. Figure 3 - page 30 provides an example of safety limits in terms of T all versus lateral
resistance for a range of track curvatures, and an example of an application to evaluate safety.

A.4 - CWR Buckling Safety Based on Risk Methodology - Extension to

risk-based safety assessment

There is always some degree of buckling risk in field conditions, owing to the variability of track
parameters which can significantly influence the CWR buckling potential. Thus, even though the
requirement of T all > ( T max T n ) is satisfied for the majority of track conditions, on occasions an
unexpected reduction in lateral resistance, stress-free temperature, or an unanticipated high rail
temperature can result in T all , being exceeded, and buckling can then occur. In these situations, the
evaluation of the probability of buckling failure for CWR track in terms of risk severity is required. Such
risk assessments can be used to compute a "reliability index" or a "buckling risk index" (BRI) which
indicates the degree of safety of CWR tracks in a probabilistic sense.

A.4.1 - Risk Assessment Approach

Existing buckling safety analysis models such as "CWR-BUCKLE" and CWERRI programs predict the
upper and lower critical temperature increases, and the safe allowable temperature increase values
for given user input parameters. These parameters are at present considered to be deterministic
inputs by the user, whereas it is well known that in field conditions, some of the parameters will vary
over a considerable range. Salient parameters are lateral resistance, misalignment and the rail stress-
free temperature. Knowing the frequency distributions of these parameters, one can determine the
probability of track buckling or a "risk severity index" through risk analysis. Risk analysis is important
because:

1. the track cannot be designed, maintained and operated for every possible scenario of its
parameters, nor for the worst-case situations, and

2. there is always some finite probability associated with some of the user inputs, especially in the
absence of reliable (and real time) measurements.

24 720
R
Appendices

The risk assessment of CWR track buckling can be based on known (measured) or assumed
(experience-based) distributions of the key parameters referred to above, and on the buckling criterion
presented in point A.2 - page 21, namely:

= T all ( T max T n )

If 0 , then buckling will occur. Thus, the probability of becoming zero or less needs to be
evaluated. Figure 4 - page 31 illustrates the process based on a set of assumed probability
distributions of lateral resistance ( F p ), misalignment ( 0 ), and stress-free temperature ( T n ). For
chosen combinations of ( 0 ), ( F p ) and ( T n ), the resulting values of can be computed using the
buckling model for a given maximum rail temperature, T max .

A.4.2 - Definition of "Buckling Risk Index" (BRI)

A risk index for buckling can be defined for the given distributions based on the percent failure ratio, or:

Buckling Risk Index (BRI) = n/N

where N is the total number of cases considered (5 x 5 x 3 = 75 for the illustration in Figure 4) and n
is the number of cases for which 0 for a given T max . This can be plotted as a risk severity curve,
i.e. buckling risk index versus maximum rail temperature as illustrated in Figure 4.

A.4.3 - Buckling Safety Based on Risk Severity (Risk Acceptance)

Each operator must then determine his own acceptable level of risk (i.e. 0%, 5%, 10%, etc.), based
on specific operating conditions and maintenance practices. Or alternatively, maintenance
practices can be based on the "uniformity" of parameters required for a defined risk level. For
the case presented in the illustrative example, if "absolute safety" is required, i.e. 0% risk, then for the
track conditions characterised by the given frequency distributions, the maximum rail temperature
permitted is 50C.

A.5 - Special Track Equipment: Switches, Turnouts and Approaches to

Bridges

In principle, the above approaches for the assurance of lateral stability for conventional CWR tracks
are also applicable to special situations such as switches, turnouts and bridge approaches. However,
the development of applicable safety criteria is awaiting the results of current research work on
theoretical model developments, verification studies, and data on the governing parameters required
for strength, stability and failure evaluations.

A.6 - Safety Criteria for High-Speed Applications: Track Lateral Shift

The criteria for CWR buckling safety developed above apply to all tracks. However, due to the high
tolerance requirements for track geometry, to the generally more stringent requirements for track
strength, and to a better control of rail stress-free temperature variation, high-speed track safety is
usually assured from the buckling point of view. The lateral stability issue affecting high-speed track
safety, however, is the formation and growth of lateral track misalignments due to the high ratio of
Lateral to Vertical (Y/Q) loads and longitudinal forces, or what is defined as track lateral shift. The
distinction between track lateral shift and track buckling as parts of the overall track stability
mechanism is made in Figure 5 - page 32. The resulting lateral misalignments caused by track shift

25 720
R
Appendices

tend to be "small" in magnitude but, in combination with other parameters, can lead to unsafe
conditions of wheel climb, dynamic wide gauge, bogie hunting, track buckling, or to conditions leading
to unacceptable ride quality.

Hence the safety issue for high-speed track is the determination of "critical misalignment" c
at which high-speed operation becomes unsafe.

A.6.1 - Track Lateral Shift Mechanism, Parameters and Allowable Misalignments

Figure 6 - page 33 illustrates the fundamentals of track shift as a "moving load problem" in terms of
the number of axle passes required to produce a permanent (residual) deflection, which may be either
stable (as shown by 1 and 2 ) or unstable (i.e. progressively increasing with the number of passes).
The key influencing parameters are also identified. For high-speed track operations, three levels of
lateral misalignment are typically defined:

- initial misalignment after construction or realignment, 0 ,

- maximum allowable "pre-maintenance" misalignment, m ,

- critical misalignment at which operations are impacted and safety is potentially compromised, c .

The initial misalignment after re-alignment or construction tolerance for new tracks is represented by
0 . This misalignment is typically of the order of one to four millimeters for high-speed tracks. The
maximum allowable misalignment prior to maintenance operations according to individual railway
practices is represented by m and may be in the range of 4-8 mm. The critical misalignment
amplitude at which vehicle operational safety is impacted is represented by c . Several possible
"failure modes" and/or design requirements may have to be considered to determine the lowest value
of c . These include sudden track shift potential, wheel climb, rail roll, buckling, inadequate ride
quality, and exceedance of vehicle design loads.

Displacements 1 and 2 are stable deflections at a number of passes, and are reached (or allowed)
only if c is larger than these. For some high-speed track conditions and parameters, stabilisation may
not be reached prior to c , hence the determination of c for high-speed tracks is a key requirement
for the development of applicable safety criteria. This critical misalignment c can be determined from
mechanistic considerations for given vehicle parameters, speed, track conditions, lateral resistance
characteristics and thermal loads, based on vehicle/track dynamics model predictions of the Y/Q
loads. These Y/Q loads are then used as input into a "moving load" track lateral response model for
the determination of the residual deflections under many axle passes.

The pre-maintenance misalignment m can be determined on a trade-off basis between the

frequency of maintenance (number of safe vehicle passes or MGT) and the margin of safety based on
critical misalignment c . The construction tolerance 0 is usually determined by the railways as a
trade-off between the cost of construction (tolerance and quality assurance), maintenance and the
number of passes that can be obtained between maintenance cycles.

26 720
R
Appendices

A.6.2 - Track Shift Criterion for High-Quality Tracks

In line with the above, requirements imposed by high-speed operating conditions, on permissible
misalignments dictate "allowables" to be determined based on track shift considerations, more
specifically in terms of prescriptions based on 0 , m and c . The safety criteria for limiting track shift
to acceptable values can be stated as the following:

- High-speed tracks under maximum expected vehicle and thermal loads should have a minimum
lateral strength to limit the development of lateral misalignments to within a specified value:
L c , or alternatively,

- Lateral loads generated by high-speed vehicles operating under maximum speed, cant deficiency,
thermal load, and initial misalignment conditions should not cause the exceedance of an allowable
deflection limit: L c .

The quantification of this criterion in terms of permissible net axle-loading (NAL/V) ratios when the
"allowable" deflection limit is based on the track lateral yield strength (or elastic limit) of L 1 mm
is available in "Recent Investigations on Track Lateral Shift Limits For High-Speed Rail
Applications" (see Bibliography - page 43). Depending on the specific high-speed track
parameters and vehicle characteristics, these permissible net axle-load ratios fall in the range of
NAL/V = 0,4 - 0,6. The key influencing parameters are the track lateral resistance characteristics,
sleeper/ballast friction coefficient, vehicle vertical axle-load, track curvature, thermal loads, and
constant versus variable, lateral axle-loads - "Fundamentals of Track Lateral Shift For High Speed
Rail Applications" (see Bibliography - page 43).

27 720
R
Appendices

A.7 - Figures

T T
Tb,max

Tb,max

Tall
Tb,min Tb,min
Tall
T

Level 2 Safety
Level 1 Safety Tall = Tb,min + T
Tall = Tb,min 50% BET - Tb,min
{
T= or
0,25 (Tb,max - Tb,min)

LATERAL DEFLECTION LATERAL DEFLECTION

Fig. 1 - Safety criteria definition in terms of "allowable temperature increase" for safety levels 1 and 2

28 720
R
Appendices

T (C) T (C)
120 120
Emin = 0
Tb,max
100 100
50% buckling energy
temperature (BET)
80 80

60 60

40 40
Tb,min
Emax

20 20

0 0
0 100 200 300 400 500 600 700 800 0 200 400 600 800 1000
Lateral deflection (mm) Buckling energy (Joules)

Parameters
Rail: UIC 60 Sleeper type: concrete
Misalignment: 12 mm in 8 m R = 300 m
Torsional resistance: medium Longitudinal resistance: medium
Axle loads: UIC/D4

Fig. 2 - Buckling energy concept illustration

29 720
R
Appendices

60

55

50

Tall (C)
45

40 Tangent
450 m
35 200 m

30
10 15 20 25 30
Peak lateral resistance, Fp (kN/m)

Parameters
Rail: UIC 60 Axle loads: UIC/D4
Misalignment: 12 mm in 8 m Sleeper type: concrete
Torsional resistance: medium Longitudinal resistance: medium

NB: Estimates on lateral resistance ranges:

Tamped to 0,2 MGT 12-18 kN/m
Stabilized (0,2 to 1 MGT) 19-22 kN/m
Consolidated 23-28 kN/m

Buckling Safety Evaluation

Safety Criterion: = Tall - (Tmax - Tn) 0

where:
Tmax = maximum rail temperature
Tn = stress-free temperature = fastening temperature - SFTN
SFTN = safety factor for Tn variation 5C

Example: For a weak (tamped) R = 450 m track fastened at Tf = 20C, and assuming
a 12 mm line defect, what is the permissible Tmax ?

Answer: Determine Fp 15 kN/m;

then, from figure: Tall = 45C ;
then, = 45 - (Tmax - 15) 0
Tmax (permissible) = 60C

Fig. 3 - Buckling safety limits for CWR tracks based on Level 1 safety criterion

30 720
R
Appendices

0.4 RISK SEVERITY ASSESSMENT

Lateral 0.8
probability

0.3
resistance
0 .2 0.7
0 .1
0.6
0 .0

Buckling Risk Index (BRI)

14 16 18 20 22 24 26
0.5
Fp (kN/m)
0 .4
Stress-free 0.4
temperature
probability

0 .3
0.3
0 .2

0 .1 0.2

0 .0
5 10 15 20 25 30 35
0.1
Tn (C)
0 .6 0.0
40 45 50 55 60 65 70
probability

0 .5 Lateral
misalignment Maximum Rail Temperature (C)
0 .3

0 .2

0 .0
3 6 9 12 15 18 21
0 (mm)

Parameters
Rail: UIC 60 Longitudinal resistance: medium
Sleeper type: concrete Vertical modulus: 41 Mpa
Radius: 300 m Axle loads: UIC/D4
Torsional resistance: medium

Fig. 4 - Buckling safety based on risk severity assessment

31 720
R
Appendices

Stage Event Cause Result

(1) Initial imperfections (welds) and
construction defects
Formation of initial
(2) High Y/Qs due to bogie hunting Lateral
1 line defects
misalignment
("small") (3) Localized weak ballast conditions
(4) Longitudinal forces
(1) Net axle Y/Q increase due to line defects
Growth of
misalignments (2) High Y/Qs due to bogie hunting
2 Track lateral shift
("small" to (3) High longitudinal forces due to T
"moderate")
(4) Multiple wheel passes
(1) High longitudinal force
(2) Reduced Tn (stress-free temperature)
Sudden formation
3 of "large" (3) Misalignments generated by track shift Track buckling
misalignments
(4) Dynamic uplift wave
(5) Weakened lateral resistance

Fig. 5 - Track lateral-stability mechanism description

32 720
R
Appendices

PROGRESSIVE
LATERAL LOADS

Y1 < Y2 < Y 3
Residual Deflection,

Y3

c
Stabilized Y2
m Y1
1 2
o
Number of Passes, N

Y
= f n ---- , N, TLR, P 1, .............
Q
Residual
deflection
Number of passes

"Loaded" and "Unloaded"

lateral resistance

Y Rail Longitudinal Force

Axle ---
- ratio is a function of:
Q
. initial line defect size,
. vehicle characteristics,
. vehicle speed,
. track curvature,
. track lateral stiffness.

Fig. 6 - Track shift mechanism and key influencing parameters

33 720
R
Appendices

Appendix B - Case studies

B.1 - Use of CWERRI and CWR-BUCKLE Models

Appendix A - page 20 defines track safety regarding buckling and the philosophy behind this. Under
the auspices of ERRI Committee D 202 (see Bibliography - page 43) TU Delft developed CWERRI to
assess track buckling quantitatively. It allows the longitudinal, vertical and lateral behaviour of CWR
track to be modelled and calculated integrally in a user-friendly environment. The basic features of the
"Stability of CWR track" model (see Bibliography - page 43) are as follows:

- three-dimensional modelling and calculation tool,

- longitudinal, lateral and vertical forces,

- lateral and vertical track stability - "Theory of CWR track stability" (see Bibliography - page 43),

- thermal and mechanical loads,

- complete train loads, taking uplift waves into account,

- three-dimensional ballast yield, taking the influence of vertical loads into account,

- track/bridge interaction, including the effects of end rotation,

- multi-span bridges with parallel tracks.

A major application of CWERRI is its use as a tool for safety analyses. Together with the DOT's CWR-
BUCKLE program, a large number of calculations have been performed in developing the safety
criteria implemented in this leaflet.

The results of the calculations with CWERRI are expressed in terms of the temperature of the rail
above the normal or stress-free temperature. Normally, there is an area between two limits ( T b ,min
and T b ,max ) wherein buckling may occur.

34 720
R
Appendices

plastic and irreversible

plastic and irreversible
deformation line
deformation line
certainty buckling
Tb,max certainty buckling
Temperature (above normal)

Temperature (above normal)

Tb,max
(plastic branch)

risk of buckling risk of buckling

Tb,min Tb,min
no risk of buckling no risk of buckling
Tb,max PB
(progressive buckling)
elastic and reversible elastic and reversible
deformation line deformation line

Lateral displacement Lateral displacement

Fig. 7 - Temperature vs. lateral displacement for good ballast (left) and poor ballast (right)

It is widely assumed that track buckling in this area between T b ,min and T b ,max depends on the
external addition of energy to the track, e.g. by a moving train when the track is in a "ready-to-buckle"
condition. For temperatures close to T b ,min , the amount of energy is much larger than for
temperatures close to T b ,max .

This amount can be calculated by a program called CWR-BUCKLE which was developed in the USA:
"Parametric analysis and safety concepts of CWR Buckling" (see Bibliography - page 43) for the same
purpose as CWERRI: track safety against buckling.

The parametric studies on track buckling - "Parametric study and sensitive analysis of CWERRI" (see
Bibliography - page 43) - were carried out for the following four types of structures with both programs:

High-speed track: design speed > 200 km/h

Main-line track: design speed 160 km/h
Secondary-line track: design speed 120 km/h
Freight-line track: design speed 80 km/h

Each type of structure has its specific properties in terms of curve radii, sleeper type, ballast,
fastenings and misalignments (as a result of maintenance specifications) and railway vehicles passing
over it. Design speeds have only been added to indicate and classify track structures. The main results
are summarised below in two examples.

35 720
R
Appendices

B.2 - Example 1: High-speed and main-line tracks

In both track structures, an ICE trainset was used for the calculations. Compared to other high-speed
vehicles, the ICE power car causes minimum total rail deflection between the two inner axles. This
makes the track structure sensitive to lateral buckling. Calculations have been performed with UIC 60
rails and a vertical track stiffness of 100 MN/m/m.

Different track properties are represented:

Lateral Torsional
Structure type Curve
Sleepers ballast resistance of Misalignment
and number radius
resistance fastening
(mm) (tan ) (kN/m) (kNm/rad/m) (mm)
High-speed track Tangent Concrete 10/10, 75, 150 8/12/16
(1) friction 0,86 15/12,
20/16
Main-line track 900 Concrete 10/10, 75, 150 10/14/18
(2) friction 0,86 15/12,
20/16
Main-line track 900 Wooden 7/7, 150, 250 10/14/18
(3) friction 1,2 10/10,
15/12

Observations concerning properties and values:

1. Concrete and wooden sleepers have different properties in terms of lateral resistance and
interaction with ballast material. With wooden sleepers, ballast grains penetrate and increase
lateral stiffness (represented by tan values). On the other hand, wooden sleepers are generally
laid in lower qualities of ballast and wooden sleepers weigh less. These two factors reduce lateral
stiffness.

2. Ballast material is characterised by two values: F p and F 1 . The first (higher) value indicates the
maximum lateral force to be applied on a one metre track panel, causing an elastic lateral
displacement (generally between 2 and 5 mm). The second value represents a constant
(asymptotic) force value for friction of a track panel through and over the ballast grains.

3. Misalignments are in the lateral direction. The above mentioned values are magnitudes (double
amplitudes) of a cosine shaped wave, respectively 8, 10 and 12 metres in length.

4. Torsional resistance varies with fastening type. It is commonly found that wooden sleepers show
higher resistance than concrete sleepers. The influence on the results is, however, quite small.

36 720
R
Appendices

Overview of Tb,max and Tb,min of track structures Nr. 1, 2 and 3

140

130
concrete sleepers:
average ballast quality
Temperature (above stress-free) C

120

110
concrete sleepers:
100 low ballast quality
concrete sleepers:
90 high ballast quality

80

70

wooden sleepers:
60
low ballast quality

50

wooden sleepers: wooden sleepers:

40 average ballast quality high ballast quality
PB PB
30

Caption:

Tb,max Nr. 1
Tb,min Nr. 1
Tb,max Nr. 2
Tb,min Nr. 2

Tb,max Nr. 3
Tb,min Nr. 3

NB: For each ballast quality, two alternatives are shown with increasing torsional stiffness and
three alternatives with increasing misalignment length and amplitude.

Fig. 8 - CWERRI temperature results for high-speed and main-line tracks

In this example T b ,max and T b ,min values are calculated and displayed below.

For track safety calculations (which are in fact more precise estimations), the following concept of
safety should be maintained as stated in Appendix A - page 20.

37 720
R
Appendices

For all CWERRI calculations, first T is calculated by ( T b ,max T b ,min ) .

If T > 20C : T all = T b ,min + 25% of T

If 5C < T < 20C : T all = T b ,min
If 0C < T < 5C : T all = T b ,min 5C
If T < 0C : not allowable on main lines

Overview of safety temperatures Tall of track structures Nr. 1, 2 and 3

85
Allowable rail temperature (above stress-free) C

80
concrete sleepers:
75 average ballast quality

concrete sleepers:
70 low ballast quality

65
concrete sleepers:
wooden sleepers:
high ballast quality
60 low ballast quality

55
wooden sleepers:
50 high ballast quality

45
wooden sleepers:
40 average ballast quality

PB PB
35
Caption:
Tall CWERRI Nr. 1
Tall CWR-BUCKLE Nr. 1
Tall CWERRI Nr. 2
Tall CWR-BUCKLE Nr. 2
Tall CWERRI Nr. 3
Tall CWR-BUCKLE Nr. 3

NB: For each ballast quality, two alternatives are shown with increasing torsional stiffness and
three alternatives with increasing misalignment length and amplitude.

Fig. 9 - CWERRI and CWR-BUCKLE results for safety temperatures

for high-speed and main-line tracks

38 720
R
Appendices

In the last case progressive buckling (PB) occurs, which means that elastic and plastic lateral
deformation easily fade into each other (Figure 7 - page 35). PB is strongly related to low ballast
quality track structures.

For all CWR-BUCKLE calculations, the following criteria are always maintained:

Tall = T{0,5 Emax} = 50% BET

B.3 - Example 2: Secondary and freight-line tracks

An EC1 freight vehicle was used for the calculations for these two track structures. Compared to other
vehicles or coaches, this vehicle causes minimum total rail deflection between inner axles, which
makes the track structure highly susceptible to lateral buckling. Calculations have been performed
with UIC 60 rails and a vertical track stiffness of 70 MN/m/m.

Different track properties are represented:

Lateral Torsional
Structure type Curve
Sleepers ballast resistance of Misalignment
and number radius
resistance fastening
(mm) (tan ) (kN/m) (kNm/rad/m) (mm)
Secondary-line 600 Concrete 10/10, 75, 150 14/18/22
track (4) friction 0,86 15/12,
20/16
Secondary-line 600 Wooden 7/7, 150, 250 14/18/22
track (5) friction 1,2 10/10,
15/12
Freight-line track 300 Corrective 10/10, 75, 150 14/22/30
(6) friction 0,86 15/12,
20/16
Freight-line track 300 Wooden 7/7, 150, 250 14/22/30
(7) friction 1,2 10/10,
15/12

For remarks concerning this Table, the reader is referred to the remarks in the Table of Example 1 -
page 36 of this point. The same is noted for track safety concepts and calculation criteria. In the
present example, T b ,max and T b ,min values are calculated and displayed below:

39 720
R
Appendices

Overview of temperatures Tb,max and Tb,min of track structures Nr. 4, 5, 6 and 7

100

concrete sleepers:
90 average ballast quality
Temperature (above stress-free) C

80

concrete sleepers:
70 low ballast quality

concrete sleepers:
60 high ballast quality

wooden sleepers:
50 low ballast quality

40

PB wooden sleepers:
30 high ballast quality
wooden sleepers:
PB average ballast quality
20
Caption:
Tb,max Nr. 4
Tb,min Nr. 4
Tb,max Nr. 5
Tb,min Nr. 5
Tb,max Nr. 6
Tb,min Nr. 6
Tb,max Nr. 7
Tb,min Nr. 7

NB: For each ballast quality, two alternatives are shown with increasing torsional stiffness and
three alternatives with increasing misalignment length and amplitude.

Fig. 10 - CWERRI temperature results for secondary-line and freight-line tracks

40 720
R
Appendices

Finally, the T all values are depicted below also:

Overview of safety temperatures Tall of track structures Nr. 4, 5, 6 and 7

70

65
Allowable rail temperature (above stress-free) C

concrete sleepers:
60 average ballast quality

concrete sleepers:
55 low ballast quality

wooden sleepers:
50 low ballast quality

45 concrete sleepers:
high ballast quality

40

wooden sleepers:
35 high ballast quality

30
wooden sleepers:
average ballast quality
25 PB PB
Caption:
Tall CWERRI Nr. 4
Tall CWR-BUCKLE Nr. 4
Tall CWERRI Nr. 5
Tall CWR-BUCKLE Nr. 5
Tall CWERRI Nr. 6
Tall CWR-BUCKLE Nr. 6
Tall CWERRI Nr. 7
Tall CWR-BUCKLE Nr. 7

NB: For each ballast quality, two alternatives are shown with increasing torsional stiffness and
three alternatives with increasing misalignment length and amplitude.

Fig. 11 - CWERRI and CWR-BUCKLE results for safety temperatures

for secondary-line and freight-line tracks

41 720
R
Appendices

B.4 - Application of results

For track safety, T all is the maximum allowable temperature above the stress-free temperature of the
rail that is considered safe as far as track buckling is concerned. T all can be seen as a buffer with
regard to many phenomena that increase rail temperature or an equivalent axial compressive stress
in the rail, pushing it into the dangerous "buckling" range. These phenomena are:

- air temperature,

- sunlight,

- eddy-current brakes,

- interaction with other structures, such as bridges.

The first and second items generally take up about 30-40C of the available temperature buffer,
depending on local circumstances. This means that specific structures with a Tall value lower than
40C will certainly encounter buckling problems in summer or under high temperature conditions. With
T all above 40C, track structures can be exposed to eddy-current brakes or to external forces from
other structures, as long as the equivalent temperature increase remains within the calculated
temperature buffer.

B.5 - General remarks

Figure 9 - page 38 and Figure 10 - page 40 show the results of the calculations for several different
track structures, calculated with two different programs and with two different criteria for the safety
temperature. It is noticeable that the differences in the results for similar track structures with small,
moderate and large misalignments are quite small (< 5C), while the differences for low, average and
high ballast quality are almost constant (8.5-10C for CWERRI results, 6-9C for CWR-BUCKLE
results). These differences between two ballast qualities are larger than the differences between
T b ,min values for two ballast qualities. It can be concluded that the values of T b ,max (CWERRI) and
energy absorption (CWR-BUCKLE) for specific track structures increase more than linearly if ballast
quality and T b ,min increase. The differences in the results for concrete-sleeper and wooden-sleeper
track structures are almost negligible for equal ballast qualities.

42 720
R
Bibliography

1. UIC leaflets

International Union of Railways (UIC)

UIC Leaflet 518: Testing and approval of railway vehicles from the point of view of their dynamic
behaviour - Safety - Track fatigue - Ride quality, 2nd edition, January 2003

UIC Leaflet 717: Recommendations for the design of bridges to satisfy track requirements and reduce
noise dimensions, 1st edition of 1.7.95

UIC Leaflet 774-3: Track-bridge interaction. Recommendations for calculations, 2nd edition, October
2001

2. ERRI reports

International Union of Railways (UIC)

ERRI D 117/RP 8: Optimum adaptation of the conventional track to future traffic - A study of factors
influencing the resistance to transverse displacement of unloaded track, October 1976

ERRI D 202/RP 1: Improved knowledge of forces in CWR track (including switches) - Proposals for
theoretical model investigations concerning CWR, August 1994

ERRI D 202/RP 2: Improved knowledge of forces in CWR track (including switches) - Review of
existing experimental work in behaviour of CWR track, November 1995

ERRI D 202/RP 3: Improved knowledge of forces in CWR track (including switches) - Theory of CWR
track stability, February 1995

ERRI D 202/RP 4: Improved knowledge of forces in CWR track (including switches) - Stability of
continuous welded rail track, TU DELFT Model CWERRI, 1997

ERRI D 202/RP 5: Improved knowledge of forces in CWR track (including switches) - Creep behaviour
of CRW track, April 1999

ERRI D 202/RP 6: Improved knowledge of forces in CWR track (including switches) - Interaction
between welded turnouts and adjoining CWR track zones TURN program and analysis of factors that
influence turnout response under a thermal load, March 1999

ERRI D 202/RP 7: Improved knowledge of forces in CWR track (including switches) - Lateral and
longitudinal resistance measurements: synthesis report, April 1999

ERRI D 202/RP 8: Improved knowledge of forces in CWR track (including switches) - Long-term
investigations of CWR, April 1999

ERRI D 202/RP 9: Improved knowledge of forces in CWR track (including switches) - Methods for the
non-destructive measurement of forces in CWR track. Summary, April 1999

ERRI D 202/RP 10: Improved knowledge of forces in CWR track (including switches) - Leaflet 720 R.
Laying and Maintenance of CWR Track, April 1999

43 720
R
ERRI D 202/RP 11: Improved knowledge of forces in CWR track (including switches) - Parametric
study and sensitivity analysis of CWERRI. Verification of an FEM for the buckling resistance of tracks
with continuous welded rails (CWR), April 1999

ERRI D 202/RP 12: Improved knowledge of forces in CWR track (including switches) - Final report,
April 1999

3. Miscellaneous

Kish A. and Samavedam G.

Dynamic Buckling of Continuous Welded Rail Track: Theory, Tests and Safety Concepts, TRR 1289,
Conference on "Lateral Track Stability", May 1990

Samavedam G., Kish A., Purple A. and Schoengart J.

Parametric Analysis and Safety Concepts of CWR Track Buckling, DOT/FRA/ORD-93/26, December
1993

Kish A. and Samavedam G.

CWR-Buckle, Version 2.0 Software and Users Guide, 1996

Kish A., Samavedam G. and Wormley D.

Recent Investigations on Track Lateral Shift Limits for High-Speed Rail Applications, Proceedings of
World Congress on Railway Research, November 1997

Kish A., Samavedam G. and Wormley D.

Fundamentals of Track Lateral Shift for High-Speed Rail Applications, US DOT/Volpe Centre Report,
December 1998

Van M.A.
Stability of Continuous Welded Rail Track, Delft University Press, June 1997

44 720
R
List of figures

Fig. 1 - Safety criteria definition in terms of "allowable temperature increase"

for safety levels 1 and 2...................................................................................................... 28

Fig. 2 - Buckling energy concept illustration ................................................................................... 29

Fig. 3 - Buckling safety limits for CWR tracks based on Level 1 safety criterion ............................ 30

Fig. 4 - Buckling safety based on risk severity assessment ........................................................... 31

Fig. 5 - Track lateral-stability mechanism description ..................................................................... 32

Fig. 6 - Track shift mechanism and key influencing parameters ..................................................... 33

Fig. 7 - Temperature vs. lateral displacement for good ballast (left) and poor ballast (right) .......... 35

Fig. 8 - CWERRI temperature results for high-speed and main-line tracks .................................... 37

Fig. 9 - CWERRI and CWR-BUCKLE results for safety temperatures

for high-speed and main-line tracks ................................................................................... 38

Fig. 10 - CWERRI temperature results for secondary-line and freight-line tracks ............................ 40

Fig. 11 - CWERRI and CWR-BUCKLE results for safety temperatures

for secondary-line and freight-line tracks............................................................................ 41

45 720
R
 Warning

No part of this publication may be copied, reproduced or distributed by any means whatsoever, including
electronic, except for private and individual use, without the express permission of the International Union of
Railways (UIC). The same applies for translation, adaptation or transformation, arrangement or reproduction by
any method or procedure whatsoever. The sole exceptions - noting the author's name and the source - are
"analyses and brief quotations justified by the critical, argumentative, educational, scientific or informative nature
of the publication into which they are incorporated".
(Articles L 122-4 and L122-5 of the French Intellectual Property Code).
International Union of Railways (UIC) - Paris, 2005

Printed by the International Union of Railways (UIC)

16, rue Jean Rey 75015 Paris - France, March 2005
Dpt Lgal March 2005

ISBN 2-7461-0525-X (French version)

ISBN 2-7461-0526-8 (German version)
ISBN 2-7461-0527-6 (English version)

720
R